mammnaM-i iBHJumii irnm cf^^T'% i- MBLA^raOI Library r PHOTOSYNTHESIS AND RELATED PROCESSES VOLUME II Part 1 By EUGENE I. RABINO WITCH Research Professor, Photosynthesis Research Labora- tory, Department of Botany, University of Illinois. Formerly Research Associate, Solar Energy Research Project, Massachusetts Institute of Technology. VOLUME II . Part 1 Spectroscopy and FInoreseence of Photosynthetie Pigments; Kinetics of Photosvntiiesis J/. rO PHOTOSYNTHESIS ^:( and Related Processes 1951 INTERSCIENCE PUBLISHERS, INC., NEW YORK Interscience Publishers Ltd., London Copyright, 1951, by Interscience Publishers, Inc. ALL RIGHTS RESERVED. This book or any part thereof must not be reproduced in any form without permission of the publisher in writing. This applies specifically to photostat and microfilm reproductions. INTERSCIENCE PUBLISHERS, Inc., 250 Fifth Avenue, New York 1, N. Y. For Great Britain and Northern Ireland: INTERSCIENCE PUBLISHERS Ltd., 2a Southampton Row, London W. C. I. PRINTED IN THE UNITED STATES OF AMERICA BY MACK PRINTING COMPANY, EASTON, PA. PREFACE to Volume II, Part 1 The manuscript of ^'oUlme II of this monograph was ready in draft foi-m when the first vokime was dehvered for pubHcation in 1944. After the intermption caused by war, I felt reluctant to publish the second volume without thorough revision in the light of new research data and of my own better understanding of some of the phenomena discussed. I began, in 1947, a revision of the manuscript; while I was revising the material, research in the field of photos^Tithesis picked up after the war- time slack, and the writing became something of an Achilles vs. turtle race. When, finally, the text achieved a temporary completeness, the count of the galleys revealed that it had become too long to be published under one cover. It was therefore divided into two half-volumes. The division cut through the part dealing with the kinetics of photosynthesis. Because of this, it seemed inadvisable to provide this half-volume with a separate subject index; an index for the whole work will be found at the end of the second part.* The latter will complete the treatment of the kinetics of photosjmthesis (temperature effects, flashing light experiments, induction phenomena, and the function of the pigments, especially the energy transfer between them). The last three chapters will constitute an addition to Volume I, and will deal particularly with the new work on photochemistry of pigments (in solution and in chloroplasts) , and with studies of the chemistry of carbon dioxide reduction by means of radioactive carbon. The hopeful advance in these two fields has changed the appearance of the whole field of photosjTithesis. The analysis of kinetic data, to which many pages in this half -volume are devoted, now seems somewhat like an attempt to reach a treasure chamber by drilling through steel walls while keys have been found to unlock the door. However, photosjTithesis is not only a biochemical process in which all we want to learn is the chemical composition of the intennediates and the nature of the enzymes involved. It is also a most interesting physicochemical phenomenon; its kinetics will be worth continued study even after organic chemists * As in Volume I, an index of the most important investigations described is pro- vided at the end of this half-volume. VI PREFACE and biochemists have disentangled its chemical processes as thoroughly as they did those of respiration — which is still a long way to go. As a matter of fact, the kinetic aspects of respiration themselves are not ade- quately known, and will have to undergo hard study sooner or later. Mr. Earl E. Jacobs not only kindly checked the derivation of kinetic equations in Chapters 26, 27, and 28, but has contributed much original thought and work to their development and interpretation ; it is a pleasure to thank him for his unstinted assistance. Much of the work on this volume was carried out while I was a member of the Solar Energy Research Project at the Massachusetts Institute of Technology. My thanks are due to the Project Committee and its chair- man, Professor Hoyt C. Hottel, for generous assistance. I am equally indebted to the Photosynthesis Research Project, Department of Botany, University of Illinois, and Professor Robert Emerson, whose understand- ing and help have made the termination of the work possible. Mrs. Carolyn Baer, Mrs. Marjorie Goodrich, and Mr. T. R. Punnett have given me valuable aid in the reading of the proofs and the checking of the bibliography. Eugene I. Rabinowitch Urhana June 1951 CONTENTS Preface . Page V PART THREE— SPECTROSCOPY AND FLUORESCENCE OF PHOTOSYNTHETIC PIGMENTS Chap. 21. Absorption Spectra of Pigments in Vitro 603 A. Absorption Spectra of Chlorophyll and Its Derivatives 603 1. Absorption Spectra of Chlorophylls a and b 603 2. Absorption Spectra of Chlorophylls c and d, Bacteriochlorophyll and Protochlorophyll 614 3. Relation between Absorption Spectrum and Molecular Struc- ture of Porphin Derivatives 619 4. Some Theoretical Remarks on the Spectrum of Chlorophyll. . . 630 (a) The Term System 630 (b) Life-Time of the Excited States of Chlorophyll 633 B. Influence of JVIedium on Absorption Spectrum of Chlorophyll and Bacteriochlorophyll 635 1. Solvent Effect in the Spectra of Chlorophyll and Bacterio- chlorophyll 637 2. Absorption Spectrum of Colloidal and Adsorbed Chlorophyll. . 649 C. Absorption Spectra of the Carotenoids 656 1 . Experimental Results 656 2. Theoretical Considerations 662 D. Absorption Spectra of the Phycobilins 664 Bibliography 668 Chap. 22. Light Absorption by Pigments in the Living Cell 672 A. Light Absorption by Plants 673 1. General Remarks 673 2. Average Transmittance and Reflectance of Leaves and Thalli in White Light. Intensity Adaptation and Movements of Chloro- plasts 677 3. Absorption by Nonplastid Pigments 684 B. Spectral Properties of Plants : 686 1. Empirical Plant Spectra 686 2. Band Maxima of Chlorophyll and Bacteriochlorophyll in the Spectra of Living Plants 697 (a) Red Band of Chlorophyll a 697 (b) Red Band of Chlorophyll b 701 (c) Red and Infrared Bands of Bacteriochlorophyll 702 (d) Blue-Violet Bands of Chlorophyll 705 (e) Protochlorophyll 705 vii Vlll CONTENTS Chapter 22, contd. PaGE 3. Absorption Bands of Accessory Pigments in Live Cells 705 4. True Absorption Spectrum of the Pigment Mixture in the Liv- ing Cell 709 C. Distribution of Absorbed Energy among Pigments 717 L Effects of Spatial Distribution of Pigments in the Cell 717 2. Apportionment of Absorption in Uniform Mixture 718 3. Absorption by Chlorophylls a, 6, c and d 719 4. Absorption by Carotenoids in Green Plants 721 5. Absorption by Carotenoids in Brown Algae 723 6. Absorption by Carotenoids and Phycobilins in Blue-Green Algae 728 Appendix. Natural Light Fields 730 Bibliography 736 Chap. 23. Fluorescence of Pigments in Vitro 740 A. Fluorescence of Chlorophyll in Vitro 740 1. Fluorescence Spectra of Chlorophyll and Its Derivatives in Solution 740 2. Yield of Fluorescence and Life-Time of the Excited States of Chlorophyll 751 3. Factors Limiting the Yield of Fluorescence 755 (a) Internal Conversion (Physical Dissipation of Excitation Energy) 756 (b) Isomerization or Dissociation ("Mononiolecular" Chemical Quenching) 756 (c) Reaction with Foreign Molecules ("Bimolecular" Chemical Quenching) 756 (d) Bulk Transfer of Electronic Energy 757 (e) Self-Quenching 759 4. Influence of Solvent on Yield of Chlorophyll Fluorescence 763 5. Influence of Concentration on Yield of Fluorescence. Self- Quenching. Fluorescence of Chlorophyll in Colloids and Adsorbates 772 6. Quenching and Activation of Chlorophyll Fluorescence by Admixtures 777 7. Long-Lived Active States and Afterglow of Chlorophyll 790 8. Summary — A Scheme of Fluorescence and Sensitization 795 B. Fluorescence of Carotenoids and Phycobilins in Vitro 798 1. Fluorescence of Carotenoids 798 2. Fluorescence of Phycobilins 799 Bibliography 801 Chap. 24. Fluorescence of Pigments in Vivo 805 1 . Fluorescence Spectra of Plants 806 2. Fluorescence Yield and Sensitized Fluorescence in Vivo 812 3. Effects of Heat and Humidity on Chlorophyll Fluorescence in Vivo 817 4. Variations of Chlorophyll Fluorescence Related to Photo- synthesis 819 Bibliography 826 CONTENTS 1>^ PART FOUR— KINETICS OF PHOTOSYNTHESIS Page Introduction ^^^ Chap. 25. Methods of Kinetic Measurements 833 1. Material 833 2. Light Measurements 837 3. Measurements of Oxygen Evolution 844 4. Measurements of Carbon Dioxide Consumption 851 5. Measurements of Carbohydrate Production and Energy Con- version °'^'^ 6. Application of Isotopic Indicators 854 Bibliography 855 Chap. 26. External and Internal Factors in Photosynthesis 858 1. The "Cardinal Points" and the "Luniting Factors" 858 2. Photosynthesis Not a Homogeneous Reaction 864 3. Some General Kinetic Considerations 866 (a) Sources of Saturation in Photosynthesis 866 (b) Origin of Kinetic Curve Systems of Different Types 868 4. Internal Factors and the "Physiological Concept" of Photo- synthesis 8< 2 5. Rate of Photosynthesis under Constant Conditions. Midday Depression and Adaptation Phenomena 873 6. Aging and Self-Inhibition 880 Bibhography ^^ Chap. 27. Concentration Factors 886 A. Experimental Carbon Dioxide Curves 886 1. Carbon Dioxide Molecules and Carbonic Acid Ions 886 2. General Review of Carbon Dioxide Curves 891 3. Carbon Dioxide Compensation Point 898 4. Carbon Dioxide Fertilization and Inhibition 901 5. External Supply and Exhaustion Effects 903 6. Role of the Stomata 910 7. Interpretation of Carbon Dioxide Curves 916 (a) The Carboxylation Equilibrium 917 (b) Diffusion Factors 921 (c) Slow Carboxylation 924 (d) Nondissociable ACO2 Compound. The Franck-Herzfeld Theory 927 (e) Back Reactions in the, Photosensitive Complex 930 (f) Acceptor "Blockade" • 933 (g) Calculation of Carboxylation Constant from Carbon Di- oxide Curves 934 (h) Are Experimental Carbon Dioxide Curves Hyperbolic? 937 B. Carbon Dioxide Concentration and Fluorescence 939 C. Concentration of Reductants 943 1. Effect on Rate of Carbon Dioxide Reduction in Bacteria 943 2. Effect on Yield of Fluorescence 949 X CONTENTS Chapter S7, contd. PaGE D. Concentration of Inhibitors 951 1. Inorganic Ions 951 2. Poisons and Narcotics 954 Bibliography 960 Chap. 28. The Light Factor. I. Intensity 964 A. Light Curves of Photosynthesis 964 1. General Review 965 2. Linear Range 979 3. Compensation Point 981 4. Saturating Light Intensity 985 5. Absolute Maximum Rate 989 6. Maximum Rate and Average Rate of Photosynthesis under Natural Conditions 996 7. Interpretation of Light Curves 1007 (a) Influence of Inhomogeneity of Light Absorption 1007 (b) General Shape of Light Curves 1012 (c) Analytical Formulation: Effect of Preparatory Dark Re- actions 1017 (d) Analytical Formulation: Effect of Processes in the Photo- sensitive Complex 1020 (e) Analytical Formulation: Effect of "Finishing" Dark Re- actions 1033 (f) Analytical Formulation: Narcotization 1041 B. Light Curves of Fluorescence I 1047 1. Relation between Light Curves of Photosynthesis and Fluo- rescence 1047 2. Effect of Various Factors on Light Curves of Fluorescence. . . . 1051 (a) Carbon Dioxide 1051 (b) Reductants 1052 (c) Temperature 1055 (d) Cyanide 1057 (e) Hydroxylamine and Azide 1062 (f ) Ion Concentration 1063 (g) Narcotics 1063 (h) Oxygen 1063 3. Interpretation of Light Curves of Fluorescence 1067 Bibliography 1078 Chap. 29. The Light Factor. II. Maximum Quantum Yield of Photosynthesis 1083 1. Quantum Yield Measurements by the Manometric Method. . . . 1085 2. Nonmanometric Measurements of Quantum Yield 1118 (a) Chemical Methods 1118 (b) Polarographic Methods 1120 (c) Calorimetric Method 1123 3. Quantum Yield of Bacterial and Algal Photoreduction 1125 4. Quantum Yield of Oxygen Liberation by Isolated Chloroplasts 1129 CONTENTS XI Chapter 29, Contd. PaQE 5. Maximum Quantum Yield in Relation to Lighit Curves as a Whole 1132 (a) Extrapolation of Maximum Quantum Yield from Measure- ments at Higher Light Intensities 1132 (b) Quantum Yields in Strong Light 1136 6. Theoretical and Actual Maximum Quantum Yield 1137 Bibliography 1139 Chap. 30. The Light Factor. IIL Photosynthesis and Light QuaUty; Role of Accessory Pigments 1142 1. Action Spectrum 1142 2. Quantum Yield and Wave Length in Green Plants. Role of Carotenoids 1147 3. Photosynthesis of Green Plants in Ultraviolet and Infrared .... 1152 4. Monochromatic Light Curves, and the Action Spectrum of Photosynthesis in Strong Light 1158 5. Quantum Yield and Action Spectrum of Photosynthesis in Brown Algae 1168 6. Quantum Yield and Action Spectrum of Red and Blue Algae. Role of PhycobUins 1178 7. Action Spectrum of Purple Bacteria 1187 Bibliography 1188 Index 1193 PART THREE SPECTROSCOPY AND FLUORESCENCE OF PHOTOSYNTHETIC PIGMENTS Chapter 21 ABSORPTION SPECTRA OF PIGMENTS IN VITRO A. Absorption Spectra of Chlorophyll AND Its Derivatives* 1. Absorption Spectra of Chlorophylls a and b The spectra of chlorophylls a and b have been studied in detail because of their theoretical interest, as well as because of their usefulness for the spectrophotometric determination of these pigments. Until recently, the results of different authors did not agree ver}^ well, either in the exact posi- tions of the band maxima, or in the values of the extinction coefficients. Lately, improved chromatographic purification methods have enabled Zscheile and co-workers (cf. Zscheile 1934, 1935; Zscheile and Comar 1941; Comar and Zscheile 1941; Harris and Zscheile 1943) and Mackinney (1938, 1940, 1941) to obtain preparations of chlorophylls a and b that could meet high standards of spectroscopic purity and reproducibilitj'. Zscheile, Comar and Mackinney (1942) studied samples of chlorophyll prepared by the first two investigators at Purdue and by the third one at Berkeley, measuring the extinction coefficients by means of two different photoelectric spectrophotometers. For Zscheile and Comar's "wet" preparation of chlorophyll a (cf. page 60-1), the two instruments gave practically identical extinction curves. (In the region between 430 and 660 m/x, all deviations were within 2%.) This shows how successfully large photo- metric errors (which are common in visual and photographic determinations of absorption curves) can be eliminated by the use of photoelectric devices. The spectra of the solutions of chlorophylls a and b prepared by Mackinney in the dry state were, on the whole, similar to that of Zscheile's moist preparation; but dif- ferences up to 10% in chlorophyll a and 15% in chlorophyll b did occur between the ex- tinction curves determined in the two laboratories, as well as between these two curves and that of Zscheile's preparation. The deviations varied irregularly with wave length, indicating the probable presence, in Mackinney's preparation, of an admixture of colored components. One may regret that no measurements were made below 430 m^, since earlier experiments have shown a particularly strong variability of the absorption curve in this spectral region [cf. page 607). The differences between the absorption curves of Zscheile and Mackinney appear minor when compared with the discrepancies that existed between the curves published by earlier investigators {cf. Table 2LI). These discrepancies must have been due to the use of less reliable photometric devices, and to the inferior purity of the earlier chloro- * Bibliography, page 668. 603 604 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 phyll preparations — the latter clearly indicated by the relatively high absorption in the green {cf. last column in Table 21. IB). A spectroscopically important impurity likely to be present in many chlorophyll preparations is the magnesium-free pheophytin, formed from chlorophyll whenever the latter comes in contact with acids. Elimination of magnesium from chlorophyll may take place even in living plants, e. g., under the influence of acid fumes (cf. Stern 1935, 1938; and Tiegs 1938); it can easily occur during extraction, when the pigments are exposed to the action of acids contained in the cell sap. (Harris and Zscheile added mag- nesium carbonate to the extracting solvent to neutralize these acids.) This "primary" pheophytin is removed during chromatographic separation (according to Zscheile 1941, a pheoph>'tin layer in the chromatogram was responsible for his earlier belief that leaf extracts contain a "chlorophyll c," cf. Vol. I, p. 402); but some pheophytin can again be formed afterward, e. g., under the influence of atmospheric carbon dioxide. As shown in figures 21.18 and 21.19, the pheophorbides (and this ap- plies to pheophytins as well) have rather strong absorption bands in the green. Zscheile and Comar (1941) and Harris and Zscheile (1943) found that the ratios of the extinction coefficients in the maxima of the red chlorophyll bands (660 mn for component a in ethyl ether and 642.5 mfx for component b in the same solvent) and of the green bands of pheophytin (505 and 520 niyu, respectively) reach 52 in solutions of the purest prepara- tions of chlorophyll a, and 19 in similar preparations of chlorophyll b, but may drop to as low as 20 and 4.5, respectively, after these preparations have been allowed to stand for as little as a single day in the dry state. Zscheile and Comar (1941) recommended therefore that drying be avoided altogether in the preparation of spectroscopically pure chlorophyll solu- tions. More recently, Zscheile, Comar and Mackinney (1942) succeeded in preparing dry chlorophyll a which could be stored and still showed, upon dissolution, the high ratio of extinctions in the red and in the green indicative of high purity; but no standard procedure for obtaining such stable preparations could be given. Zscheile, Comar and Harris (1944) found that the spectra of ethereal solutions of pure preparations of chlorophyll a show signs of deterioration after about one week storage at 0-5° C. in darkness. Crude ether extracts from leaves, on the other hand, proved to be comparatively stable — some gave no evidence of spectroscopic change even after 14 weeks storage (at — 20° C). Fresh corn leaves could be stored at —20° C, for a whole month without deterioration of chlorophj'-ll. Another problem of chlorophyll purification is the elimination of traces of chlorophyll a from chlorophyll b. According to Zscheile, supposedly "pure" chlorophyll b, used by many earlier observers, did contain up to CHLOROPHYLLS tt AND h 605 10% of chlorophyll a. Its presence can easily be recognized by increased light absorption at 614 mju. According to Biermacher (1939), the fluores- cence spectrum of chlorophyll b is even more senstive to contamination with chlorophyll a than the absorption spectmm (cf. chapter 23, page 744). He recommended extraction with hexane (which dissolves chlorophyll a much more easily than chlorophyll b) as a means of final purification of the 6-component . Extraction is repeated until the fluorescence spectrum of the residue no longer shows the chlorophyll b band. Meyer (1939) asserted that the band at 535 ni/u, which is noticeable in most if not all extinction curves of pure chlorophyll a {cf. fig. 21. IB, and Table 21. lA), is not found in the spectra of fresh leaf extracts, and concluded that a mixture of the purified chloro- phylls a and b is not identical with what he designated as "native" chlorophyll (a + b). This conclusion was criticized by Mackinney (1940, 1941), who found, to the contrary, that by mixing the two pure chlorophyll components one can reproduce the spectrum of a fresh leaf extract in all its details (except, of course, for the blue-violet region, where the absorption of the extracts is partly due to the carotenoids). Fjgiu-es 21.1 A and B show the extinction curves of the two pure chloro- phyll components in ethyl ether, according to Zscheile and Comar (1941). (The second figure is an enlarged detail of the first one.) 180 160 140 120 100 80 60 40 20- r — 0- Component - - i- Component - } II ■J 1 1 1 1 1 1 :i - n \ \ \ - .X "f-^ 1 1 1 1 i^ Iv 16 14 10- 640 380 420 460 500 540 580 620 660 440 480 520 560 600 WAVE LENGTH, m>i WAVE LENGTH, m/i A B Fig. 21.1 Extinction curves of chlorophylls a and b in ethyl ether (after Zscheile and Comar 1941). Ordinates are specific extinction coefficients: log (h/I) = aspcd, where c is in g./l. and d in cm. To obtain molar extinction coefficients, multiply the data for chlorophyll a by 893 and those for chlorophyll b by 907. (These factors are uncertain to the extent of =p 1% because of the unknown degree of hydration of chlorophyll, cf. Vol. I, chapter 16.) 606 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 The absorption curves given by Mackinney (1938, 1940, 1941) and by Winterstein and Stein (1933), although less detailed, agree satisfactorily with figure 21.1; but those of Sprecher von Bernegg, Heierle and Almasy (1935) and of Hagenbach, Auerbacher and Wiedemann (1936) show con- siderable deviations, illustrated by Tables 21.1. Table 21.1. Chlorophyll Solutions in Ethyl Ether A. Absorption Maxima. Main bands in italics. Sprecher Zscheile Harris von Bernegg Hagenbach and and Band Rudolph et al. et al. Egle« : Mackinney Comar Zscheile no. (1933) (1935) (1936) (1939) (1940) (1941) (1943) Chlorophyll a 1 660 657. S 655.6 662.2 660 660.0 660.0 2 613 612.5 608.5 613.6 — 612.5 614.0 3 577 575.4 574.6 574.1 — 572.5 — 4 — 534.8 531.8 530.6 — 527.5 — 5 — — 510.8 . — . — — — 6 — 503.8 494.2 490.5 — 497.5 — 7 — — 464.1 — ■ — — — 8 — 425.8 430.7 431.1 430 427.5 429.0 9 — 407.0 — — — 410.0 410.0 Chlorophyll 6 1 643 642.2 637.6 644-6 642.5 642.5 642.5 2 597 594.7 589.1 595.6 — 592.5 594.0 3 — — 566.7 567.3 — 567.5 — 4 558 — 559.4 544.6 — — — 5 — — ■ 552.6 — — — — 6 — 537 540.4 — — 547.5 — 7 — — 499.9 ■ — — 502.5 — 8 — 451.8 449.9 457.3 453 452.5 453.0 9 — 425.8 — 426.5 — 430.0 428.5 B. Molar Extinction Coefficients, a = log {h/r)/cd, where c is in mole/1, and d is in cm. Most probable values ; in italics. Chlorophyll a in ethyl e ther a X 10-4 Ratios of 1 joefficients blue peak red peak Red peak Blue peak Green min. Observers (660 m/a) (430 m/i) (472 m/i) red peak green nun. Sprecher von Bernegg et al. (1935) 7.4 8.5 0.27 1.15 27 Hagenbach et al. (1936).... 7.23 14.3 0.25 1.98 29 Mackinnev C[Q4(Y\ 7.65 9.75 0.11 1.28 70 Zscheile and Comar (1941). 9.10 12.0 0.08 1.32 114 Zscheile, Comar and Mac- kinne> ' (1942)* 9.00 11.7 0.08 1.35 108 Same'' . . 8.34 10.4 0.09 1.39 85 Same''. . 8.78 11.5 0.09 1.31 84 Harris and Zscheile (1943). — — — 1.33 — Table continued CHLOROPHYLLS a AND 6 607 Table 21.1 — Continued Chlorophyll b in ethj'l ether Observers (643 m/i) (453 m>i) (510 m^) Sprecher von Bernegg et al. (1935) 4.7 9.4 0.28 2.0 17 Hagenbache^aZ. (1936).. 7.10 20.9 0.34 2.91 21 Mackinney (1940) 5.00 13.6 0.26 2.72 19 Zscheile and Comar (1941) 5.15 15.5 0.24 3.01 21 Zscheile, Comar and Mac- kinney (1942)^ 4.80 12.9 0.24 2.65 20 Same'' 4.98 13.9 0.24 2.79 21 Harris and Zscheile (1943) -- — — 2.98 — " Band "axes" — i. e., arithmetic means of the wave lengths of the hmits of blacken- ing on a photographic plate (all other data in this table were obtained by photoelectric photometry) . '' Zscheile and Comar's moist preparation measured by Mackinney. " Mackinney's dry preparation measured by Zscheile and Comar. <* Mackinney's dry preparation measured by Mackinney. The reproducibility of the red band encourages its use for the spectrophotometric assay of the two chlorophylls. This method was developed by Ghosh and Sen-Gupta (1931), Zscheile (1934, 1935), Sprecher von Bernegg, Heierle and Almasy (1935), Haskin (1942), Comar and Zscheile (1942), Comar (1942) and Comar, Benne and Buteyn (1943). Measurements at two different wave lengths are required to calculate the concentrations of the two components. The use of the absorption maxima at 642.5 and 660 mix permits the most sensitive determination, but requires precise work, since the extinction values in the sharp absorption peaks are very sensitive to variations in the width of the spectrometer slit or to sUght errors in the adjustment of the monochro- mator. Cross-checks at other wave lengths are therefore desirable. All errors could be eliminated by the use of monochromatic Ught; but the spectrum of the mercury arc — the usual source of monochromatic light in the laboratory — does not contain suitable Unes in the red and orange regions. As pointed out once before, particularly wide discrepancies can be noted between the different extinction curves in the blue-violet region {cf. Table 21. IB). It was noted by Albers (1941) that the subsidiary violet band (situated, in ethereal solution, at 410 mju in chlorophyll a and 430 m/i in chlorophyll b) sometimes appears as a slight hump on the main band, and sometimes as a separate peak, almost as prominent as the main one. The observed variations in the maximum height of the violet peak may be caused by the more or less complete separation of this doublet structure. Differences of this type cannot be attributed to the presence of caro- tenoids, or other impurities. A renewed photometric study of this spectral region is desirable. If the deviations will not disappear upon further puri- fication of the material and improvement of the photoelectric methods, one may have to consider, as a possible explanation, the existence of tautomers. 608 ABSORPTION SPECTRA OF PIGMENTS 7.V VITRO CHAP. 21 H2C=CH CH3* HT rn .C, ,C C i H I C N N C V ., \. _/ H,c-c< El I ir |c-CH, „^ o< m \ \ m\ /\ \ H \ c c ^c' H^i I,,".: CH2 HC C. H I lio 9. u 1 . CH2 HC C. ^ 1. .iio 9 ^^^y'°'^ I I \ (Phytol) I ' T X H39C20OOC-CH2 COOCH H r mr L i 0 H39C20OOC — CH2 COOCH A B H2C=CH CHj* ' H I H3C-C(r^ I J C jj >C-C2H, C N N C HC S Mg fl CH C N N C 3 " \/ ^J< V CHg HC C (Phytol) I I \ H39C20OOC— CHj COOCH Fig. 21.2. Chlorophyll a structure according to Hans Fischer. A, B and C are three isomeric or tautomeric (or mesomeric) structures, distinguished by the routing of the all-roimd conjugated ring system (heavy line) and the positions of the "semi-isolated" double bond and of the Mg— N bonds (all of which depend on this routing). The asterisk designates the position of a carbonyl group in chlorophyll h. A, semi-isolated double bond in nucleus III ; Mg bound to nuclei I and II. B, semi-isolated double bond in nucleus II; Mg bound to nuclei I and III. C, semi-isolated double bond in nucleus I; Mg bound to nuclei II and III. We recall that the three structures of chlorophyll, A, B and C, described in chapter 16, Vol. I, (c/. fig. 21.2), were characterized there as probably tautomeric rather than mesomeric. We also recall that Strain and Mann- ing found {cf. page 403) in the chromatograms of leaf extracts, two new CHLOROPHYLLS a AND h 609 chlorophylls, which they called a' and b' and interpreted as tautomers of chlorophylls a and b. The forms A, B and C have different double bond ar- rangements in the nonhydrogenated pyrrole nuclei, but the same hydro- genated nucleus IV. Since the red absorption band is somehow associated with the hydrogenation of this nucleus (cf. page 621), modifications A, B and C may have identical red bands. They could, however, differ in the positions or shapes of the blue-violet bands, associated with the conjugated porphin system as a whole. Erdman and Corwin (1946) noted the spectroscopic similarity of etio- porphyrin and A'"-methyl etioporphyrin, and deduced from this that the two "central" H-atoms in the porphin system must be fixed at definite nitrogen atoms for » 10"* sec. This supports the hypothesis that struc- tures such as those represented by the three formulas in figure 21.2 are tautomeric rather than mesomeric. We have used so far only data obtained with ethyl ether as solvent, since they alone offered the possibility of comparison between the results of several observers. Determinations of extinction curves of chlorophyll in solvents other than ether are Usted in Table 21.11. Table 21.11 Chlorophyll Extinction Measurements in Various Solvents Solvent Reference Methanol, Rabinovvitch and Weiss (1937) (ethyl chlorophyl- Ude); Harris and Zscheile (1943); Mc Brady and Livingston (1948, 1949) Ethanol Meyer (1939); Sprecher von Bernegg, Heierle and Almasy (1935) Butanol (71- and iso-) Harris and Zscheile ( 1943)" Octanol Harris and Zscheile (1943)" l-Propyl ether Harris and Zscheile (1943)" Dioxane Harris and Zscheile (1943)" Benzene .... Winterstein and Stein (1933); Hausser {cf. Fischer and Stern, 1940; Harris and Zscheile (1943) Cyclohexane Harris and Zscheile (1943) Acetone Mackinney (1940); Harris and Zscheile (1943) Carbon tetrachloride Harris and Zscheile (1943)" Methyl oleate Harris and Zscheile (1943)" Olive oil Harris and Zscheile (1943)" " Several of the curves of Harris and Zscheile (1943) are reproduced in figure 21.26. The absorption spectra of the two chlorophylls in the ultraviolet are shown in figure 21.3. The absorption remains considerable all the way down to 200 m/u; the most prominent band is a double band of chlorophyll b at 310 and 335 mju. In the spectrmn of chlorophyll a, distinct maxima 610 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 appear at 325 and 375 mn. Below the region shown in figure 21.3, both components have an absorption maximum at 250 m/i (asp. ^^ 30). An absorption band at 330 mix was first noticed in the spectrum of alcoholic extracts from nettle leaves by Lewkowitsch (1928). The absorption spectra of chlorophyll, ethyl chlorophyllide and phytol in the infrared were described by van Gulik (1914) and Stair and Coblentz 80 72 64 56 - 48 40 32 24 16 \ Chlorophyll b \_'' 260 280 300 320 340 360 WAVE LENGTH, m/i 380 400 420 Fig. 21.3. Ultraviolet spectrum of chlorophylls a and b in ethyl ether (after Harris and Zscheile 1943). Specific extinction coefficients; c in g./l. (1933). Chlorophyll is transparent between 0.7 and 3 n (this may be useful in preventing the overheating of leaves in direct sunlight). It has several absorption bands at 3-4, and 5.8 yu; most of them are found also in the spectrum of phytol, and are absent from that of ethyl chlorophyllide (c/. fig. 21.4 and Table 21.IIA); they thus belong to the phytyl chain rather 90 80 70 z* 60 o <" 50 UJ on I- z UJ o u. u. UJ o o Q. tr o (n m < Normal 390 430 470 510 550 590 630 670 710 WAVE LENGTH, m,i Fig. 21. 4A. Al)sorption spectrum of allomerized chlorophyll a in methanol (after Livingston 1948). A word must be said about the absorption spectrum of allomerized chloro- phyll. When chlorophyll in alcoholic solution is permitted to stand in air, it is "allomerized," ?'. e., according to Conant and Fischer, oxidized at the r(10) iitom (cf. Vol. I, page 4(i()). This reaction is catalyzed, according to Livingston and co-workers, by salts such as LaCls and BaCl2. Spectro- scopic evidence indicates that a similar, or identical, oxidation occurs also under the influence of iodine or bromine even in the absence of air: Fischer's chemical observations indicate that quinone has the same effect. According to Livingston (1948, 1949), allomerization of chlorophyll is characterized by a spectral change, illustrated by figure 21. 4A. Tlu^ two 614 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 curves in this figure were obtained by chromatographic separation of a partially allomerized solution; a curve identical with curve h is obtained by leaving a methanolic solution of chlorophyll a stand in air at 80°C. for two days, or by adding to it 2-3 equivalents of iodine, or by adding traces (10"^ ml.) of LaCls or CaCls. Standing in air, or addition of traces of iodine, has no effect on the spectrum of chlorophyll a in ether or carbon tetrachloride, except for slow general bleaching. If some methanol is added to the nonpolar solvent, allomerization proceeds more slowly, but ends with being as complete as in pure methanol; transfer of allomerized pro- duct into pure ether or carbon tetrachloride does not reverse the change. In cresol, allomerization also seems to occur, but is complicated by other changes, probably due to the acid nature of this solvent. Livingston and co-workers also have measured the absorption spectrum of the yellow and brown unstable intermediates in the reactions of chloro- phyll (in methanol) with FeCls (Vol. I, page 464) and with alkali (see "phase test," Vol. I, page 459). The results will be described in chapter 37, in the section dealing with new observations on the chemistry of chloro- phyll. 2. Absorption Spectra of Chlorophylls c and cl, Bacteriochlorophyll and Protochlorophyll The absorption spectrum of "chlorophyll c," also called chlorofucin (cf. Vol. I, page 406), is characterized, according to the earlier authors (e. g., Tswett), by a band in the region of 630 m/x. Strain and Manning (1942) determined the extinction curve of this compound, first by subtraction of the extinction curve of pure chlorophyll a from that of the chlorophyll ex- tract from brown algae, and later by direct spectrophotometry of isolated chlorophyll c. Figures 21. 5A and B show that the results of the two meth- ods are in approximate agreement. Two "chlorofucin" bands are situated in the orange and red — with peaks at 575.5 and 627 m/i, respectively, in methanol, and at 581 and 631 m/x, respectively, in 80% acetone. Figure 21. 5C also shows a band in the blue (at 446 m^u) almost ten times stronger. The ratio between the intensities of the bands in blue and red is thus much larger than in chlorophyll b (where it is about 3), not to speak of chlorophyll a (where the two bands are approximately equal in intensity, cf. Tables 21. IB and 21.VIII), but the general pattern of the spectrum is similar. Wa.ssink and Kersten (1946) and Tanada (1951) gave similar absorp- tion curves for a "chlorophyll c" fraction from chromatographic frac- tionation of the pigments of diatoms. The three absorption peaks ap- pear, in methanol, at about 450, 590 and 635 m/u, with the first band about ten times more intense than the other two (cf. p. 623). CHLOROPHYLLS C AND d, BACTERIO- AND PROTOCHLOROPHYLL 615 c o u Egregia Extract Extract minus a I I I I 580 610 (A) A « /A X "/^ / \ c o o Chlorofucin - -Ads. -Partition 1 1 1 1 0.1 500 600 -ii " " - X U5 c o o - L^^ :i J ' 1 1 500 600 : i i3 X \ \ c o o \\ \ -Vy \ I I I I I I I I ' T 580 610 WAVE LENGTH, m/i 400 500 600 WAVE LENGTH, m/i (C) 400 500 600 700 WAVE LENGTH, m^ ID) I I I I I I I I I I I I - - IsochI d' Chid' V Chi a' f\ « - V // \ o> \l 1 \ o w II w w // \\ +• VI » W .»_ w n \\ in _ "\\ tt M c \\ // i\ o w n \\ o - i / Chl d « ~ IsochI d Chio -i 0.025 _ 1 1 1 1 1 1 1 1 1 1 400 500 600 WAVE LENGTH, m/i 700 660 670 680 690 WAVE LENGTH, m/i 700 Fig. 21.5. Spectrum of chlorophyll c (chlorofucin) in methanol. Absorption coeffi- cients, a, in relative units (after Strain and Manning 1942). (A) Chlorofucin from Egregia, a brown alga (spectrum calculated by difference). (5) Same spectrum measured with chlorofucin prepared by adsorption (dots) or partition (solid line). (C) Chlorofucin from Nitzschia, a diatom (measured spectrum). {D, E, F) Absorption spectra of chloro- phylls a, d, d', and of isochlorophylls d and d'; absorption coefficients, a, in relative units (after Manning and Strain 1943). Open circles in E, absorption coefficients of isochlorophyll d. 61G ABSORPTION SPECTRA OF PIGMENTS 7.V VITRO CHAP. 21 The absorption spectrum of the '' chlorophyll d" of red algae also was determined by Manning and Strain (1943). Since this investigation was j only briefly mentioned in Vokime I, a few words may be said here about this newly discovered pigment. Its presence is revealed by a bulge on the red side of the chlorophyll a band, observed in the spectra of methanol extracts from red algae. In pure chlorophyll a solution in methanol, the ratio of extinction coefficients at 665 and 700 m^u is 90; in extracts from twenty red algae, this ratio was between 15 and 65. Short extraction leads to products with even lower I'atios ; r. g. , t wo minutes extraction of Gigarlina agardhii gave a product Math a ratio of only 10:1. Apparently, chlorophyll d is much more easily extracted by methanol than chlorophyll a. Chromato- graphic purification can be used for the preparation of pin-e chlorophyll d. Figure 21.51) shows the spectrum of this pigment in methanol. In ethyl ether, the band maxima of chlorophyll d lie at 686 and 445 myu, and below 395 niju. In spectrum as well as in solubility and other chemical properties, chlorophyll d resembles chlorophyll a more than chlorophyll h. Careful search for chlorophjdls h and c in the chromatograms of pigments from red algae gave negative results (the upper limit foi- tlie content of chlorophyll h is 0.3% of that of chlorophyll a). The isornerization of chlorophyll d was mentioned in chapter 16 (Vol. I). It occurs in the methanolic solution upon standing in the dark, in the presence or absence of air, and leads to three new pigments, which seem to be interconvertible. They were designated chlorophyll d', isochloro- phyll d and isochlorophyll d'. The spectra of the components d and d', respectively, are very similar, and the same is true of those of the isomers iso-d and iso-c?'. The latter two spectra are almost identical with those of the chlorophylls a and a' (cf. fig. 21. 5E) ; but the maximum of chlorophyll iso-d and iso-d' lies about 5 mfu. further toward the blue (cf. fig. 21.5 F). The red band of chlorophyll d (and d') lies about 37 m^ further toward the infrared than that of chlorophylls a and a' (cf. fig. 21.5F). Isochlorophylls d and d' are not found in fresh extracts from algae. The conversion d -^ iso-d appears to be slower than the conversions d-^ d' and iso-d -^ iso-d'. Four interconvertible pheophytins with diffei'ent spectra were pro- duced by the action of acids on the four d-pigments; but successive treat- ment with alkali and acid led to spectroscopically identical products for all four d-pigments. These products are distinct from the compounds ob- tained by a similar treatment of either of the two a-pigments. Table 21. Ill illustrates the relationships between the six pigments a, a', d, d', iso-d and iso-d' and their transformation products. The main absorption band of the bacteriocMorophyll of purple bacteria CHLOROPHYLLS C AND d, BACTEKIO- AND PROTOCHLOKOPHYLL 017 Table 21. Ill ISOMERIZATION AND OtHER REACTIONS OF CHLOROPHYLLS d AND a " Chlorophyll d' It Chlorophyll d Isochlorophyll d' Gochlorophyll d CH-Mgl Pheophytin d KOH KOH — Isopheophytin d KOH KOH 681 m^ 698 m;i 658 m^ 636 m^ CHjCOOH CHjCOOH 672 m^ 691 Wfi 660 m/i 672 m/i HCI HCI HCI HCI SSOmjii " After Maiming and Strain (1943). Chlorophyll u' Chlorophyll a ,. HCI CHjMgl Pheophytin a KOH KOH 662 m^ 649 my. CHjCOOH 667 m^ 652 m^ HCI Ihci — f — 667 m^ 0.20 Fig. 21.6. Relative absorption curve of green pigment from bacterium Spi- rillum rubrum in methanol (after French 1937). 0.2 ml. moist cells extracted in the dark at 0° with 5.2 ml. absolute methanol Extract kept in dark and measured at room temperature with very weak light. A similar curve was given by Vermeulen, Wassink and Reman (1937) for alcoholic extract from Chromaiium (major peak at 790 m/n, minor peaks at 705, 600, 510, 470 and 440 mM). 600 800 WAVE LENGTH, m/i lies in the near infrared and has not yet been measured precisely; its gen- eral shape is shown by figure 21.(). According to this figure, the absorp- tion spectrum of bacteriochlorophyll contains three main bands — one in the infrared (770 m/i in methanol, shown also in fig. 21.21), one in the orange (005 m/x) and one in the violet (~400 ni/i). One may assume {cf. below, page 622) that the two latter bauds are analogous to the red and the blue- 618 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 violet bands of chlorophyll a, respectively (the whole system having been shifted by about 60 m^t toward shorter waves), while the band in the near infrared has no analogue in the spectrum of ordinary chlorophyll. The status of the "orange" bacteriochlorophyll band needs additional clarification. Our interpretation is based on figure 21.6. A strong band on the short-wave side of the main red band is also recognizable in the spec- trum of bacteriopheophytin as observed by French (fig. 21.21), but is situated much further toward the green, at 530 m^; in addition, there is a weaker band at 680 m/x and indications of a still weaker one at about 630 m/n. Dutch observers state (see page 702, chapter 22) that "alcoholic 1.0- r;0.5- c o u - 1 / iK 1 ■ 3 ^ ■yj \ 2 ENERGY.ev. \s,/3 i,9 ISX^I 1 1 1 1 620 640 660 680 700 WAVE LENGTH, m^ 720 Fig. 21.7. Red absorption bands of al- coholic extracts of (1) green alga, Chlo- rella, (2) green sulfur bacteria, and (3) a blue-green alga, Oscillaloria (after Katz and Wassink 1939). Q. to 200 700 300 400 500 600 WAVE LENGTH, m^ 21.8. Absorption spectrum of protochlorophyll after Rudolph (1933) in the insert, and after Koski and Smith (1948) in the main figure. Fig. extracts from purple bacteria show only one absorption maximum at 774 m/x," but it is not certain how wide a region this statement is supposed to cover. The extract absorption curves in figures 21.30A and B extend only down to 730-740 m/x. One question to be clarified is the possible contribution to the ab- sorption curves of bacteriochlorophylls of derivatives analogous to chloro- phylls h, c and d. The existence of pigments of this kind was suggested by Seybold and Egle (1939) (cf. Vol. I, page 407), who gave some provisional figures for the positions of their absorption bands. The absorption spectrum of bacterioviridin — the pigment of green bac- teria (cf. Vol. I, page 445) — was observed bj'' Metzner (1922) and, more recently, by Katz and Wassink (1939). Apparently, it is very similar to that of chlorophjdl a. Figure 21.7 shows the red adsorption band in SPECTRUM AND STRUCTURE OF PORPHIN DERIVATIVES 619 alcoholic extracts from ChJorcUa (green alga), Chlorobium limicola (green bacteria) and Oscillatoria (blue alga). The difference between the curves 1 and 3 can be attributed to the absence of chlorophyll h in blue algae; while the larger difference between the curves 3 and 2 indicates a chemical distinction between chlorophyll a and bacterioviridin. The absorption peak of the latter pigment in ethanol lies at 668 m/x, while that of the former one is situated at about 662 m/x. The absorption spectrum of protochlorophyU (from squash seeds) was described by Noack and Kiessling (1929, 1930, 1931) as well as by Rudolph (1933), Seybold (1937) and Koski and Smith (1948). The long-wave bands are listed in Table 21.IV; the whole spectrum is shown in figure 21.8. Table 21. IV Absorption Bands of Protochlorophyll In chloroform-pyridine In ether Noack and Kiessling Rudolph (19.33) Seybold (1937) (1929, 1930) Component a Component b X, m/i Order of intensi- ties X, rufji X, m/i Order of intensi- ties X, m)i Order of intensi- ties 641-621 582-567 540-521 1 2 3 621 (602) 571 536 650-620 620-603 592-572 572-555 545-530 <480 1 3 2 5 4 645-632 632-605 620-605 587-570 545-520 <500 3 1 4 2 5 3. Relation between Absorption Spectrum and Molecular Structure of Porphin Derivatives To understand the role of chlorophyll in photosynthesis, it would be important to have a detailed knowledge of the nature of the lowest excited state of the chlorophyll molecule, since sensitization must be due to the interaction of chlorophyll in this excited state with the primary sensitiza- tion substrate or substrates (e. g., with the C02-acceptor complex {CO2}, or with the oxidant {H2O} ; cf. Vol. I, chapter 7). If this interaction is in the nature of a reversible oxidation-reduction (which is probable) analysis of the nature of the excited state may permit conclusions as to the type of oxidations (or reductions) most likely to be involved. Theoretical analy- sis of porphin spectra was initiated too late for use in the following discus- sion, which is based on empirical relationships only. The papers by Kuhn (1949), Simpson (1949), and Piatt and co-workers (1950) will be summar- ized in the last chapter. 020 l.bxiu^ tt - 1.4 1- 2 // V, in benzene 5 '2 ' v u. ' V L. 1 I, uj 10 r \. o r o z 0.8 _ , o t- E^ 06 1 O 11 A in V /\ CO 0.4 Y /, \^ < V / ' V EC V /\ /' V 3 0.2 V'N // \v o x. ^/^' >* s 1 1 ^^"^^ 1 l^'-r'^^ k-«*4^_ 480 640 520 560 600 WAVE LENGTH, m/i Fig, 21.9. Absorption spectrum of porphin (;ifter Stern, Wenderlein and Molvig 193G). green X red green \ red green X red Fig. 21.10. Typical spectra of porphin derivatives above 500 m^i (not showing the strongest band, at 420 niju) (after Stern 1938). (1, 2, 3) porphyrin spectra; (4. 5) chlorin spectra; (6) bacteriochloriu spectrum. SPECTRUM AND STRUCTURE OF PORPHIN DERIVATIVES 02 1 The conjugated double bond system of porphin (which is the basic structure of all chlorophyll pigments as well as of the porphyrins) is a chromophore, capable of pi'txlucing sti'ong absorption bands in the visible and near ultraviolet. This is iUustrated b}' the absorption spectrum of the parent substance of the group, porphin (c/. fig. 21.9). This spectrum has a typical pattern of four bands between 480 and 700 mn, generally increasing in intensity' toward the violet, but with the third band from the red weaker than the second one. Stern and Wenderlein (1936) called this pattern the "phjdlo type" (fig. 21.10, 3); it is exhibited by many porphyrins. Other Electronic States B Vibrational States 3 o ' 1 o 1 V o> c s O o > t m Fig. 21.11. Interpretation of porphin spectrum in terms of transitions from the ground state, X, to two excited electronic states A (vibrational states Ao, Ai, A2, A3...) and B. 5.0x10 4.2 D ■5 2.6 E — Chloroporphyrin-e4 dimethyl ester — Chlorin-e4 dimettiyi ester 480 520 560 600 640 670 WAVE LENGTH, m/i Fig. 21.12. Spectroscopic effect of transition from the porphin to the dihydroporphin (chlorin) sys- tem (in dioxane) (after Stern and Wenderlein 1935). porphyrins have similar foiu'-band spectra with a somewhat different dis- tribution of intensities; these arc called by Stern the "etio t^^pe" and the "rhodo type" (fig. 21.10, 1 and 2). Compounds of these three types trans- mit freel}^ in the red; their color is red or purplish, hence the name "por- phyrin." In addition to the series of bands shown in figiue 21.9 and 21.10, all porphin derivatives have the so-called "Soi'et"band in the blue- violet (simi- lar to the blue- violet band of chlorophyll ). The distances between the foiu' l)ands in the green, yellow and red are such as to make plausible their in- terpretation as vil)rational l^ands, corresponding to a common electronic transition; while the blue-violet band stands apart and probably corre- sponds to a different electronic excitation (cf. Table 21. V and fig. 21.11). The porphin spectrum vmdergoes a far-reaching change (cf. spectra 4. and 5 in fig. 21.10) upon the hydrogenation of one pj^role nucleus, i.e., 622 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 upon the transition from the prophin system to the dihydroporphin {chlorin or rhodin) system. ("Rhoclins" are chlorins derived from chlorophyll 6; c/. page 447, Vol. I.) Chlorins are green (as their name implies) and this Table 21.V PoRPHiN Bands X (him) V (cm."0 613 16300 560.5 17820 517.5 19300 487 20530 430 23260 Ar (cm.~i) 1520 1480 1230 (2730) indicates strong absorption in the red, as well as in the blue and in the violet, and transparency in the middle of the visible spectrum. Figure 21.12 shows the spectrum of a compound of the etio type, chloroporphyrin- e4 dimethyl ester, together with that of its hydrogenation product, chlorin- e4 dimethyl ester. Wliile the bands in green and yellow appear weakened and displaced toward shorter waves in consequence of hydrogenation, a new and very intense band arises in the red (at about 660 m/x) , which com- pletely overshadows all the other bands of the orange-yellow system (its intensity is about equal to that of the blue-violet band, not shown in fig. 21.12). The hydrogenation of a second pyrrole nucleus, which is charac- teristic of bacteriochlorophyll and other derivatives of tctrahydroporphin (cf. Vol. I, page 447), causes a renewed transformation of the spectrum (fig. 21.10, 6). As stated on page 617 {cf. fig. 21.6), the strongest band of bacteriochlorophyll is situated in the near infrared. According to Stern and Pruckner (1939) the "bacterio" type bands in the visible region (red, yellow and green) are generally weaker (with the notable exception of one comparatively strong green band), and situated at shorter waves than the corresponding bands of the chlorin type. It thus seems that the ef- fect of the second hydrogenation is similar to that of the first one, i. e., the majority of the previously existing bands (at least, of those in the region above 450 m/x) are weakened and shifted toward the violet, while a new band of dominant intensity arises at the long-wave end of the spectrum We made use of this interpretation on page 618, when we suggested that the orange (rather than the infrared) Ijand of bacteriochlorophyll be considered as analogous to the red band of ordinary chlorophyll (see also page 631). The rule that the dominant red or infrared band in the absorption spectrum of porphin derivatives is associated with the presence of one or two hydrogenated nuclei may be very significant from the point of view of the mechanism of the photosensitizing action of chlorophyll and bacterio- chlorophyll, and it is therefore important to mention some apparent excep- tions to this rule. One such exception was mentioned on page 619 — SPECTRUM AND STRUCTURE OF PORPHIN DERIVATIVES 623 prolochlorophyll, a green compound which, according to H. Fischer {cf. Vol. I, page 445), nevertheless is a prophyrin rather than a chlorin or phor- bin. The spectrum of protochlorophyll (cf. fig. 21.8 and Table 21. IV) does not show the predominance of the red band over the bands in yellow and green to the same extent as do the typical spectra of dehydro- or tetra- hydroporphin derivatives; but it resembles these spectra somewhat more than it does the typical porphyrin spectra in figure 21.10. A re-examina- tion of the structure of the protochlorophyll molecule is therefore desirable. It is noteworthy that "protopheophytin," obtained from protochlorophyll by the action of acids, was found to have a typical porphyrin spectrum. A similar case is presented by chlorophyll c. As stated on p. 014, this pigment has an arrangement of bands similar to those of the chloro- phylls a and b, but the red ))and is very weak compared to the Soret band ; in fact, the spectra of chlorophyll c (insert in fig. 21. 5C) and protochloro- phyll (fig. 21.8) are extremely similar. It is therefore significant that Granick concluded, from chemical evidence, that chlorophyll c, too, is a porphin rather than a chlorin derivative (cf. chapter 37). Another interesting problem of the same character was raised by an investigation of Aronoff and Calvin (1943). They prepared (by condensa- tion of benzaldehyde with pyrrole) several compounds that tliey interpreted as isomeric hexaphenylporphins. Some of these compounds had porphyrin spectra of the "etio type" (fig. 21.10, 1), but others had spectra with a pre- dominant sharp band in the red (fig. 21.13, curve B). The authors thought at first that these green isomers may contain one pyrrole nucleus turned around, placing its N atom on an outside corner. Rabinowitch (1944) suggested they could perhaps be interpreted as chlorins: Chlorins isomeric with hexaphenylporphin could be formed, e. g., by attachment of one phenyl group to a pyrrole nucleus, and shifting of the two liberated hydrogen atoms to another pyrrole nucleus, thus: Tetrophenylporphin Tetrophenylchlorin 624 ABSORPTION SPECTRA OF PIGMENTS 7.V VITRO CHAP. 21 Subsequently, Calvin, Ball and Aronoff (1943) found indications that the two "isomers" whose spectra are shown in figure 21.13 actually l)elong to two different reckiction levels of the porphin system. Thus, in this case at least, the dominant red band was confirmed as an indicator of partial hydrogenation. 3.0 X 10 2.5 2.0 « 1.5 H 50Q 650 550 600 WAVE LENGTH, m^ Fig. 21.13. Molar absorption spectra of two tetraphenylporphin "iso- mers" (after Aronoff and Calvin 1943). No such explanation can as yet be given to another observation of Aronoff and Calvin — that addition of hydrochloric acid to tetraphenylpor- phin solutions with spectra of the etio type causes a reversible transition to chlorin type (fig. 21.14). Aionoff and Calvin attributed this to salt formation: porphin + 2HC1-^ (porphin H2)++(C1~)2. The addition of 2 hydrogen ions thus appears to have the same effect on the porphin spectrum as does the addition of two hydrogen atoms. If this is true, the problem of the spectroscopic difference between prophins and chlorins becomes cognate SPECTRUM AND STRUCTURE OF PORPHIN DERIVATIVES 625 to the proljlem of acid-base color (;hanges (concerning the latter, see Epstein, Kariish and Rabinowitch 1941, and Lewis and Bigeleisen 1943). According to Pruckner (1942), imidoporphyrins (which differ from por- phin derivatives by the substitution of an XH gi-oiip for a bridge carbon) 0.800 0.700 0.600 0.500 3 0.400 0.300 0.200 0.100 500 550 650 700 600 WAVE LENGTH, m/x Fig. 21.14. Effect of increased acidity on spectrum of a tetraphenylporphin (after Aronoff and Calvin 1943). 10 ml. alcohol, containing 5 ml. 3 X 10~' ^^ solution of free base and variable amounts of hydrochloric acid. (1) no acid, (2) 0.0204 A'' HCl, 0.5 ml., (5) same, 1 ml, (,/,) same, 2 ml, (5) same, 3.50 ml, (6) 6.3 A HCl, 5 ml. also have a strong absorption band in the red. She suggested tliat the appearance of this band is generally associated with increased molecular symmetry; but it is not quite clear why hydrogenation of one or two pyrrole nuclei, or substitution of NIT groups for (' atoms, should lead to higher symmetry. 626 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 The "c'hlorin type" spectrum remains almost unaffected by all trans- formations leading from the parent substance chlorin to chlorophyll a. This is in accordance with general experience as far as the introduction of methyl and ethyl groups is concerned. It is noteworthy, however, that the introduction of an unsaturated (vinyl) substituent in nucleus I also causes only a slight shift of the bands, as illustrated by figure 21.15. In other words, a difference of two hydrogen atoms in a side chain is almost without influence on the spectiaim, while a similar difference in the nucleus has a strong effect. The spectroscopic effect of carboxyl groups in chlorophyll also is small. The alcohols esterifying these carboxyls have a certain influence on the 6.4 X 10* 62 5.6 Rhodochlorin dimethyl ester Mesorhodoctilorin dimethyl ester (both in dioxane) o o E 470 490 520 550 580 610 640 670 WAVE LENGTH, m^ Fig. 21.15. Effect of vinyl group on chlorin spectrum in di- oxane (after Stern and Molvig 1937). intensity of the absorption bands (but none on their position) : The shorter the alcohol, the sharper the absorption peak. As an example, figure 21.16 shows the extinction curve of phytyl pheophorbide (pheoph^^tin), together with that of methyl pheophorbide. It is further noteworthy that closure of the carbocyclic ring, i. e., the transition from chlorin to phorbin, also hardly affects the spectrum at all, as shown by figure 21.17. The carhonyl group in nucleus II, whose pres- ence distinguishes chlorophyll b and its derivatives from the corresponding compounds of the a series, has a much stronger effect on the spectrum: The two chlorophylls have distinctly different colors — one blue-green and the other yellow-green. Figure 21.1 shows that this difference is caused by SPECTRUM AND STRTICTItrE OF PORPIIIN DERIVATIVES 027 different positions of the bhie-violet bands: That of chlorophyll a is con- fined to the violet and ultraviolet regions, allowing free transmission of blue light, whereas the a])S()rption peak of the 6 compound is situated in the blue, so that this compound transmits only green light. The arrangement of the weaker bands in the middle of the visible ii^gion is also affected by the carbonyl in position 5, to such an extent that Stern classified the spectra of chlorophyll b and its derivatives as a separate "rhodin" type (fig. 21.10, 5) distinct from the "chlorin" type (fig. 21.10, 4). 480 520 560 600 640 WAVE LENGTH, m>i Fig. 21.16. Effect of esterification on chloriri spectrum (after Stern and Mol- vig 1937). 670 480 520 560 600 WAVE LENGTH, m/i 640 670 Fig. 21.17. Effect of closure of car- bocyclic ring (transition chlorin — *■ phorbin) on chlorin spectrum in di- oxane (after Stern and Wenderlein 1935). The other carbonyl group of chlorophyll, the one in the carbocyclic ring (in position 9), has no pronounced influence on the spectrum of chlorin derivatives. This is striking because, in the case of porphyrins, a carbonyl group in a similar position affects the spectrum to a considerable degree. Stern (c/. Fischer and Stei-n 1940, page 343) suggested an explanation of this difference, based on Fischer's earlier assignment of the two extra hy- drogen atoms to nucleus III. By this assignment, the C=0 double bond in position 9 was removed from conjugation with other double bonds in the molecule (cf. Volume I, page 441), and this could explain why its effect 628 AnSORrTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 on the spectrum is less pronounced than that of the conjugated C=0 bond in position 3. As stated on page 441, Vohime I, Fischer later concluded from chemical degradation experiments that the two extra hydrogen atoms are located in nucleus IV. A certain difference between the conjugation in nuclei II and III exists, however, also in the latter structure (formula A, page 608) — namely, the double bond 3-4 in nucleus II is part of the all- round "aromatic" ring system, while the double band 5-6 in nucleus III is 1 0.0x10^ 9.0 Methyl chlorophyllide a Methyl pheophorbide a < ■A 480 520 560 600 640 670 WAVE LENGTH, m^ Fig. 21.18. Effect of magnesium on a porphin spectrum (in dioxane) (after Stern and Wenderlein 1936). merely in "one-sided" conjugation with this system. In structure B, on the other hand, bond 5-6 is part of the all-round conjugated system, and bond 3-4 is in one-sided conjugation. Finally, in structure C, both bonds, 3-4 and 5-6, participate in all-round conjugation. Thus, a difference be- tween the chromophoric effects of carboxyls in nuclei II and III appears plausible in structures A and B, but not in structure C. We note further that, according to Stern and Pruckner (1939), a carbonyl group in nucleus I also has no strong effect on the spectrum. This indicates that, with re- SPECTRUM AND STRUCTURE OF rORPITIN DERIVATIVES 020 spect to conjugation, the role of nucleus I is similar to that of nucleus III — and this points to structure B as the tme structure of chlorophyll, in prefer- ence to structure A, advocated by Fischer (page 443). (As mentioned on page 444, Vol. I, structure B also has the advantage of providing a direct link to bacteriochlorophyll.) If this interpretation of the spectroscopic data is correct, it means that the introduction of a carbonyl group has a stronger effect on the spec- 7.0 60 Dihydropheophorbide Pheoporphyrin 05 30000 25000 20000 60 5.0 4.0 3.0 6.0 Chlorophyll a Pheophorbide a 30000 25000 20000 ^' _i L ■ [ ■ ■ ■ I ■ • r • I ■ r ■! ■ I' 350 400 450 500 600 WAVE LENGTH, m/i 4.0- ' ' 'I ' ■ ■ I ■ ■ I' ■ I ■ r 'I — r- 350 400 450 500 600 WAVE LENGTH, m^ Fig. 2 1.1 9 A. Effect of porphin -> chlo- rin transition on the visible and ultraviolet spectrum (after Hagenbach, Auerbacher and Wiedemann 1936). 3.0 Chlorophyll b Pheophorbide b 30000 __i i_ 25000 20000 -I I ■ ' ■ ■ 1 ■ ■ ' . ' * -1 1 1 1 • ]• '\ • \ 350 400 450 500 600 WAVE LENGTH, m^ Fig. 21.19B. Extinction curves of chlorophyll and pheophorbides in the visible and near ultraviolet (after Hagenbach, Auerbacher and Wiede- mann 1936). trum if this group comes into conjugation with a C=^C bond, which is merely in one-sided conjugation with the "aromatic" system, than if it is attached directly to the latter system. Introduction of magnesium, into the molecule (transition from phorbin to phyllin) has the effect of enhancing further the main red absorption band, and of weakening the bands in the green, as illustrated by figure 21.18. The result is the beautiful pure green color of chlorophyll — so different from the dull olive-green of pheophytin. While the porphyrins, chlorins, phorbins and rhodins differ in their absorption spectra in the green, yellow, orange and red, their spectra in the G30 ABSORPTION SPKCTRA OF PIGMENTS IN VITRO CHAP. 21 violet and ultraviolet all show the same pattern. As pointed out by Stern (1939), neither the transition from porphin to chlorin nor the introduction of magnesium has much effect on the intensity of the blue-violet absorption band. This is shown by figures 21.19A and B. The blue-violet band is shifted by hydrogenation toward shorter waves, but suffers no appreciable change of intensity. Comparison of figure 21. 19 A and B shows that, while both magnesium and the extra hydrogen atoms enhance the main red absorption band, these two substituents have antagonistic effects on all the rest of the spec- trum, below GOO m.fx. 4. Some Theoretical Remarks on the Spectrum of Chlorophyll (a) The Term System In section 3 (c/. fig. 21.11) we interpreted the four absorption bands of porphin in yellow and green as vibrational bands belonging to the same band system (electronic transition X -^ A). A similar interpretation has been suggested by Prins (1934) for the bands of chlorophyll in the red, orange, yellow and green ; it is made plausible by the magnitude of the Ai* values given in Table 21. VI. (c/. the infrared absorption frequencies in Table 21.IIA). The blue-violet and the two ultraviolet bands are best inter- preted as separate electronic transitions. Table 21. VI Chlorophyll a Bands v (cm.~0 660 15100 612.5 16300 572.5 17500 527.5 18900 497.5 20100 427.5 23400 375 26700 325 30800 Ac (cm.~') 1200 1200 1400 1200 (3300) (3300) (4100) According to this interpretation, the term scheme of figure 21.11 could apply also to chlorophyll. However, instead of a (more or less) luii- form change in probability in the series of transitions X — > Ao, X -» Ai, X ^ A2 (as revealed by a gradual increase in intensity of the corresponding bands in porphin and its derivatives), one would have to postulate in the case of chlorophyll a predominant probability of the transition X —> Ao (to account for the outstanding intensity of the first absorption band at 660 m/i). Furthermore, the relationship between the spectra of porphins and dihydroporphins, illustrated by figure 21.12, suggests a shift of the bands in the hydrogenated compounds toward the shorter, rather than to- ward the longer, waves. Therefore, if the red band of the dihydroporphins is X -^ Ao, the first absorption band of the porphins must be interpreted as ANALYSIS OF CHLOROPHYLL SPECTRUM 631 X -^ Ai, and the band X -^ Ao must be considered missing because of low intensity. However, this interpretation is made untenable by the observa- tion— to be discussed in chapter 23— that both in porphins and in dihydro- porphins the main fluorescence bands are close to the first absorption band in the red — whether this band is the weakest of all absorption bands, as in porphin, or the strongest one, as in chlorophyll. This shows that in both cases, the first observed absorption bands lead to the vibration-free upper level Ao. If the first absorption band of porphin were X -> Ai (as tenta- tively suggested above), we would expect to find the main fluorescence B «3 ^1 Yo o > a> m Fig. 21.20. Hypothetical term system of the chlorins. Band X -♦ Ao is submerged by band X — »■ Yi. band some distance toward the red from it — in the approximate position of the "invisible" X -^ Ao absorption band (Rabinowitch 1944). Since this is not the case, an alternative hypothesis must be considered. It is represented in figure 21.20, and suggests that the hydrogenation of one pyrrole nucleus creates a new low electronic excitation state Y, situated a little lower than the level Ao, "inherited" from the nonhydrogenated system.* Figure 21.12 makes one suspect that the second chlorin band (612.5 m/i in chlorophyll), which is stronger than the corresponding band in the nonhydrogenated compound, may also belong to the system X — > F (as a second band of this system, X ^ Yi, . . .), and that it masks the weak l)and A^ -^ Ao. A similar interpretation can be suggested for the infra- red band of bacteriochlorophyll and other tetrahydroporphin dei'ivatives : We can attribute the strongest infrared band of bacteriochlorophyll (fig. * Another possibiHty — suggostetl by the spectra of protochlorophyll and chloro- phyll c — iR that the band A' -* I'o exists also before hydrogenation, but is strongly enhanced by the latter. 032 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 21.6) to a new electronic transition, X — > Z, added (in consequence of the hydrogenation of a second pyrrole nucleus) to the three transitions X ^ Y, X —^ A and X -^ B present in the spectrum of chlorophyll. Simultane- ously with the creation of a new excited electronic level Z, the "old" levels Y, A and B are shifted upward, thus accounting for the "violet shift" of all bacteriochlorophyll bands "inherited" from chlorophyll. The in- frared band X — > Z dominates the spectrum of bacteriochlorophyll to an even greater degree than the red band, X —> Y, dominates the visible spec- trum of ordinary chlorophyll. 500 600 700 800 WAVE LENGTH, m/i 900 Fig. 21.21. Absorption spectrum of bacteriopheophytin from Spirillum rubrum (after French 1940). Specific absorption coef- ficients, c, in mg./i., d in cm. Since the absolute extinction coefficients of bacteriochlorophyll are as yet unknown (fig. 21.6 gives only the optical densities), it is not certain whether the predominance of the 770 niM band is eau.sed by its great absohite lieight, or tjy a relatively low intensity of tlie otlier bands. A specific extinction curvt; of bacteriopheophyiin was given by PVench in a later paper (1940), and is reproduced in figure 21.21. It shows that the molar absorption coefficient of bacteriopheophytin (in methanol) reaches 2.7 X 10^ ANALYSIS OF CHLOROPHYLL SPECTRUM 033 in the maximum of the orange band, while the infrared peak is almost exactly twice as high. According to figure 21. Hi, tiic maxinmni absorption coefficient of ordinary pheo- phytin a in the red is about 4.2 X IQ' (in dioxane, wliere the peaks are usually sharper than in methanol). It thus seems that the dominant position of the infrared band is due both to its own outstanding intensity and to the comparative weakness of all other bands. The addition of a new low olectronic level in consefiiienee of each hydro- jienatioii step of Ihe poriihin system offer.s an iutei-estiMg inobleni for theo- retical (liscussit)n. OlThand, one would cxpecl increased satui'ation lo de- crease ratlier than to increase the niiniher of excited electronic states. If the red bands of dihydroporphin (and the infrared bands of tetra- hydropoiphin) are brought about (or enhanced) by the presence of electi'ons associated with the ad(Htional carbon-hychogen bonds, it seems pkiusible that hght could specifically activate the "extra" hydrogen atoms. This would make excited chlorophyll (or bacterioehlorophyll) an effective hydrogen donor — a property which may be of decisive importance for the photochemical function of this pigment. In Volume I (chapter 19, pages 552-554) the primary photochemical oxidation of chlorophyll was dis- cussed as a possible mechanism of sensitization in photosynthesis. This hypothesis would gain consideral)ly in plausi]:)ilit3^ if it could be j^roved that absorption of light actual)}' activates chlorophyll as a hydrogen donor. The effect of light on the chlorophyll-ferric iron equilibrium (c/. Vol. I, page 488) is the only observation at present that lends experimental sup- port to the concept of chlorophyll as a light-activated reductant. Stoll (1936) thought that the excitation of chlorophyll by hght activates especially its "odd" hydrogen atom in position 10. Studies by Krasnovsky {c.j. chapter 35) indicate the capacity of chloro- phyll to act also as a light-activated oxidant. (b) Life-Time of the Excited States of ChloropMjll The natural life-time of the state Y can be calculated from the integral area of the red absorption band Xo -^ Fo. Strictly speaking, one should take into account also the probabilities of transitions from Fo to the vibrating states Xi,2. . . , which could be derived from the relative intensi- ties of the successive bands in the fluorescence spectrum (c/. fig. 23.2); but we are con- cerned here with orders of magnitude only. Prins (1934), who made this integration, obtained for the number of "absorption electrons" (i. e., the number of harmonic oscillators with the charge e that could account for the observed intensity of absorption ac- cording to classical electromagnetic theory) : 634 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 (21.1) /-C 0.24 per molecule chlorophyll a (in ethanol) 22 per molecule chlorophyll b (in ethanol) In the quantum theory, / is the measure of the transition probability between the states Xo and Fo; and the reciprocal of the transition probabil- ity is the mean life-time, t, of the excited state (as far as the latter is limited only by the fluorescent transition Fo -^ Xo) . The relation between / and r is: (21.2) 3 ^, mc 8 e^ir 2 1 _ 1.96 X 10- where m, c, e and tf have the usual meaning. The theoretical mean life- time of chlorophyll molecules in the lowest excited state (reached liy ab- sorption of red light) is therefore : (21.3) = { 8.2 X 10~8 sec. for chlorophyll a (in ethanol) 8.9 X 10~* sec. for chlorophyll h (in ethanol) The higher intensity of the blue-violet absorption band (particularly in chlorophyll b) indicates that the natural life-time of the excited state B, is somewhat shorter than that of state F — probably 5 X 10 ~^ sec. or less. NUCLEAR DISTANCE Fig. 21.22. Crossing of potential curves. The actual life-times of chlorophyll in states B, A and F are considerably shorter than the "natural" life-times. This is indicated by the complete absence of fluorescence originating in the levels B and A and the relative weakness of fluorescence originating in level F. From the complete absence of blue-violet fluorescence in chlorophyll solutions (c/. page 748) it follows that the energy of state B must be dissi- pated within 5 X 10 ~'^ sec. or less. (With the natural life-time of 5 X INFLUENCE OF MEDIUM 635 10~^ sec, energy dissipation within 5 X 10~^^ sec. would reduce the yield of fluorescence to <0.01% and thus make it practically unobservable.) Similar considerations apply to state A. The ease with which states .4 and B are transformed into state Y (as shown by the excitation of red fluorescence with yellow or blue light, cf. page 748) indicates that the po- tential energies of states A, Y and B (plotted against some appropriate "configiH'ation co-ordinate") give curves of the type shown in figure 21.22. At point M, the electronic excitation energy of state B is easily transformed into the (smaller) electronic excitation energy of state Y, plus a large amount of vibrational energy. The yield of red fluorescence of chlorophj'll in solution is of the order of 10% (cf. chapter 23). This shows that the actual life-time of state Y in solution is of the order of one tenth of the above calculated "natural" life- time, i. c, about 5 X 10"^ sec. The various "quenching" processes that may contribute to this shortening of the life-time of excited molecules will be discussed in chapter 23 (page 755). B. Influence of Medium on Absorption Spectrum of Chlorophyll and Bacteriochlorophyll* We have spoken so far of the absorption spectra of chlorophyll antl its derivatives in solution as though they were determined only by the chemi- cal structure of these compounds. However, the absorption spectra also are affected by changes in the nature of the solvent, and even more strongly by adsorption on solids, or by the formation of colloidal aggregates. These spectroscopic changes are caused by interaction between the light-absorbing molecules and their neighbors. Kundt had noticed as early as 1878 that the absorption bands of many dyestuffs are shifted toward longer waves with increasing refractivity of the solvent. This relation appears plausible in the light of London's theor}^, which establishes a connection between the capacity of molecules to refract light and the intensity of intermolecular forces — both properties being determined by polanzability of the molecules. Parallelism between molecular attraction and polarizability presumes the absence of other, chemical or physical forces between the interacting molecules. Such forces may arise if the molecules bear electric charges, dipole moments, or possess incompletely saturated valencies; therefore, we can expect Kundt 's rule to apply primarily only to neutral, nonpolar, saturated molecules in nonpolar, saturated solvents. The rule may also apply to series of polar or nonsaturated molecules in which the additional interactions due to dipoles or residual valencies are approximately constant (e. g., to a homologous series of alcohols). * Bibliography, page 669. 636 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 London's theory of molecular attraction predicts that solvents with a high refractive index (i. e., strong polarizability) should exercise strong at- traction on solute molecules, thereby causing considerable deformation of their electronic systems and considerable shifts of their energy levels. The direction of the resulting displacement of the absorption bands depends on the comparative polarizability of the solute molecules in the ground state and in the excited state. Figure 21.23 shows that the excitation \ i Chi* \ ^ a: Hi z u \ / s* hv o z o A \ chr.„, / Chi St ^ / y' a: >s^ ^ / _ 1- o UJ _I UJ \ L ''"sol \ Chi, ol / s V^ '0 NUCLEAR DISTANCE Fig. 21.23. Influence of solvation energy on energy of excitation (if B* > S,hp > hv.oi). S + hi' = S* + hv^o\ -5. d + AS = hv,o\ - hv. energy of a bound (e. g., solvated) molecule, /ij^soi.j is related to that of the free molecule, hv, by the equation : (21.4) hv,o\. = hp + S - S* = hv + AS where AS is the difference between solvation energies of the normal and the excited molecule. If A*S < 0, i. e., if the excited molecule is attracted by the solvent more strongly than the normal one, hv^oi. is smaller than hv and the absorption band is shifted toward the red (as postulated by Kundt). That A.S should be negative is plausible, since excited molecules usually have a looser electronic structure and are therefore more easily polarizable than the normal ones. Figure 21.23 shows that equation (21.4) is only correct if the equilibrium distances r and ro are identical. This will not generally be the case (r * is likely to be somewhat larger than ro). The e.xact equation for the "red shift" according to figure 21.23 is: SOT.VKNT EFFECT 037 (21.4a) ^"soi. - hv = AS + 8 the red shift thus being decreased by the amount 8. Probably no other compound has been so often studied from the point of view of Kundt's rule as chlorophyll. The origin of this interest was the fact, first noticed by Hagenbach in 1870, that the maximum of the red band of chlorophyll in living plants is displaced by about 20 mn toward tbe red end of the spectmm. compai-ed to its position in solution, (lerland (1871) found that a similar displacement occurs in the case of the absorption bands of chlorophyll in the yellow and green. It was early suggested that this position of the bands indicates a peculiar state of chlorophyll in the living cell, and numerous attempts have been made to reproduce this state in vitro. We will see, however, in the following review of experimental data, that the "red shift" is not a specific effect, and could be caused by various types of aggregation or complexing. We will first deal with chlorophyll solutions in different organic sol- vents, and then with colloidal solutions, complexes and adsorbates, in which chlorophyll is associated with proteins, lipides or other carrier;-. 1. Solvent Effect in the Spectra of Chlorophyll and Bactericchlorcphyll Chlorophyll was one of the dyestuffs whose stutiy caused Kundt (1878) to postulate a relation between the refractive index of the solvent and the position of the absorption bands, which became known as "Kundt's rule." Since then, numerous observations have been made of the spectrum of chlorophyll in different solvents, e. g., by Baas-Becking and Koning (1934), Hubert (1935), Wakkie (1935), Katz and Wassink (1939), Bicrmacher (1939), Egle (1939) and Harris and Zscheile (1943). The agreement be- tween the different authors is not very satisfactory— for example, Hubert found 662.5 mju for the position of the absorption peak of chlorophyll a in methanol, and 604 m/x for its position in ether, while Katz and Wassink obtained, for the same solvents, 604 and 661 m^, Harris and Zscheile 664 and 660 mfi, and Bass-Becking and Koning 656 and 666 m/x, respectively. Many discrepancies probably have been caused by the use of poor spectrophotometric equipment; Mackinney (1938) stressed, for example, the errors inherent in the identification of the band maximum with the so- called band "axis" (cf. above, page 607, and chapter 23, page 744). Other, and perhaps more important, differences may have been caused by the preparations used — often leaf extracts containing chlorophylls a and b in unknown proportions. It is difficult enough to obtain spectroscopic reproducibility even with purified preparations of a single chlorophyll com- ponent! Small solvent impurities, too, may strongly change the spectrum (c/. p. 647). 638 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 to IS > a o w. H K w O o M o o Ph o H « o n M o 03 o CO O 0) X) O o ^ j3^ 0jO5 m 3 lO lO N a C3 OS — « T3 H- 03. Sm ^.^; + bO + a) CO «2 CO O CO CO % 1 ■CO o CO »o lO iC lO lO CO ST) CO CD CO 00 CO CO CO CO CO 1 CO ' CO 658 CO CO CO CO >0 CO CO 00 IM "0 lO lO no -H lO TJH •>* Ttl lO CO CO CO CO CO C3 CO CO Oi lO CO tH (N (m' lO CO CO CO CO CO CO CO CO CO CO lO -* ■* O (M (M -^ lO lO CO CO CO iM CO CO CO CO r^ CO -0> CO CO CO CO o o ^ CO O lO CO CO o o CO (M o o •^ ^ o CO > A ^ o3 S w to a O(NO(NC> CO CO ^^^'CilOOOiOiOCDO^ CoCoCOCoCOCDCOCOt^^. cocococo^^cocococo CD ' t^ O CO TO CO lO O CD CO CO CD -r CD CD CD CD CD CD CDOO— it^C5(NTOCDCiOCO CDClOOO-*iiCOOiOTOiO Tto 00 CD CO TO TO CO CO (M CD CO 00 CD CO CO CD 05 .2. > O o > c3 a m > O fl m 03 w o ■*^ e a o o 00 be s c O +3 (B hJ O 56 Oh o 3 -« oi aj CD ■O -C c3 M sq C c3 640 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 Tables 21. VII and 21. VIII contain a summary of the experimental re- sults. Considering the former table as a whole, one finds only a very rough confirmation of Kundt's rule — it consists mainly in the fact that ^max. values greater than 670 ran are found only with solvents whose refrac- tive index is greater than 1.5. Between Ud = 1.33 and 1.5, the variations of Xmax. appear small and irregular. However, if one sorts out solvents of the same chemical type, a more regular shift at least of the red absorption maximum becomes apparent. Figure 21.24 summarizes, as an example, the data of Katzand Wassinkand Harris and Zscheile for nonpolar solvents (curve A) and for alcohols (curve B). 672 670 I" 668 iLl 666 UJ 664 662 660 Carbon disulfide. NONPOLAR SOLVENTS 2-Et-l-hexanol POLAR SOLVENTS P l-Amylol g JJ I'Butanol '^2-Me-l- "Ethanol P^opanol Methanol Xylene Benzene ^Toluene Pentane ® ©■iCyclohexane Hexane 1.50 1.60 Fig. 21.24. Absorption maxima of chlorophyll a in solvents of different refractivity. The shift of the blue band remains irregular, even in selected series of solvents. The highest value of X^ax. of the blue band was found in the solvent with the lowest refractive index, methanol. (This fa(;t may per- haps be attributed to a specific sensitivity of the blue-violet band to tautom- erization— a hypothesis discussed on page 607; tautomerization equilibra are known to be strongly affected by solvent changes.) Livingston and co-workers (1949) noted that, in the series of chlorophyll a solutions in alcohols (from methanol to octanol), the relation between the two peaks in the blue-violet region changes systematically. In methanol, they are equally high and separated onl}' by a dip; in octanol, the short- SOLVENT EFFECT 041 wave peak apiiears as a lower liiimp separated from the main peak by a trough. Wakkie (1935), too, divided solvents into classes and gave four separate X^ax. = /(^) curves — one for nonpolar solvents, one for weakly polar solvents (ethers, ketones), one for alcohols and one for colloidal solu- tions in water and glycerol. However, the last curve is not directly com- parable with the other three, since, as we shall see later, the position of the X »- o z Ijj I 1 2 760- 640 770- ■650 2 N \ \. 5 o 780- •660 • • -- 3 5 • 4 2 sOO 4 • 790- ■670 1.3 1.4 "H, •2^ 1.5 1 1.6 800 1. 1 1 1 1 1 6 1.8 1 1 2.0 2.2 1 2.4 1 2.6 1 2.8 1 3.0 1 3.2 1 3.4 3. (/7^-|)/(/7^-h2)' Fig. 21.25. Wave lengths of absorption maxima of green pigment extracts in different organic solvents, in relation to their refractive indices. (1) Carbon di- sulfide, (2) pyridine, (3) chloroform, (4) l)enzene, (5) carbon tetrachloride, (6) ethanol, (7) ether, (8) methanol, (9) acetone (after Katz and Wassink 1939). Top scale: extracts from purple sulfur bacteria (strain D) (O). Bottom scale: extracts from ChlorcUa (•). absorption bands of colloidal chlorophyll solutions depends not only on the solvent, but also on the degree of dispersion of the sol. Wakkie's curves (as well as our two curves in fig. 21.24) are displaced toward longer waves with increasing dipole moment of the solvent — a rela- tion that can be explained by the superposition of attraction forces be- tween solvent and solute caused by permanent 'polarization upon forces due to polarizabiliiy. This dipole effect is stronger for chlorophyll b than 642 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 for chlorophyll a — the former being the more po'ar of the two compounds. For example, according to Harris and Zscheile, the red band of chlorophyll a lies at 660 mn in ether and 664 m^t in methanol (AX = 4 m^), while the corresponding values for chlorophyll b are 642.5 and 651 m^t, respectively (AX = 8.5 m/x). Hydrogen bonding, too, may have to be taken into con- sideration. If one compares the effects of varying refractivity of the solvents on the spectra of different homologous solutes, one may expect the solute with the stronger polarizability to exhibit the strongest shift. Polarizability in- creases with the intensity and wave length of the first absorption band. Thus, the spectra of dyestuffs should be more sensitive to solvent changes than the spectra of noncolored substances, and the sensitivity of dyes of dif- ferent color should increase with the shift of the main absorption band to- ward longer waves. This is confirmed by the finding of Pruckner (1940) that the solvent effect increases strongly from porphins through dihydro- porphins (chlorins and phorbins) to tetrahydroporphins (bacteriochloro- phyll). The first absorption band of chlorophyll is situated further toward the red and is more intense than the first absorption band of the porphyrins ; and the same is true of the first absorption band of bacteriochlorophyll compared to that of chlorophyll. The two (or four) additional hydrogen atoms contained in these compounds contribute electrons that are easily excitable by light (thus giving rise to long-wave absorption bands) and easily displaceable in electric fields (thus producing strong polarizability). Figure 21.25 shows the shifts of the band maxima of bacteriochlorophyll and ordinary chlorophyll in the same solvents. This figure indicates that on the w^ave length scale the solvent effect is about twice as strong for the first band of bacteriochlorophyll as for the first band of ordinary chloro- phyll. Katz and Wassink (1939) extrapolated the curves in figure 21.25 to vacuum (refractive index 1) and predicted that the absorption peak of free chlorophyll molecules — if it can ever be determined — will be found at 648 ± 5 m/i, and that of free bacteriochlorophyll molecvdes at about 740 m/x. In piperidine solution, the absorption peak of chlorophyll lies at 642 niyu, i. e., be- yond the extrapolated Hmit for the free molecule. This demonstrates the existence of exceptions to Kundt's rule, probably caused by specific chemical interactions between solvent and solute. Other (less striking) exceptions from Kundt's rule have been dis- cussed by Mackinney (1938, 1940) and Egle (1939). Theoretically, it would be more appropriate to plot, in figures 21.24 and 21.25, wave numbers (or frequencies) since these are proportional to energies and therefore bear direct relationship to the terms in equation (21.4). In a narrow spectral range, such as is used here for a given band of a single pigment, linear extrapolation on a wave number scale would give results not significantly different from those obtained by linear extrapola- SOLVENT EFFECT 643 tion on a wave length scale. In the comparison of the shifts of bands in different parts of the spectrum, on the other hand, the use of wave lengths may be quite misleading; for example, a shift by 10 m/x at 440 m^ is equivalent, on the energy scale, to a shift by 22.5 m^ at 660 mju. The effect of solvents on absorption maxima of chlorophyll other than the main red peak has not been studied systematically, but it is known that in general all of them experience shifts toward longer waves with increasing refractivity of the solvent (see e. g., Egle 1939). Exact measurements of this shift may prove useful in the interpretation of the spectrum, since ab- sorption bands that lead to the same electronic upper state can be expected, according to page 636, to show the same shift. Krasnovsky et al. (1949) found bands II and III to be shifted, in pyridine, by 30-35 m^x (to 643 and 622 m^), while other bands were shifted by 8-15 m/x only. The solvent effect is not restricted to hand shifts, but also involves changes in the width and shape of the bands and (perhaps as a conse- quence of these changes) alterations in the absolute and relative intensities of the band maxima. In this case, too, bands leading to the same excited electronic level must show a similarity of behavior (cf. Pruckner 1940). In the case of chlorophyll, the ratio between the intensities of the blue- violet and the red peak is quite different in different solvents. While Table 21. IB showed a value of 1.3 for ether solutions of chlorophyll a, the ratio drops to 1 in methanol (Mackinney 1940, Albers 1941, Harris and Zscheile 1943) and rises to approximately 1.5 in dioxane (Harris and Zscheile 1943) and perhaps also in benzene (according to Hausser's meas- urements, cf. Fischer and Stern 1940; not confirmed by Table 21. VIII). Figure 21.26 shows the absorption curves of chlorophylls a and h in a variety of solvents, according to Harris and Zscheile. Table 21. VIII Relative Intensity of Chlorophyll Absorption Peaks in Different Solvents (after Harris and Zscheile 1943) Blue max. /red max. Solvent Chi. a Chi. b Methanol 1.00 2.85 2-Ethvl-l-hexanol 1 .00 — 2-Methyl-l-propanol 1 . 02 — 1-Butanol 1 . 03 — Methyl oleate 1.25 — Acetone 1.26 2.95 Olive oil 1.29 — Isopropyl ether 1 29 Carbon tetrachloride 1 . 32 2 . 54 Benzene 1 . 33 2 . 45 Ethyl ether 1.33 2.98 Cyclohexane 1 . 36 — 1-n-Dioxane 1 .46 — G44 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 — 1 0} ' ^■■■■, o sz ■^rz ■:•*'» rrr s.",i: -^■^^ -,..,." IS r-:^ ■ —" " 1 1 1 1 1 1 liorophyll 0 in methanol carbon tetra benzene 1,4-dioxane "~ =r= '=^ i^~- ^ =?=■ ^ L._ N.~~ ;-i-N V, 0 V ^\t' 1 o i : 1 ^ J - 1 1 1 - 1 4 1 z 'r- — r^ -q ^'_ :zti ..«— ' ■■m^ ^' c- -- — - -" — '"- ' — • -Ui" V -i^ '-.. ^ \.. ^. ) ■v^ ■X. 6 UJ > '*'D S^I*M :r- — ---• 1^/ hlorophyll a in methanol isobutanol octanol (2-ethy|-l-hexano '-'.'-'. <^ _^ ^ -- ^^^ -liMW - — ~ ss ^^ rt.",~ s». ' I f ) k \ \ . 3 y T o i j ■ i i — J < ^, w^rti STTT? ^ rrtr; rrrr: !!2^ C .' ' > ■<, \ •^. ) > < s SOLVENT EFFECT G45 V . - c c c JC - a E c ■c 3 _ X c i c o o D c c o s^ In- c- i o 1 4 — r- — ■<■'; i • __ s ri*-; "^ rrrr; rr- -- ir.".' ':E: .::::■ .^^ c "-- • -— — - — — -■; :'Z' ".*.— — .-.; .T^. [^ ~:= ^ STS •^_ -1"; ^.- ^ ■^.^ '--■"C \ K 0) l»'U >. J3 O. o ^1 o JA o ■ ,— N -^ rt S ^— ' o > ■• o '-^ CO 1943 rent ^^. ^-^ o a ■n •-" S] J3 o ^^ o -O -r r^ :: >5 3 ^ a, .22 o o t- o u> ■^ "5 u> ffi Q i? ^-^ o -2 3 CM c3 10 „ 5^ ^ a o 00 a. m E solven rent ty 1- ^ _J If) rent diffe ">. ^ Q..2 o O w, 10 fc, e * o „ " J t^ CL o o o fH * ^ o C .J3 .2 O o ^ a> ft~! ■* O *- _ -* o o C ^ o o ■■S -5 o O c ^ ^' -^ UJ ir TT I 1 A Dry benzene Benzylamine Benzyl alcohol Benzene activated with benzylamine or benzyl alcohol 450 550 WAVE LENGTH, m/i 650 Fig. 21.26A. Absorption spectra of chlorophyll a in pure hydrocarbon and hydro- carbon containing an alcohol or amine (after Livingston et al. 1949). on page 607 that these differences may perhaps be indicative of tautomeric eciuilibria; but this is merely a conjecture. It w^ould be interesting to evaluate the total areas under the different curves to find whether the transition probabilities are changed by the sol- vent, or whether the latter merely affects the shapes of the bands, without changing their total areas. An as yet httle investigated subject is the absorption spectrum of dye- stuffs in general, and of chlorophyll in particular, in mixed solvents. Ob- servations of this type could give information about the occurrence, and SOLVENT EFFECT 647 energy, of complex formation of chlorophyll with different organic mole- cules. Livingston, Watson and McArdle (1949), in a study devoted pri- marily to the strong effect of small admixtures of polar solvents on the fluorescence of chlorophyll solutions in hydrocarbons, noted that these ad- mixtures also changed the absorption spectrum. As an example, figure 21.26 B shows the effect of traces of water on the absorption spectrum of chlorophyll b in benzene. In the dry solution, both main peaks are lower and a shoulder appears at about 670 m^t on the long-wave side of the red LlJ O Ll. Ll. LlI O O a. cr o (/> CD < LlJ > Wet J I I l_ _I I I I I I 450 650 550 WAVE LENGTH, m/x Fig. 21.26B. Absorption spectra of chlorophyll 6 in dry and wet benzene. peak. Figure 21.26 A shows the effects of two other polar solvents, benzyl- amine and benzyl alcohol, on the absorption spectrum of chlorophyll a in dry benzene. In this case, the main absorption peaks are higher in dry nonpolar solvent, and no "shoulder" appears on the long-wave end of the spectrum. A remarkable fact shown by this figure is that the spectrum of the activated solution (the term "activated" refers to fluorescence, which is absent in pure benzene), is the same whether activation is due to amine or to alcohol, although the absorption spectra of chlorophyll a in pure benzylamine and in pure benzyl alcohol are quite different. This could mean that polar molecules associate preferentially with a certain tautomeric form of chlorophyll, and in this way stabilize it; the presence of a small 648 ABSORPTION SPECTRA OF PIGMENTS IiY VITRO CHAP. 21 number of such molecules thus converts the spectrum of "normal" chloro- phyll into one of "tautomerized" chlorophyll. In pure polar solvent, on the other hand, polar molecules surround the chlorophyll molecule from all sides and thus cause a diffei-ent and more radical change in its absorption spectrum. The absorption spectrum of the "activated" solution is affected by in- creasing temperature, indicating a shift toward dissociation of the equilib- rium (M+~ = polar molecule, tChl = tautomeric chlorophyll). The interpre- tation of these interesting results will be discussed in chapter 23 (page 769) after presentation of the corresponding fluorescence data. Evstigneev, (xavrilo^'a and Krasnovsky (1949") noted that polar mole- cules have no effect on absorption spectrum and fluorescence of magnesium- free compounds (pheophytin and phthalocyanine) and therefore ascribed this effect to the binding of these molecules bj^ the residual valencies of magnesium. A few words can be added here on the effect of dissolved gases on the absorption spectrum of chlorophyll solutions. Padoa and Vita (1932) described changes in the ab- sorption spectra of chlorophylls a and b (in benzene solutions) in contact with nitrogen, oxygen, carbon monoxide and carbon dioxide. A strong effect was observed in the case of carbon monoxide — a result taken as an indication of the existence of a chlorophyll- carbon monoxide complex, similar to carboxyhemoglobin. However, the spectra re- produced in the paper of Padoa and Vita are so different from the true spectrum of chlorophyll, that they must have been obtained with some decomposition products rather than with the intact pigment. Katz and Wassink (1939) found practically iden- tical extinction curves for colloidal aqueous bacteriochlorophyll extracts in atmospheres of oxygen, hydrogen sulfide, nitrogen, hydrogen and air. Evstigneev, Gavrilova and Krasnovsky (1949^) asserted that the pres- ence of oxygen does have a certain effect on the spectrum of chlorophyll (a + b, or pure h) in toluene. Upon evacuation, the absorption coefficient decreased in both maxima, which were shifted slightly toward the red. In chlorophyll b, a new maximum of absorption appeared, when air was re- moved, at 670 mju. These changes were reversible ; but irreversible changes were noted in the ultraviolet part of the spectrum. Similar changes were observed in carbon tetrachloride and heptane, but not in pyridine, ethanol, acetone or benzene. Addition of one drop of alcohol, pyridine or acetone to 10 cc. toluene destroyed the effect of evacuation. Later (1949^) the same investigators found that the effects they had ascribed to the admission of air were actually caused by the admission of water vapor. These results obviously bear a relation to Livingston's conclusions that chlorophyll is present, in nonpolar solvents, in a state different from that to I COLLOIDAL AND ADSORBED CHLOROPHYLL 049 which it is converted by the presence of even traces of an alcohol or water. Evstigneev et al. (19490 suggested that in the absence of polar molecules chlorophyll is dimerized, and that the dimer is dissociated by oxygen molecules, dimerization being due to unsaturated magnesium valencies, which can also be saturated by oxygen. Their subsequent results (19492) made this interpretation of ox.ygen action unnecessary. In solvents of intermediate type, or in mixed (or simply not extremely purified) solvents, both types of interaction may occur simultaneously. Changes in the absorption spectra of chlorophyll can be caused, accord- ing to the observations of Livingston and co-workers (1948, 1949), also by small ciuantities of admixtures other than polar solvents. Examples are iodine, bromine, ferric and eerie salts (Rabinowitch and Weiss, 1937), and, in the case of chlorophyll h (but not chlorophyll a!) , phenylhydrazine. Some effects of this type might be similar to those of polar solvents, i.e., they may be caused by reversible formation of molecular complexes. Mostly, how- ever, they are due to irreversible chemical changes such as allomerization (or, more generally, oxidation), or reduction (which is likely in the case of chlorophyll h and phenylhydrazine); these phenomena do not belong under the heading of "effects of the medium on the absorption spectrum of chlorophyll," but rather under that of "chemical reactions of chlorophyll, revealed by spectroscopic measurements" {cj. pages 450-467 in Vol. I, and chapter 37 in this volume). 2. Absorption Spectrum of Colloidal and Adsorbed Chlorophyll Colloidal aqueous solutions of chlorophyll are obtained by mixing a molecular solution of the pigment (in alcohol or acetone) with water. The spectrum of the resulting solution depends on the conditions of mixing; this is the reason that earlier investigators could not agree on the position of the band maximum of colloidal chlorophyll. Herlitzka (1912), Will- statter and Stoll (1918) and Baas-Becking and Koning (1934) reported that this maximum coincides with that of chlorophyll in leaves, i. e., lies close to 680 m^- Ivanovski (1907, 1913) and Hubert (1935), on the other hand, found the maximum at 668 m/x, i. e., in the same region as in many true solutions. Hubert noticed, however, that the position of the maxi- mum was affected by the degree of dispersion of the colloidal sj^stem : Ad- dition of magnesium chloride, which caused a growth of the particles (and finally led to flocculation) shifted it by as much as 8 mn — from 668 to 676 m/z. Later, Wakkie (1935) and K. P. Meyer (1939) found that the es- sential factor is not the .size of the colloidal ])articles, but their internal density, i. e., the concentration of chlorophyll molecides in them. 650 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 A colloidal solution prepared by Meyer by rapid addition of 3 volumes of water to 1 volume of a chlorophyll solution in ethanol was clear, trans- parent and nonfluorescent. Its particles were 0.5 to 3 /i in diameter. The band maximum was at 670 m/x, and the shape of the extinction curve was similar to that of the original solution in ethanol. On the other hand, a colloidal solution prepared by adding quickly 0.6 volume of water to 1 volume of ethanol and then diluting by 6.4 volumes of water, was turbid and opalescent. Its particles had a diameter of 1-3 n, i. e., were not sub- stantially larger than those of the first, transparent colloid, but they con- tained more pigment. The absorption maximum of this colloid was sit- uated further toward the red, at 673 m^, and the whole shape of the extinc- tion curve was more like that of the leaves, as shown by figure 21.27 (Meyer described the spectrum of this colloidal preparation as "identical" with that of the leaves, but figure 21.27 does not justify this statement). By counting the particles of the colloid, Meyer found that the concentra- 48 400 450 650 700 Fig. 21.27. 500 550 600 WAVE LENGTH, m/i Transmission curves of leaves (1 and 2) and of colloidal chlorophyll solutions (3) (after K. P. Meyer, 1939). tion of chlorophyll in the particles of the turbid solution was of the order of 0.13 mole/1., i. e., similar to that in the chlorophyll grana in the leaves (cf. Vol. I, chapter 15, page 411). Even these ''densely packed" colloid par- ticles are still far from "sohd," but contain up to 90% solvent. Because of the intensity of the absorption bands, the extinction curves of dyestuffs usually are measured with concentrations of the order of 10"^ to 10~'* mole/I. No deviations from Beer's law (i. e., no changes in the ex- tinction curve with concentration) were observed in chlorophyll solutions in this range of concentrations. With very thin glass cells (~0.1 mm. deep), dyestuff solutions containing 10 ~^ mole/1, can be investigated, but even this is a hundred times more dilute than the 0.1 mole/1, reached in Meyer's colloidal particles and also present in the chlorophyll grana in COLLOIDAL AND ADSORBED CHLOROPHYLL 651 leaves. Wakkie (1935) emulsified a chlorophyll solution in ether in a satu- rated water-ether mixture, bubbled air through it and watched the changes of the absorption spectrum as the ether evaporated and the chlorophyll in the drops grew more and more concentrated. At a certain concentration (not estimated in the paper) the absorption band began to shift to the red, from 666 to 676 m^. This in(Hcates that a shift similar to that caused by the accumulation of chlorophyll in colloidal particles can be produced also by an increase of its concentration in true molecular solution. When the ether was completely evaporated, the remaining dry chlorophyll had an absorption maximum at 679 m^u. This agrees approximately with the measurements of Hubert (1935), who gave 680.5 m/x for the absorption maximum of solid chlorophyll (thin film of dried chlorophyll on glass). The possibility of energy exchange between excited and normal chloro- phyll molecules and the spectroscopic effects of this exchange — which must increase with increasing concentration of the pigment — will be discussed in chapter 32, in connection with the concept of the "photosynthetic unit" and similar hypotheses. A similarity between the absorption spectra of solid chlorophyll in suspension and of chlorophyll in the living cell was first claimed by Ivanovski (1907, 1913). It thus appears that a shift of the red chlorophyll band toward the longer wave lengths (approaching its position in the leaf spectrum) can be achieved not only by interaction with a solvent of high polarity or polariz- ability, but also by interaction with other chlorophyll molecules. This of- fers several alternatives for the interpretation of the state of chlorophyll in vivo. In the case of bacienochlorophyll, Katz and Wassink (1939) noted that the absorption band of evaporated pigment was shifted by not more than 2.5 m^ from its position in solution, while in live bacteria (and in colloidal extracts from bacteria) the same band is shifted toward longer waves by as much as 80-100 mn. In this case, the position of the band in the spectrum of the living cells definitely indicates interaction with other cell components and not merely close mutual proximity of the pigment molecules. A "red shift" of the absorption bands of chlorophyll probably can be ob- tained also by adsoi-plion on appropriate carriers: According to Seybold and Egle (1940), the red absorption band of chlorophyll adsorbates on starch is situated at 662 m^i, i. e., in the same region as in solution. Figure 21.27A, taken from Seybold and Weissweiler (1942), shows the absorption peaks of chlorophylls a and b, adsorbed on sugar, in the following positions: Chlorophyll a, 670 and 450 m/x, and chlorophyll b, 662 and 488 m^— values that correspond to shifts by 10 and 20 m/x for the a-component and 19 and 35 mn for the 6-component (compared to band positions in ether). The 652 ABSOnrTION RPKCTRA OF TIOMENTS /,V VITRO CHAP. 21 100 90 80 70 2 O 1- o UJ o 10 Sugar / / /a I \ I I \ I I \i I t I i\ I \t 1 \ I \ I Ol 1 1 I 1 I I 1 I I I I r I 400 500 600 WAVE LENGTH, m/x 700 Fig. 21.27A. Tran.smission and reflection curves of chlorophylls a and h adsorbed on sugar (after Seybold and Weissweiler 1942). 1.5 (A) Spinach leaf extract in water 0.0 1 1 ^ (B) Spinocti leaf extract in 2% digitonin, diluted 1: 10 with water 400 500 600 700 WAVE LENGTH, m^x 400 500 600 700 WAVE LENGTH, m,i Fig. 21.28. Absorption curves of spinach leaf extracts (after Smith 1941). COLLOIDAL AND ADSORBED CHLOROPHYLL 653 figures for the blue-violet bands are so high as to call for a recheck (see chapter 22, page 705, for the position of the blue-violet bands in the living cells). Eisler and Portheim (1922) and Noack (1927) stated that the ab- sorption spectra of chlorophyll adsorbates on proteins are "similar to those of the living cells," but gave no figures. -(A) Distilled water A ^1 ENERGY, e.v. ..._2_ 1.9 1.8 1 r-^-. r^^- 1.7 640 680 72C WAVE LENGTH, m>i 640 680 720 WAVE LENGTH, m^ 2.U (of) I. Cells in distilled water 2. Colloidal extract in fresh egg albumen ENERGY, e.v. 1.9 1.8 T T "T 17 (e) 1. Cells in distilled water 2. Alcoholic extract ENERGY, e.v. 19 "1.8 1.7 T 640 680 720 WAVE LENGTH, m^ 640 680 720 WAVE LENGTH, m>t ■(c) Knop's culture nfiedium 640 »- 680 720 WAVE LENGTH, m^A Fig. 21 .29. Comparative absorption .spectra of cell suspensions and pigment extracts of Chlorella in different media (after Katz and Wassink 1939). Curve 1, cells; curve 2, ex- tracts. See page 65G. Absorption curves ai'O available for "natural" chlorophyll-protein colloids extracted from leaves, algae and bacteria. Smith (1938, 1941) gave an absorption curve of the crude extract from spinach leaves, which contains broken chloroplasts or grana, and another curve for the same ex- tract clarified by digitonin (r/. fig. 21.28, curves A and B). The maximum of curve A is at 678 m;u, that of curve B at 675 m/x. 654 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 The relatively strong absorption of the crude extract in the violet may be due to the presence of yellow pigments ; its stronger absorption in the far red (>700 m^u) was attributed by Smith to scattering. However, in- creased absorption in the far red has also been observed by Noddack and 2.0 O (a) Phosphate buffer,^ pH6.6, 1/15 M n — -| — r — I — I — r 820 860 900 940 WAVE LENGTH, m/i 1.5 o o (c) Culture medium No. 231 1.6 740 1.4 — 1 — r" — I — I I 860 900 940 {b) Distilled water 740 1 r 780 820 T 1 — r — i — I — r 860 900 940 WAVE LENGTH, m/i (.d) I. Cells in distilled water 2. Alcoholic extract 740 780 820 wavE LENGTH, m/i — 1 — I — r 900 940 780 820 WAVE LENGTH, vnfi Fig. 2 1.30 A. Comparative absorption spectra of cell suspensions and pigment ex- tracts of strain D purple sulfur bacteria in different media (after Katz and Wassink 1939). Curve 1, cells; curve 2, extract. See page 656. Eichhoff in Chlorella suspensions (see fig. 22.21), although these investiga- tors used an integrating method, which w^as supposed to give true absorp- tion values, free of scattering effects. Rabideau, French and Holt (1946) gave absorption curves (obtained with an Ulbricht sphere) and transmission curves (obtained with a Beck- man spectrophotometer) for chloroplast dispersions prepared by means COLLOIDAL AND ADSORBED CHLOROPHYLL 655 of supersonic waves. These curves, which can be found in figure 22.15. indicate that enhanced extinction in the far red is characteristic of trans- mission much more than of true absorption— thus supporting Smith's ex- planation and contradicting the results of Noddack and Eichhoff. 2.0 (o) Phosphate buffer, AJH 6.6, \/\5M ENERGY, e.v. 1.5 1.4 1.3. 740 780 820 WAVE LENGTH, m/i T 1 — r — I — I — r-^ 860 900 940 {b) Distilled water 1.6 I ENERGY, e.v. 1.5 1.4 1.3 740 780 820 860 900 940 WAVE LENGTH, m^ 2.0 (c) Culture medium No. 231 1 — I 1 — I — I — r— 860 900 940 {.d) \. Cells in distilled water 2. Alcoholic extract 740 780 820 WAVE LENGTH, 'm/i — 1 r — I — r 860 900 940 820 WAVE LENGTH, m/i Fig. 21. SOB. Comparative absorption spectra of cell suspensions and pigment ex- tracts of Rhodospirillum rubrum in different media (after Katz and Wassink 1939). Curve 1, cells; curve 2, extract. Concerning the difference between the cell (and the colloidal extract spectra), in A and B, see page 703. According to Smith, the molar extinction coefficient of chlorophyll in the maximum of the red band is approximately the same in the aqueous extract containing digitonin and in ether or acetone solution. In other words, the shift in the position of this band from GGO to G75 m^ occurs without a change in its intensity. 050 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 Absorption curves of colloidal chlorophyll-protein extract were given also by Katz and Wassink (1939). Figure 21.29 shows the red band in the spectra of Chlorella suspensions, and in colloidal extracts from the same cells (a) in phosphate buffer pH 6.0, (6) in distilled water, (c) in Knop's culture medium and (d) in fresh egg albumen. The position of the maxima are but slightly different in all these ciu'ves (approximately 080 m/x); the shapes of the curves are, however, affected by the nature of the medium, particularly in extracts a and b. The solution in egg albumen has an ex- tinction curve practically identical with that of the living cells. Fig. 21.29e shows, as a contrast, the strong shift occurring upon extraction of the pig- ment with alcohol. Similar curves were given by Katz and Wassink (1939) and French (1940) for colloidal extracts from purple bacteria (c/. figs. 21.30). A transmission curve of a water extract of the blue-green alga Chroococ- cus can be found in figure 22.48B. In these extracts, the phycocyanin- protein forms a true colloidal solution, while the other pigments probably are in the same state of dispersion as in extracts from green algae and leaves. C. Absorption Spectra of the Carotenoids* 1 . Experimental Results The extinction curves of carotenoids in organic solvents have been in- vestigated by numerous authors, among whom we may mention Willstatter and Stoll (1913), Pummerer and Rebmann (1928), McNicholas (1931), Smakula (1934), Gillam (1935), Sprecher von Bernegg, Heierle and Almasy (1935), Miller (1935, 1937), Strain and co-workers (1938, 1942, 1943, 1944), French (1941), Beadle, Zscheile and co-workers (1942, 1944, 1945) and Zechmeister and co-workers (cf. review by Zechmeister 1944). Recently, a number of absorption curves were determined by Karrer and co-workers, and were reproduced in a monograph by Karrer and Jucker (1948). The absorption spectra of carotene and "leaf xanthophyll" in the in- frared were observed by Stair and Coblentz (1933). They show a series of absorption bands characteristic of long unsaturated carbon chains; many of them coincide closely with the absorption bands of phytol (cf. Table 21.IIA). The spectra of all carotenoids in the visible are characterized by two or three intense bands near the violet end of the spectrum. Depending on how far these bands extend into the blue and green, the color of the pig- ments may be yellow, orange or even red. The position of the absorption bands depends, often even more strongly than in the case of chlorophyll, * Bibliography, page 670. ABSORPTION SPECTRA OF THE CAROTENOIDS 657 on the state of the pigment and the surrounding medium. Table 21. IX shows that the direction of the band shift in different solvents is the same as for chlorophyll — i. e., they are displaced toward the longer waves with increasing polarity and polarizability of the solvent. From ether to carbon disulfide, the "red shift" amounts to 43 m/x for carotene /3 and 35 m/x for luteol — as compared to only 10 m/z for chlorophyll. However, the effect of transition from a nonpolar to a polar solvent of approximately the same polarizability — e. g.,hom ether to ethanol — seems to be smaller for the carot- enoids than for the more polar chlorophylls (c/. Table 22. IX). A still stronger displacement sometimes occurs in aqueous colloidal solutions. According to Karrer and Strauss (1938), the maximum of band I of carotene is shifted, from 480 m^ in hexane and ether, to 510 m/x, or even 535 m^, in hydrosols. This, too, is a much wider shift than was ob- served in chlorophyll colloids. (Because of this shift, some carotene sols are red, while their molecular solutions are j^ellow.) In addition to a shift of the band maxima, changes of medium may also cause a broadening of the carotenoid bands. This effect appears to be particularly strong in the case of some algal carotenoids. For example, the curves given for fucoxanthol spectrum by Strain, Manning and Hardin (1944) show a considerable flattening of the absorption peaks and extension of the absorption band toward the longer waves in ethanol as compared to petroleum ether. The spectrum of peridinol, a pigment of the dinoflagel- lates, shows a similarity strong solvent effect. The spreading of the absorption bands of the carotenoids into the green, rather than a shift of their peaks, probably explains the color of broA\Ti algae and diatoms. The striking difference between their color and that of green plants appears inexplicable if one considers only the solution spectra of fucoxanthol {cf. fig. 21.35A) or peridinol, since these are almost identical with the spectra of the carotenols of green plants (e. g., luteol and zeaxan- thol). Recently, Karrer and co-workers (1943, 1948) published an absorption curve of fucoxanthol in hexane (fig. 21.36) which shows a comparatively slow decline of absorption toward the longer waves. The absorption re- mains marked up to 550 mju, while that of most other carotenoids drops to zero at 500 m^u. The reason for the difference between this curve (which togethei- with that of chlorophyll could explain the brown color of fucoxan- thinol-bearing algae) and that given l)y Wald for the same solvent, remains unexplained. The curves given by Wassinkand Kersten (1946) for the absorption spec- trum of the "yellow" and the "orange" fraction of the carotenoids from the diatom Nitzschia dissipata also show an extension of the absorption in the second fraction (in methanol) to about 550 m/z, the peaks being situated 658 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 X (M H m ^ o 3 ^ -G S 3 rt C3 O 05 O I GO O «D lO »0 lO lO -* '^ CO t^ 00 I (N 00 ■* Tjl Tfl Tfi a a X 00 t- _ . . Tf* ^ TjH lO -3 ni I — I 3 CO ci y. o c c3 o o "o x; >> -u ^ G O -i-i ^ c3 s O H O O o3 G is ■sa oi G a G G c3 G G cd c3 o3 a a a t+_( ^4-1 t+- 3 3 3 ej 03 o3 Li! ►> KA H-H l-H hH -73 73 -3 G r^ G 03 c3 o3 G C^ .s G '53 Tt< iD 'S -u 05 S-i ■^-i CO 1-H 73 tn CO ;..! V ^ ^ i~ CO (D O -— ' "o £m 3 v^^ 3 N • £ ^ w N3 m W CO M 1 o ^ •^ CO '^ o ^H o (N (N (N (M IM CO rfH ^ '^ •* ■* TfH '^ ^ Tf* Tf< lO o t- ~H o o CO iC l:^ CO •* ■^ •^ lO IC -+ >c »o »o t^ t^ ^ "* '^ ^ Tfl Tt< tp Tjl Tt< Tt^ "3 O C (M CO IC lO o t^ t- t^ 00 r- 00 00 00 o ^ Tt* TjH ■^ '^ ^ TfH Tf< 4 Ifide 5 73 U ti c r^ r< 3 3 CD O ^ G G M CQ 03 O S^ -S o oj o "o "o O U-t •^^ G c "Fi"?! 03 % G a P o 2 g O ^ o ^ jG -G -tJ -tJ -4^ +3 r^ ~ -^^ 2 -G 03 03 w w a w w 1—1 O o o O G o3 > 0 o a> 3 1-1 I ABSORPTION SPECTRA OF THE CAROTEN c O 05 e S3 C G o3 a 1^ O a Xi a & o > o3 03 2^ o n ^^ bo a o3 oj S O P-l to tS3 O c «3 CO a; > o3 o 03 o o 03 03 03 SI CO a o3 ■ . -C ^— ' bO ■^ .2 -^ C ~ 2 « 'S , ^ § fe§ 3 -5 "3 . W -3 ^ .S ^ c3 »-. Lh O > O -D > !> > 02 CD o ;-! 03 3 3 CO o O --H t^ CO — ' o lO »-0 lO 10 lO 10 rfi T}i rfH ^ Tti ^ rt< r^ (N 1 00 0 «) r^ 0 t^ t^ ^ '^ Tf^ Si 03 -3 +3 s 03 ^ n) 3 — ' p 0 03 Ethanol Ethanol Pctroleu Methano 03 o3 a p o3 o o 3 659 c3 3 : o3 bC 3 's - 3 • c3 3 CO 1^ o 00 CO ■^ "^ "^ 10 ^ ^ Tj< ,^ —I CO (M •* CO 'ti ^ ■^ ^ 00 T-l lO -rJH r-H JO T-l t^ 00 t^ o i^ o t^ M< Tt< rt< T}H rt< -* -^ ooooooooo 333GGC333 c3C3c3o3c3o3s3o3c3 ^-3-3j=:-3^j3j::^ to bC 03 m K td ai m 03 03 O; 03 03 -*-^ -^^ -*^ -»-3 -*-3 o3 o3 c3 c3 o3 n-H ^y '-^ w cj ra to 23 03 o3 to s O 0; 03 03 03 03 b£ bC M bO bO 03 rt o3 o3 c3 52 03 53 OS 53 00000 pqpGqGqQa < pq 000 . -5 •S -^ — e3 *J X a o 03 3 3 ;3 o o3 ;-! .3 03 C Q o J3 O 3 J3 o3 *^ X 3 O 03 3 3 O 03 03 -j-i .rt S 03 .3 03 03 O 03 "3 CO - - ^ C5 - - - 03 to O (13 03 !© O -ti 00 •^ 05 Tfl I> ^ rfi -fH 00 CO iM o CO r^ o -i< t^ T^ 10 -t 10 -* ■* 10 CO CO »o t^ IM (N (N t^ 0 -+i r^ 0 iO lO »o •o ^ 10 03 03 03 '"O "O T) «3 U2 53 3 3 3 C/J 7) ro -o TS T-l 0 1— ' 0 r^ 0 3 3 0 a 0 3 0 03 _o C3 Xi c3 -n .3 1-. -3 ^ t-, -fc^ ^j 03 -♦J 03 3 o a o 3 u .2 s o 03 03 03 3 Ph .2 03 O o3 J2 03 'o. Si 3 Ah O CO 03 "o -s > o o 43 P4 o o O 43 T3 O o > 03 GfiO ABSORPTION SrECTRA OF PIGMENTS IN VITRO CHAP. 21 300 - 280 - 240 A 200 / ^A jPl80 /^ \ 120 / \ 80 / \ 40 v 1 1 1 1 1 1 1 1 1 1 1 1 1^ a - Carotene /3 - Carotene Lycopene 380 400 420 440 460 480 500 520 220 WAVE LENGTH, m/i 280 340 400 460 WAVE LENGTH, m/x 520 Fig. 21.31. Absorption spectrum of j8- carotene in hexane (after Zscheile, White, Beadle and Roach 1942). Fig. 21.32. Absorption spectra of a- carotene, /3-carotene, and lycopene in 80% ethanol + 20% diethyl ether (after Miller 1937). 280 4i _ r\ -^ . r-! 240 - // \ \ \ \ \ \ \ \ \ \ 0 t\ 200 - It It p \ \ \ \ \ \ \ \ ISO - / 1 / / / / / / / / \ \ \ \ \ 120 - / / / / // // y / y / - I I \ 1 80 / / V \ \ \ \ 40 -/ 1 1 1 1 \ 1 I 1 \ \ \ \ 1 r~- 2.5 2.0 o Q 1.5- r * 0 A. f * \ II \ y /^" XN \ ~ - \ 1 1 1 2.4 -20 380 400 420 440 460 480 500 520 WAVE LENGTH, m/i 400 450 WAVE LENGTH, m/i 1.5 500 Fig. 21.33. Specific absorption spectra Fig. 21.34. Specific absorption coef- ofluteol (solid line ) and zeaxanthol (broken ficients of luteol in ethanol (after Strain line) in ethanol (after Zscheile, White, 1938). I, dried luteol. II, luteol with Beadle and Roach 1942): c is in grams 2.65% petroleum ether. Dots, luteol in per liter; d is in centimeters. ethanol as reported by Smakula (1934). Ill, pure luteol. IV, same after heating 3 hrs. at 77°. Values of log a for I and II at left, for III and IV at right. ABSORPTION SPECTRA OF THE CAROTENOIDS 661 400 420 440 460 480 500 520 540 WAVE LENGTH, n\^ Fig. 2 1.35 A. Alisorption spectrum of fucoxanthol in hexane (after Wald 1942). (For a different curve for the same solution see fig. 21.36.) c o o 400 450 WAVE LENGTH, m,i 500 Fig. 21.35B. Absorption spectra of diatoxanthol, diadinoxanthol, neodia- dinoxanthol, zeaxanthol and luteol (after Strain, Manning and Hardin 1944). 3.2- o S 2.7 a o o 2.2 L7 • A /\ - Violoxanthol / \ . in alcohol / \ ■ / '-^ \ • / ''^' ~^^ \ - / '' H • / / r " / /' \^' " 1 1 \ \ . / / Fucoxanthol \ * - J / / '" '^sxone \ \ I ^'^ \ ■y ' \ \ ^ * * \ /v \ / / 1 ^ ■ \ \ \ \ \ \ I ^' \ \ ■ \\ ''^ \ ^^ " _] — 1 — 1 — 1 — 1 — L-. 250 500 550 300 350 400 450 WAVE LENGTH, m/x Fig. 21.36. Specific absorption coefficients of fucoxanthol in hexane and violaxanthol in ethanol (after Karrer and Wtirgler 1943). Concentration in g./lOO cc. 662 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 at 426, 450 and 476 mn (Table 21. IX). The curves given by them for Hve diatoms, colloidal extracts and pigment solutions in organic solvents indicate a considerably increased absorption, in vivo, in the region 500- 560 mju. Table 21. IX shows that the absorption bands of many bacterial carote- noids are situated further toward the red than those of green plants and algae — a relation that reminds one of that between chlorophyll and bac- teriochlorophyll. Figures 21.31 and 32 show the extinction curves of three carotene iso- mers and figures 21.33 and 21.34 those of luteol — the most common carote- nol of green plants (c/. Vol. I, page 415). Figures 2 1.35 A and B repre- sent the spectra of several carotenols of broAvn algae, diatoms and dino- flagellates. 300 260 220 180 140 100 60 4510 A -o4780 4680 >4920 J I L J I L 0 4 8 12 16 20 24 28 SPECTRAL REGION ISOLATED, m^ Fig. 21.37. Effect of width of spectral region isolated on the specific absorption coefficient of /3-carotene at selected wave lengths (after Zscheile, White, Beadle and Roach 1942). The changes in a observed by Miller with single monochromator (•) were not found in the work with a double monochromator (O). The absorption bands of the carotenoids are much broader than those of chloro- phyll; they are therefore less sensitive to changes in the width of the slit, as illustrated by figure 2i.37. 2. Theoretical Considerations Carotenoids are 'polyene dyestuffs, i. e., their color is due to the chromo- phoric properties of a straight chain of conjugated double bonds. Syn- thetic polyenes, and the cyanine dyestuffs used in photographic sensitiza- tion, are other examples of the same type. With increasing length of the ABSORPTION SPECTRA OF THE CAROTENOIDS 663 conjugated chain, the absorption bands are shifted regularly toward longer wave length, and their intensity becomes greater. These simple relations between color and molecular structure make polyene dyes particularly suitable objects for theoretical studies. This problem was treated by Paul- ing (1939), who used the method of atomic orbitals and the concept of resonance, and by Mulliken (1939, 1941), who used the method of molecular orbitals. According to Paul- ing, the fundamental resonance possibilities of polyene molecules are provided by the shifting of the double bonds, which results in the transfer of an electron from one end of the molecule to the other. If we consider, e. g., a straight conjugated chain with an even number of carbon atoms {A), the shift of all double bonds to the left will produce structure B and a shift to the right, structure C: {A) CH3CH=CHCH CHCH=CHCH3 (B) +CH3=CHCH=CH CH=CHCHCH3 (C) CHaCHCH^CH CH=CHCH=CH3 + Each of the states B and C has a large dipole moment; but, since the two moments have opposite directions, and B and C have equal probabilities, the molecule will show no dipole moment at all, both in the ground state of the molecule and in the lowest ex- cited states formed by resonance between the same three structures. However, the transition from the normal state to an excited state of this type lias a "transition mo- ment" whose order of magnitude is that of the dipole moment of the individual structures B and C This is a very large moment, and it increases with length of the chain. In wave mechanics, the probability of a spectroscopic transition between two states {i. e., the intensity of the corresponding absorption hne or band) is determined by the magni- tude of the "transition moment." This explains why polyene molecules have strong absorption bands, and why their intensity increases with the greater chain length. The second approach to the same problem is that of the theory of "molecular orbi- tals." It considers the actual state of the molecule without decomposing it into imagi- nary resonating components. It tries to assign the electrons not to definite atoms or bonds, but to definite lA-functions (orbitals) of the molecule as a whole. In a long chain of conjugated double bonds, some of these orbitals include the nuclei of all atoms in the chain, and electrons assigned to them can be considered as moving freely through the whole chain (this being the counterpart to the "shifting of double bonds" in the reso- nance theory). A conjugated double bond chain has, in this theory, a certain similarity to a metallic wire. An investigation of a molecule by this theory consists in the determination of the qualitative characteristics of available orbitals and the evaluation of the relative ener- gies of the states obtained by different assignments of the electrons to the orbitals. Let us consider (Mulliken 1939) a straight chain of n carbon atoms (n = even num- ber) and n/2 double bonds (the presence of symmetrical end groups on both ends of this conjugated chain — which is common in carotenoids — does not alter the problem). It contains n "unsaturation orbitals," sweeping over the whole conjugated chain, of which n/2 are "bonding" (i. e., electrons assigned to them stabilize the molecule), and n/2 "antibonding." Each orbital can, as usual, hold two electrons, so that the n available "unsaturation electrons" are just enough to fill the n/2 bonding orbitals, thus giving a singlet normal state. The transfer of any one of these electrons into any one of the n/2 antibonding orbitals leads to an excited state; there are therefore nV4 groups of ex- cited states. Each group consists (because of the interaction of orbitals with the elec- tron spin) of one triplet and one singlet state; however, because of the prohibition of 664 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 singlet-triplet transitions in light atoms, only the n^/4 singlet states are of importance for absorption. The longer the chain, the more numerous are the excited states. It can be shown that the "center of gravity" of these states in the energy diagram remains more or less unchanged, while the lowest excited states shift closer and closer to the ground state as the chain grows longer. This is shown schematically in figure 21.38. At the left we have the energy diagram of an A2-molecule, at the opposite end, that of an infinite chain of A nuclei; the long- wave absorption limits are represented by the ar- rows; they become shorter and shorter, i. e., the absorption shifts further and further to the red, with increasing chain length, until it extends into the infrared. MuUiken has shown that, if the chain is a straight as possible {i. e., if all carbon atoms in the chain are in trann positions), the transi- tion to the lowest excited state (arrow in fig. 21.38) is more piobable than all the other transi- tions together. This means that the intensity of the absorption band with the lowest fre- quency must increase steadily with increasing chain length. We have thus obtained theo- retical interpretations, both of the gradual shift of the absorption band to longer waves, and of the increase in its intensity with the growing length of the chain. Fig. 21.38. Shift of first absorp- tion band with increasing length of conjugated chain (thickness of ar- row indicates intensity). The two or three separate maxima observed in carotenoid spectra may mean as many distinct electronic transi- tions; but more probably they correspond to coexcitation of one or several vibrational quanta. The distance between maxima (ca. 1500 cm.~0 is of the order of magnitude of vibrational quanta in organic molecules. D. Absorption Spectra of the Phycobilins* The absorption spectra of the phj'cobilins have been observed in living algae, in aqueous colloidal extracts of chromoproteids and in organic solu- tions of chromophores. The results are somewhat confused because both phycocyanin and phycoerythrin apparently occiu* in several modifications of slightly different color. (These modifications might be due either to minor variations in the structure of the chromophores, or to the association of the same chromophore with different proteins.) The first extensive data on the absorption spectra of the phycochromo- proteids were given by Schiitt (1888). Among the more recent papers on this subject are those of Lemberg (1928, 1930), Svedberg and Lewis (1928), Svedberg and Katsurai (1929), Dhere and Fontaine (1931), Svedberg and Eriksson (1932), Roche (1933), Katz and Wassink (1939) and French and co-workers (1948,1951). Bibliography, page 67 1. ABSORPTION SPECTRA OF THE PHYCOBILINS 065 The only availal)le data on the absorption spectra of the isolated (^. e., protein-free) chromophores are those of Lemberg (1930). He gave the following wave lengths for the maxima of the absorption bands: Compound Medium Wave length 598 m/i 606 mju 498 niM Cyanobilin (from Porphyra tenera) Erythrobilin (from Porphyra tenera) HCl (cone.) Acid CHCI3 HCl (cone.) Combination of the pigments with protein shifts the bands toward the red; l)ut the amoimt of this shift is not adequately described by compari- son of the above-cjuoted figures with the positions of the absorption maxima of aqueous exti'acts from algae, because the data on free pigments refer to strongly acid solutions, while those on chromoproteids relate to neutral or only weakly acid solutions. The main absorption maximum of the phyco- cyanin-protein complex lies at 615 m^u in the pH range between 3.5 and 7, but is displaced, in concentrated hydrochloric acid, by as much as 41 ran toward the red (to 656 m/x). Comparing this strongly acid solution of the chromoproteid with an equally strongly acid solution of the chromo- phore, we find a "red shift" by as much as 58 m^u; comparison with a neu- tral chromoproteid solution would indicate a shift of only 17 m/x. Similar figures were given more recently by Wassink (1948) for cyano- bilin from blue-green Oscillatoria (Xmax. = 620 m^i for the chromoproteid, 610 mn for the solution of the cyanobilin in chloroform, and 600 mju for its solution in HCl) . Extinction curves of aqueous chromoproteid colloids were given by Svedberg and Lewis (1928), Svedberg and Katsurai (1929), Svedberg and Eriksson (1932), and French and co-workers (1948,1951). Figure 21.39 shows the extinction curves of the phycoerythrins from five different algae. Three maxima (566, 540 and 498 mix) are always present, but with variable relative intensities, pointing to the existence of three different forms of the pigment (perhaps the same chromophore linked to different proteins). Van Norman et al. (1948) found only two absorp- tion peaks (550 and 495 m/x) in aqueous extract from Iridaea. It also has several bands in the ultraviolet. Similar observations were made with phycocyanin. In phycocyanin from a Rhodophycea (e. g., Ceramium rubrum. and Porphyra tenera), Sved- berg and Katsurai (1929) found two bands in the visible, at about 615 and 550m)u and ultraviolet bands at 355, 271 and 240 m/x. In the phycocyanin from a Cyanophycea (e. g., Aphanizomenon flos aquae) they found only one visible band, at 615 mju, and ultraviolet bands at 368 and 277 mix. GG6 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 Ceramium 450 500 550 600 650 450 500 550 600 650 450 500 550 600 650 WAVE LENGTH, m/i WAVE LENGTH, m/i WAVE LENGTH, m/i Fig. 21.39. Specific absorption spectra of phycoerythrins from different algae (after Svedberg and Eriksson 1932). 80 60 » 40 20 - 615 fTi/iA \ 277 m,i / ^ 1 368 m/i J \ 1 1 200 300 600 700 400 500 WAVE LENGTH, m/x Fig. 21.40. Specific absorption spectrum of phycocyanin from Aphanizo- menonflos aquae (after Svedberg and Katsurai 1929). Figure 21.40 shows the extinction curve of the phycocyanin from Aphan- izomenon flos aquae, according to Svedberg and Katsurai (1929). Using Lemberg's estimate of 2% pigment in the chromoproteid (with a molecular weight of 636 for the chromophore, this corresponds to one mole pigment in 3.2 X 10^ g. of the complex) we can convert the specific extinction coefficients, given by Lemberg (1928, 1930) and Svedberg and Katsurai (1929) into molar extinction coefficients. The resulting values (cf. Table 21. X) are exceptionally high — up to 3 X 10^, as against only 4 X lO'* in the maximum of the red band of chlorophyll and 1.5 X 10^ in the maximum of the main absorption band of the carotenoids. This makes the correctness of Lemberg's analytical data somewhat doubtful. Lemberg ABSORPTION SPECTRA OF THE PHYCOBILINS 667 (1930) himself noted that the specific extinction of the phycobiHns is ten times stronger than that of hemoglobin — while his analysis indicated the presence of only one half mole pigment per Svedberg unit of protein in phycobilins, as against one mole pigment per unit of protein in hemoglobin. If Lemberg's analysis is in error, and the content of phycobilins in the chromoproteids is as high or even higher than that of hemin in hemoglobin, the molar extinction coefficients of the phycobilins, given in Table 21, X will have to be proportionally reduced. Table 21. X Estimated Molar Extinction Coefficients of Phycobilins (in the Band Maxima) Pigment State X, m^ °mol. Observer" Phycoery thrill Pigment in HCI Chromoproteid from 495 1.8 X 10* L Ceramium 565 2.6 X 105 S,K Chromoproteid from Ceramium 565 2.5 X 10* L Chromoproteid from Porphyra 565 2.5 X 10* L Phycocyanin Pigment in HCI Chromoproteid from 598 ca. 105 L Ceramium 615 1.3 X 105 S,K Chromoproteid from • Ceramium 615 2.0 X 10= L Chromoproteid from Porphyra 615 3.1 X 105 L Chromoproteid from Aphanizomenon 615 2.6 X 105 S,K « L = Lemberg (1928, 1930). S,K = Svedberg and Katsurai (1929). A comparison of the intensity of the phycocyanin band at 615 m/x with that of the chlorophyll band at 680 m/Li in the absorption spectrum of live Oscillatoria cells (cf. fig. 22.18) leads to similar doubts concerning Lemberg's analytical data. According to Lemberg (1928), the content of the chromoproteids in algae (determined with another species, Ceramium rubrum) is of the order of 0.5%, with only 2% chromophore in the complex. This corresponds to as little as 0.01 % phy cobilin in the dry matter of the algae — while the concentration of chlorophyll in red algae usually is of the order of 0.1% {cf. Table 15.11, Vol. I). The predominance of the phycocyanine band over the chlorophyll band in figure 22.18 therefore leads to the improbable conclusion that the molar extinc- tion coefficient of the phycobilins is at least ten, and perhaps one hundred, times higher than that of chlorophyll — unless we prefer to assume that Lemberg's analytical figures are too low. In addition to the fact that the estimate of the chromophore content in the complex (~2%) is probably too low, the content of the chromoproteid in the algae (~0.5%) may also have been underestimated {cf. chapter 15, page 418). 668 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 Bibliography to Chapter 21 Absorption Spectra of Pigments in Vitro A. Absorption Spectra of Chlorophyll and Its Derivatives 1913 Willstatter, R., and StoU, A., U titer suchungen ilber Chlorophyll. Springer Berlin, 1913. 1914 van Gulik, D., Ann. Physik, 46, 147. 1918 Willstatter, R., and StoU, A., Unter suchungen iiber die Assimilation der Kohlensdure. Springer, Berlin, 1918. 1922 Metzner, P., Ber. deut. botan. Ges., 40, 125. 1928 Lewkowitsch, E., Biocheni. J., 22, 777. 1929 Noack, K., and Kiessling, W., Z. physiol. Chem., 182, 13. 1930 Noack, K., and Kiessling, W., ibid., 193, 97. 1931 Noack, K., and Kiessling, W., Z. angew. Chem., 44, 93. Ghosh, J. C, and Sen-Gupta, S. B., /. Indian Chem. Sac, 8, 581. 1933 Winterstein, A., and Stein, G., Z. physiol. Chem., 220, 263. Stair, R., and Coblentz, W. W., /. Research Natl. Bur. Standards, 11, 703. Rudolph, H., Planta, 21, 104. 1934 Prins, J. A., Nature, 134, 457. Zscheile, F. P., /. Phys. Chem., 38, 95. Zscheile, F. P., Botan. Gaz., 95, 529. 1935 Zscheile, F. P., Cold Spring Harbor Symposia Quant. Biol., 3, 108. Stern, M., Kleine Mitt. Mitglied. Ver. Wasser-, Baden-, u. Lujthyg., 11, 47. Sprecher von Bernegg, A., Heierle, E., and Almasy, F., Biochem. Z., 283 45. Stern, A., and Wenderlein, H., Z. physik. Chem., A174, 81. 1936 Stern, A., and Wenderlein, H., ibid., A176, 81. StoU, A., Naturwissenschaften, 24, 53. Hagenbach, A., Auerbacher, F., and Wiedemann, E., Helv. Phys. Acta, 9, 3. Stern, A., Wenderlein, H., and Molvig, H., Z. physik. Chem., A177, 40. 1937 Stern, A., and Molvig, H., ibid., A178, 161. Seybold, A., Pla7ita, 26, 712. Vermeulen, D., Wassink, E. C., and Reman, G. H., Enzymologia, 4, 254. French, C. S., J. Gen. Physiol., 21, 71. Rabinowitch, E., and Weiss, J., Proc. Roy. Soc. London, A162, 251. 1938 Stern, M., Kleine Mitt. Mitglied. Ver. Wasser, Boden- u. Lufthyg., 14, 39. Tiegs, E., ibid., 14, 34. Mackinney, G., Plant Physiol., 13, 123. Stern, A., Z. physik. Chem., A182, 18G. 1939 Seybold, A., and Egle, K., Sitzber. Heidtlberg. Akad. Wiss. Math, naturw. Klasse, 1939, No. 1, 7. Katz, E., and Wassink, E. C., Enzymolgia, 7, 97. Stern, A., and Pruckner, F., ibid., A185, 140. Biermacher, 0., Thesis, Univ. Fiil)ourg (Switzeiiand). Egle, K., Sitzber. Heidelberg. Ak. Wiss. Math.-Nat. Klasse, 1939, No. 1, 19. Meyer, K. P., Hclv. Phys. Acta, 12, 349. EIELIOGRAPHY TO f'TT APTF.R 21 600 1940 Mackinney, G., J. Biol. Chem., 132, 91. French, C. S., /. Gen. Physiol, 23, 483. Fischer, H., and Stern, A., in Fischer and Orth, Die Chemie des Pyrrols, Vol. II. 2, Akad. Verlagsges., Leipzig, 1940. 1941 Mackinney, G., /. Biol. Chem., 140, 315. Zscheile, F. P., and Comar, C. L., Botan. Gnz., 102, 463. Comar, C. L., and Zsclicile, F. P., Plant Physiol, 16, 651. Albers, V. INI., Gibson Island A. A. A. S. Symposiion, on Photosynthesis (unpubhshed). Epstein, L. F., Karush, F., and Rabinowitch, E., /. Opt. Soc. Am., 31, 77. 1942 Strain, H. H., and Manning, W. M., /. Biol. Chem., 144, 625. Comar, C. L., and Zscheile, F. D., Plant Phjsiol, 17, 198. Haskin, H. H., /. Biol Chem., 144, 149. Zscheile, F. P., Comar, C. L., and Mackinney, G., Plant Physiol, 17, 666. Comar, C. L., Ind. Eng. Chem., Anal. Ed., 14, 877. Pruckner, F., Z. phydk. Chem., A190, 101. 1943 Comar, C. L., Benne, E. J., and Buteyn, E. K., ibid., 15, 524. Manning, W. M., and Strain, H. H., /. Biol Chem., 151, 1. Harris, D. G., and Zscheile, F. P., Botan. Gaz., 104, 515. Aronoff, S., and Calvin, M., J. Org. Chem., 8, 205. Calvin, M., Ball, R. H., and Aronoff, S., /. Am. Chem. Soc., 65, 2259. Lewis, G. N., and Bigeleisen, J., ibid., 65, 1144, 1944 Zscheile, F. P., Comar, C. L., and Ilariis, D. G., PlaJit Physiol, 19, 627, Rabinowitch, E., Rev. Modern Phys., 16, 226. 1946 Wassink, E. C, and Kersten, J. A. H., Enzymologia, 12, 3. Erdman, J. G., and Corwin, A. H., J. Am. Chem. Soc, 68, 1885. 1948 McBrady, J., and Livingston, R., J. Phys. & Colloid Chem., 52, 662. Koski, V. M., and Smith, J. H. C, J. Am. Chem. Soc, 70, 3558. Livingston, R., et al, First Annual Report on ONR Project 059028. 1949 Livingston, R., in Photosynthesis in Plants. Iowa State College Press, Ames, 1949, pp. 179-196. Kuhn, H., /. Chem. Physics, 17, 1198. Simpson, W. T., ibid., 17, 1218. 1950 Piatt, J. R., ibid., 18, 1168. Longuet-Higgins, H. C, Rector, C. W., and Piatt, J. R., ibid., 18, 1174. 1951 Tanada, T., A7n. J. Botany, 38, 276. B. Influence of Medium on Absorption Spectrum of Chlorophyll and Bacterio- chlorophyll 1870 Hagenbach, A., Ann. Phys. (Poggendorf) , 141, 245. 1871 Gerland, E., ibid., 143, 585. 1878 Kundt, A., Ann. Phys. {Wiedemann), 4, 34. 1907 Ivanovski, D., Ber. deut. botan. Ges., 25, 416. 1912 Herlitzka, A., Biochem. Z., 38, 321. 1913 Ivanovski, D., Ber. deut. botan. Ges., 31, 600. Ivanovski, D., Biochem. Z., 48, 328. 1918 Willstatter, R., and Stoll, A., Untersuchungen iiber die Assimilation der Kohlensaure. Springer, Berlin, 1918. 670 ABSORPTION SPECTRA OF PIGMENTS IN VITRO CHAP. 21 1922 Eisler, M., and Portheim, L., Biochem. Z., 130, 497. 1927 Noack, K., Biochem. Z., 183, 135. 1932 Padoa, M., and Vita, N., ibid., 244, 296. 1934 Baas-Becking, L. G. M., and Koning, H. C, Proc. Ac. Set. Amst., 37, 674. 1935 Hubert, B., Rec. trav. botan. neerland., 32, 323. Wakkie, J. G, Proc. Acad. Sci. Amsterdam, 38, 1082. 1937 Rabinowitch, E., and Weiss, J., Proc. Roy. Soc. London, Al62j 251. 1938 Smith, E. L., Science, 88, 170. Mackinney, G., Plant Physiol, 13, 427. 1939 Katz, E., and Wassink, E. C., Enzymologia, 7, 97. Biermacher, 0., Thesis, Univ. Fribourg (Switzerland). Egle, K., Sitzber. Heidelberg. Ak. Wiss. Math.-nat. Klasse, 1939, No. 1, 19. Meyer, K. P., Helv. Phys. Acta, 12, 349. 1940 Mackinney, G., Plant Physiol., 15, 359. French, C. S., /. Gen. Physiol, 23, 483. Fischer, H., and Stern, A., in Fischer and Orth, Die Chemie des Pyrrols, Vol. II.2, Akad. Verlagsges., Leipzig, 1940. Pruckner, F., Z. physik. Chem., A187, 257. Seybold, A., and Egle, K., Botan. Arch., 41, 578. 1941 Smith, E. L., /. Gen. Physiol, 24, 565. Albers, V. M., Gibson Island A. A. A. S. Symposium on Photosynthesis (unpublished). 1942 Seybold, A., and Weissweiler, A., Botan. Arch., 43, 252. 1943 Harris, D. G., and Zscheile, F. P., Botan. Gaz., 104, 515. 1946 Rabideau, G. S., French, C. S., and Holt, A. S., Am. J. Bot., 33, 769. 1948 Livingston, R., et al, First Annual Report on ONR Project 059028. 1949 Livingston, R., et al, Second Annual Report on ONR Project 059028. Livingston, R., Watson, W. F., and McArdle, J., /. Ayn. Chem. Soc, 71, 1542. Evstigneev, V. B., Gavrilova, V. A., and Krasnovsky, A. A., Compt. rend. (Doklady) acad. sci. USSR, 66, 1133. Evstigneev, V. B., Gavrilova, V. A., and Krasnovsky, A. A., ibid., 70, 261. Krasnovsky, A. A., Brin, G. P., and Vojnovskaja, K. K., ibid., 69, 393. C. Absorption Spectra of the Carotenoids 1913 Willstatter, R., and StoU, A., Untersuchungen ilber Chlorophyll. Springer, Berlin, 1913. 1928 Pummerer, R., and Rebmann, L., Ber. deut. chem. Ges., 61, 1099 1930 Kuhn, H., Winterstein, A., and Kaufmann, W., ibid., 63, 1489. 1931 McNicholas, H. J., /. Research Natl Bur. Standards, 7, 171. 1932 Kuhn, R., and Brockmann, H., Z. physiol Chem., 206, 41. von Euler, H., Karrer, P., Klussman, E., and Morf, R., Helv. Chim. Acta, 15, 502. 1933 Kuhn, R., and Brockmann, H., Ber. deut. chem. Ges., 66, 407. Rudolph, H., Planta, 21, 104. Stair, R., and Coblentz, W. W., /. Research Natl. Bur. Standards, 11, 703. BIBLIOGRAPHY TO CHAPTER 21 671 1934 Smakula, A., Angew. Chem., 47, 657. 1935 Gillam, A. E., Biochem. J., 29, 1831. Karrer, P., and Solmssen, U., Helv. Chim. Acta, 18, 1306. Sprecher von Bernegg, A. S., Heierle, E., and Almasj^ F., Biochem. Z. 283, 45. Miller, E. S., Botan. Gaz., 96, 447. 1937 Miller, E. S., Plant Physiol, 12, 667. 1938 Karrer, P., and Strauss, W., Helv. Chim. Acta, 21, 1624. Strain, H. H., Leaf Xanthophylls, Carnegie Inst. Wash. Publ. No. 490. 1939 Mulliken, R. S., /. Chem. Phys., 7, 121, 364, 570. Pauling, L., Proc. Natl. Acad. Set. U. S., 25, 577. 1940 French, C. S., Botan. Gaz., 102, 406. 1941 Mulliken, R. S., and Rieke, C. A., Phys. Soc. of London, Reports on Prog- ress in Physics, 8, 231. 1942 Zscheile, F. P., White, J. W., Beadle, B. W., and Roach, J. R., Plaiit Physiol, 17, 331. Beadle, B. W., and Zscheile, F. P., J. Biol Chem., 144, 21. Wald, G., unpublished. Strain, H. H., and Manning, W. M., J. Am. Chem. Soc, 64, 1235. 1943 Strain, H. H., and Manning, W. M., ibid., 65, 2258. Strain, H. H., and Manning, W. M., Carnegie Inst. Wash. Yearbook, 42, 79. Karrer, P., and Wiirgler, E., Helv. Chim. Acta, 26, 116. 1944 Strain, H. H., Manning, W. M., and Hardin, G., Biol Bull, 86, 169. Zechmeister, L., Chem. Revs., 34, 267. Nash, H. A., and Zscheile, F. P., Arch. Biocliem., 5, 77. 1945 Nash, H. A., and Zscheile, F. P., ibid., 7, 305. 1946 Wassink, E. C., and Kersten, J. A. H., Enzijmologia, 12, 3. 1948 Karrer, P., and Jucker, E., Carotinoide, Birkhauser, Basel, 1948. D. Absorption Spectra of the Phycobilins 1888 Schiitt, F., Ber. deut. botan. Ges., 6, 36, 305. 1928 Lemberg, R., Ann. Chem. (Liebig's), 461, 46. Svedberg, T., and Lewis, N. B., /. Ain. Cliem. Soc, 50, 525. 1929 Svedberg, T., and Katsurai, T., ibid., 51, 3573. 1930 Lemberg, R., Biochem. Z., 219, 255. 1931 Dhere, C, and Fontaine, M., Co7npL rend., 192, 1131. Dhere, C., and Fontaine, M., Ann. inst. oc'anog., 10, 245. 1932 Svedberg, T., and Eriksson, L B., /. Am. CJiem. Soc, 54, 3998. 1933 Roche, J., Arch. phys. biol, 10, 91. 1939 Katz, E., and Wassink, E. C., Enzymologia, 7, 97. 1948 Wassink, E. C., Enzymologia, 12, 362. Van Norman, R. W., French, C. S., and Macdowall, F. D. H., Plant Physiol, 23, 455. 1951 French, C. S., Proc. Soc. Exptl Biol (in print). Chapter 22 LIGHT ABSORPTION BY PIGMENTS IN THE LIVING CELL The determination of light absorption in sokitions or other homogeneous media is a routine measurement, and the results permit a simple interpre- tation (based on Beer's law) in terms of molecular absorption coefficients (also called "extinction coefficients" — since attempts to discriminate be- tween these two terms have not been successful in practice). The experi- mental determination of the absorptive power of plants is less simple, and often the exact meaning of the results is problematical. The measurement of light energy absorbed by leaves, algal thalli or cell suspensions is com- plicated by scattering, which is significant not only in multicellular tissues, but even in suspensions of single cells (because the dimensions of the cells, ~10~^ cm., are larger than the wave length of visible light, ~5 X 10~^ cm.) . The interpretation of the results in terms of the absorption constants of the pigments is complicated, not only by the light scattering on phase boundaries, but also by inhomogcneous distribution of pigments in cells and tissues, and by the shifting and deformation of the absorption bands caused by adsorption and complexing. Let us assume, for example, that we have measured the energy, /, of a beam of light falling on a vegetable object — leaf, thallus or cell suspension — and the energy, /', emerging from this object, taking care to integrate the latter over all directions so as to include both the light transmitted forward (T) and the light reflected backward (R), and thus to avoid the "gross" errors that may be caused by scattering. If now we try to apply to the results Beer's law : (22.1) I' (= T + R) = 7 X 10-"^'' with the intention of calculating an absorption coefficient, a, we find, first of all, that scattering has made the length of the path of the light in the ab- sorbing medium — d in eq. 22.1 — indefinite (even its average — Mestre's "de- tour factor" — is not constant, but depends on wavelength, cf. Kok 1948). In the second place, we note that the local accumulation of pigments in the chloroplasts has made the concentration of absorbing molecules in the path of the individual light beams — c in equation (22.1) — variable: Some light l)eams pass between the chloroplasts and encounter no pigment molecules at all (a phenomenon to which we will refer later as the "sieve effect"). G72 GENERAL REMARKS ON LIGHT ABSORPTION BY PLANTS 073 In the last place, if, overcoming these two difficulties, we succeed in obtain- ing a reliable value of a, it is an average absorption coefficient of a mixture of several pigments, whose individual absorption spectra in solution we may- know, but whose bands are variously shifted and deformed, in the living cell, by adsorption and complexing. The task of apportioning the total absorption at a given wave length to the component pigments (which re- (juires the knowledge of their individual absorption coefficients and of their distribution in the cell) often proves impossible of achievement, except by gross simplifications. We shall deal first, in part A, with the determination of the amount of light energy absorbed })y plants, and then, in part B, with the spectroscopic properties of individual pigments in vivo and their contribution to the total absorption. A. Light Absorption by Plants* 1. General Remarks In working with solutions in plane-parallel glass cells, the detennination of the absorbed light energy (A) requires two measurements: Either one measures the incident light flux (/) and the transmitted light flux (T), or, more commonly, one compares T with the flux To transmitted by a blank cell containing pure solvent. A is calculated by one of the following: (22.2a) A = I - T or (22.2b) A = To - T Both Sive first approximations. Equation (22.2a) neglects all reflections; a second approximation can in this case be obtained by subtracting from / the light flux reflected from the front wall of the absorption cell : (22.3) A = 1(1 - r) - T where r is the reflection coefficient of the cell material. However, reflec- tion from the front wall is only part of the total reflection in the cell; to make our equation exact, we should write, in place of (22.3) : (22.4) A = I - T - R meaning by R the total reflected flux. Equation (22.2b) is a better first approximation than (22.2a), because it neglects only the difference between the reflections from the solution cell and the blank cell. The fluxes reflected from the front walls of both cells are identical, but those reflected from the back walls are different (because * Bibliography, page 736. 674 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 of the weakening that hght suffers in passing through the absorbing me- dium). A second approximation, which can be substituted for (22.2b), is: (22.5) A = To - T + I{1 - r)[r(l - lO-^^-^j] where a is the absorption coefficient of the soUition and d the thickness of the cell. In this equation, consideration has been given to one reflection from the front wall (factor 1 — r) and one reflection from the back wall (factor in brackets) . However, these reflections are only the beginning of an infinite series (as with two mirrors on opposite walls) . The exact equa- tion for A in terms of the properties of the blank cell is : (22.6) A = {To - T) + {Ro - R) where Ro and R are the total light fluxes reflected by the blank cell and the solution cell, respectively. If a and r are known, exact values of To, T, Ro and R can be obtained by the summation of infinite power series. If the light "trapped" between the walls leaves the cell after an odd number of pas- sages, it is added to the transmitted ilux; if it escapes after an even number of passages, it is added to the reflected flux. Consequently, the series for T contains only even powers of r and odd powers of 10""'^, and the series for R only odd powers of r and even powers of lO""''. (The ratio of the sums for T and To is given in equation 22.12.) If a and r are unknown, R (and Ro) must be determined experimentally. Since repeated reflection lengthens the average path of the light in the absorption cell, it increases absorption. In the case of homogeneous solu- tions in plane-parallel glass cells, this increase represents only a minor cor- rection (to be estimated on page 711); we mention it merely to illustrate the complications in the measurement of light absorption that arise from the presence of phase boundaries. In nonhomogeneous systems, the com- plications are similar in principle, but much more important quantita- tively. Reflections are not only more numerous, but also stronger (because of the varying angles with which the light strikes the interfaces) ; and they are supplemented by refractions and total inner reflections, which all af- fect the length of the path of the light beam and the direction in which it leaves the medium. Leaves and thalli are heterogeneous systems, with numerous phase boundaries between ah- channels, cell w'alls, cytoplasm, vacuoles, plastids and starch grains ; and the passage of light through plants or plant organs is, therefore, a very complicated phenomenon. It has been repeatedly dis- cussed—by Willstatter and Stoll (1918), Briggs (1929), Mestre (1935), Seybold and co-workers (19321-2, 1933'-'-, 1934, 1943), Schanderl and Kaempfert (1933), Meyer (1939) and Loomis (1941, 1949), among others— but these discussions have not gone far beyond the qualitative stage. GENERAL REMARKS ON LIGHT ABSORPTION BY PLANTS 675 A suitable statistical theory (c/. page 713) may permit the calculation of A from measurements of light transmission in one direction, made with two or more different optical densities of the scattering material (e. g., with a series of several leaves, or with several cell suspensions of different concentration or layer thickness) . However, it is better not to rely on such theoretical equations, but, particularly in working with leaves or thalli, actually to measure the light fluxes transmitted and reflected in all direc- tions. Having determined experimentally both T and R, one can use the exact equation (22.4) for the evaluation of A. The time to use theoretical equations for combined absorption and scattering comes when one is not satisfied with the knowledge of the amount of absorbed energy, but wants also to know the absorption coefficients, e. g., as indicators of the molecular state of the pigment in the living cell. Attempts have been made to use "blanks," for example, white parts of variegated leaves (cf. Linsbaur, 1901, Brown and Escombe 1905, Weigert 1911, Meyer 1939, and Seybold 1932i'2, 1933i, 1934), or algal thalli from which the pigments had been ex- tracted {cf. Reinke 1886), or tissues bleached by long exposure to light (c/. Wurmser 1926), and to imitate in this way the method usually applied to transparent media. In the latter case, the blanks provide an automatic correction for reflection (cf. page 673) ; in the case of plants, they were intended to provide a correction also for scattering. However, the approximation (22.2b), which is generally satisfactory in work with trans- parent media, may give entirely erroneous results when applied to optically inhomogene- ous systems. This was pointed out by Willstatter and Stoll (1918) and Warburg (1925) when they criticized the absorption calculations of Weigert (1911). The error is caused by the large difference between the fluxes R and Ra {cf. equation 22.6) reflected by the green and the colorless leaf. A green leaf may transnut about 10% and reflect another 10% of incident white light, while a similar, pigment-free leaf may transmit 50% and reflect the other 50%. If the absorption of the green leaf is calculated from these figures by means of equation (22.2b), the result is A = 40%, which is only one half the correct value (80%)! Therefore, if one wants to determine absorption, A, by comparison of a green leaf with a pigment-free leaf, one has to use the complete equation (22.6), i. e., to measure the four quantities To, Ro, T and R, while measurement of only three quantities, /, T and R, is sufficient to make the same determination with a single leaf, according to equation (22.4). Furthermore, in plant work, one is never certain whether the "blank" is entirely free of pigments: Accordmg to Seybold and co-workers (1933S 1942), so-called "white" leaves of Acer negundo absorb 10-20% of incident white light; this absorption may be caused by nonplastid pigments, or by a small quantity of residual chlorophyll or caro- tenoids. The transmitted light flux (T) and the reflected light flux (R) can both contain a coUimated component, T, or R^ (light transmitted in the direc- tion of the incident beam, or reflected according to the laws of specular re- flection), and a diffuse component, T^ or Ra, so that equation (22.4) can be written more explicitly as follows : (22.7) A = I - {r. + Ti) - {R. + Ri) G76 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 Measurements of T and R must include both the colhmated and the diffuse components. If the leaf is sufficiently thick, and not glossy, it acts as an ideal scatterer, i. e., the intensity of light scattered in a given direction is proportional to o UJ u. UJ o a: LiJ I- UJ < < =*'**;S^. 1 1 Cosine curve ""•N ^ ^/ ^ V>^ L, X ^^ 1 >^„^Reflection Trans missio n N ^^ V ^s^^ 0 10 20 30 40 50 60 70 80 90 ANGLE, degrees Fig. 22.1. Angular distribution of light transmitted and reflected liy a leaf of Coleus blumei (after Loomis, Carr and Randall 1941). Reflection is diffuse and obeys the cosine law; transmission is only partly diffuse and therefore deviates from the cosine law. II0J00_901_80 70 Fig. 22.2. Scattering of light by a dense suspension of Chlorella (in vessel ()) (after Noddack and Eichhotf 1939). Direction of incidence A ^ B; area a represents backward scattering, or reflection (R); area b forward scattering or transmission (7'). the cosine of its angle with the direction of the incident light. Figure 22.1 shows the angular distribution of the light scattered by a comparatively thin leaf of Coleus. This leaf transmits some colhmated light (as evidenced by the deviation of the angular distribution of T from the cosine law), but its reflection is entirely diffuse. Thicker leaves, with similarly dull sur- faces, obey the cosine law with respect to both reflection and transmission. TRANSMITTANCE AND REFLECTANCE OF LEAVES 677 whereas thick leaves with glossy surfaces may show a completely diffuse transmission, but a marked specular reflection. Figure 22.2 shows the forward and back scattering of light by a small vessel containing a suspension of Chlorella cells, as observed by Noddack and Eichhoff (1939). The sharp peak (C) at 180° is caused by specular reflection from the glass wall. In practical work, one can often di'op the distinction between T and R and measure the total scattered flux .S' ^ {T -{- R) b\' means of some inte- grating device: (22.8) A = I - S An example is the study by Rabideau, French and Holt (1946) of the ab- sorption spectra of leaves and pigment extracts. 2. Average Transmittance and Reflectance of Leaves and Thalli in White Light. Intensity Adaptation and Movements of Chloroplasts The first measurements of the proportion of white light transmitted by leaves were carried out by Sachs in 1861. Later this magnitude was meas- ured by Detlefson (1888), Linsbauer (1901), Brown and Escombe (1905), Purevich (Purjewitsch) (1914), Schanderl and Kaempfert (1933), Seybold (1932i'2, 19331-2, 1943), Loomis, Carr and Randall (1941,1947,1949). The transmittance oj algae was investigated by Reinke (1886), Wurmser (1921) and Seybold and co-workers (1934, 1942). The first measurements of the reflectance of leaves were made by Co- blentz in 1912, and were followed by those of Pokrovski (1925), Shull (1929), Seybold and co-workers (19322,1933i'2,1942,1943) and Loomis, Carr and Randall (1941,1949). The only data on the reflectance of algae are those of Seybold and co-workers (1934, 1942, 1943). Brown and Escombe (1905) and Purevich (1914) found comparatively high values — of the order of 20% — for the transmission of (infrared-free) white light by average leaves. Seybold (1932) suggested that these re- sults were falsified by the inclusion, in the measured transmitted flux, of the thermal radiation of the leaves. In agreement with Pokrovski (1925), Seybold found that an average fully green leaf transmits not more than 10% of infrared-free white light. Leaves are almost transparent in the far red and near infrared (c/. figs. 22.30 and 31). Therefore, transmission values obtained by means of thermopiles (or other infrared-sensitive in- struments) are deceivingly large if the light used for the measurements contains a large proportion of infrared radiations. According to Loomis, and co-workers (1941,1949), an average leaf transmits 30% of total sun- light, including the infrared. With artificial light sources of lower tempera- 678 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 ture, the over-all transmission may be much greater. Selenium barrier layer photocells ("photronic cells"), because of their low sensitivity in the infrared and in the far red, indicate, in direct sunlight, a transmission of only 5 to 10%, depending on the thickness of the leaf (c/. Seybold 1932^, and Egle 1937). The light absorption by a leaf depends on its thickness and the concen- tration of the pigments. As mentioned above, fully green leaves transmit and reflect only 10 or 15% of incident visible light, and absorb as much as 85 or 90%; on the other hand, green onion skins, one cell thick, transmit 85%, reflect 10% and absorb less than 5% (Seybold 19322). In chapter 15 70 - I 80 OC O A 1/ Shade// N>. — '/ ^v> / //san ^- i // // // >SN. // ^y ^S.^/ ^ ^^^■^■"— '^ ^ ^ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 500 600 WAVE LENGTH, m/i 700 Fig. 22.3. Absorption spectra of a shade leaf and a sun leaf of Fagus sylvatica (the shade leaf contains 50% more chlorophyll and 80% more carotenoids) (after Seybold and Weissweiler 1943). (Vol. I) we described the adaptation of plants to the intensity and compo- sition of the incident light. We found there that typical "shade plants" contain two or three times more pigment than typical "sun plants." The probable purpose of this "intensity adaptation" is to ensure an adequate supply of energy to plants that grow in the shade, or to algae that live deep under the sea, and, inversely, to prevent light injury to plants or algae ex- posed to direct sunlight. There is no doubt that the presence of fucoxanthol or of the phycobilins in brown, red and blue algae has considerable influence on the amount of light adsorbed; this will be brought out in detail in part C. The effects of variations in the concentration of chlorophyll are much smaller. Even "light-green" plants absorb so large a proportion of incident light that a doubling of their chlorophyll content can increase the absorption only com- paratively little. Three examples can be found in our illustrations: Fig- ure 22.3 shows the very slight enhancing effect that a 50% excess of chloro- phyll and 80% excess of carotenoids have on the absorption of light by a MOVEMENTS OF CHLOROPLASTS 679 shade leaf of Fagus, as compared to a sun leaf of the same species. The two lowest curves in figure 22.14 indicate a sightly larger difference between the spectral transmission curves of a dark-green and a light-green Hibiscus leaf. Finally, figure 22.10 illustrates the effect of extreme variations in chlorophyll content, such as occur in aurea leaves. Here, a few per cent of the normal chlorophyll content (cf. Table 15.1, Vol. I) suffice to produce from two thirds to nine tenths of normal absorption in the region between 520 and 700 m^. Without varying the concentration of the pigments, many plants have it in their power to adjust the light absorption by the displacement or re- orientation of the chloroplasts. These tactic reactions, discovered by Bohm in 1856, were investigated by Stahl (1880, 1909), Senn (1908, 1909, 1917, 1919), Liese (1922) and Voerkel (1933), among others. It was found that each chloroplast moves independently, i. e., it is not carried by streaming of the protoplasm. In moderate light, the chloroplasts gather on the illumi- nated front walls and orient themselves so as to present their large cross- 1 I Apostrophe Antistrophe Diastrophe Porastrophe Fig. 22.4. Schematic representation of different chloroplast orientations (after Benecke and Jost 1924). Arrows show direction of light incidence. sections to the light ("antistrophe," "epistrophe" and "diastrophe" in figure 22.4; the first one is produced by one-sided, and the other two by two-sided, illumination). In strong direct light, on the other hand, the chloroplasts turn their axes parallel to the light beams and line the side walls of the cells ("parastrophe" in fig. 22.4). During the night, they often assume characteristic "night positions" — congregate around the nuclei, or disperse throughout the cytoplasm, or line the internal walls ("apostrophe" in fig. 22.4). The largest variety of chloroplast movements has been ob- served by Senn in green and brown algae, and in diatoms. In leaves, the chloroplasts in the parenchyma cells assume positions similar to those shown in figure 22.4, but the chloroplasts in the palisade tissue usually re- main arrayed along the side walls, leaving the end walls free. Instead of moving bodily these chloroplasts merely change their shape : In strong dif- fuse light, they spread flat against the walls, whereas in weak light they protrude into the cytoplasm, without losing contact with the walls. lUu- 680 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 mination with strong parallel light may, however, produce antistrophe in these cells as well. Obviously, the effect of parastrophe is to decrease, and that of epis- trophe, antistrophe or diastrophe to increase the absorption of hght. How Fig. 22.5. Chloroplasts in Funaria (after Voerkel 1933): above, in liglit (epistrophe) ; below, in darkness (apos- trophe). en en en < a. < UJ o 130 120 no 100 90 80 70 60 50 40 30 20 10 ° 1 i H- O - c — o o 1/ +■ 1 - C /— a \ \ Starch formation 1 1 Profile orient ctiloropiosts \ \ / 1 . \ i"/ \ I \ i hi) . \ \ \ 1 ■ ^X ' N \ m \ \ 20 40 60 TIME, min. 80 Fig. 22.6. Changes in light transmission through leaves of Tradescantia viridis, caused by chloroplast orientation and starch formation (after Schan- derl and Kaempfert 1933). successfully this can be achieved is illustrated by figure 22.5, which shows how Funaria cells change their appearance upon transition from epistrophe to apostrophe. Differences in the transmittance of leaves in light of different intensity, caused by the regrouping of chloroplasts, have first been actually observed by Detlefson (1888) and Stahl (1880, 1909). A quantitative investigation was made by Schanderl and Kaempfert (1933); typical results are shown MOVEMENTS OF CHLOROPLASTS 681 in Table 22.1. This table indicates that in blue-violet light increase in transmittance may be by as much as one-third (from 19 to 25%). Table 22.1 Transmittance of a Leaf of Adiantum cuneatum (after Schanderl and Kaempfert 1933) Light exposure In diffuse room light, % After 4 hr. exposure to sun, %. White light (inchiding infrared) Red and infrared 31 37 34 42 Yellow and green 16 18 Violet and blue 19 25 With some plants {e. g., T radescantia ?>mrfis). Schanderl and Kaempfert found a reversal of the effect after the first half hour of illumination {cf. fig. 22.6) ; they attributed it to increased scattering, caused by the forma- tion of starch grains. Schanderl and Kaempfeit did not prove that increased transparency of sun-exposed leaves was due entirely to reorientation of the chloroplasts, and not, e. g., to a partial bleaching of the pigments. However, the results of Willstatter and StoU (1918), which showed no change in chlorophyll concentration after strong illumination {cf. Vol. I, chapter 19, page 549), argue against the second explanation. 100 90 80 »? 70 - BLUE LIGHT . X««1(-X-)H( "X^X- / /BLUE-GREEN LIGHT YELLOW-GREEN LIGHT/ L5 20 2.5 3.0 3.5 4.0 LOG INTENSITY, lO^cal/cm.^ hr. 45 Fig. 22.7. Orientation of chloroplasts in Funaria in light of different color (after Voerkel 1934). The phototaxis of the chloroplasts is caused mainly by blue-violet Hght and is entirely absent in red light (fig. 22.7). It must thus be sensi- tized by the carotenoids rather than by chlorophyll. This is not a proof 682 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 (a) (b) Fig. 22.8. Leaves of Naegelia zebrina (after Mecke and Baldwin 1937). Normal leaves, left; water-filled leaves, right; photographed on (a) panchromatic plate and on (6) infrared-sensitive plate (max. 850 mju)- Increase of transmittance by decreased scattering best recognizable on the infrared photograph. Fig. 22.9. Leaf of Tussilago farfara on black background (after Schan- derl and Kaempfert 1933). Dark sections imbibed with water, therefore more transparent than the air-filled section. TRANSMITTANCE AND REFLECTANCE OF LEAVES 683 that the whole phenomenon is unrelated to photosynthesis, since in natural light fields, red light does not occur without the presence of some blue- violet hght as well. Detlefson (1888) and Purevich (1914) found that leaves become less transparent in the presence of carbon dioxide, but Ursprung (1918) was unable to confirm this result. This effect, if at all real, could be attributed to increased formation of starch grains. Scattering of light in leaves can be decreased by injection of water into the air channels (e. g., by evacuation under water; cf. figs. 22.8 and 22.9). According to Seybold (1933-), water-filled leaves of land plants (as well as those of aquatic plants of the type of Elodea) transmit about twice as much light as air-filled leaves. For example, a submerged leaf of Potomageton alpinus was found to transmit as much as 22% of white light (Seybold 1932); similar figures were obtained by Seybold (1934) for the transmis- sion by algae {Chloj-ophyceae, Phaeophyceae and Rhodophyceae). In- creased transparency of water-filled leaves or thalli is due to reduced dif- fuse reflection (5% instead of >10%), rather than to weaker absorption. As a rough approximation, it can be assumed that average land leaves transmit 10% of (infrared-free) white light (400-700 m/x), reflect 10%o and absorb 80% ; while average aquatic plants or algal thalh reflect 5%, trans- mit 15% and absorb 80%. Loomis (1947) found, for 28 species of leaves, transmissions between 2 and 9%, and reflections of the same order of magnitude. Yellow tobacco leaves transmitted -^ 30% of visible light, and light yellow leaves, 36% of visible and 53% of total sunlight. Normal leaves of Nicotiana and Quercus transmitted 5-7% of visible, and 25-30% of total sunlight. According to Seybold, a single chloroplast transmits from 30 to 60% of visible light (depending on spectral distribution), absorbs 30 to 60% and reflects about 10%. As mentioned in chapter 19 (Vol. I) an average leaf contains the equivalent of from five to ten complete layers of chloroplasts; these can easily account for the above-estimated absorption of 80% of the incident white light. For a detailed discussion of the light absorption in successive layer of chlorophyll molecules or plastides, see Seybold and Weissweiler (1943). In regularly patterned systems of colored and colorless materials, the absorption sometimes depends upon the angle of incidence. WTien this is the case the absorption of diffuse light may be different from that of collimated light. Seybold (1933'') made a search for such an effect in leaves, but without suc^esSi The reason probably is that leaves are such strong scatterers that collimated light is converted into practically com- pletely diffuse light, long before it emerges from the leaf. 684 LIGHT ABSORPTION BY PIGMENTS JN VIVO CHAP. 22 3. Absorption by Nonplastid Pigments When the total absorption of hght by a leaf, thallus or cell suspension has been determined, the question arises as to what part of this absorption is due to the chloroplast pigments. It has been assumed by many authors — ^from Reinke (1886) to Noddack and Eichhoff (1939) — that a certain part of the absorption of white light in plants is due to the "colorless" parts of the tissue, the cytoplasm, cell sap, nuclei, starch and cellulose. Seybold arbitrarily ascribed one eighth of the total absorption to these components, and seven eights to the chloroplast pigments. An absorption curve of a white Pelargonium leaf, given by Seybold and Weissweiler (1942), shows considerable absorption near the blue-violet end of the visible spec- trum. Of course, no really colorless substance can absorb visible light. But the plant cells contain coloring materials associated with the cell walls or with the cell sap rather than with the plastids, such as flavones, tannins, etc. Some of these substances are only weakly colored — ^usually yellow; others, although intensely colored, are present only in small quantities, as com- pared to the plastid pigments. In some species, however, flavones and anthocyanines are so abundant as to give the leaves a striking red color (leaves of the purpurea variety, and many young leaves in the spring). The color of these leaves advertises the fact that much of the light energy they absorb goes to nonplastid pigments. Figure 22.10 shows, as an example, the spectral transmission and re- flection curves of three varieties of Corylus avellana, normal, aurea and purpurea. Curves 1 and 2 illustrate the effect on light absorption of ex- treme variations in the concentration of chlorophyll (c/. page 678), while curve 3 shows the considerable increase in absorption, particularly in the green, caused by the presence of anthocyanines in purpurea leaves. (Light absorption in aurea and purpurea leaves will again be discussed in chap- ters 28 and 30, in relation to the yield of photosynthesis.) As far as green leaves and algae are concerned, the participation of non- plastid pigments in light absorption remains a moot question. It probably varies widely from species to species. For example, according to Thimann (unpublished), leaves of Phaseolus vulgaris contain a large quantity of yel- low, water-soluble pigments. The same is true of the needles of the coni- fers (Burns 1942). It was mentioned above (page 675) that "white" leaves may absorb 10 or 20% of incident white light (Seybold 1933S 1942, cf. fig. 22.12). This absorption, too, is probably due to nonplastid pigments. Most investigators who measured the yield of photosynthesis in rela- tion to the amount of absorbed light silently assumed that the absorption NONPLASTID PIGMENTS 685 by the nonplastid components was negligible; but this assumption is not, or is not always, justified, particularly in work with blue-violet light. It is therefore advisable to ascertain, before undertaking quantitative work on photosynthesis, whether the plants chosen are free from, or at least poor in, nonplastid pigments. iS o u _l (xl g tn < Q. a: o en OQ < 400 30 20 10 WAVE LENGTH, m^ 500 600 700 C 30 20 10 lOOF 90 80 70 60 50 40 ' ' ' ■ 1 ' -x-" (0) / ..; 1 ■ ■ . . „ / s , 2/ ^ ^~\ / 'A - -J 3 ^\-^ z' -*«* (^) 2/ ""^--x ; / \ / / / ""^ 'A ^*>>«,^_^ \/ 1 1 / «s>^^ / J J T^^\ / / / •'*' **"'*^'^. y ^ -^ 3 _...-•"" _ ...— - - 1 3" "■■■■-.. *^^ \ — "^ — * \ , > n\ **• 'y^A*. (c) \\ / \ \\l y^ \ \ \ 2\\ \ \ / 1^^ 1 \ I ««-- / /; \ / \ t \ /'" \ / \ / ^ y \ y \ / — . — 1.. 1 1 1 1 , " 400 500 600 WAVE LENGTH, m/i 700 Fig. 22. 10. Spectra of normal (curves 1), aurea (curves 2) and "pur- purea (curves 3) leaves of Corylus avellana (after Seybold and Weiss- weiler 1942). (a) Transmission curves, (6) reflection curves, (c) ab- sorption curves. New absorption curves of yellow, red and white air- filled and air-free leaves can be found in the paper by Seybold and Weissweiler (1943i). Noddack and Eichhoff (1939) attempted to determine the absorption of light by the "colorless" components of Chlorella, and concluded that it is negligible. This con- clusion may be correct, but the method employed — comparison with the absorption Ijy 686 LIGHT ABSORPTION BY PIGMENTS IiV VIVO CHAP. 22 cells bleached with bromine — does not appear reliable, since bleaching by bromine may not be restricted to plastid pigments. It was suggested that the proportion of the total light absorption by leaves due to plastid pigments can be determined by extracting the latter and comparing the absorp- tion of the extract with that of the original leaf. Timiriazev (1885) and Lazarev (1924, 1927) based this suggestion on the assumption that the absorption of white light by ex- tracted pigments is about equal to the absorption by the same pigments in the leaf. However, light scattering, "sieve effect" and band shifts can make the two magnitudes quite different. True, some of these factors enhance the absorption, while others weaken it, but it would be an unusual coincidence if the net result were exactly nil. Experimentally, Noddack and Eichhoff (1939) found that the increased absorption by live Chlorella cells in the far red almost exactly compensates for the decreased absorption in the region 520-680 m/x (c/. fig. 22.21); but Seybold and Weissweiler (1942) could not confirm this result, and found, to the contrary, that throughout the visible spectrum the absorption by live cells is more complete than that by the extracts. B. Spectral Properties of Plants* 1. Empirical Plant Spectra It is clear from the preceding discussion that what is usually called a "leaf spectrum," or even a "spectrum of a cell suspension," is not the true spectrum of the pigments contained in these materials (meaning by "true spectrum" the plot of the absorption coefficient of the pigment mixture 16 - "A 14 \ »« \ r\ \ 1 ■£ / / \ \ o 12 _ / / \ \ ►- // y \ o / f \ UJ 1 1 \ \ 1 il^ >o "s T / / J / \ \ / iij \ ,'r \ \ / o 8 - ' ^^ \ / o \ /' V, 1. ^n\ 1 (n - >A 1 (/) 6 ~ V^ A 5 s. \ 1 (/5 N \ 1 1 4 - ^ a: 1- 2 - 0 . . . 1 . . , . 1 . . . . 400 700 500 600 WAVE LENGTH, m>i Fig. 22.11. Transmission and reflection spectrum of a leaf of Parie- taria officinalis (after Seybold and Weissweiler 1942). * Bibliography, page 737. EMPIRICAL PLANT SPECTRA 687 100 o UJ _l li. tlJ a. ■£ o in (/) z < 400 700 500 600 WAVE LENGTH, m,i Fig. 22.12. Transmission (Jg), reflection {Rg) and absorption (A^,) of light by a green leaf of Pelargonium zonale, and the corresponding constants for a white leaf (index w) (after Se.ybold 1933). The figure indicates considerable absorp- tion by "white" leaves, particularly in the blue- violet region. For other absorp- tion curves of "white" leaves, see Seybold and Weissweiler (19430. 4.4 o o 4.01—1 400 450 500 550 600 WAVE LENGTH, m^ 650 Fig. 22.13. Transmission spectrum (log Ta/T) of leaves of Fatsia and Acuba (after Meyer 1939). Absolute values of ordinates adjusted to give best agreement with spectrum of extracted pigments. against wave length). One may measure /, T and R {or S = T -{- R), plot T/I, R/I, or A/I (= [/ - (T + R)]/I = [I - S]/I) against wave length, X, and call the resulting curves "transmission spectra," "reflection 688 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 en en ir 350 400 450 650 700 750 500 550 600 WAVE LENGTH, m/x Fig. 22.14. Transmission (T/I) of dark-green, light-green and yellow leaves of Hibiscus rosa-sinensis, and of equivalent quantities of pigments in ether (after Loomis, Carr and Randall 1941). With allowance for reflection, the leaves absorbed more light than the extracted pigments in the green part of the spectrum, and less in the blue and red. •^"^^v^OOOSAmg./ml. , J ^ Supersonic Absorption \ ,y \ treated V "'''/ 1 choroplost \\^ ,.' ^ A suspension \ (Spinacti) ■Reflection ^__ — 1 ---. 1 1 \ <- — ■ — 00084 mg /ml Clioroplast suspension (Spinach) 400 800 500 600 700 800 400 500 600 700 WAVE LENGTH, m/i WAVE LENGTH, m^ Fig. 22.15. Absorption and reflection spectra of leaves (G, H), of chloroplast suspen- sions (I, K) and of the supernatant fraction after centrifugation of disintegrated chloro- plasts (J, L) (as measured with the Ulbricht sphere) (after Rabideau, French and Holt 1946). The chlorophyll concentrations of the suspensions are given on figures. EMPIRICAL PLANT SPECTRA 689 (f) o Chlorella (Emerson and Lewis) -1.4 - Sphere 400 500 600 700 WAVE LENGTH, m;i 800 400 500 600 700 WAVE LENGTH, m/i 800 Fig. 22.16. Leaf and chloroplast absorption spectra (on a log density basis and ad- justed to the same height at 670 m/u.) (after Rabideau, French and Holt 1946). The Beckman curve of the chloroplastin solution (chloroplasts disintegrated by supersonics) is an / - T plot (;'. e., it represents absorption plus scattering); the Ulbricht sphere curve is an / — T — R plot (pure absorption). An absorption curve of Chlorella (after Emerson and Lewis) and a photosynthesis action curve (after Hoover et al.) are given for comparison. g en if) z < a: 100 90 80 70 60 50 40 30 20 10 E, Elodeo - P, Potomageton | Er, Elodeo after Reinke e, green cells, after Engelmann ■■ - " .* ^ '' / ■■ .* /^ Atn^ '• / 1 n rt v^ ./ / - vT""' / / : j ^V \ i " ../■'■' / V ^, 'S-yy 1 1 100 0 400 700 500 600 WAVE LENGTH, m/i Fig. 22.17. Transmission spectra of water plants (after Seybold 1934). a. o CQ < O CO en en z < cr H 500 600 700 WAVE LENGTH, m/i Fig. 22.18. Transmission T, absorp- tion A and reflection R by green alga Monostroma. Absorption curve D of red alga Delesseria given for comparison {cf. fig. 22.20) (after Seybold 1934).* * For an absorption curve of Ulva lactuca and its extracted pigments, see Seybold and Weissweiler (19430- 690 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 100 90 80 G, Fucus single cells according to Gaidukov R, Phyllitis occordingto Relnke £, Brown cells according to Engelmonn 500 600 700 WAVE LENGTH, m/i Fig. 22.19. The transmission spectra of brown algae (after Seybold 1934). Maximum of transmission is at 600 m^i instead of ^^550 mju as in green plants and algae (figs. 22.17-18) because of the presence of fucoxanthol. For ab- sorption curves of Fucus and Laminaria, see Seybold and Weissweiler (1943')- •- 400 500 600 700 WAVE LENGTH, m/i Fig. 22.20. Transmission T, reflec- tion R and absorption A of the red alga Delesseria (according to Seybold 1934). More recent curves for Delesseria and Porphyra are given in Seybold (1943); for Iridaea and Gigartina, see Van Norman, French and Macdowall (1948). 500 600 700 800 WAVE LENGTH, m/i Fig. 22.21. Absorption spectra, (/ - S)/I = {I - T - R)/I, of Chlor- ella suspension (after Noddack and Eichhoff 1939). About 1.1 X 10' cells and 1.6 X 10"^ g. chlorophyll a and 6 in 1 ml. A, 35 ml. pigment extract in a vessel 3.90 cm. thick; B, cell suspension, same volume, same vessel. EMPIRICAL PLANT SPECTRA 691 spectra" or "absorption spectra," respectively; or one may use colorless specimens as "blanks," determine To and i^o (or So = To + ^o), plot T/To, R/Ro or .Sy,So against X and designate these curves as "transmission," "reflection" or "absorption" spectra, respectively — but, although each of these plots is legitimate as representation of a certain property of the specimen investigated, they are all different. (For example, fig. 22.11, p. 686, shows the transmission spectrum and the reflection spectrum of the same leaf.) The true absorption spectrum on the other hand is an intrinsic property of a molecular species (or, in the case of a mixture, the average of intrinsic properties of several molecular species). We will discuss the quantitative analysis of the empirical "leaf spectra" 1.4 L2 1.0 §0.6 _i 0.4 0.2 1 1 1 u ■— Intact cells -- Combined extracts 1 \ ~~ 1 \ 1 1 \ \ \ \ ^ \_ \ / / / v_ / 1 \ J V \ 4 V \ \ \ N ^' - 0.7 0.6 0.5 04 0.3 0.2 0.1 f 1 1 1 r I 1 1 Intact cells 1 \- Combined extracts / / A A y / ^ t \ \ vj / \ 1 \ i\ 400 440 480 520 560 600 640 680 720 WAVE LENGTH, m/i 400 440 480 520 560 600 640 680 720 WAVE LENGTH, m/i Fig. 22.22. Tran.sniission .spectrum of Fig. 22.23. Transmission spectrum {To/T) Chhrella suspension (after Emer- {To/T) of suspension of Chroococcus son and Lewis 1943). (blue alga) (Emerson and Lewis 1943). or "cell spectra" below, in section 4. First, we will present a selection of experimental results, and discuss their qualitative aspects. Spectral data on leaves of land plants have been collected by Ivanovski (1907, 1913), Willstatter and Stoll (1918), Ursprung (1918), Wurmser (1921), Lubimenko (1927), Seybold and co-workers (19321-2, 1933, 1934, 1936, 1942i'2, 1943), Meyer (1939), Loomis et al. (1941, 1949), Iljina (1946) and French et al. (1946), among others. The curves on pages 686 to 689 are from some of the more recent investigations. Figure 22.17 shows the transmission spectrum of the water-filled leaves of the aquatic plants Elodea and Potomageton, according to Seybold (1933). Transmission spectra of algal thalli were measured by Reinke (1886), Engelmann (1884, 1887), Gaidukov (1904), Wurmser (1926) and Seybold and co-work- ers (1934, 19421--, 1943). Figures 22.18-22.20 show curves given by 692 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 the last-named observer. Transmission spectra of suspensions of unicellu- lar algae were studied by Noddack and Eichhoff (1939), Katz and Wassink (1939), Emerson and Lewis (1941), Wassink and Kersten (1946), Van Norman, French and Macdowall (1948) and Tanada (1951). Several such curves are reproduced in figures 22.21-22.24. Albers and Knorr (1937) measured the absorption spectra of single chloroplasts, in the narrow region 664-709 mix (cf. fig. 22.35). Vermeulen, Wassink and Reman (1937), Katz and Wassink (1939), Wassink, Katz and 540 1 — r-T — I — r— 1 — r—r- 800 900 960 600 700 WAVE LENGTH, m/i. Fig. 22.24. Transmission spectrum (To/T) of Oscillatoria (l)lue alga) (after Katz and Wassink 1939). Curve 1, cell suspension; curve 2, chlorophyll extract. Dorrestein (1939) and French (1937, 1940) observed the transmission spec- tra of purple bacteria (figs. 22.25 to 22.28). Egle (1937) and Loomis, Carr and Randall (1941) investigated the trans- mission and reflection of leaves in the infrared. Figures 22.29 and 22.30 show that, from 800 to 1300 ni/x, T -\- R accounts for 85 or 90% of the inci- dent light in the (comparatively thin) potato leaf, and for 75 or 85% in the (0.6 mm. thick) leaf of Ficus elastica. This region includes most of the infrared radiation of the sun that reaches sea level. (About 75% of the latter belong to the region 700 to 1500 m^.) Absorption bands at and above 1.5 n, shown in figures 22.29 and 22.30, are due to water and carbon EMPIRICAL PLANT SPECTRA 093 Fig. 22.25. 400 500 600 700 800 WAVE LENGTH, m^ 900 1000 Transmission spectrum {To/T) of a Chromatium suspension (after Vermeulen, Wassink and Reman 1937). 800 900 WAVE LENGTH, m>. 1000 400 R, bonds of red pigment spirillaxonfhin G, bonds of green pigment boctenoctilorophyll 600 800 WAVE LENGTH, m/i Fig. 22.26. Absorption curve of ex- tracted pigment (broken line) compared with transmission curve of pigment in live cells of Streptococcus varians (strain Cll) (after Frencli 1940). Absorption bands of extracted pigment shifted to let maxima coincide with those of the live cells. Fig. 22.27. Relative absorption curve of the pigments of Spirillum rubrum in living bacteria (measured photoelectric- ally using a scattering control of bleached bacteria) (after French 1937). 694 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 dioxide. Altogether, green plants absorb comparatively little light above 700 m/x — in the region where absorption would probably be useless for photosynthesis, but could cause strong heating. Cell suspension 640 700 WAVE LENGTH, m/x 800 840 Fig. 22.28. Absorption spectra of pigments of green sulfur bacteria in the red and infrared region (after Katz and Wassink 1939). 600 1000 1400 1800 WAVE LENGTH, m/i 2200 2600 Fig. 22.29. Transmission and reflection of near infrared by potato leaf {Sol- anum tuberosum) (after Loomis, Carr and Randall 1941). Note absorption bands due to water, > 1300 m^, but very low absorption (reflection + transmission = 90-100%) in the 800 to 1300 mju region, which contains most of the near infrared rays of sunlight. Figure 22.31 illustrates the "whiteness" of leaves and conifer needles in the near infrared, by a comparison of the reflection of infrared light by leaves with that by a sheet of white paper. According to Mecke and Bald- win (1937) this lack of absorption (rather than infrared fluorescence) causes the striking brightness of vegetation on infrared landscape photo- EMPIRICAX, PLANT SPECTRA 695 60 50 aJ 40 < I- u 30 o a: iij °- 20 / 1 ^-^Reflect 1 on-Bottom ^ \ / n —A/ / / 1 •** L 1^ "^xTransmission I 1 1 / '~^ \ \ /" s V/' k/- 600 1000 1400 1800 WAVE LENGTH, m/i 2200 2600 Fig. 22.30. Transmission and reflection of near infrared by a rubber leaf (ficm elastica) (after Loomis, Carr and Randall 1941). This thick (0.8 mm.) leaf has normal reflection, but shows rather general absorption in the transmission spectrum. Fig. 22.31. Leaves photographed in infrared light on white background (after Loomis, Carr and Randall 1941). 696 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 Fig. 22.31 A. Summer landscape photographed in infrared Hght (after Mecke and Baldwin 1937). graphs (fig. 22.31 A). Figure 22.32 shows the reflection spectra of three leaves of the same species, but of different ages, as given by Shull (1929). Table 22.11 Reflection Transmission and Absorption of Light by Leaves OF Fraxinus excelsior (after Pokrovski 1925) Wave length, m,. 480 500 550 600 620 650 Reflected, % 2.5 4.1 11.5 7.7 6.7 5.4 Transmitted, % 6.4 8.6 17.5 12.4 11.1 9.8 Absorbed, % .91.1 87.3 71.0 79.2 82.2 84.8 The figures of Pokrovski (1925), given in Table 22.11, illustrate the fate of light of different wave lengths falling on a leaf of Fraxinus excelsior. Figure 22.33 shows the reflection spectra of autumnal leaves of different colors (averages for 20-80 different species), as given by Loomis, Carr and Randall (1941). Spohn (1934) observed that the position of the absorption maximum of chlorophyll is shifted in autumnal leaves, from 670 to about 660 niju, probably indicating the liberation of chlorophyll from the pigment-protein-lipide complex present in photosynthetically active cells. CHLOROPHYLL BANDS IN PLANTS 697 Two characteristics strike the eye in all the above-reproduced curves; the shift of band maxima toward longer waves (as compared with their position in solution spectra) , and the diffuse appearance of all bands, leading to considerable absorption (or apparent absorption) in those regions (green and extreme red) where pigment solutions are almost completely trans- parent. Of these two characteristics, the first one can be safely attributed to changes in the intrinsic absorption curves of the pigments in the cell (cf. page 698). The shapes of the bands, on the other hand, are affected largely, but probably not exclusively, by scattering, "sieve effect," and other phenomena of geometrical optics. 40 35- Young J_ _!_ 40 35 30 25 20 15 10 5 - Yellow/ ^' ■^•>, / Orange/ / j< - / 1 / / 1 / ..- /Red k V s<=: / / Gree"n"~ 1 1 1 1 430 460 500 540 580 620 660 700 400 450 500 550 600 650 700 WAVE LENGTH, m/i Fig. 22.32. Reflection (/?//) of leaves of Tilia americana (after Shull 1929). WAVE LENGTH, rufi Fig. 22.33. Average reflection from green and fall-colored leaves (after Loomis, Carr and Randall 1941). 20- 80 species averaged for each curve. 2. Band Maxima of Chlorophyll and Bacteriochlorophyll in the Spectra of Living Plants (a) Red Band of Chlorophyll a Chlorophjdl is responsible for practically all absorption of green plants above 550 m/i {cf. page 719), and the main red absorption peak of chloro- phyll a is recognizable in all plant spectra. Hagenbach in 1870, noticed that this peak is displaced by about 20 m/x toward the infrared compared with its position in ether, alcohol or similar solvents; and Gerland, in 1871, found that a similar shift also affects other, less prominent, chlorophyll bands in the j-ellow and orange part of the spectrum. 698 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 Timiriazev (1872), in an early discussion of the effect of scattering on the spectrum of leaves, suggested that scattering may not only broaden the absorption bands, but also shift their maxima. However, the attribu- tion of the "red shift" to scattering is not permissible, as shown, e. g., by the return of the red band to the position it occupies in solution, after soak- ing the leaves with ether, or immersing them in boiling water (fig. 22.34) (Willstatter and Stoll 1918, Seybold and Egle 1940, Seybold and Weiss- weiler 1942). These treatments do not dissolve the pigments, and do not make the tissues more homogeneous; they merely destroy the association o z < q: 60 - / / / 50 - ; ; Leat iOQked in ether / y -v / 40 / 30 - ' Fresh leaf \ < 20 10 *"«fc // 7^v7 _<^"^Leof soaked in boiling wafer >-/ n . . 1 .... 1 ... . 400 500 600 WAVE LENGTH, m/i Fig. 22.34. Effect of boiling and immersion in ether on transmission spectrum of a Parie- taria leaf (after Seybold and Weissweiler 1942). 700 Q. q: o m < 670 680 690 700 WAVE LENGTH, m>i Fig. 22.35. Absorption spectra of single Protococcus chloroplasts (after Albers and Knorr 1937). of chlorophyll with proteins and lipides. Drying, on the other hand, which changes the scattering of light by leaves to a much larger extent than immersion in hot water, does not affect the position of the red band maximum (Seybold and Egle 1940). The conclusion that the position of the red absorption peak in the liv- ing cell is not affected by scattering, is confirmed by determinations of the absorption spectra of single Euglena cells (Baas-Becking and Ross 1925), and of single chloroplasts (Albers and Knorr 1937). Figure 22.35 shows that the absorption maximum of single Protococcus chloroplasts lies close to 680 m/x, i. €., in the same region as in whole plants. Scattering effects must be weaker than in leaf spectra not only in the spectra of single cells, but also in those of cell suspensions. Several such spectra were reproduced above (cf. figs. 22.21 and 22.22 for Chlorella, CHLOROPHYLL BANDS IN PLANTS 699 Table 22.III Absorption Maxima in the Spectra of Living Cells Position of Plants Observer Type Main red Other of maximum, bands spectrum" X, mtt (approx.) Land Plants Fatsia (fig. 22.13) Meyer (1939) Hibiscus (fig. 22.14) Loomis (1941) Pelargonium (fig. 22.12) Seybold (1932) Tilia (Fig. 22.32) Shull (1929) T T A R 677 665 670 678 Eight spp. (figs. 22.15, 16). Ralideau, French and Holt (1946) A, R 670-680 486, 468, 440 470 580, 480, 440 500-480 460-430 660-690 600-620 588-580 Fifty different spp Lubimenko (1927) Various spp Seybold (1942) T 663-690 Cf. below T 678-684 R 680-684 Aquatic Plants Potornageton (fig. 22.17) Seybold (1934^) 667 440, 420 Algae Chlorella (fig. 21.29) Katz, Wassink (1939) Chlorella (fig. 22.21) Noddack, Eichhoff (1939) Chlorella (fig. 22.22) Emerson, Lewis (1941) Chlorella Wassink, Kersten (1946) Chlorella Seybold (1942) Laminaria (fig. 22.19) Seybold (1933) Nitzschia Button, Manning (1941) Nitzchia Wassink, Kersten (1946) Delesseria (Fig. 22.20) Seybold (1933) Iridaea Van Norman et al. (1948) Chroococcus (fig. 22.23) Emerson, Lewis (1941) A A T T A T T T T T 680 668 672 488, 473, 432 675,625(7)475, 430 680 669 680 675,630(7) 580,470,440 668,605 540-490-^ 675,620'' 550s 480 683,625* 436 Chloroplasts Protococcus, Spirogyra, Zyg- nema (fig. 22.35) Albers, Knorr (1937) 681-684 698, 687, 676, 679, 673, 675, 667, 669 « T = transmission spectrum, R = reflection spectrum, A = absorption spectrum (transmission + reflection). * Phycocyanin. ' Phycoerythrin. 700 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 22.23 for Chroococcus, and 22.24 for Oscillatoria). The "red shift" is recog- nizable in all of them, particularly in the last three figures, in which the spectra of extracted pigments are compared with those of live cells. True. Noddack's figure (fig. 22.21) indicates only a comparatively small shift —from 660 to 668 mju— in live Chlorella cells, but Katz and Wassink (1939) and Seybold and Weissweiler (1942) found, for the wave length of the red band in Chlorella, the value generally given for leaves — about 680 111^. Finally, it was stated on page 653 {cf. figs. 22.15, 22.16 and 21.29) that the red absorption peak is situated at about 675 m/z also in aqueous chloro- plastin extracts. The position of the red absorption peak in live cells is thus determined, beyond doubt, by the state of the pigment (which is pre- served, to a certain extent, in colloidal extracts), and not by geometrical- optical conditions. It has often been suggested that the position of the red absorption maximum of chlorophyll (and the number and position of secondary maxima) varies in different species. Table 22. Ill gives a summary of ex- perimental results. It shows the red peak at 675 mn ± 15 mn, with the average position corresponding to a "red shift" of 15 m^u, or 370 cm.-^ (compared to the position of the corresponding absorption band in ethereal solution of chlorophyll a). The extreme limits of variation of X^ax. in Table 22. Ill are 665 and 690 m^ — quite a wide range. However, because of the diffuse character of the spectra, the position of the maximum often cannot be determined precisely. It is therefore not certain whether any of the tabulated variations in X,„«j are real. (It will be noticed that four values are given for Chlorella, 680, 668, 672 and again 680 m/z!) Lubimenko (1927), who made photographs of the absorption spectra of a large number of leaves, insisted that they do exhibit real differences — not only in the positions of the main peak, but also in the number end positions in the secondary absorption maxima in the yellow, green and blue. According to Lubimenko, the spectra of some species ("group 1") contain eight bands in the visible region. An example is nettle {Urtica dioica), with the following bands: I, 680-660; II, 650-645; III, 630-606; IV, 600-570; V, 550-540; VI, 512-480; VII, 450-430; and VIII, below 420 m^- The bands III, IV and V become visible only when two leaves are used. Leaves of "group 2," which includes the largest number of species, show seven bands (band VII of the above list is missing). Elodea canadensis and Hedera helix belong to this group. "Group 3" (e. g., Prunus laurocerasus) shows only six bands (bands I and II are fused). Finally, plants of "group 4" {e. g., the alga Ulva lactuca) have onW four bands: I, II, V, VI and VIII. The absence of band III in these spectra is particularly remarkable. Table 22. IV shows the limits of variation in the positions of the above-listed bands, as given by Lubimenko, and also gives a tentative identification of these bands with the bands of chlorophylls a and b in solution. According to Lubimenko, in passing from species to species, different bands are displaced by different amounts, or even in different directions. CHLOROPHYLL BANDS IN PLANTS 701 Table 22.IV Absorption Bands in Leaves (after Lxbimenko 1927) I II III V VI VII Band (o) (fe) (a) IV (0,6) (a,6) (0,6) (6.0) VIII Chi {a + h) in ether", X . . . . 669- 648- 619- 599- 571- 547- 506- 467- 439- 415— 655 638 605 570 559 523 489 446 424 Leaves, X from . . . 700- 660- 630- 600- — 555- 510- 450- — 418-^ 680 650 620 570 — 542 490 430 — to 680- 648- 620- 595- — 545- 505- — — 430— 655 638 610 575 — 535 480 — — ° c/. Table 21.1. Photometric curves prove that many of Lubimenko's visually recorded "secondary bands" are only shoulders on the slopes of the main bands or slight ripples on almost uniformly high absorption plateaus (c/., for ex- ample, figs. 22.13 and 22.15). Apparent shifts in the positions of these "maxima," or even the disappearance of some of them, could easily be caused by changes in the ratios of chlorophylls a and h, as well as by varia- tions in scattering. Lubimenko saw in these differences an evidence of variability of chemical composition of "natural chlorophyll" (a name given by him to a hypothetical complex formed by proteins with all the plastid pigments) . Albers and Knorr (1937) found that the number and position of the ab- sorption maxima in single chloroplasts of Protococcus, Spircgyra and Zy- gnema vary not only from species to species, but also from specimen to specimen. In addition to the maxima at 681-683 m^ and 672-675 ui/jl, which may be attributed to chlorophylls a and b, respectively, some chloro- plasts showed secondary maxima at 667-669 m^u, 678-679 m/x and — rather unexpectedly — also in the far red, at 698 niju (c/. fig. 22.35). Albers and Knorr considered these results as indicating variations in the chemical nature of chlorophyll (e. g., oxidations or reductions) that they thought might be associated with the participation of chlorophyll in photo- synthesis (c/. Vol. I, chapter 19). (b) Red Bund of Chlorophyll h The main red absoiption Ijund of chlorophyll b is noticeable, according to Lubimenko (1927), in some spectra of leaves (groups 1 and 2, cf. page 700) but not in others (groups 3 and 4). There is no doubt that this l)and, too, is shifted toward the red from its position in solution (642.5 niju in ether), but the extent of the shift is difficult to measure, because the band appears merely as a hump on the short-wave side of the main chloro- phyll a band. Lubimenko's table (Table 22. IV) gives, for the band axis, 702 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 values from 643 to 655 ihm- This seems to indicate a somewhat lesser shift than is found with the a-component; Table 21. VI gives the impres- sion that the same may be true also for the spectra of the two chlorophylls in different solvents in vitro. The clearest indication that chlorophyll b retains its identity in live cells is provided by the fluorescence spectrum. According to Tables 24.1 and 24.11, the axis of the main fluorescence band of chlorophyll b in vivo lies at about 656 m/x, i. e., only about 5 mn further toward the red than in ether solution (while the axis of the corresponding a band is displaced at least twice as much). This, too, indicates that the absorption band of chlorophyll b in the living cell probably lies at 647-648 mju. (c) Red and Infrared Barids of Bacteriochloi'ophyll It was stated on page 642 that the bands of bacteriochlorophyll are more susceptible to shifts under the influence of the solvent than are those of chlorophyll. We are therefore not astonished to find that the two bands of bacteriochlorophyll, which are situated in methanol solution at about 605 and 770 m/x, respectively (c/. page 017, fig. 21.6), are replaced in live bacteria by bands situated much further in the infrared — at approximately 800 and 850-870 m/x in figs. 21.30, 22.26 and 22.36, and at 825 and 875 m/x in fig. 22.25. Figures 21.30A and 22.26 {cf. also Table 22. V) indicate a third absorption band close to or beyond 900 m/x, while figure 22.27 shows only one, very strong infrared band beyond 900 m/x. Wassink, Katz and Dorrestein (1939) found that the absorption spec- tra of purple bacteria vary from strain to strain, as illustrated by Table 22.V and figure 22.36. Alcoholic extracts from all organisms listed in Table 22. V showed only one absorption maximum in the red (at 774 m/x, cf. fig. 21.30B) in place of the two maxima of most Athiorhodaceae (a weak one at 800 and a sharp one at 875 m/x), and three of some Athiorhodaceae and all Thiorhodaceae (at approximately 800, 850 and 895 m/x). Analysis of the band shapes made it probable that the spectra of all purple bacteria contain three infrared bands, even if one of them may sometime be concealed by the other two. Similar variations in the spectra of different species and strains of purple bacteria were observed by French (1940). The correlation of the "cell bands" with the "solution bands" of bac- teriochlorophyll (fig. 21.6) is not clear. Shall the cell band at 800 m/i be considered as the displaced solution band at 605 m/i, and the cell band at 850-870 m/x as the displaced solution band at 770 m/i? This would mean a shift by as much as 4000 cm.-^ for the "orange" and 1500 cm."'^ for the "red" band. The first shift is so large that one is inclined to doubt the BACTERIOCHLOROPHYLL BANDS IN BACTERIA 703 correlation. An alternative is to consider both cell bands, that at 800 as well as that at 850-870 m/x (and perhaps that at 900 mn, too) as related to the one solution band at 770 m^t. This means "red shifts" varying be- tween 500 and 2000 cm.~^, and implies that bacteriochlorophyll is present, in purple bacteria, in at least three different pigment-bearing complexes. _ strains 740 ENERGY, e.v. .6 1.5 1.4 ""■ ' — I 1 H 1 1 r 1.3 800 2.0 _ strains; ( 1 ) Rhv. I (2) Rhv. 2 0.5 ENERGY, e.v. 1.6 1.5 14 1.3 — I T 1 r- — I 1 r — I 1 r- 900 960 740 800 900 960 WAVE LENGTH, m/x 2.0 _ strains: (1) Phm C\2 (2) Phm.4sot'd r\\ with Oz ^V\ (3) Phm. 5 ^, f 3\\ 1.5 - M y w - 1.0 o 3 1 ^V \ 3^ 0.5 — ENERGY, e.v. 1 0 1.6 , i-p. 1.5 1.4 1 i' 1 1 i' 1 1.3 -1 H-l WAVE LENGTH, m/x Strains: ( 1 ) Rhsp. 1 A 2.0 - (2) Rhsp. 4 (3) Rhsp. 8 A 1.5 2 -A 1.0 3 1 r/y \ ^^^ 0.5 - ENERGY, e.v V 0.2 1.6 1 ' 1 1.5 1.4 r^— , 1 r^-r- 1.3 740 800 WAVE LENGTH, m^ 900 960 740 800 900 960 WAVE LENGTH, m/t Fig. 22.36. Infrared absorption spectra of suspensions of different strains of Athiorho- daceae (after Wassink, Katz and Dorre.stein 1939). The wide variations in the relative intensities of the three absorption peaks (illustrated by the several curves in figure 22.36, and even more strikingly by the curves in figures 21.30A, 21.30B, 22.26 and 22.7) can then be attributed to differences in the relative amounts of the several 704 LIGHT ABSORPTION V,Y PIGMENTS 7.V VIVO f'HAP. 22 Table 22.V Absorption Maxima of Purplk Bacteria (after Wassink, Katz and Dorrestein 1939) m/i 854 S50 . 5 SoG" 852.5 865 1 858 / Species and strain -~ Thwrhodaceae Strain D (Roelefson) 895 Strains 1 , 4, 7,12 (v:ui \i<-l) 895 Strains 9, b, 19 (van Niel-AIuller) 895 Strains 101, 201 Weak Si mill, ;!01, 401 895 AOtiorltoddceae Rhoduvibrio (2 strains) 865 — 864 — Rhodohadllus palustn's (3 strains) 881 ] 873 [ — 862. 5j Phaeonwnas varians (Streptococcus varians) (3 strains) 885 1 885 [ 850 . 5 892.5] Rhodospir ilium rvhrum (2 strains) 875 \ 878 / III 803.5 79f) 796 803.5 802.5 803 802 r799 799 (798.5 800 complexes in the individual species and strains. The complexes may be formed by combination of the same pigment (bacteriochlorophyll) with different proteins, each complex being perhaps specifically adapted to the utilization of one reductant, such as hydrogen, sulfide, thiosulfate or an or- ganic hydrogen donor (this hypothesis was suggested by Wassink, Katz and Dorrestein). Other possibilities include several isomeric or tautomeric forms of bacteriochlorophyll, or small differences in chemical composition, or in the reduction level, of the pigment. If this interpretation of the three cell bands is correct, the question arises as to the reasons for the absence in live cells of a counteipart to the 605 mp solution band of bacteriochlorophyll. No answer can be given to this question — except that the matter requires renewed, and more syste- matic, investigation. It was mentioned in chapter 21 (page 618) that the role of the 605 niM band is not quite clear even in solution spectra. According to Katz and Wassink (1939), live green sulfur bacteria have two absorption maxima (at 740 and 810 m^, respectively) instead of the single band found in bacterioviridin extracts (at 668 m^, cf. fig. 22.28). According to the statement on page 642 the width of the "red shift" is in agreement with the hypothesis that bacterioviridin is a derivative of tetra- hydroporphin (as assumed on page 445 in Vol. I). CAKOTENOID BANDS IN PLANTS 705 {d) Blue-Violet Bands of Chlorophyll The position, in live cells, of the second main band of chlorophyll or bacteriochlorophyll — that situated in the blue-violet part of the spectrum — has received much less stud}^ than that of the red band, the main reason being that the presence of carotenoids and other yellow pigments tends to make absorption in this region very heavy and diffuse. Table 22. VI con- tains some values read from figures reproduced earlier in this chapter. This table indicates a "red shift" by 5 or 10 mju. On the frequency scale, this shift (250-500 cm.~0 is about equal to that of the main red band. Table 22. VI. lii.rE-VioLKx Absorption Maxima in Living Plants Organism ''max.' ™'' Fatsia {rf. fig. 22.13) 438 Chlorella {cf. fig. 22.22) 432 Chroococcus {rf. fig. 22.23) 436 Chlorophyll a in ether 427.5 (c) Protochlorophyll French (1951) estimated, from the action spectrum of chlorophyll formation, that the red absorption band of protochlorophyll lies, in vivo, at 650 m/x. Because of its weakness, it has not yet been directly observed. 3. Absorption Bands of Accessory Pigments in Live Cells Lubimenko (1927) could not find, in the spectra of green leaves, ab- sorption maxima identifiable with the absorption bands of the carotenoids (which could easily be observed, at 510-480 m^u, in the spectra of yellow leaves or yellow parts of variegated leaves) . He pointed out that the first carotenoid maximum may be masked by band VI of chlorophyll (cf. Table 22. IV) and that the second one, at 470-450 mix, should be visible between the chloroph3'll bands VI and VII. Lubimenko concluded that the yellow pigments do not exist "as such" in the green plastids. This conclusion, ob- tained b}^ a purely ciualitative examination of the spectra, is untenable. The spectra of leaf extracts, in which the chlorophylls and the carotenoids certainly exist as separate entities, also do not alwaj's show the individual maxima of all these pigments. In figure 22.47, for example, the absorption curve of the extract from barley leaves shows only three maxima between 400 and 500 m^ — at 410, 428 and 450 niju — instead of the six maxima of the separated green and mellow ])igmeiits. The two main carotenoid 706 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 maxima (at 422 and 467 m^t) and one chlorophyll maximum (at 500 m^) are lost by the superposition of the absorption curves. Upon closer examination, carotenoid maxima can be identified at least in some plant spectra. Peaks, which probably belong to the carotenoids, are noticeable, for example, at 468 and 486 niju in the spectrum of the leaves of Fatsia (Fig. 22.13, cf. also Table 22.III), and at 473 and 488 m/x in that of Chlorella (Fig. 22.22). If one attributes both these peaks to luteol, whose absorption bands in ethereal solution are situated at 442 and 472 m/i (cf. Table 21.1), it follows that the carotenoid bands in live cells are shifted toward the red by as much as 25-30 m/x. (Av = 1250 cm.-^ or three times the shift of the red and blue-violet absorption bands of chloro- phyll.) Emerson and Lewis (1942) postulated, in their interpretation of the Chroococcus spectrum (cf. page 723), a shift of the carotenoid bands in vivo by 14 m/x from their position in ethanolic solution; according to Table 21.1, this corresponds to a shift by 24 m/x relative to ethereal solu- tion, in good agreement with the preceding estimate (25-30 niju) . Menke (1940) extracted chlorophyll from chloroplast preparations (made from spinach leaves; cf. Vol. I, page 369) ; and found that a brick- red residue was left. A suspension of this residue in water showed absorp- tion bands at 490 and 540 m/x — much further toward the red than the caro- tenoid bands have been observed in live green plants (470 and 490 m/x were the positions quoted above for Fatsia leaves and Chlorella cells). Moistening with ether led to a change of color, and a shift of the absorption bands to their usual positions in carotenoid solutions (442 and 472 m/x) . The absorption spectrum of the brown alga Laminaria was found by Menke to exhibit bands— probably due to fucoxanthol— at even longer waves, namely, 499 and 545 m/x- Heating to 70° C. led to a shift of these bands to below 510 m/x, and to a color change from brown to green. The transmission curves of diatoms (Nitzschia dissipata) published by Wassink and Kersten (1946), as well as the absorption curves of brown algae given by Seybold (1934, 1943), clearly show an increased absorption (in comparison to the green algae, such as Ulva or Chlorella) in the region 500- 580 m/x, but give no indication as to the position of the absorption peak (or peaks) of the carotenoid responsible for this absorption. From the comparison of the transmission curves of live diatoms (and of aqueous col- loidal cell extracts, whose brownish color is similar to that of cell suspen- sions) with the transmission curves of the (green) pigment exti-act in an or- ganic solvent (methanol and petroleum ether), Wassink and Kersten esti- mated that the fucoxanthol bands are shifted in vivo by about 20 m/x (corresponding to about 700 cm.^^), toward the longer waves; but this esti- mate is not at all reliable because of the absence of pronounced maxima. These curves, and Karrer's absorption curves of fucoxanthol in solution, make it appear uncertain whether the increased absorption of diatoms and brown algae in the green (500-560 m/x) is due mainly (or exclusively) to a PHYCOBILIN BANDS IN ALGAE 707 strong red shift of the fiicoxanthol band as a whole, or to a broadening of this band toward the longer waves. These observations give some information about the strength with which the carotenoids are bound to the protein-pigment-lipide complex of the chloroplasts. According to equation (21.4) and figure 21.23, the "red shift" can be caused by association of the light-absorbing molecule with other molecules (by adsorption, solution or complexing). The shift is ap- proximately equal (on the energy scale) to the difference between the bind- ing energies of the normal and the excited pigment molecule. In the case of green leaves and algae, we found the chlorophyll bands to be shifted in vivo by about 370 cm.-\ while the luteol bands were shifted by as much as 1250 cm.-i (according to Meyer, and Emerson and Lewis), perhaps even by 2220 and in some cases 2530 cm.-^ (according to Menke). The fuco- xanthol bands in brown Laminaria are, according to Menke, shifted still more widely — by 2585 and 2815 cm.-^ All these shifts are relative to the band position in ether solution. A better idea of the binding energies could be obtained by comparison with extrapolated positions of the bands of isolated pigment molecules. Such an extrapolation was made for chloro- phyll and bacteriochlorophyll on page 642, but it is not yet possible for the carotenoids. The strong red shifts of the carotenoid bands— particularly those of fucoxanthol— indicate clearly that these pigments form part, in chloro- plasts, of some complex, and that the binding becomes particularly strong when the carotenoid molecules are electronically excited. This fact may be relevant for the transfer of electronic excitation energy from the caro- tenoids (particularly fucoxanthol) to chlorophyll, a phenomenon that is revealed by the occurrence of fucoxanthol — sensitized fluorescence of chlorophyll in vivo (cf. chapter 24, page 814)— and that probably explains also the participation of carotenoids in the sensitization of photosynthesis (cf. chapter 30). The absorption peaks of the carotenoids appear especially clear on French's (1937) spectral transmission curves of purple bacteria {cf. fig. 22.27), because in this case the carotenoid bands fall between the two ab- sorption bands of bacteriochlorophyll, at about 600 and 400 m^- Differences in the carotenoid bands of brown and red varieties of Streptococcus arians were described by French in a later paper (1941). The absorption bands of the phycobilins are clearly discernible in the spectra of blue and red algae, e. g., in figure 22.20 at about 550 m/x (phyco- erythrin) and in figure 22.23 in the neighborhood of 625 m/x (phycocyanin) . According to Emerson and Lewis (1942), the phycocyanin maximum is shifted by about 6 m/x toward shorter waves in the aqueous cell extract (figs. 22.23 and 22.48), while the absorption peaks of other pigments retain the positions they had in living cells. This indicates that in the extract, 70S LIGHT ABSOHrTION BY PIGMENTS IM VIVO CHAP. 22 i o l-t El CM O o CO O .< Pi o 03 St o 3 o bB bC bC o 00 10 ITO o o 10 CO O -H o o C5 s O O o tn SI CO C2 00 .OJ CO ' 'i .s -^ in C3 CO CO 'o c 42 g + 2 + a a (H 002 o o.S (M CO <£ bC S •as o ^ P (S CO o o CO bi e^ c>i c- be M bD be (M CO O CI o3 + O CO o CO o CI lO 00 '■r- d --^ -O O Co 00 C4 CI CI CI CO >. o s CO en CO a; C5 C2 !2 .rf 'l£ m a 73 CO ^ o fl s bC f3 03 '53 c o'S. H CO O CO G bp 'a I .3 o p m o cr o 00 o ^ <-^ CO CI d Tf< lo lO CO o o O O O o o CO CI CI 1-H CI cT Cl 00 1-H 00 CI CI CI Cl 0 CI Cl CI CO ^ 2 ^ CI to O A O w -a" a cS O CO C5 1—1 . -H §2 CO 2 ■§ h^ CO M CO > o3 OJ CO > ,. o3 bC bC to > _^ o3 S O •2 St. 2 -*^ ' — a r^ &. a, ti5 tr^ o — ' CO OJ > 03 h-l 00 ci CI .2 03 CO CI CO C _o '-*3 o o3 o3 o CO CO q3 CD CO s 03 II .2 E-i a CC 03 a 03 o O o TRUE ABSORPTION SPECTRUM 709 the cyanobilin-proteid becomes dissociated from the pigment-protein- Upide complex present in the Hve cell. It was stated in chapter 15 (Vol. I, page 399) that, of all plastid pigments, only the phycobilin chromopro- teids pass into true colloidal solution upon extraction with water. To sum up, it is certain that all the pigments found in extracts from plants retain their spectroscopic identity in live cells, despite their prob- able close association in a common complex. The association causes, however, considerable band shifts, and probably also changes the shape of the bands, particularly those of accessory pigments, such as fucoxanthol. 4. True Absorption Spectrum of the Pigment Mixture in the Living Cell In the two preceding sections, we tried to derive as much information as possible from the positions of the absorption maxima in the empirical plant spectra described and reproduced in section 1. It was mentioned re- peatedly that the difference between the "plant spectra" and the spectra of extracted pigments is not limited to band shifts, but includes also changes in the height, width and shape of the individual bands. However, as stated on page G97, the latter changes are, to a large extent, the product of scattering and other geometrical-optical phenomena. In the present sec- tion, we will deal first with a more detailed description of the appearance of the absorption bands in living plant cells, and then with the possibility of deriving from the empirical plant spectra the true absorption curves of the pigment mixtures contained in them. General diffuseness was stated on page 697 to be the most striking char- acteristic of plant spectra as compared to the absorption spectra of the pigment extracts. The ratios of the "optical densities," log (I/S) or log (To/T) in the absorption "peaks" and "valleys" can serve to illustrate this statement. Table 22.Vn shows that the ratio of the optical densities in the "green minimum" and the "red maximum," which, according to Table 21. IB is less than 0.01 in pure chlorophyll a, and about 0.05 in an ether extract from barley leaves (which contains all the chloroplast pigments), is as high as 0.3 in live algae and may reach 0.6 in green leaves. The ratio (violet peak: red peak) also is changed: it declines from 1.75 in ethereal extract, to 1.6 in colloidal aqueous extracts, 1.4 to 1.5 in hve algae, and 0.9 to 1.2 in green leaves. It ^^•ill be noted that the ratios derived from true absorption spectra, log (I/S), are not very different from those derived from the transmission 710 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 spectra, log (Tq/T), obtained by comparison of colored with colorless tis- sues (or, in the case of cell suspensions, by comparison with pure water). Similar data for a variety of green, yellow and red leaves can be found in Seybold and Weissweiler's paper (1943). The question arises: can the leveling off of the absorption peaks and the filling in of the absorption valleys be attributed entirely to geometrical- optical effects (scattering and "sieve effect") or do they indicate a genuine deformation of the absorption curves of the pigments (which could be caused by close packing of pigment molecules in the chloroplasts, as well as by their association with proteins and lipides) ? The fact that the leveling off is much stronger in the spectra of leaves than in those of algae indicates that geometrical-optical effects account for a considerable part of the phenomenon. Tanada's (1951) observations with glycerol (table 22.VII) support this interpretation. Clearly, scat- tering and "sieve effect" must lead to an apparent decrease in the selec- tivity of absorption, as far as plots of log {I/T) or of log (To/T) are con- cerned, since, in these two representations, losses of the weakly absorbed (green and extreme red) light by scattering obviously simulate absorption. One could attempt to explain in this way the results under 4, 5, 7, and 8a- 8d in Table 22.VII, which were obtained by the use of blank cells with pure water. On the other hand, the similarity of results of the experiments listed under 6 and 8, in which the true absorption of Chlorella cell suspensions was determined, with the results listed under 7, in which the transmission of a similar suspension was measured, cannot be explained in this way. In the case of leaf spectra too, it is not immediately obvious why diffuse scattering should lead to transmission curves, log (To/T) (obtained by com- parison of green with white specimens), characterized by high optical den- sity in the regions of weak pigment absorption (green and far red) . If one would assume, for example (c/. Meyer 1939), that the weakening of the trans- mitted beam by passage through a green leaf can be represented by the equation: (22.9) T = I X 10 "('^"^+''^^ where K is a proportionality constant, equivalent to the product (concentration X thick- ness of the absorbing layer) in Beer's law, and a a "scattering coefficient"; and assuming also that the corresponding equation for the white leaf is: (22.9A) To = I X 10~''>^ then the "transmission curve" would be given by the equation: (22.10) log To/T = -Ka-^ In other words, the "transmission curve" would follow faithfully — except for a proportionality factar K — the true absorption curve of the TRUE ABSORPTION SPECTRUM 711 pigments in the cell. Plotted on a semilogarithmic scale (log log [Tq/T] as function of X), the curves with and without scattering would run parallel. If this were the case, the ratios in table 22.VII derived from the transmission curves, log (To/T), would be unaffected by scattering. Closer examination shows why scattering can produce a distortion of the absorption curves in the sense observed. This can be shown with the help of the simple example discussed on page 673 — that of repeated reflections in a plane- parallel absorption cell. The equation usually applied for the calculation of the absorption coefficient, a, is: (22.11) a = (1/d) log i^To/T) where d is the depth of the absorption cell, and T and Tq the fluxes trans- mitted through the solution and the pure solvent, respectively. This equation was sho\^Ti on page 674 to be a first approximation, neglecting the difference in the reflectances of the two cells. As mentioned on page 674, correct expressions for T and To can be obtained by summation of infinite series. This summation leads to the following relationship : (22.12) T = To 10-"^ (i - ,rio-.«.) and thus to : „ = l[,o.^'-,o.(L^i;i5;=^)] Since the term in parentheses is > 1, equation (22.13) shows that the value of a, calculated in the usual way from (22.11), is too large and that the rela- tive error increases with decreasing absorption. With r = 0.1 and To/T = 1.01, i. e., an absorption of only 1%, the error in a is 2%. Such an error can be neglected in most absorption measurements. What matters to us, however, is the fact that the percentage error depends on the value of a and therefore changes with wave length. In other words, it causes a distor- tion of the transmission curve. The character of this distortion — increase of apparent a values for all wave lengths, but particularly for those in which the true absorption coef- ficient is small — remains the same in scattering media, where the effect becomes much stronger than in plane-parallel cells filled with transparent materials. The loss of selectivity by scattering can be even more pronounced in absorption spectra, log (//*S), than in transmission spectra, log (To/T). Radiations that would be only weakly absorbed by straight passage through a leaf (e. g., green or extreme red light) become more strongly absorbed 712 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 when their optical path in the leaf is increased by scattering. On the other hand, a certain proportion of radiations that would be almost completely o n "BLACK" 'white" 2.0,7 'M'M/ ///V ^^v 'i^i ( yy V'A l\ f\\ (\V^V\ \^ \\ w j^\\\^ >\\^ f^ — ..9:; Wi^W// //A 'M )vA Xm r\ y \)^ \v ^\ 1 V\Y ^1/ , /// ) \ y |/ ( ){y J^V\ A \t\ IV'N /\\ \\i y. y. 1 ■\r\\i m <7d = 7C 0 0 0 0 1.8 '. MK^ />V V a )(w X '/^ \y\/ Ai^\ vW'^X ■.'^.A'v; It \\Y\ A/ ■ % 1 1 ord = 6C 1 1 '4/v // Y V (x'/^ \ ){ V \\/ A\\ \\\\^ L \ Vj!\ Ai A \\fy •f\A A' \V 1 A 17 '■^- 22, /]/[ IV, '/// \A/ Ak /\Y\| "/W \u\\ A, ^ w •■''Y' VI 1 1 (j-d = 50 1 I if 0 %W°^^ /I/ aIa iXy ylK (\W \AA Na }\}\ r\v'\\ ■ "t'"\ \6'/. ^7 Y ^/ '^ A/ (y(/ X Y'' (y\ \Xy ^1/^^^ m l( X / aX YS} Ay ,s^\ AV) \\N/\\ M ^^ \ A i i. ^ zji M/ ,,0 nj r^ A' K A ciA )x\^ v\^' ANN' \\\A \^ •A ^VA ^ \. lfo-d=25.0 C '-' ^ '444 A A X y Mv VA) N\A # i^AA V\ X'V A*1 /l 1 1 fffd=20.0 N. 1^ !; ^/ /TO rH ^^ '\K/ \ ,i )(^\A ^oV Nx\ ^\A A 0 '^^ //l^,- yfU / ^ \^ a" X \ \ \X\ VNN vA/ .\vv;2 ^!N ;Xl\ - A 1 O* \ 9 ,'- '^/uT/ Y / V\/ s A,, )/ ^ /'A WY ^^ s\V> A'-/'^ A^X w tc(= 15.0 — 0 '■'^ ' ^ i> '' , \\ AX x<^N >.^ \^ l(\ V-. \ \ \\ t-^ t \ * ' '' /1V L y y K k\ A/ N ^^; h ' y V\ hS Av /\ v/ \ K'' CTd = l2.0 — % k\,\/(] ^ /' vA, 0 \^ sV '-,'- "■ 1 ~i 1 n ^ 4- Vv /* ^u' \/ AN \ A \-/ 7aa= lU.I- dj/V Vl / ;' A A' ^ y y\\ \ \' \ ■ y ■■{^ if tH = 8 (^ 1 LlI HQ 1. Yk'W ;^ / \A N x' ^ / V ^1 ^ \ s\ 4'- \:^6 fad- ^n H °9' ]/ ^J/\ />! / X A ,' K X \' ^\'^ U or : t '1 /'/C. \\' / X < \' ^ : \\ *^ \ \ /A^ S^ to 1 •^ '^"Ia V\ \ ^ / / \' A ^ .^ \ < \ 'A'\ y. T \^^ /" ) ^ n7 .- \y\Jr( '\-^ ^ \ A S ,^ ^ < x' ( \ \ J' A A:0 Zd--4 0 1 1 < 0./ / IP rJr -^ v^ oV ■^ \ ^ ,/ r , / X V ^ X ^ ^> s\ ^o'^ ^ 06^- ip'XP\ ^ X J (^ X / \ ,^ ^ x^ V \ ^ iv. ^1 s ■ vV ■^n l*^^^ ^ ^ \ A "M^ ^ \ k'H K, > \ / ^ . '«• ^ ■, >^' 1 1 n R - Lt y ^ >; f y- \ ^ / \ ^ X ^ > < s, ^, ^x \ \>^ *' Jy \^ ^^ \ )i > ■iN / < > ■\ > •* ^ ^ad'^ZO il- v ^ ^ \ <^ ■> s JxT"' \ / 'N ^ (*'' ^ '" V >! s >j r^ \lj> ?^ -^ k* 0^^ H" <;^ -^ i;;^ x;^ .A ^Li J>. .<^ ^^= 1.0- ^ (V>-.^ V A, A V < ^. >< s > ■^ ^ \ < > ^ ^ n \; .^ >< '- \ »-^ rrf- =0.5 n 1 r \ ^ •^ >■ n -= jL-'fHd = 0 L .. 0 010 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 "TRANSPARENT" REFLECTED LIGHT, ( R' + P')/q' Fig. 22.37. Nomograph for determination of absorption and scattering coefficients in turbid media for diffuse reflection and transmission measurements (after Duntley). (22.14) R' + P' Q' ad sinh Kd (22.15) log — + log Q' = log {ad + ad) sinh Kd + Kd cosh Kd {ad + ad) sinh Kd + Kd cosh Kd Kd (22.16) Kd/K = Vad {ad + 2ad) absorbed if forced to pass straight through the leaf (e. g., blue-violet or red light) will be enabled to escape from it by diffuse reflection, which shortens the optical path in the leaf. Therefore, in a plot of log (I/S) versus X TRUE ABSORPTION SPECTRUM 713 in a scattering medium, we will find that not only are the valleys less deep, but also the peaks are less high than in a similar plot in a nonscattering system. This qualitative explanation of the effect of scattering on selective trans- mission and absorption can be replaced by an exact analysis if the system satisfies certain conditions — namely, random distribution of scattering centers and— in the simplest forms of the theory— equal probability of scattering in all directions. Equations for combined absorption and scattering in systems of this t}T)e have been derived by several authors; we will mention here, as ex- amples, the investigations by Wurmser (1941), Duntley (1942, 1943) and Saunderson (1942). Figure 22.37 represents a nomograph constmcted by Duntley (1942). In this diagram, the abscissae are the expressions: {R' + P')/Q', and the ordinates the expressions: log {l/T) + log Q' , whose relation to the absorption coefficient (per unit path), a, and the scattering coefficient (per unit path), a, is shown in the inserted formulae {d being the depth of the layer measured). The constants P' and Q' are characteristic of the asymmetry of the scattermg. If scattering is iso- tropic, and the incident, transmitted and reflected fight are perfectly dif- fuse, P' = 0 and Q' = 1, and the abscissae in the figure are simply the measured reflectances, while the ordinates are the measured optical densi- ties (logarithms of inverse transmittance) . The nomograph remains ap- proximately correct also if the incident light is collimated, if only the re- flected and transmitted fight are completely diffuse {cf. page 676). Under these conditions, the absorption coefficient, a, and the scattering coefficient, a, can both be read from the graph, if reflectance and transmittance have been determined. Another method of determination of the two coefficients was described by Saunderson (1942). It requires two reflection measurements — one with a thin layer and one with a layer of "infinite thickness" {i. e., having practically negligible transmission). Wurmser (1941) suggested that transmission measurements with two layers of different optical density can be used to determine the coefficients of absorption and scattering, and gave a sample nomograph for two specific depths of the cell (0.035 and 0.1 cm.), which is reproduced in figure 22.38. However, he did not point out that his two-constant formula is correct only if both layers scatter sufficiently to ensure complete diffuseness of the transmitted light. Procedures of the above-described type can be appfied with a fair de- gree of reliance to cell suspensions, and it is desirable that future investiga- tions of spectra of such systems make use of them. The application of Duntley's nomograph to leaves or thalli also is possible, but one has to re- 714 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 member that the statistical theory deals with random distributions of large numbers of small colored particles, and therefore does not take into account the "sieve effect," which may be caused by regular alignment of the com- paratively large chloroplasts. The result of this effect is the admixture 0.8 0.7 0.6 0.5 h > 0.4 en o 0.3 0.2 0.1 O 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 log (Tq/T^)- loglTo/r,) Fig. 22.38. Nomograph for calculation of absorption and scattering coefficients in turbid media from two transmission measurements (after Wurmser 1941). (22.17) To/T = cosh dVaia + a) + „ " , " ^ sinh dV«(a + / "• / /j^^y 7^ - loo/ 1/ i ^ ^ ^ -^ X i^5 1 1 1 I 1 d\ dz 1 -- 0.035 = 0.1 1 1 1a{a + cr) of white light to the transmitted flux, i. e., a decrease in absoption at all wave lengths. In the determination of the apparent absorption coefficient, this decrease must have the largest effect in the absorption peaks, and a lesser effect in the regions of low absorption. It thus helps to make ab- sorption less selective. The sieve effect is negligible in Chlorella suspensions, since the cells in such a sus- pension— unlike the chloroplasts in a leaf — are distributed at random. The suspensions used in the experiments of Noddack and Eichhoff contained about 1 X 10' cells/ml. in a plane-parallel vessel 3.9 cm. thick. The average diameter of a Chlorella cell is about 5 m, and its average cross section is 10 "'^ cm. 2, so that 4 X 10' cells would cover a surface of 1 cm. 2 with eight complete layers. Statistics predict that the probability of a beam's traversing such a suspension without striking a single cell is e~^ = 0.0003 and thus negligible. According to page 683, the number of chloroplasts in a fully green leaf is sufficient to form five or ten continuous layers; the statistical probabil- TRUE ABSORPTION SPECTRUM 715 ity that a beam of light passes between all these chloroplasts, distributed at random, is between 2 X IQ-^ and 5 X IQ-^. Consequently, if the dis- tribution of the chloroplasts were random, the sieve effect would be negli- gible. However, the effect can become important if the alignment of chloroplasts attains a high degree of regularity. The results of Schanderl and Kaempfert (1933) (Table 22.1) point toward a measure of success that nature has achieved in this regulation {cf. fig. 22.5). In very young or thin tissues, the sieve effect is important even without an alignment of the chloroplasts. Meyer (1939) mentions that he was unable to measure the absorption spectra of the seedlings of Tradescantia, and of oat, because "chlorophyll in these seedlings was so granulated that they appeared in transmitted light not green but checkered, consisting of white and dark spots." One way of viewing the sieve effect is to consider the mutual "shading" of molecules in each colored particle, which prevents them from exercising their full absorbing capac- ity; this interference obviously cannot become effective unless the light is markedly weakened by the passage through a single particle. This condition is satisfied in the chloroplasts, since in the peaks of the absorption bands of chlorophyll a single chloro- plast absorbs more than 50% of incident light {cf. page 683). Until quantitative theories have found actual systematic application to cell suspensions, if not to leaves and thalli, the question— what, if any, changes in the true shape of the absorption bands can be deduced from the spectra of live plants— will beg detailed answer. Several qualitative indices that such changes do occur can be noted even now, but none of them is entirely reliable. These indications are: enhanced absorption in the far red and near infrared (to which we referred on page 654), decreased absorption in the maximum of the red band, and the comparatively weak absorption in the blue- violet region. Increased absorption in the far red is exhibited not only by leaves {cf., for example, fig. 22.15) but also by Chlorella and Chroococcus suspensions (figs. 22.21, 22.22 and 22.23) and by aqueous protein-pigment suspensions (fig. 21.28A). In the experiments of Smith (1941), this excess absorption was observed to disappear upon the addition of a detergent, digitonin (compare figure 21.28A with B); he therefore attributed it to scattering. A difference that may be explained in the same way was noted by Rabi- deau, French and Holt (1946) between the transmission and the absorption spectra of chloroplastin dispersed by ultrasonic waves (cf. fig. 22.15). However, as mentioned before, a strongly enhanced absorption in the far red by live Chlorella cells is recognizable also in Noddack and Eich- hoff's figure (fig. 22.21), which (unless the integrating apparatus failed to function as intended) could not be affected by scattering in the way as- sumed by Smith. It is true that, as explained on page 711, scattering, by 716 LIGHT ABSOKrTION BY PIGMENTS IN VIVO CHAr. 22 changing the average length of the hght path in the mediinn, could also change the absorption spectrum (log I/S as function of wave length, which is what fig. 22.21 represents). However, it appears impossible that the light path in the Chlorella suspension could be so lengthened by scattering as to replace the practically complete transparency of the pigment extract at 800 mjLi, by an absorption of over 10%. If Noddack and Eichhoff's re- sults can be relied upon (which is not certain) we arc led to consider the spread of the chlorophjdl absoriition in living cells into the far red and in- frared, a genuine change in the absorption curve of the green pigment. It may be noted (cf. fig. 22.48B) that the absorption curve of Chroococcus (a Cyanophycea) , calculated by superposition of the absorption curves of all the extracted pigments, although it is very close to the transmissi(jn curve of a living cell suspension (probably, because of the absence of chloro- plasts and consequent reduction of scattering), nevertheless also shows enhanced absorption in the far red. Increased absorption in the green, which is so striking a feature of leaf spectra, may or may not indicate a genuine change of the chlorophyll spec- trum. The above-mentioned Chroococcus curve shows no similar effect. In interpreting the absorption in this region, one has to consider, in addi- tion to scattering, also the possible presence in living cells of protochloro- phyll, pheophytin or other relatives of chlorophyll that have absorption bands in the middle of the visible spectrum. The possible extensive spread toward the longer waves of the absorption bands of the carotenoids also has to be taken into account. Decreased absorption by live cells in the maximum of the red band, shown in figures 22.21 and 22.22, cannot be explained by a sieve effect (which is negligible in cell suspensions); diffuse reflection, too, probably is insuf- ficient to account for this effect. A certain flattening of the red band may therefore also be characteristic of the true absorption curve of chlorophyll in the living cell. Table 22.VII shows that (as first noticed by Wurmser in 1921) the ratio of the apparent absorptions in peaks of the blue-violet band and of the red band is low in algae (about 1.5) and particularly in leaves (about 1.0), as compared with pigment extracts (1.75). The blue-violet rays are scat- tered more strongly than red ones; but this could not explain a decrease in their absorption. (Such an effect would only be understandable if the diffuse reflection of blue- violet light by leaves were stronger than that of red light; but no such effect is revealed by figures 22.12 or 22.32.) The comparatively weak absorption of blue-violet light by plants is made even less understandable by the fact that yellow, water-soluble pigments present in leaves must enhance the absorption in this region. However, Seybold and Weissweiler (1943) opposed Noddack and Eich- SPATIAL DISTRIBUTION OF PICMENTS 717 hoff's assertion (fig. 22.21) that live Chlordla cells absorb less light than the pigment extract in the region 550-680 mn; they found the absorption of the cells to be the same as that of extracts in the peaks of the bands, and greater every^^here else. The weaker absorption by living colls, compared with the pigment ex- tract, in the blue and violet region was noticed by Emerson and Lewis (1942) also in the spectrum of the blue-green alga Chroococcus (cf. fig. 22.48B), which is almost free of scattering effects. C. Distribution of Absorbed Energy among Pigments* The allotment of absorbed light energy to the several pigments is very important for the interpietation of the quantum yield of photosynthesis and, in particular, for the understanding of the role of the accessory pig- ments— carotenoids and phycobihns. 1. Effects of Spatial Distribution of Pigments in the Cell The first step in the apportionment of absorbed energy is separation of the absorption by the "photosjTithetic" pigments— chlorophylls, carote- noids and phycobilins — from that by pigments such as the flavones and anthocyanines, which probably bear no relation to photosynthesis at all. This question w^as discussed before (cf. page 685) ; figure 22.10 was given as illustration of the extreme case of leaves of the "purpurea" variety, in which a very considerable part of incident hght, particularly in the green, is absorbed by the water-soluble red pigments. The presence of pigments of this type complicates matters not only by adding new components to the composite absorption spectmm, but also by raising the problem of "color filters": Generally the apportionment of the absorbed light energy to different pigments, in the region of common absorption, requires the knowledge not only of the true absorption curves of the pigments in the state in which they are present in the living cells, but also of their microscopic and submicroscopic distribution. In the case of flavones and anthocj^anines, it is definitely known that their dis- tribution is different from that of chlorophyll— they are concentrated, not in the chloroplasts, but in the cell walls and vacuoles, and thus form "color filters," before or between the chloroplasts {cf. the calculations of Noddack and Eichhoff 1939). This makes it particularly advisable to use, for quan- titative study of photosjTithesis, plants containing as little nonplastid pigments as possible. Even if the flavones and anthocyanines are absent, and the object stud- ied contains no separate carotenoid-bearing bodies, it is by no means certain *Bibliography, page 738. 718 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 that the siibmicroscopic structure of the plastids (discussed in Vol. I, chapter 14, and iUustrated by fig. 14.6) does not place some pigments in a different position with respect to light absorption than others. We may hope that studies with the electron microscope and investigations of the optical properties of the plastids (birefringence, dichroism etc.) will reveal more about the arrangement of molecules in these bodies. Pending these developments, all estimates of the relative contributions of different pig- ments to the absorption of light by plants must be based on the assump- tion of an identical proportional composition of the pigment mixture in every point of the cell or tissue. 2. Apportionment of Absorption in Uniform Mixture If it can be assumed that the pigment mixture in the cell or tissue under investigation has uniform composition — meaning that, wherever the mix- ture is present, it has the same relative composition (but not that the same absolute concentration of mixture is present everjovhere)— the contribu- tion of the zth pigment to the total absorption by the mixture at a given wave length, Ai, is proportional to the product CjOJi, where Ct is the concen- tration and «< the absorption coefficient of this component. (22.18) Ai = A (CiaifSiCiai) Whatever the length and shape of the path of the light beam in the medium, equation (22.18) applies to absorption in every infinitely small element of this path ; therefore, it applies also to the integral light absorp- tion, independently of scattering or other geometrical-optical phenomena. The relative concentrations, Cf, can be determined, e. g., by extraction and photometric estimation; the correct value of the total absorption, A (i. e., the value corrected for reflection and scattering), can be determined, for each wave length, by the methods discussed in part A. The applica- tion of equation (22.18) therefore hinges primarily on the knowledge of the true absorption coefficients of all pigments in the state in which they are present in the cell, and here the difficulty comes in. In part B, we referred to statistical theories whose application may per- mit the determination of the average absorption coefficients of the pigment mixture, from measurements of transmission and reflection (or two reflec- tion measurements, or two transmission measurements with different opti- cal densities). In figures 22.37 and 22.38 we gave examples of nomographs that could be used for this purpose, provided both the transmitted and re- flected light fluxes are perfectly diffuse; and we suggested that these (or other similar theories) be used in the future in the optical study of cell sus- pensions, leaves and thalli. APPORTIONMENT OF ENERGY TO CHLOROPHYLLS 719 Even after the average absorption curve of the pigment mixture has been determined, this will not give us the desired knowledge of the absorption curves of the individual pigments; but it will be a step in the right direc- tion : Some sections of this average absorption curve will be due to a single pigment or a small group of related pigments (e. g., the part above 550 mM in green plants, to chlorophylls a and h, and in brown algae, to chlorophylls a and c). The changes in shapes of the absorption bands, found in these regions, may be considered, by analogy, as valid also for the bands of the same pigments in the regions of composite absorption. (However, the dif- ferent polarizabilities of a molecule in different electronic states, and the possibility of resonance effects between molecules with overlapping absorp- tion bands call for caution in the use of such analogies.) By constructing true absorption curves of cells or plastids with var>ang contents of the individual pigments, one can hope to assemble material whose analysis will permit derivation of the absorption curves of the indi- vidual components. (Here, too, caution will be needed because resonance phenomena may destroy simple additivity of absorption coefficients.) In the light of these considerations, all attempts undertaken so far to apportion the light energy absorbed by plants, among individual pigments, are but first crude approximations. In some of these studies, the analysis was made entirely on the basis of comparison of the spectra of extracts with those of the solutions of separated pigments. In others, a certain improvement was achieved by assuming that all bands were shifted in vivo by the same amount, without change of shape. In a third group of inves- tigations, the analysis was further improved by assuming individual values for the shifts of the bands of different pigments; but still no attempt was made to take into consideration the possible changes in shapes of the bands. 3. Absorption by Chlorophylls a, b, c and d All absorption of light by chlorophyllous plants or plant organs at wave lengths longer than 550 m/x can be attributed to the chlorophylls, except in red and blue algae, where phycobilins may absorb light up to 650 or 700 mfji {cf. figs. 21.39 and 21.40), and purple bacteria, which contain carotenoids with absorption maxima at 550-570 m/x {cj. fig. 22.27 and Table 21.VIII). No serious attempts have been made to apportion the absorption be- tween chlorophylls a and h. A crude idea of this distribution in an ex- tract from a green alga is given by INIontfort's figure (fig. 22.39). (It will be noted that the abscissae in this and other figures of Montfort and Sey- bold are absolute values of absorption in each separate solution, not per- centages of the total absorption by the mixture, and thus do not add to 720 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 100%.) This figure is too crude to reveal an increased participation of chlorophyll b in the region of its band maximum in the red; but shows 450 700 750 Fig. 22.39. 500 550 600 650 WAVE LENGTH, m/x Absorption of pigment extracts from Ulva tacluca (green alga) (after Montfort 1940). 9? o _J _l >- X a. o a: o _i X o > m z g \- a. tr. o CD m < 100 500 600 WAVE LENGTH, m/i Fig. 22. -40. Light absorption by cliloi'ofucin (chlorophyll c), in per cent of total absorption, by a meth- anol extract from diat(jms (after Strain and Manning 1942). 100 700 500 600 WAVE LENGTH, m>i Fig. 22.41. l'roi)ortions of light absorbed hy cliloroi)hylls a and d in a methanol ex- tract of Erythrophyllum delesserioides (after Manning and Strain 1943). Absorption by red and yellow pigments is not considered in the calculation. clearly such an effect in the blue, l^etween 450 and 530 m^i, where the ab- sorption by the 6-component is up to five times stronger than that by chloro- phyll a. ABSORPTION BY CAROTENOIDS IN GREEN PLANTS 721 jMiich more precise measurements were made b}'- Strain and Manning (1942) with the chlorophyll a -\- c mixture extracted from diatoms, and bj- Manning and Strain (19-43) with the chlorophyll a -i- d mixture from red algae. Figiu-es 22.40 and 22.41 show that chlorophyll c (chlorofucin) contributes about 90% of total chlorophyll absorption at 570 mix; while chlorophyll d account.s for 60% of the total chlorophyll absorption at 470 and 90% at 710 m/x. These figures arc for methanol extracts. The relative roles of the chlorophyll components in the absorption of light in vivo are uncertain. In the case of chloroph}-!! h, not even the position of its red absorption peak in vivo is kno^vn with any certainty (c/. page 701). It was suggested (cf. page 612) that alternation of the absorption maxima of chlorophylls a and b in tlie red, orange and 3'ellow, best shown by figures 21.1 A and B, may be nature's means to ensure most effective light absorption throughout this region; but all these bands are so broadened in leaf absorption spectra that it is doubtful whether the absorption by the natural mixture a -\- b is, at any wave length above 550 mn, markedly different from what would prevail if either of the two components were alone present in equivalent concentrations. The situation is different below 550 m^i. The region of prevalent absorption by the 5-component, which we noted in extract at 450-530 mix (fig. 22.39), in all probability exists also in hve green cells. Because of the "red shift" it probabh^ extends in vivo from about 460 to about 540 m/x. In this region, the presence of chloroph^dl b may be of considerable importance from the point of view of enhanced light absorp- tion by green leaves and algae. 4. Absorption by Carotenoids in Green Plants The distribution of light between the chlorophylls and the yellow caro- tenoids of green plants in the region below 550 m/x has been much discussed. The first estimate was made by Warburg and Negelein (1923). In their calculations of the quantum yield of photosj^nthesis (cf. chapter 29), they decided, from extract spectra, that the carotenoids of ChlorcUa accoimt for 30% of the total absorbed light at 436 m/x. Figure 22.42 shows the results of the first more detailed estimate by Seybold (1936), based on spectroscopic measurements with an extract from leaves of Phaseolus vulgaris. Figure 22.39 represented similar data for the multicellular green alga Ulva lacluca; this graph shows the absorp- tion by carotene and the carotenols separately (as well as that by the chloro- phylls a -\- b, a and b). The two figures show the absorption by the caro- tenoids becoming noticeable below 530 m/x (in solvents of low polarizabil- ity) and below about 570 m/x in carbon disulfide. The largest percentage 722 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 participation of the carotenoids in total absorption Is indicated between 450 and 520 m/x in methanol, and up to 550 m/x in carbon disulfide. A considerably more precise analysis was made by Emerson and Lewis (1941, 1942, 1943). They measured the transmission of a Chlorella sus- 400 700 500 600 WAVE LENGTH, m/i Fig. 22.42. Absorption by all pigments {a + b -\- c + x) from 100 cm.* of leaves of Phaseolus vulgaris in methanol, compared with absorption by an extract of the chlorophylls a + 6 in methanol, and by extracts of the caro- tenoids c + X in methanol (1) and carbon disulfide (2) (Seybold 1936). 1.2 1.0 p 0.8 b 0.6 o o _) 0.4 0.2 /s — e I Dtal extracted 1 1 ninmpnt?; in t thonol 1 J \ — — Carotenoids (in ethanol, after saponification and chlorophyll ft \ 1 / \'. \ / ••A ^ — '' ^ 400 440 480 520 560 600 640 680 WAVE LENGTH, m^ Fig. 22.43. Absorption spectra of ethanol solutions of pigments extracted quantitatively from Chlorella cells (after Emerson and Lewis 1943). At wave lengths over 520 m^, the spec- trum of chlorophyll fraction coincides with that of the total extract. o o LJ I- o < O m 80 c 60 Q. tr o IT) CD < O 40 20 r 1 \ 1 -i.- tntnrt rpllc; 1 r " 1 \ Combined extracts 1 elotive absorption by corotenoids J 1 \ \ 1 / vJ 'r \ .... \ ■. J \ \ . _■'— - <^ ^^ ^ - X 400 440 480 520 560 600 640 680 720 X WAVE LENGTH, m/i Fig. 22.44. Comparison of absorption spectra of extracted pigments (shifted as described in text) and of intact Chlorella cells (after Emerson and Le\vis 1943). Curve for intact cells shows absorption due to cell suspension 1.4 cm. thick, con- taining 0.96 mm.^ cells/ml. pension and the spectra of the total pigment extract and of the individual pigments. The results are represented in figures 22.43 and 22.44. The first one refers to conditions in the extract, and shows that, in ethanol, the absorption by the Chlorella carotenoids becomes marked below 520 m^t ABSORPTION BY CAROTENOIDS IN BROWN ALGAE 723 and accounts for about 50% of total absorption between 500 and 450 m^; below 450 m^u, the absorption by the chlorophylls again becomes predom- inant. Figure 22.44 is an attempt to interpret the conditions in living cells. It shows the transmission spectrum, log (To/T), of the cell suspension — it would be better if it were the absorption spectrum, log I/S. Also shown is a composite absorption spectrum of the pigments obtained by shifting the chlorophyll bands, derived from the preceding figure, above 550 m^u, toward the red by 10 m^, and below 580 m^, by G m^ (it would be simpler to make the plot on the frequency scale and use a uniform shift!); the carotenoid bands, derived from the preceding figure, were shifted by 14 m/x. (The values used for chlorophyll may be a little low; cf. page 706.) Two breaks in intact cell curve are at points where cell suspension was stirred and filters in monochromator were changed. The dotted curve shows fraction of total absorbed light absorbed by carotenoids, based on the curve for extracts after introduction of wave length shifts. The difference between the composite absorption curve for total pig- ments and the — much more diffuse — transmission curve of the cell sus- pensions can be due (as discussed in part B) partly to scattering and partly to intrinsic changes in band shapes. A theoretical treatment, as de- scribed on pages 711-714, could eliminate the scattering effect and reveal more clearly the intrinsic alterations of the pigment spectra; but no at- tempt was made to take scattering into account, and the proportion of light absorbed by the carotenoids was calculated simply by comparison of the absorption curve of the combined extracts in figure 22.44 with those of the individual extracts in figure 22.43 (after appropriate "red shifts" of the latter). The results, indicated by the dotted line, show 40% absorp- tion by the carotenoids at 520 m^, a maximum of 75% at 500 mju, a second- ary maximum of about 55% at 460 m/x and a decline to about 25% below 440 niM- According to these results, carotenoids contribute significantly to the total absorption of plastid pigments in Chlorclla from 530 m^t down- ward. 5. Absorption by Carotenoids in Brown Algae Brown algae (including diatoms) contain no chlorophyll b, and should thus, according to page 720, transmit more freely than green plants in the region 460-540 mix. Instead, as their color shows, they absorb consider- able amounts of green light (510-580 mju) — much more than do the green cells. This must be ascribed to the presence of a specific carotenoid, fucoxanthol. If figure 21.35A is correct, and the absorption spectrum of fucoxanthol in solution does not extend toward the red any further than 724 LIGHT ABSORPTION BY PIGMENTS IN VI VU CHAP. 22 that of luteol — i. e., not much beyond 510 niju — the brown color can be attributed to fucoxanthol only by assuming a very wide shift of absorption bands (or their strong broadening toward longer waves). According to page 706, Menke had in fact found indications that the absorption peaks of fucoxanthol in live Laminaria cells are situated as far toward the red as at 499 and 545 mji (instead of 457 and 492 m^i in ethanol solution, according to Table 21. IX). Attempts to analyze the absorption by brown algae, described below, luive not taken into account the possibility of such a wide shift — not to speak of any changes in the width of the bands. On page 657 we referred to a more recent measurement of the extinc- tion curve of fucoxanthol in hexane by Karrer and Wiirgler, which showed a considerably broadened band, extending to or even beyond 530 van (fig. 21.36). 100 80 5 60 Q. < 20 ^~N g, Total pigments -f.x^ f+x, Total carotenols - A 0, Chlorophyll a ""' \ f, Fucoxanthol c, Carotenes - \ N^ N \ 1 y \, "vV- 1 ^ f^ \ N V 9 ^^ \ f \ /^ \ \ w //a \ — o\ \x \\ / - \ \ \\ y^ A - \ \ ^^/ A \ \ 'c, Carotenes I c N, V, \ \ / 0, Ctilorophyll a \ V. \\-' g, Total pigments \ - V" c^'\\ X, Other caretenols\ - \ \V f, Fucoxanthol oloneX 400 500 600 WAVE LENGTH, m/x 700 Fig. 22.45A. Absorption by extracted pig- ments from Fucus vesiculosus (brown alga, extreme sun form) (after Montfort 1940). 400 500 600 700 WAVE LENGTH, m/i Fig. 22.45B. Absorption of e.\- tracted pigments from Laminaria digi- tata (brown alga, 12 m. depth) (after Montfort 1940). Figures 22.45A and B are analyses of the absorption of light by pigment extracts from brown algae, as published by Montfort (1940). They are reproduced here, despite their crudeness, as examples of variations in the relative importance of fucoxanthol as liglit-al)SO)-bing agent in the "helio- philic" surface forms of brown algae, and the "umbrophilic" forms of the same algae that live in considerable depth. In the latter, fucoxanthol ac- counts for most of the extract absorption between 450 and 540 m^u, while in the former its importance is very much smaller. Seybold and Montfort discussed their results in terms of the increase in light absorption in the spectral regions 550-500 and 500-450 m/x caused by the presence of the carotenoids. Table 22. VIII shows that, in extracts from green leaves and green algae, the presence of the carotenoids increases ABSORPTION BY CAROTENOIDS IN BROWN ALGAE 725 this absorption by 40 to 00% in the region 500-550, and by 30% in the region 450-500 mn. In bro^vn algae and diatoms, the absorption increase due to the carotenoids is much larger, from 160% to 400%. This is a consequence of the absence of chlorophyll b and the presence of fucoxanthol ; these two pigments substitute for each other, despite their difference in color. Table 22.VIII Ratio of Absorption by Individual Pigments in Methanol or Benzene and by THE Total Pigment in Methanol" (after Montfort 1940) 550-500 m^ 500-450 iiim X. X. (including f including Organisms Clil. F.) pl> Clil. F.) C. /3 b Diatoms 0.34 0.40^ 3 0.20 0.82 0.24 5 Phaeophijceae Laminaria digitata taken from 12 m. depth 0.38 0.69= 2.6 0.32 0.93 0.33 3.1 Fucus vesiculatus from the surface . 0.24 0.80 4.2 0.22 0.91 0.30 4.5 (0.33)'= (0.47)"= Chlorophyceae Ulvalactuca 0.73'' 0.34 1.37 0.75'' 0.56 0.25 1.33 (0.19)^ Leaves Phaseolus vulgaris^ . — — 1.6 — — 1-3 " The figures in the table do not add to unity because they do not represent the proportions in which the energy absorbed by the pigment mixture is divided between the individual pigments, but rather the ratios (.absorption by the separated pigments): (, absorption by the mixture). Chi. = chlorophyll (in methanol), X. = carotenols (in methanol), F. = fucoxanthol (in methanol), C. = carotene (in benzene). * /3 = factor by which total absorption is increased by the presence of carotenoids. " F. alone. '' Chi. a + b. « Chi. a alone. / According to page 684, the leaves of this species contain a considerable quantity of water soluble yellow pigments. A considerably more detailed analysis of the absorption by diatom pig- ments was made by Button and Manning (1941); the results are repro- duced in figure 22.46. They are based on measurements with acetonic pig- ment extracts from Nitzschia closterium. In constructing the figure, Button and Manning postulated a shift of all absorption bands, those of chlorophyll as well as those of the carotenoids, by 20 m/z. (They based this assumption on the observation that, in the spectrum of this diatom, the maximum of the red chlorophyll band was recognizable at 680 m^; the maximum of the blue-violet band, although not distinct in the cell spectrum 72G LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 itself, could be recognized in the spectrum of a bile salt extract and ap- peared to be shifted by about the same amount— 15 m/x.) We know, how- ever, that the blue-violet band of chlorophyll is shifted by approximately the same amount as the red band on the frequency scale (and not on the wave length scale), and the bands of the carotenoids— particularly those of 1 — [—1 — I — r— 1 — I — I — I — r— r- T— I — I — r— I — 1 — I — I — I I I — I — I I I I I Chlorophyll o Total carotenoids Fucoxanthol Caretenoids other than fucoxanthol 400 500 600 WAVE LENGTH, m/i 700 Fig. 22.46. Distriliution of alisorption in acetone extracts from Nitzschia closterium (after Button and Manning 1941). Lowest curve, chlorophyll a. Middle curve, chlorophyll a + carotenoids other than fucoxanthol. Top curve, all pigments. fucoxanthol— probably are shifted much further than those of chlorophyll. Thus, figure 22.46 must contain considerable errors. The absorption by fucoxanthol above 500 m/x probably is much stronger than shown, with the consequent increase in general absorption in this region. (Figure 22.46 represents a green rather than a brown mixture!) ABSORPTION BY CAROTENOIDS IN BROWN ALGAE 727 A still more detailed analysis of the spectrum of a diatom, Navicula minima, was undertaken by Tanada (1951). By extracting the pigments stepwise with aqueous methanol of different concentrations, he obtained evidence of the following "red shifts" in the blue- violet region (compared to methanolic solution) : chlorophyll a, 8 m^t, chlorophjdl c, 20 m^t; fucox- anthol, 40 m/x; other caroteaoids, 20 m/x. A comparison of the cell spec- trum with the sum of so adjusted pigment spectra can be found in fig. 30.9B. How wary one must be in drawing conclusions based on a uniform shift of the bands of several pigments is illustrated by figure 22.47, taken from Strain (1938): It shows the absorption curves of barley leaf extracts in ethanol and in ether, and the corresponding curves for the separated yellow and green pigments, in the same solvent. In ether, up to 90% of the total absorption in the region 470-500 m^ is due to the yellow pigments, and less than 10% is absorbed by chlorophyll; in alcohol, on the other hand, the green o 400 440 480 520 560 600 640 680 400 440 480 520 560 600 640 680 WAVE LENGTH, m^i WAVE LENGTH, m,i Fig. 22.47. Effect of solvent on absorption spectra of barley leaf pigments (after Strain 1938). a referred to 1 g. fresh leaves in 1 1. solution. I, all ether-soluble pig- ments; II, chlorophyll a and b; III, carotenoids. pigments account for about one half of the total absorption in the same region. The difference is due to the fact that the transition from ether to ethanol causes a stronger shift of the bands of the more polar pigment — chlorophyll — than of those of the less polar carotenoids (11 and 5 m^i versus 1 and 4 m/x, respectively). 728 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 6. Absorption by Carotenoids and Phycobilins in Blue-Green Algae Emerson and Lewis (1943) analyzed absorption by the pigments in the blue-green alga Chroococcus, in the manner described on page 722 for Chlorella. Figure 22.48A illustrates the absorption in alcoholic extract containing chlorophyll and the carotenoids, and in aqueous extract con- taining the phycocyanin-protein complex. The separation of the ab- sorption regions is much neater than in green plants: above 650 m/x, 0.9 0.8 0.7 0.6 ^ 0.5 if g 0.4 0.3 0.2 0.1 1 1 1 \ — « Alcohol-soluble pigments ■•— Chlorophyll — • Carotenoids —a Phycocyanin X Phycocyanin (Svedberg) 0.9 0.8 I \ \ -" Intact cells -« Combined extrocts -« Water extract 0.6 0.5 400 440 480 520 560 600 €40 680 400 440 480 520 560 600 640 680 720 WAVE LENGTH, m/x WAVE LENGTH, m/i A B Fig. 22.48. Analysis of spectrum of Chroococcus (blue alga.) (after Emerson and Lewis 1943). (A) Absorption spectra of extracted pigments in ethanol, and of phyco- cyanin in water. (B) Absorption spectra of intact cells (1.26 mm.' cells/ml., layer 1.4 cm. thick), combined extracts (calculated from A by shifting bands as described in text) and of water extract (containing all pigments). Water extract curve has changed shape, indicating alteration of the phycocyanin-protein complex by extraction. absorption is due mainly to chlorophyll; between 530 and 640 m^c, to phycocyanin; and between 460 and 510 m/x, to the carotenoids (because of the absence of chlorophyll b) . Below 440 m^, chlorophyll again becomes the main absorber. Figure 22.48B is the attempt to reconstruct conditions in the cell. The chlorophyll and carotenoid bands were shifted, as in the treatment of Chlorella, the first ones by 10 m/x in the red, and 6 m/x in the blue and violet, and the second ones by 14 m/x. The phycocyanin band was shifted by 6 m/x compared to its position in aqueous extract. In the case of chloro- phyll and the carotenoids, the extraction was assumed to be complete; in that of phycocyanin, the fact that the 620 m/x absorption peak in the ex- CAROTENOIDS AND PHYCOBILINS IN BLUE-GREEN ALGAE 729 tract was considerably lower than in the cell spectrum was taken as an indi- cation of extraction losses {cf. Vol. I, page 418, about difficulties of quanti- tative extraction of phycobilins). Therefore, in the construction of the ''combined extract" curve in figure 22.48B, the phycocyanin curve of figure 24.48A was not only shifted, but also increased in height by a factor sufficient to make the phycocyanin peak of this curve coincide with that of the cell curve. Emerson and T.ewis pointed out that the agreement between the "cal- culated" absorption curve and the empirical transmission curve (fig. 22.48) is much better than in a corresponding construction for Chlorella. They saw the explanation of this fact in the absence of chloroplasts in Cyano- phyceae, and consequent reduction of scattering. The remaining differ- 100 80 a. (T O CO < 20 ^~ "N. ••,, \ / 1 \ :' \ w Y/ Chlorophyll \ \ i " Carotenoids ! M 'f\ ... \ , ■ \ Phycocyanin / ■ \ , 1 J ■ %_ •X - o on UJ UJ > UJ q: - / . \ /lO" \ / ; ■ r / s./ \ ■ / 3^ ^ - 1 y ■ / "■■•-. ,^ ^ ^ X ■ . ■•■/ / f , jf / / V: .--"7> /^ / / / / > \ . ,.'■ / /> 7^ y -^6^ - zey'' // y / ' / A ■ ,'' / / / . y / ^ / // y" y^' *■ / / 4^ •^ \ ' 2I°--'' / /// 0 ''"^ . y /^^ y ..•■• /' ' / /y -""^ / __ ■ /' //• / ./, «,.^___^ X- 'O'..- ^ Y/ r .---^] ^ ^^ ^: \^ " Fig. 22.50. Energy distribution curves for solar radiation at differ- ent heights of the sun over the hori- zon (after Seybold 1936). For 38°, 21° and 14° curves, black square = 2.50 X 10-3 cal./(cm.2 min.); for 10° to 2 "curves, black square = 0.25 X 10-3 cal./(cm.2 min.). If) iDO O O OO O If) O O U) tOif) (J) OJ iDh- O in OD — ro ^"^ t m if)io ID IS) IS) ^ WAVE LENGTH, m;i Plants live in "light fields" whose normal character depends on the clima- tic zone and habitat, and which are also subject to variations with the season of the year, the time of day and the meteorological conditions. Three characteristics of natural light fields are important to plants : their total intensity, their spectral composition and their periodicity. There are strong indications that plants adapt their life processes in general, and their photosynthetic apparatus in particular, to all these three factors. The intensity adaptation reveals itself in the different characters of "umbro- philic" and ''umbrophobic" (or "heliophilic") plants (shade and sun plants); chromatic adaptation is most strikingly shown by the occurrence of red algae Bibliography, page 738. NATURAL LIGHT FIELDS 731 in the depths of the sea ; and the periodicity adaptation shows itself in the distinction between long-day plants of the arctic zone and short-day plants of the temperate and tropical zones. The spectral composition and intensities of light fields in different habi- tats of plants have been measured and discussed by several authors, for example, Wiesner (1907), Ursprung (1918), Seybold (1934, 1936) and Egle (1937). The factors that determine the total intensity of the available light (in the photosynthetically important region 400-700 m/x) are the height of the sun over the horizon, the clearness or haziness of the air, cloudiness and the position of the plant in direct sunlight or in the shade. < Q. UJ > I- < _i LJ q: Fig. 22.51. sun ov 20 30 40 HEIGHT OF THE SUN, degrees Total irradiation and relative intensity of sunlight in relation to height of or the horizon (after Seybold 1936). Inner scale at left, cal./(cm.- miu.). Reflection by the surrounding surfaces may also be of importance, particu- larly for the evergreens (reflection by snow!) and shore plants (reflection by water!) (cf. Egle 1937). Figure 22.50 shows the intensity distribution of the combined light of the sun and the sky for different heights of the Sim over the horizon. The curves show the transition from the red light of the setting sim (ma.ximum intensity at 680 m/x) to the yellow light of the sun in the zenith (maximum intensity at 520 m/i). Figure 22.51 shows the increase in the total energy of the light, from 0.1 cal./cm.^ min., at 10° elevation, to 0.6 cal./cm.^ min., at an elevation of 50°. (The figures are for a surface normal to the direction of the incident light; for a horizontal surface the variations are much wider.) When the sun is obscured by light clouds, the total light intensity decreases to 10 or 20% of its full value (i. e., to 0.1-0.2 cal./cm.- min. at midday). When the sky is entirely overcast. 732 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 the light intensity can drop to as low as 0.02 cal./cm.'- min., or even less, i. e., to only 1% of full sunshine. At the same time, the spectral composi- tion changes in the way characterized by figure 22.52, taken from a paper by Taylor and Kerr (1941); the light intensity becomes almost uniform throughout the visible spectrum, and the common designation of such days as "gray" proves to be correct. 400 440 480 520 560 600 640 680 WAVE LENGTH, m/i Fig. 22.52. Average energy distribution curves for daylight (after Taylor and Kerr 1941). (A) Zenith sky, color temperature 13,700° K.; (B) north sky on 45° plane, color temperature 10,000° K.; (C) totally overcast sky, color temperature 6,500° K. ; (D) sun plus sky on horizontal plane, color temperature 6000 ° K. ; (E) direct sunlight, color temperature 5335° K. 400 500 600 700 WAVE LENGTH, m/i Fig. 22.53. Energy distribu- tion in the shade (after Seybold 1936). (G) Edge of wood (black square = 2.5 X 10 "^ cal. per (cm.2 min.); (1), (2), (3) three shade habitats of Oxalis (blaf;k square = 0.025 X 10 "^ cal. per (cm.^ min.). When the plant is in the shadow of a rock, house or mountain, and re- ceives light mainly from the blue sky ("blue shade"), the spectral composi- tion of its light field is entirely different from that to which it is exposed in direct sunlight, as shown by figure 22.52. The intensity of radiation from a clear blue sky is of the order of 20% of that of full sunlight (i. e., about 0.1 cal./cm.2 jj^jn ) at sea level. It decreases mth increasing altitude, as the scattering air layer above becomes thinner and the color of the sky a deeper blue. The plants that live in the shadow of other plants, e. g., the floor vege- tation in the forest, receive their light filtered through the chlorophyll layers NATURAL LIGHT FIELDS 733 of the overhanging foliage. They Hve in the "green shade." Figure 22.53, taken from Seybold (1936), shows the spectral compositions of the light field in the midst of a forest, compared with that at the edge. Character- istic is the minimum at 650 m/x, clearly corresponding to the absorption maximum of chlorophyll. A large part of radiations reaching the floor of a forest are either infrared or deep red, scarcely visible to the eye and use- less for photosynthesis. The total intensity of the light field under the trees (400-700 m^u) is less than 10% of that of full sunlight above the forest and can drop to as low as 1% in a dense pine forest (Seybold). An even stronger alteration in the intensity and spectral composition of the light field occurs when sun rays pass through thick layers of water. Table 22. IX gives some data on the decrease in total intensity with depth. The main cause for this drop in light intensity is the absorption by water Table 22.IX Decrease in Light Intensity with Depth 0 50 100 200 ft. Atlantic Ocean Brightness" 1076 114 37 4.4 0 1 2 5 10 20 m. Titisee Lake Intensity-* 100 57 32 21 13 10 Bodensee Lake Intensity*. 100 54 30 15 9 6 « Beebe and HoUister (1930), Hulburt (1932). '•Seybold (1936). itself. Some higher bands (overtones) of the vibrational spectrum of water lie in the visible spectrum (the fundamental frequencies are in the near infrared). They decrease in intensity from red to blue; in the violet and near ultraviolet, however, the absorption increases again, probably due to weak electronic bands. Figure 22.54 shows the extinction curve of water in the region 360-800 m^u, according to the measurements of Aschkinass (1895) and Sawyer (1931) (c/. Dorsey 1940). A water layer 10 m. thick reduces the light intensity at 640 m/x by the factor of 10; a layer 1 m. thick does the same at 760 mju ; 100 m. are required for a similar reduction at 440 m^. Because of this absorption in red, yellow and violet, thick layers of pure water are bluish green in color. The absorption by natural waters in the blue and in the violet is usually much larger than that by pure water, due partly to the presence of certain inorganic ions (e. g., iron) and partly to that of organic matter (e. g., the chlorophyll of the phytoplankton). This increased absorption at the short-wave end of the spectrum gives natural waters a pure green or even 734 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 a yellowish-green tinge, as compared with the bluish-green color of water as such. Figure 22.55 shows the spectral composition of light in different depths of the Mediterranean on a clear day according to Seybold (1934). 100 360 680 720 440 520 600 WAVE LENGTH, m^ Fig. 22.54. Absorption curve of pure water (a = specific absorption coefficient). S, Sawyer (1931). A, Aschkinass (1895). 400 500 600 700 WAVE LENGTH, m/i Fig. 22.55. Energy distribution under 1-50 m. water (after Seybold 1934). Table 22.X Spectral Composition of Light under Water" Italics indicate position of maximum of transmission. A. transmission by a five meter thick layer* IXlfl Water 700 650 600 550 500 450 400 B. SPECTRAL COMPOSITION OF LIGHT IN DIFFERENT DEPTHS'' Depth Per cent 350 Pure water (blue) 7 27 44 83 90 95 79 62 Danish Sea (blue-green) 5 15 27 37 41 18 7 5 Pudget Sound (green) 3 10 18 33 30 28 17 8 Lake Mendota (green) 0 3 8 8 6 4 2 0 Red Orange Yellow Green Blue Violet 0 28 16 15 13 18 16 1 18 20 21 12 15 14 3 9 17 27 19 16 12 5 5.5 17.5 36 22 12.5 6.5 7 0 13 46 27 10 4 9 0 6 52 29 10 3 " For more information, see e. g., Schmidt (1908), Knudsen (1922), Hulburt (1928), Atkins (1932) and the book by Dorsey (1940). * Shelford (1929). ' Trout Lake, Wis- consin, according to Manning, Juday and Wolf (1938). NATURAL LIGHT FIELDS 735 Similar sets of curves, showing a less rapid decline in intensity at the red end of the spectrum, were given by Johnson and Kullenberg (1946) and Levring (1947) for sea water at the West Coast of Sweden. Levring at- tributed the absorption in violet and blue — which often converts the blue- green color of pure water into a yellow-green — to the presence of suspended particles. Some additional data can be found in Table 22.X. Algae are found to a depth of 120 m. These deep water species live in a light field that contains almost exclusively green radiation (practically no red light is available below 10-20 m.). The total intensity of light avail- able to them is only a few per cent of the light enjoyed by the species living close to the surface of the sea. 400 500 600 700 WAVE LENGTH, m/i 500 600 700 WAVE LENGTH, mM Fig. 22.56. Light absorption by algae in different depths (in meters) in per cent of light incident on the surface (after Seybold 1934). We have discussed in chapter 15 (Vol. I) the ways in which plants adapt themselves to the chromatic composition of the light fields. There is no doubt that the brown algae, containing fucoxanthol, are capable of absorb- ing more light in the middle of the visible spectrum than the green plants, and that the presence of phycobilins in the Rhodophyceae and Ctjanophyceae increases their absorbing capacity in this spectral region even more strongly. Table 22.VIII showed what these changes in the pigment system mean for the absorption capacity of the plants in different spectral regions. However, this table referred to illumination with light whose intensity is constant throughout the spectrum. The consequences of the same changes for the absorption of natural light were discussed by Seybold (1934), Mont- fort (1934) and Levring (1947), and we must refer to their papers for de- tailed results. We merely mention, as an example, that, according to Sey- bold, a Rhodophycea will absorb up to 95% of the light available in a 50 ra. 736 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 depth, where a Chlorophycea will be able to absorb only 30^0% (cf. fig. 22.56). Bibliography to Chapter 22 Light Absorption by Pigments in the Living Cells A. Light Absorption by Plants 1856 Bohm, J. A., Sitzber. Akad. Wiss. Wien, Math, naturw. Klasse, 22, 479. 1861 Sachs, J., ibid., 43, 265. 1880 Stahl, E., Botan. Z., 38, 297, 321, 345, 361, 377, 393, 409, 868. 1883 Reinke, J., Ber. deut. botan. Ges., 1, 395. 1885 Timiriazev, C, Cornpt. rend., 100, 851. 1886 Reinke, J., ibid., 44, 161, 177, 193, 209, 225, 241. 1888 Detlefson, E., Arbeit, des botanischen Instituts, Wilrzburg, 3, 534. 1901 Linsbauer, L., Botan. Zentr., Beihefte, 10, 53. 1905 Brown, H., and Escombe, F., Proc. Roy. Soc. London, B76, 29. 1908 Senn, G., Die Gestalts- und Lageveranderungen der Chromatophore, Engel- mann, Leipzig, 1908. 1909 Senn, G., Ber. deut. botan. Ges., 27, ( 12.) Stahl, E., Zur Biologie des Chlorophylls. Fischer, Jena, 1909. 1911 Weigert, F., Die chemischen Wirkungen des Lichts. Enke, Stuttgart, 1911. 1912 Coblentz, W. W., Bull. U. S. Bureau Stand., 9, 283. 1914 Purevich, K. (Purjewitsch), Jahrb. wiss. Botan., 53, 210. 1917 Senn, G., Verhandl. naturforsch. Ges. Basel, 28, 104. 1918 Willstatter, R., and Stoll, A., Untersuchungen uber die Assimilation der Kohlensdure. Springer, BerHn, 1918, p. 122 ff. TJrsprung, A., Ber. deut. botan. Ges., 36, 122. 1919 Senn, G., Z. Botan., 11, 81. 1921 Wurmser, R., Arch. phys. bioL, 1, 33. 1922 Liese, W., Beitr. allgeni. Botan., 2, 323. 1924 Lazarev, P. P., /. Applied Phys. USSR, 1, 142. Benecke, W., and Jost, L., Pflanzenphysiologie, Vol. 1, Jena, 1924. 1925 Pokrovski, G. I., Biochem. Z., 165, 420. Warburg, 0., ibid., 166, 386. 1927 Lazarev, P. P., ibid., 182, 131. 1929 Shull, C. A., Botan. Gaz., 87, 583. Briggs, G. E., Proc. Roy. Soc. London, B105, 1. 1932 Seybold, A., Planta, 16, 195. Seybold, A., ibid., 18, 479. 1933 Schanderl, H., and Kaempfert, W., ibid., 18, 700. Seybold, A., ibid., 20, 577. Voerkel, S. H., ibid., 21, 156. Seybold, A., ibid., 21, 251. 1934 Seybold, A., Jahrb. wiss. Botan., t9, 593. 1935 Mestre, H., Cold Spring Harbor Symposia Quant. Biol., 3, 191. BIBLIOGRAPHY TO CHAPTER 22 737 1937 Mecke, R., and Baldwin, W. C. G., Naturwissenschaften, 25, 305. Egle, K., Planta, 26, 546. 1939 Noddack, W., and Eichhoff, H. J., Z. physik. Chem., A185, 241. Meyer, K. P., Helv. Phys. Acta, 12, 349. 1941 Loomis, W., Carr, and Randall, H. M., Gibson Island A. A. A. S. Sym- posium on Photosynthesis (unpublished). 1942 Burns, G. R., Am. J. Botany, 29, 381. Seybold, A., and Weissweiler, A., Botan. Archiv, 43, 252. 1943 Seybold, A., and Weissweiler, A., Botan. Arch., 44, 102,456. 1946 Rabideau, G. S., French, C. S., and Holt, A. S., Am. J. Botany, 33, 769. 1947 Looniis, W., A.A.A.S. Symposium on Photosynthesis, Chicago (unpub- lished). 1948 Kok, B., Enzymologia, 13, 1. 1949 Loomis, W., Photosynthesis in Plants, Iowa State College Press, Ames, 1949, p. 5. B. Spectral Properties of Plants 1870 Hagenbach, A., Ann. Phys. (Poggendorf), 141, 245. 1871 Gerland, E., ibid., 143, 585. 1872 Timiriazev, C, Ber. deut. chem. Ges., 5, 329. 1884 Engelmann, T. W., Botan. Z., 42, 81, 97. 1886 Reinke, J., ibid., 44, 161, 177, 193, 209, 225, 241. 1887 Engelmann, T. W., ibid., 45, 393, 409, 425, 441, 457. 1904 Gaidukov, N., Ber. deut. botan. Ges., 22, 23. Gaidukov, N., Hedwigia, 43, 96. 1907 Ivanovski, D. (Iwanowski), Ber. deut. botan. Ges., 25, 416. 1913 Ivanovski, D. (Iwanowski), Biochem. Z., 48, 328. 1918 Ursprung, A., Ber. deut. botan. Ges., 36, 86, 111. Willstatter, R., and Stoll A., Untersuchungen ilber die Assimilation der Kohlensdure. Springer, Berlin, 1918. 1921 Wurmser, R., Arch. phys. biol., 1, 33. 1925 Pokrovski, G. I., Biochem. Z., 165, 420. Baas-Becking, L. G. M., and Ross, P. A., /. Gen. Physiol, 9, 111. 1926 Wurmser, R., /. phys. Radium, 7, 33. 1927 Lubimenko, V. N., Rev. gen. botan., 39, 547. 1929 ShuU, C. A., Botan. Gaz., 87, 583. 1932 Seybold, A., Planta, 16, 195. Seybold, A., ibid., 18, 479. 1933 Schanderl, H., and Kaempfert, W., ibid., 18, 700. Seybold, A., Planta, 21, 251. 1934 Seybold, A., Jahrb. tuiss. Botan., 79, 593. Spohn, U., P/anfa, 23, 241. 1936 Seybold, A., Jahrb. loiss. Botan., 82, 741. 1937 Albers, V. M., and Knorr, H. V., Plant Physiol, 12, 833. Vermeulen, D., Wassink, E. C, and Reman, G. H., Enzymologia, 4, 254, 738 LIGHT ABSORPTION BY PIGMENTS IN VIVO CHAP. 22 Egle, K., Planta, 26, 546. French, C. S., /. Gen. Physiol, 21, 71. Mecke, R., and Baldwin, W. C. G., Naturwissenschaften, 25, 305. 1938 Strain, H. H., Leaf Xanthophylls, Carnegie Inst. Wash. Pub. No. 490. 1939 Katz, E., and Wassink, E. C, Enzymologia, 7, 108. Wassink, E. C, Katz, E., and Dorrestein, R., ibid., 7, 113. Noddack, W., and Eichhoff, H. J., Z. physik. Chem., A185, 241. Meyer, K. P., Helv. Phys. Acta, 12, 349. 1940 French, C. S., J. Gen. Physiol., 23, 483. Menke, W., Naturwissenschaften, 28, 31. Seybold, A., and Egle, K., Botan. Arch., 41, 578. 1941 Smith, E. L., /. Gen. Physiol, 24, 565. Zscheile, F. P., and Comar, C. L., Botan. Gaz., 102, 463. Emerson, R., and Lewis, C. M., Gibson Island A. A. A. S. Symposium on Photosynthesis (unpublished) . Button, H. J., and Manning, W. M., Am. J. Botany, 28, 516. Loomis, W., Carr, and Randall, H. M., Gibson Island A. A. A. S. Sym- posium on Photosynthesis (unpublished). Wurmser, R., Rev. brasil biol, 1, 325. 1942 Emerson, R., and Lewis, C. M., J. Gen Physiol, 25, 579. Duntley, S. C, /. Optical Soc. Am., 32, 61. Saunderson, J. L., J. Opt. Soc. Am., 32, 727. Seybold, A., and Weissweiler, A., Botan. Arch., 43, 252. 1943 Seybold, A., and Weissweiler, A., ibid., 44, 102. Seybold, A., and Weissweiler, A., ibid., 44, 456. Duntley, S. C, J. Opt Soc. Am., 33, 252. 1946 Rabideau, G. S., French, C. S., and Holt, A. S., ibid., 33, 769. Wassink, E. C, and Kersten, J. A. H., Enzymologia, 12, 3. 1947 Iljina, A. A., Zhurnal fys. Khimii, 21, 145. 1948 Van Norman, R. W., French, C. S.. and Macdovvall, F. D. H., Plaid Physiol, 23, 455. 1949 Loomis. W.. Photosynthesis in Plants. Iowa State College Press. Ames, 1949, p. 5. 1951 Tanada, T., Am. J. Botany, 38, 276. French, C. S., and Koski, V. M., Proc. Soc. Exptl Biol, (in press). C. Distribution of Absorbed Energy among Pigments 1923 Warburg, 0., and Negelein, E., Z. physik. Chem., 106, 191. 1936 Seybold, A., Jarhb. iviss. Botan., 82, 741. 1938 Strain, H. H., Leaf Xanthophylls. Carnegie Inst. Wash. Publ. No. 490. 1939 Noddack, W., and Eichhoff, H. .J., Z. physik. Chem., A185, 241. 1940 Montfort, C, ibid., A186, 57. 1941 Button, H. J., and Manning, W. M., Am. J. Botany, 28, 516. Emerson, R., and Lewis, C. M., Gibson Island A. A. A. S. Symposium on Photosynthesis (unpublished). BIBLIOGRAPHY TO CHAPTER 22 739 1942 Strain, H. H., and Manning, W. M., /. Biol Chem., 144, 625. Emerson, R., and Lewis, C. M., /. Ge7i. Physiol., 25, 579. 1943 Emerson, R., and Lewis, C. M., Am. J. Botany, 30, 165. Manning, W. M., and Strain, H. H. J. Biol. Chem., 151, 1. 1951 Tanada, T., .4//;. J. Botany, 38, 276. Appendix. Natural Light Fields 1895 Aschkinass, E., Ann. Phys. (Poggcndorf) , 55, 40L 1907 Wiesner, J., Der Lichtgenuss der Pflanzen. Engelmann, Leipzig. 1908 Schmidt, W., Sitzber. Akad. Wiss. Wien Math, naturw. Klasse, Aht., lid, 117, 237. 1918 L^rsprung, A., Ber. dent, botan. Ges., 36, 111. 1922 Knudsen, M., Conseil permanent intern, exploration mer, Publ. No. 76. 1928 Hulburt, E. 0., /. Opt. Sac. Am., 17, 15. 1929 Shelford, V. E., in E. Abderhalden, Handbuch der biologischen Arbeits- methoden. Part 2, 9, p. 1495. 1930 Beebe, W., and Hollister, G., Bull. N. Y. Zool. Soc, 33, 249. 1931 Sawyer, W. R., Contrib. Canad. Biol. Fisher., 7, 75. 1932 Hulburt, E. O., /. Optical Soc. Am., 22, 408. Atkins, W. R. G., /. conseil permanent intern, exploration mer, 7, 171. 1934 Seybold, A., Jahrb. iviss. Botan., 79, 593. Montfort, C., ibid, 79, 493. 1936 Seybold, A., ibid., 82, 741. 1937 Egle, K., Planta, 26, 546. 1938 Manning, W. M., Juday, C, and Wolf, M., /. Am. Chem. Soc, 60, 274. 1940 Dorsey, N. E., Properties of Ordinary Water Substance. Reinhold, New York, 1940. 1941 Taylor, A. H., and Kerr, G. A., J. Optical Soc. Am., 31, 3. 1946 Johnson, N. G., and Kullenberg, B., Svensk. hydr.-biol. Komm. Skr., Ill, 1, No. 1. 1947 Le\Ting, T., Goteborgs Kungl. Vetenskaps. Vitterhets Samh. Handl. (VHB), 5, No. 6. Chapter 23 FLUORESCENCE OF PIGMENTS IN VITRO Fluorescence phenomena have two aspects. In the first place, the fluor- escence spectrum offers a welcome addition to the absorption spectrum in the study of the term system and molecular structure of a chemical com- pound. In the second place, the yield and duration of fluorescence gives significant information as to the fate of the excitation energy and thus provide clues to the mechanism of photochemical reactions of the light- absorbing compound. In the case of chlorophyll, we are particularly in- terested in the second, photochemical aspect of the fluorescence phenomena. Because of the division of this treatise into a chemical and a physical part, the photochemistry of chlorophyll already was dealt with in the first volume (chapters 18 and 19), while fluorescence could first be discussed in the present, second vohune. Certain conclusions derived from fluores- cence studies had to be anticipated in Volume I ; some of them will have to be repeated and amplified here. Some new facts and considerations have been added to this field since the appearance of Volume I ; their fluorescence aspects are discussed in the present chapter, and their photochemical aspects in chapter 35. A. Fluorescence of Chlorophyll in Vitro* 1. Fluorescence Spectra of Chlorophyll and Its Derivatives in Solution The fluorescence of chlorophyll was discovered by Brewster more than a hundred years ago (1834). It was first studied spectroscopically by Stokes in 1852, and a by-product of this study was the discovery that the leaf pigment consists of two green and two yellow components. Dher^ (1914) and Wilschke (1914) contributed the first photographs of the fluorescence spectrum. However, because the fluorescence bands of chlorophyll are situated in the far red and infrared, and the sensitivity of red- and infrared-sensitized plates varies strongly with wave length, the photographic method is not very suitable for the quantitative study of chlorophyll fluorescence. * Bibliography, page 801. 740 FLUORESCENCE SPECTRUM OF CHLOROPHYLL 741 Figure 23.1 A shows the appearance of the fluorescence spectrum of chlorophyll on a panchromatic plate, and figure 23. IB, on an infrared- Wave length, m/i Reference spectrum (He + Hg) ' Absorption Chloroptiyll a Fluorescence Reference spectrunn (He + Hg) ' Absorption Chlorophyll a ■ Fluorescence with self- absorption Exciting light ( \ < 470 m/i) Calibration spectrum (Nernst lamp) Reference spectrum (He + Hg) oo o o inoui o r-- 1^ lO lO O in iT) O o O ID o o Wave length, m^ Reference spectrum (He + Hg ) Absorption Chlorophyll a + b Fluorescence ■ Chlorophyll o (Fluorescence) Reference spectrum (He + Hg) Chlorophyll ( Absorption a + b I Fluorescence Chlorophyll 0 (Fl. with self-obs.) Exciting light ( X< 470 m/i) Calibration spectrum (Nernst lamp) Reference spectrum (He + Hg) oo o o o ino >n O >n o O m o m 1 O O r " X 1 7iri|iiii II 1 1| 1 to a> f-- o X X OJ 1 1 I Fig. 23.1 A. Fluorescence spectra of chloroj)!!}'!! (a at top and a + b below) in ether on panchromatic plates (after Dhere and Fontaine, 1931). sensitive plate. Fig. 23.2 shows spectrophotometric curves of the fluores- cence of chlorophylls a and 6 in ether, determined by Zscheile and Harris 742 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 (1943) by means of a photoelectric spectrophotometer. The maximum of the first emission band Hes, in this figure, at 664.5 mn for chlorophyll a and CVJ>0 00 «) aou>K C>JO<0 OD in •• • ^"^^^ Chlorophyll a in ether excitation by Hg line 365 m/i) Chlorophyll a Chlorophyll 6 Chlorophyll a Chlorophyll b dissolved in carbon disulfide (excited by carbon arc) ojirtcD (0 oo«3r- 1^ (MOID 00 l^f^ (O If) to o Wove length, A. Helium Etioporphyrin Phylloerythrin Pheophorbide a Pheophorbide b Chlorophyll a Chlorophyll b Helium pyridine ethyl ether Fig. 23. IB. Fluorescence spectra of chlorophyll and some related compounds on infrared-sensitive plates (after Dhere and Raffy 1935). Top: (2) Chlorophyll a in ether (excited by X36-5 m/x); (4) and (7) chlorophyll a in CS2 (e.xcited by car- bon arc); (5) and (8) chlorophyll 6 in CSo (excited by carbon arc). Bottom: (2) etioporphyrin in pyridine; (3) phylloerj^hrin in pyi-idine; (4) pheophorbide a in ether; (5) pheophorbide b in ether; (6) chlorophyll a in ether; (7) chlorophyll b in ether. 648.5 m/i for chlorophyll 6— only very slightly on the long-wave side of the maxima of the corresponding absorption bands (c/. Table 23. IC). The FLUORESCENCE SPECTRUM OF CHLOROPHYLL 743 first fluorescence band is followed bj^ a second one, of lower intensity, situ- ated in the far red, which is visible in figure 23. IB as well as in figure 23.2; and by a third, weak one, situated in the near infrared, which is not re- corded in these figures. In an earlier investigation of Zscheile (1935), an additional band was observed in the fluorescence spectrum of chlorophyll b at 672.8 m^; but Dhere and Biermacher (1936) and Biermacher (1936) ascribed it to contamination with chlorophyll a, and this ex- planation was accepted by Zscheile and Harris (1943). According to Biermacher (1936), the fluorescence spectrum provides the most sensi-* five test for the purity of chlorophyll b. A purification method based on this test was described in chapter 21. 100 >- cn z UJ o z LD o (/) Ijj q: o iij > UJ — ^— Chlorophyll o Chlorophyll b 620 660 700 740 WAVE LENGTH, m/i 780 Fig. 23.2. Fluorescence spectra of chlorophylls a and b in ether. Photo- metric curves corrected for self-absorption (after Zscheile and Harris 1943). Table 23. lA shows the positions of the main fluorescence bands of the two chlorophylls in ethyl ether, as found by several investigators, and Table 23. IB, the positions of the same bands in various solvents. A spectrophotometric curve of the fluorescence of a benzine extract from Brassica (containing both chlorophyll components) can be found in a paper by Vermeulen, Wassink and Reman (1937). In Table 23.1, some values represent band maxima, Xj^^x., as determined by photoelectric photometry, and others band axes, X, i. e., the arithmetic means of the extension limits of the bands on spectrum photographs. These limits depend strongly on the spectral sensitivity curves of the photographic plates used, and on the length of the exposure. Biermacher (1936) insisted, however, that, as long as only the peaks of the bands are photographed, e. g., by using suitably short exposures, the axes coincide 744 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 Table 23.1 Fluorescence Bands of the Chlorophylls a and b A. BAND POSITION IN ETHYL ETHER Maxima (X^ax .). rnn Axis (X; >. m/* Baas- Knorr, ] Becking, Zscheile T>h6r6. Bier- Albers Koning Zscheile Harris Raffy mac her Component Band (1933, 1935)« (1934)" (1935) b (1943)'' (1935)" (1936 )t a I (6330, 672 675 668.5 664.5 665 663 II — 723 720 736 — III — — — — 801 — b I (6370, 657 — 648.5 648.5 646 647 II — 705 708 713 — III — — — — 789 — B. BAND POSITIONS IN DIFFERENT SOLVENTS Chlorophyll a Chlorophyll b Axis ^max., Axis, '^max., ^^ (X), m^ m^ (X). mu Baas- Bier- Bier- Becking Knorr Zscheile, macher Knorr, macher Koning Albers Harris (1936) Albers (1936) Solvent np (1934) (1933) (1943) (1933) Hydrocarbons Pentane 1.357 — — — 663 — 644.5 Hexane 1.375 — — — 663 — 644.5 Cyclohexaiie 1.431 — — 665.5 • — — — Benzene 1.501 675 677 673 666.5 657 650 Paraffin (liq.)- • • • — 678 — — 668.5 — 645.5 Vaseline (white). — — — — 672.5 — 647.5 Ethers Ethyl ether 1.353 675 672 664.5 663.5 657 647 Isopropyl ether. . 1.37 — — 670 — — — Dioxane 1.42 — — 668 666 — 699.5 A Icohols Methanol 1.320 666 675 674 667 657 — Ethanol 1.362 675 — — 666 — 654.5 2-Methyl-l-pro- panol 1.39 — — 674 — — — 1-Butanol 1.402 — — 673.5 — — — 2-Ethyl-l-hexanol 1.43 — — 676.5 — — — Cyclohexanol .... 1.466 — — — 669 — 653 Halogenides and Sulfides Carbon tetra- chloride 1.466 — — 671.5 668 — 648 Chloroform 1.446 680 — — 670 — 650 Carbon disulfide. 1.630 681 — — 676 — 656 Methylene iodide 1.756 682 — — d — d Tetralin — — 688. 5^ — — — Miscellaneous Acetone 1.359 668 672 670 665 657 653 Methyl oleate . . . 1.46 — — 670 — — — Pyridine 1.509 — — — 674 — 658.5 Olive oil 1.48 — — 672 — — Aniline 1.586 675 — : - 676 — 662 Lecithin — 677 « — ^- , — — Table continued FLUORESCENCE SPECTRUM OE CHLOROPHYLL 745 Table 23.1 {Continued) C. BAND SHIFTS IN DIFFERENT SOLVENTS" Zscbeile and Harris (1943) Biermacher (1936) Chi. a, Solvent AX„ m/i Solvent Chi. o Chi. 6 AX, m^ Methanol 10 2-Ethylhexanol 10 2-Methylpropanol 9.5 Benzene 9 Isopropyl ether 9 Acetone 8.5 1-Butanol 8.5 Methyl oleate 8 Carbon tetrachloride 8 Olive oil 7.5 Dioxane 7 Cyclohexane 5.5 Ethyl ether 4.5 Pentane 4.5 4.5 Hexane 4.5 4.5 Pyridine 4.5 3.5 Carbon tetrachloride 4 3 Methanol 3.5 — Aniline 3 5 Benzene 2.5 4.5 Dioxane 2.5 5 Ethanol 2.5 6.5 Cyclohexanol 2 4 Chloroform 2 2.5 Carbon disulfide 2 6 Acetone 1.5 8.5 Ethyl ether 0.5 5 Paraffin (liq.) 0 0.5 " Photographic. * Photoelectric photometer. " Concerning subsidiary maxima, see page 748. <* Biermacher (1936) found no fluorescence at all in methylene iodide (as well as in nitrobenzene). ^ This figure is quoted by Seybold and Egle (1940) from Stern. ^ After Stewart, Knorr and Albers (1942). " AX = X(fluorescence) — X( absorption). with the true band maxima. He therefore denied that the difference be- tween the X vahies measured by him and the X^ax. vahies found by earher investigators covild have been due to the dechne in the sensitivity of his photographic plates in the far red; he suggested instead that this differ- ence was caused by the faihire of other observers to avoid "self-absorption," i. e., reabsorption of fluorescent light befoi-e its escape from the chlorophyll solution. (Because of the position of the fluorescence band of chlorophyll close on the red side of the absorption peak, self-absorption must cause an apparent shift of the fluorescence band maximum toward longer waves.) The correctness of this interpretation was acknowledged by Zscheile and Harris (1943), who made a spectrophotometric redetermination of the fluorescence bands, varying the chlorophyll concentration systematically, and using a capillary vessel to reduce self-absorption. The extent of the self-absorption effect is illustrated by figure 23.3. It shows how large a part of the fluorescence band is overlapped by the (shaded) absorption band, and how, in consequence of this overlapping, the position of the maximum of the fluorescence band can be displaced, by self-absorption, by as much as 12 mju. The position of the second fluorescence band (at 720 m^) is found to be practically unaffected by reabsorption; this is natural, since this band leads to a vibrating state of the chlorophyll molecule and consequently does not occur in absorption — at least not with a marked in- tensity. 746 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 Table 23. IC compares the most reliable figures in Table 23. IB — those of Biermacher, and Zscheile and Harris — with the wave lengths of the ab- sorption bands as listed in Table 21. IV. This comparison shows that the fluorescence maximum remains on the long-wave side of the absorption maximum in all solvents, but that the distance between the two maxima sometimes falls to as little as 1 m/x. ^^— Fluorescence absorbed by minimum layer of chlorophyll solution Fluorescence obsorbed by V4mm. loyer of chlorophyll solution Fluorescence absorbed by 3 V* mm. layer of chlorophyll solution Fluorescence obsorbed by 10mm. layer of chlorophyll solution >////^^. Absorption (from fig. 21.1) 620 780 660 700 740 WAVE LENGTH, m/i Fig. 23.3. Effect of self-absorption on fluorescence spectrum of ciilorophj'll a in ether (after Zscheile and Harris 1943). It is difficult to say whether any of the variations in the shift, AX, indi- cated by Table 23. IC, are significant; in particular, whether the conspicu- ous difference in the order of solvents found for the two chlorophylls, a and &, is real. Seybold and Egle (1940) suggested that AX is abnormally large (^^15 myu) in chloro- phyll solutions in Hpides such as lecithin. (This assumption was necessary for their "two-phase theory" of the state of chlorophyll in vivo; cf. Vol. I, page 393). This sug- gestion is not plausible in itself, and therefore cannot be accepted without confirmation by reliable measurements. As to the reason for the "red shift" of the fluorescence bands compared with the absorption bands, the explanation must lie in the loss of vibra- tional quanta in the interval between excitation and re-emission, or after re-emission. Quite generally, the molecule has a somewhat different nu- clear configuration in the excited and in the normal state. Therefore, ac- cording to the so-called Franck-Condon principle, electronic excitation is accompanied by the excitation of a certain amount of vibrations. A large part if not all these vibrational quanta are dissipated before re-emission of FLUORESCENCE AND ABSORPTION BANDS 747 light. After the emission, the molecule finds itself for a second time in a deformed state, and, for a second time in the fluorescence cycle, some energy is converted to vibrational energy. The magnitude of AX indicates that the vibrational quanta concerned must be of the order of 100 cm.-^, much smaller than the quanta (1000-1400 cm.-^) postulated on page 630 to account for the sequence of the visible absorption bands of chlorophyll. Another explanation of the red shift could be derived from the hy- pothesis (c/. page 631) that the main red absorption band, Xo -^ Yo {cf. fig. 21.20), conceals a weak band, Xo ^ Ao, which belongs to the yellow-orange band system. If this is true, and if the red fluorescence band is the pure Fo -^ Xo band, the somewhat different position of its maximum is under- standable. This explanation is less likely because the displacement of the fluorescence band toward the red is a general phenomenon, while the over- lapping of the bands Xo -^ Fo and Xo ^ Ao, if it exists at all, can be only an accidental occurrence. The influence of the solvent on the position of the fluorescence band must be attributed to the same cause as its influence on the absorption spectrum, i. e., to the difference in the solvation energy of the pigment in the ground state and in the excited state. Table 23.IC shows that within a homolo- gous group of solvents an approximate parallelism exists between the posi- tion of the fluorescence band and the refractive index of the medium. This regularity already was noted and discussed in chapter 21, when we dealt with the absorption spectra of chlorophyll in different media. A new light on the effect of solvents on the fluorescence of chlorophyll was thrown by the observations of Livingston, Watson and McArdle (1949), which will be described further below. These experiments indicate that the solvent effect is twofold: In the first place, the presence of at least a small amount of solvent molecules of a certain type (Avater, alcohols, amines) appears to be needed to bring out the fluorescence (presumably, by converting chlorophyll from a nonfluorescent into a fluorescent tauto- meric form). Alter the fluorescence had been "activated" in this Avay, its spectrum and intensity are independent of the specific nature of the "ac- tivator," and determined only by the nature of the bulk solvent. In other words, whether the fluorescence of chlorophyll a in benzene is "activated" by methanol, or piperidine, or water, its spectrum and intensity are char- acteristic of benzene as medium. Of course, when larger quantities of the "activator" are added, the spectrum must sooner or later approach that characteristic of the chlorophyll solution in the pure activator; but these transitions have not yet been studied. Like the two main chlorophylls, a and b, chlorophyll c {chlorofucin) also has a red fluorescence band; its axis lies at 631.5 mju in ether (Dhere and Fontaine 1931), and at 635 m/x in ethanol (Wilschke 1914). 748 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 Chlorophyll d, too, was described by Manning and Strain (1943) as exhibiting a deep red fluorescence. Its spectrum shows a first maximum at 693 mju (in ethereal sokition) and indications of a diffuse second maximum at 750 mM- No quantitative data appears to be available on the fluorescence spec- trum of allomerized chlorophylls. Proiochlorophyll has a fluorescence band at 626.5 mju, in ether (Dh6r6 1930). The fluorescence bands of pheophytin a lie at 676 and 730.5 mju in ether (Dhere and Raffy 1935) and at 677.5, 717, 750.5 and 804 mju in dioxane (Stern and Wenderlein 1936). A photometric curve of the fluorescence spectrum of bacteriochlorophyU in solution was obtained by Vermeulen, Wassink and Reman (1937). It showed two bands, a weaker one at 695 mju, and a stronger one at 810 m/j (fig. 23.4). The relationship of these two bands is not clear (c/. p. 751). The fluorescence of these and many other porphin derivatives was re- viewed by Dhere (1937, 1939). c J3 O 650 700 750 800 WAVE LENGTH, m/i 850 Fig. 23.4. Fluorescence spectrum of bacteriochlorophyU in solution (after Vermeulen, Wassink and Reman 1937). No counterpart of the strong blue-violet absorption band appears in chloro- phyll fluorescence, even if white, violet or ultraviolet light is used for excita- tion (c/. Dhere and Raffy 1935, Prins 1934, and Vermeulen, Wassink and Reman 1937) . The same is true of the weaker absorption bands in the mid- dle of the visible spectrum. True, Prins (1934) said that the red band oc- curs in fluorescence without the yellow and orange bands only if fluores- cence is excited by red light (660-680 m/x); and Ivnorr and Albers (1933, 1935) observed "subsidiary" fluorescence bands on the short-wave side of the main one— at 633 and 637 mju in chlorophyll a and b, respectively (c/. Table 23.1) ; but Zscheile and co-workers (1935, 1943), who excited fluores- cence with white light, and Vermeulen, Wassink and Reman (1937), who used ultraviolet and blue exciting light, obtained fluorescence curves with- out any indication of such additional bands (cf. fig. 23.2). Zscheile and Harris (1943) foimd that the fluorescence spectmm of chlorophyll was exactly the same whether excited by the mercury lines FLUORESCENCE OF CHLOROPHYLL DERIVATIVES 749 365, 404.7, 435.8 or 546 m/x, or by white light filtered through a red, orange yellow, violet, blue or green filter. Whether chlorophyll is capable of emittmg a weak, but long-lasting infrared fluorescence (originating in a metastable state, cj. page 753) is uncertain. Calvin and Dorough (1947) have described such a "phosphores- cence," but Livingston and co-workers (1948) could not confirm their ob- servations (c/. page 795). Figure 23. IB shows, beside the fluorescence spectra of the two chlorophylls, also those of the two pheophorbides, and of two porphyrins. The pheophorbides {i.e., chlorophyllides in which hydrogen has been substituted for magnesium) fluoresce not less strongly than the chlorophylls or chlorophyllides themselves; but certain other sub- stitutions in the same position in the molecule {e. g., copper instead of magnesium) cause complete disappearance of fluorescence. Stern and Molvig (1935, 1936^), Stern and Dezelic (1936) and Stern (1938) have investigated the fluorescence of numerous porphyrins and chlorins. They found that, similariy to chlorophyll, all of them fluoresce with red light, even when excited by violet or ultraviolet radiation. The main fluorescence band always lies close to the first absorption band in the red — whether this band is the weakest of the whole absorption spectrum (as in some porphyrins) or the strongest one (as in chlorins and phorbins). Stern (1938) found that tetrapyrrole compounds without the closed por- phin ring system (e. g., the bile pigments), as well as compounds in which the conjugation in the porphin ring is interrupted, do not obey this rule, and do not show sharp fluorescence bands at all. He therefore considered a sharp red fluorescence band as an important characteristic of the all- round conjugated porphin ring system. According to the term systems given in figures 21.9 and 21.25, the ap- pearance of the red band in fluorescence, to the exclusion of all the other bands, means that all excitation energy in excess of that corresponding to the lowest, nonvibrating, excited electronic state is dissipated before fluorescence can occur — probably first by internal distribution of this energy among vibrations within the pigment molecule, a process known as "internal conversion," and then by gradual transfer of vibrational quanta to the medium. The dissipation is interrupted at the lowest excited level, whether this is level .4 (in porphyrins), Y (in chlorins and phorbins) or Z (in bacteriochlorophyll), long enough to allow a significant proportion of the excitation energy to escape as fluorescence. It was suggested by Franck and Herzfeld (1937) that the capacity of chlorophyll to convert rapidly quanta of larger size into smaller red (quanta may be important for the function of this pigment in photosynthesis, be- cause it prevents the occurrence of undesirable photochemical reactions that could be sensitized by the larger quanta. This surmise may or may not be correct, but since the same property is shared by all porphyrins and chlorms, it cannot explain the special suitability of chlorophyll as photo- catalyst in photosynthesis. 750 FLUORESCENCE OF PIGMENTS /iV VITRO CHAP. 23 Table 23.11 A. Fluorescence Bands of Porphin in Dioxane Band no. -1 0 1 2 3 Interpretation Ai->'Xi Ao —*■ Xo ^0 -* Xi Ao —*■ Xi Ao -^ Xs X, niM V, cm. ' 591 16,900 616.5 16,200 644 15,500 669.5 14,950 684 14,600 Ai*, cm."' 700 700 550 350 B. Vibrational Quanta in Fluorescence Spectrum of Chlorophylls a and b Value V, cm. Chlorophyll a 665 15,000 736 13,580 801 12,480 1420 1100 Chlorophyll 6 646 15,500 713 14,020 789 12,670 1480 1350 The sequence of several bands observed in the fluorescence spectra of chlorophyll and its derivatives (as well as in those of the porphyrins) prob- ably corresponds to transitions from the excited electronic state ^o (or Fo) to different vibrational levels of the ground state, Xo, Xi^ X2 . . . As an example, we consider first the fluorescence spectrum of porphin, as ob- served by Stern and Molvig (1936). It consists of the five bands listed in Table 23.11. The band at 616.5 mju is the strongest. Comparison with the absorption spectrum of the same compound (compare Table 21. V) shows that all five fluorescence bands probably are related to the one weak band in the absorption spectrum at 613 m/x. According to the term system in figure 21.9, this is the 0-^0 band of the A -^ X system. The band designated as " — 1" in Table 23.11 must then be an "anti-Stokes band," originating in the next-to-lowest vibrational level of the excited electronic state A. According to Table 21. V, the level Ai is situated 1520 cm.~^ on the short-wave side of the 0-^0 band; while the fluorescence band " — 1" is displaced only 700 cm. ~^ in the same direction. This differ- ence can be interpreted by assuming that the state in which the " — 1" band originates belongs to a vibrational sequence not excited by light absorption. Or, one can postulate that the " — 1" band does originate in the vibrational level A, but terminates in a vibrational level of the ground state situated about 800 cm. ~^ above the Xo state. Bands 1, 2 and 3 probably all originate in the non vibrating excited state Ao, and terminate in the successive levels, Xi, X2 and X3, of the ground state. According to this interpretation, the vibrational quanta of the ground state of porphin, which are excited by fluorescence, are compara- tively small and rapidly declining in size (700-800, 550 and 350 cm.^). In chlorophyll, on the other hand, the two sets of vibrational quanta, derived from the absorption and the fluorescence spectrum, respectively, are of the same order of magnitude (1100-1500 cm. ~^). This is shown by a compari- son of the Av values in Table 21. VI with those in the Table 23.IIB. (For INTERPRETATION OF FLUORESCENCE SPECTRUM 751 the sake of uniformity, all figures in Table 23.IIB are based on the data of Dhere and Raffy in Table 23.IA.) In bacieriochlorophyll the distance between the two fluorescence bands (shown in fig. 24.4) is much larger than in chlorophyll— 2400 cm.-^; and the short-wave (red) band is weaker than the long-wave (infrared) band. This points to two different electronic transitions, rather than two vibra- tional bands in a common band system. The two fluorescence bands may even belong to two different molecular species. Only the stronger of them —that at 810 myu- appears to be correlated with a known absorption band of bacteriochlorophyll, that at 770 m/x (c/. fig. 21.7). (This correlation im- plies that AX, the displacement of the fluorescence band relative to the absorption band, is of the order of 40 mn, as against < 15 mn in ordinary chlorophyll.) This may be the place to mention the luminescence that occurs when a chlorophyl solution in tetralin is heated to 125° C. This phenomenon was first described by Rothe- mund (1938) and investigated spectroscopically by Stewart, Knorr and Albers (1942). The maximum of the luminescence band was found at 677.5 m^. After the tetralin solution was heated for five minutes, chlorophyll showed a change— its fluorescence band was shifted from its original position at 688.5 to 671.0 niyu and was reduced to one third its original intensity; the absorption spectrum also had undergone a transforma- tion, especially in the blue-violet region. The origin of this luminescence is as yet un- known, and its interpretation as chemiluminescence, suggested by the investigators, although plausible, requires confirmation. 2. Yield of Fluorescence and Life-Time of the Excited States of Chlorophyll The fluorescence yield can be defined either as the proportion of ab- sorbed energy re-emitted in the form of radiation or, more significantly, as the proportion of re-emitted photons. The two figures coincide only in the case of resonance fluorescence; usually (particularly in condensed sys- tems) the emitted light is of a lower frequency than the absorbed light ("Stokes' rule"), and the "energy yield," e/, is therefore smaller than the "quantum yield," (p. The relation €f < f holds true for the main fluorescence bands of all derivatives of porphin and chlorin, and the difference becomes particularly large if blue, violet or ultraviolet light is used for excitation. It was stated above (page 748) that only red light is emitted in the fluorescence of these compounds ; this means that the absorbed ultraviolet, violet or blue energy ciuanta are transformed into much smaller red quanta, while up to 50% of the absorbed light energy is dissipated. Since the "theoretical" life- time of the excited state B (upper state of the blue-violet band system) is of the order of 5 X 10 "« sec. {of. page 634), the absence of even 0.01% fluorescence in this system shows that the electronic energy of the state B is dissipated m less than 5 X IQ-^^ sec, i. e., after less than one hundred molecular vibrations. 752 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 It appears that all porphyrins, chlorins and phorbins, when brought into an electronic state with an energy higher than that of their lowest ex- cited state, rapidly lose this excess energy, and revert into state A (por- phyrins), Y (chlorins and phorbins) or Z (bacteriochlorophyll and its derivatives). Differences in the yield of red fluorescence among various compounds of these three classes must then be due to variations in the longevity of the lowest excited states {A, Y or Z). In ''nonfluorescent" compounds (e. g., copper pheophorbide, wdth

oChl + rQ (23.1B) Chi* + Q > rChl + oQ (where o stands for oxidized, r for reduced, Q for quencher and Chi* for excited chlorophyll). The second tj^pe is described by equation (23.2) : (23.2) Chi* + Q > Chl*Q * ChlQ > Chi + Q the quenching effect being due to accelerated internal conversion of ex- citation energy into vibrational energy in the complex Chl*Q. If Q is the solvent, only reactions (23.1 A and B) can be classified as "chemical quenching," while reaction (23.2) becomes identical with "phys- ical" energy dissipation in the solvated pigment molecule. Photochemical reactions with foreign molecules interfere with fluores- cence only if they take place directly, i. e., by encounters of electronicallj'' excited molecules with the quencher. If, on the other hand, reactions of this kind are preceded by monomolecular steps such as isomerization (or dissociation) of the excited molecule, their effect on fluorescence may become negligible (since the molecules that take part in the photochemical reaction are the ones lost for fluorescence anyhow; (c/. second scheme on p. 483, Vol. I). We will use this concept below; cf. page 788 in interpreting the nonquenching of chlorophyll fluorescence by certain compounds whose autoxidation is sensitized by this pigment. Sometimes the isomeric molecule, formed after light absorption, has a certain chance of reverting into the original electronicalh' excited state (e. g., with the help of thermal energy) . Similarly, photochemical dissocia- tion products may have a certain chance of forming an electronically ex- cited molecule by recombination. If this exact reversal of the primarj^ photochemical process occurs after a period that is long compared to the duration of ordinary fluorescence (10~'^ sec), we obsei-i'e the emission of "delayed fluorescence" or "phosphorescence." Chemical reactions of the metastable, isomeric photoproduct (or of the dissociation products) will cause quenching of this delayed emission. This is the mechanism of the strong quenching of phosphorescence of many dyestuffs by oxygen {cf. page 789). {d) Bulk Transfer of Electronic Energy Transfer of electronic excitation energy "in bulk" to another type of molecules can lead to the quenching of the fluorescence of the originally excited molecular species, either by substitution of a secondary (stronger 758 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 or weaker) "sensitized" fluorescence of the quencher, or (if the quencher is nonfluorescent) by complete conversion of the excitation energy into heat. Three types (or rather three Hmiting cases) of electronic energy trans- fer mechanisms are known. The first, which is the only one possible when the distance between the excited molecule and the quencher is >10~^ cm. (for visible light quanta), is trivial — the emission of a light quantum by the primarily excited molecule and its reabsorption by a molecule of the ciuencher, a process similar to the "self-absorption" of fluorescence (page 745). The second mechanism is energy transfer by kinetic collisions (so-called "colhsions of the second kind"), or "encounters," to use a term more appropriate for molecules in solution. It is associated with the mutual disturbance of the electronic structures of the two molecules in contact, and requires approach to within the kinetic collision diameter (10~^ to 10"'' cm.). In this case, the energy exchange is not contingent on "resonance" between the electronic excitation states of the two part- ners, since a considerable fraction of electronic energy can be converted into vibrational or kinetic energy in the collision. A third and perhaps most interesting possibility is the "resonance transfer" of electronic excitation energy between two practically undisturbed molecules, which can occur when these molecules are within a distance smaller than the wave length of the exchanged quantum (^^10~-' cm. for visible light), and does not require an actual "contact" between them. The probability of this kind of transfer depends decisively on resonance between the energy-exchanging molecules (i. e., on the mutual overlapping of the fluorescence band of the donor and the absorption band of the acceptor). The phenomenon was first discussed by Kallman and London in application to sensitized fluores- cence in gases. Similar considerations were afterward applied to solutions by J. Perrin (1926, 1927), who used classical electrodynamics, and by F. Perrin (1929, 1932), who first attempted a quantum-mechanical treatment. F. Perrin used this energy transfer mechanism to interpret so-called "concentration depolarization" of fluorescence in solution (decrease in the degree of polarization with increasing concentration). Subsequently, several other phenomena in fluorescence and photochemistry have been ascribed to energy exchanges of this type, and improved theoretical treat- ments were evolved by Vavilov and co-workers (1942, 1943, 1944), Forster (1946, 1947, 1948) and Arnold and Oppenheimer (1950). Because of the importance of the resonance transfer concept for the photochemical mechanism of photosynthesis (in particular, for the possible participation of phycobilins and carotenoids in it), these papers will be discussed in greater detail in chapters 30 and 32. Here, we are concerned only with the possibility of quenching (or excitation) of chlorophyll fluorescence being FACTORS LIMITING THE YIELD 759 due, in some cases, to resonance transfer of- excitation energy not requiring molecular contact. Examples will be found on p. 778 (quenching of dye fluorescence by other dyes), p. 790 (chlorophyll q fluorescence sensitized by b) and in chapters 24 and 32 (energy transfer between pigments in vivo). Self-quenching, too, may be caused by resonance (p. 797). (e) Self-Quenching Experience shows that quenching by molecules identical with the ex- cited one often is particularly strong. This is revealed by rapid decrease in the yield of fluorescence of many substances with increasing concentra- tion of the fluorescent pigment. This strong self -quenching probably is due to very close resonance between the fluorescent molecule and the quencher. However, resonance transfer of electronic energy does not in itself explain self-quenching, because, from the point of view of the yield of fluorescence, it should be irrelevant whether the excitation energy stays with the originally excited molecule or is transferred to another molecule of the same kind. Nevertheless, self-quenching can result from resonance, if some additional phenomena are taken into account. Effective energy dissipation can result either from kinetic encounters of excited and normal pigment molecules ("kinetic self-quenching"), or from their close average proximity ("static self-quenching"). In the first case, we can pos- tulate transient formation of dimeric molecules during the encounter. Resonance between the two structures, D*D ^ DD*, creates an attraction force, leading to a more intimate contact of the two electronic systems than that established in an encounter of two nonresonating molecules. This can bring about accelerated conversion of electronic into vibrational energy; the molecules which met as D* -|- D will then separate as D + D. Resonance transfer of excitation energy over distances wider than a collision diameter also can explain self-quenching, if one makes certain auxiliary hypotheses. Forster (1947, 1948) suggested, for example, that even dyestuff solutions which reveal no equilibrium dimerization (i. e., show no effect of concentration on the absorption spectrum; cf. below), contain a small proportion of nonfluorescent, dimeric molecules. If the resonance exchange of excitation energy is so fast that this energy visits. during its life time, a considerable number (say > 100) pigment molecules, the presence of even a single dimer in this series of "hosts" may suffice to "trap" the excitation energy, and dissipate it into heat. (Of course, for this mechanism to be effective, the absorption band of the dimer and the fluorescence band of the monomer must overlap sufficiently to permit reso- nance exchange.) Franck and Livingston (1949) suggested another possibility — that 760 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 energy traps are provided by monomers which are in the state of excep- tionally strong thermal agitation. (This hypothesis explains also the de- cline in the intensity of fluorescence, usually observed with rising tempera- ture.) Accelerated dissipation of electronic energy in "hot" molecules is plausible, since in order to convert electronic energy into vibrational energy, a configuration of the nuclei must be reached in which the electronic sys- tem has the same energy in the excited and in the normal state ("crossing point of two potential curves" in the diatomic model). This configuration usually can be achieved only by combination of electronic excitation with vibrations of appropriate kind; the excited molecule must wait until an accidental fluctuation of thermal agitation supplies the critical degree of freedom with the amount of vibrational energy required to make internal conversion possible. The higher the temperature, the shorter will be this waiting period, and the greater the probability of internal conversion oc- curring during the electronic excitation period, and competing successfully with fluorescence. One may ask: how can resonance migration of excitation energy assist this mechanism of dissipation? Does it make any difference whether the excitation stays with one molecule and awaits there the thermal fluctua- tion that will permit it to be dissipated, or whether it visits a thousand molecules during the same total life-time, spending a correspondingly short time with each of them? We said elsewhere in this book that a man cannot change his life expectancy by sleeping every night in a different bed! Whether this analogy applies here or not depends on the relative duration of a thermal fluctuation and electronic excitation. If the fluctuation is short-lived, in comparison not only with the total duration of electronic excitation, but also with the time during which the excitation remains with a single host molecule, then migration can have no effect — the chance of being hit by lighting is the same whether one spends the thunderstorm under a single tree or shifts every minute from one to another (identical) tree. If, however, the state of abnormal thermal agitation lasts long com- pared to electronic excitation of a single molecule, then resonance exchange will increase the chances of the two meeting in one molecule. (If one house in a hundred in a town is quarantined for smallpox, then a visitor who comes to town will be much safer if he stays the whole time in the first house he has entered than if he visits a hundred houses, spending a corre- spondingly short time in each of them.) In mathematical form, the probability of two independent events, one lasting T sec. (electronic excitation) and another t sec. (thermal fluctuation) overlapping each other in a single molecule, is changed, by subdividing T into n periods of T/n sec. each, by the factor: (nt/T) + 1 FACTORS LIMITING THE YIELD 761 which is significantly different from 1 if f is not < T/n, i. e., if the duration of the thermal fluctuation is not much shorter than that of electronic ex- citation of a single molecule. The former can be postulated to last for a period of a few molecular vibrations, thus t ^ 10 -^^ sec. The total period of electronic excitation is T ^ 10 -« sec. (for example, in alcoholic chloro- phyll solution, the natural life-time of excitation is ^5 X 10-^ sec; the actual life-time must be ten times shorter, as indicated by a fluorescence yield of about 10%). Under these conditions, for t to be not much shorter than T/n, the number n must be higher than 10^ i.e., excitation energy must be exchanged more than ten thousand times before its dissipation (staying < 10-^^ gee. at each molecule visited). The role of thermally excited ("hot") monomeric dyestuff molecules in the concentration quench- ing of fluorescence thus is predicated on this minimum length of energy exchange chains, and on the possibility of internal conversion occurring during the extremely short sojourn of the electronic energy in the hot mole- cule. It may be suggested that conversion requires (at least) a period of a single molecular vibration (^10-^^ sec). This would restrict quenching by hot molecules (in the case of chlorophyll) to exchange chains not shorter than 10^ and not longer than 10^ molecules. In addition to the various physical mechanisms of self-quenching which were considered so far, two chemical mechanisms also are feasible, analogous to the two chemical mechanisms of quenching by foreign substances, dis- cussed in section (c). They are: an oxidation-reduction reaction between the excited and a normal molecule (photodismutation) : (23.3) Chi* + Chi > oChl + rChl and formation of nonfluorescent dimers by two normal molecules: (23.4) Chi + Chi , Chl2 Dimerization of dyestuff molecules is favored by the fact that marked resonance attraction must occur not only between an excited and a normal molecule (as discussed above), but also between two molecules in the non- excited state. (This application of London's theory of intermolecular forces to pigment molecules was suggested by Rabinowitch and Epstein 1941.) The tendency of dyestuff molecules to dimerize (or polymerize) in solution may be attributed to such resonance phenomena. In many cases— perhaps the majority of those observed so far—self-quenching of fluorescence appears to be due to "permanent" dimerizations (or polymeri- zations) rather than to reaction (23.3) or dimerization after excitation. The dechne in yield of fluorescence with increasing concentration is in this case a direct measure of the degree of association. Like the unstable dimers formed in light, the stable dimeric molecules formed in the dark can be 762 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 nonfluorescent either because of rapid internal conversion of excitation energy, or because of the occurrence of internal oxidation-reductions (dis- proportionations). It was already mentioned above that combination of dimerization with resonance exchange of energy can lead to strong quenching even when the number of dimeric molecules is very low. Whether self-quenching is due to pre-existing dimers (or polymers) or to encounters between excited and nonexcited monomers (in which dimers are formed) can often be deduced from observations of the absorption spectra: In the first case the absorption spectrum of the dyestuff must change with concentration (as observed, e. g., by Rabinowitch and Epstein with thionine and methylene blue) ; in the second case. Beer's law must be obeyed {i. e., the absorption spectrum of the dye must be independent of concentration). Quenching by a chain of energy exchange reactions, with a dimer as occasional link in the chain, suggested by Forster, also does not require marked deviations from Beer's law, since the number of dimers present can be very small. Which of the several possible quenching or self-quenching processes actually limits the yield of fluorescence in a given solution is not easy to say. If the pigment is stable in light, and its fluorescence is unchanged after long illumination, "physical" quenching is the likely mechanism. True, even when chemical quenching does occur, the dyestuff may be photostable, if the quenching reaction is reversible; and the yield of fluorescence may remain unchanged with time, even when quenching ini- tiates an irreversible sensitized chemical reaction, if the products of this reaction do not quench fluorescence stronger (or weaker) than the originally present molecular species. Usually, however, chemical quenching is not entirely reversible, but causes a more or less rapid chemical change of the fluorescent pigment; and sensitized chemical reactions often do lead to the formation of products whose presence changes the intensity of fluores- cence. When this is the case, the fluorescence yield must change with time — either because the original fluorescent compound is converted into a new one, with different properties, or because the fluorescence yield of the original species is changed by the accumulation of the products of the sensitized reaction. Therefore, whenever the yield (or the spectrum) of fluorescence changes with time, the indication is strong that chemical fac- tors account for at least part of the quenching (for examples, see page 764). A systematic study of the effects of solvent, concentration, admixtures, temperature and other factors on the yield of fluorescence is needed to elucidate the quenching problem, which was discussed above on the basis of general possibilities more than on the basis of actual observations with chlorophyll. Studies of this kind would be' of particular interest for under- standing the mechanism of sensitized photochemical reactions, such as INFLUENCE OF SOLVENT ON YIELD 763 photosynthesis. Since this summary was written, a number of pertinent data have been collected by Livingston and co-workers. A discussion of these will be found below. If we consider the few presently available data on the intensity of chlorophyll fluorescence in different media, we acquire the impression that physical energy dissipation and chemical quenching must both play a part in these systems; but much remains to be done before their relative roles will become clear. The possibility that the yield of chlorophyll fluorescence in solution may be limited by photochemical dissociation of chlorophyll {e. g., into "monodehydrochlorophyll" and a hydrogen atom) was suggested by Franck and Wood (1936). This hypothesis was discussed in chapter 18 (Vol. I, page 484), and it was pomted out that a quantum of red light (with an energy of about 40 kcal/emstein) is unlikely to disrupt a carbon- hydrogen bond (whose standard energy is about 100 kcal/mole), even if some energy might be gained by the solvation of the hydrogen atom. Thus, if light absorption does cause a reversible photochemical change of chloro- phyll, it is more likely to be either tautomerization, or reaction with the sol- vent. Dissociation becomes more likely when excitation occurs in the blue- violet band, with quanta of about 60 kcal./einstein; it was mentioned be- fore that this is one possible explanation of the lower yield on fluorescence in this region. It is not unlikely that energy dissipation by internal conversion is the basic factor limiting the yield of fluorescence of chlorophyll in condensed systems; it is also possible that— as suggested by Franck and Livingston (1941)— tautomerization occurs as a more or less regular intermediary stage in this dissipation (c/. Vol. I, p. 490). Chemical interactions with solvent or admixtures, as well as self -quenching, are then to be considered as contributmg factors, which further depress the yield of fluorescence under certain conditions. (According to Lewis and Kasha — cf. pp. 790-2 —formation of a metastable triplet state could play the role ascribed above to tautomerization.) 4. Influence of Solvent on Yield of Chlorophyll Fluorescence Chlorophyll fluoresces in all (or most) organic solvents (as well as in wax and paraffin), but with different intensity. No precise measurements of the fluorescence intensity in different solvents are available, but Franck and Levi (1934), Albers and Ivnorr (1935), Ilnorr and Albers (1935), Knorr (1941) and Zscheile and Harris (1943) gave some preliminary results; all these investigators agree that wide differences occur both in the initial intensity of fluorescence in different solvents, and in its change with time (cf. Table 23.III). 764 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 Table 23.III Changes of Fluorescence Intensity of Chlorophyll with Time in Different Solvents" (after Zscheile and Harris 1943) Fluorescence intensity (relative units) Time of illumination, min. Ethyl ether Isopropyl ether Acetone Cyclo- hexane Benzene ecu 0.25 1 2 3 57.5 54.0 51.5 144 139 134 130 178 167 163 161 61.0 49.0 44.0 42.8 110 71.0 55.6 44.0 90.5 44.5 29.0 21.0 » According to Franck and Levi (1934), the fluorescence of chlorophyll is about twice as strong in ethanol as in aniline. More recently, Livingston and co-workers found that fluorescence is very weak or entirely absent in nonpolar solvents if these are free from traces of water, alcohols or amines. An effect of solvent on the yield of fluorescence is to be expected what- ever the mechanism of quenching. Even the probability of "monomolecu- lar" chemical quenching by dissociation (or tautomerization) of the ex- cited molecule probably depends on the degree of stabilization of the dis- sociation products (or of the tautomeric form) by solvation. (Well known is the effect of solvent on enolization — a tautomeric transformation that can occur in chlorophyll; cf. Vol. I, page 444.) The occurrence of self- quenching, too, may differ strongly depending on the solvent, as indicated, e. g., by the vastly different stabilities of dimeric dyestuff molecules in water and alcohol. The velocity of physical dissipation of excitation en- ergy by "internal conversion" also must depend on the nature of the me- dium to which the vibrational quanta are to be transferred in order to make dissipation irreversible. Finally, the probability of "bimolecular" chemi- cal quenching by reaction with the medium obviously is a function of the nature and purity of the solvent. The "physical" quenching of fluorescence by the solvent can be ex- pected to be least efficient in solvents of symmetric, nonpolar nature, such as cyclohexane or carbon tetrachloride. However, recent evidence shows that chlorophyll solutions of this type fluoresce less strongly than those in more polar or polarizable solvents, and this makes it probable that chemical interactions of the pigment molecules may often be at least as important as physical dissipation. The rapid decline of fluorescence with time, found in many chlorophyll solutions (e. g., those in benzene and car- bon tetrachloride), also points to chemical interaction; but whether it in- volves the solvent or impurities (e. g., dissolved oxygen) remains to be es- tablished. It is also uncertain whether the rapid decline of fluorescence is caused by a chemical transformation of chlorophyll, or by chlorophyll- sensitized formation of substances with strong quenching properties. INFLUENCE OF SOLVENT ON YIELD 765 Obsei-vations of Albers and Ivnorr (1934, 1935) and Kjiorr (1941) con- cerning the changes in the fluorescence spectra of chlorophylls a and b with time, in different solvents (ether, acetone, benzene and methanol), and under different atmospheres (air, oxygen, carbon dioxide and nitrogen) revealed a bewildering variety of shifts in positions, shapes and intensities of the fluorescence bands, which do not lend themselves to easy mterpreta- tion, but indicate complex chemical changes. Apparently, both the sol- vent and the dissolved gases participated in chemical reactions with ex- cited chlorophyll molecules. In some systems, these reactions led to a complete disappearance of fluorescence after less than one hour of illumin- ation. One reason for the complexity of the results of Knorr and Albers may have been the use of unfiltered light from a powerful mercury arc. Strong ultraviolet irradiation may have caused chlorophyll to react with substances that would not have affected it in visible (particularly red) light. One strange observation of Ivnorr and Albers is that the fluorescence of chlorophyll a (but not that of chlorophyll h) in acetone (but not in other solvents) is best preserved under an atmosphere of oxygen (where it dis- appears only after twelve hours of illumination, whereas, in nitrogen or carbon dioxide, it vanishes completely in less than one hour). As de- scribed in chapter 18 (Vol. I, page 491), this may mean that the quenching of fluorescence is caused, in acetone, by a reduction of the pigment (and oxidation of the solvent) and that oxygen restores the reduced pigment to its original fluorescent form (c/. chapter 36) . We have considered so far only those changes in the intensity of fluores- cence which could follow from the interaction of the chlorophyll molecule in the excited state with the medium. It w^as mentioned, however, in the introduction that another type of fluorescence effects is possible — one in which the state of the chlorophyll molecule is altered already in the dark, prior to excitation. This alteration must manifest itself in a change of the absorption spectrum. One possibility of this type is that chlorophyll may dimerize (or polymerize) in some solvents, and remain monomeric in others. Dimerization is known to cause the disappearance of fluorescence of many dyestuffs (such as methylene Ijlue) in aciueous sohition. In the case of chlorophyll, no similar effect has as yel Ix'on discoN^crod— unless owv con- siders the nonfluorescence of colloidal solutions and solid chlorophyll as the result of "quenching by polymerization." Another type of associa- tion, how^ever, appears to be important in this case — association of chloro- phyll molecules with hydroxyl groups or amine groups present in solvent molecules; in contrast to the quenching effect of dimerization, association of this type seems to be necessary to bring out the fluorescence of chloro- phyll. 766 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 This conclusion follows from some remarkable observations described by Livingston, Watson and McArdle (1949). Contrary to what was gener- ally assumed before, they found that chlorophyll solutions in nonpolar organic solvents do not fluoresce at all (or only very weakly), but that traces of polar admixtures, such as water or methanol, are sufficient to "activate" their fluorescence. In this study, Livingston and co-workers used a mercury arc for excita- tion (mainly the lines 436 and 405 m/x) . The fluorescent light was filtered through a deep-red filter, so that only the second fluorescent band of chloro- phyll a (720 m/i) was measured. This eliminated self-absorption. (It was also stated that the position of the second band is less strongly affected by the solvent than that of the first band, which can be shifted by as much as 7.6 m^u, cf. table 23. IC; but this seems strange, since one would not Fig. 23.5. UJ o z UJ o (/) UJ cr o _I U. u. o z UJ 0.2 0.4 0.6 0.8 1.0 SQUARE ROOT OF MOLE FRACTION OF ALCOHOL Intensity of fluorescence of chlorophyll a in the system octanol-benzene (after Livingston 1948). anticipate substantial differences in solvent effects on two emission bands originating, presumably, in the same excited state and leading to two ad- joining vibrational levels of the same lower state.) Chlorophyll solutions in dry hydrocarbons were obtained by evaporat- ing in vacuum a solution of chlorophyll a or h, dissolving the residue in dry hydrocarbon, evaporating again, and repeating this operation until all water (which may have been present in chlorophyll from its preparation) had been removed. (Disappearance of fluorescence was used as criterion of dryness.) Various polar "activators" were then added, dissolved in the same dry hydrocarbon. In purest dry benzene, the intensity of fluorescence (F) was <3% of that ordinarily observed in the same solvent (Fo) ; even this weak fluorescence ACTIVATION OF FLUORESCENCE 767 could perhaps be due to residual moisture (or other polar admixtures). Addition of 0.01% water (6 X 10"^ mole/1. HoO) brought F back to the usual level, Fo- Other solutions that proved nonfluorescent were those in n-heptane, iso-octane, styrene, chlorobenzene, carbon tetrachloride and diphenyl ether. Solutions in methanol, ethanol, octanol, dimethyl ether and diethjd ether, on the other hand, remained fluorescent even after drying. It was Fig. 23.6. IQ-'^ 2x10"'* MOLALITY OF ACTIVATOR Intensity of fluorescence of chlorophyll a in a hydrocarbon as function of the concentration of an activator (after Livingston 1948). Curve number 1 2 3 4 5 Solvent n-Heptane Benzene Benzene Isooctane Benzene Activator Phenylhydrazine Benzyl alcohol Cetyl alcohol Methanol Piperidine Chlorophyll ... a a or 6 a a a considered possible, however, that the fluorescence in the last two solvents was due to residual impurities. Figure 23.5 shows the intensity of fluorescence as function of composi- tion in a mixture of benzene and octanol. The fluorescence is completely activated by 0.0016 mole alcohol in a mole of hydrocarbon, corresponding to a concentration of about lO"^ mole/1. Similar relationships were found in mixtures of benzene wuth other polar solvents — alcohols and amines. Figure 23.6 shows the initial parts of five activation cui-ves. Table 23.IIIA gives, under [Ac]i/j, the molar concentrations of the activators needed to raise the fluorescence intensity to V2 ^0; they range from 6.8 X 10 "^ 708 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 mole/1, for dimothylamine in benzene, down to only 6.5 X 10~^ mole/1, for piperidine in benzene. Water is about half as effective as piperidine. Although amines generally are the strongest activators, diphenylamine and diphenylhydrazine are without effect. It is noteworthy that phenyl- hydrazine acts as activator in low concentration, and as quencher in high concentrations. The maximum intensity of fluorescence to which chlorophyll in a given nonfluorescent solution can be raised by activators is independent of the specific activator used, at least in the first approximation. Activation seems to be completely reversible; in other words, fluores- cence disappears again if the activator is distilled away, and reappears upon its renewed addition. Table 23.IIIA Efficiency of Activation of Fluorescence of a 5 X 10^*^ Mole per Liter Solution OF Chlorophyll a (after Livingston, Watson and McArdle, 1949) Solvent Activator [Ac]i,5 Ki Benzene Dimethykniline 6.8 X lo-i* 1.05 X 10 Benzene Phenol 6.0 X 10-2 1.55 X 10 Benzene Aniline 2.75 . X 10-2 4.55 X 10 n-Heptane Phenylhydrazine 8.0 X 10-^ 1.70 X 10=" Benzene Phenylhydrazine 6.5 X 10-^ 1.78 X 10=" Benzene Formamide 4.3 X 10-^ 2.50 X lO'' Benzene Benzyl alcohol 4.2 X 10-^ 2.90 X 10' Benzene Benzyl alcohor 4.2 X 10-^ 2.90 X 10^ Benzene Benzoic acid 3.6 X 10-4 3.15 X 10» Benzene Cetyl alcohol 2.9 X 10-4 4.15 X 10' Benzene Octyl alcohol 2.0 X 10-" 4.57 X 10' Iso-octane Methyl alcohol 1.1 X 10-4 1.03 X 10* Benzene Benzylamine 3.9 X 10 -s 2.67 X 104 Benzene Benzylamine" 3.3 X 10-^ 3.00 X 104 Benzene Water 3.5 X 10-^ 2.95 X 104 Benzene n-Heptylamine 7.5 X io-« 1.36 X 10^ Benzene Piperidine 6.5 X io-« 1.56 X 10^ " Chlorophyll b. The fluorescence spectrum does not change significantly with progres- sive activation. At least, in benzene-water mixtures, the positions of the two band peaks are the same at F/Fo = 0.2, 0.8 and 1.0. As described earlier, the absorption spectrum of chlorophyll does change with increasing admixture of the activator (c/. fig. 21. 26, A and B). The absorption spec- trum of a fully activated solution, although it is different from that of the ACTIVATION OF FLUORESCENCE 769 nonactivated one, is independent of the nature of the activator, and bears no relation to the absorption spectrum of chlorophyll in pure activator. The fluorescence of activated solutions is quenched by rising tempera- ture (15-70° C), particularly strongly in the region of partial activation. The maximum intensity, Fo, is a linear function of temperature over a comparatively wide range (a similar relationship was found also by Zscheile and Harris, 1943). In pai'tly activated solutions, the temperature curve is not only steeper than in fully activated ones, but also shows — with some activators at least — a definite curvature. The concentration of the strongest known activators, required to achieve complete activation, is of the same order of magnitude as that of chloro- phyll itself, which in the experiments of Livingston et al. was 5 X 10 ~^ mole/1. This shows that the effect cannot be due to kinetic encounters between chlorophyll and activator (which are too rare at such low concen- trations) ; nor can it be ascribed to a change in properties (such as dielec- tric constant) of the solvent as a whole. Rather, the effect must be caused by the association of chlorophyll molecules with the molecules of the activa- tor. The change in absorption spectrum supports this assumption. If the nonassociated form is totally nonfluorescent, the intensity of fluorescence can be used to calculate the proportion of chlorophyll molecules in the as- sociated form — assuming that the absorption coefficient of associated chlorophyll is the same as that of the nonassociated pigment. Figvire 21. 26. A shows that this is not quite true for chlorophyll a at 436 ni/x; but Livingston neglected this difference. Assuming a one-to-one complex [ChlAc ] the equilibrium constant of association : can be calculated from the half-activating concentration [Ac]i/,: K[Chl]o[Ac]o (23.4B) [ChlAc] = 1 + K[Ac]o ^^^•■*^^ Fo " 1 + A'[Ac]o (23.4D) A' = ^ Ac],/, {cf. chapter 27, eq. 27.12). This is a simplified solution, based on the as- sumption [Ac] ^^ [Ac]o; in other words, it assumes that the amount of activator bound in the complex is small compared to the total amount added to obtain activation. In Table 23.IIIA, the concentration [AcChl] at half-activation is 2.5 X 10 ~^ mole/1, for piperidine; since [Ac Jo = 6.5 X 10-^ [Ac]i/, = 4 X 10"" molc/1. In other words, the simplified equations are not quite applicable to piperidine (and four other systems 770 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 at the bottom of Table 23.IIA. In this case the approximate equation (23.4D) should be replaced by the exact equation: (23.4E) K = [Ac]y, -0.5[Chl]o Livingston used, however, the simplified equation for all the systems studied by him. Furthermore, he calculated the constants K, not from the value of [Acl^/^, but from the average slope of the activation curve. He obtained in this way the K-values given in Table 23.IIIA, which differ somewhat (although not in the order of magnitude) from values that could be calculated by means of equation (23.4E). (For example, the latter would give, for piperidine, K = 2.2 X 10^ instead of 1.6 X 10^). Deviations from straight line in the plot of log [(Fo/F) — 1] against [Ac]o were noted by Livingston and co-workers at low concentrations of the activator, and were ascribed by them to the presence of an "adventi- tious" activator, Ac' (probably water). They held the latter responsible for the weak fluorescence still noticeable even at [Ac]o = 0 {e.g.,F ^ 0.03Fo in purest benzene). By using the same assumptions as before («chiAc = achi for excitation light and [Ac']o » [ChlAc']), and assuming the adventi- tious activator to be water, Livingston and co-workers obtained theoretical curves fitting well the experimental results. From the equilibrium constant K and its change with temperature, values of AF, AH and AS were calculated for three systems shown in Table 23.IIIB. Table 23.IIIB Thermodynamic Constants for the Association of Chlorophyll a WITH "Activators" in Benzene Activator AFo, kcal./mole A//. kcal./mole ASo, E.U. Heptylamine Cetyl alcohol -7.1 -5.0 -6.0 -2.5 -1.5 3.3 5.0 Aniline -2.3 2.7 Since Fo seems to be independent of the nature of the activator, Livings- ton suggested that the essential effect of association is isomerization (or, rather tautomerization) of chlorophyll, e. g., conversion of the keto form into an enol. If — as assumed by Livingston — this is the transformation discussed in Volume I, page 459, allomerized chlorophyll should be inca- pable of it, and should therefore remain fluorescent in pure hydrocarbons. This consequence remains to be tested. Livingston attributed the postulated stabilization of the keto form by ACTIVATION OF FLUORESCENCE 771 alcohols and amines to the formation of hydrogen bonds with the amino or hydroxy group, shown in formula 23.1. I I I ^c^\^ ^c^'^c^ -c<\^ H H I Chelated Enol Form Keto Form Stabilized Keto Form (noiifluorescent, stable (unstabilized) (fluorescent, stable in the in pure hydrocarbons) presence of amine or alcohol) Formula 23.1. Tautomerization and fluorescence of chlorophyll (after Livingston, Watson and McArdle, 1949). The fluorescence of ethereal chlorophyll solutions is not covered by this hypothesis, and further experiments are needed to show Avhether this fluorescence would persist upon more stringent purification. As an alterna- tive explanation of the phenomenon of activation, Li\nngston considered the possibility that chlorophyll forms nonfluorescent dimers in pure hydro- carbons, and that these dimers are dissociated into fluorescent monomers by association with polar molecules. If this were the case, however, one would expect the extent of activation to depend on concentration of chloro- phyll; while a few — admittedly preliminary — experiments showed no dif- ference in the values of F/Fo between partially activated chlorophyll solu- tions in benzene containing 4.8 X lO'S 1.5 X 10"* and 2.3 X 10"^ mole/1, chlorophyll, respectively. Observations bearing obvious relation to those of Livingston and co- workers, have l)een made also by Evstigneev, Gavrilova and Krasnovsky (1949^. In studying the effect of oxygen on the absorption spectrum and fluorescence of chlorophyll, they first found this effect to depend on the solvent. In toluene, heptane and carbon tetrachloride, oxygen increased the absorption (page 648) and activated the fluorescence, while in pyridine, ethanol, ethyl acetate, acetone and (commercial) benzene, it had no effect on absorption, and quenched fluorescence. Later (1949^) the same investigators found that these effects were due not to oxygen, but to water vapor. In moist toluene, oxygen quenched fluorescence in the same way as in ethanol or other polar solvents. Evstigneev and co-workers also discussed two conceivable mechanisms of the action of polar solvents. One was the same as Livingston's alterna- tive hypothesis — solvent effect on dimerization. The other differed from Livingston's preferred hypothesis: it assumed attachment of polar mole- cules to the free co-ordination places at the central magnesium atom. This 772 FLUORKSCENCK OF PIGMENTS IN VITRO CHAF. 23 last hypothesis is supported by the observation that polar molecules do not affect the absoi-ption spectrum and fluorescence of pheophytin; even more convincing is the fact that a similar difference in behavior occurs between phthalocyanine and its magnesium complex, although these compounds contain no cyclopentanone structure, so that Livingston's interpretation cannot be applied to them. 5. Influence of Concentration on Yield of Fluorescence. Self- Quenching. Fluorescence of Chlorophyll in Colloids and Adsorbates The self-quenching of fluorescence in chlorophyll solutions was studied by Weiss and Weil-Malherbe (1944). They found that with a constant intensity of illumination, the intensity of fluorescence of an ethyl chloro- phyllide solution in ethanol increased with increased concentration, be- tween 1 X 10"^ and 1 X 10^^ mole/1., due to increased absorption. The latter became practically complete above 1 X 10^^ mole/1.; instead of be- coming constant from there on, the fluorescence intensity declined rapidly, dropping at 1 X 10 -^ mole/1, to one sixth of its maximum value. Six points were measured between [Chi] = 2 X 10"* and 1 X 10 --^ mole/1., and found to lie on a hyperbola: (23.5) F = C/{l+k[Ch\]) where F is fluorescence intensity and k and C are constants. The constant k was evaluated graphically and fc = 2 X 10'^ sec.-^ was found. Making the— unjustified— assumption that the intensity of fluorescence is limited exclusively by self-quenching, one can easily show (c/. equations on page 546, Vol. I) that the constant k has the meaning: (23.6) k = ke/kf where h is the bimolecular rate constant of self-quenching: K (23.7) Chi + Chi* > Chl2 ( > 2 Chi) and kf the monomolecular rate constant of fluorescence • (23.8) Chi* > Chi + hu The reciprocal, l/kj, is the average life-time of the excited molecule when limited only by fluorescence. According to page 634, k/ ^ 10^ sec.-i; with fc = 2 X 10-^ sec.-\ this gives, according to (23.6): (23.9) A-c ^2 X 1012 sec. -1 This calculation is based on the assumption that reaction (23.7) is the only one that limits the yield of fluorescence. However, equation (23.5) would remain correct, i. e., self-quenching would follow a hyperbolic curve, CONCEJSTKATION QUENCHING 773 also if fluorescence were competing, not only with the bimolecular reaction (23.7), but also A\-ith one (or several) mononiolecular (or pseudomonomolec- ular) transformations of the excited molecule, such as energy dissipation by internal conversion, or tautomerization, or a reaction with the solvent. In this case, the constant k in (23.5), instead of meaning (23.6), would have the meaning (23.10): (23.10) k = kc/ikf + kd) where ka is the mononiolecular rate constant of the competing energy- dissipating process (or processes), (corresponding to the sum fc,- -{■ kt, internal conversion plus tautomerization, in equation 19.5 in Vol. I, page 546). Furthermore, in this case, the constant C would not be equal to Fo (the fluorescence intensity with quantum yield 1), as assumed by Weiss and Weil-Malherbe, but would have the meaning: (23.11) C = FoA7/(A7 + kd) With the assumption C = Fo, the data of Weiss and Weil-Malherbe indi- cate a yield of fluorescence (

(n 0.9 ^ 0.8 0.7 0.6 0.5 0.4 o „ fQ05 M 1005 M KHCOj MaHCOj y^O.\M KHCOj A \ ©. \ {0,042 /W \ 10.058 M KHCOj NaHCOj ^0035 M KHCOj + On ■•© 0.065 /W Nat oA % ^^0>5-^ MaHCOj ^n"* ^ o ••. • ...... 1 1 III! 1 ' 1 ti 1 • 6 8 10 12 14 16 18 20 22 24 TIME IN BUFFERS, hr. Fig. 25.1. Relative rate of photosynthesi.s in Chlorella vulgaris as a function of the lengthof e.xposure of the cells to 0.1 M solutions of KHCOg, NaHCOs, or varying mixtures of the two bicarbonates (after Pratt 1943). gas supply during growth (Burk), or the "oligodynamic" effects of small quantities of rare elements icf. Emerson and Lewis 1939). However, some investigations on the quantum yield of photosynthesis (cj. chapter 29) indicate that these conditions may influence mainly the gas exchange in the first few minutes of exposure to light, and have little effect on the steady rate of photosynthesis (cf. Table 25.1). According to Emerson and Green (1938) the photosynthesis in Chlorella is insensitive to Avide variations in pH (Vol. I, p. 339) ; consequently, these algae can be used in acid as well as in alkaline solutions (Con- cerning the difference between quantum yields of Chlorella in aci'd and alkaline buffers, see pp. 1096, 1107.) Pratt (1943^) observed that the rate of photosynthesis of Chlorella in strong light declined to 40% of its initial value after nine hours in a 0.1 M NaHCOs solution and then remained 836 METHODS OF KINETIC MEASUREMENTS CHAP. 25 steady; while, in a 0.1 M KHCO3 solution, there resulted an increase in rate by about 30% within six hours, followed by several hours of steady oxygen production and then a rapid decline. In a mixture of 0.035 M KHCO3 and 0.065 M NaHCOs, the rate remained unchanged for about fifteen hours, after which the stimulating effect of the potassium salt ap- parently wore off, and the rate declined rapidly (fig. 25.1). In all these experiments, the cells spent about one half of the total time in light and one half in darkness. Among other algae used in quantitative photosynthetic studies were the Chlorophyceae: Scenedesinus {cf., for example, Gaffron 1942), Hormidium flaccidum (van der Honert 1930 and van der Paauw 1932), Stichococcus hacillaris (Aufdemgarten 1939); the Rhodophycea, Gigartina harveyana (Emerson and Green 1934) ; the Cyanophycea, Chroococcus (Emerson and Lewis 1942); the brown alga, Fucus serratus (Steemann-Nielsen 1942); and the diatoms, Nitzschia dosterium (Barker 1935, Button and Manning 1941), A^. dissipata (Wassink, Kersten 1945), Navicula minima (Tanada 1951). Advantages of algae as material for photosynthetic measurements (e. g., the convenient use of manometric methods) are shared to some extent by the higher aquatic plants. Among them, Elodea has been the most popular in photosynthetic studies, because of its widespread occurrence in stagnant waters. Other aquatic plants used in photosynthetic work were Cahomba caroliniana (Smith 1937) and Potomageion (Gessner 1937). Detached leaves provided the material for most of the earlier investigations of photosynthesis ; they were used by Blackman and co-workers, by Brown and Escombe (1905) and by Willstatter and StoU (1918) in their pioneer investigations of the quantitative aspects of photosynthesis. Because of the interruption of natural translocation processes, the time course of photosynthesis in detached leaves may differ from that in similar leaves attached to the stem (Vol. I, p. 332). Whole land plants, enclosed in glass vessels, were long used in investiga- tions of the rate of photosynthesis under field conditions. A group of workers at the Smithsonian Institution in Washington {cf. Hoover, Johns- ton and Brackett 1933 and McAlister 1937) showed that this method can give results equivalent, as to precision and consistency, to those derived from experiments with algae. The material used in their studies were single young wheat plants. The culture of purple bacteria that can be used for studies of bacterial photosynthesis has been described by van Niel and co-workers (1931, 1944), French (1937), Gaffron (1933-1935) and the Dutch group (Eyniers, Wassink, Katz, Dorrestein et at. 1938, 1942) ; the bacteria used included Rhodospirillvm rubrum, Streptococcus varians, and strains of Chromatium and Rhodovihrio. LIGHT MEASUREMENTS 650 m^; the two "warm whites," little energy <500 mju. The quanta of visible light are from 5 X 10"^^ to 2.5 X lO"^'' erg each. In sunlight, the intensity maximum lies at about 575 niM, corresponding to /^ = 3.5 X 10~i2 erg. Thus an illumination of 1 lux from the sun corre- sponds to 1.2 X 10^2 quanta or 2.0 X lO'^^ einstein/(cm.2 sec.) of photo- synthetically active light. In artificial light, X is much nearer the red end of the spectmm. If one assumes, for this light, hv = 3.2 X 10"^^ erg, 1 lux becomes equivalent to 1.4 X 10^^ quanta or 2.3 X lO^^^ einstein/ (cm.^ sec.) of photosynthetically active light. In direct sunlight at noon, about 10" visible quanta impinge per second on one square centimeter of horizontal surface. A pigment molecule situ- ated on this surface, whose average molar absorption coefficient for visible light is of the order of 3 X 10^ will absorb about twelve quanta every second. The frequency of absorption acts is determined, for a molecule with the molar ab- sorption coefficient a, by the equation: (25.1) ?i = 4 X lO-^'aA^;,;, where Nh^ is the light flux in quanta/ (sec. cm.''). Instead of calculating, as above, the energy flux and the number of quanta impinging on the illuminated surface from the intensity of illumina- tion in lux, it is of course much better to measure this flux directly, by means of a thermoelement, bolometer, photocell or actinometer. This is also the only way to define the intensity of colored light, which cannot be measured in lux. The energy flux can be expressed in ergs or calories (per unit surface and unit time) or watts (per unit surface). The relation be- LIGHT MEASUREMENTS 839 tween these units is shown in Table 25. III. However, without information as to spectral composition, the indication of the energy flux in ergs, or calories, is even less revealing than that of the intensity of illumination in lux, because 60% of direct sunlight and about 95% of the energy flux from incandescent lamps belong to the far red and infrared, and are not used by plants for photosynthesis. Unless the proportion of these radiations is known, quoting the energy flux may easily give an entirely erroneous con- cept of the quantity of light available for photosynthesis. Table 25.III Energy Flux Units* Units Watt/cm.2 Erg/(cm.2 sec.) Cal/(cin.2 sec.) Watt/cm.2 Erg/(cm.^ sec.) Cal./(cm.2 gg(j.) 1 10' 10-' 1 4.19 4.19 X 10' 0.239 2.39 X 10^» 1 Of the common photometric devices, only thermoelements and bolom- eters react uniformly to radiations of all wave lengths. All other instru- ments—vacuum photocells, barrier layer cells, actinometers— possess a selective spectral sensitivity. Some investigators suggested that instru- ments insensitive to infrared light, e. g., selenium barrier layer cells ("pho- tronic cells"), should be used in preference to thermopiles or bolometers in the measurement of light intensities in the work on photosynthesis, in order to avoid measuring infrared together with visible light. However, this is almost equivalent to a return to visual photometry, since the eye, too, can be described as an infrared-insensitive photometer. In fact, the sensitivity curve of the selenium barrier layer cell is quite similar to that of the human eye. Both drop rapidly above 600 m^— right in the middle of the main red absorption band of chlorophyll (c/. fig. 25.2A) . Vacuum type photocells also show strong variations in sensitivity with wave length (fig. 25.2B). No photocells, actinometers or photographic plates possess uni- form sensitivity throughout the visible spectrum (although some of them are reasonably constant in the region of the shorter waves). Therefore, no instruments of these types can give reliable photometric data in nonmono- chromatic light, (c/. Mestre 1935). Warburg and Schocken (1949) have developed an actinometer based on Gaffron's earlier observations in Warburg's laboratory of sensitized autoxidation of allyl thiourea (chapter 18, page 509). Thiourea was sub- stituted for allyl thiourea as oxidation substrate, and pyridine for acetone as solvent; it was found that, with ethyl chlorophyllide (or protoporphyrin) as sensitizer, a quantum yield equal to 1.0 ± 0.1 (molecules oxygen con- * We will use the abbreviation kerg foi- 1000 erg. 840 METHODS OF KINETIC MEASUREMENTS CHAP. 25 sumed per quantum absorbed) can be obtained over a considerable range of wave lengths and intensities. The convenience of this actinometer is the possibility of using it in conjunction with Warburg reaction vessels in a manometric sj^stem. Assuming equal absorption of light in the reaction 300 400 500 WAVE LENGTH, m/i 600 700 Fig. 25. 2A. Spectral sensitivity curve of a selenium barrier layer cell (Wes- ton Photronic Cell). 20 .^ 16 C D > 12 > i B lu V) 4 - G.E. Pj 22(Cs) G E. ly 405 (No) RC.A. 910 (photomultiplier) 200 300 400 500 600 WAVE LENGTH, m/i 700 800 Fig. 25.2B. S'lHictral .sen.sitivit.y curves of three typical vacuum-type photocells. vessel and in the actinometer vessel, and arranging for identical light fluxes to reach both vessels (e. g., by placing them side-by-side in a uniform light field, or by alternating the reaction vessel and the actinometer vessel in LIGHT MEASUREMENTS 841 the same position), quantum yield determinations can be made simply by comparing the pressure changes in the two vessels. Average quantum yields can be determined in this way for prolonged periods of illumination, or for illumination with diffuse light, more easily than with instruments which measure momentary light intensity, and require a collimated beam. This actinometer promises to become a very useful tool in the study of photosynthesis; hut despite its convenience it, too, must l)o used with cau- tion, and checked from time to time against a physical light -measuring in- strument such as a thermopile or bolometer. The nature and mechanism of the reaction in the Warburg-Gaffron-Shocken actinometer is unknown, and the exact dependence of its rate on light intensity, wave length, nature of the solvent, presence of impurities, oxygen pressure, concentrations of the reductant and the sensitizer, and rate of stirring, remain to l)e investi- gated. Available measurements indicate that the quantum yield declines slowly with increasing illumination in the "middle range" (3-7 X 10""' einstein/(sec. cm.-)), but changes faster both at the higher light inten- sities (>10 X lO-i*^ einstein/(sec. cm. 2)) and at very low light intensi- ties (<3 X 10-1" einstein/(sec. cm. 2)). It remains to be seen, however, whether the factor determining the change is intensity of the beam — i. e., energy per unit cross section — or its energy (more probably, energy ab- sorbed in unit volume of the liquid). Some decline in quantum yield with increasing illumination may be caused by exhaustion of oxygen in the il- luminated layer; however, this is likely to account only for a part of the observed trend. Another possible cause of this decline is competition of back reactions between the intermediate oxidation and reduction prod- ucts, with the "forward" reaction which leads to the consumption of oxy- gen. If they are bimolecular in respect to the intermediates, the back re- actions must be favored by a higher concentration of the latter and there- fore can become more effective at light intensity increases. It must be recalled that photometers with selective spectral sensitivity cannot be relied upon not only in the determination of absolute light inten- sities, but also in the comparison of two light sources or in the determina- tion of the proportion of light absorbed by passage through a colored sys- tem. Unless the absorption is very weak throughout the spectrum, the spectral composition of the transmitted light will be different from that of incident light, and the selectively sensitive instrument will react differ- ently to these two fluxes. It will tend to exaggerate the absorption if the latter takes place in the region of maximum sensitivity, and underestimate it if it occurs in the region of low sensitivity. The steeper the spectral sensitivity curve, the larger will be the errors that occur in absorption measurements in nonmonochromatic light. Selenium barrier layer cells, for example, cannot be used for such measurements even in "monochro- 842 METHODS OF KINETIC MEASUREMENTS CHAP. 25 matic" red light, isolated b}^ filters (or monochromators with wide slits), since their sensitivity droj s by a factor of 10 between 600 and 700 m^ (fig. 25.2A). For all these reasons, if white light is used for photosynthetic work, the best way of characterizing its intensity is to measure it by means of a ther- mopile protected from infrared light by a suitable filter. This will give an adequate picture of the quantity of light available for photosynthesis, and enable one to determine correctly the proportion of this light absorbed by the plants. The desire for greater sensitivity often will force the investigator to use a photoelectric cell, instead of a thermopile, despite all the shortcomings associated with its selective sensitivity ; this should be done only in full realization of the errors that can be introduced in this way. Only in work with truly monochromatic light are the photocells entirely reliable (as- summg that the linearity of their response has been ascertained by fre- quent comparison with a thermoelement). In addition to the problem of a reliable photometric instrument, dif- ficulty arises in the determination of the quantity of light absorbed by leaves, algae or cell suspensions, because of the scattering phenomena dis- cussed in chapter 22. The scattering by cell suspensions is comparatively weak, and that by leaves can be reduced by injection with water, or— still better— with gly- cerol (by evacuation under the liquid), thus eliminating the most effective source of scattering— the liquid-air interfaces (c/. fig. 22.8 and 9). How- ever, figure 22.2 shows that, even in Chlorella suspensions, enough scatter- ing is present to cause a marked error in the determination of absorbed light energy — an error of only a few per cent in the region of strong absorption, but of 100% or more in green or in the far red, where tme absorption is very weak. In measurements in these spectral regions, as well as in precision ex- periments in other parts of the spectrum, it is necessary to measure the total scattered flux {i. e., the diffusely transmitted and diffusely reflected fluxes, Ta and Ra, together with directly transmitted and specularly re- flected fluxes Ts and Rs) and to use the complete formula: (25.2) A = h - Ts - Td - Re - Rd = h - S for the determination of the absorbed light energy, A. The neglect of both T^ and R^ in the determination of A must lead to entirely erroneous results, because, for all leaves and thalli, Ts is only a fraction — and often a small one — of the total transmitted flux. The fact that optical work with leaves requires a consideration of scat- tering was clear to Maquenne (1860) and Simmler (1862) ; but some inves- tigators—not only botanists like Sachs (1864) and Detlefsen (1888), but LIGHT MEASUREMENTS 843 even physicists like Vierordt (1871) and Lazarev (1924, 1927)— thought that they could neglect it. In most measurements, however, an attempt was made to include at least the diffusely transmitted light, Ta, by the simple device of placing a large collecting surface immediately behind the absorbing system. Seybold (1932) pointed out that this procedure brings the risk of measuring the thermal radiation of the tissue together with the transmitted flux. (A similar error could be caused by fluorescence, but the latter usually can be neglected.) To avoid errors, one may inter- pose an infrared-absorbing filter between the leaf and the collecting ther- mopile. The measurement of the diffusely reflected flux Ed requires more elabo- rate devices and has often been omitted. The resulting error in the deter- mination of the absorbed intensity can be considerable, since leaves of land plants reflect about as much, or more, light as they transmit — namely, from 10 to 15,% of (infrared-free) white light (c/. page 683). Submerged algae or water-filled leaves have a lower reflectance — they transmit about 20% and reflect from 5 to 10% of white light. However, diffuse reflection cannot be entirely neglected even when working with algal suspensions, as shown by the results of Noddack and Eichhoff (1939) in figiu-e 22.2. The sharp reflection peak at 180° is due to the walls of the vessel; but, in addi- tion to this specular reflection, the figure shows a small, but not negligible, diffuse reflection ; integrated over all angles, it adds 3 or 5% to the trans- mitted flux and reduces correspondingly the absorbed energy, A. In work with cell suspensions in spherical or cylindrical vessels, the dis- tinction between reflected and transmitted light becomes irrelevant, and an integral measurement of light scattered in all directions, S, can be sub- stituted for the separate measurements of R and T. A small vessel con- taining the suspension can be placed inside an "integrating" box or cell, or in the focal point of a mirror, illuminated by a narrow beam of light entering through a hole in the mirror, and the light scattered in all direc- tions can be collected and measured. For the determination of /, a "white" scatterer can be substituted for the suspension cell. A device of this type was Noddack and Eichhoff's (1939) "ellipsoid photometer," in which the light scattered by a small cell was collected on a thermopile sensitive to light falling from all directions. The scatterer was placed at one focus of an ellipsoidal mirror, and the collector at the other. In speaking of the methods of determining the hght energy absorbed by cell suspen- sions, we must also mention Warburg and Negelein's method of total absorption (1922, 1923). These authors used a very concentrated Chlorella suspension; the vessel had a silvered back wall so that no light was transmitted; and the absence of diffusely re- flected hght was ascertained by experiment. (The correctness of this last assertion was questioned by Mestre 1935, and this criticism is supported by the above-mentioned re- sults of Noddack and Eichhoff.) Thus, \\'arl)urg and Xogelein assumed, simply, A = / 844 METHODS OF KINETIC MEASUREMENTS CHAP. 25 It was mentioned on page 673 that, in working with solutions, the neces- sity of estimating R is usually avoided by using a blank cell, whose reflec- tion is assumed to be equal to that of the solution cell. Many authors have hoped to get around the necessity of measuring Ra for leaves or thalli in a similar way, by using as "blanks" plant tissues deprived of pigments. This idea has been carried out in different ways: Reinke (1886) used algal thalli from which the pigment had been extracted by alcohol; Linsbauer (1901), Brown and Escombe (1905), Seybold (1932, 19331-2) and Meyer (1939) compared the transmission by green parts with that by white parts of variegated leaves; Wurmser (1921) determined the transmission of thalli before and after bleaching by prolonged illumination. However, the interpretation of results obtained in this way presents considerable dif- ficulties. It has already been said (page 673) that equation (22.2b) is only an approximation, although a satisfactory one, even in the work with transparent media. Weigert (1911) thought that it could also be used, as such, for leaves, and applied it to the data of Brown and Escombe; but his calculation led to absurdly low values of A, and its fallacy has been pointed out by Willstatter and StoU (1918) and Warburg (1925). In all precision experiments on light absorption by plants, measure- ments of the three quantities I, T and R cannot be avoided. The determi- nation of T and R can be carried out either by means of integrating devices that collect the reflected and the transmitted light, or by differential "gonio- photometric" methods, i. Y^ y ^ ■''>' ^/ ^ y / j.-?^-; '4 ,■'' f- y x' / y 1 I , t / / y '/ / / /. i // // 7 1 1 0 10 20 30 40 50 60 70 80 90 100 TRANSMISSION, % Fig. 26.5. Weakening of light in transmission through a leaf of Cyclamen persicum (after Srhanderl and Kaenipfert 1933). Scale at right of leaf cross-section, depth in microns. spongy parenchyma cells, for example, is under all circumstances consider- ably weaker than that in the paUsade cells (cf. fig. 26.5). Thus, in curves representing the rate of photosynthesis (P) as a function of carbon dioxide concentration or light intensity, the abscissae are mean values (averaged over the whole cell or over many cells). This alone must prevent these cur^^es from following the course suggested by Blackman, even if the law of limiting factors were exactly valid for the ideal case of a uniformly illum- inated and uniformly supplied homogeneous system. Katz, Wassink and Dorrestem (1942) derived an equation by means of which the experimen- tally obtained "light curves," P = /(7o) (where h is the intensity of the light falling on the front wall of a vessel with a suspension of algae or bac- 866 EXTERNAL AND INTERNAL FACTORS CHAP, 26 teria), can be recalculated to represent P in relation to the average light intensity, /, actually falling on an individual algal or bacterial cell (see p. 1009 and figs. 28.22A, B and C). 3. Some General Kinetic Considerations Some investigators, who realized the inevitable distortion of light curves and carbon dioxide curves of photosynthesis by the "depth effects" dis- cussed in the preceding section, assumed that in the measure in which these effects can be eliminated (experimentally, by the use of very dilute systems, or mathematically, by applying adequate corrections for inhomogeneity) the kinetic curves would approach the ideal "Blackman type." Undoubt- edly, the elimination of depth effects shortens the transitional region be- tween the ascending part of the curves and the saturation plateau; but figure 28.22C shows that it does not make the breaks sharp. Only a frac- tion of deviations from "Blackman behavior" can be attributed to inhomo- geneity; even with all "depth effects" eliminated, the kinetic analysis of photosynthesis will still have to contend with kinetic curve systems of all three types exemplified by figures 26.2, 26.3 and 26.4. The common feature of all these curves is the occurrence of saturation, i. e., of states in which the rate of photosynthesis is independent of the vari- able Fi. The saturation level may depend on the parameter F2 — as in the Blackman type and Bose type curve systems (figs. 23.2 and 23.3) — or it may be independent of F2, in which case curve systems of the "third type" (fig. 23.4) are observed. (a) Sources of Saturation in PhotosTjnthesis The over-all rate of a process consisting of a series of successive chemical or physical stages — sometimes referred to as a catenary series — cannot ex- ceed the rate of any of its individual steps. "Saturation" of such a process with respect to a given kinetic variable, Fi, is therefore reached whenever the over-all rate becomes equal to the maximum rate of a single step (a step which in itself must be independent of this variable). For example, under given conditions of external carbon dioxide pressure and temperature (perhaps also humidity and other factors affecting the colloidal structure of the cell), carbon dioxide can be supplied to the photosynthesizing cells at not more than a certain maximum rate, which is reached when the station- ary concentration, [CO2], at the site of photosynthesis is zero, and the dif- fusion gradient between the medium and the chloroplast has therefore the maximum possible value. This maximum rate of carbon dioxide supply by diffusion is independent of illumination (except for possible indirect rela- tions of the type mentioned on page 863). Therefore, "light saturation" SOME GENERAL KINETIC CONSIDERATIONS 867 should occur whenever the over-all rate of photosynthesis approaches the maximum rate of carbon dioxide supply by diffusion. In this considera- tion, the diffusion of carbon dioxide could be replaced bj^ a preliminary chemical reaction the rate of which is proportional to the concentration, [C02],e-g-, the formation of the compound !C02} from carbon dioxide and an "acceptor," which was postulated in chapter 8 (Vol. I). In this case, light saturation is determined by the maximum possible rate of formation of {CO2} that is reached when all acceptor molecules are free, (i. e., when all complexes { CO2 1 are utilized for photosynthesis practically instantane- ously after their formation). Similarly, "carbon dioxide saturation" must occur whenever the over-all rate of photosynthesis approaches the rate of supply of light energy, and the quantum yield assumes its highest possible value. There are other factors, besides carbon dioxide supply and the supply of light energy, which also can impose "ceilings" on the over-all rate of pho- tosynthesis and thus cause saturation phenomena. This role can be played, e. g., by the concentration of any one of the several catalysts participating in photosynthesis (including the "photocatalyst" chlorophyll). For ex- ample, if one reaction step in the "catenary series" of photosynthesis is the monomolecular transformation of a catalyst-substrate complex: (Catalyst + Substrate) > Catalyst + Product the maximum rate of this reaction step is reached when all the available catalyst molecules are loaded with substrate molecules. When photosyn- thesis proceeds at a rate that requires such maximum utilization of one cata- lyst, variations in most kinetic variables (such as the concentrations of the reaction partners or of the other catalysts, or light intensity) cannot increase the rate any further. Only a rise of temperature can lift the "ceiling" im- posed by the maximum velocity of such a catalytic transformation. The part of a "rate-limiting" catalyst can be assumed by any of the several catalysts the existence of which was postulated in Volume I {cf. chapters 6, 7 and 9) — for example, the "carboxylase" Ea, the "stabihzing catalyst" (a "mutase" ?) Eb, or the "deoxygenases" Ec and Eq. Chlorophyll can play a similar role, c. g., if the primary photochemical process involves a chemical change of this pigment, and a certain time is required for its restor- ation. "Acceptors" or "carriers," such as the carbon dioxide acceptor postulated in chapter 8 (Vol. I) are catalysts, too, and the available quan- tity of any such auxiliary compound also can serve as a "limiting factor" in photosynthesis. All these factors can — and most of them probably do — contribute to the limitation of the rate of photosynthesis under different conditions, thus causing the "saturation" of this rate wath respect to vacious kinetic vari- 868 EXTERNAL AND INTERNAL FACTORS CIL\P. 26 ables. There is no theoretical reason to expect, and no actual experience to indicate, that saturation with respect to, say, light intensity, or carbon dioxide concentration, is always caused by the same "limiting factor." To the contrary, clear indications can be found of saturation phenomena caused by slow diffusion, slow carboxylation, various catalytic deficiencies, limited light supply and limited supply of reductants (in bacterial photosynthesis) . One may consider this as evidence of good adjustment of the photosynthetic process as a whole, since it means that the different parts of the complex mechanism of photosjTithesis have approximately the same maximum capacity. One understands, in the light of this multiplicity of possible rate-limiting steps, why repeated attempts to represent the kinetics of photosynthesis by means of models consisting of a small number of reac- tions, e. g., of one light reaction and one dark reaction only, could not have led to more than very limited success. Saturation with carhon dioxide is reached at pressures three or four times higher than the partial pressure of carbon dioxide in the free atmosphere (c/. Table 27.1); while saturation with light is reached at light intensities equivalent to 10-100% of full midday sunlight (c/. Table 28.1) and thus about equal to the average light intensity to which freelj^ growing plants are exposed in nature. The optimum temperature of photosynthesis is some- what above the average summer temperature, at least in temperate zones. An approximate adjustment of the photosynthetic mechanism to natural conditions is thus obvious. Perhaps, this adjustment was achieved in times when both the average temperature and the carbon dioxide content of the atmosphere were somewhat higher than they are now. However, this inference is by no means certain ; nature has seldom been able to de- velop ideal solutions of its adaptation problems, and is usually satisfied with more or less rough approximations. The present kinetics of photo- synthesis may have been the best plants were able to evolve in response to the now prevailing climatic conditions. (h) Origin of Kinetic Curve Systems of Different Types The distinctive feature of "Blackman type curves," described on page 860, are closely coincident initial, linear parts of the cui'ves. This charac- teristic distinguishes them from the two other t}^es of curve systems, repre- sented in figures 26.3 and 26.4. On the other hand, Blackman curves (fig. 26.2) and Bose curves (fig. 26.3) have in common a wide spacing of saturation plateaus, v.hile curves of the "third type" in figure 26.4 all ap- proach a common saturation level. Coincidence or separation of the saturation levels depends on whether saturation is imposed by the same parameter ¥-2, to which the set of curves SOME GENKRAL KINETIC CONSIDERATIONS RHO refers, or by some other parameter, Fg, F4 . . . For example, in a set of curves, P = /[CO.], for various values of light intensity, I, the saturation levels will be well separated (as in figs. 26.2 and 26.3) if saturation is im- posed by the rate of light supply, but will coincide (as in fig. 26.4) if satura- tion is due to the limited rate of a dark, catalytic reaction. Coincidence or divergence of the initial slopes depends on the qualitative and quantitative relationships between the variables Fi and F2. If both these variables affect the rate of the same reaction step, the rate will gener- ally depend on both of them. For example, if Fi is the concentration, [CO2], and F2 is temperature, and if the slope of the ascending part of the curves P = f [CO2] is determined by the velocity of carbon dioxide m\)p\y by diffusion, this slope will depend on temperature. If, on the other hand (a) the factors Fi and F2 affect different steps in the "catenary series" (e. g., if Fi is light intensity and F2 is temperature) and (6) the relative values of Fi and F2 are such that the process affected by F2 is far below its maximum rate when that affected by Fi approaches its maximum rate- then (and only then) the reaction rate will be a function of Fi alone, and practically independent of F2. The second condition (b) is important. Contrary to the way in which the concept of "limiting factors" or "rate-determining reaction steps" often is used, the existence of a reaction step of limited maximum efficiency af- fects the rate of the over-all reaction long before the rate actually "hits the ceiling." Therefore, if several reaction steps have maximum rates which are not too different in their order of magnitude, the rate of the over- all reaction must be affected by all these "potential bottlenecks" and not only by the "narrowest" one. This fact was recognized in the alterations of Liebig's "law of the mini- mum," which we have mentioned on page 862, and some quantitative illus- trations of it will be given later in our analytical discussions. To make it plausible, we may use a mechanical analogy; the flow of water through a system of pipes with several strictions depends on the diameter and length of all of them, and not only on the one that has the greatest flow resistance. In the plot of P against Fi, for different values of F2, we can expect all cui-ves to be to the right of, and below, two "limiting" lines: one a slant- ing "roof" imposed by the maximmn rate of the "slowest" reaction step the rate of which is proportional to Fi (or, more generally, is a function of Fi) ; and the second a horizontal "ceiling" imposed by the maximum rate of the slowest reaction step the maximum rate of which is independent of Fi. For example, in the case of curves P = f [CO2], for different tempera- tures, the two limiting curves may be determined by the maximum rate of diffusion of carbon dioxide and the rate of light absorption, respectively (cf. fig. 26.6). These two limiting lines together form a typical "Blackman 870 EXTERNAL AND INTERNAL FACTORS CHAP. 26 curve"; but they are merely limits within which the actual kinetic curves are confined. The point we are trying to make — in advance of its analyti- cal proof — is that it would be wrong to imagine that the existence of the "roof" OA and the "ceiling" BC will leave unaffected the curves situated entirely within the "permitted area" OXC, and merely force back into this area the curves (or section of curves) that would otherwise cross the limit- Fig. 26.6. "Roof" and "ceiling." OA, maximum rate of a diffusion step (proportional to [CO2]); OA ', maximum rate of carboxylation (proportional to [CO2]); BC, maximum rate of primary photochemical process ( = rate of light absorption) (independent of [CO2]); B'C, B"C", maximum rates of catalytic reactions. ing lines. To the contrary, the existence of the "roof" and the "ceiling" — as well as that of the potential, higher "roofs" and "ceilings" {OA' , B'C, B"C" in fig. 26.6) — will push down and toward the right even the kinetic curves that would not have approached the limiting values if the limitations were absent. We will have opportunity for analytical proof of these statements in subsequent chapters. For example, in chapter 27 we will derive "carbon dioxide curves" of photosynthesis from simple models of the mechanism of entry of carbon dioxide into the photosynthetic reaction ; and we will find there that the sequence of the two reactions: SOME GENERAL KINETIC CONSIDERATIONS 871 (26.1) CO: + A . ' ^ ACO2 — > A + {HCO2} (i. e., a reversible binding of carbon dioxide by an acceptor A followed by reduction of the complex ACO2 by a photochemically produced reductant {H}) leads to a "Bose type" system of curves P = /[CO2] (where Pis the rate of the over-all reaction). If carboxylation is so rapid (compared with photosjTithesis) that its equilibrium is not disturbed even in strong light, these "carbon dioxide curves" will be hyperbolae, which diverge from the origin, and remain in a constant ratio up to saturation. Saturation corre- sponds, in this case, to complete carboxylation of all the available acceptor, and is therefore reached at the same value of the variable [CO2] in all curves. The saturation rate rises with increasing concentration of the reductant {H} (and therefore also with light intensity, since w^e assume that { H } is produced by light) . We will further see that, if carboxylation is a slow process the rate of which is proportional to [CO2], a [C02]-proportional "roof" is imposed on the rate of the over-all reaction, namely : (26.2) Pmax. = A-„Ao [CO2] where Ao is the total concentration of the acceptor*, and ka the rate constant of carboxylation {cf. equation 26.1). We will show that, because of this "roof," the curves P = f [CO2], which would otherwise begin with the slope k^Ao {i. e., which would stay just inside the "permitted area") will be reduced to an initial slope half as large, i. e., kaAo/2. More gener- ally, curves that, without limitation, would have begim with a slope ak^Ao will be reduced to an initial slope k^Aoa/(a + 1). Thus cui-ves that in the case of rapid carboxylation would begin with slopes between 10 k^Ao and 100 kaAo, would all be confined, in consequence of slow carboxylation, to slopes between 10/11 fc^Ao and 100/101 fc„Ao, and would thus present a "Blackman picture." If the carboxylation product, ACO2, in equation (26.1) has to undergo a monomolecular transformation before it can react with {H}, this would impose a [C02]-independent ceiling on the rate of the overall reaction, namely : (26.3) Pmax. = kiAo where ki is the rate constant of the postulated monomolecular transforma- tion, and Ao the total available quantity of A. As a result of this "ceiling," the curves P = f [CO2], which would otherwise reach a saturation value of P = kiAo, will be reduced to a saturation level half as high, P = A;iAo/2. * We omit square brackets in the designation of constant concentration, i. e., we write Ao, Chlo, etc., instead of [A]o, [Chl]o, etc. 872 EXTERNAL AND INTERNAL FACTORS CHAP. 26 The curves that would other\vise reach a saturation value high above fciAo will be crowded into a narrow space immediately below fciAo, and the curves with saturation values below fciAo, will all be depressed. A curve the satu- ration level of which would otherwise be iSfciAo will be reduced by the im- position of the ceiling to a saturation level of ^kiAo/(0 + !)• It follows from this formula that, even if the saturation rate without the limitation were only one tenth of the maximum rate of the postulated monomolec- ular transformation of ACO3, it would nevertheless be reduced by 10% by this "potential bottleneck." 4. Internal Factors and the "Physiological Concept" of Photosynthesis The belief, referred to on page 860, that the rate of photosynthesis must obey a simple and rigid kinetic law has been partly responsible for the conclusion, reached by some plant physiologists, that it does not obey any recognizable kinetic law at all. These physiologists, disillusioned by the over-simplifications in which the preceding generation had indulged, turned their attention to the complex relations between photosynthesis and other phenomena of plant life, such as nutrition, respiration, growth or aging. Factors often referred to as "internal" or "plasmatic" are supposed to be responsible for these relations. Stressing the prime importance of these factors in the regulation of photosynthesis, Kostychev and his pupils minimized or denied the direct influence of the easily regulated "external" factors, such as light intensity, CO2 concentration and temperature. Their revolt against the apphcatioa of kinetic laws to photosynthesis and similar processes would perhaps be less violent if these physiologists would have reaUzed that the "law of hmiting factors" is by no means the last word in the physicochemical approach to photosynthesis, that, in fact, the concept of "hmiting factors" is foreign to reaction kinetics. The belief that no other kinetic laws are possible led Chesnokov and Bazyrina (19.30) to argue that since, according to Harder and Lundegardh, the factors "carbon dioxide concentration" and "light intensity" are not "truly limiting" (z. e., that often tne change in either of these factors can affect the rate) the "true" limiting factors must be sought inside the plant. It did not occur to them that photosynthesis may have no "true" limiting factor at all. Kostychev, in an article called "A New Concept of Photosynthesis" (1931), suggested that "external" factors affect photosynthesis mainly, if not exclusively, in an indirect way, by stimulating or inhibiting certain unknown ])lasmatic activities. All the conclusions obtained by Blacknian (and others l^efore and after him) on the basis of the "ph\sieochenii(ud" approach were rejected Ijy Kostychev as spurious. A similar point of view was taken by van der Paauw (1932) and by Kostychev's co-workers, Chesnokov and Bazyrhia (1930, 1932), who sought to prove by experiments that two external factors — carbon dioxide concentration and temperature — have no direct effect on the rate of photosynthesis at all, and that the third RATK UNDER CONSTANT CONDITIONS 873 one — light intensity — affects this process only partly by direct action, and partly through plasmatic stimulation. In the U.S.S.R., there is a tendency now to consider this point of view as the only one in accord with dialectic materialism, and attempts to isolate photosynthesis from other functions of the Uving organism and study it as an independent photochemical reaction are criticized as "mechanistic."* Such dogmatic assertions, practically banished from physics and chemistry, but still recurring in biological sciences, particu- larly in the U.8.S.R., are strangely beside the point. A reaction in a living organism is distinguished from that in a test tube by the complexity of the system in which it takes place. This complexity is due to three causes: the impossibility of separating the reacting system from other components of the organism; its inhomogeneous structure; and its complex and largely unknown chemical composition. Unprejudiced experiments alone can prove whether, despite these handicaps, direct relationships can be established between external kinetic variables and the rate of the specific process under investigation. If this proves possible, a promising approach to the understanding of the process is opened, and it would be foolish to refuse to use it because of dogmatic objections. 5. Rate of Photosynthesis under Constant Conditions. Midday Depression and Adaptation Phenomena Obsei-vations that lend support to the "physiological" concept of photo- synthesis include, among others: the difference in photosynthetic activity, under identical external conditions, of plants grown in various habitats; the adaptation of photosynthetic activity to changed conditions (stronger or weaker light, higher or lower temperature — cf. for example, Harder 1933 and Brilliant 1940); the effects of aging; and the changes of photosyn- thetic activity under constant external conditions (fatigue, midday depres- sion etc.) . Not only do plants of different species behave differently under identical external conditions, but variations are found also between "sun plants" and "shade plants" of the same species, "sun leaves" and "shade leaves" on the same branch and even between different parts of the same leaf (c/. Drautz 1935). The photosynthetic activity of a plant often changes strongly in the course of a smgle day, not to speak of a whole season. Obsei-vations of diurnal changes have played an especially important role in the development of Kostychev's "new concept" of photosynthesis. Offhand, one would expect photosynthesis to increase steadily after sunrise until the light intensity has reached the saturating value, and then remain more or less constant, unless cloudiness decreases the illumination below the saturating intensity, until the evenmg decline sets in. The actual behavior of the plants often follows, however, a much more complicated * An interesting monograph, Photosynthesis as Life Process of the Plant, by Miss Brilliant (1947) seems to be the only comprehensive review of photosynthesis published in Russian in recent years. It contains a survey of about 200 Russian and 350 other papers, many of them not utilized in the present book. 874 EXTERNAL AND INTERNAL FACTORS CHAP. 26 pattern. Thoday (1910) discovered that the production of organic material by leaves may show a temporary decHne in the middle of the day; the plant takes an "afternoon nap" (c/. fig. 26.7). McLean (1920) observed that this decline may even result in the release of carbon dioxide. The gas exchange measurements of Kostychev and co-workers (1926- 1931) (cf. also Chesnokov, Bazyrina and co-workers 1932), carried out under a large variety of climatic conditions, from Central Asia to the Arctic Sea, and with algae as well as with land plants, showed that the phenom- enon of "midday rest" is widespread in the plant world, but that it assumes 8 II 13 14 17 TIME AFTER MIDNIGHT, hr. 20 Fig. 26.7. Diurnal course of photosynthesis of two leaves of Erio- botrya japonica under natural conditions (after Kursanov 1933): solid line, May 30; broken line, June 6. the extreme form of a reversal of photosynthesis and evolution of carbon dioxide only under special conditions, particularly in hot climates. The diurnal course of photosynthesis has received the attention also of numerous other investigators, among whom we may mention Geiger (1927), Montfort and Neydel (1928), Maskell (1928i), Hiramatsu (1932), Harder, Filzer and Lorenz (1932), Bosian (1933), Kursanov (1933), Stalfelt (1935), von Guttenberg and Buhr (1935), Monch (1937), Filzer (1938), Neuwohner (1938), Neubauer (1938), B. S. Meyer (1939) and Bohning (1949). Harder (1930) and co-workers, Schoder (1932) and Drautz (1935), as well as Neuwohner (1938), have attempted to explain the diurnal curves by combined variations of several external factors (light intensity, humidity and temperature) . They succeeded only partially, and had to admit that a considerable part of the observed variations remained unexplained and had to be attributed to unknown "plasmatic" factors. This is particularly clearly demonstrated by the observations of Filzer (1938), who found that leaves, picked from trees at different times of the day and then investigated under constant conditions in the laboratory, showed the same periodic changes in photosynthetic production as did leaves left attached to the plant and exposed to the natural change of night and day. Similarly, Mas- MIDDAY DEPRESSION 875 kell (1928^) found that detached cherry laurel leaves, illuminated with con- stant light for 24 hours, showed a deep depression of photos>Tithesis during the night hours; thus, not only the "midday nap," but also the "night sleep" appears to be influenced by internal factors. Geiger (1927), Maskell (19281-2) and St^lfelt (1935) considered the closure of the stomata as the immediate cause of the middaj^ depression. Maskell (1928^) observed that the nightly depression of photosynthetic activity of steadilj^ illuminated leaves can be avoided by increasing the jmrtial pressure of carbon dioxide ("pressing carbon dioxide through half- closed stomata"), and that steadily illuminated leaves of Hydrangea (the stomata of which are almost rigid) showed only a slight decline of photosynthesis during the night hours. Both Maskell (1928^) and St§,lfelt (1935) found a parallelism between the average aperture of the stomata and the rate of photosynthesis (of. chapter 27, page 910). It thus seems plausible that the diurnal rhythm of photosynthesis of the higher land plants is to a large extent conditioned by stomatal move- ments. The question remains, however, what causes the stomata to close at certain times of the day, even though the illumination and the carbon dioxide supply are kept constant? One "internal factor" that has been much discussed in connection with this problem is the accumulation of (soluble or insoluble) carbohydrates. (Concerning the effect of excess carbohydrates on the rate of photosynthe- sis, see chapter 13, Vol. I.) The midday depression may be a pause during which these materials are translocated or partially combusted. This ex- planation, first accepted by Kostychev, Kudriavtseva, Moisejeva and Smirnova (1926) and Kostychev, Bazyrina and Chesnokov (1928), was later rejected by Chesnokov and Bazyrina (1930^), w^ho found that plants with entirely different diurnal course of translocation may nevertheless show the same diurnal course of photosynthesis. It was on the basis of results such as this that Kostychev (1931) finally reached his extreme con- clusion concerning the purely physiological regulation of photos\Tithesis. Against these findings of Kostychev and co-workers, Kursanov (1933), von Guttenberg and Buhr (1935) and Monch (1937) confirmed the existence of a relation between the accumulation of sugars and starch and the diurnal rhythm of photosynthesis. However, according to von Guttenberg and Buhr (1935) and Neuwohner (1938), no smgle ex-planation can be made to fit all cases of midday depression. In some cases (e. g., in young leaves in spring) it is clearly traceable to the choking of the photosynthetic ap- paratus with carbohydrates. In other cases {e. g., in summer leaves on hot days) the loss of water and the ensuing closure of stomata provide the most plausible explanation. Accumulation of half-oxidized products which "narcotize" the photosynthetic apparatus, as suggested in Franck's theory 876 EXTERNAL AND INTERNAL FACTORS CHAP. 26 of induction (chapter 33) , is another type of mechanism which must be taken into consideration. Drop of the CO2 content of the air (Bohning 1949) and enhanced CO2 supply through the roots (p. 910) also have been blamed for the midday depression. The phenomenon of midday depression was found by Montfort and Neydel (1928) also in stomata-free ferns, and by Kostychev and Soldaten- kov (1926), Kursanov (1933) and Neubauer (1938) in algae. Gessner (1938) found no pronounced midday decline of photosynthesis in the higher aquatic plants (Elodea, Potomageton etc.), except with shade- adapted species or individual plants in which it could be interpreted as "inhibition by excess light" {cf. Volume I, page 535). Although minor fluctuations of the rate remained unexplained, the rate of photosynthesis in Gessner's submerged plants generally followed the changes in the intensity of illumination. The maximum of photosynthesis was often found in the early afternoon rather than at noon, but this could be explained as a tem- perature effect. Chesnokov, Giechikhina and Jermolayeva (1932) found that respiration, too, has a complicated diurnal rhythm. Because of this, the true rate of photosynthesis cannot be obtained by applying a uniform respiration correction to the net rate of oxygen liberation measured at different times of the day. They also found that the respiration of leaves often is much stronger than was generally assumed before. In young leaves, in particu- lar, the rate of respiration may approach that of photosynthesis. This explains why the rate of carbon dioxide liberation by some plants during the midday depression was found to be almost as large as the rate of the carbon dioxide consumption by photosyn- thesis before and after this rest period. However interesting the phenomena of the diurnal rhythm of photo- synthesis, and similar "physiological" effects (such as aging, fatigue, etc.) may be, the primary question for a kinetic study of photosynthesis is not whether these variations can be explained, but whether they can be elimi- 7iatcd, and photosynthesis made to proceed at an even and reproducible rate. It is difficult to realize such steadiness in field experiments. Boysen- Jensen and Miiller (1929), Boysen-Jensen (1933) and Mitchell (1936) said that, if conditions are reasonably constant, rate of photosynthesis in natural surroundings remains steady ; but Maximov and Krasnosselskaja-Maximova (1928) and Waugh (1939) observed that the photosynthetic production of leaves on the tree fluctuated, under constant external conditions, in suc- cessive four minute periods, by as much as =*= 100% of the hourly average. Kostychev (1931) concluded from these observations that measurements of photosynthesis over short periods have no meaning, and that the minimum time over which photosynthesis should be measured is a whole day! Stocker, Rehm and Paetzold (1938) found that rapid changes of the rate of jjhotosynthesis under natural conditions (fluctuation period: RATE FLUCTUATIONS 877 several minutes) are associated with similar fluctuations of the carbon dioxide concentration in the air. Scarth, Loewy and Shaw (1948) observed that the photosynthesis of detached leaves, determined by measuring the infrared absorption of carbon dioxide, occasionally showed unexplained, regular fluctuations (with a period of the order of 1 hour) . When plants are investigated under natural conditions, the apparently erratic behavior can be attributed to the difiiculty of controlling all the relevant factors. However, considerable doubt has also been expressed as to the capacity of plants to carry out photosynthesis at a constant rate under controlled conditions in the laborator>\ Experiments with lower plants, e. g., unicellular algae, such as Chlorella, have given comparatively satisfactory results: If certain prescriptions concerning culture and treat- ment were adhered to, these algae could be relied upon to maintain a con- stant and reproducible rate of oxygen production for several hours (leav- ing aside the short time induction phenomena to be discussed in chapter 33). According to Pratt (1943^), when alkaline buffers are used, the con- stancy of the rate depends on the nature of the cation present: Thus, in 0.1 M NaHCOs, the rate declined during the first 10 hours and then be- came steady; in M KHCO3, it increased during the first 10 hours, remained steady for the next 5 hours and then declined rapidly; in 0.065 M Na- HCOs + 0.035 M KHCO3, the rate remained steady for the first 15 hours and then began to decline (see fig. 25.1). Noddack and Eichhoff (1939) stated that for a given suspension of Chlorella, the rate of photosynthesis is reproducible within ±10%, and that these variations can be further reduced by preliminary adaptation of the cells to the light intensity in which they are to be studied. Experiments with higher land plants or aquatic plants have been con- tradictory, and at first rather discouraging. Tnie, Willstatter and StoU (1918) found that detached leaves, properly supplied with water and car- bon dioxide, maintain a constant rate of photosynthesis (within a few per cent) for 4 or 6 hours, even in strong light (40,000 lux) ; but Harder (1930, 1933), Arnold (1931) and Jaccard and Jaag (1932) asserted that strong trends as well as irregular changes develop in the photosynthesis of aquatic plants kept imder constant external conditions. Arnold obsei^ved, e. g., that in moderately strong light (18,000 lux) the rate of photosynthesis of Elodea dropped, in 2 or 3 hours, to one fifth or one tenth of its origmal value; at 4000-6000 lux, it increased for the first 2 or 3 hours and then decreased slowly ; only at 2000-3000 lux did it remain approximately con- stant for several hours. Harder (1930, 1933) made similar observations and stated that light intensity must be measured relative to the intensity to which the plants have been "acclimated" before the experiment. By 878 EXTERNAL AND INTERNAL FACTORS CHAP. 26 changing the ratio of these two intensities, he obtained the family of curves represented in figure 26.8 — some showing a steady increase in photosynthe- sis with time, others exhibiting an equally steady decline. He interpreted these time phenomena in terms of three plasmatic effects: "activation," Fig. 26.8. Changes of photosynthesis with time in aquatic plants under constant conditions. Numbers indicate increasing ratios between the conditioning and the illuminating intensity (after Harder 1930). "deactivation" and "exhaustion" (the last two being distinguished by the duration of the dark period required for recovery) . Experiments by Gessner (1937) and Steemann-Nielsen (1942) indicated, however, that only the phenomenon of Induction (Harder's "activation") is of fundamental nature (it will be dealt with in chapter 33), whereas "deactivation" and "exhaustion" can be avoided — at least as far as water plants in light of about 40,000 lux are concerned — ^by preventing the stag- nation of water, and the consequent dwindling of the carbon dioxide supply. o....^^ s L 3 TIME, hr. Fig. 26.9. Constant photosynthesis of a light-adapted plant (L) and a shade-adapted plant (S) of Elodea canadensis (after Gessner 1937). That stagnant water in the immediate neighborhood of assimilating plants can easily be depleted of carbon dioxide (even if it contains reserves in the form of carbonates and bicarbonates) was first i)roved by the calculations of Romell (1927). Figure 26.9, taken from Gessner, shows the time course RATE UNDER CONSTANT CONDITIONS 879 of photos3^lthesis of two twigs of Elodea canadensis, one adapted to strong light and another to weak Hght, in a steadily renewed medium. Both show good constancy for many hours of uninterrupted illumination (after an initial induction period in the shade plant) . Some of Gessner's experiments were extended over 6 days, with rate variations remaining within ±25%. If to these results of Gessner with aquatic plants we add those of War- burg and his successors with unicellular algae, and of Hoover, Brackett, and Johnston (1933), Mitchell (1936) and Bolming (1949) with higher land plants, there appears to be no fundamental difference between plant 300- N O n E e in >- 0, 0.9- UNITS EXTRACT, log scale 0.1 0.5 1.0 5 10 100 0 - (A) 1 9 ^"*"~~~-^^^^ 8 ^^~~--- -;.^ 7 - • 6 'i 1 1. 1 1 1 1 0.7- -1.0 -05 0.0 as 1.0 1.5 LOG UNITS EXTRACT A 2.0 UJ X 0.9 - ^ , 5- — ° y\^,^ - 1— 0.8 - (B) /^ ^^^^^ z >- // ^ ■^^ CO2 + OH- or partially, according to the equation : 2 HCO3- > CO32- + H2O + CO2 In Volume I (page 157), we said that Arens' results are in need of ex- perimental verification, and that, if they prove to be correct, they may be explained (a) by the diffusion of ions through the leaf without penetration into the interior of cells, and (6) by cell wall penetration l:»y neutral salt molecules, such as KHCO3. In connection with the latter possibility, it would be important to obtain quantitative information on the rate of pene- tration of bicarbonate as compared to that of free carbon dioxide ; conceiv- ably, the observations of Arens could be explained even if the first rate is only one hundredth or one thousandth of the second one, and thus negligible from the point of view of the kinetics of photosynthesis. By considering the penetration problem as a quantitative rather than qualitative one, we can anticipate that the influence of the external con- centrations [HCOa"] and [COs^"] on the carbonic acid system inside the cell will depend on whether we deal with an approximate equilibrium (z. e., work in the dark, or in low light), or with a photostationary state in which carbonic acid is rapidly consumed by photosynthesis. In the first case, even a very slow penetration of salt molecules may result in considerable changes in the composition of the cell fluids, while, in the second case, the effect of such slow penetration may be completely negligible in comparison with that of the much more rapid flow of CO2 molecules. More recently, Steemann-Nielsen (1946) revived Angelstein's (1911) argument. He studied the rate of photosynthesis as a function of the con- centration [CO2] in the medium, using the two aquatic plants Myriophyl- lum spicatum and Fontinalis antipyretica. In Fontinalis, the rates of oxygen liberation found in alkaline solutions (pH 8.3, containing from 0.5 X 10-^ to 5 X 10-3 mole HCO3- per liter, together with from 0.5 X 10"^ to 5 X 10-^ mole CO2 per liter) were hardly different from those observed in acid solutions with the same amount of CO2, but practically no HCO3- ions. A significantly different result was obtained with MyriopMjllum: In this plant, the yield of photosynthesis in alkaline bicarbonate solutions was ten times higher than in acid solutions with the same content of CO2 mole- cules! In some alkaline solutions the rate of oxygen liberation by Myrio- phyllum was as much as one third of the rate found in acid solutions with the same total concentration ([CO2] + [HCO3-]). This was interpreted by Steemann-Nielsen as indication that Myriophyllum uses IICO3- ions directly with about one third the efficiency with which it uses neutral CO? molecules. The difference between the two species, Fontinalis and Myri- ophyllum, was tentatively related by Steemann-Nielsen to the fact that CARBON DIOXIDE MOLECULES AND CARBONIC ACID IONS 889 Myriophyllum grew in a locality where the pH in summer was as high as 9-10 (corresponding to a ratio of 100 HCOs" to 1 CO2), while Fontinalis was gathered in a locality where the water contained very Httle bicarbonate, but as much as 30 X 10"^ mole per liter of free CO2 molecules (a remarkably high figure, if one recalls that water equilibrated with the free atmosphere contains only about 1 X 10 ~^ mole per liter CO2). The behavior of the two species is illustrated by figures 27.1 A and B. One can either accept these figures as evidence of direct participation of E E Q. O c 0) u 0) Q. o 3 O o Q. wu 1 / t5= 0 50 1 / ^ / o\° ! / ^ ; / f 1 1 '•" c?;^' (6) / "N ° / ./c,* 1 1 1 1 1 05 10 10 20 30 40 50 60 CO, (total), in 10'' mole/l 0.5 10 5 10 CO2 (total), in 10"' mole/l. 15 Fig. 27.1. (A) Photosynthesis in Fontinalis antipyretica, at 15,000 lux, 22° C, as function of total carbonate concentration; (B) Photosynthesis in Myriophyllum spicatum, at 37,000 lux, 20° C, as function of total carbonate concentration (after Steeraann- Nielsen 1946). HCOs" ions in the photosynthesis of Myriophyllutn (or, more exactly, of comparatively easy penetration of HCOs" ions into the cells of this plant), or one must assume that, for some reason yet unknown, the buffering action of bicarbonate was much more effective with Myriophyllum than with Fo7itinalis. It will be noted that the "free carbon dioxide" curves are practically identical for both species; in other words, if we attribute the shapes of these curves to carbon dioxide depletion, we must assume that the latter has been equally strong in both experiments. We must then assume that addition of about 100 HCOs" to 1 CO^ has had very little effect on (lei)letion in the case of Fontinalis, but had reduced it by about one third in that of Mijrioplujllum. Whether this explanation is plausible is difficult to say without the knowledge of various relevant factors, such as the absolute rate of photosynthesis per unit area, shape.s of the two plants (ratio of volume to surface; cf. page 905) and the efficiencj' of stirring. In chapter 13 (Vol. 1) we mentioned the effect of canons on the rate of photosynthesis, in particular the difference between the rates in sodium carbonate and potassium carbon- ate buffers, as observed by Pijson and later by Pratt. Some additional information on the unfavorable effect of sodium bicarbonate was given in chapter 25. Steemann-Nielsen 890 CONCENTRATION FACTORS CHAP. 27 (1946) enlarged these observations by determining the carbon dioxide curves of photo- synthesis in solutions containing different cations. He found that, with Myriophyllurn spicatum, in acid solution, the presence of sodium or calcium (in the form of chlorides) had no effect on the rate of photosynthesis. In alkaline solutions, on the other hand, the rate was lowest in sodium bicarbonate, higher in potassium bicarbonates and still higher in calcium bicarbonate. The highest rate could be obtained in a solution con- taining K+, Na+, Ca2+, Cl~, and S04^~ ions in the same proportion as the water of the lake that was the natural habitat of the plants. In lake water, the rate was independent of pH between 8.5 and 10.5, while in potassium carbonate-bicarbonate buffer (10 ~' mole HCOs" per liter) the rate increased slowly between pH 8.4 and 10.5 and dropped sharply to zero at pH 1 1 . These results can be interpreted in terms of Steemann-Nielsen's concept of direct participation of bicarbonate ions in photosynthesis (for example, by assuming different rates of penetration of different neutral molecules, MeHCOs, through the cell membrane; cf. Vol. I, page 197), but they may also be of a more indirect and complex origin. Ac- cording to Pratt {cf. fig. 25. 1 ) the cation effects are largely irreversible, a complication not considered by Steemann-Nielsen. Tseng and Sweeney (1946) studied the red alga Gelidinium cartilagineum and found that the rate of its photosynthesis was determined exclusively by the concentration of free carbon dioxide molecules, [CO2], and not affected by the simultaneous presence of a large number of bicarbonate ions, [HCOs"] > 10 [CO2]. Ruttner (1947) compared the limiting pH values established in water as the result of prolonged photosynthesis of different aquatic plants. Elodea (canadensis or densa), Photomageton, Myriophillum prismaium, Lemma trisuUa and several other aquatic phanerogams continued to reduce carbon dioxide until the pH rose considerably beyond pH 9; while several mosses (such as Fonlinalis antipyretica) ceased to assimilate carbon dioxide when pH reached 9.0. At the latter pH, [CO2] = 0.4 X IQ-^ mole/1.; this is the region in which the "carbon dioxide compensation point" was found previously with land plants {cf. page 899). Ruttner suggested that the capacity of aquatic higher plants to carry out net positive photosynthesis at [CO2] equilibrium values <^ 1 X 10~^ mole /I., if bicarbonate is present, indicates their capacity to utilize bicarbonate ions di- rectly, and not merely as source of CO2 molecules in the medium. In a second paper, Ruttner (1948) gave evidence supporting the assumption that the cessation of photo- synthesis of Fontinalis at pH 9 is the result of low CO2 concentration (0.4 X 10 -'^ mole /I., corresponding to about 0.01 vol.%), and not of excess alkalinity. Earlier observations of Shutov (1926), Bode (1926) and Dahm (1926), who found that many aquatic plants can raise the pH of the medium to values as high as 11.8 (Spirogyra), can then be in- terpreted as meanmg that these plants, too, can use bicarbonate ions directly as source of carbon for photosynthesis (or, more exactly, as a vehicle to transport carbon dioxide from the medium into the cells). Osterlind (1948,1949) went even further than Steeman-Nielsen and Ruttner, and asserted that certain plants use bicarbonate ions more effectively than carbon dioxide molecules. He noted that the alga Scenedesmus quadricauda did not grow at all at pH 5.5 (in a solution aerated with ordinary air). It reached a high rate of growth at about pH 6.5, and increased it slowly up to pH 9. According to Osterlind, this increase is not an effect of alkalinity, but a consequence of the presence of bicarbonate ions. He based this conclusion on the observation that, in air containing 5% CO2, good growth could be obtained even at pH 3-4. The maximum rate of growth was reached when [HCO3-] exceeded 9 X 10-^ mole /I.; between 2 and 8 X lO^^ moIe/1., the rate was proportional to [HCO3-]. Osterlind thought that, with the cell populations used, no CARBON DIOXIDE CURVES 891 carbon dioxide exhaustion could occur even in the solutions which contained no bicar- bonate ions. He also noted that, with 10 X 10 ^ mole/1, of HCOs" present, growth was as much as twenty-five times faster than with 10 X 10 "« mole /I. of carbon dioxide; and concluded that Scenedesmus quadricauda uses bicarbonate ions (for growth, and thus presumably also for photosynthesis) twenty-five times more efficiently than free carbon dioxide molecules. Later (1950'- 2) Osterlind found that Chlorclla pyrenoidosa does not use bicarbonates; since he found no difference in the carbonic anhydrase content of the two species, he suggested that their cell membranes arc different. Pending further analysis concerning the role of carbonate ions (a ques- tion which the above-described experiments have reopened), we will pro- ceed on the old assumption that the rate of photosynthesis is primarily a function of the concentration of the molecular species CO2 in the immediate surroundings of the cells, and that the main effect of the presence of HCO3- ions is to prevent this concentration from depletion during photosynthesis. Figures collected in Table 27.1, apart from those given for Myriophyl- lum by Steemann-Nielsen, give no indication of a large, direct contribution of carbonate ions to photosynthesis. We note, for example, that Emerson and Green (1938) were able to achieve carbon dioxide saturation of photo- synthesis in an acid phosphate buffer when the medium contained only 0.7 X 10 -^ mole/1. CO2. In experiments with carbonate buffers, in which each carbon dioxide molecule was accompanied by 1000 bicarbonate ions (and as many or more carbonate ions), saturation usually was observed either at approximately the same or at an even higher value of fC02]. 2. General Review of Carbon Dioxide Curves The necessity of carbon dioxide ("fixed air") for photosynthesis was discovered by Senebier in 1782 (c/. Vol. I, chapter 2). The earliest quanti- tative studies of the relation of the rate of photosynthesis to the concentra- tion of carbon dioxide were made by Kreusler in 1885 and 1887, Brown and Escombe in 1902 and Treboux, and Pantanelli, both in 1903. Since these observations showed an increase of the rate with increasing fC02] in the region of low concentrations, and a decline at high concentrations, they were interpreted on the basis of the then popular "optimum theory" (fig. 26.1), until Blackman and Smith suggested in 1911, that they can better be ex- plained by the concept of "limiting factors." Blackman pointed out that no evidence existed of a "minimum" [CO2] required for the beginning of photosynthesis, or of a sharp "optimum"; instead of the latter, experi- ments showed a wide range of [CO2] values over which the rate remained approximately constant. As described in chapter 26, Blackman claimed that correctly determined carbon dioxide curves must be broken lines of the shape shown in figure 26.2; and many investigations have been carried out with the expressed purpose of "proving" or "disproving" this "law." 892 CONCENTRATION FACTORS CHAr 27 to !/l H o H O M ^1 09 > 0 O a Q o < ^ o n O P^ o m O M H H P 2| I- fa 3 d o .« *r +j c « t. e O (NO (NCOiO O O ^1 S\ S\ S\ SI CO 00 CO O O 1— I o o I— I 5|. _ CO lO CO lO lO lOiO ooOiot>-cot--r-ot>-r>-t~ _ O"*CRi0 0000C0>O0Ci0000 '^ (N CO I— 1 ■— 1 1— 1 1-H ry2 53 I O o 03 3 o3 I o u I S e to S 0) -1-3 03 I o o 0) a o 03 CQ O ^ o oj ;h W lO ^ CO a o ^_^ o 1—1 >o 03 ""— ' CO o CO ^ *-i CO 3 W aj^-^ ^ ^ > -fj X. r-i O -1-3 hC 'S) o IM O 1 O O ,—1 -H ^ (N -H ; f^ 00 ^ ^ (M CO -* V CO CO '^ CO to '3'S 3 3 ■^ t-- CO Ol O 'S,'SdiMQooooocd'-<'c;?<=>^^;r;?2 t-t Sh ^ CB 0)3 -u -IJ -7; ^ > ^ a cO — o ooffi OOM "o "o a> CO m ->-= o3 00 am MM o o3 o 0) 3 -♦-3 o o3 o s^ CO i- ?^ p^ .^^ -*-;> ^ -^^ see 60 S CO « ^ C^ S e S •»-i l>* -Kj s § s o Q 0 *■>» e e e g 2: C3 e Si 'r-i *C«a a. 0 0 CO ^ e V* •^~ e e a. e .0 -0 rS S r-» ^ '"^ 0 5 .0 -0 3 5 e e Otin 0 0 a T-H (N ^ C5 a r-H C1 V ^ G ^ ^ ^ 0) 03 pq ffi 00 t^ CO (M CO CO 0 CTi Oi .— ( f— 1 r-H ^- — ' - — ^ CO J3 ji -1-3 -1-3 s •3 ' r* 03 r^ a 1-5 CO m CARBON DIOXIDE CimVES 803 c l> «o 'O o — ■? > O O O lO lO '-' lO fO O" o o c^) '^^ — 1 lo ^ d o >oo "2 •' 3 >- ^5 ? Al V ^ of. a: 3 '2'3 C -«-3 .5 o' ^ £ >c lO ■* CO (Ml© 00 T-H 5 d r — ^ r- o 1— 1 ^^ r— « d d — ' o o d 00 'o o 1 lOQO Oi I I 1 (N "d d S S g tn 0) O o 5 S o «3 o o 2^1 o o; G t« G c O G ^ intensi ux if ni lerwise ?nated) C^ (N lO lO bC ■- G H p fc, 3 5 ==i5 J § -2 G bii ^ -G o 0; CT (N (N >, Light (in kl oth desij ^ g'-^^e3t^c300& 30C1 ?'':St^G ti • -O c« c3 o =^ j3 (M CO lO O 03 1; SCO o a 1 CO m i72 ^ CO "^ !>. & ^-^ 00 00 -3 6 > > IS -BW K a 0! t/j ^i. o o to H 3 ^ ■^ % -i CO 1 s <-^ h fl^ e '-s e e S S e^ S -3 G O.^^ 8 5 1 2 a s tf 3 S Si. e o s 1 Q Q J3 U)'„ C J a. 1? ^f o 1 1 ^ CS d 3 O o!3: t::; ^ ^5 ^30 3 g"^ .s - jii.Go 00 ii~ i I i C2 S 2 2 '53 '53 03 03 05 CO Qi Qi G S" u i-H 1-H t t 4J > CO 6 a CI _to h 03 S 1— 1 1— 1 c OJ O o 03 Co; ;C 0) c CO r— ) o o G G JS 3^ u (U OJ _ Oi C ^ ^3 Ci wo 3 O <; c 1 a C C G C 0) G c3 o3 S M W > > H H 02 t 894 CONCENTRATION FACTORS CHAP. 27 It hardly needs repeating that the problem must be treated, not from the point of view of an apodictic "law," but on the basis of the general princi- ples of reaction kinetics, and that these principles admit of "limiting fac- tors" only as approximations, useful under certain extreme conditions. Table 27.1 gives a summary of the most important experimental de- terminations of the carbon dioxide curves of photosynthesis, since the time of Blackman and Smith. As a general rule, these curves rise rapidly at first, then more slowly and finally go over into "saturation plateaus." At excessively high [CO2] values, the rate may decline again. Table 27.1 gives, in the last two columns, the concentrations found necessary to pro- duce full carbon dioxide saturation and half saturation, respectively. (When the approach to saturation is gradual, the second figure can often be given with more precision than the first one.) We note that the observed saturating concentrations vary all the way from 0.5 X 10 -s, to 400 X 10"^ mole/1. CO2. It will be shown in section 5 that the higher values are beyond doubt due to depletion of carbon dioxide in the medium surrounding the plants, and consequent establishment of large external concentration gradients. They can be strongly reduced by accelerated circulation or buffering. The lower values, on the other hand, may be determined either by diffusion resistance which is not affected by buffering or stirring (e. g., that of the stomata, air channels, of adsorption layers, cell walls and cytoplasm), or by intrinsic kinetic characteristics of photosynthesis (such as the carboxylation equilibrium and the rate of car- boxylation). In the general discussion of the kinetic curves of photosynthesis in chapter 26, three types of curve sets, P = f[Fi], with F2 as parameter, were considered and designated as the first (or "Blackman") type, the second (or "Bose") type, and the third type (see figs. 26.2, 26.3 and 26.4, respec- tively). It was stated that curves of {he first type must arise when the parameter, F2, determines the maximum rate of a partial process that does not depend on the independent variable, Fi, and therefore imposes a hori- zontal "ceiling" on the curves P = f(Fi), without affecting the initial slope of these curves. Curves of the third type are found when the parameter af- fects only the initial slope of the light curves, for example, if it codetermines the rate of a process that is also proportional to the independent variable, Fi. In curve systems of the second type, the parameter affects both the initial slope and the saturation level. Carbon dioxide curves of all three types can be expected under appropriate conditions; theoretical examples were given in chapter 26. So far, however, the only experimentally deter- mined carbon dioxide curve sets have been obtained with light intensity as parameter. Four sets of such curves, which appear comparatively reliable as far as the measuring technique is concerned, are reproduced in figures CARBON DIOXIDE CURVES 895 Fig. 27. 2A. Carbon .dioxide curves of Foniinalis aniipyretica at various light intensities (in lux) (after Harder 1921). Abscissa, [CO2] in (mole/1.) X 10^ o Intensity 6.2 (20°C) + Intensity 6.2 (I2''C ) A Intensity 20(20°C ) ^ 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 COz, 0.001 vol.% Fig. 27.2B. Carbon dioxide curves of Hormidium flaccidum at two light in- tensities (in relative units), and two temperatures (after van der Honert 1930). 0.01 volume per cent COo corresponds at 20 °C. to 3.37 X 10 -« mole/1, (c/. Vol. I, p. 174). 896 CONCENTRATION FACTORS CHAP. 27 PC02>'°''°°'^ 0.2 0.3 0,4 0.5 CO 1 1 ' 1 en UJ n 20 t- z > CO o 1- o I 10 ' u. o UJ < "^ n 1 1 1 50 100 150 [G02],(mole/l)x I0« 200 Fig. 27.2C. Carbon dioxide curve of Chlorella pyre- noidosa (after Emerson and Green 1938). ^//25 phosphate buffer; pH 4.6; 25° C; rate in mm.' 02/hr. per mm.' of cells. 27.2A, 27.2B, 27.3 and 27.4. Figure 27.2B (taken from van der Honert's work on Hormidium, 1930) closely resembles Blackman's prototype: At low concentrations, all curves merge into a single straight line; close to saturation they sharply turn horizontal. Emerson and Green (1938) gave a single [CO2] curve for Chlorella in saturating light, which showed an even earlier and more sudden saturation: It rose linearly up to [CO2] — 0.7 X 0.082 0164 0246 f COjj , volume per cent 0328 0410 Fig. 27.3. Carbon dioxide c^urves of whole Triticum (wheat) plants at different light intensities (after Hoover, Johnston and Brackett 1933). Parameters in kerg/(cm.^ sec). CARBOX DIOXIDE CURVES 897 10"^ mole/1, and then abruptly became horizontal (fig. 27. 2C). In this figure the maximum yield corresponds to one volume oxygen per volume of cells each three minutes. The rate values were obtained by admitting a known amount of carbon dioxide into a Warburg vessel, shaking vigor- ously and measuring the pressure changes at short intervals until all carbon dioxide was used up. 200 0 6 31 klux .74 klux b8-8 — 8- 0.407 klux _i_ 25 30 5 10 15 20 [C02],in 10'^ mole/ 1. Fig. 27.4. Carbon dioxide curves of Cabomba caroliniana (after Smith 1938). Figures 27.3 and 27.4, obtained with wheat and with the water plant Cabomba, respectively^ show a more gradual approach to saturation, but they, too, indicate a coincidence of all curves at low [CO2] values, which is characteristic of the Blackman type. Harder's Fontinalis curves (fig. 27. 2A), on the other hand, are of a pronounced "Bose type": Curves corresponding to different light intensities diverge from the beginning, and 898 CONCENTRATION FACTORS CHAP. 27 remain in an approximately constant ratio; saturation is approached very gradually, and is not quite reached at 32 X 10 ~^ mole/1., even on the curve that corresponds to an illumination of only 2000 lux. Harder's curves indicate that, in his plants, the rate of oxygen liberation did not become limited entirely by carbon dioxide supply even at the lowest used bicar- bonate concentrations, and did not become independent of [CO2] even at the highest used concentrations. The range studied, 0.03 to 0.3% KHCO3, i. e., from 4 X IQ-^ to 40 X 10-^ mole/1. CO2 (c/. Vol. I, page 178), was, however, a rather narrow one. The curves in figs. 27.2-4 clearly tend to coincide only below 1 X 10 ~^ mole/1. In the theoretical discussion later in this chapter, we will see that carbon dioxide curve? that diverge from the origin can be expected if the carbon dioxide-acceptor complex, ACO2, is not fully saturated with carbon di- oxide, at low [CO2] values, even in the equilibrium state; while carbon dioxide curves that coincide at low [CO2] values can be predicted if the carbon dioxide dependence of photosynthesis is due entirely to the limita- tion of the rate of processes by which carbon dioxide is made available for photosynthesis (such as liberation of CO2 from HCOa", diffusion, and car- boxylation of an "acceptor"). Table 27.1 shows that, with increasing light intensity, the "half satura- tion point," which we will designate by 1/JCO2], generally shifts toward the higher concentrations; so that, in very intense light, it may fall con- siderably beyond 10 X 10 ~^ mole/1. This fact can be significant, indi- cating certain kinetic conditions (as will be shown later in this chapter, see p. 934 ff); often, however, it merely means increasing depletion of carbon dioxide in the neighborhood of the cells when photosynthesis pro- ceeds at a faster rate. 3. Carbon Dioxide Compensation Point For each light intensity, there must exist a carbon dioxide concentration at which photosynthesis just compensates respiration and the net gas ex- change is zero, and below which respiration exceeds photosynthesis. This "carbon dioxide compensation point" has not been studied in the same sys- tematic way as was the "light compensation point" (c/. Table 28. Ill); Miller and Burr (1935) first devoted an investigation to it. In their experiments, a large variety of potted plants were enclosed in vessels filled with gas mixtures of different composition and illuminated with white light of about 20,000 lux, until all observable gas exchange stopped, i. e., until the carbon dioxide concentration had declined to the compensation point. It was found that this occurred — at temperatures from 5° to 35° C. — when the carbon dioxide content was down to about 0.01%. At low CARBON DIOXIDE COMPENSATION POINT 899 temperatures, this gas composition remained unchanged for several hours; at 35-37° C, after a short period of constancy, the carbon dioxide pressure began to increase again — probably because photosynthesis suffered slow thermal inhibition (c/. chapter 31), while respiration remained constant. Thomas, Hendricks and Hill (1944) found that in beet plants, at 15° C, photosynthesis compensated respiration at fC02] — 0.003% — about one third of the value found by Miller and Burr. Gabrielsen (1949) found a value of 0.009 vol.% for the CO2 compensation point of Sambucus leaves at 10 klux. In submerged plants, one has to distinguish between the compensation point at constant pH and the steady state reached after prolonged photo- synthesis with limited carbon dioxide supply: In the latter case, both [CO2] and pH change with time (OH" ions being left behind when CO2 is withdrawn from HCO3-), and the final steady state may be determined by either one or both of these factors. We have referred above to the experi- ments of Dahm (1926), Shutov (1926) and Ruttner (1947, 1948), which were interpreted by Ruttner (1948) and Osterlind (1948, 1949) as indicating that some aquatic phanerogams and algae can use bicarbonate ions so efficiently that the presence of a minimum concentration of free CO2 molecules is not needed to maintain their photosynthesis; these plants are able to continue -net synthesis even after [CO2] has been reduced to 10-^ mole/1, or less, and pH had risen above 10 or 11. (The pH of the sap inside the cells remains approximately neutral.) Aquatic mosses, such as FontinaJis, on the other hand, cease to liberate 0? when [CO2] is reduced to some such value as 0.4 X 10"^ mole/1. (Ruttner 1948). This corresponds to 0.01 vol. % CO2 in the atmosphere, and indicates a compen- sation point similar to that found with land plants. In the steady state reached by photosynthesis of aquatic plants of this type, the reaction of the medium is below, or about equal to pH 9. A rather striking observation of Miller and Burr was that the carbon dioxide compensation point did not depend on temperature. This contrasts with the strong dependence on temperature of the light compensation point (c/. page 984). The reason for this difference is that temperature has a strong influence on respiration, as well as on photosynthesis in strong light, but only a weak effect (or none at all) on photosynthesis in light of low intensity (c/. chapters 29 and 31). In the measurement of the "light com- pensation point," photosynthesis is in the "light-limited" and therefore temperature-independent state; while in the measurement of the carbon dioxide compensation point, it is in the "carbon dioxide limited" state and therefore depends on temperature. However, the exact coincidence of the temperature coefficients of respiration and photosynthesis, implied in the results of Miller and Burr, is unlikely to be more than an accident. 900 CONCENTRATION FACTORS CHAP. 27 We assume that the carbon dioxide curves of photosynthesis, if properly corrected for respiration, continue smoothly below the compensation point and reach zero when the carbon dioxide concentration is zero. However, their exact determination in the region of very low carbon dioxide concen- trations is difficult because of the production of carbon dioxide by respira- tion. It is difficult to remove this carbon dioxide completely (e. 1%, given by some investigators, are unlikely. We will quote two examples of more reliable determinations: Verduin and Loomis (1944) found that, in a maize field, the concentration of carbon dioxide 100 cm. above ground, was 0.055- 0.080% at night, and rapidly declined to 0.045% in the morning. Fuller (1948) found that CO2 concentration near the ground (0-1 cm.) reached EXTERNAL SUPPLY AND EXHAUSTION EFFECTS 903 (at 1 P.M., in June), 0.13% in a forest, 0.10% in grassland, and 0.18% in a river bottom. The concentration declined steeply with height above ground in all these habitats, dropping to near average (~0.04%) 8 or 10 cm. above ground. The decline of photosynthesis at excessively high concentrations of carbon dioxide {e. g., 10 volumes per cent CO2 or more, corresponding to over 300 X 10-^ M), which, before Blackman, was considered a confirmation of the "optimum theory," was reinterpreted by Blackman as an inhibition effect, alien to the intrinsic kinetic mechanism of photosynthesis. It was dis- cussed as such in chapter 13 (Vol. I) which dealt with various inhibitors and stimulants. Referring the reader to this chapter, we merely repeat here references given there to the work of de Saussure (1804) (who dis- covered the effect), Boussaingault (1865), Bohm (1873), Ewart (1896), Chapin (1902), Pantanelli (1903), Jaccard and Jaag (1932) and Livingston and Franck (1940). A recent study by Ballard (1941) with leaves of Ligustrum can be added to the list. It showed that, at 17° C, inhibition occurred (at 35,000 lux) at [CO.] = 2%, while at low temperatures (6° C.) no inhibition was noticeable up to 5%. We recall that Chapman, Cook and Thompson (1924) found that high carbon dioxide concentration induces closure of the stomata; it was therefore suggested in chapter 13 that sto- mata may account for some of the observed carbon dioxide inhibition ef- fects. Other phenomena, which also may contribute to the inhibiting influence of excess carbon dioxide, are its adsorption on catalytic surfaces ("narcotization"), and possibly also acidification of the cell fluids (shift of intercellular buffer equilibria). That the closure of stomata is not the only reason for carbon dioxide inhibition is illustrated by the observation of Osterlind (1949) that it also occurs with algae such as Scenedesmus quadricauda. An inhibition of the growth of this alga became noticeable at 2 X 10"^ mole/1., and reached 50%, at 10 X 10-3 mole/1. CO2. 5. External Supply and Exhaustion Effects In commenting on Table 27.1, we noted wide variations in the numerical values of the saturating carbon dioxide concentration' and suggested that these variations may be due largely to the exhaustion of carbon dioxide in the immediate neighborhood of the plants. We will now consider this as- pect of the problem more closely. The experimental results fall roughly into three classes. One group, which includes the results of Blackman and Smith (1911) and Singh and Kumar (1935), the somewhat less extreme data of James (1928) and the figures given by Steemann-Nielsen (1946) for Fontinalis, and by Wassink 904 CONCENTRATION FACTORS CHAP. 27 and co-workers (1941-1942) for purple bacteria, is characterized by con- tinued increase of the rate of photosynthesis with increasing carbon dioxide concentration until the latter has reached 50, 80, 200 (Singh and Kumar) or even 400 X 10""^ mole/1. (Blackman and Smith) ; the last value corre- sponds to 12% carbon dioxide in the air! Earlier measurements of Kreusler (1885, 1887) and Brown and Escombe (1902), not included in the table, fall into the same category. An intermediate group of results, including Harder's (1921) and Smith's (1937, 1938) on higher aquatic plants, and Emerson and Green's (1934) on Gigartina, place carbon dioxide saturation at 20-30 X 10"^ mole/1. CO2. Finally, in several careful investigations, the rise of photosynthesis with increasing carbon dioxide concentration was found to cease as early as be- tween 0.5 and 5 X 10"^ mole/1. CO2 (Hoover and co-workers 1933, and Singh and Lai 1935, higher plants; van der Honert 1930 and van der Paauw 1932, Hormidium; Emerson and Green 1938, Chlorella; and Barker 1935, diatoms). It will be noted that results of this low order of magnitude have been obtained both with land plants in rapidly circulating gas, and with algae in well-stirred acid or alkaline solutions. There is little doubt that most if not all results of the first type were due to insufficient circulation and consequent depletion of carbon dioxide in the medium surrounding the plants. It is by no means certain that con- centration gradients in the external medium did not affect significantly also the results in group 2, or even in group 3. And, in addition to gradi- ents in the external medium (which can be reduced by intense circulation), we also must consider those in the stomata, air channels, cell walls and cytoplasm. The importance of rapid circulation can be understood by considering that green cells, such as Chlorella, can consume, in strong light, up to one half their own volume in carbon dioxide each minute. In cell suspensions, the volume of the cells usually is from 0.1 to 1% of the volume of the medium. Consequently, the suspension as a whole will use up its own volume in carbon dioxide in from 200 to 2000 min. In other words, the rate of consumption of carbon dioxide will be from 2 X 10 "Ho 2 X 10"* mole CO2/I. min. Consequently, if the concentration of carbon dioxide in the medium is a; X 10"^ ilf, it will be all used up in from 0.05 x to 0.5 X minute, or from 3 a; to 30 x second. Consulting Table 8.II, we note that in water equilibrated, at 25° C, with an atmosphere containing 0.01% CO2, X = 0.4; 0.1% CO2, a: = 4.1; 1% CO2, x - 41, and so on. Conse- quently, a cell suspension in an acid medium that contains no significant amounts of HCO3- ions, containing 0.1 to 1% cells by volume, will con- sume all its carbon dioxide in from 1.2 to 12 seconds, if it has been equili- brated with air containing 0.01% CO2; in from 12 to 120 seconds, if the CARBON DIOXIDE EXHAUSTION EFFECTS 905 atmosphere contained 0.1% CO2, and so on. Bicarbonate solutions con- tain, for each GO2 molecule, about 100 HCOs" ions; they should therefore last one hundred times longer than acid solutions with the same value of [CO2]. Finally, 0.1 M carbonate-bicarbonate buffers containing from 2 X . 10^ (buffer No. 1) to 330 (buffer No. 11) carbonate and bicarbonate ions for each CO2 molecule provide sufficient reserves to maintain full photosyn- thesis, in suspensions containing 0.1 to 1% cells by volume, for from 500 to 5000 minutes, or S to 80 hours. These figures lead to several conclusions. First, measurements of the rate of photosynthesis at low carbon dioxide concentrations (e. g., less than 1% CO2 in the air, or 30 X 10 -^ M in solution), if they are to last for more than a few seconds, require an ample supply of carbon dioxide, either in situ, in the form of carbonate and bicarbonate ions, or from outside, in the form of large amounts of circulating liquid or gas, which must be kept well supplied with fresh carbon dioxide to replace losses. Second, whenever leaves or multicellular algae are used, very intense stirring or circulation is required to prevent the establishment of a carbon dioxide concentration gradient around the plants. The required stirring depends on the ratio of surface to volume. This is well illustrated by the following example: Gessner (1938) measured the oxygen liberation by two varieties of Proser- pinaca palustris — one with large leaves and one with finely divided, feather- like leaves. In stagnant water, the first variety produced much less oxygen than the second one; stirring improved strongly the efficiency of the large- leafed, but did not affect the oxygen production by the feather-leafed variety. In other words, in the absence of circulation, external carbon dioxide supply must have been the rate-limiting factor for the large-leafed, but not for the feather-leafed, variety. It seems that, with multicellular objects, even the provision of a strongly stirred bicarbonate-buffered medium does not always guarantee the absence of carbon dioxide exhaustion effects. Wassink (1946) found, for example, that the photosynthesis of 5 mm. discs cut out of leaves, sus- pended in carbonate buffer No. 9 (7.9 X 10 -^ mole/1. CO2) and shaken in a Warburg apparatus, still was largely "carbon dioxide limited." The equilibrium concentration of carbon dioxide in the atmosphere above this buffer is 0.25% (cf. Tables 8.V and 8.II). By increasing the initial carbon dioxide content in the air space to 2% in some cases, and to as much as 9% in others, Wassink was able to obtain carbon dioxide saturation. Without exact knowledge of the dimensions of the apparatus, it is difficult to esti- mate the final carbon dioxide concentration and the pH of the solutions treated in this way. Closure of the stomata in punched leaf discs may have been one of the reasons for apparent extreme carbon dioxide requirements observed in 906 CONCENTRATION FACTORS CHAP. 27 these experiments. From this point of view, and from the point of view of favorable ratio of surface to volume, unicellular algae offer much better conditions. In brief experiments, or in weak light, they can be used in .acid solutions previously equilibrated with carbon dioxide of sufficiently high partial pressure (>1%; it was calculated above that a suspension containing 1 volume per cent of cells will use, in saturating light, all the carbon dioxide contained in water equilibrated with 1% CO2, in 1.5 minutes). If stronger illumination or longer duration of experiments is desired, acid solutions can be used only if the carbon dioxide content is continuously renewed, e. g., by stirring with a gas the carbon dioxide con- tent of which is maintained by contact with an alkaline carbonate buffer. More efficient should be the provision of carbon dioxide reserves in situ by using carbonate buffers directly as suspension media, as first sug- gested by Warburg. However, a certain difficulty arises from their un- physiological and variable alkalinity. In progressing from M/IO buffer No. 1 (0.5 X lO-'^ mole/1. CO2) to M/10 buffer No. 11 (29 X 10"^ mole/1. CO2), we find the pH declining from 11 to 8.5. Since all living cells are more or less sensitive to excess alkalinity (even if Chlorella appears to be remarkably resistant to it), this drop of pH could cause continued increase of the rate of photosynthesis in a range where this rate is intrinsically in- dependent of carbon dioxide concentration. This may explain, for ex- ample, the difference between the carbon dioxide curve of Chlorella as de- termined by Warburg (1919) in carbonate buffers, and the same curve ob- tained by Emerson and Green (1938) in a phosphate buffer. The first one continues to increase up to and beyond 9 X 10^^ mole/1. CO2, while the second one is perfectly flat above 0.7 X 10"^ mole/1. CO2. On the other hand, observations of Ruttner (1947, 1948) and others on the maximum pH reached in non-renewed media after prolonged photosyn- thesis by aquatic plants (c/. above page 890), tend to discount the damaging effect of alkalinity on algae and submerged phanerogams (as contrasted to water mosses), by indicating the continuation of photosynthesis up to pH 11 or 12; pH measurements on cell sap showed it to maintain its ap- proximately neutral reaction even in such highly alkaline media. (For other possible explanations of the difference between the CO2 curves of Chlorella as observed by Warburg, and by Emerson and Green, see page 908.) Another pertinent question is whether the rate of conversion of HCOa" ions into CO2 molecules always is high enough to provide effective replenish- ment of used-up carbon dioxide. In chapter 8 (cf. Vol. I, page 175) we dis- cussed the finite rate of hydration and dehydration of carbon dioxide, and estimated that, in acid solution at room temperature, an H2CO3 molecule lives ca. 0.1 sec. before dissociating — the monomolecular rate constant EXTERNAL SUPPLY AND EXHAUSTION EFFECTS 907 of dehydration being about 10 sec.-^ at 18° C; (c/. Table 8.III). The rate of dehydration of HCOa" ions was not given there, but we can esti- mate it from the rate of addition of OH" to CO2, determined experimen- tally by Brinkman, Margaria and Roughton (1933) : CO2 + OH- , HCO3- k' k = 2.05 X 103 (18° C.) To obtain k', first calculate the equilibrium constant of the above reaction from the known constants of dissociation of water (1.04 X 10~"), ionic dissociation of H2CO3 into H+ and HCO3- (1.8 X 10""), and hydration of CO2 (2.2 X 10-^), and obtain: K = k/k' = 4.4 X 10' This, together with the above value for k, gives (for 18° C.) : A;' = (2.05 X 103)7(4.4 X 10') = 0.47 X 10"* This indicates that an HCOa" ion lives, on the average, at 18° C, as long as 2.7 X 10^ sec. before being dissociated into OH ~ and GO2. A bicarbonate buffer containing y mole HCOs^/l. can therefore supply a maximum of 4.7 X 10 ~^ y mole C02/l./sec. by this dehydration process. At pH < 10, dehydration via H2CO3 must be added; at pH 9 it can double the rate of conversion of HCOs" to CO2 (assuming the association of HCOs" and H+ to H2CO3 to be practically instantaneous). A solution containing y = 0.02 mole/1. HCO3- (Warburg's il//10 buffer No. 2, pH ^ 10.7) is thus able to supply a maximum of 9 X 10"'' mole CO2/I. sec. The cor- responding figure for buffer No. 9 (0.085 M HCO3-, pH ^ 9.4) is 5 X 10 ~^ mole CO2/I. sec. Comparing these figures with the above-estimated maximum rates of photosynthesis in strong light (from 2 X 10~Ho 2 X 10-^ mole CO2/I. min., or from 3.3 X 10 ^^ to 3.3 X 10 "^ mole CO2/I. sec. for suspensions containing from 0.1% to 1% cells by volume), we note that the maximum supply exceeds maximum consumption in the 0.1% suspen- sion by a factor of about three in buffer No. 2 and by a factor of about fifteen in buffer No. 9. In 1% suspension, the supply is quite insufficient in buf- fer No. 2 and barely sufficient in buffer No. 9. Considering the roughness of the calculation (e. g., the use of concentrations instead of activities), the margin is by no means secure even in the dilute suspension. Assuming the calculation to be exact, a supply process with a maximum rate equal to only 3 times the noninhibited rate of reaction is bound to cause a marked inhibition (c/. chapter 26). It is therefore an open question whether the limited rate of reproduction of CO2 molecules from HCOs^ ions can play a role in the determination of the rate of photosynthesis of dilute suspensions in strong light, at least in the more alkaline carbonate buffers. This 908 CONCENTRATION FACTORS CHAP. 27 • "bottleneck" may well have contributed, c. gr., to the decline in rate ob- served by Warburg (1919) in Chlorella at [CO2] < 9 X 10"^ M. (As mentioned before, Emerson and Green have noted no such decline until [CO2] was down to 0.7 X 10~* M, and have suggested that damage caused by increased alkalinity of the lower carbonate buffers could provide an explanation of Warburg's results.) Carbon dioxide exhaustion effects are not restricted to experiments in liquid media, but affect also measurements made with land plants in a car- bon dioxide atmosphere, if it is stationary (c/. Lundegardh 1921), or in- sufficiently agitated (Kreusler 1885,1887; Singh and Kumar 1935); this was demonstrated by Kostychev ef aZ. (1927) and Chesnokov and Bazyrina (1932). Here again, not only the rate of gas circulation, but also the size and shape of the plants may be of importance and the opening of the stomata constitutes an additional complication. To sum up it seems safe to assume that, whenever the rate of photo- synthesis was found to continue its increase with the external concentra- tion of carbon dioxide much above [CO2] = 10 X 10 "^ M, the reason was slow outside supply of carbon dioxide to the photosynthesizing cells, and consequent exhaustion of the reduction substrate. Experiments in vigor- ously stirred solutions, or in rapidly circulating gas mixtures, regularly showed the photosynthetic apparatus to become saturated with carbon dioxide at concentrations not much higher, or even lower, than 1 X 10 ~^ M. Even in experiments of this type, one cannot be certain whether all diffusion effects have been eliminated, particularly in higher plants, where the diffusion resistance of the stomata, epidermis and air channels cannot be destroyed by stirring or gas circulation. The diffusion resistance of the cell walls or protoplasmic layers also remains unaffected by all me- chanical means (although it may perhaps be changed by chemical agents). Another source of distortion of the carbon dioxide curves of photosyn- thesis was noted by Howies (unpublished) and Whittingham (1949) in Brigg's laboratory. They observed that the photosynthesis of Chlorella in carbonate buffers with low [CO2] values was time-dependent, if the cells had been transferred into the CO2 deficient medium from a culture medium of higher concentration (such as 4% CO2). The initial rate was low; it increased by a factor of 3 in the course of two or three hours, and then be- came constant. If the cells were cultured in air (0.03% CO2), the rate was high and constant from the beginning. Obviously then, with cells "incubated" at high [CO2], the shape of the carbon dioxide curve will de- pend on the duration of the measurement. If the carbon dioxide curve of Chlorella is determined at low [COu] values, with cells "adapted" to low carbon dioxide concentration, the value of VJCO2] is as low as 0.5 to 1.0 X 10-« mole/1. EXTERNAL SUPPLY AND EXHAUSTION EFFECTS 909 The initial inhibition of photosynthesis in low [CO2], shown by cells previously exposed to high [CO2] values, could be related to the photoxida- tion phenomena observed in CO2 starved plants (c/. Vol. I, chapter 19). Using Franck's picture, it can be suggested that CO2 satiated cells, placed in CO2 deficient medium and exposed to light, develop a large quantity of a "narcotic" (perhaps, because they were full of metabolites), which settles on chlorophyll and holds photosynthesis down. The autocatalytic removal of this inhibition seems to require 2-3 hours (as against a few minutes in ordinary induction, cf. chapter 33). That such cells in fact are inhibited is confirmed by the observation that if, after brief exposure to light in low [CO2], they are brought back into a medium of high [CO2] (such as buffer No. 9), they show a reduced rate of photosynthesis in this medium as well. Since these experiments were carried out in carbonate buffers, the ob- served effects can be attributed either to changes in [CO2], or to those in If all carbon dioxide activity gradients between the outside medium and the site of photosynthesis could be avoided, we would still anticipate, on theoretical grounds, that carbon dioxide concentration will retain an influence on the rate of photosynthesis: first, because of dissociation, under low partial pressure of carbon dioxide, of the carbon dioxide-acceptor compound that we assume is formed as an intermediate in photosynthesis (cf. chapter 8) ; and second, because of the dependence of the rate of forma- tion of this compound ("carboxylation") on the factor [CO2]. These two relationships will be discussed theoretically in sections 7b and c ; but there can be no certainty, until much more precise measurements have been carried out, that any of the observed carbon dioxide curves actually reflect one or both of these intrinsic kinetic relationships, rather than the more incidental diffusion phenomena. As long as a [CO2] effect can be made to disappear by improved stirring, it reveals itself as due to external diffusion; but, when no further improvement in rate can be achieved in this way, this does not mean that the remaining [CO2] effect is not caused by diffusion in those parts of the gas path were external stirring can do no good. In estimating the supply of carbon dioxiile to plants under natural conditions, the possibility of caibon dioxide siipplij through, llir rools nuist not be overlooked. It was mentioned in chapter 2 (N'ol. I) that the doclrine of the aerial nijurishment of plants was the second accomplishment of the discoverer or, more exactly, co-discoverer of ])hot()- synthesis, Ingen-Housz. Since the time of Liebig, this doctrine has become the basis of the science of plant nutrition. However, under certain conditions, Senebier's concept that carbon dioxide can be supplied by the soil water to the roots and thence to the leaves, scorned l)\' Tngcn-TTousz, may be correct. This may affect, field determinations of the rate of pliotosynthcsis, based on measui'ements of the carbon dioxide consumption from the air, and may also influence the results obtained by other methods, if the ob- served rates are considered in relation to the external carbon dioxide concentration. For 910 CONCENTRATION FACTORS CHAP. 27 recent discussions of this question, we refer to Bergamaschi (1929), Livingston and Beall (1934), Suessenguth (1937), Overkott (1938, 1939) and Hartel (1938). The experiments of the two last-named authors, in particular, have confirmed unambiguously that a cer- tain amount of carbon dioxide can be conveyed from roots to leaves by convection (and to a smaller extent by diffusion), and that this supply can be utilized by leaves for the synthesis of carbohydrates. It was even suggested that this "invisible" CO2 supply, brought about by increased transpiration during the hot hours of the day, may be the cause of the decline of the carbon dioxide absorption from the air, which is often ob- served at midday (c/. page 873). Whether this is a valid hypothesis cannot be judged without quantitative investigations; but in any case, it cannot explain all the aspects of the so-called "midday depression," (o) because these also include a decline of oxygen liberation, (6) because they have been observed not only in the higher land plants but also in aquatics. 6. Role of the Stomata It was stated above that, in experiments with the leaves of the higher land plants, a special problem is posed by carbon dioxide passage through Fig. 27.5. Diagram of a section through stoma and substomatal cavity of a leaf to show direction of diffusion of gases in photosynthesis (after Robbing and Rickett). Arrows with black balls represent carbon dio.xide; those with tri- angles, oxygen. the stomata and air channels, through which it has to flow in order to reach the photosj^nthesizing cells of the palisade tissue and of the spongy paren- chyma. A controversy as to whether the carbon dioxide enters the leaf only through the stomata or also through the cuticle was decided by Blackman (1895). He proved, by ROLE OF THE STOMATA 911 experiments with paraffined leaves, that gas exchange takes place almost exclusively through the stomata, in the way indicated in figure 27.5. Only under very high pres- sure of carbon dioxide did Blackman observe a slight penetration of the gas through the cuticle. According to St&lfclt (1935), in the free atmosphere, carbon dioxide pene- trates the cuticle at a rate of only between 3 and 6 X 10~^ mole/cm. ^ hr.; the gas flow through the stomata may be as much as one hundred times faster, i. e., of the order of 5 X 10"'' mole/cm. 2 hr., despite the fact that their openings occupy only about 0.1% of the total leaf surface. Freeland (1946) found more recently that, in some leaves, the relative rate of passage of carbon dioxide under pressure through the lower and the up- per surface is so low as to suggest predominance of diffusion through the epidermis over passage through the stomata. The thickness of the epidermis may be an important factor in the determination of the relative role of stomata and epidermis as routes for the entry of carbon dioxide into the leaf. Ferns and other lower land plants possess no stomata, and therefore must receive all their carbon dioxide supply through the epidermis. Stomata also are absent in aquatic plants and algae, where their main function — regulation of evaporation — is not required. Between 10,000 and 30,000 stomata are present on each square centi- meter of the leaf surface, either on both sides or on the lower side only. They are elongated slits, usually from 10 to 15 ix long, flanked by two "guard cells" {cf. figs. 27.5 and 27.6), which are capable of changing shape so as to effect the opening or closing of the slit {cf. fig. 27.7). Fig. 27.6. A portion of the lower epidermis of a geranium leaf (after Robbins and Rickett). This mechanism is brought into operation by shifts in the sugar-starch equilibrium, which increase the turgor when the slits are to be opened, and decrease it when they must be closed. The problem of the diffusion resistance of stomata has been considered from two points of view: First, it was asked: Is it possible for a diffusion flow of up to 10-5 mole (0.24 cc.) COa/hr. (c/. chapter 28, Table 28.5) to pass through the stomata on 1 cm.- of the leaf surface, when the total open 912 CONCENTRATION FACTORS CHAP. 27 area is less than 1 mm.^ and the concentration drop is not more (and often less) than 1 X 10~'^ mole/1, (which is the normal C(\ concentration in the open air)? The second question was : Granted a remarkably low diffusion resistance of the stomata, is this resistance nevertheless an important "limiting factor" in photosynthesis of higher plants, particularly at low carbon dioxide concentrations? To understand why the first question had to be asked, suffice it to recall the experiment of Brown and Escombe (1900), who showed that a leaf takes up carbon dioxide from quiet air almost as rapidly as an eciually large surface of an alkali solution! It Avas soon found that this unexpectedly high rate of diffusion has nothing to do with the physiological properties of the leaf but is a general property of multiperf orate septa, i.e., barriers con- taining many small openings. Model experiments on transpiration showed {cf. Sierp and Seybold 1929, 1930) that the rate of evaporation from a ves- Fig. 27.7. Stoma of Hellehorus sp. in transverse section. Darker lines show shape assumed by guard cells when stoma is open; lighter lines when stoma is closed (from Strassburger et al, after Schwendener). In closed state, vacuole (shaded area) con- tracts because of water loss caused by decreased turgor (produced by polymerization of sugars). sel, covered with a septum, can be as high as three fourths of that from an equally large open vessel— even if the aggregate area of the holes is less than 1% of the total surface of the liquid! The theoretical solution of this apparent paradox was given (for the case of evaporation) as early as 1881 by the Austrian physicist Stefan. He used, for this purpose, the formal similarity of the equations describing the diffusion flow of matter from an extended surface and from a point source, with the equations de- scribing the hues of force in the electrostatic field in front of an extended conducting surface, and around a small conductor. In this formal analogy, the diffusion flow corresponds to the electrostatic capacity of the conductor; and it is known that the capacity of a large flat condenser is determined by the area of its plates, while the capacity of a small spherical conductor is determined by its radius. In the same way, the amount of DIFFUSION THROUGH SEFTA 913 evaporation from an extended surface is proportional to its area, while the amount of evaporation from a small sphere is proportional to its radius; the same applies to the comparison of diffusion across an extended plane (the case usually considered in the derivation of diffusion equations) with the diffusion through a small hole. Diffusion through a multiperforated septum can be treated in the same way as that through a single hole as long as the distance between the holes is large enough (compared with the radius of the holes) for the half-spherical surfaces of equal concentration (and the radial lines of flow, which are normal to these surfaces) to be established around each hole without marked interference by the neighboring holes. This principle was first applied to the penetration of carbon dioxide through the stomata by Brown and Escombe (1900) upon advice of the physicist Larmor. Renner (1910, 1911), Brown (1918), Freeman (1920), Sierp and Noack (1921), Sierp and Seybold (1927, 1928, 1929), Huber (1928) continued the study, being, however, mainly concerned with the trans-piraiion of plants. As a typical result, we reproduce a table from the paper by Sierp and Seybold (1929). Table 27.11 shows the rates of evaporation of water through septa with different numbers of holes but a constant total open area. The next-to-last row shows the flow-retarding effect of an inadequate distance between the holes. The table indicates that a maximum rate of diffusion is reached asymptotically when the holes are reduced to 20-10 n in diameter. Although the aggregate area of the holes (3.14 mm.2) is less than 1% of the total area of the vessel (-400 mm.^), the evaporation rate through the septum with holes 10 n in diameter is as high as 70% of that from the open vessel. These figures indicate that the dimensions of the stomata (5-15 n) may be appropriate to ensure the de- sired rate of gas exchange through the smallest possible number of openings. Table 27.11 Evaporation through Septa (After Sierp and Seybold 1929) Distance between holes, Total Total multiples Rate of Number Diameter open circumference of pore evaporation of holes of hole, M area, mm.^ of all holes, mm. diameter in quiet air 1 2000 . 3.14 6.28 — 1.0 400 100 3.14 125.6 9.5 7.7 1,600 50 3.14 251.2 9.2 11.1 10,000 20 3.14 628.0 9.1 11.7 40,000 10 3.14 1256.0 9.0 12.1 10,000 20 3.14 028.0 4.0 5.5 Open surface 400 — — 16.3 The theorj^ of gas diffusion through multiperforate septa was further advanced by Verduin (1949), by mathematical analysis of the mutual 914 CONCENTRATION FACTORS CHAP. 27 interference of the openings. He calculated that interference should be inversely proportional to the square of the distance between pores, d: logQ/Qi = -k/d^ where Qi is the diffusion rate at d = oo . This equation agrees well with experimental results of Verduin (1949) and Weishaupt (1935). At a given ratio of pore diameter and pore distance, the interference must be stronger the smaller the pores. The stomata are so small that the diffusion through each of them is reduced significantly by interference — ^sometimes by > 50% of the theoretical value for an isolated opening of the same size. As stomata close gradually, interference weakens; and the diffusion rate therefore declines slower than proportionally to the open area. These experiments and their theoretical interpretation explain how the tiny stomata can allow a large volume of carbon dioxide to diffuse into the leaf, thus permitting a high rate of photosynthesis. We now turn to the second question; does the resistance of the stomata impose a significant limit on the carbon dioxide supply and, with it, on the rate of photosynthe- sis? Closed stomata undoubtedly must curtail photosynthesis drastically (restricting it to the utilization of the carbon dioxide that can reach the chloroplasts by diffusion through the cuticle, or is produced in the leaf by respiration). The question is: How far must the stomata be open to cease exercising a restrictive influence on photosynthesis? May this restriction be significant even when slits are fully open? Are they the bottlenecks responsible for the "Blackman features" of many carbon dioxide curves? It will be recalled that, in the preceding chapter, it was shown that the restrictive influence of a reaction step generally becomes felt long before the rate of the over-all process closely approaches the "ceiling" imposed on it by this step. Therefore, the resistance of the stomata may affect the shape of the carbon dioxide curves even when the rate of photo- synthesis is not more than one half or one quarter of the maximum possible flow of carbon dioxide through the stomata. For an experimental study of the influence of stomata on photosynthesis, one must measure the rate of photosynthesis under constant external con- ditions, but with varying apertures of the stomata. Unfortunately, treat- ments used to enforce partial closure of the stomata (such as incubation in darkness or in dry air) may also directly affect the efficiency of photosyn- thesis, so that caution is required in the interpretation of the results. In order to arrive at reliable conclusions, the width of the stomata and the rate of photosynthesis must be determined with the same leaf^a condition that has not always been fulfilled. The relation between stomatal openings and the rate of photosynthesis ROLE OF THE STOMATA 915 has been the subject of study by several investigators, among them Iljin (1923), Geiger (1927), Maskell (1928), Johansson and Stalfelt (1928), Kostvchev, Bazyrina and Chesnokov (1928), Boysen-Jensen (1932), Schoder (1932), Stalfelt (1935), Newton (1936), Heath (1939) and Heath and Penman (1941). Of these, Kostychev, Bazyrina and Chesnokov (1928), and Schoder (1932) could find no correlation between the two magnitudes. All other observers concluded that, under certain conditions, a clear-cut relationship can be noted between them. Thus, Maskell (1928) found a parallelism between the diurnal and seas- onal course of stomatal apertures (as measured by a porometer) and the 16 - 12 - Oat 3,8000 lu> • • • • • • • • • — 8 • / • • •L • 4 • / / , 1 1 1 02468 U2468 OPENING,/! OPENING,,! Fig. 27.8. Relation between photosynthesis and stomatal openings in ordinarj^ air (0.03% CO2) (after Stalfelt 1935). One-sided illumination. Air flow 4 ± 1 m./min. Average errors indicated by crosses. Ordinates, P in mg. CO-/(100 cm. 2 hr.). rate of photosynthesis. He made a theoretical estimate of the maximum rate of carbon dioxide passage through the stomata, and concluded that it is of the correct order of magnitude to operate as a bottleneck in photosyn- thesis. Stalfelt (1935) determined the opening of the stomata by microscopic measurements, using excised pieces of leaves of wheat or other cereals. The same leaves also were used for the determination of photosynthesis. The width of the stomata was varied by exposure to dry air, and preillumin- ation by light of varying intensity. Figure 27.8 shows typical results. It will be noted that in strong light (26,000 lux) the limiting effect of stomata does not disappear even when they are fully open ; at 8000 lux, on the other hand, this effect already ceases to be noticeable when the stomata are only 916 CONCENTRATION FACTORS CHAP. 27 one quarter open (2 ju). This difference is understandable since at 8000 lux the maximum rate of photosynthesis is only 8 mg. COo/cm.- hr., while in 26,000 lux, it rises to > 20 mg. C02/cm.2 hr. Stalfelt concluded that stomatal openings easily may limit the carbon dioxide supply in ordinary air, and therefore also the rate of photosynthesis under natural conditions, particularly in strong light. Like Maskell, she supported this view by cal- culations of the rate of diffusion through the stomata, based on equations of Brown and Escombe (1900). These calculations confirmed that the maximum rate of carbon dioxide flow from ordinary air through wide open stomata is of the same order of magnitude as the maximum rate of photo- synthesis. These results, while clearly showing the possible "bottleneck" role of the stomata, do not mean that other parts of the path between atmosphere and chloroplasts do not contribute commensurable — or even greater — terms to the total diffusion resistance. In the face of these results, one must disagree with Renner (1910), who thought that the resistance of the stomata represents only a negligible fraction of the total diffusion re- sistance on the carbon dioxide path from the atmosphere to the chloroplasts, as well as with Schroeder (1924), who attempted to prove that the diffusion resistance of the air channels is the rate-limiting influence in the photosynthesis of the higher plants, and in this proof altogether omitted the resistance of the stomata. Romell (1927) pointed out that Schroeder neglected, not only the flow resistance of the stomata, but also that of the gas-liquid interface, and of the liquid phase between the cell wall and the chloroplasts. Romell calculated that the gradient of the carbon dioxide concentration in the air channels must be smaller than in the protoplasm (be- tween cell wall and chloroplast), and that both these gradients should be negligible in comparison with the drop of concentration at the phase boundary, caused by the rela- tively small accommodation coefficient of carbon dioxide on water (as calculated from Bohr's measurements of the velocity of escape of carbon dioxide from aqueous solution). The maximum theoretical rate of diffusion, calculated by Romell by taking all these factors into account, proved to be considerably lower than the maximum rate of photo- synthesis that the leaves actually can reach in open air. One is thus led to assume (c/. van der Honert 1930) that the accommodation coefficient of carbon dioxide is larger on the cell wall than on a water-air interface. 7. Interpretation of Carbon Dioxide Curves The preceding pages show that reliable experimental material for analyt- ical interpretation of the carbon dioxide curves of photosynthesis is hardly available at present, and will not be easy to obtain. We have stated that at least two intrinsic kinetic factors could make the rate of photosynthesis a function of the external carbon dioxide pressure: the probable reversi- bility of the primary carbon dioxide absorption (carboxylation) step, and a finite rate of carboxylation. The difficulty is to recognize the workings of these "intrinsic" or chemical factors behind the more incidental, physi- INTERPRETATION OF CARBON DIOXIDE CURVES 917 cal flow plienomena outside and inside the plant. We cannot be sure at present whether any of the observed carbon dioxide curves reflect reason- ably well the effect of carboxylation equilibrium (or of the rate of carboxy- lation), or whether practically all carbon dioxide dependence of photosyn- thesis, known so far, is due to diffusion phenomena, with possible addi- tional distortions by the time effects noted on page 908. Despite this unsatisfactory state of our experimental knowledge, we will go through with some kinetic derivations leading to general equations for the shape of carbon dioxide curves, as affected by the several factors of slow diffusion, limited rate of carboxylation, reversibility of carboxylation and limited supply of light energy. We will thus obtain a kind of skeleton analytical theory of the carbon dioxide curves, which could prove useful for devising and interpreting future kinetic measurements — if only inves- tigators of the kinetics of photosynthesis would change their present habit of considering only their own limited data, and ignoring all but their own ad hoc derived equations. (a) Carboxylation Equilibrium Two steps in photosynthesis, the rate of which depends directly on the factor [CO2] are: First, diffusion of carbon dioxide from the medium to the reaction site, and second, the first chemical reaction of carbon dioxide. In chapter 8 (Vol. I), we decided that this reaction is a nonphotochemical, catalytic carboxylation. We usually described it by the formula CO2 -^ {CO2}, but since the concentration of the "carbon dioxide acceptor" (until now symbolized by braces) enters explicitly into many of the following ki- netic equations, we will from now on designate it as A, and the product of carboxylation as ACO2. (Franck and Herzfeld, 1941 , used the more specific symbols RH for acceptor and RCOOH for the product.) The two [CO.j]-dependent steps can then Ijc written as follows: Ki (27.1) CQ2 . ^ (CO,)a (where (C02)a refers to carbon dioxide in the immediate neighborhood of the acceptor, and k^ is a diffusion constant) and: (27.2) (C02)a + A , '^ ^ ACQ., ( > reduction) a The reduction of ACO2 may be either a direct photochemical process (as assumed by Franck and Herzfeld; cf. scheme 7.VA), or a nonphotochemi- cal reaction with an intermediate, as postulated in many other schemes in chapters 7 and 9. Even in the latter case, the rate of reduction is hkely to be a function of light intensity, because the partner with which the com- 918 CONCENTRATION FACTORS CHAP. 27 pound ACO2 reacts must be a — direct or indirect — product of the primary photochemical process. Two different premises can be used in the kinetic analysis of the effect of carboxylation. One alternative (indicated by arrows in equation 27.1, is to assume that the carboxylation is markedly reversible, i. e., that k^ is of the same order of magnitude as A-a[C02]a. In this case, the association of the acceptor A with carbon dioxide is not complete even without any dislocation of the equilibrium by the consumption of ACO2 in light. The other alternative, preferred by Frank and Herzfeld, is to assume that the carboxylation equilibrium lies entirely on the side of association (meaning A-a[C02] » A-a), so that in the dark practically all acceptor is "saturated" by carbon dioxide molecules (at all practically significant partial pressures of carbon dioxide), unless the carbon dioxide molecules are displaced by other association partners, such as reduction intermediates, narcotics, etc. Free molecules A occur in this picture only during (or immediately after) intense photosynthesis, when reduction of ACO2 is (or has been) too rapid for the recarboxylation to keep step with it. As described in chapter 8 (vol. I) the known equilibria of carboxylation in vitro correspond to practically complete dissociation. Only few cases are known in which the carboxyl group is thermodynamically stable with respect to decarboxylation (at least, under sufficiently high carbon dioxide pressures). The "saturation" of photosynthesis with carbon dioxide, which occurs under partial pressures as low as 0.1%, indicates that conditions may be different here, perhaps in consequence of a coupling of carboxylation with another reaction, such as degradation of a "high energy phosphate," or an "endergonic" oxidation-reduction (c/. Vol. I, page 201). However, there is no experimental or theoretical reason — except convenience in ana- lytical formulation— to postulate that in photosynthesis the carboxylation equilibrium lies completely on the side of synthesis, even at the lowest practi- cally significant carbon dioxide pressures. We will therefore begin our analysis by assuming that the degree of saturation of the acceptor with car- bon dioxide does depend on the external concentration of carbon dioxide. If one molecule of carbon dioxide is taken up by one molecule of ac- ceptor, the carboxylation equilibrium is determined by the equation: (27.3) [ACO2] = (A'aAo[C02]a)/(l + i^a[C02]a) where ^0 is the total available concentration of the acceptor: (27.4) Ao = [A] + [ACO2] and iCa is the equilibrium constant of carboxylation: (27.5) Ka[C02]a[A] = [ACO2] INTERPRETATION OF CARBON DIOXIDE CURVES 919 If the carboxylation mechanism is as simple as postulated in (27.1), the equilibrium constant Kg, is equal to the ratio of the two rate constants ka and h\. In equation (27.4) it is assumed that the acceptor, A, is either free or occupied by CO2. This may not be the complete description for two rea- sons: In the first place, the first reduction product of ACO2, designated by us as AHCO2, may require time for its dissociation into A and HCO2; part of the acceptor is then "blockaded," during photosynthesis, by this reduction product {cf. section / below). In the second place, the photochemical re- duction may have to be repeated several times, e. g. : (27.5A) ACO2 — ^^ AHCO2 — — > AH2CO2 ^ AH3CO2 ^ AH4CO2 > A + H2O + {CHoO! before the reduction product can separate itself from the carrier A (as in the Franck-Herzfeld mechanism discussed in section d below.) If AHCO2 is assumed to be the only product of photochemical reduction, the completion of its reduction to AH4CO2 — i. e., to the carbohydrate level- must be ascribed to dismutations : 4 AHCO2 > AH4CO2 + 3 ACO2 cj. Vol. I, p. 158. The simplest assumption that can be made in the interpretation of the carbon dioxide curves of photosynthesis is that they are, at least basically, saturation curves of the acceptor A. This means that one assumes (a) that equilibrium (27.3) is not strongly dislocated during photosynthesis, at least under moderate conditions, and (6) that the rate of photosynthesis is given by the rate of reduction of the compound ACO2, and the latter is proportional to the concentration [ACO2]: (27.6) P = nkf X [ACO2] where the constant k* depends on the intensity of illumination (as indi- cated by the asterisk). We will deal with the possible limitations of as- sumption h later (see, e. g., section e). The condition for the correctness of a (i. e., for the maintenance of equilibrium 27.3 in light) is (cf. formula 27.2) : (27.7) k* < K That this condition is not always satisfied is demonstrated by the "pick up" phenomena, described in chapter 8 (Vol. I). These observations show that in very intense light (i. e., when k* is very large) or in the presence of certain poisons (when k'a is very small) the acceptor A becomes "denuded" of carbon dioxide and afterward "picks it up" in the dark. 920 CONCENTRATION FACTORS CHAP. 27 The factor n in (27.5) is either 1 or a fraction of 1, depending on how many molecules CO2, at best, can be reduced to the carbohydrate level, { CH2O } , for each molecule CO2 which undergoes the first reduction step (to AHCO2). If the mechanism of reduction involves only one photo- chemical step (ACO2 -^ AHCO2), followed by a two-step dismutation: 4hv 4 AHCO, > 3 ACO2 + HoO + (CH2OI + A then /; is 1/4. If an energy dismutation step of the tj^pe discussed in chapter 9 (page 264) also is involved, n is reduced to 1/8. If, on the other hand, CO2 is reduced to the H4CO2 level in a straight series of photochemi- cal reduction steps: AGO. ^"' ) AHCO, -^ -^ -^ ATI4CO, > A + H,0 + CH,0 then 71 is equal to 1. (As a compensation, the constant /;;* can be equal in the dismuation model to the number of the absorbed light quanta, but must be l/n times smaller in the straight reduction model, where 1/n quanta are needed to carry a single CO2 molecule through all four reduc- tion steps.) Assuming that the conditions (27.6) and (27.7) are satisfied, we can insert into (27.6) the equilibrium value (27.3) and obtain, for the carbon dioxide curves of photosynthesis, the equation: (27.8) P = 7lkt KaAo[C02]a/(l + i^a[C02]a) For various values of k* (e. g., for various light intensities, 7), equation (27.8) represents a family of curves of the "Bose type." These curves are hyperbolae : (27.9) P/(Pxnax. - P) = K,[C02] At the saturating concentrations of carbon dioxide, P approaches asympto- tically the maximum rate: (27.10) Pmax. = nk* Ao All carbon dioxide curves separate from the beginning, their initial slopes being : (27.11) {dP/d[C02].)a = nk*,K,Ao They all reach "half saturation" at the same carbon dioxide concentration: (27.12) l/JC02]a = 1/i^a The empirical carbon dioxide curves deviate more or less widely from this simple type: Even the "Bose type" curves, shown in figure 27. 2A, do not all reach half saturation at the same value of [CO2]. We can attempt to consider the curves (27.8), determined exclusively by static conditions, CAKBOK DIOXIDE DIFFUSION 921 as the "primary" carbon dioxide curves, and treat the empirical curves as if they were basically such curves, deformed by superimposed kinetic in- fluences. In the region of low carbon dioxide concentrations (or in the presence of poisons such as hydrogen cyanide), these additional influences comprise supply reactions of limited efficiency — slow carbon dioxide dif- fusion and slow carhoxylation. These two processes of limited, but [CO2]- proportional, maximum rate tend to impose a slanting "roof" on the P — /[CO2] curves and thus to convert Bose's several divergent hyperbolae into Blackman's single and almost straight line. In the region of high CO2 concentrations, the additional kinetic influences must be due to [CO2]- independent factors, such as catalyst deficiencies, which tend to impose a horizontal "ceiling" on the over-all rate, i. e., to produce "carbon dioxide saturation" of photosjiithesis even before the acceptor A has become sat- urated with carbon dioxide. (6) Diffusion Factors When diffusion and carboxylation are slow processes (more exactly, when their maximum rate under the given conditions is not rapid compared with the actual rate of photosjoithesis, P) the concentration [C02]a of the carbon dioxide molecules in the immediate neighborhood of the acceptor may decline during photosynthesis considerably below the concentration of the same species in the medium, [CO2]; while the concentration of the carboxylated acceptor, [ACO2], may decline markedly below the value corresponding to the thermodynamic equilibrium, as determined by equa- tion (27.3). The stationary concentrations, [C02]a and [ACO2], estab- lished under such conditions, can be calculated by the application of the law of mass action to the reactions (27.1) and (27.2): (27.13) [CO^la = (fcdlCOo] + A-;[AC02])/(A:aAo - A^JACO,] + k,) and: (27.14) [ACO2] = (A-a[COo]aAo)/(A;* +K + k^CO^U) Combined, these two equations give a quadratic equation for [ACO2] (and thus also for P) as a function of [CO2]. Its one physically significant solution is: k^k* Ao + Kkd + k*kd + A;,fc4COo] (27.15) [AGO,] = — - — ■ ' — ' kJc^Ao + = nKMk*kd/{K^k^k* + kd) The limiting slanting line (roof") is: (27.24) Piim. = nkd[COi] — an expression obviously representing the maximum possible supply of carbon dioxide by diffusion (multiplied by n) . The initial slope of the particular carbon dioxide curve, which, without the diffusion limitation, would have started with the slope equal to that of the limiting line (equation 27.24), i. e., according to equation 27.11, the curve corresponding to /r* = ka/K^Ao, is reduced, by slow diffusion, to one half its former value : (27.25) {dP/d[C02])o = nkd/2 More generally, a primary curve wdth an initial slope aka is reduced by the diffusion limitation to an initial slope of : (27-26) {dm]). = (^) "'^ Thus, primary curves with initial slopes between 10 and 100 ka, will be confined, in consequence of slow diffusion, between ^%\ka and ^^%Q\kd, i. e., their initial parts will practically coincide with the limiting straight line, and thus present a picture of the "Blackman type." On the other hand, primary curves that would have exceeded the limiting rate only by factor of the order of unity, as well as those that would have merely ap- proached, but never exceeded, this limit, will retain their individuality and nonlinear shape, and will show a gradual transition from the "Blackman type" to the "Bose type." A certain depressing effect of diffusion will be felt even in a curve the original slope of which was as low as 0. 1 ka- (The 924 CONCENTRATION FACTORS CHAP. 27 slope of this curve will be reduced by 10%.) This example of the "advance effect" exercised by a "limiting factor" according to the general laws of reaction kinetics has already been quoted in chapter 26. (c) Slow Carboxylation This case is contained in the aboA'c-derived general equations (27.16- 27.19), if one makes the assumption: (27.27) k^Ao<^kj This implies another inequality : (27.28) /crf [CO2 ]»/(-.' [ACOol (since A-,.-lo[CO,] > AvUiCO,], > A-JA]lC"().], = /.{[AC'O,] + A*, the last equation being the steady state condition for the complex ACO2). Conditions (27.27 and 27.28) reduce (27.13) to: (27.29) [COaJa = [CO2] as it should be when the diffusion supply is ample. Consequently, (27.14) and (27.15) are replaced by: (27.30) [ACO2] = A-.,[C02]Ao/(A:: + k* + A:, [CO-,]) and (27.16), by the much simpler equation: (27.31) P = nkM[CO,]k*/ik: + k* + k^iCO,]) This equation can be written as: (27.32) P/(Pn>ax. - P) = k,[CO,]/{k: + kt) These hyperbolae reach half saturation at: (27.33) ./JCOd = ^ (1 + I) (as could be derived also directly from equation 27. 17) . Their initial slopes are (c/. 27.18): (27.34) {dP/d[C02])a = nk^k^*/{K + k*) All curves (27.31) are confined under the "roof": (27.35) Pli:n. = nfcaAo[C02] an equation that can also be derived from (27.19), and obviously represents the maximum possible rate of carboxylation. By reasoning similar to that employed just above, it can be sho^vn that a "primary" carbon dioxide curve with an initial slope ak^A^in will have its slope reduced, in consequence of slow carboxylation, to a/ {a + l)/Cai4on. In other words, in this case, too, INTERPRETATION OF CARBON DIOXIDE CURVES 925 the influence of the Hmiting process is felt long before the rate approaches the limit. Carboxylation was treated so far as a one-step reaction, with a rate proportional to the concentration [C02]a- I* is known, however (c/. Vol. I, page 203), that carboxylation in photosynthesis is cyanide-sensitive, and thus undoubtedly a catalytic reaction. It must therefore consist of several steps, such as the formation of a substrate-catalyst complex, and the transformation of this complex. We have no information as to the precise nature of these steps, but it can be assumed that at very low [CO2] values, the rate-determining step will be one with a rate proportional to [CO2], e. g., reaction (27.36). Under these conditions, the above-given derivations will be valid for the catalyzed as well as for direct carboxyla- tion. At the higher [CO2] values, however, other steps or the carboxyla- tion process may become rate-determining — steps limited in their maxunum efficiency by the available amount of the relevant catalyst (carboxylase). A [C02]-independent "ceiling" will thus be imposed on the rate of the over- all process, which will be determined by the product of the amount of the catalyst available and the average time a catalyst molecule requires to complete the desired transformation. The specific form of the corresponding kinetic equations will depend on the postulated mechanism of the catalytic action, and, as stated above, we have at present no reasons to favor any one mechanism among the sev- eral compatible with our general knowledge of the mechanism of enzymatic processes. As an illustration, we will go through a calculation based on the simple mechanism: K (27.36) COo + Ea . Ej, ■ CO2 *.' (using Ea, as in Volume I, as symbol for the carboxylase). (27.37) Ea -CO, + A :^i=^ Ea + ACO2 (For the sake of simplicity, we neglect diffusion and use [CO2] where [C02]a should be used.) The equilibrium constants K, and K,^ are subject to the condition: (27.38) ifeA'ea ( = p >< ^ ) = K. We assume that equilibrium (27.36) is established practically instantaneously, and remains undisturbed during photosynthesis, but that equilibrium (27.37) is less rapidly attained, and can therefore be displaced in light.* Designating by El the total avail- * The assumption that (27.36), too, is a slow reaction would lead to more com- plicated quadratic equations for [ACO2] and P, similar to those obtained in section b for the combined effects of slow diffusion and carboxylation. 926 CONCENTRATION FACTORS CHAP. 27 able quantity of the enzyme Ea we can derive the following equations for the steady state: /( k*K k* \ [l + KACO2] + ^' [CO2] + y^o) (27.40) P = n/c*KaAo[C02] /( 1 + K.iCO^] + p^ [CO2] + ^ ) The term proportional to A;;.[C02] in the denominator imposes on these carbon di- oxide curves an "absolute ceiling" (i. e., a maximum rate independent of both [CO2] and A;*, and thus of 7): (27.41) PStJ: = nA^eaE^Ao (The lower index refers to [CO2], the upper to I.) This obviously is n times the maximum possible rate of carboxylation according to the mechanism (27.36 and 27.37).) For relative saturation, as a function of [CO2], we obtain: [C02 (27.42) P/(Pmax. -P) = '^.iiEt^L. 1 + {k*IK.^l) and for half saturating carbon dioxide concentration: (27.43) „JC0,1 = i (i|, + 1)/ (1 + ^J which reduces itself to: (27.44) ,/JC02] = UK, for high A;* values (strong light), and to: (27.45) ,/JC02] = 1/ifa for low kr values (weak light). Depending on whether K& < Ke, or vice versa, 1/JCO2] will shift upwards or down- wards with increasing light intensity. The initial slope of the curves (27.40) is: (07 An / rfP \ nktK^Ao ^"^'■^^^ \d[C02])o 1 - AfAe'aE^ The reduction in rate caused by Ea deficiency is, by comparison of (27.40) with (27.8): K ^^ A'e[C02] + 1 For saturating [CO2] values: (27-46a) ^ = VV+E!E^r:[C02] + l. (27.46b) /3 = V (' + sL) and is thus equal to 3^ for a light intensity k* — ^A^ea, which without enzyme deficiency would have given the saturation value (equation 27.41). Generally, the saturation INTERPRETATION OF CARBON DIOXIDE CURVES 927 level is reduced from nakeaE°AAo to { ) A-ea^^A^o. Thus, here again, the effect of fe) the limiting process becomes felt when the nonlimiting rate is still far below the imposed "ceiling." (d) Nondissociahle ACO2 Corn-pound. The Franck-Herzfeld Theory So far, we have considered the carbon dioxide curves as, basically, ACO2 saturation isothermals, merely distorted by slow diffusion, slow carboxylation and limited quantity of the carboxylase Ea. An alternative was mentioned several times before. The carboxylation equilibrium may lie practically completely on the side of association, and the effect of the factor [CO2] on the rate of photosynthesis may be due entirel}^ to kinetic phenomena, such as the limited rates of diffusion and carboxylation. The corresponding kinetic equations can easily be derived from the more general formulae given in sections h and c by putting fc^ = 0, i. e., assuming that the rate of decarboxylation is negligible. For example, if the carbon di- oxide limitation comes exclusively from slow carboxylation (while the sup- ply of carbon dioxide by diffusion is ample), we can start directly with equa- tion (27.31); omitting the k^^ term in the denominator of this equation, we obtain: (27.47) P = nKAo[C02]k* /(K + ^a[COo]) (If k* «: ks[C02], this equation is reduced to P = nA;*Ao as expected.) This equation, too, represents hyperbolic carbon dioxide curves: (27.48) P/(Pmax. - P) = K[C02]/k* with half saturation at: (27.49) ./JCOa] = A;*Aa and the initial slope: (27.50) {dP/d[C02])o = nkM It will be noted that, in this case, all carbon dioxide curves make the same angle with the [CO2] axis, independently of A:^ (i. e., of the light intensity). Half saturation, on the other hand, occurs at [CO2] values that are propor- tional to k* (i. e., increase with increasing light intensity). The same method can be extended, to account also for saturation ef- fects caused by limited amount of the carboxylating catalyst ^ai e. g., by assuming fc^ = 0 in formula (27.37). Equation (27.40) is reduced, by this assumption, to: (27.51) P = n2fefceaA:?AoE° [C02]/(KefceaEi[C02] + Kek* [CO2] + k*) The "absolute ceiling" of this family of hyperbolae, approached when both 928 CONCENTRATION FACTORS CHAP. 27 k* and [CO2] are high, is the same as in formula (27.41), i. e., equal to n times the maximum rate of formation of ACO2 by reaction (27.37). The equation of the individual carbon dioxide hyperbolae is : (27.52) P/(Pn,ax. - P) = {keJil + fc*)K%[C02]A* They reach half saturation at : (27.53) ,/JC02] = k*/K,(k,,El + fc*) and have the following initial slope (independent of k*, and thus of light intensity) : (27.54) {dP/d[C02])o = nKeke.A^El The assumption of a stable ACO2 compound was used by Franck and Herzfeld (1941) in their detailed kinetic theory of photosynthesis— the most elaborate to be found in the Uterature. The rather complex chemical mechanism on which the kinetic analysis was based was illustrated by scheme 7.VA in Volume I. It would have been inconvenient for Franck and Herzfeld to use a different postu- late concerning the stability of ACO2, because they assumed that the carrier is associated not only with carbon dioxide molecules, but also with seven reduction intermediates. To consider the equilibrium ACO2 ^ A + CO2 as reversible, while postulating a firm attachment to the same carrier of the reduction intermediates, would have meant added mathematical complications; while to assume a different reversible association equilib- rium for each intermediate (as was once suggested in chapter 9) would have been still more cumbersome. Thus, Franck and Herzfeld chose the simplest way when they as- sumed the complexes of A with CO2 as well as with all seven intermediates to be practi- cally undissociable (except by light — a complication discussed in Vol. I, p. 167, and in chapter 29). Because of the assumption of an undissociable complex, the derivation of Franck and Herzfeld contains no equivalent of section a. Since they did not take into account the effects of slow diffusion, it also contains no equivalent of section b. The aspects of the carbon dioxide supply problem that Franck and Herzfeld did consider were the phenomena treated in section c, i. e., slow carboxylation, caused either by low carbon dioxide concentration, or by carboxylase deficiency. Their treatment of these two effects was somewhat different from that given in sec- tion c because they postulated a different mechanism of catalysis. Instead of the reac- tion sequence (27.36, 37), Franck and Herzfeld assumed the following three reactions: K (27.55a) CO2 + A , A • CO2 (formation of a "loose" complex) (27.55b) ACO2 + Ea > AGO, + EI (catalyzed formation of a "stable" complex and inactivation of the catalyst) k' (27.55c) E; ^—^ Ea (reactivation of the catalyst) FRANOK-HERZFELD THEORY 929 Franck and Herzfeld assumed equilibrium (27.55a) to be established practically instantaneously, whereas reaction (27.55b) was assumed to have a finite velocity, and thus to be capable of becoming a "bottleneck" of photosynthesis. This occurs when either the substrate concentration [A-C02], or the enzyme concentration [Ea] is low, or, more generally, when the product of the two concentrations is small. According to (27.55), the concentration of the "loose" complex is: (27.56) [A-COs] = K [AJICO^] (For the sake of consistency — cf. equation 27.5- we use as equilibrium constant the in- verse of Franck and Herzfcld's K). The rate of the bottleneck reaction (27.55b) is: (27.57) {dlAC02]/dl) = ^•a[A•C02][E] = k^K[A][C02][EA] We assume— as we did in all our derivations so far— that no kinetic factors other than those connected with the supply of carbon dioxide affect the rate of photosynthesis. Under these conditions, the equations of the carbon dioxide curves can be derived by calculating the stationary concentrations [A] and [Ea], inserting them into (27.57) and then calculating the stationary concentration [ ACO2] by equalizing the rate of production of [ACO2] given by equation (27.57) and the rate of reduction of this product by light. In the Franck-Herzfeld theory, all carrier molecules, A, may be considered, for kinetic purposes, as attached to a molecule of the sensitizer (chlorophyll), so that the substrate molecules bound to A can undergo direct photochemical change; putting it more cautiously, only those A molecules are taken into consideration in kinetic equations that are attached to chlorophyll. Their total number can be designated as Ao. Simi- larly, all chlorophyll molecules are supposed to carry acceptor molecules, A; putting it more cautiously, only the absorption of light by those chlorophyll molecules that are as- sociated with A is taken into consideration. We designate the total number of such molecules as Chlo. This number probably can be reduced by certain inhibitors, such as urethan or other narcotics, that displace the acceptor A from chlorophyll. (The same applies, according to Franck, to "self-narcotization" by the unfinished product of photosynthesis to which reference was made in chapter 24.) In Franck's picture, the rate constant k* in equation (27.6) can be written as k*I: (27.58) P = -nid[AC02]/dt) = A;*7[AC0.2] Here, A;*7[AC02] is the rate of absorption of light by the acceptor A in combination with CO2, k* being essentially an average absorption coefficient of the specimen under investigation. The factor n is equal to 1 in the Franck-Herzfeld mechanism; but k*I is, in the steady state, not more than one eighth of the total rate of absorption. If one assumes, as Franck and Herzfeld did, that this absorption coefficient is the same whether chlorophyll is associated wath ACO2 or with any of the seven reaction intermediates {cf. scheme 7.VA), then in the steady state the amounts of [ACO2] and of seven intermediates must all be the same. This means that: (27.59) Ao = ( = Chlo) = [A] + [A -002] + 8 [ACO2] where Ao is the total available quantity of the acceptor A bound to chlorophyll. Equalizing (27.57) and (27.58), we obtain: (27.60) IACO2] = (k,K[\][EA][C02])/k*r The stationary value of [Ea] can be calculated by equalizing the rates of reactions 27.55b) and (27.55c); this gives: (27.61) [Ea] = A;:E°/(A:.[A-COo] + k:) 930 CONCENTRATION FACTORS CHAP. 27 After inserting (27.61) into (27.60), one can use (27.56) and (27.59) to eliminate [A] and [A.CO2] and to obtain the desired equation for [ACO2] — and according to (27.58) also for P — in terms of [CO2], Ao and E^. Because of the assumed two-step mechanism of formation of [ACO2] (with the loose complex A-C02 as an intermediate), the resulting equation is quadratic. Its solution — which again represents a family of hyperbolae — is a rather complex expression, and we do not need to quote it here. For high values of both [CO2] and /, the expression for P approaches asymptotically the value: (27.62) PZTr.: = nKKElAo/ikaAo + K) (with the lower index referring to [CO2] and the upper to /). In the two extreme cases, when either kaAo ]^ K, or vice versa, expression (27.62) reduces itself either to nkj^l (t. e., the maximum rate of reaction 27.55c), or to nA-aE^An (the maximum rate of reaction 27.55b). The relations are much simpler if the rate of reaction (27.55c) is assumed to be rapid (compared with the rate of photosynthesis P), so that [^a] ' is practically equal to zero, and [Ea] to E^. In this case, one obtains an equation containing only first power of P, the solution of which is: (27.63) P = k,,KElA,[C02]k*l/{k*I + k*IK[C0.2] + 8 k^KEUCO,]) This is an equation of a family of hyperbolae (with / as parameter) : Half saturation (P = i/jPrnax.) is reached when: (27.65) VJCO2] = k*mKkn + 8 k^KEl) = 1 (^ + ^Sk.El/k^I)) It is shifted, with increasing light intensity, toward the higher [CO2] values. The initial slope of the curves (27.63) is: (27.66) {dP/d[C02])o = nkaKElAo Because, in formula (27.55b), the formation of ACO2 was assumed to be irreversible, the slope (27.66) is independent of light intensity. (In other words, at very low carbon di- oxide concentrations, all ACO2 formed is reduced by light, whatever the intensity of the latter.) Thus, as stated before, the carbon dioxide curves represented by equations such as (27.47) or (27.63) have more pronounced "Blackman characteristics" than the curves obtained with the assumption of a dissociable ACO2 complex. Of course, the hypothesis of a stable association of the acceptor A. with chlorophyll is independent of the other postulates of the Franck-Herzfeld theory; the latter can be combined also with the assumption of a dissociable ACO2 complex. (e) Back Reactions in the Photosensitive Complex So far, we have discussed the carbon dioxide curves only from the point of view of the "preparatory" dark processes at the "reduction end" of photosynthesis, since these are the stages of photosynthesis most closely related to the "carbon dioxide factor." The influence of the preparatory reactions at the "oxidation end" INTERPRETATION OP CARBON DIOXIDE CURVES 931 (such as the possible preUminary enzymatic binding of water) as well as the influence of "finishing" reactions (such as conversion of AHCO2 to a carbohydrate, and hberation of oxygen) have so far been neglected; while the role of reactions within the photosensitive complex proper was taken into account only by assuming the rate of reduction of ACO2 to be equal to the product fc*JAC02], where k% was considered a function of the illu- mination intensity, I. No new source of carbon dioxide saturation was added by this assumption; saturation remained determined entirely by the four factors treated in detail in sections a-d (limited quantity of A, slow diffusion, slow carboxylation and limited quantity of the carboxylase, Obviously, however, carbon dioxide saturation can also be produced by limitations of any of the other partial reactions in photosynthesis. For example, msufficient amount of a catalyst needed for the preliminary trans- formation of the reductant (H2O, H2, H2S. . .) might have the same "ceil- ing" effect on the carbon dioxide curves as the limited amount of the en- z>Tne (carboxylase) that catalyzes the prelmiinary transformation of car- bon dioxide itseK. Little could be gained by trying to write out equations for carbon di- oxide curves that would include the effect of a Umited supply of the reduc- tant. On the other hand, a few words may usefully be said about the in- fluence on the carbon dioxide curves of a limited supply of light. The kinetic equations that follow from the consideration of the prob- able forward and back reactions within the photosensitive complex proper will be derived in chapter 28 (p. 1020 ff.). They show, as expected, that the rate of absorption of light by this complex imposes a limit on the rate of photosynthesis that cannot be raised by increased supply of carbon dioxide (or change in any other external factor). Consequently, the "light factor" is in itself capable of producing a saturation effect in the carbon dioxide curves. For example, if we consider equation (28.14), derived from reac- tion mechanism (28.11), as an equation of carbon dioxide curves (i. e., if we treat 7 as a parameter), we find that, at high [CO2] values, the rate approaches the maximum: (27.67) Pmax. = nk*ICh\o which is the rate of supply of light quanta multiplied by the number n of carbon dioxide molecules that can be transfomied by a single quantum — perhaps 3^. Half saturation is reached, according to scheme 28. lA, at: (27.68) ,/JAC02] = k'/kr One could thus ask whether the interpretation of carbon dioxide curves really requires the assumption of an acceptor A, supposed to be present in 932 CONCENTRATION FACTORS CHAP. 27 limited quantities and carboxylated by a slow dark reaction; or whether one could perhaps explain all carbon dioxide saturation phenomena by ref- erence to limited supply of light. The occurrence of phenomena such as carbon dioxide "pick-up" after intense photosynthesis (Vol. I, page 206) provides, however, direct evidence that a dark carboxylation reaction actu- ally does occur, and that it has an effect on the rate of the over-all reaction of photosynthesis. On the other hand, it is undoubtedly true that some, at least, of the carbon dioxide curves, particularly those measured in weak light, owe their hyperbolic shape entirely or preponderantly to the limited rate of supply of light quanta. Reaction mechanism (28. lA, eqs. 28.20), which we have used, provides that the reaction of the primary photoproduct, HX-Chl-Z, with the oxi- dant, ACO2 (rate constant, A,), competes with the "deactivating" reaction that converts HX • Chi • Z back to X • Chi • HZ (rate constant k'). An alter- native mechanism (28. IB, eqs. 28.21) also is discussed in chapter 28, in which the "primary" back reaction of the photoproduct, HX-Chl-Z, is ehminated by immediate reaction of HX • Chi • Z with (free or bound) water, converting it to HX-Chl-HZ. This simplified mechanism will be used in chapter 28 to analyze another possible kinetic effect within the photosensi- tive complex — the accumulation, during strong photosynthesis, of the chlorophyll complex in the changed, photoinsensitive form (HX • Chi • HZ in the mechanism used). In deriving equation (28.14), the simplification [HX • Chi • Z ] «; [X • Chi • HZ ] ^ Chlo was made ; we do not make a similar assumption in respect to HX-Chl-HZ, but postulate that the "reduced" form accumulates and brings about saturation of light curves. (With in- creasing light intensity more and more chlorophyll complexes will be pres- ent, in the steady state, in the photoinsensitive form, HX-Chl-HZ.) This assumption leads to equation (28.27) for P as function of / and [ACO2], and to eciuation (28.28), if the equilil)rium value (27.3) is substituted for [ACO2] in (28.27). Considering (28.28) as equation of carbon dioxide curves (/ = constant), we obtain the following expressions for these curves: P — P k*I '^'•*'' "p.„. " f/(l + K.[C0,1 ) + k,K,A,lCO,] ("•™' V.ICOJ - J; (s;j^,) Equation (27.70) shows that the half-saturating carbon dioxide concentra- tion rises with increasing light intensity, and that the equilibrium constant Kg^ can be obtained by an extrapolation of 1/JCO2] to high light intensities (linear extrapolation with 1/7 as abscissa) . It will be recalled that in the case of the carbon dioxide supply limitation (c/. equation 27.22 or 27.33) we had to obtain the same value by extrapolation to low light intensities INTERPRETATION OF OARRON DIOXIDE CURVES 033 (/ = 0, and thusfc* = 0). We will return to the evaluation of K^ from car- bon dioxide curves in section g. (/) Acceptor "Blockade" In comparing the equations obtained in the preceding sections, with the empirical carbon dioxide curves, one has to keep in mind that, despite the considerable complexity of some of these equations, they all embody certain simplifjdng assumptions, and therefore cannot be valid except within a Hmited range of conditions. The only kinetic factors taken into account so far were slow diffusion of carbon dioxide, slow carboxylation and limited quantities of the acceptor A and of the carboxylase Ea. In section e we discussed the additional com- plications that may be caused by the deactivation of the primary photo- chemical product, HX-Chl-Z, in competition with its reaction with ACO2, or by the accumulation of the photosensitive complex in the reduced form, HXChl-HZ. A single slow preparatory dark reaction with a rate proportional to [CO2] plus the limited quantity of a single catalytic agent (e. g., Ea, A or Chi) could suffice to account for the increase of P at low values of [CO2], for the individual saturation of each carbon dioxide curve and for the oc- currence of "absolute" saturation (i. e., saturation with respect to both [CO2] and /). We know, however, that several catalytic steps of limited maximum efficiency play a part in photosynthesis ; and, although some of these steps are not directly associated with the assimilation of carbon dioxide, a limitation of the rate of the over-all reaction, whatever its source, must be reflected in the shape of the carbon dioxide curves, particularly in the region where they approach "absolute saturation." In Volume I, we outlined a general scheme of photosynthesis that in- cludes, in addition to preparatory supply reactions, two other main types of catalytic processes — "finishing" reactions, associated with the conversion of the first reduction products into carbohydrates and with the production of molecular oxygen. What happens to the first reduction product (which we will now desig- nate as AHCO2) can affect the rate of photosynthesis in various ways. If, for example, this product has to undergo a chemical transformation before it can be separated from the carrier A, this transformation may re- ciuire a certain time, so that, in intense light, a considerable fraction of the acceptor A can be "blocked" by AHCO2. Or else, the product AHCO2 may require a catalyst (^b) for its stabilization, and unless this catalyst is available within a sufficiently short time, AHCO2 may react back (with the oxidized photosensitive complex X-Chl-Z, or with the intermediates 934 CONCENTRATION FACTORS CHAP. 27 of the oxidation of water, such as A'HO, or with some other celkilar oxi- dants) to form ACO2 (or, according to a hypothesis of Franck, A + CO2). The latter possibihties will be discussed in section le of chapter 28. We will briefly consider here the effects of the "acceptor blockade." The conserva- tion equation (27.4) must in this case be replaced by: (27.71) Ao = [A] + [ACO2] + [AHCO2] Assuming that the restoration of the acceptor is a monomolecular process k' (27.72) AHCO2 '- — > A + HCO2 we can write the equations for the stationary concentration of ACO2 and (assuming the validity of 27.6) also for the rate of photosynthesis. Taking the simplest case — that discussed in section a (carboxylation equilibrium undisturbed by slow diffusion or slow carboxylation) — we obtain: (27.73) P= nA.K.[CO.W. l+TValCOo] (27.74) (27.75) The carbon dioxide curves (27.73) are thus hyperbolae the half saturation of which is shifted, with increasing light intensity, toward the lower concentrations of carbon dioxide (a shift opposite to that caused by slow diffusion or slow carboxylation; and apparently not encountered in experimental curves). Because of the occurrence in the denomina- tor of (27.73) of the product A;*[C02], the rate cannot exceed the "absolute maximum": (27.76) PZ^. = n/cj'Ao i. e., n times the maximum rate of restoration of the acceptor according to (27.72). (g) Calculation of Carhoxijlation Constant from Carbon Dioxide Curves If one wants to use the carbon dioxide curves for the determination of the carboxylation constant K^^, as was done in chapter 8 (Vol. I), one should avoid the region of "absolute saturation," and determine systematically, and with a high degree of precision, the carbon dioxide curves in the region where the yield is still proportional to light intensity. In this way, one could perhaps determine which of the several above discussed mecha- nisms of carbon dioxide supply provides the best interpretation of the facts. The experimental data available at present are neither exact nor systematic enough for such an analysis. INTERPRETATION OF CARBON DIOXIDE CURVES 935 Among the characteristics of the various theoretical equations derived above for the carbon dioxide curves, which might be used for comparison with the experiment, are the relations between half-saturating carbon di- oxide concentration and light intensity (c/. equations 27.12, 17, 22, 33, 44, 45, 49, 53, 65, and 75), and between the initial slope and light intensity (equations 27.11, 18, 23, 34, 46, 50, 54 and 66). The value of 1/JCO2] is independent of k* {i. e., of light intensity) and equal to \/K^ only to the extent to which the carbon dioxide curves are proportional to the saturation curves of the ACO2 complex (eq. 27.12). It has been suggested by some that this assumption is legitimate whenever the carbon dioxide curves are found to be hyperbolic. Thus, Burk and Lineweaver (1935), having satisfied themselves that the carbon dioxide curves of Warburg, Harder, James, van der Paauw, and Emerson and Green (Table 27.1) follow the hyperbolic saturation law, proceeded to calculate from them the carboxylation constants K^ and obtained values ranging from 1 X 10^^ to 10 X lO"'' l./mole (at room temperature). The corresponding free energies of carboxylation, AFa(= RT loge K^) are be- tween — 6.9 and —8.3 kcal/mole. It was mentioned on page 908 that, according to Whittingham (1949), the half-saturating CO2 concentration of Chlorella is shifted down to 0.5 or 1.0 X 10 -^ mole/1, if care is taken to avoid inhibition phenomena at low CO2 concentration, by allowing 2-3 hour induction period (if the cells were grown in high [CO2]) or, still better, by using cells gro^vn in low [CO2] (e. g., in air). The heat of carboxylation also was estimated by Burk and Lineweaver. From the absolute rate values found by Emerson and Arnold, at 6° and 24° C. in flashing light, they calculated AH a = 1.3 to 6.2 kcal/mole. However, Table 8. VIII shows that the fixation of carbon dioxide by organic molecules is accompanied — as is natural in reactions in which small molecules are attached to larger ones— by a decrease in entropy, —TAS being as large as +8 or even +16 kcal/mole at room temperature. Thus, if — AF of carboxylation is 7-8 kcal/mole, - AH should be of the order of 15-20 kcal— considerably larger than the estimate of Burk and Lineweaver. Our derivations in sections b, c and d show that the carbon dioxide curves may be strongly affected by slow diffusion, or slow carboxylation, without losing the hyperbolical shape. Equations (27.17, 22 and 33) show that these two factors cause i/2[C02] to increase linearly with increasing light intensitij. The constant K^ can in this case be obtained by linear ex- trapolation of 1/JCO2] to / = 0. Figure 27.9 shows that the data of Harder, Hoover and co-workers, and Smith (compare Table 27.1) extra- polated in this way, give 1/2CO2 values in the neighborhood of 5 X 10"® ilf, and thus K^ values of about 2 X lOS corresponding to AF = —7.9 kcal. /mole. 936 CONCENTRATION FACTOHS CHAP. 27 However, the linear extrapolation procedure is not always reliable. While slow diffusion and slow carboxylation tend to shift the half-saturat- ing concentration upward with increasing light intensity, other influences, such as the "acceptor blockade" (equation 27.75), or the deficiency of catalysts, may shift it downwards. Equation (27.43) shows, for example, that a limited amount of the carboxylase Ea may cause 1/JCO2] either to increase or to decrease with increasing light intensity, depending on whether Kf, is smaller or larger than K^. In this case, extrapolation of 1/2 [CO2] to zero light intensity should still give the correct value of K^, but systematic data will be needed to make the required nonlinear extrapolation possible. < llJ o z _ O V. O a, O E O < < CD < X Doto of Hoover et al I r I I I 2 4 6 8 10 12 14 16 18 20 22 24 LIGHT INTENSITY, klux Fig. 27.9. Extrapolation of 1/JCO2] to / = 0. Finally, under certain conditions the value of i/^ [CO2] may have no rela- tion to the carboxylation constant at all. This is obviously true if the ACO2 complex is practically undissociable. Equation (27.49), for ex- ample, contains only the kinetic constants k^, and 14 ■ Similarly, in section e, where we postulated that the carbon dioxide saturation is caused by "detautomerization" of the photocomplex HX-Chl-Z, equation (27.68) showed 1/2 [CO2] to be a function of the kinetic constant k^ and kr only. To sum up, the constant 1/Ks, may be ^ 1/JCO2] or )„ In this case, the thermodynamic carboxylation equilibrium condition (as- suming that the several carbon dioxide molecules in the complex are inde- pendent of each other) is: (27.83) [A(C02)„] = XaAo[C02]V(l + A^fCOa]") CARBON DIOXIDE CONCENTRATION AND FLUORESCENCE 939 consequently : (27-8^) [ACCOO-l.a.. - [A(CO.)] - ^"^^^^J One may use this derivation as starting point for speculations in which the "condensation" of carbon dioxide into a (7„ compound (perhaps, with n = 3 or 6) is assumed to be achieved, not after photochemical reduc- tion (or between successive photochemical reduction steps), but already in the preliminary, nonphotochemical carboxylation stage. One could quote in this connection the observations of Van Rysselberghe and co- workers (1946), who found indications that carbon dioxide is reduced, at the cathode of a polarograph, after preliminary formation of a poljaneric adsorption complex containing six CO2 molecules. Studies of the Hill reaction (chapter 35) and of the C(14) uptake in light (chapter 36) indicate three possible additional complications of the kinetics of carbon dioxide uptake in photosynthesis: (i) There may be several carhoxylations involved, e.g., one of the C2 -^ Cs and one of the C3 -^ C4 type. Since, in the steady state, both must proceed at the same rate, this complication may not change the kinetic derivations too radically. When, for some reason, one carboxyla- tion becomes a rate-limiting step, the other must be slowed doAVTi too, by the blocking action of its accumulated products. {2) Carboxylation may he coupled with hydrogenation ("reductive carboxylation," typified by direct conversion of pyruvic to malic acid, and already mentioned on p. 937). In other words, reaction (27.2), instead of being kinetically independent, may be combined with the first step in (27.5A). This, too, should not necessarily change the kinetic derivations in a radical way, since "coupling" probably means a fast sequence of two reaction steps, perhaps catalyzed by two active groups in the same protein molecule. (3) The acceptor. A, 7nay he a product of photosynthesis, disappearing in the dark. This would make Ao a function of the rate of photosynthesis, P, and would necessitate reconsideration of the kinetic equations based on a constant value of Ao. B. Carbon Dioxide Concentration and Fluorescence* In chapter 24, when discussing the fluorescence of chlorophyll in the Uving cell, we mentioned the close relationship often found between the in- tensity of fluorescence and the rate of photosynthesis; we stated that, because of this relationship, the measurement of fluorescence has become an important tool in the study of the kinetics of photosj^nthesis. Conse- * Bibliography, page "JGS. 940 CONCENTRATION FACTORS CHAP. 27 quently, we proposed to review the effects of various external factors on the intensity of chlorophyll fluorescence in vivo parallel with the presentation of the influence of these factors on the rate of photosynthesis. In following this plan, we have now to describe the changes in the in- tensity of fluorescence of living plant cells, associated with variations in the supply of carbon dioxide. The general finding appears to be that reduction or complete stoppage of the carbon dioxide supply usually affects the yield of fluorescence, ^, in a certain range of illuminations. It has no effect on the yield of fluorescence in very weak light; and probably also none in very strong light, but the latter generalization is in need of experimental confirmation. In purple bacteria, the commonly observed increase of ^ upon removal of carbon di- oxide sometimes gives place, with increasing light intensity, to the opposite effect (c/. fig. 28.30) . Since no systematic measurements of ^ for variable [CO2] have been carried out, the plotting of "carbon dioxide curves" of fluorescence, F = /[CO2], is not possible. Instead, "hght curves" have been drawn, showing the intensity of fluorescence as a function of light intensity, with [CO2] as a parameter, usually for two different concentra- tions of carbon dioxide only (or simply "with carbon dioxide" and "with- out carbon dioxide"). Several such curves will be reproduced in chapter 28 (c/. figs. 28.25, 28, 29 and 30). We anticipate here some of the facts to be presented there : The yield of fluorescence, ^p, generally remains constant (^^i) up to or beyond the intensity region in which photosynthesis becomes saturated with light; but sooner or later, it increases more or less gradually, finally to reach a new steady level, ^2- We designate the intensity at which the transition begins, as /', and that at which it ends, as I" . The effect of removal of carbon di- oxide api^ears to be a downward shift of this transitional range. Thus, McAlister and Myers (1940) found that in 4% CO2, the yield of fluorescence of wheat leaves was constant up to 600 kerg/(cm.^ X sec.) (kerg = 10^ erg) while, in 0.03% CO2, the yield increased (by about 15-20%) between 200 and 600 kerg/(cm.2 X sec.) {cj. fig. 28.25). Similarly, Franck, French and Puck (1941) found that the steady state fluorescence of a leaf of Hydrangea, at 7 = 7 kerg/cm.- sec. was about 20% higher in carbon dioxide-free air than in air containing 5% CO2. At 71 kerg/cm.- sec, on the other hand, the yield of fluorescence in 5%C02liad increased so strongly that now a change to 0.03% CO2 had no noticeable effect. Working with a culture of diatoms (Nitzschia sp.), Wassink and Kersten (1945) found the yield of fluorescence in intense light, 50 kerg/cm.- sec, to be higher in the absence than in the presence of carbon dioxide. Figure 28.28 indicates that, in this case, the difference is brought about by a de- CARBON DIOXIDE CONCENTRATION AND FLUORESCENCE 941 dine of

HX.Chl.Z). C. Concentration of Reductants* 1. Efifect on Rate of Carbon Dioxide Reduction in Bacteria The ultimate reductant in ordinary photosynthesis of green plants is water. The activity of water in the cells can be changed by direct hydra- tion and dehydration; or by immersion into solutions of different osmotic pressure. Both treatments have a considerable effect on photosynthesis. However, this effect cannot be treated as a kinetic phenomenon obeying * Bil)liography, page 963. 944 CONCENTRATION FACTORS CHAP. 27 the mass action law, since it is related to changes in permeability and other colloidal properties of the protoplasm and the cell membranes, which af- fect, to a varying degree, all activities of the living cell. Dehydration ef- fects were therefore discussed in chapter 13 (Volume I, page 333), where we dealt with various physical and chemical inhibitions and stimulations of photosynthesis. It was mentioned there (page 334) that one of the ways in which de- hydration may inhibit photosynthesis is bj'^ its influence on the stomata. Since the primary purpose of the stomata is to regulate transpiration, they 100 2 o < 75- < o 50- < b 25 - LlI q: - o 0 0 / y^ o 1 5% H2 1 1 1 1 10 20 30 40 50 HYDROGEN PRESSURE, mm. Hq 60 70 Fig. 27.10. Effect of hydrogen pressure on rate of hydrogen assimilation by Streptococcus varians (after French 1937). 114 mm.^ cells. Incandescent lamp below conical vessels; 25° C; argon + 5% CO2; rate in 95% H2 = 100. naturally close upon dehydration (by an increase in osmotic pressure in the guard cells; cf. page 912). In this way, the reduction in the activity of one of the reactants in photosynthesis (water) may result in the inter- ruption of the supply of the other reactant (carbon dioxide) . A more favorable opportunity for quantitative kinetic study arises when hydrogen, hydrogen sulfide, thiosidfate or other inorganic or organic substances replace water in the role of the reductant reacting with carbon dioxide in purple bacteria or so-called "adapted" algae (cf. Vol. I, chapters 5 and 6). Figure 27.10 (after French 1937) shows the effect of changes in the concentration of hydrogen on the rate of photoreduction of carbon dioxide by Streptococcus varians. The over-all reaction in this case is, according to Volume I (page 104), 2 H2 + COo ^ {CH2O} + H2O. The "hydrogen CONCENTRATION OF REDUCTANTS 945 curve" has the same typical "saturation shape" as the curves representing the rate of photosynthesis as function of carbon dioxide concentration; half saturation is reached at ca. 20 mm. (2.5%) H2, full saturation above 75 mm. (10%). Wassink et al. (1942) found, with Chromatium in nitrogen containing 5% CO2, no signs of hydrogen saturation up to about 2% (cf. fig. 27. 11 A); single experiments at higher pressures indicated that saturation was reached probably at about 10%, certainly below 15%. (The rates at 15% and 30% H2 were identical.) These hydrogen concentrations are approximately 100 times higher than those required for saturation of photo- synthesis with carbon dioxide (about 0.03% CO2 is needed for half satura- tion). In analogy with the interpretation of carbon dioxide saturation. E E O o Ll. O UJ < I- 0. 3 200 Scale of concentratjon 10 times reduced 100 200 300 400 AMOUNT OF Hj/VESSEL, mm^ Fig. 27.11. Rate of carbon dioxide reduction by Chrornatium (after Wassink 1942): (A) effect of [H2]; {B) effect of [H2S]. 5% CO2 in N2; pH 6.3; 29° C; strong light. given in the first part of this chapter, the simplest explanation of hydrogen saturation is to assume reversible formation of a hydrogen acceptor com- pound, with a dissociation constant of the order of 0.02 atmosphere (as compared with 3 X 10 ~^ atmosphere for the carbon dioxide acceptor com- pound). Thermodynamically, reversible hydrogenation (i. e., hydrogena- tion with energy close to zero) presents no difficulties, since the free energies of hydrogenation of organic compounds can be either positive or negative, depending on the degree of resonance stabilization of the individual com- pound in the hydrogenated and the dehydrogenated form (c/. Vol. I, page 217). Kinetically, however, the problem is less trivial, since in vitro, no example is known of organic compounds behaving like metallic palladium, i. e., taking up hydrogen under high pressure and releasing it when the pressure is reduced. Quantitatively, the results of French with Streptococcus varians and of Wassink with Chromatiujn are similar enough to justify the suggestion that the hydrogen acceptor is the same in both species. This common ac- ceptor may be either the enzyme hydiogenasc, or the compound, desig- 946 CONCENTRATION FACTORS CHAP. 27 nated in chapter 6 by Ah (i. e., the compound assumed there to be hydro- genated to AhH2 by the mediation of the hydrogenase). However, in interpreting the hydrogen saturation of photoreduction, one has to keep in mmd that hydrogenation equilibrium is only one of the two possible explanations. As in the case of the carbon dioxide curves, the hydrogen curves may also be affected — or completely determined — by kinetic influences, such as slow rate of hydrogen supply and slow hydrogena- tion (in the linearly ascending part), and limitations of light supply, of oxidant supply, or of the availability of a catalyst, in the horizontal part of the curves, which follows saturation. ro . e E O o u. o UJ a. 3 iVJVJ 150 100 y 50 / ' 1 1 1 1 .J 0 0.2 0.4 0.6 0.8 1.0 1.2 THIOSULFATE CONCENTRATION, % P^ig. 27.12. Thiosulfate concentration and rate of photo- reduction of carbon dioxide in Chromatium (after Wassink, Katz and Dorrestein 1942): 5%C02; pH6.3; 29° C; strong light. Wassink (1942) made reaction rate studies on the same organism, Chro- matium, also with varying concentrations of gaseous hydrogen sulfide as re- ductant. As shown in figure 27.1 IB, the (initial) rate was found to be proportional to [H2S] up to about 2% (in nitrogen) ; no signs of saturation were noticeable in the investigated range of concentrations. Higher doses of hydrogen sulfide could not be used because of rapid poisoning. The influence of the concentration of thiosulfate as reductant was also studied on Chromatium. Eymers and Wassink (1938) first found the rate to be constant between 0.16 and 2% thiosulfate, indicating that the satu- rating concentration was <0.16%. More detailed measurements were made by Wassink, Katz and Dorrestein (1942), who detennined a complete "thiosulfate curve," P = f [thiosulfate], reproduced in figure 27.12. It shows full saturation near 0.5% and half saturation a little below 0.1% thiosulfate (corresponding to 0.06 mole Na2S203/l.). CONCENTRATION OF REDUCTANTS 947 As with variable [CO2], the effect of variations in the concentration of the reductants disappears in weak Hght when the supply of light quanta becomes the rate-determining factor (c/. fig. 28.5B). Photoreduction with mixed reductants offers an interesting kinetic problem. Wassink, Katz and Dorrestein (1942) conducted some experi- ments in which Chromatium cells were supplied with both hydrogen and thiosulfate (at pH 6.8). They used each reductant "in excess" (meaning that the quantity of each alone would have sufficed to produce saturation) and calculated (indirectly, from manometric measurements) the relative amounts in which the two reductants were consumed. The average was about two molecules of carbon dioxide reduced at the cost of thiosulfate for one molecule of carbon dioxide reduced at the cost of hydrogen. If the two acceptor systems (for H2 and H2S2O3) are separate, this result may mean either that the cells contain twice as much of the thiosulfate acceptor system as of the hydrogen acceptor system, or that the first acceptor re- acts (in the hydrogenated state) twice as rapidly with the activated photo- complex as the second one. On the other hand, if only one hydrogen ac- ceptor is present, and the two donors compete in supplying it with hydro- gen, then the result, as reported, is not very significant, since, in this case, the relative utilization of hydrogen and thiosulfate would depend on rela- tive concentrations (even if both reductants are present "in excess" — the only information provided). In chapter 22, we mentioned the spectroscopic experiments of the same investigators that made them think that the composition of the photocom- plex X-Bchl-HZ (Bchl = bacteriochlorophyll) may itself be specific for each reductant, i. e., that bacteria contain a multiplicity of complexes, X- Bchl' -HZ', X- Bchl" -HZ" • • •, adapted to the utihzation of the several reductants R'H, R"H, • • • • (This hypothesis was suggested to explain the multiplicity and varying relative intensities of the absorption bands of bacteriochlorophyll m vivo.) Wassink and co-workers now argued that, if this were true, the photoreduction of carbon dioxide by a mixture of reduc- tants would be additive, rather than competitive. Since rate measurements indicated competition (the rate of total gas consumption in the presence of both reductants always was lower than in pure hydrogen and larger than in pure thiosulfate), the Dutch investigators concluded that their earlier explanation of the spectroscopic phenomena w^as incompatible with the results of kinetic studies. However, the argument would only be fully con- clusive if it were definitely known that, under the conditions of the experi- ment, the rate of photoreduction was limited by the amount of the reduc- tant available for reaction with the photocomplex. It was mentioned above that saturation of the over-all rate with respect to the reductant can often be due to a limitation elsewhere in the photosynthetic apparatus, e. g., to 94 S CONCENTRATION FACTORS CHAP. 27 the deficiency of a "finishing" catalyst such as Eg. Whenever this is the case, the rate of photoreduction cannot he increased by increasing the avail- able quantity of the reductant or by adding a second reductant and thus put- ting to work the (otherwise idle) photocomplexes specifically adapted to it. In kinetic work with purple bacteria, one has to keep in mind their capacity to utilize organic materials — including intercellular ones — as re- ductants. Competition rather than additivity seems to be the rule in this case too. The "sigmoid" shape of the light ciu'ves of hydrogen consump- tion, noted both by French, with Streptococcus varians, and by the Dutch observers, with Chromatium (cf. figs. 28.8), can perhaps be interpreted as a consequence of such competition: In weak light, intracellular reductants (pei'haps sugars or their derivatives) supply all the hydrogen necessaiy for the slowly i^roceeding ])hotoreduction of carbon dioxide, and, therefore, only very little external hydrogen is used. In stionger light, the diffusion supply of internal hydrogen donors proves insuffi(;ient, and the more rapidly diffusing molecular hydrogen takes over as the main reductant. To minimize the role of internal reductants, and thus to obtain light curves without a sigmoid initial section, Wassink (1942) has attempted to starve the bacteria before the experiment; however, he found no significant change in results. The experimental material on the "reductant curves" of j^hotoreduc- tion is as yet rather limited and no attempts have been made to represent these curves analytically. If one accepts the general scheme of photo- synthesis given in scheme 7.1 (Vol. I, page 153), the kinetic role of the re- ductants appears symmetrical to that of carbon dioxide. An analytical treatment of the effect of reductants on the over-all rate would thus have to deal with the same type of partial processes as were treated in the analysis of the carbon dioxide factor, namely, supply by diffusion, preparatory catalytic dark reactions (such as binding of hydrogen by an acceptor with the help of the hydrogenase), lilieration of the acceptor from the primary oxidation product, and "finishing" dark reactions (such as stabilization of the primary oxidation product). According to Franck's concept, re- peatedly mentioned before, the finishing dark reactions on the oxidation side (which consist in the elimination of the primary oxidation products — ■ "photoperoxides" — either by their conversion to molecular oxygen or by their reduction with reductants such as hydrogen or hydrogen sulfide) have the peculiar property that their failure to keep pace with the primary photoprocess leads not merely to the loss of a large proportion of primary products by back reactions but also to a reaction between them and oxidiz- able metabolites. This side reaction produces, according to Franck, a "narcotic," capable of enveloping chlorophyll and stopping the primary photochemical process. EFFECT OF REDUCTANTS ON FLUORESCENCE 949 The question whether the participation of water as reductant in photo- sj'nthesis involves some preUminary transformations similar to those of carbon dioxide remains open. The abundance of water in all cells may be advanced in favor of the convenient assum]:)tion that, even if a transforma- tion of this kind — c. g., hydration of a "water acceptor" — is needed before water can act as a hydrogen donor, the rate of this reaction is high enough to prevent it from playing a limiting role in photosynthesis. However, if the binding of water, which we have sj'mbolized by H2O — > {H2O) in scheme 7.1, and which we may now describe by the equation: (27.87) H2O + A' , A'H.O in analogy to equation (27.2), is an enzymatic process, the limited amount of the enzyme (particularly in an appropriately inhibited state) may well ])ecome a rate-limiting factor, despite the overabundance of the reactant H2O in the cell. 2. Effect on Yield of Fluorescence Hydration and dehydration of plant cells have a strong influence on the intensity of chlorophyll fluorescence in vivo; but in this case, as in that of gas exchange, it is difficult if not impossible to distinguish between the (undoubtedly possible) direct kinetic effects, and the indirect disturbances caused by changes in the colloidal structure of the pigment-protein-lipide complex. We must therefore refer here to the description of the relevant phenomena in chapter 24. The study again becomes much more fruitful when purple bacteria are used. The supply of reductants, such as H2, H2S or H2S2O3, has been found to strongly affect the yield of fluorescence of bacteriochlorophyll in these organisms. The light curves of fluorescence given in figs. 28.31, 28.32 and 28.33 (taken from the work of Wassink, Katz and Dorrestein 1942, on Chromatium) illustrate this phenomenon. The plots represent the fluores- cence intensity, F, (not yield, O K O o o S P5 a P 00 ca iC c o 3 4) » K w as E In " o" O O w It C— ' O) o o. 03 O O lO to to o iM o t^ r^ CO o c:,!3 A ^ -^^ ■^ O to »0 t^ CO (N O — ■ -- ,H ? A O»0i0i00t0000 to --^ CO to -— . .'^ O CO CO "^ "^ O •* CO i — ' ro CO OeO'^0>Ot^CDOOiOOi-''~-05(N ^ ' A M M M ACIC o o , V ' — ^ -• — ^ ^"^ f — V ,-^ y^ r — » J — V r^^ ^ CI iOtOiO,-^iOOOtotOOtO(MM (Xl\/ -to •«.-■ • -INM •-^ ^ ^^ ^ 2 2'^ I 7 a S a g 3 ^ p: o o to O o c c c O 3 S 3 >a m m m lO 'O CO I o o (M O) i-H I ,-1 c^i O O to to • ■ l^.-tiOtOtOcOtNtOtO'^iOo-*^ ;Vo'io '(nm'A.^ '^ I I II ^ a. a lO ■* _:_:— :— ;— ;— ;t^«o &; ^' ^»' ^ ? ^ OOOOOO^"*" COiOtotOtOiO^^O «S«E«XXXXXX5S 3 3 3 3 3 j^ 'S OOOo00050)050505 CO „ ^ 0) S a s 6 03 i: ■ ~ 0 ^ _» Vi > >d c« c o p3 T} ~ Oi o m o '■^ j: o w ^ < C ?: c > c fe J (In J Ph Gfl I— < 00 cc 03 a; o ►-! s o ■^ a> 1-1 to- to t^ co CO CO Ol O) 03 o PQ cQ S m K S 5 ^ fi O S a o a CO 3 ■ts J3 U C3 O 13 CO LIGHT CURVES OF PHOTOSYNTHESIS 967 o to I o CO CO I A A A A ,-v^-,^,,-,^-,.^'-nOOOOOOOOOOOO>00'C oooooioiOinoovO'-cccoomMoOh-asro^cC'-i O Ci ;0 O "O t^ r^ 2-'~"£i ? "T A A A A -^ ^ O 00 IM O O 'O lO o 1-H m o M '^ —' ■# o IN m f- OS o o o o ■* in O 00 O 7 ^^ 5 a o o o o m o »n ea « C-) -^ —' CO r-i A A t-oc^iooinooco CO (N 1-t t-< caaaaaaa ^c3e333cSo3cJcJe3ds3^c3s3e3 CO as .h .h .i; OS o3o903^^cdo303aS j<<<0»nt^iOmiOt^lOiCt^iC»nt^iCiO COCOCOCO(NC^i-ilMcnP3ffl 0 2 3 5 a t « S = s C c3 S a ^^ ■^ [B CO ~ .5 s ^^ 60 IN 99 ^ >. IN .a O 0) « a 03 p s o JS O Q M2 E-i U 03 3 K (U > > V oi cs « di hJ H CO o OS •a J2 a (N "*i •* ■* 02 05 i—t •^ U V -M o a a Q a CO tH IV •c Fi ^ dl c9 ui O M is ooooooooo — 'OCOOOOOO 1 C-l c-l C-1 -r ca CO lO o lO o lO ^'^ • CO CO CO to > — ■ — ' — ' o I I CO CO t- b- co a a a a c3 cj ^ o3 ^ ':5 is; •? a ^' ^ '^ ^' o o o o - o o o o =^ lO iC in O H- j o o o o o o o o iC iC lO o ■^^ "^ ^1 ^ji '^^'^'^ eocococo C0C0COC0Oi»OiOiOiO -H ■* o (N O O r-> CO e a O 0 o — 1 •* o> •-• '- o o o o 6 6 6 6 o o o o S K K K ap water uffer No uffer No uffer No uffer No tiJWWWHmcofnm s o s o ,-A in t~ ■N CO CO as 02 Oi 03 o ^-« COOOOOOOOOOO'0 0 N »0 «0 O o .-' LO »o O O O O '-O o A u-^ O O O O C*5 O O O O O O 'O o i0 0i000000000»000 ^ ^ f <^ A ^ A ' ^ aaaas-o.acE.o.ao.aa ES2SSSSSS = c£S£ ^ ^' ^ ^' ^' ^ i i ^ i i W ^ ^ oooooooooooooo OOOOOOOOOOOOOO in lO >0 "O 1.0 "O IM IM (N C-l (N C-1 C> IN oooooooo COMMMCOCCf^'M'MO'NlM'NlM iO»ft»0»nkO»Oc3c3c3oicdo3c3o3 oj a; a> CJ c; i-i C^ •* O 05 ^^ 6 d d 6 d d z z z z z z 2. a. +i -tJ +i 03 3 33 .^2 o 33::;^32^^^^1 0. n. C C « M So 0 so » Ci 3 a ft. 0 ^; •— c> c^ s tS -c S § § s g c 0 Tl ■w CO at O "M ■0 05 00 00 "^ A z: O O A IM --I 000 '^ 0-\ Xl .00 ^ iC ^ a a a s e S S S 03 o3 ^ ^ d 10 >o O •* CO M a a S^^^' a a a a [^ ^ c3 o3 33 33 C3 cS o3 33 000 00 g22 22ZZZZ 03 moo (M (N Cl e a s Oi Oi Oi 6 6 6 z z z k. u u Oi f^ IN IN C) ^ -? Oi 05 CO as 00.... O O o o o o ^ Z Z Z Z .— t .— < 1) ^ qj aj . . tc in tt3 to ^ t- — — — — :.^ < ffl ffl m (5 000 z z z U kl ti a; 0/ ^ to to to 333 pa oa P3 s 0 0 p s <, =0 e 0 3 5 7: 0 « 0 !- V e e «1 0 0 C n a ^-^ <1) & a 3 3 f i. 3 P. s s> 3 ^n M ^ 0 e 0 0 0 ■c -T1 s C3 J- > >, ^ cj 33 03 T3 -O ro (N CO c^ -^ 0> 05 (N 00 o> -H c^ CO CO CO 0> Oi 0 02 o> ■a c F^ W > 0 Pi N 0« fi « ^ C' -ii G -^ C3 c tn OJ C3 C3 s H C3 is ts LICITIT f'X^RVES OF PHOTOSYNTHESIS OfiO A --»■.!• ,^ UO ^ 0 -1 O M "S 01 O A ^ in o rt CM A o »n '-^ t^ '-^ 00 ^ < O ■< LT •— > i~ o o o I': o IM O CO 6 (U B3 3 PS a-- m a W O O O 05 ^ C-l C^ — 1 o 2; ta 3 o . — o ^^ o 8=^ t^ P 00 (N 00 si si) o o c 5 i5 ■>*< a v z a --. ^ o o ic o CO CO 332 A I ■^ o o o o O »C r-H ^ A >n lo IN C<1 ^ ^ O O lO \ '» C-l CVJ , ■* t^ CC "O lO to '•' -^ rt "^ c-j -- 0 *? in _: *- L- ^ - - 1 o 6 Sou i5 >--; 1,- ci ^ — d o to — 3 .2 to 5 O U H ."— V o "? rH 00 ^-' ITI 0 si ^« 5- ^ » U C! fe; ^ ^ ■* lo C) >-o CO CO ■* -* a a o> 03 c Z UJ o ,^ E C8 W PQ o o o o CO CO CO CO ^A I ( C O 00 c o CI CO CI CO O CJ 00 00 CO ^ ) ; z. s. c~ — 5 5 5 5 c3 7j cj ci z z z ?; Oi Oi Ci O". (M C) IM C^l »0 lO lO »0 I c '•^ o CO o 05 CO d "3 <1> o to w T3 J3 00 a> o CI 3 S3 bi Xi "< CO T3 ;« ? 0 fH > -P 'S CO s CO •-' -: s .3 CO I— < y-l CO h 00 3 d *'* >1 o tj o O O -^ . e g c C C B M g ^ o o C O ct « s tw .« O O O O •6^ o i o CI ■* c. r-< ^ a c "35 CO *S 03 ss ~ N 0 Q 00 3 si O to O ^ ^<2 970 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 12.5 0078,.^ _-::::r=: c 6 6 10.0 o .. s-s=:==^^I!r 'v- " ^^^——' . — -'■■■■■'0075 OI27 q' X ^^^"""^ 0 0496 !5 7.5 _J s w 5.0 00326 X y^ ^^-^ — " 00277 < 00262 8 25 ^^ ■ 0-0131 1^ 1 1 1 1 1 1 1 1 1 400 800 1200 600 LIGHT INTENSITY, foot candles 2000 Fig. 28.1. Light curve.s of photosynthesis in whole young wheat plants at 19° C. (after Hoover, Johnston and Brackett, 1933). Parameters: Vol. % COi; one foot candle equals about 10 lux. 8 X 10^ INCIDENT INTENSITY, erg/cm'^ sec. ' Fig. 28.2. Light curves of strawberry leaves (after Wassink 1946). Leaf disks floating on water in equilibrium with air containing 9% COj, at 17"'and 25°. Broken line: same, in Warburg buffer No, 9, at 25 °. LIGHT CURVES OF PHOTOSYNTHESIS 971 11 V) UJ I >- 1.0 - (/5 p 0 0.5 - /: X Q. / 0 - /" (.> 0 _l /o 2.5 2.2 1.9 1.6 1.3 1.0 100 120 Log r, lux Fig. 28.-4. Light curves of the water plant Cabomba caroliniaiia for various [CO2] values (left) (after Smith 1938). Warburg buffers, white light, 25.3° C. In the right- hand figure log P is plotted against log /. Ordinates are correct only for the uppermost curve; others have been displaced downward in steps of 0.2 log unit each, with correct positions given at right of figure. Parameters are buffer numbers (Table 8.V and foot- note o, p. 969). Left abscissa, klux. 972 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 150 £ 100 E CM O <3 o o /with CCj / ^ / ^ A^'"'^" without COz r 1 1 1 1 1 A 1 50 ^ I 23 4567x10* LIGHT INTENSITY.erg/cm^ sec. Fig. 28.5. Light curves of diatoms (Nitzschia sp.) in Richter solution (after Wassink and Kersten 1945). In 5% CO2 and in C02-freeair, 17° C. 10 , E E o o tl. o liJ < a. 3 250 - 200 - 57o COz ^•— °~ IbU 100 / / / 50 /' CQ 1 7„ COz 1 1 1 250 ^-'^■" y / / / 200 /|0% thiosulfote / / / ^_ . .' ^''ooe? thiosulfote 150 // 100 // / Q^ / 50 / / / / /ill Fig. 28.5 A. Light curves of CO2 re- duction by purple bacteria (Chroinatium D) at two CO2 concentrations (after Wassink, Katz and Dorrestein 1942). 1% thiosulfate, pR G.3, 29° C. Fig. 28.5B. Light curves of COo re- duction bj' purple bacteria (Chromatium D) at two different thiosulfate concen- trations (after Wassink, Katz and Dorre- stein 1942). 5% CO-,, pU 6.3, 29° C. Abscissa for both figures, incident intensity in (erg/cm. ^ sec.) X 10" LIGHT CURVES OF PHOTOSYNTHESIS 973 Fig. 28. G. Light curves of Chlorella at different temperatures (after Wassink, Vermeulen, Reman and Katz 1938). * 1.75 X IC* INTENSITY, erg/cm. sec. Pig. 28.7. Light curves of Chlorella at different tem- peratures (after Noddack and Kopp 1940). [CO2] = 7.87 X 10 ^ mole per Uter. 1 HK = 940 erg/(cm.^ sec). Up- per figure, white light; lower figure, red light. 25,000 10 15 INTENSITY, HK 974 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 250 200 - I 150 o 100 50 pH 6.3, l%thiosulfate - . ^o ^° 28° o / / . " /'■' ' A ■/ / n ... n r 1 ] J pH 76, l5%fhiosu Ifafe 500 - — — O ^- ■ 28° 400 / 300 1 22° 200 I /^u-u a 0 16° 100 1 . 1 3 X 10" LIGHT INTENSITY, erg/cm^ sec. 3 X 10* Fig. 28.8. Light curves of purple bacteria (Chromatium D) at different temperatures (after Wassink, Katz and Dorrestein 1942). (5% CO2, concn. 1, corresponding to ca. 80% absorption of sodium light; cf. fig. 28.22.) 200 0 05 ml 0 03% KCN/ml J L i 1.45x10* INTENSITY, erg/cm sec Fig. 28.9A. Light curves of HCN-inhibited Chlorclla cells showing that HCN is ineffective in weak light (after Wassink, Vermeulen, Reman and Katz 1938). GENERAL REVIEW 975 a: iLl > 600 400 < ^ 200 O no inhibitor • 2.5 X 10"" mole/hter NHjOH-HCl 20 40 60 80 RELATIVE LIGHT INTENSITY 100 Fig. 28.9B. Light curves of inhibited Chlorella cells showing that NH20H-HC1 is effective at all light intensities (after Weller and Franck 19-11). 200 E E 100 O no inhibitor • 0 05 ml. 50% urethan/ml. 150 X 10' INTENSITY Fig. 28.9C. Light curves of in- hibited Chlorella cells showing that ethylurethan is effective at all light intensities (after Wassink, Vermeulen, Reman and Katz 1938). Intensity is in erg/cm.^ sec. 2.00 F n S lOOi- 0.50 - o O 0.25 - X Q. 0.13 350 700 1600 4000 22000 350 700 1600 4000 22000 LIGHT INTENSITY, lux Fig. 28. 9D. Light curves of inhibited Chlorella cells sliowing that CuSO^ is effective, NiSO^ ineffective at low light intensities (after Greenfield 1942). 976 TflE LIGHT FACTOE. I. INTENSITY CHAP. 28 i 1.45 X 10* INTENSITY, erg/cm sec Fig. 28.9E. Light curves of Chlorella cells showing O2 effect (after Wassink, Vermeulen, Reman and Katz 1938). 250 - 200 - E E z" o o > LlJ 250 - 200 o without cyanide A 0003% KCN 3 xlO" INCIDENT INTENSITY, erg/cm.^ sec. 3 xlO* Fig. 28.10. Light curves of inhibited diatoms (Nitzschia) at 25° C. in Warburg buffer No. 9, showing both ethylurethan and cyanide (?) to be effective at low light intensities (after Wassink and Kersten 1945). LIGHT CURVES OF PHOTOSYNTHESIS 977 500 15% Hj, pH 6 3 ,o' INCIDENT INTENSITY, erg/cm sec 250 o o 200 150 I % H2S2O3, pH 6 3 \- (S) ^, without KCM o 100 < H .50 3 tr A .9/ o'/0.00l5 % KCN xlO" 0 I INCIDENT INTENSITY, erg/cm sec. INCIDENT INTENSITY, erg/cm sec 600 0 0187% NaNj xlO" INCIDENT INTENSITY, erg/cm' sec. 300 250 1% H2S2O3, pH 6 3 /-.^^" 1 2'7o eti./lure'^i" ^ X IC INCIDENT INTENSITY, erg/cm sec Fig. 28.11. Light curves of inhibited purple bacteria (Chromatium) (after Wassink, Katz and Dorrestein 1 942 ) . ( H. or H2S2O3 as reductant, 5 % CO2, 29 ° C . ) HCN shows no effect in weak light with Ih. With HoSaOj, effect is shown also iu weak light. Nn^OII-IICl shows no effect in weak light (c/. fig. 28.9B). With ethylurethan, the effect is particularly strong in weak light. 978 THE LIGHT FACTOR. INTENSITY CHAP. 28 300 250 £ 200 £ O o fe 150 LlI < t 100- O pH 6 3 A pH 76 / / / / / 50^ /.■ k I X 10' E e o o 600 500 400- i~ 300- u. o UJ < 200 y- Q. 13 100 15% Hs o pH 63 A pH 7.6 / / / / / y / / / / / / // // // X 10^ i_ INCIDENT INTENSITY, erg/cm^ sec Fig 28 12 Effect of pH on rate of COo reduction by Chromatium (after Wassink, Katz and Dorrestein 1942). 15% Ho (at right) or 1% H2S0O3 (at left) as reductant, 5%C02, 29°C. 300 20 40 60 80 100 INCIDENT INTENSITY 120 140 Fig. 28.13. Light curves in relation to age in Chlorella (after Wassink and Katz 1939). Gas pha.se air, 29° C. LIISTEAR RANGE 979 centimeter. For relationships between iV/,^ and intensity of illumination see chapter 25, page 838. Table 28.1 does not list the measurements of light curves in the presence of various poisons, such as potassium cyanide, hydroxylamine or azide, of narcotics, such as urethan, or of salts, such as copper sulfate. Several curves of this type are, however, reproduced in figures 28.9-11; for addi- tional information, we refer to chapters 12 and 13 in volume I, and to chapter 37. In the latter, we will also describe the light curves of algae in the state of (almost complete) anaerobic inhibition, which Franck, Prings- heim and Lad (1945) were able to measure by the very sensitive phos- phorescence method. Figures 28.1-28.13 contain a selection of typical light curves. Attempt was made to include curves for all types of plants — higher land plants, aquatic higher plants, green and colored algae, diatoms and purple bacteria. Figures 28. 1-28. 5A represent families of curves in which the carbon dioxide concentration, [CO2], serves as parameter (strangely enough, no such set is available for Chlorella). Figure 28. 5B shows light curves of purple bac- teria for two concentrations of the reductant (thiosulfate) ; figures 28.6- 28.8 represent curve sets with temperature as parameter. Figures 28.9- 28. 1 1 illustrate the effect of inhibitors. The effect of pH (in purple bacteria) is shown in figure 28.12, while figure 28.13 shows the influence of age. Later in this chapter, some additional sets of curves will be given to illus- trate the influence of inherited or acquired conditioning to strong or weak light. In chapter 26 we discussed three types of curve sets, P = f[Fi] with a parameter F2, which can be anticipated in photosynthesis. Examples of conditions under which each type can occur were given, for carbon dioxide curves, Fi = [CO2], in chapter 27. 2. Linear Range We will now consider some of the details of the light curves : the linear range, the compensation point, the saturating light intensity and the maxi- mum yield. Perhaps the most important quantitative characteristic of the light curves is the initial slope, which determines the maximum quantum yield; it will be discussed separately in chapter 29. Figures 28.1,28.7, 28.9A,28.10and 28.14A,B show that many lightcurves exhibit a practically exact proportionality between rate and light intensity over a considerable range of intensities. This "linear range" is less clearly delimited in figures 28.2-28.6. In the light curves of purple bacteria, it is often obscured by an inflection (c/. figs. 28.8 and 28.11A-D). Data col- lected in Table 28.11 indicate that (at room temperature and with an ample 080 THE LTGHT FACTOR. I. INTENSITY CHAP. 2S H m Z Z I— I H P3 C a H 03 CO a z Sh 02 O H O „ a PQ cu ^- & < (6 O Z >— I P3 '^con -f oooo, o o o o ' oooo, -^ IM -^ IM - lO lO lO 't 00 oo o o , o o CO -f ' GO X) • ■ '^ -H (M- CO cr. O GO IM IM I I I O O 'O iM — ' ^ -f t^ IM d CO IJ X 1, ^ 5 — "^ ■*"" "•"■ „X ^ "^ — ^ *'"' • p ■ p • ;^ — ^ ;=^ ^ -— . TJ '^i; S >T3'^ 1^ ^ ;^ >H >- >H >H Pi M IM ■ -—23 Qj a; ^ ^ ^ rs c c c b Oi PP OO POO ooo 5? lOiO ^^&^ 'o'c'c ■g'S > > > C T3:2 f , 1^ (N CO o o CO 00 -t bC a -tj oo— 1 66 z :- ooo •*-^ o CJ aJ aj rt +j -tj > c-1 e-1 l^^ ^A *■•* *-y> ^<^ ^> P9 PO CO OOOOO ooooo ^^ ^^ '^^ ^^ ^^ I-Ih HH HH HH HH WWW wwwww h-H l-H 1^ ^ X' + + + 5i^ „ „ _ c3 cS PPP -§,-§, OOO §■& hH h-i h-i Ci O l-l-l l-M HH t^ r^ f> '^ ►^ ,^ , ^^ t-H I— I h-" h-^ hH s s ^.5 OO (D (U 0) -r m iT^ IX) D S 3 « « =3 e e e •" QQ o O ^ ^ ^ ^ Q^ o o o o o o oooooooo 00 CO a o c ^ O IM '-5 . !- yn o X?" W CO C5 ..lii to fe: oo CO c?J2-i O ^ ■-*-, 2 s"!^ bio hO bC 03 o3 Cj (35 CO •5 --CO oD*--' o oj bc-n 05 CO o .a o W M T3 -a o o G5 o w ^" o -o o "A < 00 bb ^ CO 05 O to CO c c o o to to u ^ a s WW PQ" IM rJH bb2 t?.s LO -^^ 1^ C^ CO Jh ^^— I o si''- "73 to •*-' "iflw^ ,pil> 00 -G £ CIM C £ ^ bc 05 6 3 O o "o lO o d P O c s LO 00 o d COMPENSATION POINT 981 suppl}' of carbon dioxide) the linear range usually extends up to 5 or 10 kerg/cm.2 sec, corresponding to 1-2 klux of white light. In some cases, however, the first signs of curvature have been observed — despite ample supply of carbon dioxide — as earl}'- as at 1 kerg/cm.- sec, or 200 lux; while in others, the linear increase continued up to 50 or even 100 kerg/cm.'^ sec, i. €., 10-20 klux (c/. figs. 28.1 and 28.14B). Theoretically, no exact definition of the linear range can be given, since all light curves probably are hyperbolae (or curves of a higher order) and can only approach straight lines asymptotically. A formal definition of the upper limit of the linear range could thus be given only in terms of a definite deviation from linearity. 200- 5 10 15 20 LIGHT INTENSITY, einstein/cm.^ mm Hg. 28.14A. Approximate linearity of light curves of Chlorella in white light up to ca. 1300 lux (or G.5 kerg/cm.- sec.) (after Emerson and Lewis 1941). E 6 o I- < C/5 < _J < I- < 2 4 6 8 LIGHT INTENSITY Fig. 28. 14B. Light curves in purple hactciia in soilium light (after Eymers and Wassink 1938) (showing linearity up to GO kerg/(cm.2 sec.)). Light hitensity in (erg/cm.^ sec.) X 10^ Wassink (1946) gave incident intensities of monochromatic yellow light at which the yield of photosynthesis of nine horticultural plants showed 16% deviation from proportionality {cf. Table 28.11). Kok (1948,1949) and van der Veen (1949) found that the linear range may consist of two segments, the lower one up to twice as steep as the up- per one {cf. chapter 29, p. 1113). 3. Compensation Point The compensation point is the light intensity Ic at which photosynthesis is balanced by respiiation. so that the net gas exchange is zero. 982 THE LIGHT FACTOR. INTENSITY CHAP. 28 One could also call "compensation point" the carbon dioxide concentration at which the gas exchange becomes zero at a given light intensity {cf. chapter 27) ; or the tem- peratui-e at which the gas exchange becomes zero at a given combination of the param- eters I and [CO2] (cf. chapter 31); but the word is seldom used in either of these two ways. Sometime, the designation "upper compensation point" is applied to the second crossing of the curves of photosynthesis and respiration, which may occur either at very high light intensities or at "superoptmial" temperatures (cf., for example, fig. 31.1). When the carbon dioxide supply is not too low, compensation occurs within the linear range of the light curve, where the slope of the latter is determined by the maximum quantum yield of photosynthesis and the rate of light absorption {i. e., the optical density of the specimen). Probably {cf. chapter 29) the maximum quantum yield is approximately the same for all species (at least, when all cells are fully active — which is not always the case, e. g., in "aged" cultures). Differences in the compensation points found under these conditions must therefore depend mainly or exclusively on two factors: rate of respiration and optical density of the specimen. Respiration is proportional to the concentration of cells in a suspension ; Table 28.III Compensation Point of Leaves and Thalli Authority Plant 7 , lux Temp. HIGHER LAND PLANTS Boysen-Jensen, Mliller (1929) Fraxinus excelsior (shade leaves) 200 20° C. (sun leaves) 700 20° C. Fagus silvatica {shside leases) 150 20° C. (sun leaves) 500 20° C. MOSSES Boysen-Jensen, Miiller (1929) Marchavtia pohjmorpha Stalfelt (1939^) 6 species (winter), average 6 species (summer), average 100 390" 20° C. About 11° C. LICHENS (symbiotic GROWTHS OF ALGAE AND FUNGi) Boysen-Jensen, Miiller (1929) Peltiqera canina 4200 Stalfelt ( 19391) 12 species (winter), average 1020 12 species (summer), average 1160 '' 20° C. 13° C. AQUATIC HIGHER PLANTS AND MOSSES Plaetzer (1917) Elodea (summer) 2 Elodea (winter) 18 Caboiuba caroliniana 55 Miriophyllum spicatum 128 Fontinalis antipyretica 150 CincUdotus aquaticus 140 COMPENSATION POINT 983 Table 28. Ill (continued) AQUATIC HIGHER PLANTS AND MOSSES (continued) Authority Plant 7^, lux Temp. Harder (1923) Fontinab's (shade plant) 95 After 12 days in dark 64 After 20 days in dark 27 Fontinab's (sun plant) 152 After 12 days in dark 84 After 20 days in dark 10 Harder (1924) Fontinabs grown at 4.6° —1000 18° C. Grown at 20° —580 18° C. GREEN ALGAE Plaetzer(1917) Spirogyra sp. 174 Cladophora sp. 253 Ehrke (1929, 1931) Enteromorpha compressa 457 16° C. Ulva lactuca 357 16° C. Cladophora rupestris 322 16° C. Noddack, Eichhoff (1939) Chlorella ptjrenoidosa (thin sus- pension) 400 25° C. COLORED ALGAE Ehrke (1929, 1931) Fuciis serratus (brown) 408 16° C. Laminaria saccharina (brown) 345 16° C. Plocamium coccineum (red) 299 16° C. Phyllophora hrodiaei (red) 312 16° C. Delesseria sanguinea (red) 270 16° C. " Values remarkably high for shade plants. '' The high values of h are caused by the respiration of the (photosynthetically inactive) fungus. while light absorption, in an optically dense system, increases more slowly than proportionally with the concentration (Beer's law). Therefore, if we compare a dense suspension with a thin suspension of identical cells, we can expect to find the compensation point of the second one at a lower light intensity. AVhen, on the other hand, a decrease in optical density is brought about by a decline in the concentration of chlorophyll within the cells (without a change in the number of the cells per unit volume), the compensation point will be shifted in the opposite direction (i. e., toward higher intensities), because in this case the decline in the total yield of photosynthesis will not be compensated by an even stronger decline in total respiration. The respiration of chlorophyll-deficient cells is either the same as that of normal green cells (cf. Noddack and Kopp) or even stronger (chlorotic Chlorella cells grown by Emerson and Lewis; most sun-adapted, light-green leaves). Three examples of the latter behavior will be found in Tables 28.111 and 28.IV (p. 989). 984 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 The relation between the kmetic properties of sun-adapted (heho- I phiHc) and shade-adapted (umbrophiHc) plants will be discussed on page | 987; that between warmth-adapted (thermophilic) and cold-adapted (cryophihc) plants, in chapter 31. As shown by Harder 's data in Table 28. Ill, the effects of adaptation to weak light and low temperature differ in sign — the first one reduces respiration and thus shifts h toward weaker illuminations, while the second one enhances repiration, and thus shifts h toward more intense light. Algae that live deep under the sea, particularly red algae, are adapted both to weak light and to low temperature. The effect of umbrophilic adaptation predominates, however, and the compensation points of these algae generally are lower than those of the surface algae. Without low compensation points, these organisms could not develop 100-120 meters under the sea (the lowest level from which organisms have been recovered by dragnet) because the intensity of illumination at 120 meters depth is of the order of only 200 lux (see data of Seybold 1936, in chapter 22). As discussed in more detail in chapter 15 (page 424), the deep-sea algae are adapted not only to low light intensity, but also to predominantly blue- green light. If the compensation points of the green surface algae and the colored deep-sea algae were compared in blue-green light, the lower compensation points of the latter probably would appear even more strik- ingly than in Table 28. III. In general, the compensation points of different species shown in the table are comparable only if the experiments were carried out under closely similar conditions (same carbon dioxide concentration, temperature and previous history of the plants), since otherwise the intensity of respiration of a given species may vary wddely. Plants allowed to photosynthesize efficiently for some time often accumulate assimilates and then respire many times stronger than similar plants "starved" for an extended period of time (c/. Harder's data in Table 28.III). Such special conditions may perhaps explain the very low /, values found by Plaetzer for some aquatic plants. The ratio between respiration and photosynthesis at low light intensi- ties is generally changed in favor of respiration by an increase in tempera- ture (cf. chapter 31); thus, higher temperature must cause an upward shift of the compensation point (of. fig. 28.15 and Table 31. III). Narcotics have a similar influence, since they, too, reduce photosynthesis (at all light intensities) much more effectively than respiration (c/. chapter 12). En- zyme poisons (e. g., cyanide) may have a lesser or even opposite effect, because their influence on photosynthesis in weak light usually is rather small (c/. figs. 28.9A, 28. 11 A), while most of them strongly inhibit respiration. In certain algae (e. g., some Scenedesmus strains), the effect SATURATING LIGHT INTENSITY 985 of cyanide on respiration is stronger than on photosynthesis, even in strong Hght; in organisms of this type, addition of cyanide causes a strong down- ward shift of the compensation point (c/. chapter 12). Indications of a peculiar difference between cyanide effects of photosynthesis above and below the compensation point were mentioned in chapter 12 (Vol. I, p. 308). Reduced supply of carbon dioxide decreases photosynthesis without af- fecting respiration. If, in consequence of carbon dioxide deficiency, the light curves begin to bend in very weak light, the compensation point may be shifted to high light intensities (c/. fig. 28.15), or never reached o (3) (l)(2) Fig. 28.15. Shift of compensation point with changing carbon dioxide concentration. (1) — >- (2) decreasing [CO2]; (3) -*- (1) increasing tem- perature. at all. This case was mentioned in chapter 26, when we spoke of the ex- periments of Chesnokov and Bazyrina (1932) and Miller and Burr (1935), in which gas balance was observed at light intensities of the order of 20 klux. Miller and Burr (1935) noticed that, in this "carbon dioxide-limited" range, the compensating light intensity was independent of temperature. This means that the temperature coefficient of the carbon dioxide supply process (diffusion or carboxylation?) was practically equal to that of respira- tion. 4. Saturating Light Intensity When Reinke discovered tlu^ light saturation of photosynthesis, he found it to occur at an intensity close to that of sunlight at noon (*So = ap- 986 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 proximately 60 klux in moderate zones; cf. chapter 25). However, the saturating intensity varies widely from species to species and specimen to specimen. One reason for this is difference in optical density. Satura- tion begins when the most exposed chlorophyll molecules receive a certain hght flux, and becomes complete when the most deeply shaded molecules obtain this saturating intensity. The intensity of incident light at which this complete saturation occurs obviously must depend on whether we use a thick or a thin leaf, a dense or a dilute cell suspension. This "density effect" already was described in chapter 25 and will be again discussed later in this chapter (page 1007). Even with the density effect eliminated — either experimentally, by using optically very thin objects, or by calculation {cf. fig. 28.22) — the saturating light intensity still remains dependent, for a given species, on internal factors such as age and adaptation (to strong or weak light), and external variables, such as carbon dioxide supply and temperature. The effects of carbon dioxide concentration are illustrated by figures 28.1 to 28.5, those of temperature by figures 28.6 to 28.8. Using the notion of "ceilings" introduced in chapter 26 (page 869) we can say that everything that lowers the ceiling imposed on the over-all reaction of photosynthesis must shift the saturation toward lower light intensities. This may be a decrease in [CO2], a decrease in available reductants (in purple bacteria), or a decline in the amount of one of the catalysts. The temperature effect is complex, because changes in temperature affect all ceilings simultane- ously— those imposed by diffusion, as well as those caused by enzymatic reactions of limited maximum yield. Among the internal factors affecting the saturating light intensity, the most important is adaptation to strong or weak light. Shade-adapted plants often are darker green, i. e., contain more chlorophyll (per unit area or unit volume) than the corresponding light-adapted species or individuals. This difference in optical density would in itself be sufficient to cause changes in the shape of the light curves: Darker, shade-adapted plants are more efficient light absorbers, and their light curves should there- fore have a steeper initial slope. If the higher optical density is due to in- creased concentration of the pigment (with the concentration of all other constituents of the catalytic apparatus remaining the same), the saturation rate, related to unit volume of cells (or to unit area of leaves, assuming the leaf thickness is constant), should be the same for heliophilicandumbrophilic varieties; while the saturation rate related to U7iit amount of chlorophyll should be lower in the darker specimens. In practice, conditions are more complicated, because shade leaves often are thicker, and shade cells do grow larger than their heliophilic counterparts. These relationships will be discussed in more detail in chapter 32. The experimental result we SATURATING LIGHT INTENSITY 987 want to quote now is that the saturation rate of umbrophiHc plants usu- ally is much lower than that of the heliophilic plants, even if related to unit volume or unit area (not to speak of the rate per unit chlorophjdl con- tent). This indicates that adaptation to weak light involves, in addition to an increase in pigment concentration, a decrease in the amount of one or several catalysts that exercise a rate-limiting influence in photosynthesis. Coupled with steeper initial rise, this lower "ceiling" on the rate of photo- synthesis in shade plants often leads to a very early light saturation. While the light curves of sun-adapted plants may continue to rise at or even be- yond 100 klux (cf. data of Singh and Kumar, Smith, Boysen- Jensen, and Gabrielsen in Table 28.1), the light curves of shade plants may show saturation at light intensities as low as 1 klux (cf. figs. 28. IG and 28.18). 2- I ' o If) Peltigera CSJ ^(J O o a. -I E 0,^ Marchantia _L 9 II 13 klux 15 17 19 21 23 Fig. 28.16. Light curves of net gas exchange of an umbrophiHc moss (Mar- chantia) and a hehophiUc hchen (Peltigera) (after Boysen-Jensen and IMiiller 1929). The former is hght-saturated at 1 klux; the latter at or above 20 klux. The difference between the shapes of the light curves of heliophilic and umbrophilic land plants was first observed by Weis (1903), who compared the shade plant Poly podium with the sun plant Oenothera. This phenom- enon was also investigated by Lubimenko (1905,1907,19081.2,1928,1929), Boysen-Jensen (1918, 1929), Boysen-Jensen and Miiller (1929^) and Lunde- gardh (1921, 1922), among others. Typical results are illustrated by figures 28.16, 28.17 and 28.18. The first of these figures refers to an umbro- philic moss (which is compared with a heliophilic lichen) ; the second com- pares shade-adapted specimens of two aquatic plants with sun-adapted individuals of the same species and the third contains a comparison of the light curves of a shade-adapted leaf and a sun-adapted leaf on the same plant (see also Table 28. IV). We see that the effect of phylogenetic adap- tation (fig. 28.16) is similar to that of the individual adaptation of whole 988 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 plants (fig. 28.17) or single leaves (fig. 28.18). The figures in Table 28.IV further show that the respiration of sliade-adapted plants is weaker than that of the sun-adapted specimens. Fig. 28.17. Light curves of photosynthesis of shade plates (S) and sun plants (L) of the same species (after Gessner 1937). Former saturated at 40 klux; latter onlv far above 80 klux. 9 II 13 15 20 30 40 klux Fig. 28.18. Light curves of sun leaf (a) and shade leaf (b) of Fagus silvatica (after Boj^sen-Jensen and Miiller 1929). Former saturated at or above 30 klux; latter at 3 klux. AB.SOLTITK MAXIMUM RATE 9S9 Bohning (1949) noted that the rate of photosynthesis of shade-adapted leaves on trees of Pyrus malus decHned in continuous ilkimination of 32 klux from an initial value of about 20 mg. to < 5 mg. C02/(hr. X 100 cm. 2) after 20 days. Sun-adapted trees, on the other hand, showed no dechne during a similar period of continuous illumination, even in 50 klux. Kramer and Decker (1944) compared the light curves of white pine with those of three hardwood trees, and noted that the first one behaves as a heliophile and the deciduous trees as umbrophiles. This supports a previously sug- gested explanation of the fact that young deciduous trees "squeeze out" young pine trees on the floor of a forest. Table 28.IV Photo.synthesi.'; and Respiration of Shade Leaves and Sun Leaves OP THE Same Plant (after Boysen-Jensen and Muller 1929) Species Specimen «" P kiu'.net" praax.^ net pmax./;j Fraxinus excelsior Fagus sibatica Sun leaf Shade leaf Sun leaf Shade leaf 12 0.4 1.0 0.2 1.4 2 2 2.2 1.8 9.8 4.2 6.6 2.4 8.2 10.5 6.6 12.0 " Respiration {R) and photosynthesis {P), in nig. CO^/lOO cm."- hr. Lubimenko (1928) and Montfort (1934) found that some species have "rigid" umbrophihc or heliophihc characteristics, i. e., they are unable to adapt themselves to illuminations different from those to which the species as a whole has become adapted in its phylogenesis, whereas other species are capable of individual readjustment, as shown in figures 28.17 and 28.18 and in Table 28.IV. Umbrophihc character is typical also of algae that have been adapted — phylogenetically or individually — to weak light. Chlorella cultures grown in dim hght are richer in chlorophyll than those grown in strong light (c/. Table 25.1); the light curves of these "shade-adapted cells" rise more steeply, and reach saturation earlier than those of the "light-adapted cells." Similarly, van der Paauw (1932) found that Hormidium cells grown in dim light (2000 lux) become light-saturated in comparatively weak light (5000 lux), and show light inhibition in only slightly stronger light. Algae that live deep under the sea, particularly the red ones, behave as extreme umbrophiles, and their photosynthesis, too, reaches saturation at light intensities of the order of a thousand lux. 5. Absolute Maximum Rate It was mentioned before that simultaneous increase of carbon dioxide supply and light supply usually leads to saturation of photosynthesis long 990 THE LIGHT FACTOR. INTENSITY CHAP. 28 < w Co O HI (G B >: g 00 M n < o O K P^ O to p:3 & S H ►J O m ag O m ^'^. •sh -♦^ >) ^•.J3 '^C'tl ^ o ^ OJ tH ,* Pi ° «^ -« o o u ^ lO f« Xi o C-i 2x ^ S^c kl OT3 M^ G O S cS O „ OQ fl) Cl,"" Q o 00 1 — I 02 m -to-*-* CO o CO O 00 -* fO ^ 00 lo CO CO CO 00 ■-H CO 05 O 05 oo en O O 00 ^ o r-H.-i^^.-i C^i— iT-( OiMOi— I 1— I I— I ^H o '— ' O > ;^ t35 M CO ^ S ti '-'S CO is s ^ m e > en a s &5 ?5, O o bC 3 , o . a. o bO >> O (^ C O C O 01 ^ 0^ in; ' ■ m Oj > o 9 o bC bC e == s p CO 'T3 5^ <^ ^ "ij O -C) CO 5 .1 o ABSOLUTE MAXIMUM RATE 991 ^^ H H (3 4 Eg o o o > a o o a 9 e s O 00 00 e t- CO 00 e ^ (N (N CSI T T— 1 1— 1 i—i 00 CO o > o CO S •^ -S "C t- s 9- W o faC ^ 3 3 M 3 o3 o ii (-1 3 sc it: 3 3 J Si ^ . s ^ s tf s> ^ Cn g 3 O CO 00 CO o o CO o 05 1-H c3 ■^^ — ' a 3~ CO o ^ -f 1 ■-5 1 -^ r-* IB _a 02 M o pq O S :2 o o <^ M 't' Ol 2< CTJ GO CO 3 c c3 T3 O -3 a p 2 >^ o t- '^ bi \0 oo M bb e -c- 992 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 before "carbon dioxide inhibition" or "light inhibition" becomes apparent. We have therefore assumed that, independently of any inhibition, certain intrinsic internal factors (such as limited availability of certain catalysts) impose an "absolute" ceiling {i. e., a ceiling independent of both [CO2] and /) on the maximum rate of photosynthesis. Determination of this maximum rate is of interest from the practical point of view (estimation of absolute and relative efhciency of different plants as producers of organic matter) as well as from the point of view of the kinetic mechanism of photosynthesis. However, the two aspects call for different methods of comparison. The practical problem can best be answered by using unit surface as the basis of rate determination (since what one wants to know is how much organic matter can be harvested from a unit area covered with plants of different species) . From the point of view of a theorist, comparison should be based on unit cell volume, or unit chloro'phyll content, rather than on unit area. Willstatter and Stoll (1918) designated the maximum quantity of carbon dioxide that can be reduced in unit time by unit quantity of chlorophyll in a cell or tissue as "assimilation number" {va in Table 28. V), and the shortest time in which one molecule of chlorophyll can reduce one molecule of carbon dioxide {Ta in Table 28. V) as "assimilation time." (These constants will be ana- lyzed in chapter 32.) We designated, in chapter 27, the maximum rate of photosj'nthesis, at a given light intensity, reached with saturating concentrations of carbon dioxide, by Pmax.l we can use the symbol p™^'' for the maximum rate reached, at a given carbon dioxide concentration, when light intensity be- comes saturating; and the symbol Pmax! for the "absolute" maximum rate, obtained when both carbon dioxide supply and light intensity are saturat- ing. Table 28. V shows that for the leaves of land plants the values of PmTx. generally are of the order of 20 mg. COa/hr. 100 cm.^ of leaf surface, and sometimes reach 80-90 mg. Even aurea leaves, despite their very low content of chlorophyll, constitute no exception. Only some algae and aquatic plants investigated by Kniep (1914), Emerson and Green (1934) and Gessner (1938) fell far short of this production. A yield of 20-80 mg. CO2/IOO cm. 2 hr. — assuming it is reached in light of 40 Idux, 80% of which is absorbed by the leaf — corresponds to the conversion into chemical en- ergy of 4 to 16% of absorbed light energ}^ (in the photosynthetically active region, 400-700 m/x), and thus to a quantum yield between 0.018 iyio) and 0.07 (H4)- (This estimate is based on factors given in chapter 25.) The relation of these yields, obtainable in strong light, to the maximum quan- tum yields observed in weak light will be discussed in chapter 29. In the case of aurea leaves, the quantum yield in the light-saturated state appears ABSOLUTE MAXIMUM RATE 993 to be higher, .since roughly the same yield of carbon dioxide reduction is obtained here with a lower light absorption. However, the average ab- sorption of white light by aurca leaves can vary, depending on their actual chlorophyll content, from as low as 30% or less, to as high as 75% of that of normal green leaves of the same species. An estimate of the quantum 5'ield in the light-saturated state requires therefore that absorption de- terminations and yield measurements be performed on the same specimens. The fact that aurea leaves may absorb only a slightly smaller propor- tion of incident light than normal green leaves, caused Seybold and Weiss- weiler (1942) to consider their higher "assimilation numbers" (table 28. V) as irrelevant (and not — as assumed by Willstatter and Stoll — as a sign of exceptionally high capacity for photosynthesis). However, the capacity for photosynthesis f». the lighl-saturaled state, P"^^'., is not a function of the efficiency of light absorption, but a measure of the amount of a limiting enzyme present in the cells. The values of Pmax.' for aurea leaves show that in these leaves an abnormally low chlorophyll content is not accompanied by a proportional reduction in the content of the rate-limiting enzyme. Yields obtained by Noddack and Kopp (1940) with Chlorella pyren- oidosa, if related to dry weight, are higher than those given for most land plants in Table 28. V. However, because of the high concentration of chlorophjdl in Chlorella (3-4%, instead of 0.5 to 1% in leaves), the assimila- tion numbers are not higher, but somewhat lower, and the assimilation times somewhat longer than those given by Willstatter and Stoll for the leaves of the higher plants. Like the maximum quantum yield (at low light intensity) , the maximum rate of photosynthesis (in strong light) is a constant of the plant, i. e., it is independent of the optical density of the selected material. The only ex- ternal factor that affects it (apart from the presence of poisons or inhibi- tors) is temperature (as illustrated by figs. 28.G-28.8). It is difficult, if not impossible, to define the absolute maximum rate of photosynthesis also as a function of temperature. In short experiments, the highest rates can be obtained, with plants adapted to moderate conditions, at about 35° C. ; but, in prolonged experiments, "heat inhibition" is apt to occur even at temperatures as low as 22-25° C. (c/. chapter 31). We have used, in Table 28. V, mostly values obtained at 18-20° C, which are certainly smal- ler than the highest efficiencies of which most of the investigated plants were capable at higher temperatures, at least for short periods of time. The maximum rate of photosynthesis of a species or individual plant depends on adaptation to strong or weak light. As described on p. 986, shade-adapted species or individuals generally have a lower "ceiling rate," indicating a decreased content of a rate-limiting catalyst. In addition, they often show an early onset of light inhibition (c/. fig. 28.19). 994 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 max. ma.^. Since inhibition by excess light is a time effect {cf. chapter 19), the P values of shade-adapted plants change with the duration of illumination. We recall in this connection the time curves that Harder (1933) found for Fonfinalis antipyretica (fig. 26.8). The general impression made by these complex curves was that photosynthesis declined with time {i. e., the plants suffered light injury) whenever the illuminating light was more intense than the light to which the specimens were accustomed during the growth period. It was noted on page 987 that in the shade-adapted plants the apparent lower content of the enzyme responsible for the absolute saturation of photosynthesis is coupled with a higher content of chlorophyll. We will encounter, in chapter 32, other cases in which the content of the rate- limiting enzyme appears to be independent of that of chlorophyll (Chlorella cells grown in strong or weak light, cf. Tables 25.1 and 28. V; green and aurea varieties of land plants which were mentioned above, cf. Table 28.V and fig. 32.2) as well as cases in which these two concentrations change in the same direction {Chlorella cells made chlorotic by iron deficiency, cf. figs. 32.3 and 32.4). The shape of the light curves of shade-adapted plants has been much discussed in the ecological literature, particularly in relation to the photosynthetic production of aquatic plants at different levels under the surface. Even green algae, or submerged higher plants, found only a few meters under the surface, which should not have acquired ex- treme umbrophilic characteristics, were observed to produce a maximum of oxygen when placed at a certain depth, and to show light inhibition when exposed to direct sunlight. This, however, might have been, at least in part, a thermal effect. Much more pro- nounced optima on yield vs. depth curves were reported for the photosynthetic efficiency of colored (brown or red) algae at different levels under the sea. Ruttner (1926) and Schomer (1934) observed that several aquatic higher plants {Elodea, Myriophyllum, Cerathopyllum) had a maximum efficiency 1-5 meters under the surface. Curtis and Juday (1937) found similar optima for the green algae Ana- boena and Gloethea (in 9-10 meter depth). Van der Paauw (1932) found that Hormi- dium grown in a light of 2000 lux suffered light inhibition at 5000 lux. On the other hand, Gessner (1938) found no "optimum" in the light curves of shadow-grown or sun- grown Elodea plants in lamp light up to 30,000 lux. He tried ultraviolet light (360-400 m.y.) to imitate sunlight, but this, too, produced no inhibition. He suggested that the reported depth optimum of Elodea may be caused by chromatic adaptation (to bluish- green light) rather than by intensity adaptation. However, this explanation is im- plausible since it implies that photosynthesis can be inhibited by the addition of red and blue-violet light to green light, which has never been observed. Perhaps, carbon dioxide supply conditions are more favorable at a certain depth than on the surface, and this causes the rate to increase with increasing depth, as long as illumination remains suf- ficient for light saturation. Particular attention has been paid to the maximum efficiency and light inhibition of colored algae in relation to their vertical distribution in the sea. Engelmann suggested (see chapter 15, page 420) that the color of brown. ABSOLUTE IVIAXIMUM RATE 995 and especially of red, algae is the result of chromatic adaptation to the predominantly bluish-green light that prevails deep under the sea; Berth- old (in 1882) and Oltmanns (in 1905), on the other hand, thought that colored algae are adapted not so much to the spectral composition of light in their natural habitats as to its loiv intensity. The ensuing controversy — which led to almost complete vindication of Engelmann's theory of chro- matic adaptation — will be discussed in chapter 30. However, the funda- mental importance of chromatic adaptation for the composition of the pigment system of deep-sea algae does not mean that these algae are not also adapted to low light intensity and do not use the same mechanism — shifts in relative concentrations of red, blue and green pigments — for chromatic as well as for intensity adaptation (c/. Harder 1923). 6000 5000 Cladophora rupeslris, green Fucus vesiculatus, brown Ys ~^^hodymenia palmato, red \ .^ 60 75 117 140 S/90 LIGHT INTENSITY 197 S/2 Fig. 28.19. Typical light curves of red, brown and green algae (after Montfort 1929). Light intensity in relative units and frac- tions of full sunlight. Equal fresh weights of algae used. The response of colored algae from different depths to intense illumination has been studied, among others, by Maucha (1924, 1927), IMarshall and Orr (1927, 1928), Ehrke (1931), Curtis and Juday (1937) and particularly by Montfort (1929, 1930, 1933, i.^ 1934, 1936). Figure 28.19 shows a typical "optimum" curve, obtained by the last- named investigator. Montfort noticed that algae from one and the same level often show different resistance to strong light: Some red algae, containing much phycocyanin (such as Rhodymenia palmata), continued to synthesize effectively on the surface, while others, found in the same depth, but containing mainly phycoerythrin (such as Dehsseria alata) suffered a "sunstroke" and died. The surface-living, almost pure-green form of 996 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 the blue alga Gigarlina behaved like a typical shade plant, whereas the violet, deep-water form of the .same species, I'ich in j)h3'cocyauin, maintained its jjhotosj'nthesis at full ef- ficiency even in direct sunlight. 6. Maximum Rate and Average Rate of Photosynthesis under Natural Conditions The curve corresponding to [C'02] = 0.03% i.s theoretically not more important than any other light curve of photosynthesis, but its saturation value has a considerable interest because it represents the maximum rate of production of organic matter hy land plants in the open air. (In dense growth, or under otherwise abnormal conditions, the concentration of car- bon dioxide may vary between 0.01 and 0.1%, and this must affect the maximum rate of photosynthesis in some natural habitats.) It was stated in chapter 27 that with 0.03% carbon dioxide, and in intense light, the supply of carbon dioxide has a considerable rate-limiting influence, and the saturation value may therefore be below the "absolute" maximum, P|||"' at the same temperature. Table 28. VI contains some relevant experimental data (for a more extensive table, see Stocker 1935). Most figures in this table represent the net consumption of carbon dioxide. For strongly photosynthesizing plants, the corresponding values of true photosynthesis are 10-15% higher; but for weakly photosynthesizing plants (e. g., the arctic plants investigated by Miiller) the difference may be much larger, as illustrated by the figures in parentheses. Table 28. VI contains some striking contradictions, which remain to be elucidated. There is a general contrast between the P""' values found by Boysen- Jensen and co-workers (usually 1-10 mg. C02/hr. 100 cm.-, with the largest single values not exceeding 20-25 mg.), and the much larger values reported — often for the same species and under similar climatic conditions— by Kostychev and other Russian plant ph3^siologists (usually 10 40 mg. COo/hr. 100 cm.'-, with the largest single values reaching 80 or 100 mg.). Only in the case of arctic plants is there an approximate agreement between Boysen- Jcnsen's co-worker Miiller, and Kostychev and his co-workers. In the case of sun- adapted plants from moderate zones, the average of Danish measurements (section Ba of the table) is 13 mg., that of Swedish measurements (section B6), 16 mg., and that of English, Japanese and German measurements (with the exception of the early determina- tions of Sachs carried out by the half-leaf starch method), 10 mg. The average of the Russian analyses, listed in section Be, is as liigh as 24 mg. The results obtained by Kostychev, Bazyrina and Vasiliev (1927) by the determination of the synthesized assim- ilates did not differ significantly from those obtained by the same group by determina- tion of absorbed carbon dioxide. It was mentioned on page 908 that Kostychev and co-workers attributed the lower values of Boysen-Jensen to insufficiently rapid gas circulation. PHOTOSYNTHESIS RATK UNDER NATURAL PONDTTIONS 007 This was denied by Boysen-Jensen and Miiller (1029); but one notices in Table 28. VI that the newer measurements of the Danish school have given somewhat higher values than those of 1018-1920, and thus reduced the discrepancy between the averages in sections Ba and Be to a factor of about 2. In section C, containing plants from arid zones, we find a similar dis- crepancy l>et ween the result of Harder and co-workers in Algeria, and Wood in Australia (1-10 mg./hr. 100 cm.-), and those of Kostychev and Kardo- Sysojeva in Central Asia (20-70 mg.). In section D, practically all the listed values fall into the range 1-10 mg. (no Russian measurements are listed here, c/. however, the data of Kostychev and Kui'saiiov for the subtroiiical vegetation of the Black Sea littoral in .section Be). In the group of alpine jilants (section E), Monch (1037) and the Russians agree in finding the highest yields ever recorded under natural conditions. Earlier, Henrici (1018) had reported, for the alpine plant BelUs perennis, a yield of 232 mg. C02/hr. 100 cm.^ This value appears so incredibly high that we did not include it in Table 28.VI; but even the results of Blagoveshchenskij (1035) and Monch (1037) (00-100 mg./hr. 100 cm.^) indicate remarkablj^ high quantum jaelds (of the order of one CO2 mole- cule reduced per twenty quanta, in light of 80 klux, and \vith not more than 0.03% CO2 present). It should not be assumed that the carbon dioxide concentration was exactly 0.03% in all measurements listed in Table 28. VI. In Blagoveshchenskij 's experiments in the Pamir, for example, the [CO2] assays varied between 0.01 and 0.02%, and the highest yields were obtained at the latter concentration (which is. still considerably below the normal value of 0.03%). Stocker found, in the undergrowth of the tropical forest, [CO2] values up to 0.04%. (Compare also data given in Chapter 27, page 902.) To sum up, it is certain that plants growing in moderate climates can reduce, in their natural habitats and under favorable conditions, 20 or 30 mg. CO2 per hr. per 100 cm.- of leaf surface; but it is much less certain whether any plants — desert and alpine plants not excluded — are capable of yields up to 100 mg./hr. 100 cm.^, as is suggested by the measurements of Kostychev, Monch and Blagoveshchenskij. We will now discuss the relation between the maximum yield of photo- synthesis of which leaves are capable under favorable natural conditions, and the average production of organic matter by whole plants or large plants assemblies. Land plants in the open air, exposed to the sun, can be expected to main- tain the above-estimated rate of photosynthesis (about 20 mg./hr. 100 cm.^) for a considerable part of the day (barring such phenomena as the "midday rest" ; cf. chapter 26). The intensity of illumination is sufficient, 998 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 m Z O Q O O < Pi « a o z tn H o z > ^^ 00 ^ (M O n PQ to H K M o B O X o Q >-) H O •;30 SO . ti § S E a. ^ O A CO — '^ — ' ^ a a z o Vj m —> « t) O L- ,^ (M ^ Ci > T— ( (U ^^ ' ^ t< o 0) >J :;=! CO O i^ b^ -1H OOlMOO muiz/imm OO oo O o O ~ o =£0 e 2 « « S H H S 0Q02O Sl> W3 « 0 p ?^ e e -§ CO S: .0 3 e^tej 00 S > CO a z o IS) a a a o 05 05 00 05 1 C O S3 I e tn o CO C5 a a; tc 0; 1-3 I e CQ >> O >o CO C2 T-H CO — ',— , 0 x_^ Soi u » !3 0 ?^ cr ■>ri ^ cc CO ••^ ^ 0 f! « :S ;2 e vi f/n r^ "(^ ^.) 2Q^coCQO o .-O o .CO e s oc^ o o . IM „01 (M See 3^ -j:^ -o I'? GO CO OQ^co cr2^ m PHOTOSYNTHESIS RATE UNDER NATURAL CONDITIONS 999 (N oca c CO ^ -5 COS 3 CO o -3 a 03 <9 c is COOOCSOrfOOOO— lO ^H 1—1 C*l ^H C— ' m-Jim-Jiu^xjimmvi'Si O o 00 o -00 ■to ^ ^ S S e 2,^ s £ s -is- s § s^ s c -, ^ e s ■" ■?■ (^ !~ O 2 .eo Si, <5i S 00 -s ~ S. I4j h -* '^ 00 (M l^ r— < (N O > s ^ c > 0) o ■i-H 03 "m O ^ •^ ci oj C G .^ '^ ^ >> >> s: IS3 rt rt PQ P3 ^ fc > > QJ « r^ ri "o "S >. >> -IJ -^ m rr. O o k^ '^ HH hH > O o > o > 00 C2 o3 > O a X 03 c3 O CO > O m o a o to o eo C2 bC ;-. a; M 03 O o o S IB CO C5 > o c o3 ;-! 3 O fl) M t* Cl . fl3 tf > arj -< M * * CO 00 I I 00 (M 00 00 ^H ^H Cq I I I O C5 o . «^ -H -HiOXOiO'-i^fvjiOiOOOOOOOI^ ^^. (N'l>r>o'"0 CJ I^O lO o"u-0 C5^O00C:00OCi-J (M C -— 03 .. S ^ ^ S O) _3 03 > 0) bC o3 t> 03 > 03 3 o tc ■<• d O rt fl & Oikrrl -a 00 Sir c^ .- !^0 1Z^O b£ 03 Q S ^ i= si 03 g K c>-.i 03 y M .^'3 1 = 03 O 1000 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 O ■« <» "« o e o u 00 < S P e - a. i8 S e a CO a a O 00 03 73 "a o 05 o o tn o a G, S3 CO 05 3 73 03 03 s l-t o CO o X 03 CO X o o "-H O Ci bC <1 PQ lO — I cr. t^ r: 00 -H CQ iM o LO M' o iccoco -r ■ CO T i(M-t^lMiMOiMOOO -^ -u u ^ OJ > tc 0) c -a CC '^ c >^-l .2 bO C^ T3 CO ;3 (D ^05 -^ N 1 ^ CO -^ 73 « o--^ 3 o W < 1-5 - — -i-i N^-^ ^ ;.. t. o 'Jl N -^ -U2 o (N ^ iU > < min-J2-Jim'j^:nrjimxriuiu:xuim 'n _, ~ CO 50 O Si ^ ■§ ^ £ B . . O ^J o ca el nder tera m feci "iidom a. 31 !j c a. S £5 ^ 4; £ S "S s s h to hocnix da eriurn ole ila macro imon iastr ousinia v -== .~ ■^j ^ 1, p .u •Is a^^Njhjo "t; O =c;-:--< "^■--! u^ riTOTOSYNTTnESIR RATE FNDETl NATURAL CONDITIONS 1001 [C « C 'S. /*! ^;~* ,,— ^ ^— s. a ^^ 'S > > cu c3 rf HH •-5 i-s ^-^ ^ ^"^ 0 ,,— ^ ' ^ ^ 0 10 CO o y-^ c; QO f.^ •— ' ?— t f-t >. Ol 3 5 rf .4.:^ hJ 0 CJ e3 :0 o M o < o o lO :D CO (N O CO o" 00 CO "-^ 00 00(NOC2OiCC: ^COOCO 01 M > < o 00 0 00 0 0 00 00 C~ ^ c e ess e 1: 55: 1 e 2 s to 1 ia fistul ia fistul chocarp chocarp 0 i^ «o 0 OCi « 0; JU « 0 0 -w C3 ai SC ^ w « ~ "^ -S r^ ^ u e 3 0 03 < ■z O c3 0-, 10 CO 05 C > o CO CO C5 :0 <5 -*i O C O 00 o ■* O CO O '^ C". GCtCCOCOCCCO ud (N (N CO - 05 o a o '•+3 a a m -O Tithesis, P, and the relative satura- tion, p/p™'^^-^ against the incident energy flux, I, and the absorbed energy, la. If the rate of photosynthesis, P, is plotted against the incident light intensity, /, as independent variable (curves 1), the initial slopes of the R Dense 'Thin Fig. 28.20. Effect, of optical density of a cell suspension on shape of light curves. Heavy and thin double arrows represent the "linear range" of dense and thin suspen- sion, respectivel}'; y is the angle that determines the maximum quantum yield. curves vary in proportion to optical density, but the extension of the linear range must be practically independent of optical density (since the curve of the dense suspension must bend as soon as saturation begins in the surface layer). But if P is plotted against the absorbed energy, la (curves 2), the initial slopes must be the same, but the linear range of the thin suspen- sion must be shorter than that of the dense one. On the other hand, if we plot the relative saturation, p/p"^^^-^ against either/ or/g, (curves 3 and 4), the curves of the thin suspension will remain linear much closer to full saturation. As an illustration, figure 28.21 shows the light saturation curves, INHOMOGENEITY OP LIGHT ABSORPTION 1009 p/pmax. ^ j^js^^ q£ ^^^.q chlorella suspensions of different density {cf. Table 25.1) given b}^ Eichhoff (1939). Their relationship is in agreement with the prototype of figure 28.20 (curve 5). Katz, Wassink and Dorrestein (1942) attempted to reduce analytically the Ught curves obtained with three suspensions of bacteria (Chromatium, D) of different concentration to a single curve showing the average yield per cell as function of average iUumiiiation. In a 2 cm. deep absorption vessel, the "dense" suspension, 0.8 - Q. 2 4 6 LIGHT INTENSITY Fig. 28.21. Light curves of a thin and a dense Chlorella suspension, in red hght, X = 6500 A (after Eichhoff 1939). Intensity in "energetic meter candles" (HK) (page 1098). with a concentration of 30 "Trommsdorff units"/ml. (concentration 3) absorbed about 80% of incident light of a sodium lamp, the "medium" suspension (concentration 1; 10 Trommsdorff units), about 60%, and the "thin" suspension (concentration \; 2>\ units), about 30%. Figure 28.22A shows the empirical light curves of these three suspensions, P = f{I)- They have the relative positions anticipated in figure 28.20 {!) (except for the sigmoid initial shape, which is characteristic of the Hght curves of purple bacteria). Near 7 = 0, the order of the three curves is reversed. The probable reason for this is that, at a given incident light intensity, the average illumination is lowest in the densest suspension; therefore, the deficiency of hydrogen consumption (which we think, is re- sponsible for the sigmoid shape) is maintained, in the dense suspension, up to higher in- tensities than in the dilute one, and tliis influence apparently overcompensates that of stronger absorption. 1010 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 Figure 28.22B shows the same three curves, replotted to represent yield per single cell. The densest suspension now shows the lowest yield. However, all three curves appear to approach (as expected) the same limit- ing yield at high light intensity; thus, except for the sigmoid shape, they 1100 1000 900 800 :e 700 o 600 o u. 500 cone 3 cone I cone '/3 cone 1 control, run together with cone 3 cone I control, run together with cone '/3 < a. 3 400 300 200 INCIDENT INTENSITY, erq/cm sec. Fig. 28.22. Effect of cell concentration on light curves of photosj-athesis (Katz, Wassink and Dorrestein 1942). are of the type 3 in figure 28.20. (The latter refers to P/P"''-^; but, since pmax. jg proportional to the number of cells, P/P"^^^- must be proportional to the yield per cell.) Figure 28.22C, finally, shows the yields per cell as function of average intensity of illumination throughout the vessel (in A and B, the abscissae were the incident light intensities, as measured at the INHOMOGENEITY OF LIGHT ABSORPTION 1011 front wall of the vessel). The conversion is made by reducing the abscissae in the ratio a/c, where c is the concentration and a the per cent absorption, {i. e., by 0.8/3 = 0.27 for the dense suspension, 0.6/1 = 0.6 for the medium suspension, and 0.3/0.3 = 1 for the dilute one). This treatment causes the curves for c = 0.3 and c == 1 to coincide almost exactly; but the last point on the c = 3 curve still shows considerable deviation. There are two obvious reasons why one cannot expect the reduction method used to be completely successful. In the first place, the averaging cannot be quite correct, be- cause the cells are not actually exposed to the "average" light intensity, but some are illuminated with stronger, and some with weaker light. This would not matter if the yield were proportional to intensity; but, if the yield declines with increasing intensity (as it does in the saturation region), the yield that corresponds to a given average intens- ity will be lower whea the spread of actual intensities is wider, i. e., in the more concen- trated suspension. A second complication arises from the stirring of the reaction vessel, which causes the cells to come successively into light of different intensity. The effect of this varia- tion is complex; it belongs to the group of phenomena (induction; photosynthesis in alternating light) which will be treated in chapters 33 and 34. Only if the illumination cycles are much shorter than the periods required for the completion of all dark processes of photosynthesis can one expect the cells to work, in alternating light, with the same efficiency as in steady light with the same average intensity. The known periods of dark reactions, associated with photosynthesis, include at least one with a period as short as T = 0.01 sec. at room temperature; stirring is not usually rapid enough to send each cell through the whole cycle of intensities within 0.01 sec. (c/. chapter 29, page 1106). Consequently, the cells in the stirred vessel are illuminated with an alternating light the average frequency of which is smaller than 1/r. While the frequency of intensity variations is identical for all three suspensions, their amplitude is the larger the denser the suspension. Because of induction phenomena, the highest yield at a given average illumination is obtained in continuous light (cf. fig. 34.5); consequently, the efficiency losses caused by intermittency will be highest in the densest suspension. We have thus found two reasons, each of which may explain the deviation from the average of the last point in the c = 3 curve in figure 28.22C. It may be useful to note that the changes in the illumination of individ- ual cells, caused by stirring, may be discontinuous. The absorption by a single chloroplast {i. e., in the case of Chlorella, a single cell) is so strong that the only significant intensities of illumination to which a cell is ex- posed may be those with no cells or with a very small number of cells (1 or 2) between it and the light source. The scattering of light in the suspen- sions tends to smooth over these discontinuities. With respect to the inhomogeneity of light absorption, two cell suspen- sions with the same number of cells per square centimeter, but with a dif- ferent concentration of chlorophyll within each cell, offer a case similar to that of two suspensions of identical cells, but different dilution. Whether the light curves of such two specimens will present a picture similar to that sho-wn in figure 28.20, depends on their content of the catalyst that limit 1012 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 the rate in strong light. In the case for which figure 28.20 was drawn, the catalyst content could be assumed to be proportional to the content of chlorophyll (in other words, the "assimilation numbers" could be taken as identical), since the two suspensions differed only in quantity and not in quality of the cells. When, however, changes in optical density are brought about by ^■ariations in the pigment concentration ivifhin the cells, the "as- similation number" does not necessarily remain constant. We will deal with these relationships in chapter 32; of the three cases discussed there (umbrophilic and heliophilic plants, cf. figs. 28.16-28.18; green and aurea varieties, cf. fig. 32.2; and normal and chlorotic plants, cf. fig. 32.4), only the last one is characterized by approximate constancy of the as- similation numbers, and thus leads to light curves such as those in figure 28.20. In other words, in this case only does the intracellular content of the rate-limiting catalyst change proportionally to the content of chloro- phyll. (b) General Shape of Light Curves We now leave the effects of inhomogeneity of light absorption and in- quire into the intrinsic shape of the light curves. All kinetic interpreta- tions agree that the initial, almost linear segment of the light curves corre- sponds to a state in which the primary photochemical process is so slow that the catalysts which participate in the nonphotochemical steps can suppl}'^ the substrates needed for, and transform the intermediates formed by, the primary process, without depletion of the former or accumulation of the latter. Only the light curves of purple bacteria generally show a sigmoid-shaped initial part; the probable reason for this was discussed on page 948. In the linear section, the quantum yield of photosynthesis has its high- est value along a given light curve. (In sigmoid hght curves, the highest quantum jdeld is in the point where the tangent to the light curve passes through the origin of the coordinates.) This yield may correspond to actual utilization of all absorbed light quanta for photosynthesis, or to a certain, not further reducible proportion of quanta wasted by complete or partial inactivity of a certain number of cells, or of a certain fraction of chlorophyll. A similar irreducible loss of energy can be caused by back reactions, if the proportion of photochemical products that they destroy is independent of the rate of formation of these intermediates (cf. page 1037, and chapter 29, page 1137). The light curves bend toward the hoi-izontal when the rate of the pri- mary photochemical reaction ceases to be slow compared with the maxi- mum possible rate of one or several of the nonphotochemical processes as- sociated with photosynthesis. As demonstrated in chapter 26, the limit- GENERAL SHAPE OF LIGHT CURVES 1013 ing influence of a bottleneck reaction in a "catenary series" generally be- comes felt long before the rate of the over-all process has reached the maxi- mum speed of which this hmiting reaction is capable. Consequently, the light curves must approach saturation asymptotically rather than sud- denly "hit the ceihng" (even if we forget for the time being about the ef- fects of inhomogeneity of light absorption, which further enhance the grad- ual character of saturation). For the same reason, the maximum rate reached in the light-saturated state will often be considerably lower than the "ceihng" imposed by the limiting process. As to the nature of the processes that can cause hght saturation, the general alternative is between "preparatory" and "finishing" reactions. These two types of dark processes have been fu^st discussed by Warburg, and Willstatter and Stoll, respectively. Because all transformations that occur in photosynthesis must be cychc as far as chlorophyll and other cata- lysts are concerned, the question whether a reaction takes place "before" or "after" the primary photoprocess is not always as easy to answer as one would at first imagine. We will assume that a dark reaction precedes the photochemical step, if its retardation prevents the occurrence of this step (and thus also all the succeeding ones), and that a dark reaction /o^/oius the primary photochemical process, if the latter takes place in any case, and the effect of the limited rate of the dark reaction is merely to cause an accumula- tion of the primary photoproducts. Since experience shows that no large accumulation of oxidation intermediates occurs in photosynthesis (this is evidenced by the abrupt stoppage of oxygen production after the cessation of illumination), we must assume that the primary oxidation products ("photoperoxides") are unstable; unless rapidly removed or chemically stabilized by a "finishing" process, they apparently disappear by back reac- tions. The uptake of carbon dioxide may sometimes continue for about 20 sec. in the dark {cf. Vol. I, page 200, and Vol. II, chapter 36). This may mean that some intermediate reduction products survive for that length of time, or that the carbon dioxide acceptor, A, requires it to become re- carboxylated. (It may also be that the COa-acceptor is itself a reduction intermediate of carbon dioxide cf. chapter 36.) The two alternative mechanisms of light saturation can thus be de- scribed as starvation, which causes an "idling" of the primary photochemical mechanism, and constipation, which blocks the elimination of the primary products and compels most of them to return to their initial form. As described before, the distinction between the effects of prepara- tory and finisliing dark reactions becomes still more diffi(.*ult, if we follow Franck in the assumption that one of the finishing reactions "backfires," so that its slowness, like that of the preparatory reactions, affects the composi- 1014 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 tion of the photosensitive complex. According to Franck, this is the re- action that converts the intermediate oxidation products, formed by hght, into free oxygen (or sulfur, or other final oxidation products formed in bacterial photosynthesis). When this reaction fails to keep pace with the primary photochemical process, the intermediate oxidation products ("photoperoxides") accumulate in amounts sufficient to oxidize certain metabolites, thus forming a product of narcotizing properties (perhaps an organic acid). The latter is adsorbed on the photosensitive complex, and this retards or stops altogether the primary process. Each partial nonphotochemical process of limited maximum rate im- poses its o^^^l "ceiling" on the over-all rate of photosynthesis; and, since the influence of such a ceiling is felt long before it has actually been reached, the saturation value of photosynthesis in strong light may be affected not by one limiting process, but by several such processes— particularly since the maximum capacities of different parts of the photosynthetic apparatus appear to be of the same order of magnitude (as one would expect of a well- adjusted catalytic system). In the general discussion of the kinetic curves of photosynthesis in chapter 26, three types of curve sets, P = f{Fi) with Fg as parameter, were described and designated as the first (or "Blackman") type, the second (or "Bose") type and the third type, respectively (see figures 26.2, 26.3 and 26.4) . We recall that curves of the first type must arise when the parameter Fz determines the maximum rate of a partial process that does not depend on the independent variable, Fi. This process then imposes a horizontal ceiling on the curve P = f{Fi), but does not affect its initial slope. In curve sets of the third type the parameter affects the initial slope of the light curve, but not its saturation level; this type results when F2 codetermines the rate of a process that is also a function of the independent variable, Fi. In curve systems of the second type, the parameter F2 affects both the initial slope and the saturation level. Carbon dioxide curves offered examples of all three types, depending on the nature of the parameter {cf. page 868). Since most parameters do not affect the rate of the primary photochemical process, and therefore do not change the initial slope of the light curves, the P = /(/) curve systems usually are of the first type, i. e., the various curves coincide at low light intensities, but diverge at saturation. Such are most of the curve systems observed with carbon dioxide concentration as parameter (figs. 28.1, 28.2, 28.4, 28.5 and 28.5A), the only exception being Harder's Fontinalis curves (fig. 28.3). The two light curves of Chromatium with thiosuJfate concentration as parameter (fig. 28.5B) have the same general appearance, and the curve systems with temperature as parametei-, illustrated by figures 28.6, 28.7 and 28.8, are of the same type. The efiect of inhibitors, however, is uneven and some results are contra- GENERAL SHAPE OF LIGHT CURVES 1015 dictory. It has been suggested, as a generalization of empirical results, that catalyst poisons affect only the saturation level, thus producing light curve systems of the first type, while narcotics depress also the initial slope, thus giving curve systems of the second type. However, not all experi- mental results conform to this rule. In figures 28.9 A,B, the curves of Chlorella in the presence of cyanide are, as expected, of the Blackman type; but the hydroxylamine curves (as ob- served by Weller and Franck) are of the Bose type. With the diatom Nitzschia (fig. 28.10) the effect of cyanide was different: A distinct depres- sion was observed at all light intensities between 2 and 30 kerg/cm.^ sec. (However, the per cent inhibition increased with increasing light intensity, for example, 0.003% KCN caused an inhibition by 35% at 2.4 kerg, by 45% at 13 kerg and by 57% at 27 kerg.) Other observations of cyanide inhibition of photosynthesis in weak light were quoted in Volume I (page 309). The light curves of cyanide-inhibited purple bacteria, as observed by Wassink and co-workers, show a similar "semi-Bose" behavior; re- versing the results obtained by Weller and Franck ^^^th Chlorella, the bac- teria exhibited a Blackman type behavior toward hydroxylamine! The light curves of urethan-inhibited Chlorella (as given by Warbiu'g, and by Wassink and co-workers, respectively) appear to be of the second type, according to rule; but the third type is not entirely excluded. Bose type curves were found also with urethan-inhibited Nitzschia (fig. 28.10), although in this case the per cent inhibition was not constant, but rose with increasing light intensity. Wassink's curves, showing the effect of urethan on purple bacteria (fig. 28. HE), exhibit the reverse change — the depression is more pronounced in w^eak light than in strong light. The effect of oxygen on the photosynthesis of Chlorella (fig. 28. 9E; cj. also fig. 33, page 329, Vol. I) is of the Bose type, and the same is true of the effect of copper sulfate, while that of nickel sulfate apparently is of the Blackman type (fig. 28.9D). The curves of purple bacteria with pH as parameter (fig. 28.12) are of the Bose type with thiosulfate, and of the Blackman type with hydrogen as reductant. Theoretically, one can easily understand why specific catalyst poisons such as cyanide, should affect only the saturation level and not the initial slope of the light curves: The former is determined by the rate of a dark catalytic reaction, the latter by the rate of supply of light quanta. With the poisoning becoming more and more complete, the inhibition can be expected to spread to lower and lower light intensities. If one assumes (as suggested in Vol. I, page 307) that cyanide inhibits most strongly the carboxj'lating enzyme, Ea, the observed differences in the sensitivity of dif- ferent species may be attributed to variations in the amount of this en- zyme. Some plants may contain a considerable reserve of Ea, and there- 1016 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 fore show poisoning effects only in strong light; in others, the concentra- tion Ea may be just sufficient to maintain photosynthesis in the absence of cyanide and very little inactivation is needed to cause marked retardation even in moderate or weak light. However, even in this case, the per cent reduction of photosynthesis should remain smaller in weak light than in light of saturating intensity (as was, in fact, observed \vith cyanide-poisoned Nitzschia, cf. above). This hypothesis cannot explain an apparently uni- form per cent reduction of photosynthesis at all light intensities by a typical catalyst poison such as hydroxylamine (fig. 28. 9B). An ad hoc interpreta- tion, suggested by Franck and co-workers, was described in Volume I (page 312). What was said about the effects of catalytic poisons should apply also to deficiencies of the reaction substrates (carbon dioxide in green plants, carbon dioxide and reductants in bacteria). Their effects, too, should gradually diminish, and finally disappear with decreasing light intensity. The effect of narcotics on the initial slope of light curves can be under- stood if one assumes that they are adsorbed on chlorophyll (or "chloro- plastin") in such a way as to prevent the access of reactants or catalysts. Consequently, in the partially poisoned state, only the chlorophyll mole- cules free of these adsorbents are capable of properly utilizing the ab- sorbed light quanta. If it is true that light saturation is due to the limited amount of a catalyst, such as Ea or Eb, which is kinetically independent of chlorophyll, the fact that the light curves obtained in the presence of urethan appear to be of type 2 rather than 3 (z. e., that narcotics affect also the saturation rate) requires special explanation. For example, it can be postulated that the narcotic becomes adsorbed on the molecules of Ea or Eb as well as on those of chlorophyll. Alternatively, it can be suggested that, if the light curves of narcotized plants were followed to still higher light intensities, they w^ould finally approach the same saturation level as the light curves of normal plants (i. e., they would actually prove to be of type 3 rather than 2). A third explanation, based on the idea that definite catalyst molecules are "assigned" to definite chlorophyll molecules, and become useless when the latter are "narcotized," will be discussed in chapter 32. Still another phenomenon needs to be taken into consideration. It will be shown later in this chapter that a saturation level of photosynthe- sis probably exists which is due to the distribution of the chlorophyll com- plex, during photosynthesis, between the normal photosensitive form and a changed (tautomeric, or reduced) form. The latter is formed as an inter- mediate in photosynthesis, and requires a certain time for reconversion to the original photosensitive form. While this saturation limit may not be generally apparent in non-narcotized plants, because another limit (im- EFFECT OF PREPARATORY REACTIONS ON LIGHT CURVES 1017 posed by the deficiency of a finishing catalyst, Eb) is lower, the two may be not too far apart. Therefore, in the narcotized state, when a large fraction of the chlorophyll complexes is blanketed by the narcotic and therefore inactive, the saturation level due to chlorophyll can become lower than that due to the catalyst Eb- This will cause photosynthesis to be in- hibited by narcotics even in strong light ; however, the per cent inhibition will be smaller than in weak light. (This prediction is in agreement with Wassink's findings on purple bacteria, but not mth his observations on diatoms.) (c) Analytical Formulation: Effect of Preparatory Dark Reactions In chapter 27, a rather extensive effort was made to derive equations for the function P = /[CO2] under different assumptions concerning the preparatory dark reactions on the "reduction side" of the primary photo- chemical process. The influence of light intensity was expressed in these derivations {cf. equation 27.6) by assuming that the rate is proportional to the concentration of the reduction substrate, [ACO2], and that the pro- portionality factor, k*, is a function of the light intensity, /. The resulting equations for P were then applied to the analysis of the carbon dioxide curves, by assuming constant values of the parameter / {i. e., of fc^)- The same equations can, however, equally well be considered as analytical ex- pressions of the light curves, P = f{I), with [CO2] as parameter. A speci- fic assumption must be made in this case concerning the relation of k* to I (e. g., by postulating that /:* is proportional to /, k* = k*I, cf. eq. 28.13). The simplest equation for P in chapter 27 was equation (27.8). It was based on the assumption of a dissociable carbon dioxide-acceptor complex with no limitations on the rate of its formation. The corresponding light cui'ves are linear (at least, if k* = k*I), and show no saturation effects. (This is, of course, due to the fact that, in the derivation of equation 27.8, the equilibrium concentration of the compound [ACO2] was supposed to be undisturbed even by intense photosynthesis.) The equation of these straight lines is (using k* as independent variable) : (28.1) P = {nk*K,P^,{CO^])/{l + K.[CO,]) Their slope is proportional to [CO2] at low carbon dioxide concentrations and approaches a maximum at high carbon dioxide concentrations : (28.2) (rfPM-*)max. = nAo As soon as the assumption is made tliat the stationary concentration [ACO2] is affected by the rate of consumption of this complex by photo- synthesis {i. e., that the formation of ACO2 is not infinitely fast), the light 1018 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 curves cease to be straight lines and become hyperbolae. We can refer here to equation (27.16), which describes the combined effects of slow dif- fusion and slow carboxylation, or to equations (27.21) and (27.31), which express these two effects separately. In intense light, the hyperbolae p = f(k*) approach one of the following two saturation levels, either: (28.3) P"^""- = nkdlCOi] if the rate of diffusion is the limiting factor; or: (28.4) -?•"" = nk^ Ao[C02] if the rate of carboxylation is limiting. It will be noted that both satura- tion values are proportional to [CO2]; in other words, this approximation provides for no "absolute saturation" with respect to both / and [CO2]. Half-saturation is reached, in the case of limitation by diffusion, at: (28.5) r/k*r = {K,k,[C02] + 2 ki)/2 KM and, in the case of limitation by carboxylation, at : (28.6) V/"* = ^■^' + ^-'[COj] In both cases, half-saturating light intensity increases linearly with carbon dioxide concentration. The initial slopes of the light curves are the same as those of the straight lines (28.1). This means that at low carbon di- oxide concentrations, the light curve families are of the Bose type, the Blackman type being approached at the higher values of the parameter [CO2]. "Absolute" saturation follows, as usual, as soon as we postulate a rate- limiting step, the maximum velocity of which is independent of both [CO2] and light intensity, but is determined entirely by the available amount of a catalyst. Equations (27.40) or (27.51), derived for the case of limitation by carboxylase deficiency, contain in their denominators terms proportional to the product, k* X [CO2]; when light intensity and carbon dioxide con- centration are both very high, this term becomes predominant and the yield approaches the "absolute" saturation value: (28.7) P'S,H: = nA-eaEoAo which is the maximum rate of the catalytic formation of ACO2 by reaction (27.37). For the reaction mechanism discussed in section d of chapter 27 (non- dissociable ACO2 complex attached to chlorophyll for the duration of the eight photochemical steps, as in the Franck-Herzfeld theory), with the simplification (28.8) used in deriving equation (27.03), we have: (28.8) [EaI^E: EFFECT OF PREPARATORY REACTIONS ON LIGHT CURVES 1010 (28.9) P™-- = nA-aA'E!Ao[C02]/(l + A'ICOa]) (28.10) ,//-* = (Sk.KEnCO-,])/{l + K[C02]) (28.11) PZl: = nkJRlAo (28.12) idP/dk*)o = nAo/8 with n probably equal to 1. So far, we have considered the shape of light curves as determined ex- clusively by preparatory reactions on the "carbon dioxide end" of photo- synthesis (the possible rate-limiting factors being the constant of carbon dioxide diffusion, the bimolecular rate constant of carboxylation and the available concentration of the enzyme Ea). Analogous derivations can be made for limiting influences on the "oxidation end," such as, the rate con- stant of diffusion of reductants, the rate constant of their preliminary trans- formations (e. g., of the binding of hydrogen to an acceptor) and the de- ficiency of enzymes catalyzing these reactions (e. g., the hydrogenase). In making these derivations, we could, for example, set the rate of photo- synthesis proportional to the concentration of the primary oxidation sub- strate such as the hypothetical "bound water," A'H20 or, more generally, A'HR (instead of to the concentration of the primary reduction substrate, ACO2, as we have done so far). However, we abstain from a detailed dis- cussion of these possibilities, because, in the case of green plants, there is no positive proof that a dark hydration reaction actually is needed to make water available for the photochemical process. The abundance of water in cells may make this hydration, even if it were needed, practically instan- taneous. In the photosynthesis of purple bacteria , preliminary transforma- tions of reductants are known to occur, but no definite proof has as yet been given that these transformations must be considered as preparatory reac- tions {i.e., reactions providing the oxidation substrate for photochemical process) rather than as finishing reactions removing the primary oxidation products, formed by the photochemical oxidation of water. (The second alternative is favored by van Niel, Gaffron and Franck; cf. Vol. I, p. 168.) It must, nevertheless, be borne in mind that the rather detailed considera- tion of the preparatory processes "on the reduction side," and the compara- tive neglect of the analogous processes "on the oxidation side" of the pri- mary photochemical process, which is common to most discussions of the kinetics of photosynthesis, are not justified, being based only on our in- ability to study the fate of water before its oxidation in photosynthesis, and our present insufficient knowledge of the initial transformations of hydro- gen and other reductants used by bacteria. 1020 THE LIGHT FACTOR. I. INTENSITY CHAP. 2S (d) A nalytical Formulation : Effect of Processes in the Photosensitive Complex Leaving aside the effects of hypothetical preparatory reactions "on the oxidation end," we return to equation (27.6), P = nA'*[AC02], for closer consideration from the point of view of the Ukely mechanism of light par- ticipation in photosynthesis. As mentioned before, the assumption, more or less implicit in the derivations of chapter 27, was that k* is proportional to I, the intensity of incident light : (28.13) K = ^*I and consequently: (28.14) P = nl:*l.\C(),] = 'nk*ri\C(h] One remark, limiting the practical applicability of the analytical ex- pressions derived in this section, must be made immediately. Kinetic equations are based on the law of mass action ; they presume homogeneity of the reacting system. The light intensity, /, is, however, not uniform throughout a leaf or cell suspension; it varies even within a single cell or a single chloroplast. This complication has been repeatedly mentioned be- fore, and we shall return to it again on page 1044. In the meantime, we will proceed as if light absorption were uniform throughout the region under consideration. This means that our equations will be strictly valid only for optically thin layers. In the following equations, then, I must be taken as meaning the light flux actually reaching a chlorophyll layer, and not the light flux falling on the outer surface of the system. (These two fluxes are proportional to each other, but the proportionality factor varies with depth, as well as with the wave length of the incident light.) Practi- cally, most if not all kinetic measurements have been made, not with opti- cally thin pigment layers but with leaves, thalli or suspensions absorbing a large proportion (sometimes up to 100%) of incident light. We will con- sider on page 1044 to what extent kinetic relationships derived for optically thin layers are changed through inteji,iation over the path of the light in the system (and also over the differently absorbed components of non-mono- chromatic light). The treatment of this problem is further complicated by the structural effects discussed in chapter 22 (scattering and "sieve ef- fect"). Still another complication arises in the treatment of cell suspen- sions rapidly agitated during the measurement, thus bringing the indi- vidual cells more or less periodically into light fields of different intensity. If stirring were so intense as to cause each cell to slip through all the various light fields in a time which is short compared to the "Emerson- Arnold period" (about 10 -^ sec, at room temperature, cf. chapter 34), it would PROCESSES IN THE PHOTOSENSITIVE COMPLEX 1021 have been permissible to take into account only the average ilhimination and to consider the latter as identical for all cells. In other words, the rate of absorption of light by each cell could be taken as equal to the rate of total absorption in the suspension divided by the number of the cells in it. No amount of stirring, however, can mix the contents of the chloroplasts, so that chlorophyll molecules situated deeper inside them always will re- ceive less light than those situated on the illuminated surface. What is even more important, the rate of stirring usually is quite insufficient to make legitimate even the averaging of intensity for whole cells. Such fast stirring is difficult to achieve; it is unlikely that Warburg and Burk liad the right to claim that in their experiments (c/. p. 1006) stirring was so effective that only the average intensity of illumination was important. Often, a danger exists that the periods spent b\' individual cells deep in the suspension, l^etween two exposures to full light in the surface layer, could 1)6 long enough to cause induction losses during the subsequent expo- sure (c/. chapters 29, 33 and 34). These quahtative considerations show that the way to obtain light curves of photosynthesis best suitable for kinetic interpretation is by using optically thin suspensions or tissues. A limit to this procedure is, how- ever, set by the fact that even single chloroplasts may absorb up to 50% of incident light in the absorption peaks of chlorophyll (cf. fig. 22.35); so that diluting algal suspensions until they al)8orb much less than that amount (or using faintly green tissues such as green onion skins) may merely mean allowing a part of incident light to pass between the chloro- plasts—without improving the uniformity of absorption within the plas- tids. This uniformity can only be improved by employing cells poor in pigment (chlorotic cells), or by employing light which is comparatively weekly absorbed by chlorophyll (e. g., green light). We are thus forewarned that the several equations of the light curves, which will be derived below from alternative kinetic models of photosyn- thesis, can be used for comparison with the experimental curves found in the literature, only with strong reservations. We nevertheless consider it worth while to continue with these derivations, as a step toward a more quantitative study of the problem in the future. The latter will require both improved theoretical treatment (including the effects of inhomogene- ous structure and nonuniform light absorption), and, above all, precise kinetic experiments on optically thin objects. In analyzing the validity of equations (28.13) and (28.14), two alterna- tive pictures must be considered. According to one, favored by Franck and Herzfeld, the compound ACO2 is part of the "photosensitive complex" proper and its reduction can therefore be considered the primary photo- chemical process, e. g.; 1022 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 (28.15) ACOs-Chl-A'HoO , " ^ AHCOz-Chl-A'HO In considering this alternative, it is not necessary also to adopt Franck and Herzfeld's complex mechanism, which involves eight consecutive photochemical steps; essentially the same conclusions can be reached also by considering a single photochemical step, such as reaction (28.15), and leaving the completion of the process to nonphotochemical reactions, such as dismutations and coupled oxido-reductions, as described in chapter 9, Volume I. The second alternative — for which certain arguments were adduced in Volume I (page 166)— is that the compound ACO2 (and perhaps A'H20 as well, although the two assumptions are separable) is kinetically inde- pendent of chlorophyll; its reduction is then a secondary process, a dark reaction brought about by the products of the primary photochemical re- action. The analysis is simpler if the first alternative is chosen. If we assume that the acceptor. A, is part of the chlorophyll complex, and that it takes up or loses carbon dioxide without separating itself from this complex (and that consequently, Ao = Chlo, and [ACO2] < Chlo), then all quanta absorbed by the chlorophyll molecules carrying ACO2 must be effective (as far as the primary photochemical process is concerned) — while all ciuanta absorbed by chlorophyll carrying "bare" A are lost. The rate of reduction of ACO2 is then k*I[AC02], as required by equation (28.14); ^■*/[Chl] being the rate of absorption of quanta by chlorophyll in light of intensity I (assuming that the absorbing capacity is not affected by asso- ciation of chlorophyll with either ACO2 or A). This equation already was used in chapter 27, section 7d (c/. equations 27.58-27.66). If ACO2 is kinetically independent of the photosensitive complex, the concentration [ACO2] cannot be limited to Chlo; equation (28.14) now appears to indicate that the quantum yield of photosynthesis, 7 = P/h (/a = absorbed light energy), can increase indefinitely with increasing [ACO2]. At least, this would be so if one would assume, as usual, that I^ is proportional to 7, I a, = al, so that : (28.16) 7 = P/h = ank* [ACO2] Of course, a certain limit to the increase of 7 is set by the fact that ACO2 must be 10^ or 10^ sec.~^ (the inverse of the life-time of electronic excitation states in strongly light-absorbing molecules) to perhaps as little as 10^ sec.~^; but formally, relation (28.17) remains valid as long as the back reaction follows the monomolecular law. The possibility that the "activated" complex can be deactivated before 1024 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 encountering a molecule with which it has to react (be it ACO2 or A'H20) creates a new source of dependence of the rate on the concentration [ACO2] (or [A'H20]). We now consider the second anticipated phenomenon — the breakdown, in strong light, of proportionality between irradiation and absorption. This complication does not occur significantly in "ordinary" photochemis- try, where the rate constant of the process by which the excited molecule returns to the normal state is at least a million times higher than the rate constant of light absorption, even in the strongest available Ught (order of magnitude of the maximum frequency of absorptions: 10 sec.~\ of. page 838; order of magnitude of the rate constant of deactivation, by fluores- cence or energy dissipation : > 10^ sec. "0 • In strongest available light, the photostationary concentration of activated molecules ceases to be negligible compared to that of the normal molecules if the life-time of the activated form exceeds 10"- sec. (light absorption: once every 0.1 sec; lifetime of the activated state: 0.01 sec; therefore, stationary concentration of activated molecules: 10%). If we assume such a longevity for the ac- tivated state of the chlorophyll complex, a sizable proportion of this com- plex will be present in the changed state during strong photosynthesis, and this will lead to a lack of proportionality between the "useful" light absorp- tion, Ta, and the incident light intensity, I ("useful" meaning the absorp- tion of light by chlorophyll complexes in the unchanged form — assuming that, if the changed form does absorb visible light at all, this absorption is photochemically "useless"). A new reason is thus added for the de- pendence of the yield on light intensity. We will analyze these two phenomena — (1) the competition of the "de- tautomerizing" back reaction in the photosensitive complex with the photo- chemical forward reaction; and {2) the depletion of the normal form of this complex during intense photosynthesis — by using two simple mechanisms in which the photochemical forward reaction is assumed to involve the tautomerized chlorophyll complex, HX.Chl.Z, and either the carbon dioxide acceptor compound, ACO2 (mechanism 28.20), or the hydro- gen donor, A'HR, where HR may stand for water, or for a "substitute" re- ductant (mechanism 28.21) : k*I (28.20a ami a') X-Chl-HZ , HX-Chl-Z K (28.201)) TlXChIZ + ACOo > X-Chl-Z + AHCO, (28.20e) X-Chl-Z + A'HR ^ X -Chi -HZ + A'R (28.20(1) AHCO2 + A'R > ACO2 + A'HR PROCESSES IN THE PHOTOSENSITIVE COMPLEX 1025 This mechanism is represented in scheme 28. lA. The reason for assuming the occurrence of the secondary back reaction (28.20d) will be discussed later. If we assume, as suggested on page 1019, that /vo[A'H20] ^ Av[AC02], the concentration of the oxidized form, [X.Chl.Z], can be neglected in cqa 1 XChlHZ + A (^a) (28.20a) k' (28.20a') ACO2 L HX-ChIZ (28.20 b))(/(. r 1 AHCOz X-ChIZ (£"b) {28.20c) )(*«, r ))$ X-Chl-HZ (28.20d I \ A +C02 + H2O r H20(orHR) + A A'HaO (or A'HR) ] AHO(orAR) (fc) (fo) A + {CHgO} A' + 02 (orA+R) Scheme 28.IA. Photosynthesis according to equations (28.20a-d). green plants (though perhaps not in purple bacteria), compared to that of the tautomeric form, [HX . Chi . Z ] ; the latter is then the only one the ac- cumulation of which may affect the photosynthesis of green plants in strong light. It may be argued that, if the reaction of Z with water is so much more rapid than that of HX with carbon dioxide, the sequence of the secondary reactions (28.20b and c) should be reversed, resulting in the following mechanism (shown in scheme 28. IB): (28.21a and a') X- Chi -HZ k*I ± HXChlZ k' 1026 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 ko -^ HX-ChlHZ + A'R -> X-Chl-HZ + AHCO2 (28.21b) HX-Chl-Z + A'HR (28.21c) HX-Chl-HZ + AGO. (28.21d) AHCO2 + A'R In this case, the main form in which the chlorophyll complex accumu- lates during intense photosynthesis is the reduced form, HX-Chl-HZ. k' -> ACO2 + A'HR CO, X-CHIHZ + A (^a) (28.21a) k' (28.21a') AGO, HXChIZ HgO (or HR) + A' A'H20(orA'HR) L J r HXChlHZ r)((28.2lb) A'HO(or A'R) r AHCO2 fr/)((28.2lc) 1 X- Chi- HZ (£-9) r )Q (28.2ld) (Ec) A -I- COa + HzO (£■0) A ■^ (CHgO}^ A'H-Og (orA'-hR) Scheme 28.IB. Photosynthesis according to equations (28.21a-d). We chose above equation (27.6) as starting point of our considerations (and not a similar equation for the reaction of the light-activated complex with the oxidant A'H20), and are now interested, first of all, in the possible interdependence of A* and [ACO2], caused by the primary back reaction. Therefore, mechanism (28.20) is more convenient for our purpose than mechanism (28.21), (because in 28.20 the primary back reaction, 28.20a', competes directly with the reduction of [ACO2]). PROCESSES IN THE PHOTOSENSITIVE COMPLEX 1027 In discussing the consequences of the primary back reaction, on the basis of mechanism (28.20), we will neglect the second anticipated phenomenon — • the accumulation of tautomerized chlorophyll complexes in strong Ught. (We thus introduce a new cause for carbon dioxide saturation, but no new cause for light saturation.) Later, we will use mechanism (28.21) to con- sider the second phenomenon (accumulation of HX-Chl-HZ in strong Hght, and consequent light saturation), while in turn neglecting the pri- mary back reaction. By assuming [HX • Chi • Z ] «: [X . Chi . HZ ], conditions become formally analogous to those prevailing in "ordinary" photochemistry in vitro, as dis- cussed above. From the reaction sequence (28.20a-c) we obtain, for the photostationary concentration [HX-Chl-Z], the Stern-Volmer type equa- tion: (28.22) [HX-Chl-Z] = A;*/[X-Chl-HZ]/(A;' + kAkCO-A) ^ k*ICh.Wik' + kr[A.CO,]) and hence: (28.23) P = nkrk*I [AC02]Ch\o/(k' + ^^[ACOa]) (in expected formal analogy to equation 28.1). Comparison with equation (27.6) shows that A-*, while proportional to /, is now in fact also a function of [ACO2]: (28.24) k* = krk*I Ch\o/{k' + A-JACO2]) If [ACO2] increases indefinitely, P approaches the maximum rate: (28.25) Pmax. = nA:*7Chlo = nZ aChlo which corresponds to the maximum quantum yield, n. Similarly to the yield (28.18), n is independent of Hght intensity. Equation (28.23) shows that back reactions in the photosensitive com- plex can explain the increase in yield with increasing [CO2] and the final [CO2] saturation, even if the acceptor A is available in unlimited quantities (e. g., if [ACO2] stands for dissolved carbon dioxide, the quantity of which can be increased practically indefinitely by raising the partial pressure of carbon dioxide over the system). In other words, equation (28.23) indi- cates how a limited supply of light quanta can account for hyperbolic car- bon dioxide curves, without the assumption of a limited amount of a [CO2] acceptor, or slow diffusion, or slow carboxylation. The effects of carbon dioxide supply can, of course, be superimposed upon those of limited light supply, by introducing the corresponding ex- pressions for [ACO2] into (28.23). Using for this purpose "static" equa- tion (27.3), i. e., taking into consideration only the limited amount of the [CO2] acceptor, will not lead to light saturation; the latter will be intro- 1028 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 duced if one uses for [ACO2] one of the "kinetic" expressions, taking into account the hmited rates of supply processes (carbon dioxide diffusion and carboxylation) . Absolute saturation, Pmll'. will result if it is assumed that a maximum rate of supply exists which is independent of [CO2], e. g., as a consequence of a limited amount of the carboxylating catalyst, Ea- We will now consider the second phenomenon, which is without parallel in "ordinary" photochemistry, and arises from the assumed longevity of the "activated state" of the photosensitive complex: the accumulation of these complexes in a changed form, and the consequent lack of propor- tionality between the incident light intensity, /, and the photochemically significant light absorption, I^. For this purpose, we use the alternative mechanism (28.21), since the postulated practical instantaneousness of reaction (28.21b) gives us some right to neglect in this case the primary back reaction, (28.21a) {i. e., to assume /co[A'H20] ^ A;') The steady state concentration of the reduced form, [HX.Chl.HZ], is, under these condi- tions : (28.26) [HX-Chl-HZ] = k*I Chlo/(A-*/ + MACO,]) (implying that, in the absence of ACO2, all chlorophyll complexes would go over, in light, into the reduced form, HX-C1il-HZ). The rate of photo- synthesis is: (28.27) P = /(/,v[HX-Chl-HZ][AC02] = nAvChloA-*/[ACO.>]/(A-*/ + AvlACO,]) Assuming further that the complex ACO2 is in equilibrium with free carbon dioxide, and no diffusion gradient exists, so that [CO2] — [CO^la we obtain: (28.28) P = nkrChU*IAoKAC0.2]/(k*I + A\[COo]^"*/ + A-.AoKalCOa]) an equation for the rate of photosynthesis when neither the preparatory nor the finishing dark reactions, neither on the "reduction side" nor on the "oxidation side," have a rate-limiting influence. Light saturation is in this case due entirely to the limited amount of the acceptor, Ao; and car- bon dioxide saturation, to the limited amount of chlorophyll, Chlo. The light curves (28.28) are hyperbolae: (28.29) P/(P'°^^- - P) = A-*/(l + A'alCO.,])/Av.-lo/va[CO,I The half-saturating light intensity is: (28.30) ,/./ = AvAoA\[CO,l/(A-* + A*A'JCO,]) Extrapolating to [CO2J = <», we obtain: (28.31) y/c = AvAo/A-* If the carboxyln tion equilibrium is not maintained in light, we have, more generally, instead of (28.30) ; (28.31a) y/ = kr[AC02]/k* PROOKSSES IN TTTE PTTOTOSENSTTIVE COMPLEX 1020 Comparing equation (28.31a) with equation (28.2G), we note that, in the particular picture we are considering now, "half saturation" with light corresponds to the equal distribution of chlorophyll between the forms X-Chl-HZ and HX-ChlHZ: (28.32) [HX-Chl-HZ] = Chlo/2 = [X-Chl-HZl If we assume that reaction (28.21a) has a quantum yield of unity, the constant k* must be equal to the frequency of light absorptions by a single chlorophyll molecule. We recall that / is the intensity of light actually reaching the volume under consideration, and that light absorption within this \'olume is supposed to be so small that the rate of absorption is practi- cally the same everywhere within it. Under these conditions: (2S.r>3) k* = a In 10 X 10-^ = 2.3 X 10^ « (mole/cm. 2) where a is the average molar absorption coefficient for the incident light. This equation implies that / is measured in number of einsteins of light falling per second on a square centimeter and Chlo is expressed in moles per liter. According to page 838, for white light, 1 lux ^ 1.4 X 10' ^ quanta/ sec. cm. 2 = 2.3 X 10"'- einstein/sec. cm.^ We thus have: (28.34) k* = 5.3 X IQ-^ a (mole/cm. 2) applicable when I is expressed in lux (meter candles). Average absorption coefficient of chlorophyll for visible light is of the order of 10^; this means /^* i^ 5 X 10-^ or one photochemical act per second per chlorophyll mole- cule in Hght of 20,000 lux. Experiments indicate that i//o= is of the order of 5 X 10^ lux (cf. Table 28.1) ; assuming ^^ ~ 5 X IQ-^ and Ao ~ 5 X 10"^ m./l. (roughly the con- centration of chlorophyll in the plastids; cf. Vol. I, page 411), we obtain for kr a value of 5. When both [CO2] and / are very large, the rate ac- cording to equation (28.28) approaches the "absolute ceiling": (28.35) PS: = nivChloAo (= nA:*Chlo ,/,/co) which is the maximum possible rate of reaction (28.21c) reached when a/i chlorophyll is in the state HX-Chl-HZ, and all acceptor is occupied by carbon dioxide. For low light intensities and ample supply of carbon dioxide, the light curves defined by equation (28.28) approach asymptotically the limiting straight line : (28.35a) P'/ = o = nA-*Chb/ The simplifying assumption that the carbon dioxide acceptor is, even during intense photosynthesis, in equilil^rium with external cai'bon dioxide could be dropped, and the more general expression (27.15) for [A-C02] 1030 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 could be introduced into equation (28.27) instead of the equilibrium expres- sion (27.3). However, the resulting equations, embodying the effects of carbon dioxide supply limitations together with the limitation due to light supply, would become too unwieldy for practical use. One may ask whether equation (28.35) provides a sufficient explanation of the "absolute saturation" of photosynthesis, i. e., of the maximum yields discussed in section 5 of this chapter. We can obtain from equation (28.35) the follomng estimate of the maximum possible yield of photosyn- thesis per chlorophyll molecule per second : (28.36) PSS:/Chlo = nkrAo ( = nk* yl^) ^ 0.025 or, for "assimilation time," T^: (28.36a) Ta = Chlo/PS!S: ^ 40 sec. The experimental values of "assimilation time" (c/. Table 28.V) actually range from 10 to 100 sec. (with the exception of the remarkably smaller values found for aurea varieties). It thus appears as if chlorophyll could be the component of the photosynthetic apparatus the slow rate of restor- ation of which imposes an "absolute ceiling" on the rate of photosynthesis. However, this is not the only adequate answer to the problem of absolute saturation, since flashing light experiments reveal the existence of a "finish- ing" catalyst (Franck's "catalyst B"), which appears to be present in an amount equivalent to about one two-thousandth of that of chlorophyll, and to have a "working time" of the order of 0.02 sec. at room temperature; the "ceiling" imposed by this catalyst is of the same order of magnitude as the one derived in (28.36), since 2 X 10^ X 2 X 10 ~^ is the same as 1 X 40. It is useful to show why the approximate agreement of the value (28.36) with the experimental results is not a significant confirmation of the model used in the derivation of this equation. We will see below (page 1043) that in a rectangular hyperbola, the three parameters which in general specify a hyperbola — such as (a) the initial slope, (6) the abscissa corresponding to half-saturation and (c) the satura- tion value of the ordinate — are not independent of each other, but related by equation (28.48C). What we did above was to insert the approximate experimental value of two of these parameters (n ^^ 0.1 and i/J ^^ 10^ lux) into equation (28.28), Avhich represents a rectangular hyperbola (of. equation 28.29; the equation in parentheses in 28.36 is a special case of 28.48C) ; and to derive in this way, an approximately correct value of the third parameter, p™^'^-. This result merely shows that the light curves (or, at least, the limiting light curve at high [CO2]) do approximate rec- tangular hyperbolae. Any kinetic mechanism which leads to light curves PROCESSES IN THE PHOTOSENSITIVE COMPLEX 1031 approximating rectangular hyperbolae will permit a correct estimation of the third parameter from the experimental values of the two other para- meters. We will find later (c/. chapter 32) evidence that the principal rate- limiting reaction in photosynthesis is a dark reaction catalyzed by a cata- lyst (Eb) which is present in the cells in a concentration only 0.05% of that of chlorophyll. We will also see (page 1038) that this type of limita- tion leads to light curves which are hyperbolae, but not rectangular hyper- bolae. It remains to be seen whether light curves can be measured pre- cisely enough to exclude one of the two mechanisms. Is it possible for two or more bottlenecks to exist in photosynthesis, each allowing the passage of about the same amount of reactants — the maximum rate of passage through one bottleneck being determined by the product of the concentration Chlo, the (approximately equal) concentra- tion, Ao, the (bimolecular) constant, K, of reaction (28.21c) and the quan- tum yield n (O.OS^ x 5 X 0.1 = 1.25 X 10"^); and the other, by the product of the enzyme concentration Eb and its (monomolecular) rate constant (2.5 X 10"^ X 50 = 1.25 X 10^^). Such a coincidence seems not implausible; it could even be considered as admirable economy in the allotment of catalysts to the cell. (Why have more of a certain catalyst than can be utilized because there is not enough of another one?) However, certain experimental results are not consistent with the as- sumption that restoration of chlorophyll is the bottleneck which limits (or "co-limits") the maximum rate of photosynthesis; these data indicate that a chlorophyll molecule which has taken part in the primary photo- chemical process needs much less than 4Qn (—4) sec. to return to the photo- sensitive form. We mean here the observations of Willstatter and Stoll (c/. Table 28. V; see also chapter 32, fig. 32.2), that aurea leaves have P'^^''- values only slightly lower than those of ordinary green leaves, although they contain only one third (or less) of the normal amount of chlorophyll. This is obviously inconsistent with equation (28.35), and indi- cates that P"^^"^- is determined not by the rate of restoration of the photo- chemically tautomerized chlorophyll complex (rate constant kr in equation 28.35), but by the rate of transformation of a substrate by a catalyst (such as Eb) that is kinetically independent of chlorophyll. The above-esti- mated value of kr (— 5 (sec. mole)~i) is therefore merely a lower limit; in fact, quantitative observations which aurea leaves indicate that the true value of this constant (which determines how often a given chlorophyll mole('uIe is available for the primary photochemical reaction) is at least ten times higher. This means h >50 sec.-^ (for [ACO2] = 0.05 mole/ liter), assuming that the primary photochemical reaction is (28.21c). Because of fundamental significance of these conclusions, a reinvestigation of the kinetics of photosynthesis in aurea leaves seems desirable. 1032 THE LIGHT FACTOR. I. INTENSITY CHAP, 28 It was noted that, according to equation (28.32), half saturation of photosynthesis with light should take place, in the presence of excess car- bon dioxide, when chlorophyll is distributed equally between the forms X- Chi -HZ and HX-Chl-HZ. This state should also correspond to the halfway point in the transition of fluorescence from the "low light" yield, > (f) Analytical Formulation: Effect of "Finishing^^ Dark Reactions The necessity of considering, in addition to the "preparatory" dark re- actions and the reversible changes in the chlorophyll complex, the "finish- ing" dark reactions as possible sources of light saturation phenomena in photosynthesis arises from several observations. It was mentioned above that experiments in flashing light (to be described in chapter 34) demon- strated directly the existence of a "finishing" catalyst with a "working period" of the order of lO^''^ sec. at room temperature. The maximum yield obtainable per single flash shows that this catalyst is present in a con- centration equivalent to 0.05% of that of chlorophyll. Consequently, as indicated above (page 1031 it imposes a "ceiling" on the over-all rate of photosynthesis of about one molecule carbon dioxide reduced per chloro- phyll molecule every 40 seconds — which is close to the actually observed maximum rate of photosynthesis at room temperature. A second relevant observation is made by comparing the light curves of photosynthesis of various plants with the light curves of their fluorescence (cf. part B of this chapter). If the light saturation of photosynthesis were due to a slow supply reaction (i. e., to the depletion of one of the reactants, ACO2 or A'Il20, in intense light), or to slow regeneration of the photosensitive form of the chlorophyll complex, in both cases, the light saturation would be associated with accumulation of the photosensitive complex in a chemically changed form (such as HX-Chl-Z or HX-Chl-HZ), and should therefore reveal itself by simultaneous changes in the fluorescence yield of the com- plex. This is actually the case sometimes, but not always. Figure 28.24, for example, shows light saturation of photosjoithesis of Chlorella without any change in the yield of fluorescence; and even in figure 28.26, satura- tion is almost completed before fluorescence begins to change its yield. In all such cases, saturation must be due to the failure of a finishing dark reaction to keep pace with the primary photochemical process — a failure that produces no change in the composition of the photosensitive complex, but leads to the loss of a large part of primary photoproducts by back reac- tions. It was stated before that the distinction between preparatory and finishing dark reactions is not so clear-cut in Franck's theory of "narcotic regulation" of photosynthesis. According to this theory, accumulation of primary oxidation products ("photoperoxides") leads not (or not only) to back reactions between these peroxides and the primary reduction products 1034 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 (as was assumed above), but also to the oxidation of certain metabolites (sugars?) which produces a "narcotic" (a fatty acid?). The latter settles on the chlorophyll complex, prevents any further acceleration of the pri- mary photochemical reaction, and causes a strong increase in fluorescence. In the case of purple bacteria the following two alternative descriptions of the rate-limiting and fluorescence-enhancing effect of a limited supply of the reductants (H2, H2S2O3 . . . ) are possible : Either one considers this supply as a preparatory reaction, whose slowness causes the photosensitive complex to accumulate in a changed (more strongly fluorescent) form, such as X-Chl-Z; alternatively, (using Franck's theory) one can treat this supply as part of a finishing reaction ("disposal of photoperoxides"), and explain its effect on fluorescence as a consequence of the production of the "narcotic" by accumulated peroxides. In both cases, the light curve of photosynthesis will approach, with increasing light intensity, a limit de- termined by the maximum rate of supply of the reductants (either by dif- fusion or by a preliminary enzymic transformation). Another argument against general attribution of the assimilation num- bers of green plants to the rate-limiting influence of preparatory reactions — specifically, those on the "carbon dioxide side" — was mentioned in Chap- ter 12 (Vol. I) in connection with the influence of cyanide on photosynthesis. It was stated there that the difference in the amounts of cyanide required to reduce by a certain factor the rate of photosynthesis in different plant species is most easily understood if one assumes that the cyanide-sensitive catalyst (which, in all probability, is the "carboxylase" Ea) is not rate- limiting in strong light in the absence of the poison, and becomes limiting only when a considerable part of it is inactivated. If a different ratio Ea/Eb prevails in various species and strains, the fraction of Ea that has to be inactivated in order to make this catalyst "limiting" also must change from case to case. What kind of back reactions can compete with finishing catalytic re- actions in photosynthesis? In section d of this chapter, we considered the "primary" back reaction, HX-Chl-Z -^ X- Chi -HZ, which can be called "detautomerization" of the chlorophyll complex. In equations (28.20) and (28.21), this reaction competes with the secondary photochemical forward reaction, e. g., (28.20b) or (28.21b). We noted on page 1024 that this competition can cause carbon dioxide saturation, but not light saturation, because the proportion of quanta lost by this kind of back reaction is inde- pendent of light intensity. If one would treat (28.20b) or (28.21b) as a catalytic reaction, assigning to it a catalyst with the concentration and working speed attributed above to "catalyst B," this would produce light saturation, but the latter will again be associated with the accumulation of the form X • Chi • Z or HX • Chi • HZ and thus with a change in the intensity EFFECT OF FINISHING DARK REACTIONS 1035 of fluorescence. To explain saturation not accompanied by changes in fluorescence intensity, the deficient catalyst must be placed, not between one of the two reactants (ACO2 or A'HsO) and the photosansitive complex, but between the primary and the finished products of photosynthesis. In other words, the back reactions caused by the catalytic deficiency must be "secondary" rather than "primary." A back reaction of this kind was therefore added as reaction d in mechanisms (28.20) and (28.21). We can postulate, e. g., that reaction (28.20d) comes into play because the transfor- mation of the first photoproduct, AHCO2, into a more stable intermediate requires a catalyst, Eb (perhaps a "mutase"), which is present in limited quantity. A similar postulate could be made for the effect of the catalyst Ec, which is required for the first step in the conversion of A'HO into free oxygen. Because of the symmetry we have assumed between the right and left sides in schemes 28.IA and B, a limitation in the utilization of the oxida- tion products will have the same effect on the kinetics of the process as a whole as a limitation of the utilization of the reduction products. In the first case, the secondary back reaction will be accelerated by the accumula- tion of the primary oxidation product, A'HO; in the second case, by the ac- cumulation of the primary reduction product, AHCO2. Using reactions (28.20) we can tentatively assume that Eb acts on the first reduction product, AHCO2, according to the scheme; (28.38a) AHCO2 + Eb ^ ^ ^ EbAHCO. (28.38b) EBAHCO2 > Eb + A + {HCO,} where { HCO2 } designates a stabilized reduction product. (If reaction 28.38b is a dismutation, it may require the participation of two AHCO2 radicals, but we ^nll use here the simplest possible mechanism.) The rate of photosynthesis is, according to this scheme : (28.39) P = nfcelEBAHCOo] and the "absolute maximum rate" is: (28.40) PZ^: = nkeEl where Eg is the total available quantity of the "limiting" enzyme. An equation for P as a function of / and [CO2] can be derived from this mechanism; but it is complicated, even if all the possible simplifying assumptions are made, and little could be achieved by writing it do^^^l here. The situation is simplified if we again make use of Franck and Herzf eld's mechanism (scheme 7.VA, Vol. I), in which the oxidant, ACO2, and the re- ductant, A'H20, belong to the photosensitive complex, and take part in the primary photochemical process, e. g., in the way indicated in equation 1036 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 (28.15). In this case, the primary back reaction (detautomerization) itself becomes a competitor to finisliing dark reactions, and it can therefore be assumed that the role of the catalyst Eb is to prevent this reaction from de- stroying the photoproducts. This point of view was used in the elaboration of kinetic equations of photosynthesis by Franck and Herzfeld (1937). Their derivations are comphcated by the assumption of four successive (different) photochemical steps "on the reduction side," alternating with four (identical) photochemical steps "on the oxidation side"; the product of each of these steps was supposed to require stabilization by the same "catalyst B," in order to prevent this step from being reversed by a dark reaction. Instead of trying to present here the derivations of Franck and Herzfeld we will use a simpler mechanism embodying the same basic con- cept of primary back reaction as cause of Ught saturation. This mechanism is similar to the one given in equation (7.13) in Volume I, but is further simplified by the substitution of a single primary photochemical step for the two steps (7.13a) and (7.13b). The reaction scheme is: k*I (28.41a, a') ACOs-Chl-A'H.O ^=± AHCOa-Chl-A'HO Primary forward and k' back reaction k (28.41b) AHCOa-Chl-A'HO + Eb ^— > EbHCO. + A-Chl-A'HO Catalytic stabiliza- k' \ tion of re- (28.41c) EbHCOo '- — >Eb + {HC021 duction product (28.41d) A-Chl-A'HO + H2O > A-Clil-A'HjO + (HO) V'Reloading" of (28.41e) A-Chl-A'HaO + CO2 > ACO.-Chl-A'HoO Jchlorophyll We assume that Eb is needed only for "stabilization" of the reduction product, AHCO2, and that this stabilization is achieved, in reaction (28.41b), by taking the reduced group, HCO2, away from the chlorophyll complex and thus preventing its back reaction with the oxidized group, HO. (It is not suggested that the radicals HCO2 or OH occur in the free state; these symbols can stand for corresponding functional groups in larger molecules, as indicated by braces in 28.41c and d.) In order to simplify the derivations still more and to elaborate only the effect due to the back reaction, we further assume that the reactions (28.41d and e), by which the photosensitive complex is supplied with fresh oxi- dant and fresh reductant, respectively, are practically instantaneous. Under these conditions, only two factors can cause light saturation: (a) accumulation of the photosensitive complex in the tautomeric form AHCOg-Chl- A'HO; and (5) accumulation of the catalyst Eg in the bound EFFECT OF FINISHING DARK REACTIONS 1037 form, EbHC02. The equation for P which takes into account these two effects is quadratic; its one significant sohition is: (28 42) P = I'" ^^<=^*^Chlo + A- 'A-.' 4- Kk*I + Kk^El /n\ 2A-e j-^A.A-/Chlo + A-'A,^+A-;A-/ + A-;A-.E;^y _ ,,j,o,*,chlo]'^^^ \ where n, as before, means the number of elementary photochemical steps (28.41a) required for the reduction of one molecule of carbon dioxide. AC02-Chl-A'H20 (28.41a) kl k' (28.41a') AHCOa-Chl-A' HO (28.41b) +£b K r EbHCOz _yv_ A-Chl-A'HO k. kzoo + HjO + COj Jk. Eb+(hCG2} f (28.41 d^^) ^ ACGj-Chl-A'HjO (hoJ jCHzO} I I 02 Scheme 28.11. PhotoyyuUiesis according to mechanism (28. -11). Equation (28.42) describes a hyperbolic light curve with an initial slope : (28 43) {dP/dI)o = nA:eE"BA:*Ch]o/(A;' + /ceE") If A;' C A-eEs; we thus have approximately : (28.47D) ./J - ^'°2iiichl^° (A-; JHCO2! + tOH} + X-Chl-HZ K (28-47Fb) {OHl + E > EOH A-e' (28.47Fc) EOH > E(+ O2) K (28-47Fd) I OH} + X-Chl-HZ (+ H,M) > X-Chl-HZ-HM -|- H2O (= {Chi} + H2O) K (28.47Fe) X-Chl-HZ-HM (+ 10.) > X-Chl-HZ (+ M + IHaO) where {Chi} is an abbreviated expression for the "narcotized chlorophyll," XChl-HZ-HM. The components in parentheses are supposed, for the sake of simplifica- tion, to be present in excess and to react practically instantaneously, so that their concentrations do not appear in kinetic equations. The summa- tion of the first three equations gives equation (28.47Fa) for the primary photochemical process. It is further assumed that the reduction product. 1042 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 {HCO2}, is removed practically instantaneously, but that the oxidation product, {OH}, has to undergo an enzymatic transformation (28.47Fb, c), which includes a "bottleneck" reaction imposing an absolute ceiling (28.47H) on the over-all rate of photosynthesis. If reaction (28.47Fa) runs too fast for the reactions (28.47Fb,c) to re- move the oxidation products immediately, the "narcotization" mechanism (28.47Fd, e) comes into play; accumulated {OH} reacts with ametaboHte, H2M (supposed to be present in excess in the cells), and oxidizes it to a "narcotic," HM, which is adsorbed on chlorophyll and inactivates it (supposedly without affecting the absorption spectrum, but increasing the yield of fluorescence). The narcotization is removed by reaction (28.47Fe), which completes the oxidation of the "half-oxidized" product. (At a con- stant concentration of oxygen, this reaction can be treated as monomolecu- lar.) The rate of photosynthesis is, in this scheme, equal to the rate of the limiting reaction (28.47Fc) : (28.47G) P = »^e[EOH] the maximum possible rate is (the subscript max. being justified by the above-made assumption of practically instantaneous supply of carbon dioxide) (28.47H) PZl: = nkeEo The factor n designates, as before, the number of elementary photo- chemical steps of type (28.47Fa) required for the production of one molecule of oxygen. We introduce the abbreviation: (28.47Ha) k = ^ An equation for the rate of photosynthesis can be obtained by deter- mining the photostationary concentration [EOH] and then applying equation (28.47G). The result is: (28.471) - - 2(fc - 1) Lv 2(fc - 1) ) k-i \ (One can easily ascertain that this equation gives correctly, P = 0 for / = 0, and P'"^'^- = nA-eEo for 7 = oo .) The light curves (28.471) are nonrectangular hyperbolae with the initial slope: nkji* nkjcnk* (28.47J) \dl Ji = o kekEo + k:Ch\o kUko + knCUo) SELF-NARCOTIZATION 1043 and the half-saturating intensity: k,E,{k + 1) + 2A-,:Clilo (28.48) V/ = {g) Are Light Curves Hyperholicf 2k*kCh]Q It will be noted that all the kinetic models used in this chapter lead to quadratic equations for the rate of photosynthesis, P, as function of incident light intensity, /; in other words, to hyperbolic light curves, P = /(/). The imposition of a rate "ceiling," caused by limited supply of a reactant or limited amount of an enzyme (in addition to a rate "roof," imposed by the minimum number of quanta required to bring about the reaction) changes the appearance of the light curves — for example, it can convert a rectangular into a nonrectangular hyperbola— but in the mechanisms discussed so far, general hyperbolical shape is preserved. A hyperbola is defined by three parameters. In our presentation of light curves, the axis of abscissae, I, was parallel to one asymptote of the hyperbola (the equation of the latter being P = p"''^^-)^ while the axis of ordinates, P, was chosen so as to make P = 0 at / = 0. One convenient form of writing the equation of a hyperbola in this system of co-ordinates is: (28.48A) '"^ pmax. _ p \ ' pmax. / The three parameters in this equation are 70, the initial slope of the curve, which is proportional to the maximum ciuantum yield of photo- synthesis; i/J, the half-saturating light intensity; and P'"'^''-^ the limiting rate in strong light. If the second term in parentheses vanishes, the equation becomes: (28.48B) p^£:ip = p~J = coiist- X / which is the equation of a rectangular hyperbola. For the quadratic term in (28. 48 A) to vanish, the following relation between the three parameters must be fulfilled : E>max, (28.48C) 70 W Several equations derived in chapters 27 and 28 for the light curves of photosynthesis- — such as (28.29) — actually had the simplified form (28.48B) ; however that was not always the case, as shown, e. g., by equa- tion (28.42), which cannot be represented in the form (28.48B). Since 70 is not so easily measured as the other two parameters (in fact, determination of 70 may be one aim of analytical representation of the light 1044 THE LIGHT FACTOR. I. INTENSITY CHAr. 2S curves, cf. chapter 29, p. 1132 ff.), another parameter may be chosen instead, such as ■//, the Hght intensity at which photosynthesis reaches one-quarter of its saturation vahie. This choice leads to a very simple relation : pmax. (28.48D) 70 and gives, as equation of the light curve: p r 4P "I (28.48E) / = p^,,. _ p |_6./, I - ,// + p^. (,// - 3,//)J For the hyperbola (28.48E) to be i-ectangular, it must satisfy the very simple condition: (28.48F) ,/./ = 3,// Its equation is then: (28.48G) / = p^i^^^zrp (6 v/ - v/)' "^ ]\Iore generally, a hyperbola could be represented in our chosen co- ordinates, using any three of its points (Pi,/i; PiJ-i; and P3,h) as param- eters; for a rectangular hyperbola, two points suffice. These derivations should be kept in mind in evaluating papers, in which failure to represent empirical data by equations of type (28.48B) has been taken to mean that the light curves were ''not hyperbolic" {cf Smith 1936, 1937, 1938). It was stated on page 1020 that all our derivations of light curve equa- tions were based on the assumption of uniform light absorption, and are therefore strictly applicable only to optically thin layers. The question arises to what extent the considerable optical density of most actually studied plant objects distorts the shape of light curves — for example, whether the observed "integral" light curves will be nonhyper- bolic if the "differential" light curves for each thin layer, with practically uniform light absorption, are hyperbolae. For monochromatic light, for which the absorption coefficient is a, the total light flux absorbed in a layer I (for the sake of simplicity, we assume uniform pigment distribution and absence of scattering) is : (28.481) /a = /(I - lO-«tci>'io (assuming chlorophyll to be the absorbing pigment). If the quantum yield of the over-all process is n, and if /a is expressed in einsteins per cm.^ per sec, the total rate of photosynthesis is: ARE THE LIGHT CURVES HYPERBOLIC? 1045 (28.48J) P = n/a = n/(l - 10-«icw]i) (mole CO2 reduced per second per cm.^ of illuminated area). If the illuminating light is nonmonochromatic, the first part of this rela- tion remains valid, but 7a is now expressed by an integral: (28.48K) /a = / - r Xi '\^'hen the incident light intensity increases so that light saturation sets in, in the most exposed pigment laj^er, the light curve of the integral yield bends, and does not become horizontal until saturation has become complete in the deepest laj^er. Consequently (as discussed before, cf. p. 1007 and fig. 28.20), the qualitative effect of inhomogeneous light absorp- tion must be to broaden the transitional region connecting the linearly ascending and the horizontal part of the light curves. AVhat we want to know is how the shape of the light curves is changed by this integration. Let us assume for the sake of simplicity that the "differential" light cuives are rectangular hyperbolae: p. = k*Ii pmax. _ p^ jL*r)inax. riQ-Q:(Chl]( (28.48L) Pt = I _|_ ^*j jQ-a[chi]i The integrated rate (per unit area) is then : (28.48M) P = J^ ^'^^ = U [Chi] In 10^" U[Chll In 10 1 + ^•*n0^lChlJ^ or: m dSV'^ — ^ = In [(1 + A-*/)/(l + fc*/10-'^[ch']'.) ^zs.is.n; p^^^ _ p ^^^[(^yj jj^ jQ _ jj^ j^ ^ A;*7)/(l + fc*/10-«Ichi](o) ] an ctiuation which does not represent a rectangular hyperbola. It can he developed into a series: P (28.480) pmax. p ]l 2! "^ 3!2! (1 + k*I) ^^^ (^ _ «[Chl]/oln 10 («[Chl]lo In 10) = (2 + k*I) («[Chl]Zoln 10)3 (3 + k*I){2 + k*I) 4!3! (1 + k*IY "*" ..,) This series shows that the integral light curve remains practically a rectangular hyperbola until the third term in the development ceases to be small compared to 1; then, it looses the hyperbolic shape (because of a third degree term, containing the product PP). Since the second factor in the third term decreases from 2 to 1 with increasing light intensity, the 1046 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 maximum value of this term is a[Chll/o In 10/6. This means that for the maximum deviation of the integral light curve from hyperbolical shape to exceed 5%, the integral absorption must exceed 50%: (7o// = 2; log/o// = a[Chl]ioln 10 = 0.30; a[Chl]/oln 10/6 = 0.05) At 75% absorption, the maximum deviation will reach 10%. This indi- cates that to obtain experimental light curves which will permit one to judge the true shape of the function P = f{I), objects with an optical den- sity of up to log h/I = 0.5 can be used. A corresponding derivation for a general hyperbola is much more in- volved, but conclusions are likely to be not too different from those obtained here for a rectangular hyperbola. In the literature one can find many attempts to derive analytical ex- pressions for the light curves of photosynthesis, partly by using very simple kinetic models, and partly by empirical approximation. Among the simplified kinetic equations, reference may be made to those of Ghosh (1928), Emerson and Green (1934), Baly (1935) and Burk and Lineweaver (1935); while among the empirical approximations, we may mention those of Brackett (1935) and Smith (1936, 1937, 1938). -4 O O o o -5 •P CD o 2 0 16 -8— 12 3 4 5 LOG I, meter candles Fig. 28.23. Are photosynthesis curves hyperbolae? (after Smith 1938). SoHd Hnes are derived from nonhyperbolic functions (27.77) and (28.48P); dashed line from rectangular hyperbolic function. See text page 1047. Brackett sought to expres.s the sudden approach of the light curves to saturation by an exponential curve, while Smith suggested an algebraic function of a higher order: LIGHT CURVES OF FLUORESCENCE 1047 (28.48P) p/pmax. = C//V1 + C2/2 or: (28.48Q) p/V(pmax.)2 _ p2 = CI To compare the usefulness of functions (27.77) or (28.48P) with that of the rectangular hyperbola, Smith derived, for both tj-pes of functions, equations determining the combinations of the parameters [CO2] and 7 for which the yield P has a certain constant value. Figure 28.23 shows the results: For the lowest value of P used (log P = 0.8), the dashed curve derived from the hyperbolic function gave the better fit; but for three higher values of P (log P = 1.2, 1.6, and 2.0), a very good fit was obtained by means of equations (27.77) and (28.48P). B. Light Curves of Fluorescence* 1. Relation between Light Curves of Photosynthesis and Fluorescence In vitro the intensity of fluorescence usually is proportional to the intensity of illumination. This is so because both fluorescence and the "quenching" processes that compete with it (i. e., energy dissipation, and photochemical reactions) usually are "first order" or "monomolecular" processes with respect to the concentration of excited pigment molecules. In other words, each excited molecule has a certain probability of fluores- cing, a certain probability of being deactivated by energy dissipation and a certain probability of undergoing a photochemical reaction; all these probabilities are independent of the concentration of the excited molecules, and consequently do not depend on the rate of their production and thus also on light intensity. Under these conditions, the rate of photochemical reactions too must be proportional to light intensity. Thus, both the quan- tum yield of fluorescence,

Tithesis (i. e., on fluorescence of purple bacteria in presence of reductants (after Wassink et al. 1942). pH 6.3, 29° C, The characteristic initial bend upward of the fluorescence curves of Chromatium does not quite disappear even in complete absence of reduc- tants. This can be attributed to the presence of "internal reductants," which in weak light suffice to prevent a complete conversion of the "photo- complex" into the strongly fluorescent form. The linearity can in fact be improved by preliminary starvation of the bacteria, depriving them of metabolites that could serve as internal reductants. Figure 28.35 shows that the effect of the reductants on the fluorescence of Chromatium persists even in the absence of carbon dioxide. This fact must be compared with the above-mentioned observation (fig. 28.29) that the removal of carbon dioxide had no effect on fluorescence in the absence of a reductant. We will return to the discussion of this interesting differ- ence on page 1077. 1054 THE LIGHT FACTOR. INTENSITY CHAP, 28 JNCIDENT INTENSITY, erg/cm. sec. Fig. 28.31. Fluorescence of pur- ple bacteria as function of the con- centration of reductants (after Wassink et al. 1942). 5% CO2, pH6.3, 29°C. Fig. 28.32. Influence of con- centration of thiosulfate on in- tensity of fluorescence (Wassink et al. 1942). 5% CO2, pH 6.3, 29° C. UJ o o K L 50% depression in 0.05% KCN). In the presence of reductants, fluorescence curves with and without cyanide (figs. 28.45) showed small but distinct differences, again somewhat similar to those found with varying carbon dioxide supply (cf. fig. 28.30). (e) Hydroxylamine and Azide Observations of the effect of these two poisons on fluorescence were made by Wassink and co-workers (1942) in purple bacteria. They are illustrated by figs. 28.46-28.47 (p. 1064). The influence of hydroxylamine appears similar to that of potassium cyanide — no effect up to 0.05% (a concentration at which photosynthesis is about 50% inhibited), then (at LIGHT CURVES OF FLUORESCENCE 1063 0.1%) a strong stimulation of fluorescence. The effect of sodium azide differs from that of either potassium cyanide or hydroxylamine in two respects. Fluorescence is affected (c/. fig. 28.47) even in the absence of reductants; and the typical effect is a decline, rather than a rise of the yield of fluorescence. (/) Ion Concentration In Chromatium, the yield of fluorescence can be affected by changes in pH. According to Wassink and co-workers (1942) the sign of this effect depends on whether molecular hydrogen or thiosulfate is used as hydrogen donor. (We saw on page 952 that the same is true of the influence of pH on the yield of carbon dioxide reduction by these two reductants.) This is illustrated by figure 28.48, page 1065. No experiments are available on the effect of other cations or anions on the yield of fluorescence of green plants or purple bacteria. {g) Narcotics Narcotics were found by Kautsky and Hirsch (1935) to increase the steady fluorescence of aquatic plants; this result was confirmed by Was- sink, Vermeulen, Reman and Katz (1938), cf. fig. 28.49. According to Franck, French and Puck (1941), very high concentrations of carbon di- oxide (e. g., 20%) produce a similar effect. (It was stated in chapter 13 that the effect of excessive concentrations of carbon dioxide on photosynthe- sis resembles narcotization.) The phenomenon was also studied in purple bacteria, by Wassink, Katz and Dorrestein (1942). Figure 28.50 shows typical results obtained in the presence and in the absence of reductants. The picture is similar to that with sodium azide: In the absence of reductants, the addition of increas- ing amounts of ethylurethan causes a progressive quenching (rather than stimulation) of fluorescence (although the effect is reversed at > 1.5% ure- than, at least below 15 kerg/cm.^ sec). In the presence of reductants, moderate quantities of urethan have little if any effect on the fluorescence curve, while quantities > 2% have a strong enhancing effect. (h) Oxygen Kautsky, whose theory of an exclusive transfer of excitation energy to oxygen was described in Volume I, (page 514), was naturally interested in the quenching of the fluorescence of leaves by oxygen. However, neither Kautsky, Hirsch and Davidshofer (1932), Kautsky and Hirsch (1935) nor Wassink, Vermeulen, Katz and Reman (1938) could find any distinct in- fluence of changes in the external oxygen concentration (between 1 and 1064 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 o UJ o CD UJ cr o 3 10 % Hydroxylamine o none A 0 005 D 005 V 0 5 O 1.0 INCIDENT INTENSITY, erg/cm sec. Fig. 28.46. Effect of NH4OHHCI on fluorescence of Chroniatium (after Wassink, Katz and Dorrestein 1942). 5% CO2, 15% H2, pH 7.6, 29 ° C. 14 No Reductont % sodium ozide 12 0 none A 0.0167 D 0.025 ^ 10 V 0167 0 0.5 ■3 - 8 CD ° UJ q: o 3 INCIDENT INTENSITY, erg/cm^ sec. Fig. 28.47. Effect of NaNj on fluorescence of Chromalium (after Wa.s.sink, Katz and Dorrestein 1942). 5% CO2, pU 6.3, 20° C. LIGHT CURVES OF FLUORESCENCE 1065 INCIDENT INTENSITY, erg/cm^ sec Fig. 28.48. Effect of />H on fluorescence of Chrornatium (after Wassink, Katz and Dorrestein 1942). 5% CO2, 29° C. 120 110 o not inhibited A 005 ml 60% urethon/ml 6. I V 1.6 X 10"* INTENSITY, erg/cm^ sec Fig. 28.49. Influence of ethylurethan on fluorescence of chlorophyll in suspension of Chlorella (after Wassink, Ver- meulen, Reman and Katz 1938). 1066 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 o in LlI o No Reductant INCIDENT INTENSITY, erg/cm sec. B Fig. 28.50. Effect of ethylurethau on fluorescence of Chromatium in the ab- sence and pressence of reductants (after Wassink, Katz and Dorrestein 1942). 5%C0-,, pH7.6, 29° C. 100 CO < O O U. o I- < a: Photosynthesis Fluorescence 10 20 30 40 X 10* LIGHT INTENSITY, erg/cm.^ sec. Fig. 28.51. CO2 assimilation and fluorescence vs. incident light intensity (after McAlister and Myers 1940). Solid symbols, intensity of fluorescence, F; open symbols, rate of CO2 uptake, P. INTERPRETATION OF LIGHT CURVES OF FLUORESCENCE 1067 100%) on the yield of steady fluorescence of leaves and algae. McAlister and Mj^ers (1940) found that an increase in oxygen concentration from 0.5 to 20% resulted in a marked increase in fluorescence (c/. fig. 28.51) — an effect opposite to quenching, and probably associated with the inhibiting influence of oxygen on photosynthesis, described in Volume I (chapter 13). Shiau and Franck (1947) noted that, at low light intensities, fluorescence of green algae was stronger in nitrogen than in air, but that the increase of (p with increasing light intensity began earlier in air. In some cases the two curves even crossed each other, so that in strong light the aerated sus- pension fluoresced stronger than the nonaerated one. We have spoken in this chapter only of changes in chlorophyll fluores- cence caused by internal chemical transformations associated with photo- synthesis— a relation that reveals itself indirectly, by comparison of the influences of light intensity, temperature and poisons on the yields of photosynthesis and fluorescence. In vitro, quenching of chlorophyll fluores- cence can be produced directly, by the addition of certain substances under- going autoxidation, as well as of many oxidants, including free oxygen (chapter 23, section A6). No observations have been made on the quenching of chlorophyll fluorescence m vivo by amines (or other possible substrates of sensitized photoxidation), while the effect of oxygen was described above as complex and probably mostly indirect. Shiau and Franck (1947) found that quinone depresses fluorescence in Chlorella, if added in the dark or in light after long anaerobic incubation. 3. Interpretation of Light Curves of Fluorescence In absence of positive information to the contrary, we have assumed, throughout the preceding sections, that all the observed changes of fluores- cence were increases and decreases in the fluorescence yield, tp, without significant shifts in the position of the fluorescence bands. This point could, however, profit by exact investigation. Changes in the structure of the chlorophyll complex (e. g., conversion of X- Chi -HZ to HX-Chl-Z, not to speak of reversible hydrogenation, Chi ^ rChl) could well find ex- pression in the variation of the position and shape of the fluorescence bands ; and the selective spectral sensitivity of the photometric devices used could convert these changes into apparent variations in the intensity of fluores- cence. While it is extremely unlikely that such spectral effects were re- sponsible for all or even a large fraction of the described intensity changes, it might be unwise to ignore their possibility. True changes in the intensity of fluorescence can be caused by two fac- tors: alterations in the relative probabilities of fluorescence and energy dissipation in the light-absorbing complex, and changes in the probability 1068 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 of the primary photochemical process. As described previously (c/. Vol. I, page 546, chapter 23. A8, and chapter 24, the three processes— energy dissipation (rate constant, h; quantum yield, 5), chemical transformation (rate constant, kt] quantum yield, 7) and fluorescence (rate constant, kf-, quantum yield, A-Chl-A'H.O + {OH} + {HCO2! + H2O (28.51Cb) A • Chi • A'HoO + CO2 ^ ACO2 • Chi • A'HaO or: k*I (28.51Da) X- Chi -HZ > HX-Chl-HZ + [OH] + H2O ^ ' (28.51Db) HX-Chl-HZ + CO2 — > jHCOo) + X- Chi -HZ The stationary concentration [{Chi}] (= [A-Chl-A'HaO] or fHX-- Chl.HZ]) is: (28.51E) [{Chl}]= ^*'^Chlo k'lCO-i] + k*I For the light intensity at which the yield of fluorescence, A-eChlo, when (28.5 IH) reduces itself to (28.47Ba). (This extreme case is practically identical with the one derived above from mechanism 28.21 ; in both of them, half- saturation is reached when one-half of all chlorophyll complexes are in the changed state— catalyst Eb not being fully utilized even in the light- saturated state.) 1/, / differs from ./, / when the EB-Hmitation is significant (_/. e., when A'g is not 1^ /fgChlo). 1074 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 In the other extreme case, when k^ 1 (fluorescence is half -saturated when [ { Chi } ] = 3^ Chlo, photosynthesis is half-saturated in this model when [EbOH] = 3^ Eb; the two conditions are satisfied at the same light intensity only when A; = 1). The curves representing fluorescence intensity, F, as function of light intensity can be derived from the various expressions we have obtained for [ { Chi } ] by inserting them into equation (28.51B) . The resulting equations are either second or third degree, indicating that the curves F = /(/) are either hyperbolae or third order curves. They begin at 7 = 0 with the slope ACO2 + A'11,0 or: (28.53) AHCOa-Chl-A'HO— — ^ ACOo-Chl-A'H.O and if no exhaustion of the photosensitive form has occurred at the time hght saturation has been produced by Eb deficiency, there is no reason why such exhaustion should occur if the hght intensity is stepped up still further, since the products of the back reaction are wiuly to participate again in the primary photochemical reaction. Experimentally, an increase of fluores- cence at "supersaturating" light intensities has been noted in several cases discussed above. This could be explained by assuming, with Franck, that the back reaction liberates so much energy as to cause the reversal not only of the first oxidation-reduction step, but also of the carboxylation reaction : (28.54) AHCO, + A'HO > A + CO. + A'H^O (or A + COo + A' + H.O) or : (28.55) AHC02- Chi -A'HO > AChlA'HaO + CO2 (or A-ChlA' + CO2 + H.O) In this way, the products of the back reaction are added to the pool of free carbon dioxide and water rather than to the immediately available substrates of the primary photochemical process, ACO2 and A'H20. We recall that this hypothesis was first suggested by Franck to ex- plain an entirely different observation — the "carbon dioxide burst" some- times observed in the first minutes of illumination {cf. Vol. I, page 207, and chapter 29, page 1093). When the yield of fluorescence goes up with increasing light intensity, as in figure 28.25, and reaches a new steady value, ^2, in the region of the light saturation of photosynthesis (fig. 28.26), this can be taken as a sign that saturation is due to a preparatory dark reaction; it is thus under- standable why, in McAlister's experiment represented in figure 28.25, this change was observed in a C02-deficient medium and not in 5% CO2. The results obtained by Wassink and Kersten with Nitzschia (fig. 28.28) are puzzling. The fact that above 50 kerg/cm.^ sec. tp decreases rather than increases with light intensity could be formally explained by assuming that, in this organism, the form of the chlorophyll complex that accumu- lates during intense photosynthesis has a higher value of ki {i. e., dissipates energy more rapidly), so that the sum h -\- h increases in strong light even if kt declines to zero. AVhat is more difficult to explain is that in the absence of carbon dioxide the diatoms retain the high yield of fluorescence {(Pi) in strong light, while one would offhand expect that, in this case, the lower value (^2) would prevail from the very beginning. (It was mentioned INTERPRETATION OF LIGHT CURVES OF FLUORESCENCE 1077 on page 1051 that the curves would be easier to understand if the designa- tions "with carbon dioxide" and "without carbon dioxide" were exchanged!) In the case of purple bacteria, several states seem to be needed to ex- plain the light curves of fluorescence. First of all, the low value of if in weak hght (the sigmoid shape) needs interpretation. It is probably associated with the substitution of intercellular hydrogen donors for the external re- ductants, which occurs while photosynthesis is slow. The coincidence of the two curves in figure 28.29 seems to indicate that, when photosynthesis is prevented by the absence of reductants, either chlorophyll accumulates in one and the same form in the presence and in the absence of carbon dioxide, or the two forms (e. g., HX-BChl-Z and X-BChl-Z) accumu- lated under these conditions have a practically identical rate of energy dissipation, h. In the presence of reductants, the two forms accumulated with and without carbon dioxide (perhaps X-BChl-HZ and HX-BChl-- HZ) possess, to the contrary, a very different fluorescence capacity. How- ever, as the light intensity is increased, a further change in the composi- tion of the complex occurs, leading to the crossing of the curves with and without carbon dioxide. Figures 28.31-28.35 confirm that the absence of reductants causes (in the presence as well as in the absence of carbon di- oxide), the accumulation of a form with considerably increased capacity for fluorescence (which may be X-BChl-Z, or AC02-BChl-A'). An alternative explanation of the effect of reductants on fluorescence can be given on the basis of Franck's concept of "self-narcotization." Franck assumes that reductants such as hydrogen or thiosulfate intervene in bacterial photosynthesis by reducing the "photoperoxides" formed by the primary photochemical process. If the reductants are deficient, the peroxides accumulate and produce the "narcotic," that blankets the chlorophyll and causes fluorescence to become stronger. The absence of C'Oa has less effect in bacteria because they are studied under anaerobic conditions, permitting no photoxidation. Wassink, Katz, et al. (1938, 1942,1949) explained the effect of reductants by assuming that an "energy acceptor," capable of taking light energy over from bacteriochlorophyll, thus quenching its fluorescence, can be formed exclusively by enzymatic transformation of the rc(hictants. They followed that CO2 must have no effect on fluorescence at all — whii-h is not true. The effects of cyanide and of loio temperature on fluorescence (and photo- synthesis) can often be explained by assuming that the primary effect of both is the retardation of the carbon dioxide supply processes. However, we have seen, in part A, that not all tlie experimental results on cyanide 1078 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 inhibition agree with this simple explanation; the same is true of the fluorescence measurements. It is, for example, not clear why the concen- trations of cyanide needed to markedl}^ affect fluorescence are so much higher than those needed to inhibit photosynthesis. These and the results obtained with several other poisons and narcotics require so many ad hoc explanations that we do not want to attempt them here. Parallel measure- ment of gas exchange and fluorescence in the presence of various inhibitors seems to be a very promising approach to the unravelling of the complex happenings in the photosensitive chlorophyll complex; but the presently available results are hardly sufficient to warrant a detailed attempt. Franck — to whom we owe both the fundamental concepts of fluorescence and the demonstration of how these concepts can be usefully applied to the study of photosynthesis — has WTitten two reviews of this subject (1949, 1951), which contain many observations and interpretations that could not be covered in the above presentation. Bibliography to Chapter 28 Light Factor I. Light Intensity A. Light Curves of Photosynthesis 1866 Volkov (Wolkoff), A., Jahrb. iviss. Botan., 5, 1. 1882 Berthold, G., Mitt. zool. Station Neapel, 3, 393; Jaltrb. wiss. Botanik, 13, 569. 1883 Reinke, J., Botmi. Z., 41, 697, 713, 732. 1884 Sachs, J., Arb. botan. Inst. Wurzburg, 3, 1. 1896 Ewart, A. J., J. Linnean Soc. Botany, 31, 364. 1897 Ewart, A. J., Ami. Botany, 11, 439. 1898 Ewart, A. J., ibid., 12, 363. Giltay, E., Ann. jardin botan. Buitenzorg, 15, 43. 1903 Pantanelli, E., Jahrb. wiss. Botan., 39, 167. Weis, F., Compt. rend., 137, 801. 1905 Blackman, F. F., and Matthaei, G. L. G., I^roc. Roy. Soc, London, B76, 402. Oltmans, F., Morphologie und Biologie der Algen. Fischer, Jena. Vol. II, pp. 144, 190. Brown, H. T., and Escombe, F., Proc. Roy. Soc. Londoji, B76, 29. Lubimenko, V. N., Rev. gen. botan., 17, 381. 1907 Lubimenko, V. N., Compt. rend., 145, 1347. 1908 Lubimenko, V. N., Ann. sci. not. Geneve, (9) 7, 321. Lubimenko, V. N., Rev. gen. botan., 20, 162, 253, 285. 1911 Blackman, F. F., and Smith, A. M., Proc. Roy. Soc. London, B83, 389. 1914 Kniep, H., Intern. Rev. ges. Hydrobiol. Hydrog., 7, 1. Purevich, K., Jahrb. wiss. Botan., 53, 210. Putter, A., Naturwisscnschaften, 2, 169. 1917 Brown, W. H., and Heise, G. W., Philippine J. Sci., C12, 1, So. Plaetzer, H., VerhandL phys.-med. Ges. Wurzburg, 45, 3L BIBLIOGRAPHY TO CHAPTER 28 1070 1918 Boysen-Jeiisen, P., Botan. Tid., 36, 219. Brown, W. H., and Heise, G. W., Philippine J. Sd., C13, 345. Henrici, M., Verhandl. naturforsch. Ges. Basel, 30, 43. Willstatter, R., and Stoll, A., Untersuchungen uber die Assimilation der Kohlensdure. Springer, Berlin, 1918. 1919 Warburg, O., Biochem. Z., 100, 230. 1920 McLean, F. T., Ann. Botany, 34, 367. 1921 Harder, R., Jahrh. wiss. Botan., 60, 531. Lundeg&rdh, H., Svensk botan. Tid., 15, 46. 1922 Lundegardh, H., Biol. Zentr., 42, 337. AVarburg, 0., and Negelein, E., Z. physik. C/iem., 102, 246. 1923 Harder, R., Ber. dent, botan. Ges., 41, 194. Harder, R., Z. Botan., 15, 305. Warburg, 0., and Negelein, E., Z. physik. Chem., 106, 191. 1924 Bose, J. C, Physiology of Photosynthesis. Longmans, Green, Londor, 1924. Harder, R., Jahrb. iviss. Botan., 64, 169. Lundegardh, H., Medd. Centralanst. Fdrsokwds. Jordbruksomr. Medd., No. 331. Maucha, R., Verhandl. Intern. Ver. Limnologie, II 2, 381- 1925 Dastur, R. H., Ann. Botany, 39, 769. 1926 Johansson, N., Svensk botan. Tid., 20, 107. Ruttner, F., Intern. Rev. ges. Hydrobiol. Hydrog., 15, 1. Spoehr, H. A., Photosynthesis. Chemical Catalog Co., New York, 1926, p. 33. 1927 Kostychev, S., Bazyhna, K., and Vasiliev (Wassilieff), G., Biochem. Z., 182, 79. Marshall, S. M., and Orr, A. P., /. Marine Biol. Assoc. United Kingdom, 14, 837. Maucha, R., Intern. Rev. ges. Hydrobiol. Hydrog., 18, 388. 1928 Kostychev, S., Bazyrina, K., and Chesnokov, V., Planta, 5, 696. Ghosh, J. C, Jahrb. wiss. Botan., 69, 572; /. Dept. Sci. Univ. Calcutta, 9,12. Lubimenko, V. N., Rev. gen. botan., 40, 415, 486. Marshall, S. M., and Orr, A. P., J. Marine Biol. Assoc. United Kingdom., 15, 321. Maximov, N. A., and Krasnoselskaja-Maximova, T. A., Ber. deut. botan. Ges., 46, 383. Miiller, D., Planta, 6, 22. Yoshii, Y., ibid., 5, 681. 1929 Beljakov, E., ibid., 8, 269. Boysen-Jensen, P., Dansk. Skowfor. Tidsk., 1929, 5. Boysen-Jensen, P., and ^liiller, D., Jahrb. wiss. Botan., 70, 493. Boysen-Jensen, P., and Miiller, D., ibid., 70, 503. Lubimenko, V. N., and Tikhovskaja, Z. P., Arch. biol. Stat. Akad. Wiss. U. S. S. R. Sebastopol, 1, 153. Emerson, R., /. Gen. Physiol, 12, 623. 1080 THE LIGHT FACTOR. I. INTENSITY CHAP. 2R Ivanov, L. A., and Kosovich, N. L., Planta, 8, 427. Ehrke, G., ibid., 9, 631. Montfort, C, Jahrh. wiss. Botan., 71, 52, 106. 1930 Montfort, C, ibid., 72, 776. van der Honert, T. H., Rev. Ircw. botan. neerland., 27, 149. Kostychev, S., and Berg, V., Planta, 11, 144. Kostychev, S., Chesnokov, V., and Bazja-ina, K., ibid., 11, 160. Kostychev, S., and Kardo-S^ysojeva, H., ibid., 11, 117. Lundegardh, H., Klima mid Baden in Hirer Wirkimg aiif das Pfldiizni- leben. 2nd ed., Fischer, Jena, 1930. 1931 Ehrke, G., Pla7ita, 13, 221. Stocker, 0., Ber. deut. botan. Ges., 49, 267. 1932 Boysen-Jensen, P., Stoffprodnktion der Pflanzen. Fischer, Jena, 1932. Harder, R., Filzer, P., and Lorenz, A., Jahrb. wiss. Botan., 75, 45. Chesnokov, V., and Bazyrina, K., Trav. inst. biol. Peterhof, 9, 103. Hiramatsu, K., Science Repts. Tbhokii Imp. Univ., [II], 7, 239. Miiller, D., P/anta, 16, 1. van der Paauw, F., Rev. trav. botan. neerland., 29, 497. Wood, J. G., Australian J. Exptl. Biol. Med. Sci., 10, 89. 1933 Boysen-Jensen, P., Planta, 21, 368. Hoover, W. H., Johnston, E. S., and Brackett, F. S., Smithsonian hist. Pub. Misc. Collections, 87, No. 16. Kursanov, A. L., Planta, 20, 535. Montfort, C., Biochem. Z., 261, 179. Montfort, C, Protoplasma, 19, 385. Harder, R., Planta, 20, 699. 1934 Daxer, H., Jahrb. wiss. Botan., 80, 363. Emerson, R., and Green, L., /. Gen. Physiol., 17, 817. Emerson, R., and Green, L., Nature, 134, 289. Montfort, C., Jahrb. wiss. Botan., 79, 493. Schomer, H. A., Ecology, 15, 217. 1935 Burk, D., and Lineweaver, H., Cold Spring Harbor Symposia Quant. Biol., 3, 165. Brackett, F. S., ibid., 3, 117. Ashby, E., and Oxley, T. A., Ann. Botany, 49, 309. Barker, H. A., Arch. Mikrobiol, 6, 141. Blagoveshclienskij , A. V., Planta, 24, 276. Gabrielsen, E. K., ibid., 23, 474. Miller, E. S., and Burr, G. 0., Plant Physiol. 10, 93. Singh, B. N., and Kumar, K., Proc. Indian Acad. Sci., Bl, 754. Stocker, 0., Jahrb. wiss. Botan., 81, 464. 1936 Lubimenko, V. N., Sovet. Botan. 1936, No. 5, 31. ISIontfort, C., Jahrb. wiss. Botan., 84, 1. Smith, E. L., Proc. Nat. Acad. Sci. U. S., 22, 509. BIBLIOGRAPHY TO CHAPTER 28 1081 1937 Stalfelt, M. G., Planla, 27, 30. Curtis, J. T., and Juday, C, Intern. Rev. ges. Hydrobiol. Hydrog., 35, 122. Franck, J., and Herzfeld, K. F., /. Chem. Phys., 5, 237. French, C. S., /. Gen. Physiol, 20, 711. Gessner, F., Jahrb. wiss. Botan., 85, 267. Kjar, A., Planta, 26, 595. M5nch, I., ibid., 85, 506. Noddack, W., and Komor, J., Angew. Chem., 50, 271. Smith, E. L., J. Gen. Physiol, 20, 807. 1938 Eymers, J. G., and Wassink, E. C., Enzymologia, 2, 258. Gessner, F., Jahrb. wiss. Botan., 86, 491. Manning, W. M., Juday, C., and Wolf, M., /. Am. Chem. Sac, 60, 274. Smith, E. L., J. Gen. Physiol, 22, 21. Stocker, 0., Rehm, S., and Paetzold, I.. Jahrb. wiss. Botan., 86, 556. Wassink, E. C., Vermeulen, D., Reman, G. H., and Katz, E., Enzymologia, 5, 100. 1939 Eichhoff, H. J., Biocliem. Z., 303, 112. Stalfelt, M. G., Planta, 29, 11. Magee, J. L., DeWitt, T. W., Smith, E. C., and Daniels, F., /. Am. Chem. Sac, 61, 3529. Noddack, W., and Eichhoff, H. J., Z. physik. Chem., A185, 222. Petering, H. G., Duggar, B. M., and Daniels, F., /. Am. Chem. Soc, 61, 3525. Stalfelt, M. G., Planta, 30, 384. Wassink, E. C., and Katz, E., Enzymologia, 6, 145. 1940 Gabrielsen, E. K., Dansk Botansk Arkiv, 10, 1. Noddack, W., and Kopp, C., Z. physik. Chem., A187, 79. 1941 Riley, G. A., Bull Bingham Oceanogr. Coll., 7, (4), 1. Emerson, R., and Lewis, C. M., Am. J. Botany, 28, 789. Franck, J., and Herzfeld, K. F., /. Phys. Chem., 45, 978. Katz, E., Wassink, E. C., and Dorrestein, R., paper presented at Spectros- copy Symposium, Chicago, 1941. Weller, S., and Franck, J., J. Phys. Chem., 45, 1359. 1942 Katz, E., Wassink, E. C, and Dorrestein, R., Enzymologia, 10, 269. Wassink, E. C, Katz, E., and Dorrestein, R., ibid., 10, 285. Seybold, A., and Weissweiler, A., Botan. Archiv, 43, 252. Gabrielsen, E. K., Yearbook Roy. Vet. and Agr. College Copenhagen, 28. Steemann-Nielsen, E., Dansk Botanisk Arkiv, 11, No. 2. Greenfield, S. S., Am. J. Botany, 29, 121. 1943 Emerson, R., and Lewis, C. M., Am. J. Botany, 30, 165. 1944 Kramer, P. J., and Decker, J. P., Plant Physiol, 19, 350. 1945 Wassink, E. H., and Kersten, J. A. M., Enzymologia, 11, 282. French, C. 8., and Rabideau, G. S., J. Gen. Physiol, 28, 329. Franck, J., Pringsheim, P., and Lad, D. T., Arch. Biochem., 7, 103. 1082 THE LIGHT FACTOR. I. INTENSITY CHAP. 28 1946 Wassink, E. C, Enzymologia, 12, 33. 1948 Kok, B,, Enzymologia, 13, 1. 1949 Kok, B., Biochem. Biophys. Acta, 3, 625. van der Veen, R., Physiol, plantarum, 2, 217. Bohning, R. H., Plant Physiol, 24, 222. Steemann-Nielsen, E., Physiol, plantarum, 2, 247. 1950 Lanskaja, L. A., and Sivkov, S. I., Compt. rend. (Doklady) acad. sci. USSR, 67, 1147. B. Light Curves of Fluorescence 1932 Kautsky, H., Hiisch, A., and Davidshofer, F., Ber. deut. chem. Ges., 65, 1762. 1934 Kautsky, H., and Spohn, H., Biochem. Z., 274, 435. 1935 Kautsky, H., and Hirsch, A., Biochem. Z., 277, 250. 1938 Wassink, E. C, Vermeulen, D., Reman, G. H., and Katz, E., Enzymologia, 5, 100. 1939 Wassink, E. C, and Katz, E., Enzymologia, 6, 145. 1940 McAlister, E. D., and Myers, J., Smithsonian Inst. Pub. Misc. Collections, 99, No. 6. 1941 Franck, J., French, C. S., and Puck, T. T., J. Phys. Chem., 45, 1268. Katz, E., Wassink, E. C, and Dorrestein, R., paper presented at Spectros- copy Symposium, Chicago, 1941. 1942 Katz, E., Wassink, E. C, and Dorrestein, R., Enzymologia, 10, 269. Wassink, E. C, Katz, E., and Dorrestein, R., ibid., 10, 285. 1945 Wassink, E. C, and Kersten, J. A. H., ibid., 11, 282. 1947 Shiau, Y. G., and Franck, J., Arch. Biochem., 14, 253. 1949 Franck, J., "The Relation of the Fluorescence of Chlorophyll to Photo- synthesis," in Photosynthesis in Plants, Iowa State College Press, Ames, Iowa, 1949, p. 293. Katz, E., "Chlorophyll Fluorescence as Energy Flowmeter for Photosyn- thesis," ibid., p. 287. 1950 Franck, J., A7in. Rev. Biochem. (in press). 1951 French, C. S., and Koski, V. M., Proc. Sac. Exptl. Biol, (in press). Chapter 29 THE LIGHT FACTOR. II. MAXIMUM QUANTUM YIELD OF PHOTOSYNTHESIS* In analyzing the light curves of photosynthesis in chapter 28, we did not discuss the slope of the initial linear part. This is a particularly im- portant quantity, since it determines the maximuni quantum yield of photo- synthesis and the maximum conversion of light into chemical energy achieved in this process. The present chapter will deal with these two subjects. The definition of maximum quantum yield as the limiting slope of the light curve at low light intensities implies that this curve has no inflection. An inflection has often been observed in the light curves of purple bacteria, but it has usually been assumed that the light curves of algae and higher plants show no such complication. Certain recent observations (Kok, van der Veen) lead, however, to new doubts concerning the shape of the light ciu-ves below the compensation point; we will discuss these observations and their possible significance for the determination of the maximum quantum yield later in this chapter (page 1113). If the yield of photosynthesis is expressed in moles of reduced carbon dioxide (or liberated oxygen), P, and the light absorption, la, is given in einsteins of absorbed photons, the ratio y = P/Ia is the quantum yield. (In many papers on photosynthesis, and photochemistry in general, the quantum yield is designated by the letter cp, which we reserved — (c/. Vol. I, page 546) — for the quantum yield oi fluorescence. If the yield is measured by the energy content (heat of combustion) of the produced carbohydrates, — AHc, and /„ is given in calories, the ratio e = — AHc/Ia can be called the energy conversion factor. The relation be- tween 7 and e is: (29.1) 6 = -AH,ny/NAhv = 3.9() X IQ-'' X„^7 — 4 X lO'^ X,„^7 where AH„, is the molar heat of combustion of one {CH2OI group in carbo- hydrates (approximately 112 kcal, or 4.69 X 10^^ erg); Na, Avogadro's number (6.02 X lO^^); h, Planck's constant (6.55 X 10""); and p, the frequency of light (3.00 X 10^'' /K J- The concept of the "quantum yield" of a photochemical process arose from Einstein's application of quantum theory to photochemistry in 1913. Einstein suggested, in elaboration of Planck's concept of vibrational energy quanta of electrons in atoms and molecules, that light energy, too, consists * Bibliography, page 1139. 1083 1084 THE LIGHT FACTOR. II. QUANTUM YIEL CHAP. 29 of finite quanta (photons); from this, he deduced that the number of molecules, N, changed photochemically by the absorption of a certain amount of light, must be equal to the number of absorbed photons, N^^ (Einstein's law of photochemical equivalency). In the following years, rapidly accumulating rate measurements of photochemical reactions made it clear that Einstein's principle applies only to the primary photochemical process, while the observed over-all rates of photochemical reactions usu- ally depend on the efficiency of secondary reactions, which follow the pri- mary photochemical step. Over-all rate measurements therefore only seldom lead to straightforward confirmation of the equivalency law (in other words, the empirical quantum yields usually are smaller — or larger — than unity). Photosynthesis, as the most important photochemical process in na- ture, naturally came under scrutiny from the point of view of its quantum yield. The problem appeared particularly intriguing because of the strongly endothermal character of the photosynthetic reaction. It could easily be calculated that one quantum of visible light (energy available: 40-60 kcal/einstein) is insufficient to convert one molecule of carbon di- oxide (and one molecule of water) into a link in the carbohydrate chain and a molecule of oxygen (energy needed: about 112 kcal/mole). It was obvious that several quanta must cooperate in the reduction of one mole- cule of carbon dioxide. The question was: how many (or rather: how few — since, from the point of view of reaction mechanism, we are above all interested in the maximum quantum yield obtainable under the most fav- orable conditions). According to equation (29.1), the answer to this question meant also the determination of the maximum efficiency of plants as converters of light energy into chemical energy. Over a century ago, in 1845, Robert Mayer recognized that storage of light energy by conversion into chemical energy is a most important aspect of plant activity on earth (c/. Vol. I, chapter 2). We have described in the preceding chapter several investiga- tions in which the yield of this conversion was measured over considerable periods of time, and concluded that under natural conditions it is rather low — of the order of 2-5%. It was known, however, since Reinke's investigation in 1883 (c/. page 964) that the light curves of photosynthesis are convex; the curvature sometimes becomes apparent even at very low light intensities (c/. chapter 28, section A2). This means that the energy conversion efficiency and the quantum yield increase as light intensity decreases. Warburg and Negelein (1922) set out to determine the maximum quantum yield by measuring the yield in very Ioav light. Their work marked the beginning of a new stage in the quantitative study of photosynthesis, QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1085 The inverse of the quantum yield (or quantum efficiency, which is the same thing), is called the quantum requirement. Kegrettably, the first term is often used when the second one would be appropriate — for example, it is said that "the quantum yield of photosynthesis is 4" (or 8, or some other number, where n > 1), instead of saying that it is V4 (or Vs, or, gen- erally, 1/w). 1. Quantum Yield Measurements by the Manometric Method INIost quantum yield determinations of photos>aithesis were carried out by the manometric method described in chapter 25 (see fig. 25. 3 A). In this method, the change of gas pressure is measui'cd first above a dai-kened, and then al)ove an illuminated cell suspension. The net effects obser\'ed are the result of pressure changes due to the production and consumption of both carbon dioxide and oxygen. If both the respiratory quotient and the photosynthetic quotient are unity, the net pressure changes are different from zero only because of the greater solubility of carbon dioxide in water (or, still more, in alkaline buffers), as compared with that of oxygen. To obtain a check on the two quotients, the measurement can be repeated, in darkness and in light, with a different ratio of gas-filled and liquid-filled volumes (cf. fig. 25. 3H). {a) Investigations of Warburg and Negelein Warburg and Negelein (1922, 1923) were the first to apply the mano- metric method. They worked with suspensions of the unicellular green alga Chlorella. (The species was described by them as Chlorella vulgaris, but subsequent experience makes it uncertain whether it was this species, or C. pyrenoidosa.) To avoid the difficulties of the measurement of light absorption in plants caused by scattering (cf. chapter 22), they used dense suspensions, absorbing practically all the incident light. Consequently, at any given moment, most of the cells were shaded, and their contribution to photosynthesis was small; on the other hand, all cells contributed equally to respiration. For this reason and because of the low light in- tensities used (of the order of 1000 erg/cm. ^ sec), the total volume of respiration was larger than that of photosynthesis. (In other words, Warburg and Negelein worked below the compensation point.) They noted that the respiration of Chlorella was markedly stimulated by pro- longed exposure to light (cf. chapter 20, page 564) . In order to avoid such changes in respiration during the experiment, Warburg used illumination periods of not more than 10 minutes, separated by equal or longer periods of darkness. Because of the sluggish response of the manometer to changes of gas concentration in the liquid, the determination of the gas exchange 1086 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 generally requires an interpolation, illustrated by figure 29.1. The uncer- tainty caused by this interpolation is unimportant for extended illumina- tion periods, but can markedly affect the results obtained in short experi- ments, for example, experiments of 5 or 10 minutes. Furthermore, short illumination periods increase the importance of induction phenomena. Warburg knew from his earlier work (c/. chapter 33) that an "induction loss" {i. e., delayed onset of photosynthetic activity) can occiir after dark periods of the order of several minutes, but he also knew that this effect disappears if the light intensity is reduced considerably below the satura- tion region {cf. fig. 33.8). Since in the quantum yield work very low K >, >^ c-^ B 1 ^ ■v^ 3 A ~ ^,:JfT<::-.^ in '•^'- n 1 ^O^^^ (S) •^ *- 1 ^S,,^^ 'O^ UJ ^ ^ 1 tr ^ ^1 Q. H Dork 1 Light 1 t Dark 10 25 30 15 20 TIME, mm. Fig. 29.1. Manometric determination of quantum jdeld (after Rieke 1939). Solid lines represent assumed course of photosynthesis and respira- tion; dotted curve, the pressure changes read from manometer. KH is interpolated yield of photosynthesis in 10 min. light intensities were used, Warburg assumed that induction can be neg- lected. Subsequently, Emerson and Lewis (1939, 1941) found evidence of an induction phenomenon of a different kind, consisting in photochemical liberation of carbon dioxide during the first few minutes of illumination. This carbon dioxide "gush" or "burst" does not disappear with decreasing light intensity, and could be of particular importance in measurements in very low light. An explanation of the carbon dioxide burst, suggested by Franck (1942), already was described in Volume I (page 167). This theory suggests that the "burst" is caused by the decomposition of the carbon dioxide-acceptor complex, ACO2, accumulated in the dark. This decomposition follows the photochemical reduction of ACO2 to AHCO2, and the reversal of this reduction by back reactions, e. g., in Franck's notation: ,^^ «, , • .-.^ TT^, ,, forward reaction back reaction (29.2) lACOo + HChll :r^ >|AHC02 + Ch]} > light CO2 liberation „ HChl + AGO* } > HChl + CO2 QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1087 or, in the notation used in chapter 28 (pp. 1032, 1036, etc.): , . ,TT ^ forward reaction back reaction (29.2a) ACOa-Chl-A'HjO > AHCO.Chl-A'OH > light A'H,0*Chl.ACOf ^'''^'''"''°" A-Chl-A' + H^O + A + CO^ Here, asterisks indicate that the compounds formed by back reactions contain consider- able excess energy and therefore tend to decompose into their constituents. These back reactions normally occur only in saturating light (they are, in fact, supposed to be re- sponsible for saturation); but in the first moment of illumination, practically all AHCO2 formed (even the small amounts produced in weak light) undergoes back reac- tion, because during this "induction period," certain catalysts have not yet been "re- activated," and are unable to take care of the products of the first photochemical reac- tion. A difficulty of this hypothesis is that, even with a 100% yield of the back reaction, the rate of production of ACO2 in weak light must be small compared with the same rate in strongly oversaturating light. In the latter case, all intermediates formed in ex- cess of the saturating rate are supposed to undergo back reactions; and yet, under appropriate supply conditions, no carbon dioxide limitation is observed, indicating that either the ACO2 complexes formed by back reactions do not dissociate, or the recombina- tion of A and CO2 is so fast as to prevent any exhaustion of ACO2 (in other words, the rate ceiling imposed by the formation of ACO2 must be high compared with the full rate of the primary photochemical process and not only compared with the rate of the finish- ing dark reaction). The total volume of the gush — which is about equivalent to the quantity of chloro- phyll present in the cells — is in agreement with Franck's hypothesis; but the slow re- absorption of carbon dioxide in the dark (c/. fig. 29.3B, p. 1092) requires an explana- tion, since the time course of the "pick-up" (c/. Vol. I, fig. 22) indicates that the car- boxylation equilibrium C'Oo + A -^ ACO2 usually is established in a few seconds. It may be noted that a similar difficulty was encountered in the attempt to attribute the uptake of radioactive carbon dioxide in the dark (fig. 21, Vol. I) to the same carboxyla- tion process. Another problem is presented by the necessity of a high carbon dioxide concentration (>o%) for the "saturation" of the gush, since the shape of the carbon dioxide curves of photosynthesis indicates that the acceptor must be saturated with carbon dioxide even below 0.1% CO2. These discrepancies suggest that perhaps the gush and its reversal in the dark are manifestations of a carbon dioxide metabolism related to respiration and fermentation rather than to the first step of photosynthesis; but this hypothesis, in turn, fails to ex- plain the apparent close relation of the carbon dioxide liberated in the gush to the chlorophyll complex (without such a relationship, a photochemical liberation of carbon dioxide with a high quantum yield would be difficult to understand). The "cross-linking" of respiration and photosynthesis at an intermediate reduction level, between CO2 (L = 0) and carbohydrate (L = 1), e.g., on the level of oxalacetic or malic acids (L = 0.625 and 0.75, respectively), hypothesized by Calvin and co-workers (chapter 36), if confirmed, could explain how respiratory decarboxylations might be af- fected by reactions in the photochemical reaction sequence. A carbon dioxide burst of the volume observed by Emerson and Lewis would be largely absorbed in carbonate buffers. As it was, Warburg and Negelein had decided that acid solutions (e. g., water equilibrated with an 1088 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 atmosphere containing 5% CO2) offer a better promise of full photosynthe- tic efficiency, and chose to use them instead of the "unphysiological" alkal- ine buffers. Since only a small part of the carbon dioxide liberated in the burst can be caught in pure water, the larger part must escape into the gas space, and the consequent increase of pressure will be interpreted as in- creased photosynthesis, unless a check on the photosynthetic quotient, Qp, reveals that the liberated gas is mostly carbon dioxide and not oxygen. It will be noted (c/. Vol. I, page 31) that we designate the photosynthetic quotient as Qp and define it as the (positive) ratio — AO2/ ACO2, while Warburg (and many others) designate the photosynthetic quotient as 7 (using the symbol ip for the quantum yield), and define it as the (negative) ratio +ACO2/AO2. The discovery of a carbon dioxide gush by Emerson and Lewis has made the interpretation of the results of Warburg and Negelein uncertain. The latter's quantum yield values, calculated from net pressure changes in 10 minute exposures, with the assumption Qp = 1 (or more exactly, 1.09), turned out to be close to 0.25 or V4- Offhand, this result seemed eminently satisfactory in consideration of the fact that the reduction of carbon di- oxide by water involves the transfer of four hydrogen atoms {cf. chapters 3 and 7, Vol. I). If the value M had not been so plausible chemically, the fact that this high yield could be obtained only by following a specific schedule of experi- ments, combined with special methods of cultivation of the algae, would perhaps have attracted more attention. At first, using Chlorella cells grown in full light, Warburg and Negelein obtained only quantum yields of <0.06. Later they found much higher yields are obtainable with sus- pensions adapted to weak light. These experiments were carried out in yellow -f orange light; high values of 7 (up to 0.3) were obtained by extra- polation to 7 = 0, since the light curves bent markedly even below 1000 erg/cm.2 sec. In a second paper (1923), in which monochromatic fight was used, the curvature was less pronounced and Warburg and Negelein calculated, without recourse to extrapolation, quantum yields ranging from 0.20 in blue, to 0.23 in red fight, corresponding to energy conversion factors from 0.34 to 0.59. The least number of quanta of red fight containing suf- ficient energy to cover the energy expenditure of the reaction CO2 + H2O — ^ O2 + { H2CO I is three ; the lowest plausible number of elementary reaction steps is four (corresponding to the transfer of four hydrogen atoms from water to carbon dioxide). From Warburg and Negelein's results, it ap- peared that plants are able to achieve photosynthesis with not more than four photochemical steps, leaving only a small margin to cover losses of energy by dissipation into heat, wh'n-h appear inovitaljle in a complicated chemical process. QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1089 When theorists tried to devise a detailed mechanism of photosynthesis, plausible not only from the point of view of chemistry, but also from that of energetics, they found themselves badly hampered by the straight-jacket into which the limitation to four quanta had put them. Franck and Herzfeld (1941), in particular, pointed out that, if the reduction of carbon dioxide has to be brought about by several consecutive photochemical steps (as in scheme 7.VA), the intermediates must have a certain degree of stabiUty in order to avoid reoxidatiou while waiting for the stipply of another quantum of light energy. Con- K(>quently, the ".stabilization energy" of several intermediate products must be added to the energj' requirements of photosynthesis; and a further allowance must be made for the heat of formation of the {CO2} complex and the heat of decomposition of a per- oxide (which is the probable precursor of free oxygen evolved in photosynthesis). 2 [HjOo] + (H; CO} AW. ^ 10 kcal hHj ^ 10 kcal Lhp I ^,5 kcal — Oo + (CH2O} + 3H2O AHj^ a 10 kcal CO2 + H2O 1 (CO2} + H 0 AHr= 112 kccl/mole ^^•(co^r^-"'^^'^' Fig. 29.2, Energy requirements of a linear four-stage mechanism of photosynthesis (after Franck 1941). Total energy required Ai? + A/fc + MI[C02\ + Ai/ii + Ai7i2 + A//13 + Ai/p = 210 kcal. /mole. The energy relations in photosynthesis, according to Franck and Herzfeld, are illus- trated by figure 29.2, in which Afl'i cOs) (the energy of formation of the carbon dioxide- acceptor complex), A///j, Aff/j, Aff^/^ (the "stabilization energies" of three intermedi- ates) and A/7p, the (>nergy of stabilization of the end products (which includes the de- composition energy of the peroxide, {II2O2}) are shown as additional energy terms, which together with the accumulated chemical energy, A^c, must be supplied by light. The stabilization energy required for the prevention of "backsliding" of intermedi- ates in periods of several minutes (which must pass, in weak light, in a dense suspension of green cells, between the absorption of two light quanta by one and the same chloro- phyll molecule) must be of the order of 10 kcal /mole; the heat of formation of the { CO2I complex was estimated in chapter 27 as >20 kcal/mole. Altogether, three intermediates, the ICO2} complex and the primary peroxide, must increase the energy requirements of photosynthesis from 112 kcal/mole to probably as much as 210 kcal/mole — whereas 4 einsteins of red light (X = 660 m^) provide only 170 kcal. 1090 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 The energy requirements are somewhat different if a "pyramidal" mechanism is postulated instead of the "linear" reaction sequence represented in figure 29.2 {i.e., if it is assumed that four light quanta produce four identical pairs of intermediates, which then undergo dismutation by dark reactions, finally giving one pair of finished products, [CH2O] + O2; cf. Vol. I, pages 156, 158 and 164). In this case, the four quanta re- quired to reduce one molecule of carbon dioxide can be absorbed by four different chloro- phyll molecules. (The same result can be achieved by other physical or chemical mech- anisms permitting a "collection of quanta" absorbed by several pigment molecules in one "reaction center." These "photosynthetic unit" theories will be presented in chapter 32.) In this case, the total energy requirement is obtained by adding to the accumulated energy (112 kcal/mole) approximately 20 kcal liberated in the formation of the [CO2] complex, and the energy amounts liberated in the several dismutations (or other "quanta-collecting" processes). It may be noted that one dismutation reaction (dismutation of a peroxide, yielding an oxide and free oxygen) was included also in the "linear" scheme. The dismutation of hydrogen peroxide liberates as much as 46 kcal per mole O2 (Table ll.I, Vol. I); but dismutations of organic compounds, such as the Cannizzaro reaction, are less exothermal (about 10 kcal/mole; cf. Table 9.III, Vol. I). Even so, three such dismutations, together with one dismutation of a peroxide, will bring the total energy requirement of the "pyramidal" reaction scheme up to the same 210 kcal, which were estimated above for the "linear" reaction sequence. For 16 years, the "4 quanta mechanism" of photosynthesis ^vas the ob- ject of admiration and the source of headaches for those who approached the problem of photosynthesis from the point of view of energy conversion. During this time, no serious attempts were made to check the experimental foundations of this mechanism, and the results of Warburg and Negelein were considered final. We will see in the next section that even during this time some measure- ments were made with the higher plants that gave considerably lower quan- tum yields; but because of less suitable objects and less precise methods, they were not considered to throw doubt on the validity of Warburg and Negelein's results. Beginning in 1938, however, a series of investigations appeared, in which the photosynthesis of the same algae as used by War- burg was studied by several methods (gas analysis, polarography and calori- metry) in the laboratories of the University of Wisconsin ; these measure- ments gave rather widely scattered results, but the yields were invariably much lower than 0.25. The maximum quantum yields observed in this work (to be described in some detail in section 2) were of the order of 0.1. These publications induced several investigators to repeat Warburg and Negelein's determinations, adhering as closely as possible to the original technique. Rieke (1949) used monochromatic light (mercury lines 546 and 578 m^) and an "integrating box" for the determination of light absorption. The pretreatment of the algae (adaptation to weak light), the light intens- ity (about 1000 erg/cm.- sec.) and the illumination periods (10 minutes) QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1091 were the same as in Warburg's work. The quantum yields calculated from ten experiments at 578 m/u ranged from 0.18 to 0.24, and those calculated from six measurements at 546 mn, from 0.17 to 0.20. Thus, Rieke found Warburg and Negelein's results reproducible, but only by strict adherence, not merely to the original method of culturing of the algae, but also to Warburg's schedule of illumination. Wassink, Vermeulen, Reman and Katz (1938) noted another peculiarity— the quan- tum yields determined according to the procedure of Warburg and Nege- lein, were affected by temperature — they increased from 0.11 to 0.20 when the temperature was changed from 0° to 29° C. Another peculiar observation was made by Rieke, who found that the kind of water used in the preparation of the algal culture had an effect on the quantum yield. Emerson and Lewis (1938, 1939) confirmed this. They used water from seven different sources, and obtained, under other- wise identical conditions, variations in the quantum yield from 0.16 to 0.27. The lowest values were obtained in glass-distilled water; addition of a stock niixture of "microelements" (B, Zn, Co, Mn, Mo, Cr, Ni, Co, W, Ti, V) increased it markedly; a similar result was achieved by an increase in the amount of ferrous sulfate in the nu- trient mixture (apparently, this compound contained all the micronutrients as impuri- ties). Emerson and Lewis found that, for securing the highest yields, the cells had to be grown for 5 or 10 days at 15-20° C, 20 cm. from four grouped 60 watt lamps followed by 3 days 30 cm. from a single 100 watt lamp. In agreement with Wassink and co-workers, they found a temperature effect — the highest quantum yield was observed at 10° C. At least 5% CO2 had to be present during the quantum yield measurement, and the light in- tensity could not exceed 350 erg/cm. ^ sec. These experiments could have been interpreted as indicating possible reasons for the low yields observed at Wisconsin, and thus supporting the validity of the results of Warburg and Negelein, if a new difficulty had not appeared. Emerson and Lewis found that, by combining all the favorable factors, 7 values could be obtained that were considerably above Warburg and Negelein's value of 0.25. With 10 min. illumination periods, 7 values up to 0.31 were obtained; and these were further increased by making the illumination periods even shorter-. Since even a quantum yield of M presented grave difficulties from the point of view of thermochemistry, yields of one third or higher were clearly incompatible with the accepted over-all reaction of photosynthesis. Emerson and Lewis suspected (1938, 1939) that the oxygen production in the first minutes of illumination may occur by reduction of accumulated 1092 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 intermediates of dark metabolism, rather than of carbon dioxide, and thus require less energy. This caused them to inquire into the value of the photosynthetic quotient in the first minutes of illumination. It was in this £ e < X CJ LU a: CO CO LU (T a. I- < re 12 -• — Dork » « 1 inht » »-i_,i. 10 5 3 X 10"' einstem/cm' mn 13 X lO'^einGtein/cm,^ mm. 08 - 0.6 - 0.4 - 1 02 - I 0.0 02 ^"■"•^^ — \ -^^ ^^^.^^—^ 0.4 - "^"-.- . 0.6 0.8 ' y "" -1.0 • 1 1 1 1 1 \ / N _ '' 1 1 1 1 1 1 1 r 1 10 20 30 40 50 70 90 110 TIME, min. 130 150 170 190 Fig. 29.3A. Rates of pressure change during alternating periods of light and darkness, for the same quantity of cells, in two vessels (after Emerson and Lewis 1941). Vessels differ in gas volume, thus permitting separate calculation of [O2] and [CO2] (cf. fig. 25.3B and 29.4A). t UJ o < X o X LlI \ 4 *v \ ^ ^ A \ \ 6 \j \ 6 \ ♦ * \/l/ \ \ ^ \ ^ D^ 1 8 C 8 \ ---"■jK I C^ ^f \ 10 12 14 16 10 12 14 16 /I \ ^ \ V 0 2 4 6 8 10 12 14 16 18 6 12 18 24 30 36 42 48 0 IB 24 30 36 42 48 \ Light 1 Da rk c > < =r^ „, / "^ H N \ \ 0 6 12 18 24 30 36 42 48 TIME, min. u \ Light D 2 \. s s' f 4 6 8 10 12 \ \ N, Da rk > \ ' ' N 1 i \ \ \ N 14 16 18 \ \ C ) ( 5 1 2 1 8 2 4 3 0 3 6 4 2 46 TIME, mm. Fig. 29.4. Quantum yield measurements with Chlorella (after Warburg 1948). Yellow light, 10°C. Shaded areas represent "induction losses" attributed to sluggishness of manometer. (Curve .4 shows a sliglit excess instead of deficiency of gas evolution at the beginning of illumination.) Quantum yield is determined from difference between slope in light (.45 or CD, depending on (he method of calculation) and slope in darkness. 1100 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 The value used by Warburg and Negelein, Qp = +1.09, was determined by gas analysis, in much stronger light than was used for quantum yield determinations. Warburg now redetermined the Qp value in less intense light, manometrically, by means of two sets of measurements in one vessel filled with two different volumes of liquid (5 and 8 ml.). He found a Qp value of +1.07, practically identical with the quotient used by Warburg and Negelein. Furthermore, Warburg and Kubowitz could find no evidence of a "gas burst" in the first few minutes of illumination, except for a comparatively small effect, observed at the higher light intensities, especially when foam- ing occurred in the reaction vessel (c/. fig. 29. 4A). With the belief in the validity of the original experimental procedure thus strengthened, Warburg and Kubowitz proceeded to make new de- terminations of the quantum yield by the one-vessel method. The vessel was not silvered to better observe bubble formation (which Warburg con- sidered the most serious source of error) ; readings were made without inter- ruption of shaking. The light used was mostly the yelloAV mercury lines (578 m^) with an intensity of 325-2920 erg/cm. ^ sec. To minimize effects caused by sluggish gas exchange, a smaller fluid volume was used than in 1923 and two glass beads were put into each vessel to act as stirrers. Curves such as those in figure 29.4B and C were con- sidered by Warburg as confirmation of the interpretation of pressure dis- turbances at the beginning of the light and dark periods, as consequences of this sluggishness. He described these disturbances as "symmetric," meaning that the two disturbances cancelled each other and the y values therefore were the same, whether they were determined from the steady rates, omitting the measurements in the first few minutes of light (?'. e., from slope AB), or by interpolation, as in figure 29.1 {i. e., from slope CD). This procedure is equivalent to integration of the gas exchange over the whole period of the experiment, including the two transition intervals. This obviously does not apply to the curve in figure 29. 4A, where the pressure in- crease in the first minute is faster than afterward. Calculation from the steady* state (slope AB) gives in this case a I/7 value 25% higher than that obtained by inte- gration (slope CJ)). Table 29. Ill summarize.-^ the results. Warburg concluded from these experiments that the limiting quantum yield in weak light is 0.25 and that it declines to about 0.20 at 1500 erg/cm.^ sec. He suggested, as general explanation of the smaller values found bj^ other observers, failure to cul- ture algae of the highest efficiency; but his description of the methods of culture revealed no significant difference from those used by Emerson or Rieke. QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1101 Table 29.III Quantum Yields after Warburg (1946, 1948)" / 1/7, quanta 7, molecules O2 per 10 ~' einstein X, m/j niin. 17 cm. 2 erg/cm^, sec. per molecule O2 quanta See figure 578 0.158 324 3.96^■3.60^ 0.25 29. 4D 578 0.161 330 4.35'- 0,23 578 0.177 363 4.50'' 0.22 578 0.196 403 3.60'' 0.28 436 0.201 — (5.0) (0.20) 29. 4C 578 0.330 677 4.5'';4.6^ 0.22 436 0.390 — (6.0) (0.17) 578 0.400 824 4.25'' 0.24 578 0.750 1540 5.03^4.79' 0.20 29 . 4B 578 1.42 2920 5.56'';4.45^ 0.18 29. 4A ° Chlorella suspension, 10° C, complete absorption. 450 mm.^ cells in 5 or 8 ml. liquid; bottom area 17 cm.^ The two values in parentheses probably are affected by the light absorption by carotenoids. '' P'rom the steady state. "^ By integration. In discussing Warburg's results, Emerson and his co-workers (1949, 1950) pointed out that the pressure changes in the first minutes of ilkimi- nation are affected by two factors : the sluggishness of the manometer (em- phasized by Warburg) and the carbon dioxide burst. In very weak light, the burst may be spread almost uniformly over a 10-15 min. illumination period, and the sluggish transition is then clearly revealed by the initial measurements, as in fig 29. 4B and C. In stronger light, the burst is much more sudden, and its rapid decay overcompensates the effect of sluggish gas exchange, leading to curves such as that in figure 29.4 A. The spread of the burst in low light over the whole illumination period, accentuated bj' the sluggishness of the manometer, may answer one of Warburg's objec- tions— the absence of a visible pressure burst on the low light curv^es (figs. 29.4B-D). Warburg's other (and main) objection against Emerson's criticism was that a check "under the conditions of the quantum yields measurements" confirmed the validity of the Qr value of apin-oximately -fl, and that an extreme deviation of Qf from unity would have been needed to calculate from Warburg's experimental data (obtained with the one-vessel method) a quantum yield of about 0.1 for the liberation of oxygen in light. To this, Emerson and co-workers answered that Warburg's determination of Qp was based on subtraction of the rate of pressure change in darkness from the rate of pressure change in light (this difference was called the "light effect"). From the comparison of "light effects" in experiments with different amounts of liquid in the same vessel, Warburg derived the ratio AO2/ACO2 for (he light effect; and finding it close to 1, concluded that no 1102 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 significant carbon dioxide burst could have occurred during the iUumina- tion period. However, this procedure would only be permissible if the gas exchange in light were the result of the superposition of a photochemical process (photosynthesis, with possible addition of a carbon dioxide burst) upon a dark process whose rate is the same in darkness and in light. This is usually assumed to be true of respiration (although some doubts exist even here) ; but it is not true of the reahsorption of the carbon dioxide burst (since this process occurs only in the dark). By neglecting this component of the gas exchange in the dark after a period of illumination, one automatically eliminates from the calculated "light effect" a part, if not practically all, of the carbon dioxide burst — in the same way in which the effects of the sluggishness of the manometer are eliminated in the procedure illustrated by figure 29.1; no wonder that the ratios Qp for the calculated "light effect" prove to be close to unity. (Whether the elimination of the burst is practically complete or only partial, depends on what fraction of the burst is reabsorbed during the dark period utilized in the calculation of the "light effect.") To decide whether a significant carbon dioxide burst does occur in light (and is reabsorbed in darkness), the ratios AO/ACO2 should be calculated for the illumination and the dark period separately instead of calculating them directly for the "light effect." The ratios Qdark and Qught might each be quite different from 1 — and yet, the ratio "Qp" for the "light effect" might show no significant deviation from unity. Thus, the method of calculating Qp used by Warburg (1948) to prove the absence (or at least, practical insignificance) of the carbon dioxide burst is inappropriate for this purpose — even if the experimental data used had been adequate. Emerson and co-workers argued, however, that the experiment itself was open to criticism. They pointed out that the value Qp = 1.07 was derived from measurements lasting for about 40 minutes. The plot given by Warburg shows that if only the first 10 minutes of these measurements, i.e., the period of quantum yield measurements, were taken into considera- tion, much smaller values of Qp would have been obtained. Furthermore, the Qp measurements were made in light of 3780 erg/cm.^ sec. (ten times as strong as that at which quantum yields close to 0.25 were obtained) and in red light, while the 7 measurements were made in yellow^ light, Avhich is considerably less strongly absorbed. The conditions of Qp measurement differed from those of 7 measure- ment also in temperature (20° C. instead of 10° C), carbon dioxide con- centration (8% CO2 in O2, as compared with 5% CO2 in air) and volume of respiration (twice as high in quantum yield measurements as in Qp meas- urements, indicating different culture conditions) . It was mentioned before QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1103 that the voUime of the carbon dioxide burst depends on all these conditions ; the .statement that the Qp measurements were made "under the conditions of quantum yield determinations" was therefore not justified. Rabino witch (1947) pointed out that experiments show the integrated volume of the "burst" to change only little with light intensity, the main effect of the latter being on the suddenness of the burst. This relation is to be expected for photochemical emptying, with a high quantum yield (perhaps as high as 7 = 1) of a "carbon dioxide reservoir," containing a finite volume of carbon dioxide (the exact volume being dependent on con- ditions that prevailed prior to illumination) . If the volume of the reservoir corresponds to about one molecule carbon dioxide per molecule chlorophyll, the time required for complete emptying must be of the order of the time required for each chlorophyll molecule in the suspension to absorb a quan- tum of light; in the dense suspensions and in the low light used for the quantum yield determinations, this time is of the order of ten minutes (c/. chap. 32) ; this is then the expected duration of the burst. In stronger light, the burst will be proportionally shorter. If the volume of the burst increases only little or not at all with light intensity, its importance at 3780 erg/cm.^ sec. would be much smaller than at 320 erg. /cm. ^ sec. (where the value 7 = 0.25 was found). Emerson and Nishimura (1949) criticized also other experimental aspects of Warburg's work. They pointed out that the use of equal liquid volumes and different gas volumes in the two-vessel method (Emerson and Lewis, cf. fig. 25. 3B) assured better comparability of the gas exchange than Warburg's use of a single vessel filled with different amounts of liquid (since the efficiency of gas exchange between the two phases depends on the volume of the liquid). Objection was raised also to Warburg's time schedule, which involved consecutive runs first with the more concentrated and then with the more dilute suspension. Because of continuous change in the rate of respiration of cell suspensions, only simultaneous exposure and darkening of tAvo aliquots of the same cell material could vouchsafe the required high degree of their physiological comparability. Emerson and Lewis themselves did not quite meet this requirement: they, too, worked first with one, and then with the other vessel, but in contrast to Warbvu'g, they used a fresh aliquot of the stock suspension for each experiment (con- sidering this a less objectionable compromise than the use of a single sample) first for a series of measurements in a smaller volume of liquid and, after dilution, for a second series of measurements in a larger volume. The discussion of these apparently minor details points to the great practical difficulty of the (theoretically so simple) two-vessel method: it rests on the assumption that the obsei-ved difference between the two 1104 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 pressure changes is due entirely to different relative volumes of liquid and gas. Even a very small difference in the actual amounts of gas pro- duced (or consumed) in the two vessels, or in the speed with which this gas is transferred into the manometer, can lead to large errors in calculation. Several reasons for such variations can be anticipated. Small physio- logical differences may exist between samples taken at different times from the same stock solution or may arise in the course of the experiment. C'Onsiderable discrepancies of light absorption can bo caused by different position of the two vessels in the light beam or by dilforcnces in their shape and wall material. Shaking thins out the liquid layer in the middle of the vessel, and the consequent incompleteness of absorption depends on the total amount of the liquid present and on the shape of the vessel. In 1948-1949, an unsuccessful attempt was made to settle the quantum yield controversy by a combined effort of Warburg and Emerson in the latter's laboratory. Subsequently, in the summer of 1949, Warburg, Burk and co-workers (1949''^, 1950^-^) carried out quantum yield measurements by the two vessel method (this time with equal liquid volumes) at the Na- tional Cancer Institute in Bethesda and at the Woods Hole Marine Bio- logical Laboratory. In these experiments, single yields as high as I/70 = 0.44 (70 = 2.3) were observed; even the average yield was markedly above 0.25. The conditions under which these high jnelds have been obtained were quite different from and, in some respects, opposite to those which had been recommended bj'^ Warburg and Negelein in 1923. Cell Culture. The reduction of light intensity in the last day of culti- vation, recommended by Warburg and Negelein to adapt Chlorella cells to weak light, was discarded by Warburg and Burk. Another precaution, called unimportant by Warburg and Negelein, was now found to be es- sential: fast bubbhng of the carbon dioxide-bearing gas through the culture bottle, preventing the cells from settling out, and thus assuring adequate supply of oxygen and carbon dioxide to all of them. Very concentrated suspensions were used: 0.3 cc. cells in 7 cc. culture medium (phosphate buffer, p}l 4.9, saturated with 5% CO2 in air). (In 1948, Warburg used only 0.1 cc. cells in 5 or 9 cc. solution.) This was done to ensure complete light absorption, despite an increased rate of shaking, which created a more pronounced "hole" in the center of the reaction vessel; but the respiration correction — the main source of uncertainty in measure- ments of this type — was thus made even larger than before. Response of the Manometer. Compared to the 1923 experiments, the efficiency of shaking was increased (horizontal, back-and-forth motion of a rectangular vessel with an amplitude of 2 cm., 150 times per min.). It was asserted that stirring was thus made so effective that the response of the manometer to gas production (or absorption) in the liquid was QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1105 l)ractically instantaneous. No correction (of the type illustrated in fig. 29.1) was therefore used to account for diffusion through the liciuid and the exchange between the two phases. Instead, the yields were now cal- culated from manometer readings made at the very moment of changing from darkness to light, or from light to darkness. Errors caused by "physical lag" had been considered of prime importance in 1923 and 1948; in some examples given in these earlier papers the calculated quantum yields would have been quite different without correction for this lag. Time Schedule. The two vessels were filled simultaneously with ali- quots of the same culture, and exposed alternatively to the same beam of light (e.g., 10 min. light on vessel I, then 10 min. light on vessel II, then again 10 min. light on vessel I, and so on). Both vessels (total volumes 14 and 18 cc, respectively) contained the same amount of liquid (7 cc). It was argued that whatever physiological differences may have existed between the cells in the two vessels during the first exposure (because of a "phase difference" of 10 min.), must have disappeared after several light-dark cycles. This is plausible; however, Emerson and co- workers found that at least five or six (10 min. light + 10 min. dark) cycles may be needed to eliminate the initial difference, while in many of Warburg and Burk's published experiments (cf. table 29. IV) only 2 or 3 cycles were used. The alternate exposure schedule was altogether aban- doned in almost one half of all experiments — namely those in which "background" illumination was used to compensate respiration. Light Measurement. No physical determination of light intensity was made by Warburg and Burk. (Bolometers had been used in earlier experiments, both by Warburg and by Emerson.) Instead, light intensity was determined by means of the ethyl chlorophyllide - thiourea actinom- eter, for which a quantum yield of 1.0 was previously found (bolo- metrically) by Warburg and Schocken in Emerson's laboratory (cf. chap. 35). The quantum yield of the actinometer is known to decline with increasing light flux, particularly >0.1 /zeinstein/min. Many runs of Warburg and Burk were carried out in stronger light; the intensity of the beam was reduced in these experiments to about 0.1 ^einstein/min. by means of calibrated wire screens before it was directed on the actinom- eter. Light Intensity. In Warl)urg's 1923 and 1940 measurements, the use of very weak incident, light was considered important, since the quantum yield was found to decline significantly (cf. fig. 29.8) with in- creasing light intensity, beginning as early as at 1000 erg/cm. ^ sec. (about 0.03 )ueinstein/cm.2 min.). In the Warburg-Burk work, much higher incident light intensities werc^ uschI: instead of uniform illumination (jf almosi the whole boltoin aiea, as used in earlier experiments (1923, 1948), 1106 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 a sharp beam was now thrown on the bottom of the vessel (cross section of the beam, about 3 cm. 2; bottom area, 8.3 cm. 2). The total light flux (red light, 630-650 m/x) was 0.2-0.6 Meinstein/min., i.e. 0.07-0.2 /xeinstein per cm. 2 min. — ca. ten times higher than the intensity at which quantum yields of 0.25 had been obtained in 1948. Because of the extremely high density of the suspension, practically all this light was absorbed within a 1 mm. thick bottom layer (0.3 cc.) of the suspension; thus, at any given time, >95% of the cells were in darkness, while <5% were exposed to light, the incident intensity of which was close to the saturating value (the photosynthesis of light-adapted Chlorella is saturated, in red light, in a flux of about 0.5 jueinstein/cm.^ min.). Intermittency Effect. The finding of the highest quantum yields ever observed when the incident light was of almost saturating intensity appears startling. Warburg, Burk and co-workers explained this paradox by the intermittency of illumination : because of fast shaking, individual cells remain only for a very short while in the illuminated zone, and then plunge into darkness. (Assuming uniform stirring, each cell must spend >95% of the total "illumination time" in darkness, and less than 5% in Hght). Warburg and Burk proclaimed as a "new principle" that this type of intermittency of illumination permits maximum light utilization. This assertion is not easily reconciled with the results of experiments in flashing light, to be discussed in chapter 34: According to these experiments, intermittent illumination cannot in- crease light utilization above the maximiun value possible in steady low light. All that intermittency can do is to bring the quantum yield in parti- ally or even completely saturating light close to— but never quite up to — the quantum yield in low steady light. For the quantum yield increase caused by intermittency to be at all significant, the light periods must not be longer than the "Emerson-Arnold period" (0.01 sec. at 20° C, c/. chapter 34), allowing the limiting catalyst to work in the dark, after the flash is over, for a period of time which is sig- nificant compared to the duration of the flash itself. It is doubtful, however, whether shaking at the rate of 2.5 swings per second could lead to illumina- tion flashes of 0.01 sec, or shorter. Certainly, the "dark periods" in Warburg and Burk's experiments must have been :»0.01 sec; therefore, the assumption of Warburg and Burk, that under the conditions of their experiments all cells are engaged uniformly in photosynthesis throughout the "illumination period," requires revision of the major conclusions derived from experiments in flashing light. However, this assumption is not necessary for the validity of their argument (while the duration of the light period, mentioned in the preceding para- graph, is of crucial importance). QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1 107 Background Light. Because only 5% of all cells were illuminated at any given moment, even the very high incident intensity of the red light used in the Warburg-Burk measurements did not prevent the respiration correction from being equal to or larger than the photochemical gas ex- change. It has been suggested (this hypothesis will be discussed later in this chapter), that the effect of light below the compensation point could consist in reversing respiration midway (and not after it has led to the ultimate products, CO2 and H2O). To check this hypothesis, Warburg, Burk and co-workers made experiments in which the net gas exchange in light was made positive by substituting for dark periods, periods of dif- fuse illumination of the reaction vessels by white "background light" of such intensity that photosynthesis equalled or exceeded respiration. This background illumination was maintained also during the "light" period (when a measured beam of red light was added to it), so that the "light effect," from which the quantum yield was calculated, was the increment of gas exchange caused by an increment of illumination. War- burg and Burk argued that the quantum yields obtained in this way must be those of true photosynthesis, with the storage of 112 kcal. chemical energy per mole of liberated oxygen, and could not be those of a partial reversal of respiration (with an unknown, and possibly small, conversion of light energy into chemical energy), since this reversal, if at all possible, should be accomplished already by the background illumination. It will be noted that this argument is tied up with the assumption of uniform photosynthetic activity of all cells — those that are momentarily illuminated by the flash as well as those that are momentarily in darkness. If only the actually illuminated cells (or cells <0.01 sec. out of the illumina- tion zone) can contribute significantly to photosynthesis, then only the part of the background light that falls on these particular cells is of im- portance. This part is insignificant if the background light falls from above and is absorbed in the top layer of the suspension — while measured red light enters the vessel from below and is absorbed in a thin bottom layer of the suspension. Warburg and Burk (1950) described a single experiment in which the background light, similarly to the measured light, was thrown on the vessel from below. This light was so strong as to overcompensate respiration about fivefold; nevertheless, the addition of the measured light produced an increment of oxygen production equivalent to a quantum re- quirement as low as 2.8. It is unfortunate that this particularly important experiment was carried out with a particularly unsatisfactory time schedule — three 5-min. light-dark cycles in one vessel, followed by two 10-min. light-dark cycles in the other vessel. Quantum Yield in Carbonate Bufifers. The same cells which gave, in Warburg and Burk's experiment, high quantum yields at pH 5 (culture 1108 THK LIGHT FACTOR. IT. QUANTUM YIELD CHAP. 29 medium) gave 2-3 times lower yields (7 = 0.10-0.09) in carbonate buffers (pH ^^ 9), both with and without compensating white light (last section of table 29. IV). Yields of 0.12 or less were obtained also in bicarbonate solutions equilibrated with 5% carbon dioxide in air (pH 7-8). It thus appears that for Warburg and Burk's cells (grown in acid medium) even neutral solutions were "unphysiological." Respiration in Light. The question whether respiration is affected by light is crucial for the measurement of the rale of photosynthesis in weak light. Different indirect methods (Vol. I, Chapter 20, and Chapter 36) have been used to answer it, and have given contradictory answers (including all three alternatives: "no change," "stimulation," and "in- hibition"). The simplest approach to this prol)lem is to remove (e.g., by absorption in alkali) all carbon dioxide (including that produced by respiration) and to measure the oxygen consumption in light unobscured by photosynthesis. This procedure was attempted repeatedly, but with- out success, because immediate photosynthetic reutilization of respiratory carbon dioxide competed too effectively with its absorption by the com- paratively remote external absorber. In fact, it proved difficult to reduce photosynthesis in this way much below the compensation point. War- burg et.al. (1949^) reported, however, that Avith the increased frequency of shaking, they were now able to absorb respiratory carbon dioxide in an alkali-filled side arm of the reaction vessel so effectively that the rate of oxygen uptake by a Chlorella suspension in light was exactly the same as in the dark. They saw in this experiment the proof that respira- tion as such is quite unaffected by (red) light, and refutation of all hy- potheses which postulate an exchange of intermediates between photosyn- thesis and respiration. Complete prevention of photosynthetic reutilization of respiratoiy carbon dioxide by absorption of the latter in an external absorber (although reutilization must be possible even before the carbon dioxide had escaped from the cell into the medium), is a remarkable achievement. A possible reason why Warburg and co-workers were successful where others have failed is intermittent illumination. For 95% of the "light period" each in- dividual cell is practically in darkness. Respiration goes on during all this time; all, or at least a large part of the carbon dioxide produced while the cell is in the shade may be able to escape into the medium before the cell had moved into the illuminated zone. Once a carbon dioxide molecule is in the medium, it may have a much greater chance to diffuse into the gas space than to diffuse into the small illuminated volume. In this way, 80 or 90% of respiratory carbon dioxide produced during the "light period" could perhaps escape re-utilization by the cells and reach the external ab- sorber. QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1109 It will be noted that this explanation could not be used if Warburg and Burk's concept of all cells being uniformly engaged in photosynthesis throughout the "light period" were correct. More specifically, this ex- planation requires that not only the photocatalytic mechanism responsible for the liberation of oxygen, but also the enzymatic mechanism responsible for the uptake of carbon dioxide, should cease operating within <0.1 sec. after the cells are darkened. This seems to contradict the assumptions which we used on p. 207 in the explanation of the "pick up" of carbon dioxide after intense illumination in C02-deficient medium, on p. 308 in the explanation of the effect of cyanide on yield of photosynthesis in flashing light, and will use in chapter 36 in accounting for C*02 uptake by preillu- minated cells. In all these cases, we have assumed that the capacity to take up carbon dioxide survives, in preilluminated cells, for several seconds (or even minutes) after the cells had been darkened. However, as in many such cases, apparent contradictions may arise from the use of a qualitative, "yes or no" approach, where a quantitative, "more or less" analysis is re- quired. Summary of Warburg and Burk's Quantum Yield Measurements. Table 29. IV gives a summary of the quantum efficiencies reported by War- burg and Burk (1950); several of the experiments in this table have al- ready been discussed above. Whittingham, Nishimura and Emerson (1951) were able to reproduce Warburg and Burk's results by strict adherence to the same experimental arrangement and schedule of operations. However, they concluded that these results were affected by a sj'-stematic error. Following are the major points of their criticism. 1. The two-vessel method is very sensitive to slight errors in mano- metric determinations. Thus, a difference of 0.3 mm. in the pressure change registered in one of the two vessels over a 10-min. period may change the calculated oxygen yield by a factor of two. Such a difference is well within the limits of experimental error of the method of Warburg and Burk (as contrasted to the much more precise measurements with the differential manometer, emploj^ed by Emerson and Lewis.) 2. This low precision of the method leads to random scattering of re- sults, (for example, in experiment No. 7, the I/7 values derived from in- di\'idual cycles scattered from 2.3 to 14). This can be corrected by averag- ing over a sufficiently large number of cycles. However, only in a few experiments of Warburg and Burk, as many as five or six 10 min cycles were used; in most others, only two or three. This explains why even the averaged I/7 values scattered from 2.3 to 4.9. 3. H,andom errors can explain the scattering of the results, but only a systematic error can explain the consistent finding of I/7 values considerably 1110 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 Table 29.IV Quantum Requirements of Oxygen Production (I/7) and of Carbon Dioxide Consumption (Qp/7) Calculated by Warburg and Burk (1950) for Chlorella pyrenoidosa Expt. No. Description Number, duration and order of cycles 1 = light d = dark (or background light) 1/7 Qpi A. Experiments at pH 5 (two vessels) 1 No background illumi- 6 10'1/10'd. cycles alter- nation nating in two vessels 4.62 1.25 2 ct 2 10'1/10'd cycles, alter- nating 3.6 0.84 3 a 3 10'1/10'd cj'cles, alter- nating 4.2 1.03 4 II 1 10'1/10'd; 1 20'1/20'd; 1 30'1/30'd 2.8^ 0.82 5 u 5 10'1/10'd, alternating 2.5< 0.80 7 '7-KNO3 4 10'1/10'd alternating 4.2 0.96 I+KNO3 3 10'1/10'd alternating 3.2 0.76 9 a 1 20'1/10'd 4.8 1.22 1 Same material as in exp. 1 above, slightly 5 5'1/5'd cycles in first ves- sel, then 3 5'1/5'd cycles overcompensating in second vessel 2.9 0.98 white background light Same material, 2X 5 5'1/5'd cycles in first ves- overcompensating sel, then 6 5'1/5'd cycles background light in second vessel 4.5 1.11 2 Same material as in exp. 2 above & below 2 15'1/15'd cycles in first vessel, then '2 15'1/15'd 2X overcompensat- cj'cles in second vessel 3.9 0.96 ing background light 6 Continuous back- ground illumination by white light from 1 10'1/10'd and 2 30'1/10'd cycles in one, then same in other vessel, at the begin- above. Red measur- ning of experiment 4.2 0.90 ing light from below 2 10'1/10'd cycles first in added at various one, then in other vessel. times 3 19 hr. later 10'1/10'd cycles first in one, then in other vessel. 4.8 1.11 14 hr, later* 3.4 0.78 3X overcompensating 2 10'1/10'd cycles in one white light from then same in another ves- above sel 3.0 0,90 Same material, 4X 3 5'1/5'd cycles in one, overcompensating then 2 10'1/10'd cycles in orange red light another vessel 3.5 0.98 "mostly from below" 9 Same material as in 1 20'1/20'd cycle in each exp. 9 above, 3X vessel 4.4 1.00 overcompensating background light from above Average 3.8 1.04 Table continued on page 1111 QUA NTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1111 B. Experiments at pH 9.2 (Single vessel) 2 Same material as in 5'd/15'l/15'd/15'l/10'd 10.5 exp. 2 above; respi- ration overcompen- sated bj' white light from above Same, no background lO'd/10'l/lO'd/lO'l/lO'd/- light 20 'd 9.8 Average 10.15 1 ( AOa/ACOo) for "light effect" (p. 1101). 2 Average of six cycles, single cycles give l/7-values from 2.3 to 14. ^ No significant difference between 10', 20' and 30' cycles. * Four cycles gave very closely similar results; one value off. ^ Cells washed and re-suspended in fresh medium before this measurement because of apparent drop in yield. below those which Emerson and co-workers consider correct on the strength of their own manometric experiments, and of the nonmanometric measure- ments performed in various other laboratories. Emerson sees a possible source of such an error in the decision of Warburg and Burk to reverse Warburg's earlier practice, and to make no allowance for the "physical lag" between the time of gas exchange in the chloroplasts and the time Avhen pressure changes are registered in the manometer (cf. below). 4. Warburg and Burk's claim that this lag was negligible because of fast shaking could not be confirmed by Emerson and co-workers. The fact that in Warburg and Burk's experiments the pressure change often was the same (within the limits of experimental error) in the first, and in the second 5 minutes of a 10-min. light period, can be explained, according. to Emerson, by compensation of the lag by the carbon dioxide burst. Emer- son and co-workers, too, could obtain time curves in phosphate buffer without an apparent induction period; and yet, experiments made in the same two vessels, and with the same rate of shaking, but in carbonate buffer (eliminating the CO2 burst), showed a lag of considerable duration. 5. In each vessel, taken separately, the physical lag tends to decrease the manometric effect of the transition from darkness to light (and vice versa) and thus to make the calculated "light effect" smaller. This does not mean, however, that the calculated 7 values also must be too small. The quantum yield and the ratio Qp are calculated from the light effects in the two vessels; and whether the calculated 7 value is too small or too large depends on the ratio of the two lags. With the two vessels used by Warburg and Burk, the lag is larger in the vessel with the larger gas space, and this makes the calculated quantum yields too high. The systematic error caused by this unequal lag needs to be only very small (of the order of 0.3 mm. per 10 min.) for the calculated 7 to be increased by a factor of two. 1112 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 6. While Emerson and co-workers were able to closely reproduce the measurements of Warburg and Burk (I/7 — 3-4), the results became quite different if light and dark periods were lengthened to 30 minutes or if the vessel shape h2 (fig. 29. 4A) was substituted for hi in the two-vessel combination. Either change led to 7""^ > 9, with the same cells which gave values of 3-4 by following Warburg and Burk's specifications. These changes should diminish errors due to different physical lag, though per- haps not eliminate them, nor overcome all the disadvantages in the tech- nique of Warburg and Burk. However, it is significant that the yield was found to be dependent on both timing and vessel shape. h, H Hj Fig. 29. 4A. Manometric vessels for two-vessel method of quantum yield measurements. H, vessel with small gas space; hi hi, vessels with large gas volume. 7. Warburg and Burk have calculated — AO2/ACO2 values of the order of 1(±0.2) for the ''light effect" in the two-vessel experiments. However, this does not prove that no significant carbon dioxide burst and gulp had occurred in their experiments, but — as already was explained on page 1101 — merely that the "gulp" in the dark period compensated more or less completely for the burst in the light period. This compensation is inevitable in a series of light-dark cycles in which approximately stationary conditions are established after a few cycles. Separate calculation of — AO2/ACO2 in light and in darkness (suggested on p. 1102) gives — ^in the few cases where the necessary data are provided by Warburg and Burk — values quite different from 1, with deviations in the direction required by the burst-and-gulp hj^pothesis. 8. Warburg and Burk's experiments Avith white (or red) background light have additional uncertainty because they were carried out by consecu- tive, and not alternate, measurements in the two vessels. Earlier experi- ments of Emerson and Lewis had indicated that the burst occurs not only upon change from light to dark, but also upon change from one light in- tensity to a higher one; thus, the interpretation of experiments with light- compensated (or overcompensated respiration as "base line" can be the same as suggested above for the light-dark experiments. 9. To sum up the conclusions of Emerson and co-workers, the experi- QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1113 ments of Warburg and Burk can be duplicated by strict adherence to their specifications. However, the results obtained in this way are not only of low precision (as revealed by wide scattering) , but, what is more important, contain a systematic error. The differential manometer experiments of Emerson and Lewis (1941) had been far more precise than either the experiments of Warburg, Burk et al. (1948-1950), or the experiments which Emerson and co-workers made in 1949-1950 under conditions closely imitating those of Warburg and Burk. It seems that the most reliable of the presently available data on the quantum yield remain those derived from these older measurements.* Manometric quantum yield measurements have also been reported by Kok (1948, 1949). He considered the loss of light by scattering in thin suspensions as a lesser experimental difficulty than the large respiration, the wide variation of local light intensity, and the intermittency of illumi- nation inevitable in strongly agitated, dense suspensions. He therefore worked with Chlorella suspensions that absorbed only 30-40% of the in- cident light (yellow sodium light), and used an Ulbricht sphere for the meas- urement of absorption. He found practically linear light curves up to re- markably high incident intensities — sometimes as high as 20 times the respiration-compensating light! (Compare chapter 28, section A2). Quantum yield deteraiinations were made by Kok in four different ways: (1) by measuring the carbon dioxide exchange only, oxygen being absorbed by chromous chloride in the side arm of the Warburg vessel ; (2) by measur- ing the oxygen exchange only, carbon dioxide being absorbed, in the usual way, in carbonate buffer ; (3) by measuring both the carbon dioxide and the oxygen exchange by the two vessel method, and (4) by measuring the net exchange in a single vessel, and assuming Qp = 1.09. The quantum requirements, I/7, were calculated from the slope of the straight, ascending section of the light curves, thus avoiding explicit use of a respiration correction. [The underlying assumption is, of course, that the respiration, R, is the same at all light intensities at which a straight line is obtained for the function P — R = /(/)]. Kok found the so-calculated efficiencies to depend on the age of the suspension (c/. fig. 28.13). (This probably means, primarily, dependence on the freshness of the culture medium.) The yields were almost independent of the temperature and the light intensity used in the cultivation of the algae. They were about 20% higher in acid media (water, cultvu'e liquid, or phosphate buffer) than in alkaline carbonate buffers. Lowering the oxygen pressure to 0.25% had no effect on the quantum yield, and the same seemed to be true of changing the temperature from 10 to 20 or 30°C. The 1/7-values obtained by the four methods ranged (apart from a * New results by Warburg and Burk (1951 '■2) pertain not so much to the question of the quantum yield of photosynthesis, as to that of the mechanism of utilization of the quanta. They will be described in chapters 36 and 37. 1114 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 few exceptionally high figures) from 6.9 to 12.9; the average for non- alkaline media (methods 1, 3, and 4) was I/7 = 7.85; for alkaline buffers (method 2) about 10. Kok estimated I/7 = 6.75 as the most probable lowest value of the quantum requirement. The most interesting (and controversial) finding of Kok was that linear extrapolation of the light curves to 7 = 0 consistently lead to considerably smaller values of the gas exchange than would have corresponded to the respiration of the same cells in the dark. Upon closer study, he concluded that the light curve underwent a sudden change of slope by a factor of about 2, somewhere near the compensation point (fig. 29. 4B). He took this to mean that the quantum efficiency was constant from near the saturation point down to the compensation region, and then doubled sud- denly. This shape of the light curve — according to Kok it consists of three practically linear segments — has not been found by any of the previous observers; however, Kok claimed a confirmation of the sharp break in the P = /(/) curve by new analysis of the data of Kopp and of Gabrielsen. If the slope of the light curve changes by a factor of two at the com- pensation point, the rate of respiration in strong light, determined by linear extrapolation of the light curve from above the compensation point to 7 = 0, must indicate a rate of respiration in light equal to one half of the rate of respiration in darkness. Later (1949) using a more precise manometric device (a "differential volumeter") Kok found, as an average of 50 experiments with Chlorella cells grown in Knop's medium, Rugu = 0.5 Rdnvk- (To increase /^dark, 3 O O 1 p B 500 - / 400 - / 300 - / 200 - / 100 Ay // n 7 ' 1 1 I / ^ 10 15 20 f^D / (relative units) -200 A Without glucose o O £ 200 100 0 -100 /3 :>^ lU 15 ^^y I (relative units) B With glucose Fig. 29.4B. Light curves of photosynthesis at low liglit intensity after Kok (1949), showing knick near compensation point. QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1115 these measurements were made at 30°C.) However, the break in the p = /(/) curve was now found not exactly at the compensating intensity, but at about twice this intensity (fig. 29.4B(A)). Since i^ught was equal to 0.5 Edark, the ratio of the slopes below and above the break is, for these algae, 1.33 rather than 2.0. With Chlorella cells grown in glucose solution, the break was below the compensation point (fig. 29.4B(5)) ; the slope below the break was, in this case, exactly one half of that above it, which meant that i^ught was greater than 0.5 /^dark— perhaps reflecting enhanced respiration in the cyto- plasm. A break in the same region {i.e., below the compensation point) was found also in the Haematococcus pluvialis grown in inorganic medium; but in this case, the slope below the break was less than twice that above it. Light curves obtained with Cahomba leaves (floated on carbonate buffer) indicated that the break was present there too, and that -Rught — 0.5 Rdnvk- Kok suggested that these experiments indicate the existence of two hght processes, with the quantum requirement of the "low-light process" (which he called "light respiration," cf. below) exactly one half that of the "high light process" (true photosynthesis). When the slope in the low-light region was less than twice that above it, he interpreted this as indication that the two light processes were occurring simultaneously. When the slope in low hght was exactly H of that in high light, Kok assumed that the high light process did not begin until the low-light process was saturated. Kok considered these experiments (which had indicated a probable lowest 1/7-value of 6.75 above the break), as making plausible a quantum requirement of 6 for the high light process (true photosynthesis), and 3 for the low light process ("light respiration"). However, according to Franck (1949), Rieke found in Kok's method of light measurement an error which might have reduced the calculated quantum requirements by 20%; with this correction, the results become consistent with the assump- tion of quantum requirements of 8 and 4, respectively. The sharp breaks in the light curves, found by Kok, are very improbable (cf. the discussion in chap. 26 of the impossibility of a sharp break between the ascending and the horizontal part of the light curve, postulated by Blackman). However, even if the light curves are smoothly curved rather than broken lines the possibility remains that they may decline in the low hght region more steeply than would be expected from their shape in the region of higher light intensities. Once before — in the explanation of the alleged incapacity of cyanide to reduce photosynthesis below the com- pensation point (Vol. 1, page 308)— we have been led to the hypothesis that compensation of respiration in light may not requrie complete photosyn- thesis. We will confront the same situation in the description of the study, by Calvin and co-workers of respiration in light with the help of tracer lllfi THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 carbon (Chapter 36). From the latter experiments, Calvin drew the same conclusion as Kok — that the rate of respiration in strong light is only about one half of that in darkness. (The remaining one half may represent the proportion of the total cell respiration taking place outside the chloro- plasts and therefore not affected by light.) Kok suggested that photosynthesis substitutes for respiration by produc- ing energy' carriers (such as high energy phosphates or "HEP" molecules) which the organism rcfiuires for its metabolic activity, and which it ordi- narily derives from respiration. More specifically, Kok postulated that the primary light reaction in photosynthesis has a twofold function: (1) to produce reducing and oxidizing agents (HX and Z, cf. Vol. I, scheme 7. IV) capable, respectively, of reducing CO2 to C'HoO and of oxidizing Ho(J to O2; and (2) to produce HEP-molecules by transphosphorylations coupled with back reactions between these primary products: HX + Z + phosphate > HZ + X + HEP Until respiration is fully compensated — or, rather, suspended as unneces- sary (at least, in the chloroplasts) — the absorbed light is used only or mainly to produce HEP molecules. Respiration of one ICH2O} group has been reported to produce six HEP molecules (Ochoa, Lippman) ; Kok suggested that the same number can also be obtained by recombination of six (HX -f Z) pairs, and that these six pairs can themselves be produced by three cjuanta. Thus, two HEP molecules, containing about 20 cal./mole disposable energy, are fonned by one quantum of red light (about 40 cal./ einstein). Above the compensation point, Kok assumed a quantum requirement of 6; he postulated that here, too, each c^uantum produces two (HX + Z) pairs, and that out of twelve such pairs (produced by six quanta), four (produced by two quanta) react further to reduce { CO2 } to { CH2O } and to oxidize H2O to O2, and eight (produced by four quanta) react back, con- verting eight low energy phosphates into eight HEP molecules (which, in turn, are utilized as "boosters" in the reduction process). The quantum requirement would then be 3 for the reversal of respiration and 6 for true photosynthesis. (No explanation was given by Kok why eight HEP molecules are needed in the latter case, as against only 6 in the first one.) This, obviously highly arbitraiy scheme made to fit the (supposedly) experimentally indicated l/7-values of 3 and 6, is closely related to the "energy dismutation" schemes (such as scheme 9. HI) proposed (among other possible reaction schemes of photosynthesis and chemosynthesis) in chapter 9. The assumption that one third of all quanta are used in photo- synthesis to provide oxidation and reduction agents, and two thirds for the formation of energy boosters (HEP molecules), imitates the mechanism of QUANTUM YIELD MEASUREMENTS BY THE MANOMETRIC METHOD 1117 chemosynthesis postulated for hydrogen bacteria (schemes 9. IV) in which two hydrogen molecules reduce carbon dioxide, utihzing the energy lib- erated by the oxidation of four hydrogen molecules by oxygen. According to Franck (1949) (c/. page 1115), it is permissible to interpret Kok's results as indicating quantum requirements of 8 (rather than 6) for tme photosynthesis and 4 (rather than 3) for the "low light process." If one wants to retain Kok's picture, one can, for example, suggest that four quanta procUice four oxidation and reduction agents (4HX + 4Z), while the other four produce — by back reaction of another four (HX + Z) pairs — eight HEP-molecules. (The numerical analogy with the hydrogen bacteria would be lost in this way; but it is more plausible that one ciuantum pro- duces a single HX + Z pair than that it jjroduces two such pairs, as was suggested by Kok). Van der Veen (1949), in the course of a study of induction phenomena by the thermal conductivity method (chapter 33), found, for tobacco leaves, a light curve of photosynthesis similar to that recorded by Kok — a straight line up to 450 lux, and another straight line, with about half of the slope of the first one, from 450 to 3200 lux. He combined Kok's concept of the reaction mechanism of photosynthesis with scheme 9. IV, interpreting the "energy dismutation" postulated in this scheme {cf. pages 164 and 239, Vol. I), as production of HEP molecules by recombination of a part of the primary photochemical oxidation and reduction products, and "boosting" by these HEP molecules of the reductive power of the remaining reduction products. The specific numbers used in his scheme (eight recombinations to four oxidation-reductions) taken from Kok, could equally well be re- placed by others — e.g., by four recombinations and four oxidation-reduc- tions, as in scheme 9. Ill; and the same is true of the number of quanta re- quired (six, or eight, or even twelve). A somewhat different — and perhaps more plausible — interpretation of a comparatively low quantum requirement of "anti-respiration" in weak light has been suggested (Franck, 1949) : This is the (repeatedly mentioned) possibility that intermediates of respiration can be drawn into the photo- synthetic cycle and reduced back to the carbohydrate level, and that a smaller number of quanta is required for this process than for complete photosynthesis. It is important to note that such a half-way interception of respiration would not cause a deviation of the AO2/ ACO2 ratio from its normal value of (approximately) 1 — since the only gas exchange measured in low light will be that due to residual normal respiration (e.g., respiration outside the chloroplasts). Calvin suggested, on the basis of certain C(14) tracer experiments, that a cross-link between respiration and photosynthe- sis exists on the level of malic and oxalacetic acid; however, these observa- tions are still controversial {cf. chapter 36). 1118 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 Franck (1949), in offering an explanation of Warburg and Biirk's results in terms of photochemical half-way reversal of respiration, suggested that the extent to which this process occurs depends on the capacity of respira- tion intennediates (which probably are organic acids) to penetrate from the protoplasm into the chloroplasts, and that this capacity is affected by the physiological state of the cells. It remains to be seen whether this explana- tion can suffice to explain why many careful experiments have failed to show the existence of the phenomenon. Thus, Emerson and co-workers never had observed any curvature of the light curves in the region of the compensation point, which would indicate a lower quantum requirement in very low light. Brown and co-workers (1950) found no evidence that light interferes with respiration in mass-spectrographic experiments: The uptake of 0(16)0(16) from the air continued in light, while 0(16)0(18) was evolved simultaneously by photosynthesis from algae suspended in 0(18)- enriched water. It was mentioned before (page 1108) that Warburg and co-workers (1949) arrived at a similar conclusion by observations of the rate of oxygen consumption in darkness and light under conditions assuring rapid removal of respiratory carbon dioxide from the medium; it was, how- ever, suggested that these findings might have been contingent on the inter- mittency of illumination, which prevented the utilization for photosynthesis of a large proportion of respiration products. To sum up, the possibility of photochemical utilization of respiratory intennediates remains controversial, and the effect of this re-utilization on the quantum requirement in weak light, an open question. Suggestive experimental evidence is available for both a negative and a positive answer. (On the positive side: Warburg's cyanide experiments, Kok's and van der Veen's broken light curves, Calvin's carbon tracer experiments. On the negative side: Emerson and Lewis' smooth light curves, Warburg's experi- ments in C02-free medium. Brown's respiration study with oxygen iso- topes.") Whether Franck's suggestion, that the chloroplasts sometimes are and sometimes are not permeable to respiration intermediates formed in the cytoplasm, can explain these contradictions is uncertain. A knowl- edge of the relative contribution of chloroplasts and cytoplasm to total cell respiration in the dark would be useful in this connection; but no esti- mate of this relation has as yet been made. 2. Nonmanometric Measurements of Quantum Yield The results of nonmanometric measurements of the quantum yield on the whole agree with the lower figures (I/7 = 10 ± 2) found by Emer- son and Lewis, Rieke, and others by manometric studies rather than with the higher figures (7 = 3^^ to 3^4) claimed by Warburg and Burk. NONMANOMETRIC MEASUREMENTS OF QUANTUM YIELD 1119 (a) Chemical Methods Wurmser (1923, 1925, 1926) made a few quantum yield determinations with the green alga Ulva lactuca. Each experiment lasted for several hours, and consisted in the measurement of change in oxygen concentration in solu- tion by Winkler's method. The rate of absorption of Ught was calculated from comparison of transmission by green and discolored thalli (c/. chapter 22, page 675), using a theoretical equation to take into account scattering (c/. page 713). Wurmser found, in some of these experiments, energy con- version factors up to 50%, corresponding to quantum yields up to }4- However, the calculated yield in the (weakly absorbed) green light turned out to be so much higher than in the (strongly absorbed) red light, that it indicated probable grave errors in the calculation of absorption. Warburg (1925) therefore did not consider these experiments of Wurmser as sig- nificant confirmation of his own results. Briggs (1929) obtained, with leaves of Phaseolus vulgaris, at hght in- tensities 5-10 times stronger than those used by Warburg and Negelein, yields from 7-17 cc. O2/5OO cal absorbed energy; with yellow elm leaves, from 5.3 to 8.9 ml.; with green elm leaves, from 12 to 20 ml.; and with leaves of Samhucus nigra, from 9 to 19 ml. These values correspond to quantum yields <0.1. In 1935, Gabrielsen, working with plants of Sinapis alba, calculated, also from gas-analytical measurements, by extrapolating the light curves (fig. 30.8A,B) to zero illumination, e values from 0.13 in blue, to 0.36 in red Hght, corresponding to maximum quantum yields of 0.1 ± 0.02. Gabrielsen did not question at that time the correctness of Warburg's results, and thought that his lower yields must have been due to the use of a less ef- ficient species. Later (1947), Gabrielsen repeated these experiments with Sinapis, Corylus and Fraxinus leaves, and found 70-values between 0.082 and 0.078. In the first of a series of investigations emanating from the University of Wisconsin, Manning, Stauffer, Duggar and Daniels (1938) determined the quantum yield by gas analysis, comparing the composition of a gas (containing approximately 5% CO2 and 5% O2) conveyed through a Chlo- rella suspension in the light and in the dark. The suspensions were less dense than in Warburg and Negelein's work, absorbing only 10-50% of the incident light; the intensity of the latter (green line from a mercury lamp) was somewhat higher than in Warburg and Negelein's experiments (1000- 1750 erg/cm. 2 sec.) ; 60 minute periods of illumination were used. The 7 values derived from the absorption of carbon dioxide were not very different from those calculated from the increase in the concentration of oxygen, thus indicating that the quotient Qp was close to unity. 1120 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 The quantum yields obtained in these experiments scattered consider- ably— from 0.01 to 0.1 — but never exceeded the latter limit. Still lower quantum yields (0.002 to 0.027) were obtained in experiments in white light; in this case, however, about ten times higher intensities of incident light were used, so that saturation effects appeared possible. Experiments with a different technique (closed reaction bottles, no stirring, analytical determination of the change in [O2] in solution by Winkler's method) yielded 7 values between 0.02 and 0.065. In another paper from the same laboratory, Manning, Juday and Wolf (1938) de- scribed experiments in which bottles containing Chlorella suspensions were deposited at different depths in a lake, and thus exposed to different intensities of illumination, rang- ing from full sunlight (600 kerg./cm.'' sec, not counting the infrared) down to 6 kerg/cm.* sec. The change in color of the light with depth (cf. Table 22. XI) complicated the calcu- lation of the number of absorbed quanta; the results were therefore less exact than those of the first paper. However, the approximate magnitude of 7 values was the same as in other experiments — about 0.05 at the lowest light intensities (at 10 meter depth); c/. figure 29.5. 0.07 O UJ 3 f- < 3 O 0.1 0.2 0.4 10 20 40 100 LIGHT INTENSITY, (erg/cm^ sec.) x 10" Fig. 29.5. Quantum efficiencies for Chlorella (after Manning, Juday, and Wolf, 1938). Curve A, 3.17 hr., cell concn. 3,250,000/ml.; B, 1.03 hr., cell concn. 718,000/ml.; C, 4.00 hr., cell concn. 331,000/ml.; D, 4.00 hr. cell concn. 718,000/ml.; E, 4.05 hr., cell concn. 1,900,000/ml.; F, 3.30 hr., cell concn. 1,210,000/ml. (6) Polar ographic Method In a third investigation from the Wisconsin laboratories. Petering, Duggar and Daniels (1939) applied the polar ogra phi c meOiod {cf. page 850) because it permitted the determination of respiration immediately before and after a period of photosynthesis, without the delays (illustrated by fig. 29.1) inherent in the manometric method. Figure 29.(» shows the polaro- NONMANOMETRIC MEASUREMENTS OF QUANTUM YIELD 1121 graph to respond almost immediately to transitions from respiration (in darkness) to photosynthesis (in light) and vice versa. In the top and bottom curves, the illumination is below the compensation point, and photosynthesis manifests itself in a reduced rate of consumption of oxy- gen; in four other curves, the oxygen concentration increases during photosynthesis. The curvatures of the respiration curves show the un- certainty involved in calculation of the respiration correction. The rate of oxygen consumption in the five minutes immediately following the cessa- tion of illumination was used by the authors in the calculation of this cor- rection. (This method gives the highest respiration correction and conse- quently the highest quantum yield values.) in c T3 O 41.00 - 39.00 •£ 37.00 £ o c o > o z O 35.00 33.00 - t 31.00- o o >■ X o 29.00 - 27.00 - 25.00 - 20 40 50 30 TiME.min 100 Fig. 29.6. Photosynthesis and respiration of Chlorella measured by a polarograph (after Petering, Duggar and Daniels 1939). The experiments were made with white light; from 27 to 48% of the incident light was absorbed by the suspension. The calculated quantum yields ranged from 0.045 to 0.100, clustering around 0.07, and showing no trend with light intensity in the range from 1000 to 6000 erg/cm. ^ sec. New experiments with the polarograph were conducted by Moore and Duggar (1949). In these Chlorella cells were first illuminated with light of one color (intensity, 800-2500 erg/cm. ^ sec.) and then light of another color was added, and the additional yield determined. The idea behind this procediu'e was that the uncertainty concerning the amount of respira- 1122 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 tion in light can be eliminated by subtracting from the total gas exchange in the two combined beams the gas exchange in one beam alone. (There seems to be no difference between this method and calculation of 7 from the difference in yield at two light intensities in light of the same color.) The results of these measurements are shown in Table 29. V. The 7 values Table 29.V POLAROGRAPHIC DETERMINATION OF QUANTUM YiELDS OF Chlorella IN ReD AND Red plus Blue Light (after INIoore, and Duggar 1949) Initial beam Added beam Abs. % Abs., % 1/7 6500 6500 4350 6500 6500 6500 1958 783 1288 1610 1392 1188 76 46 91 49 74 52 0.074 0.12 0.09 0.10 0.10 0.11 4358 1575 100 0.11 9.1 6500 1180 50 0.10 10.0 5461 2532 40 0.08 12.5 5461 2437 64 0.09 11.0 4047 460 77 0.10 10.0 " Intensities in erg/cm.^ sec. for the superimposed beam (0.08 to 0.11) are not significantly different from those for the single beam (0.074 to 0.12), a result that can be inter- preted as indicating two things: approximate identity of quantum yields in blue and in red light (despite the absorption of light by carotenoids in the first-named region; cf. chapter 30), and approximate linearity of the Hght curve up to the total intensity of the two combined beams. Several objections can be made (and have been made by Warburg) against the polarographic quantum yield determinations: 1. The determination of the number of absorbed photons was not satisfactory. A thermopile was placed immediately behind the reaction cell ; the energy flux falling onto the thermopile with and without the algae in the reaction vessel (illuminated by parallel light) was multiplied by the ratio vessel area:thermopile area and the difference between the two products was assumed to be the absorbed flux. This assumption im- plies that scattering out of the beam intercepted by the thermopile is compensated by scattering into this beam, and thus neglects large angle scattering. The error caused by this could lead to too high a value for absorbed light energy, and hence to too low a value for the quantum yield. 2. The suspension was not stirred. (Stirring during measurements is impossible by the nature of the method; stirring between measurements was attempted, but found not to influence the results and was therefore abandoned). The algae did not settle during a run; and stirring is obviously of less importance when oxygen determination is made in the body of the liquid than when it is carried out in the gas phase above it. On the other hand, carbon dioxide exhaustion conceivably could occur in the immediate neighborhood of the cells, and cau.se a diminution of the quantum yield. However, this danger should not be serious when measurements are made at or below the compensation point. NONMANOMETKIC MEASUREMENTS OF QUANTUM YIELD 1123 3. Although the medium (nutrient solution, pH 5.5) satisfied Warburg's require- ments of "physiological" conditions, presence of mercury drops introduced a danger of poisoning. In fact, such poisoning has been observed, but deemed too slow to affect the measurements. (c) Calorimetric Method The basis of the calorimetric determination of the yield of photosynthe- sis— which is a direct measurement of the energy conversion yield, e, rather than of the quantum yield, 7 — was described in chapter 25 (page 854). The first to carry out such measurements was Arnold in 1936-1937; how- ever, so strong was the beUef at that time in the correctness of Warburg's value, y = 14, that Arnold took his inability to obtain this yield as indica- tion of a failure of the method, and did not pubhsh his results until 1949 (reference to them was made by Franck and Gaffron 1941). Arnold used a modified Callender's radiobalance, originally designed to measure heat production by radioactive materials. Its period was so small that com- plete measurements could be made in from 1 to 10 minutes. Between 0.05 and 4 mm.^ of cells, in Knop's solution or carbonate buffer, were used. One run was made with healthy ChloreUa cells, and one run with the same cells inhibited by ultraviolet irradiation. Respiration was assumed to be unaffected by ultraviolet light (c/. Vol. I, page 344). The results are shown in Table 29.Vl. Table 29.VI Calorimetric Determination of Quantum Yield (after Arnold 1949) Extra heat evolved in light" With With 100 Aff^ inhibited healthy 1 ' cells, cells, a Cells /„ J„ - AH^ AH^ % 1/7 HEAT PRODUCTION IN MICROAVATTS C. pyrenoidosa 5^08 3^70 Os 27^ 9.5 4.60 3.80 0.80 17.4 14.8 4.46 3.90 0.56 12.5 20.6 2.24 1.90 0.34 15.2 16.9 Avocado leaf 0.786 0.656 0.130 16.5 15.6 C. vulgaris 0.912 0.700 0.212 23.2 11.1 C. pyrenoidosa 1.86 1.34 0.52 27.9 9.2 16.2 12.9 3.3 20.4 12.6 heat PRODUCTION IN ARBITRARY UNITS Scenedesmus 2.63 2.04 2.40 1.65 1.36 0.59 0.84 0.43 0.39 22.4 25.9 20.7 22.3 11.5 C. pyrenoidosa . . . 3.24 2.08 1.75 10.3 12.4 11.5 " The illuminating light was from a neon arc (ten to fifteen red lines isolated by a red filter), and had a very low intensity — from 61 to 126 erg/sec. cm.^ 1124 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 Measurements with a photocalorimeter {cf. page 854) were carried out also at Wisconsin, by Magee, DeWitt, Smith and Daniels (1939), and gave 7 values from 0.049 to 0.110 (the average of 17 experiments was 7 = 0.077 or ITy = 13). No pronounced change with light intensity was noted be- tween 1200 and 8000 erg/cm. ^ sec. More recently, Tonnelat (1944, 1946), who worked in Wurmser's lab- oratory in Paris, published similar results of an investigation initiated in 1939. Tonnelat measured the heat developed in an adiabatic microcalorim- eter under three conditions: (i) when the calorimeter contained an il- luminated black-bottomed vessel with pure water (heat evolution, £"0); {2) when it contained the same vessel with an equal volume of an algal suspension in the dark (heat evolution, Er, due to respiration) ; and (5) when it contained the same suspension and was illuminated (heat evolu- tion, E = Eo-{- Er - AHc). The energy yield, «, was then: (29.3) e = AHc/E = (^0 + Eh - E)/E and the quantum yield, according to equation (29.1), assuming X = 530 mju (green light isolated by Wratten filter No. 62) : (29.4) 7 = e/(4 X 10-5 X 530) = 6/2.12 Illumination lasted for 16 hours without noticeable deviation from linearity (t. e., presumably, from the constancy of both R and P); even in the very low Ught used (about 1.5 X lO"* einstein of green light/cm min., or 550 erg/cm.^ sec), such long duration of the experiments seems dangerous. The results of Tonnelat's determinations are shown in Table 29.VII. T.\BLE 29.VII Efficiency of Photosynthe.sis (after Tonnelat 1944) Measured with Photocalokimetkr Chlorella Energy concn., conversion Experiment cclls/0.5 cm.^ factor (1) Reflecting bottom 32 X 106 0.26 (±0.06) (2) Reflecting bottom 33 0.34(±0.05) (3) Reflecting bottom 37 0.26 (±0.06) (4) Reflecting bottom 56 0.31(±0.05) (5) Reflecting bottom 194 0.08(±0.08) (6) Black bottom 36 0.12 (±0.07) (7) Black bottom 51 0.31 (±0.05) (8) Black bottom" 48 0. 18 " With 0.1% agar. Experiment 5 shows that the yield became low in very concentrated suspensions (a result Tonnelat attributed to inhibited gas exchange in the dense layer of cells on the bottom of the vessel, but which also could be due 2 QUANTUM YIELD OF BACTERIAL AND ALGAL PHOTOREDUCTION 1125 to "self -inhibition" effects, described on page 880). Experiment 8 shows decreased efficiency in the presence of agar— an observation that Tonnelat suggested may explain some of the low values found by McGee, DeWitt et al. (who used agar to prevent the suspension from settling). Experiments were performed in vessels with black or reflecting bottom. In the first case, t values could be too low (if absorption of light by the blackened bottom was counted as absorption by the cells) ; in the second case, they could be too high (because of possible escape of reflected radia- tion). Figures in Table 29.VII indicate that the first effect is real (experi- ment No. 6 shows a lower value of y at low cell concentration), but reveal no effects due to reflection. Tonnelat concluded from these experiments that the energy conversion factor, 6, is about 0.30, and calculated from this a quantum yield of Vg. Equation (29.1) gives, however, 7 = 0.030/2.12 = 0.U5, or approximately }i. Tonnelat's error was probably due to the use of an incorrect value, 8.9 X 10-12 instead of 7.9 X lO"^" erg/mole, for the heat effect of photo- synthesis. 3. Quantum Yield of Bacterial and Algal Photoreduction French (1937^), who worked in Warburg's laboratory, used the bacterial species Streptococcus varians to study the quantum yield of the reduction of carbon dioxide by molecular hydrogen (cf. chapter 5, page 104). The rate of hydrogen disappearance was determined manometrically. The lines 852 and 894 m^ were isolated by filters from the fight of a cesium lamp; from 17 to 58% of this light was absorbed by the suspension. The quan- tum yields calculated by French ranged from 0.07 to 0.23 molecule of carbon dioxide {i. e., from 0.14 to 0.46 molecule of hydrogen) transformed per quantum, depending on the pretreatment of the bacteria. French con- sidered these experiments proof that carbon dioxide reduction by Athio- rhodaceae requires four quanta per molecule of carbon dioxide, similarly to the assimilation of green plants according to Warburg and Negelein. As mentioned before, the fight curves obtained by French in this work were sigmoid: the y,naz- values were derived from the maximum slope of these curves (reached in the inflection point). This procedure requires justification. To divide the increase in yield in a certain region of the light curve by the corresponding increase in light absorption, and to call the quo- tient "quantum yield" presupposes that the increase in light intensity pro- duces a certain additional amount of photosynthesis, characterized by its own quantum yield. Wassink, Katz and Dorrestein (1942) suggested that this may, in fact, be the case, if the bacteria use, in weak light, mainly intracellular organic compounds, instead of the externally supplied react- ants. As light intensity increases, this photochemical process is soon light 1126 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 saturated (because of supply limitations, or because of the limited amount of a necessary enzyme) ; the normal reduction of carbon dioxide with ex- ternal reductants then comes into its own. The manometrically deter- mined quantum yield of the photochemical process using intercellular sub- strates can be smaller than that of normal photoreduction, for two reasons: (a) When hydrogen is used as external reductant, any utilization of non- volatile, intracellular reductants will diminish the rate of gas consumption, even if the quantum yield of carbon dioxide reduction is the same with both types of reductants. In French's experiments (1937) with Streptococcus varians, the light curve had total gas consumption, AH2 + ACO2 as ordinate; its sigmoid shape may have been due pntirely to an initial de- ficiency in the consumption of hydrogen alone. In the experiments of the same author with Spirillum ruhrum (1937^) a nonvolatile reductant was used, and the curve showing ACO2, as function of light intensity, showed no inflection. However, in the more recent experiments of Wassink, Katz and Dorrestein (1942) with Chromatium, sigmoid curves were obtained not only with hydrogen but often also with thiosulfate as reductant {cf. fig. 28.11). A different explanation is needed in this case. (6) It was described in chapter 5 (Vol. I, page 106) how, when exter- nally supplied organic compounds are utilized by photosynthesizing purple bacteria, the proportion of "coassimulated" carbon dioxide can vary widely (or carbon dioxide may even be liberated), depending on whether the or- ganic compound is utilized mainly or exclusively as hydrogen donor (as in Foster's experiments with secondary alcohols), or serves also as the source of carbon. Wassink, Katz and Dorrestein suggested that the same applies to photochemical utilization of intracellular organic materials; here, too, the consumption of external carbon dioxide may be more or less completely suppressed by the utilization of the carbon (in the form of freshly formed carbon dioxide, or of oxidation intermediates) produced by dehydrogenation of the organic reductant. These two considerations provide a plausible explanation of sigmoid gas exchange curves, but do not fully justify calculation of the maximum quantum yield of photoreduction from the slope of the steepest section of the light curve. In the first place, the utilization of internal reductants is at present merely a hypothesis. In the second place, assuming this hypo- thesis is correct, it is still possible — and, indeed, likely — that, as light in- tensity increases, photoreduction replaces (and not merely supplements) the photochemical transformation of intercellular substrates. This can occur either because of exhaustion of the intracellular material, or because of changes in the enzymatic system (as in the "de-adaptation" of hydrogen- adapted green algae; cf. Vol. I, chapter 6). More precise measurements of the photosynthetic ratio A [CO2]/ A [reductant], at ditTerent light intensi- QUANTUM YIELD OF BACTERIAL AND ALGAL PHOTOREDUCTION 1127 ties, and investigations of the effect of intensity and duration of illumination on the shape of the light curves, could help to elucidate the situation. Un- til there is proof that the sigmoid shape of the light curves actually is due to an internal photochemical process resulting in no (or only little) gas con- sumption; and until it has been proved that this process, having become saturated in very low light, continues at a constant rate as the light grows stronger, the legitimate way to interpret the light curves is the conserva- tive one: to consider the sigmoid shape as evidence that the average quantum yield of photoreduction, y, first increases with light intensity and then decreases again. The measure of j in each point of the curve then is the slope of the straight line drawn from this point to the origin of the co- ordinates, and not the slope of the tangent. (Similarly, we do not attri- bute the convex part of the light curves to a superposition of low-yield photosynthesis upon persisting high-yield photosynthesis, but to a decrease in the average yield.) In this way, we can deduce from the sigmoid light curves only a lower limit of the maximum quantum yield (this limit being given by the slope of the tangent to the curve that passes through the origin of the coordi- nates). This limit, derived from French's light curve of Streptococcus varians, is about jnm. = 0.11. In French's study of Spirillum ruhrum (1937^), the yield was measured by the uptake of carbon dioxide. As mentioned above, the light curves showed, in this case, no initial curvature; their slope corresponded to a quantum yield of the order of 0.06 to 0.Q7. These values were termed "unreliable" by French because of inexact determinations of light absorp- tion. Subsequently, however, yields of similar magnitude were found in several investigations by the Dutch group (Wassink, Katz and co-workers). In their measurements, the initial concavity of the light curves often was only slight, and did not affect essentially the calculated quantum yield. Eymers and Wassink (1938) measured the quantum yield of photosyn- thesis by Thiorhodaceae, with thiosulfate serving as a reductant and a cesium or sodium lamp as light source. The results are shown in Table 29. VIII. One notices that these organisms have a very strong dark meta- Table 29.VIII Quantum Yields of Cakbon Dioxide Reduction by Purple Bacteria WITH Thiosulfate as Reductant (after Eymers and Wassink 1938) Light source Intensity, 10» X erg/cm.2 sec. ACOz in Hglit ACO2 in dark I/7" 7 Cesium (850- 890 ium) Sodium (590 mn) 5.2 to 83 12 to 105 +30 to -121 + 12 to -302 +47 to -1 9.2 to 35 +55 to -4 8.8 to 33'' 0.03to0.10J 0.03 to 0.114 " Quanta/molecule CO2. ^ In old cultures, I/7 values up to 175 were observed. 1128 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 holism, which makes the exact evaluation of the quantum yield difficvilt. The largest 7 values ever observed by Eymers and Wassink were about 0.11. In a subsequent investigation from the same laboratory (Wassink, Katz and Dorrestein 1942), a summary of additional 7 determinations for the same species {Chromatium D) was given, which included values obtained at two different pH values and with hydrogen as well as with thiosulfate as reductant. They are shown in Table 29. IX. The figures are described Table 29. IX Quantum Yields of Chromalium under Different Conditions (after Wassink, Katz and Dorrestein 1942) Quantum yields, \/y Reductant pH Temp., ° C. Determi- nations Limits l/y 7 Thiosulfate. . Thiosulfate . . Hydrogen. . . Hydrogen. . . Hydrogen . . . . 6.3 . 6.3 . 6.3 . 7.6 . 7.6 29° 19° 29° 29° 22° 12 1 7 6 1 8.5-13.8 8.5-16.0 10.6-15.4 10.9 12.8 11.4 12.6 12.1 0.092 J 0 . 079 0.088 0.080 0.083 as having been calculated from ACO2, and (ACO2 -|- AH2) values of the order of 50 or 100 mm.Vhr. According to the figures in text of the paper, this means light intensities of the order of 1000-3000 erg/cm.^ sec. Sapozhnikov (1937) concluded, from his own not further described measui'ements, that the quantum yield of CO2 reduction by Thiorhodaceae is 1.0. Tlie scepticism one is bound to feel about an experimental finding so in variance with all other observations in the field is not reduced by the thermodynamic treatment the author uses to make his results plausible. He argues that the light energy required to reduce carbon dioxide de- pends on the oxidation-reduction potential of the medium in which the reduction takes place. Experimentally, he found this potential (in the medium in which bacteria have lived for a while) to be positive enough for the reduction of carbon dioxide to be possible with only 40 kcal/mole extra free energy. He suggested that, for the same reason, the photosynthesis of green plants could also require only one quantum per molecule of carbon dioxide. Sapozhnikov's argument ignores two facts: (a) that free energy is also needed to establish and maintain the high positive redox potential, which he sug- gests does exist in photosynthesizing cells; and (6) that photosynthesis is not merely re- duction (of carbon dioxide) but also oxidation (of water, or of the other reductants used by bact^eria), and that whatever one can gain in energy required for reduction only means that correspondingly more energy is required for oxidation. Rieke (1949) determined the quantum yield of the green alga Scenedes- mus, which was adapted (by anaerobic incubation) to the use of molecular hydrogen as reductant {cf. Vol. I, chapter 6). He found, both in 4% CO2 and in 0.025 M KIICOs, (luantum yields of between 7 = 0.05 and 0.12, i. 6'., close to those determined for ordinary photosynthesis in the same OXYGEN LIBERATION BY ISOLATED CHLOROPLASTS 1129 species. These yields could be measured at light intensities up to 7000 erg/ cm.- sec; at the higher intensities, transition to ordinary photosyn- thesis occurred too rapidly. 4. Quantum Yield of Oxygen Liberation by Isolated Chloroplasts French and Rabideau (1945) measured the quantum yield of the "Hill reaction" (photochemical oxygen production from ferric oxalate solution, sensitized by a chloroplast suspension). This reaction was described in chapter 4 (Vol. I) as possibly representing ''one half of photosynthesis" — namely, photoxidation of water, with the ferric salt instead of carbon di- oxide serving as oxidant. Chloroplast suspensions were obtained from spinach, or from Tradescantia, by maceration and centrifugation, and ad- ded to 0.5 M K2C2O4 + 0.01 M FeNH4(S04)2 + 0.02 M K3Fe(CN)G. The solution also contained 0.20 M sucrose and 0.17 M sodium sorbitol borate buffer. A 10% NaOH solution was present in a side arm of the manome- tric vessel to absorb carl)on dioxide (which could be produced by respira- tion). Figure 29.7 shows the course of pressure changes in a Warburg ap- LU (T CO cr Q- UJ o X o 100 ^ / 80 / 60 - / -^ / ^ 40 - D -J' Dark Red 1 ght L?^i^ 1 20 - ^^ ^!)— 0— 0— 0 0 ^ 0 — Q_o -i- 1 ^ 1 ! 20 TIME, min. 30 40 Fig. 29.7. Gas liberation in light from suspension of spinach chloroplasts (after French and Rabideau 1945). paratus filled with this mixture. The light used was a red band at 660- 720 mM, with a maximum at 685 m^. The measurements were made at 10° C. Table 29.X gives some typical results obtained with chloroplasts from spinach. Material from Tradescantia gave somewhat lower yields. The quantum yields varied between 0.013 and 0.080, or between 12 and 78 quanta per molecule of liberated oxygen. The average was 7 = 0.042 for chloroplasts from spinach, and 7 = 0.030 for chloroplasts from Trades- 1130 THE LIGHT FACTOK. II. QUANTUM YIELD CHAP. 29 Table 29.X Quantum Yield of Oxygen Evolution by Illuminated Chloroplast Suspensions FROM Spinach (After French and Rabideau 1945) Light Quanta Chloroplast intensity, Fraction Quantum required/02 chlorophyll, micro cal/ of light absorbed yield. molecule. Estimated mg./vessel cm. VHiin- 7 1/7 error, % 0.15 9.0 0.72 0.068 15 10 0.15 3.15 0.71 0.0()4 16 5 0.34 3.25 0.83 0.030 33 20 0.39 3.3 0.88 0.022 46 20 0.10 3.32 0.57 0.080 12 5 0.10 3.04 0.51 0.037 27 20 0.0G3 4.13 0.56 0.033 31 20 0.103 4.18 0.545 0.013 78 20 0.09 1.6 0.33 0.058 17 20 0.045 2.8 0.47 0.031 38 5 0.090 2.8 0.79 0.03G 28.0 5 cantia (in the same set-up, a value of 7 = 0.092 was found for live Chlorella). No clear change of 7 with light intensity was noticeable in the range used (1400-6000 erg/cm. 2 sec, of which 33-72% was absorbed by the suspen- sion). These quantum yields of the Hill reaction, although markedly lower than the quantum yield of photosynthe-sis in vivo, were nearer the latter than the (much higher) yields of the chlorophyll-sensitized photoxidations in vivo {cf. Vol. I, page 513, and chapter 35). The considerable variability of the 7 values of the Hill reaction may be caused, at least in part, by rapid deterioration of the material. Several runs, made in succession with one batch, usually showed rapidly declining yields. New measurements of the quantum yield of the Hill reaction, in whole Chlorella cells and in chloroplasts from Phytolacca americana, were carried out by Ehrmantraut (1951). They were made relative to the ethyl chloro- phyllide-thiorea actinometer; two measurements of the quantum require- ment of photosynthesis in carbonate buffer served as an additional, if rough, check on the actinometer, since there is general agreement that this requirement is I/70 = 10 =±= 2. The results (table 29. XI) are much more consistent than those of French and Rabideau. They were further con- firmed by five quantum yield measurements with quinone in whole Chlo- rella cells using a monochromator (X 669 m^t) and a bolometer which gave, for 1/7, the values: 10.2; 9.9; 9.3; 10.8; and 12.0 (average: 10.4). These values, obtained in acid medium (pH 6.5), throw indirect light on the problem of the quantum yield of photosynthesis. Various kinetic evidence points to the Hill reaction having both the primary photochemical process and the rate-limiting dark reaction, in common with photosynthe- MAXIMUM QUANTUM YIELD IN RELATION TO LIGHT CURVES 1131 Table 29. XI Quantum Requirement of the Hill Reaction (after Ehrmantraut and Rabino- WITCH 1951) (Incident Light Flux 1.5 X 10"' Einstein per Min.: t = 10°C.; 200 mL Cells) Material Oxidant Quantum requirement Chlorella Chlorella Chloroplasts (from Phytolacca americana) Same Same Chlorella 0.5 mg. quinone in 3 cc. 12.4 12.6 12.6 13.1 13.1 12.8 Av. 12.8 1.0 mg. quinone in 3 cc. 13.5 13.4 14.8 13.3 Av. 13.8 1.0 mg. quinone in 3 cc. 9.3 10.2 11.5 11.4 10.6 Av. 10.6 Hill's solution 12.7 Ferricj^anide 11.0 Carbon dioxide (No. 9 11.3 carbonate buffer) 13.0 Av. 12.2 sis. It would be remarkable, under these conditions, if equality of the quantum requirement of the Hill reaction, in whole cells as well as in chloro- plast fragments, with the quantum requirement of photosynthesis in alka- line buffers, were to turn out purely coincidental. It seems unlikely that quantitatively the same damage (as measured by a supposedly "sub- standard" quantum yield) would have been inflicted on the photochemical apparatus by such diverse treatments as immersing cells into alkaline medium, poisoning them with quinone, and smashing them mechanically and separating chloroplasts fragments. A much more plasusible hypothesis is that the photochemical apparatus survives all these treatments without severe damage, and that the quantum requirement of 10 ± 2 represents the tiTie measure of the efficiency of the common primary photochemical process. The quantum requirements of A(C02)2, or Chi + 2 hv -^ Chi**] are consid- ered, and not more than two successive reaction steps are postulated be- tween the external supply of the reactant and the "rate-determining" step. (For example, in the case of the carbon dioxide factor, simultaneous con- sideration of diffusion and carboxylation leads to a hyperbolic carbon dioxide curve; but if one more supply step is interpolated, the resulting equation is of the third order.) The known light curves are much too unreliable to permit a useful in- quiry into the question whether they actually are hyperbolae {cf. section 7/, chapter 28). If it were possible to demonstrate, by new and more pre- cise measurements, that the light curves are hyperbolic, then each curve could be determined completely by three points, i. e., values of P at three known values of /. Parameters such as 70 (i- e., the initial slope), .// or P'"''^ could replace one measurement each. The general equation of a hyperbolic light curve in terms of 70, i/J and P'^^^- is : P /270 1/2/ 2_\ P^ ^ jn^ J (2*J-5) pmax. _ p "I Wpmax.)2 pm&x.J pmax. _ p pmax (This equation is obtained from the general equation of a hyperbola by transformation to a set of coordinates with origin in a point on the hyperbola and the abscissa parallel to the asymptote at a distance — p™'^''- from the latter.) In treating several particularly simple mechanisms in chapters 27 and 28, we obtained light curves (or carbon dioxide curves) that obeyed an even simpler relation: (2i).(i) p/(^pmu.. _ p) = const. X / or: P/(Pu.ax. - P) = const. X [CO.] Comparison shows that this simplification means the disappearance of the P- term in (29.5), i. e., the validity of the relation: (29.6a) 70 = P^^^/^/^I The relationship provides a simple way to check whether the kinetic mechanism is of corresponding simplicity. MAXIMUM QUANTUM YIELD IN RELATION TO LIGHT CURVES 1135 Whenever this is the case, and equation (29.6) is obeyed, differentiation with respect to / shows that the quantum requirement, I/7, is a linear function of light intensity, with i/p™'"^"- as the slope : (29.7) 1/7 = 1/70 + (I/P^-) The question arises whether the experimental value of I/7 is a linear function of the intensity of irradiation and, if so, whether the slope of the corresponding straight line is equal to the inverse of the saturation value of photosynthesis in strong light, 1/P"'^^- 7 - Emerson and Lewis (1941) Warburg (1948) 02 0.4 06 0.8 /, einstein/mole Chi x mm. 1.0 Fig. 29.8. Quantum requirement of Chiorella as function of light intensity. Figure 29.8 represents an attempt to apply equation (29.7) to the quantum yield data of Emerson and Lewis (1943) and of Warburg (1948). Warburg's values scatter too widely to decide whether they lie on a straight line or not; but, if this is assumed to be the case, the slope of the straight line is much higher than in the case of Emerson's measurements (indicating a much more rapid decline of quantum yield with the intensity of illumina- 1136 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 tion). The abscissae in figure 29.8 are frequencies of absorption of light ciuanta by each chlorophyll molecule (total number of (juanta absorbed by the suspension in unit time, divided by the number of chlorophyll molecules present). The slope of the Warburg curve, equated with 1/P'"'"'", indi- cates a maximum production of one oxygen molecule per chlorophyll mole- cule in about 10 minutes; the slope of the Emerson-Lewis curve, the pro- duction of one oxygen molecule per chlorophyll molecule in about 1 minute. Table 28. ^' shows that the actual maximum yield in Chlorella in steady light is of the ortler of one molecule oxj^gen per molecule chlorophyll in 30 seconds. A rapitl decline of the apparent quantum yield with / is consistent with the assumption that Warburg's values were affected by the inclusion of the carbon dioxide gush, since the relative importance of this gush must decrease rapidly with increasing light intensity (as was suggested on page 1103). Franck (1949) suggested that the rapid drop of I/70 in Warburg's experiments with increasing light intensity, is an indication that they re- flect a gradual repla(;ement of a 4-quanta process (half-way reversion of respiration), by an 8-quanta process (true photosynthesis). On page 1 104, we described the more recent experiments of Warburg, Burk and co-workers, who claimed that the highest quantum yields are obtainable by intermittent illumination (which prevails in very rapidly agitated, dense Chlorella suspensions). If this were true (it was men- tioned on page 1106, that the results of experiments in flashing hght do not support the contention of Warburg and Burk), then the relation.ship be- tween quantum yield and the (average) light intensity must be more com- plicated than was envisaged in the above derivations. (b) Quantum Yields in Strong Light It was suggested above that the yields of photosynthesis given for high light intensities should not contradict the results obtained in quantum yield measurements in weak light. What we meant can be illustrated by the following examples taken from the work of Willstatter and Stoll (1918). According to figure 32.2, green leaves of Sambucus nigra reduce carbon dioxide, at 6000 lux, at a rate of about 0.23 mg./cm.^ hr. or 1.44 X 10-'' mole/cm. 2 sec. Since 1 lux corresponds roughly to a flux of 5 erg/cm. ^ sec. in the region 400-700 mix {cf. chapter 25, page 838), and about 80% of this flux is a])sorbed by a single leaf, we can calculate, for the energy conversion factor : (1.44 X 10-' X 112 X 10^) cal ((■) X 103 X 5 X 0.24 X 10-' X 0.8) cal 0.28 THEORETICAL AND ACTUAL MAXIMUM QUANTUM YIELD 1137 which corresponds to a quantum yield of about 0.13 (assuming 550 m/x as average wave length) . In j&g. 32.2, the yellow leaves of Samhucus are shown to reduce, at 3000 lux, 0.067 mg. COj/cm.^ hr., or 4 X lO-^" mole/cm.^ sec. These leaves contain less than one tenth the chlorophyll present in green leaves of the same species. Measurements such as those represented in figure 22.10 indicate that the aurea leaves absorb, in the region above 500 m/i, not more than 20% of the incident energy. (Blue and violet light do not contribute much to photosynthesis in artificial light; cf. page 1163.) Thus, the energy conversion factor of the yellow leaves, can be estimated as: - ~ (4 X IQ-^" X 112 X 10^) cal = o 31 * ~ (3 X 103 X 5 X 0.24 X 10"' X 0.4) cal. corresponding to a quantum yield of about 0.14. Similar difiiculties arise in the interpretation of some of the maximum yields of photosynthesis listed in Table 28.V (0.8 or 0.9 mg. COa/cm.^ hr.). As far as can be judged from the known light curves of land plants, it seems safe to presume that saturation yields can be obtained in light of the order of 40,000 lux. A yield of 0.9 mg. C02/cm.2 hr. at 40,000 lux means an av- erage quantum yield of the order of 0.1, obtained in a region of almost com- plete light saturation! These estimates contain too many approximations to be used as quanti- tative arguments against the upper limit 0.10 ± 0.02 for the quantum yield of photosynthesis; but they show that it would be well to extend future investigations of the quantum yield to the leaves of higher plants— partic- ularly those of the aurea varieties— and to cover the entire length of the light curves. 6. Theoretical and Actual Maximum Quantum Yield All kinetic theories of photosynthesis agree that the (approximately) linear lower part of the light curves corresponds to the state in which the primary photochemical process is so slow that the nonphotochemical reac- tions—the "preparatory" as well as the "finishing" ones— can supply the materials and transform the products of this process without delay. It may thus seem as if the maximum quantum yield, calculated from the limiting slope of the light ciu-ves, should be equal to the number of quanta actually needed for photosynthesis (except for the practically negligible fraction lost by fluorescence). The fact that the experimentally determined maximum cjuantum yields often are much lower than 0.1 shows that, in many cases, the photosynthetic apparatus, or parts of it, are in a noneffi- cient state, so as to cause the loss of tiu; prodomiiinnt fraction of all the 1138 THE LIGHT FACTOR. II. QUANTUM YIELD CHAP. 29 absorbed light quanta. In "aged" cell suspensions, in particular, the maximum quantum yield can be much smaller than in healthy young cells (cf. fig. 28.13). The reasons for this inactive state are as yet unknown and may lie in nutritional or enzymatic deficiencies (we recall, for example, van Hille's experiments on the revival of photosynthesis in aged Chlorella cultures by a fresh supply of N2) or in the obstruction of catalytic surfaces by narcotizing metabolites ("chlorellin"; cf. page 880). Under the action of external narcotics (cf. fig. 28.9C) the initial slope of the light curves is clearly depressed, i. e., no full quantum yield can be obtained even in ex- tremely weak light. This must be attributed to the inactivation of a cer- tain proportion of chlorophyll complexes, probably by adsorption of the narcotic ; the light quanta absorbed by these complexes remain unavailable for photosynthesis. An interesting question— not yet investigated experimentall}^ — is whether a "substandard" quantum yield can be corrected, at least par- tially, by an increase in temperature. If the low yield is caused by some nonphotochemical process that has become so slow as to depress the rate of photosynthesis even in very weak light, heating should accelerate this process and thus improve the yield; but, if the inefficiency is caused by the fact that a certain proportion of the photosensitive chlorophyll com- plexes are inactive (partially decomposed, or obstructed by adsorption), heating might have no effect on the yield. Another interesting question is whether, even in the case when all photosensitive catalytic complexes are fully efiicient, the experimental maximum quantum yield must correspond exactly to the number of quanta actually used in the reduction of carbon dioxide. This question is raised by consideration of schemes 28.1A,B and 28.11, all of which envisage a competition between stabilizing "forward" reactions and primary or secondary back reactions. Examples of primary back reac- tions are (28.20a', 28.21a' and 28.41a') ; examples of secondary back reac- tions are (28.20d and 28.21d). In scheme 28. lA, the "rate-determining" reaction is (28.20b), and the rate is given by equation (28.23). The quantum yield is: (29.9) P/Ia = P/k*I Chlo = 7ikr[AC02]/{k' + A-JACOj]) The maximum quantum yield (reached when [ACO2] = Ao) is: (29.10) 70 = kr\nn/ik' + AvAo) In scheme 28.11, the rate is described by equation (28.42), and the maximum quantum yield is given by equation (28.43) : THEORETICAL AND ACTUAL MAXIMUM QUANTUM YIELD 1139 (29.11) 70 = dP/d(k*Ch]oI) = keEln/ik' + A;.E°) In (29.9), the "forward" reaction of HX-Chl-Z with ACO2 competes with the primary back reaction (conversion to X- Chi -HZ); in (29.11) the stabiUzing "forward" reaction of AHC02-Chl- A'HO with the catalyst Eb competes with the back reaction (conversion to AC02-Chl-A'H20). In either case, the "theoretical" quantum yield, n, can be closel}^ ap- proached only if the ratio A-'/AvA) (or k'/keE^a) is much smaller than 1. Recognition that the maximum observable quantimi yield may be smaller than the theoretical quantum yield, n, was first reached in the derivations of Franck and Herzfeld (1941) (c/. their Table 1). The reaction mecha- nism used by them, although more complicated than the ones considered here, also was based on competition between stabilizing forward reaction and back reactions of the immediate reduction products. The analytical expressions may be more complicated, but the essential result also remains the same if the effective back reaction is a secondary one, such as (28.21d). If this reaction competes with the forward trans- formation of the intermediate reduction product AHCO3 by the catalyst Eb (or of the intermediate oxidation product A'HO by the catalyst Ec), the fraction of the products that undergoes back reaction will remain finite even when the forward reactions have the maximum possible rate con- stants, i. e., when the full available amounts of the relevant catalysts are free and can be utilized for transformation (as is the case in weak light). Bibliography to Chapter 29 The Light Factor. II. Maximum Quantum Yield of Photosynthesis 1918 Willstatter, R., and Stoll, A., Untersuchungen iiber die Assimilation der Kohlensdure. Springer, Berlin, 1918. 1922 Warburg, 0., and Negelein, E., Z. phijsik. Chem., 102, 235. 1923 Warburg, 0., and Negelein, E., ibid., 106, 191. Wurmser, R., Cornpt. rend., 181, 644. 1925 Wurmser, R., AnJi. physiol. physicochimie bioL, 1, 47. 1926 Wurmser, R., /. phys. radium, [II] 7, 33. 1929 Briggs, G. E., Proc. Roy. Soc. London, BIOS, 1. 1935 Gabrielsen, E. K., Planta, 23, 474. 1937 French, C. S.,/. (?m.P%s{o/., 20, 711; 21,71. Sapozhnikov, D. I., Biokhimiya, 2, 18. 1938 Emerson, R., and Lewis, C. M., Carnegie Inst. Yearbook, 37, 216. Manning, W. M., Stauffer, J. F., Duggar, B. M., and Daniels, F., /. Am. Chem. Soc., 60, 266. 1140 THE LIGHT FACTOR. TI CHAP. 29 Manning, W. M., Juday, C, and Wolf, M., ibid., 60, 274. Eymers, J. G., and Wassink, E. C, Enzymologia, 2, 258. Wassink, E. C, Vermeulen, D., Reman, G. H., and Katz, E., ibid., 5, 100. 1939 Emerson, R., and Lewis, C. M., Carnegie Inst. Yearbook, 38, 118. Emerson, R., and Lewis, C. M., Am. J. Botany, 26, 808. Rieke, F. F., J. Chem. Phijs., 7, 238. Eichhoff, H. J., Biochem. Z., 303, 112. ^ Petering, H. G., Duggar, B. M., and Daniels, F., J. Ani. Chem. Soc., 61 /'^ 3525. Magee, J. L., De\\'itt, T. W'., Smith, E. C, and Daniels, F., ibid., 61, 3529. 1940 Emerson, R., and Lewis, C. M., Carnegie Inst. Yearbook, 39, 154. 1941 Emerson, R., and Lewis, C. M., Am. J. Botany, 28, 789. Emerson, R., and Lewis, C. M., Carnegie hist. Yearbook, 40, 157. Franck, J,, Sigma Xi Quart., 29, 80. Franck, J., and Herzfeld, K. F., J. Phys. Chem., 45, 978. Franck, J., and Gaffron, IL, Advances in Enzymology, 1, 199. Button, H. J., and Manning, W. M., Am. J. Botany, 28, 516. 1942 Wassink, E. C., Katz, E., and Dorrestein, R., Enzymologia, 10, 269. Emerson, R., and Lewis, C. M., J. Gen. Physiol., 25, 579. Franck, J., Am. J. Botany, 29, 314. 1943 Emerson, R., and Lewis, C. M., ibid., 30, 165. 1944 Wassink, E. C., and Kersten, J. A. H., Enzymologia, 11, 282. Tonnelat, J., Compt. rend., 218, 430. 1945 French, C. S., and Rabideau, G. S., /. Gen. Physiol, 28, 329. 1946 Tonnelat, J., Thesis, Univ. of Paris, Series A, No. 2092, serial No. 2959. Wassink, E. C, Enzymologia, 12, 33. Warburg, 0., and Luettgens, W., Biokhimija, 11, 303. 1947 Rabinowitch, E., A.A.A.S. Symposium on Photosynthesis, Chicago, Dec. 1947 (unpublished). Gabrielsen, E. K., Experieniia, 3, 439. 1948 Warburg, 0., Am. J. Botany, 35, 194. Kok, B., Enzymologia, 13, 1. 1949 Rieke, F. F., in Photosynthesis in Plants, Tlie Iowa State College Press, Ames, la., 1949, pp. 251-273. Moore, W. E., and Duggar, B. M., ibid., pp. 238-250. Emerson, R., and Nishimura, M. S., ibid., pp. 219-239. Arnold, W., ibid., pp. 273-276. Franck, J., Arch. Biochem., 23, 297. Kok, B., Biochim. biophys. Acta., 3, 625. Van der Veen, R., Physiol, plantaruni, 2, 217. Burk, D., Hendricks, S., Korzenovsky, M., Schocken, V., and Warbui'g, O., Science, 110, 225. Warburg, 0., Burk, D., Schocken, V., Korzenovsky, M., and Hendricks, S., Arch. Biochem., 23, 330. BIBLIOGRAPHY TO CHAPTER 29 1141 1950 \\'arl)urg, 0., Burk, D., Schocken, V., and Hendricks, S., Biochcm. Bio- phys. Acta, 4, 335. ^^'al•l)Ul•g, 0. and Burk, D., Arch. Biochem., 25, 410. Whittingham, C. P., Nishimura, M. S., and Emerson, li., Proc. Soc. Exptl. Biol, (in press). Brown, A. H., and co-workers (Columbus meeting of the Soc. Plant Physiol., October, 1950). 1951 Tanada, T., Am. J. Botan., 38, 276. Ehrmantraut, H. C. and Rabinowitch, E., Arch. Biochem. (in press). Warburg, 0. and Burk, D., Naturwiss., 37, 560. Warburg, 0., and Burk, D., Z. Naturforschung., 6b, 12. Chapter 30 THE LIGHT FACTOR. HI. PHOTOSYNTHESIS AND LIGHT QUALITY; ROLE OF ACCESSORY PIGMENTS* 1. Action Spectrum The dependence of photosynthesis on the spectral quality of Ught was the subject of much interest, long before the influence of light quantity was first investigated. As early as 1788, Senebier conducted experiments on carbon dioxide assimilation in double-walled vessels, filling the space between the walls with various colored solutions. Since then, the botanical literature has been replete with observations on the behavior of plants in light of different color. (For a review of these investigations, see, for ex- ample, Gabrielsen 1940.) This was to be expected, since everything con- nected with color has always held, and still holds a captivating interest for mankind, even though in scientific photochemistry the qualitative cate- gories of "red," "blue," "yellow" or "green," which were so dear to Goethe (he thought them to be the "primary phenomena" of optics), have been reduced to mere quantitative differences between the energy contents of the light quanta. About a hundred years ago the effect of color on photosynthesis be- came a topic of a lively discussion. Its subject was the position of the maximum of photo synthetic efficiency in the solar spectrum. In 1844, Draper found that, when the prismatic spectrum of the sun was thrown upon a plant, the largest amount of oxygen was liberated in the yellow-green re- gion ; this result was confirmed by such authorities in plant physiology as Sachs (1864) and Pfeffer (1871). Sachs pointed out that the yellow is also the region of maximum "luminosity" of light, i. e., of maximum effect on the human retina. He himself saw in this only a coincidence; but other, less cautious authors thought that such a correspondence must be sig- nificant, and attempted to explain it. The belief that photosynthesis pro- ceeds most actively in green light, which is only weakly absorbed by chloro- phyll, led to several peculiar hypotheses, such as that light energy is not used in photosynthesis at all (Pfeffer 1871), or that the role of chlorophyll in plants is merely to protect the carbon dioxide-reducing system from injury by light (Pringsheim 1879, 1881, 1882). Timiriazev (1869, 1875, 1877, 1885) vigorously fought these miscon- * Bibliography, page 1 188. 1142 INTRODUCTION 1143 ceptions. He pointed out that "luminosity" is an anthropomorphic no- tion, without meaning in objective photometry, that utihzation of light energy is the essence of photosynthesis, that this utilization cannot take place unless light is absorbed by a sensitizing pigment, and that this pig- ment cannot be anything but chlorophyll. Timiriazev was the first to use the concept of sensitization, a phenomenon then recently discovered by Vogel and Becquerel, in the discussion of photosynthesis. A similar point of view was taken by several physicists, e. g., Jamin, Becquerel and, particularly, Lommel (1871, 1872). The latter pointed out that the basic principle of photochemistry, known as Herschel's law ("no photochemical action without light absorption"), requires that the spec- tral maximum of photosynthetic efficiency coincide with the absorption maximum of the sensitizing pigment. Timiriazev (1869, 1875), Miiller (1872), Engelmann (1882) and Reinke (1884) gave experimental proofs of this coincidence, by showing that the photosynthetic efficiency of green plants decreases steadily from red through yellow to green, parallel with the decline in absorbing capacity of chlorophyll. The error of Draper, Sachs and PfefYer was attributed by Timiriazev to the use of spectrally impure light. (Timiriazev himself employed light isolated by a mono- chromator with a narrow slit, and used microanalytical methods to com- pensate for the weakness of illumination.) Engelmann suggested that the error may have resulted from the use of thick leaves or thalli, which ab- sorb light practically completely even in the minima between the absorp- tion bands of chlorophyll. (He worked with microscopic plant objects, using motile bacteria for the detection and determination of oxygen.) Engelmann (1882) noticed that, in addition to the main maximum in the red, the photosynthetic "action spectrum" of green plants has a second maximum in the blue or violet, which he associated with the strong ab- sorption band of chlorophyll in this region. This perfectly natural conclu- sion became the subject of one of the most vitriohc controversies in the history of photosynthesis; it was contested even by such enlightened plant physiologists as Reinke (1884) and Timiriazev (1885). Particularly violent were the criticisms Pringsheim (1886) directed against Engelmann's method and his results; and Engelmann (1887) answered these attacks in language seldom encountered in the pages of scientific journals, even in the quarrel- some nineteenth century. At the same time, Engelmann also sharply rebuked Timiriazev for his attempt (1885) to identify the main maximum of spectroscopic efficiency of photosynthesis with the energy maximum of the solar spectrum. Timiriazev saw in this alleged coincidence a striking example of adaptation of organisms to the prevailing conditions, and thus a triumph of the Darwinian theory. Engelmann answered that a coin- cidence of the two maxima cannot be postulated without deliberately 1144 THE LIGHT FACTOR. III. COLOR CHAP. 30 twisting experimental evidence : Direct sunlight has an energy maximum in the yellow and not in the red, as the action spectrum of photosynthesis; and the energy maximum of diffuse sky light, which is the second com- ponent of the natural "light field" of land plants, lies even further toward the blue-violet end of the spectrum. In this controversy, too, Engelmann was undoubtedly right. Looking backward, one cannot but admire the unfailing correctness of his conclusions, obtained by means of an experi- mental method most investigators would hesitate to use even for qualita- tive, not to speak of quantitative, purposes. l<]ngelmann not only estab- lished correctly the general parallelism between the action spectrum of photosynthesis and the absorption spectrum of chlorophyll ; he also clearly understood the influence on both spectra of the optical density of the specimen. His conclusions concerning the photosynthetic efficiency of the carotenoids and phycobilins, long neglected, appear well on the way to vindication sixty five years later. Much of the controversy between Engelmann and his opponents can be attributed to a lack of understanding of what is meant by "action spec- trum." A primitive definition can be based on the above-mentioned simple experiment of Draper : A spectrum is thrown on a plant or cell sus- pension, and photosynthesis is measured in different spectral bands of equal width. Such an experiment, performed with a prism in artificial light, may easily lead to the belief that the action spectrum of photosyn- thesis has only one maximum, because the energy of most artificial light sources declines rapidly toward the violet end of the spectrum, while the dispersion of the prism increases in the same direction. Both factors co- operate in causing a rapid decline of the yield of photosynthesis (related to a given spectral band width) as one proceeds from the red to the violet ; this decline may more than offset any increase caused by the renewed rise in the absorption capacity of chlorophyll in the blue-violet region. Obviously, this definition of the "action spectrum" is arbitrary and ir- relevant. An improved definition can be obtained by using Ught fluxes of equal intensity at all wave lengths, which can be achieved, e. g., by ap- propriate variation of the width of the monochromator slit, or insertion of neutral gray filters. By using such methods. Kohl (1897, 1906), von Rich- ter (1902) and Kniep and Minder (1909) were able to confirm Engelmann's finding of the existence of a second maximum in the action spectrum of photosynthesis, at the short-wave end of the visible region. However, even an action spectrum obtained by the use of spectral bands of equal energy is not universal, i. e., it cannot claim validity for all plants, not even for all specimens of a given species (e. g., all Chlorella sus- pensions). The first cause of variability of "equienergetic" action spectra is the varying composition of the pigment system (c/. chapter 15); but INTRODUCTION 1145 even for plants with identical contents of all pigments (or suspensions of identical cells) the action spectrum still depends on two individual factors. The importance of one of them — the optical density of the sample — was recognized by Engelmann : In a thick leaf or thallus, or a dense cell sus- pension, the absorption spectrum and the action spectrum both are blurred; in the limiting case of complete absorption (approximately realized in Warburg and Negelein's quantum yield experiments; cf. chapter 25, page 844), the action spectrum, too, may lose all structure. The second factor that affects the action spectrum of an individual plant or cell suspension (without affecting its absorption spectrum) is the in- tensity of illumination. If one would use monochromatic light of such high intensity as to obtain full light saturation at all wave lengths, the rate of photosynthesis (which, in the hght-saturated state, is determined only by the velocity of a dark process) will become identical in all parts of the spectrum. The structure of the action spectrum will appear as soon as the intensity of illumination is reduced below the saturating value. Since the saturating intensity depends on wave length {cf. sect. 4), the shape of the action spectrum will change ^vith decreasing light intensity, until the latter will fall within the practically linear range for all wave lengths. The initial divergence and ultimate convergence of light curves in light of dif- ferent color is illustrated by figure 30.8B obtained by Gabrielsen (1940) with green leaves of Sinapis alba. In the low intensity range, the shape of the action spectrum becomes constant, and the spectrum thus acquires a definite significance. Here — and only here— can the action spectrum be compared quantitatively with the absorption spectrum of the specimen, and the question asked whether the specific photochemical efficiency depends on wave length. First, however, a "quantum correction" must be applied. According to Einstein's law of photochemical equivalency, one has reason to expect equal numbers of absorbed quanta of different wave length to produce the same photochemical effect, but not equal quantities of absorbed energy. Thus, action spectra have to be "quantized," i.e. expressed in moles per einstein, rather than in moles per erg or calorie. In the linear region, it is legitimate to convert the "equienergetic" action spectrum into the "quan- tized" spectrum simply by dividing all ordinates by the corresponding wave lengths. In the saturation range, this is impossible, and the quantized action spectrum can be obtained only by direct experiment (i. e., by measur- ing the rate in bands of equal intensity expressed in einsteins per square centimeter per second), because here the shape of the action spectrum de- pends on intensity, and spectral bands that have equal intensity if meas- ured in energy units have different intensities if measured in einsteins. If the maximum quantum yield of photosynthesis is the same for all 1146 THE LIGHT FACTOR. III. COLOR CHAP. 30 wave lengths, the quantized action spectrum can be expected to parallel exactly the absorption spectrum; the "equienergetic" action spectrum, on the other hand, will always be askew, with blue-violet light appearing ceteris paribus less efficient than red light. If the quantized action spectrum determined from measurements in low light, differs markedly from the absorption spectrum, this is a definite indication that quanta of different wave length have different photochemi- cal effects in photosynthesis. Recent determinations of the quantum yield of green and colored algae in monochromatic light, carried out by Emerson et al., Button and Manning, and Blinks, have established the existence of such differences, and made speculations as to their origin legitimate. Similar conclusions have been drawn previously from experi- ments under badly controlled conditions, in which broad spectral bands (isolated by means of colored glass filters) and unknown light intensities were used. Conclusions drawn from experiments of this type, e. g., by Montfort, have sometimes proved partially correct, but comparison of Montfort's confused discussions with the concise presentation of Emerson and Lewis, and Button and Manning gives a most eloquent demonstration of the progress that can be achieved in plant physiology by the use of better physicochemical tools. The explanation of the effect of wave length on the maximum quantum yield of photosynthesis can be sought in three phenomena: (a) in the composite nature of the pigment system and the (qualitatively or quantita- tively) different photochemical functions of the individual pigments; (h) in the multiplicity of the excited electronic states of chlorophyll, two (or three) of which are involved in the light absorption in the visible spectrum {cf. fig. 21.20) ; and (c) in the influence that vibrational energy, acquired by the sensitizer together with electronic excitation, may have on its sensitizing action. The first factor causes the rate of photosynthesis in different spec- tral regions to be affected by the apportionment of the absorbed light energy to the several pigments, while the second and third factors can cause changes in efficiency with wave length, even in light absorbed by a single pigment. Thus, the study of the effect of wave length on photosynthesis should aim, first, at the qualitative and quantitative determination of the role of various pigments in sensitization and, second, at the analysis of the rela- tion between wave length and photochemical efficiency for each pigment. Needless to say, we are far from having achieved these aims. Even now, the greater part of newly pubUshed work on photosynthesis in colored light remains purely descriptive and unsuitable for quantitative interpreta- tion. ACTION SPECTRUM OF GREEN PLANTS 1147 In addition to wave length, the quality of light is characterized by its polarization. Almost universally, no attention has been paid to this characteristic in the study of the relationship between light and photosynthesis. Dastur and Asana (1932) and Johnson (1937) found no difference in the rate of photosynthesis in ordinary and linearly polar- ized light of the same intensity; but Dastur and Gunjikar (1934) observed a marked deficiency in the absorption by leaves of elliptically polarized light, and later (1935) also a similar deficiency in the synthesis of carbohydrates. Although we may doubt the correctness of these results, it must be borne in mind that birefringence of the chlo- roplasts {cf. Vol. I, p. 362) and the correlated dichroism may conceivably lead to dif- ferences in the capacity to utilize light of a different state of polarization. However, only a very weak dichroism has been observed in the chloroplasts in the natural state (cf. Vol. I, p. 366). 2. Quantum Yield and Wave Length in Green Plants. Role of Carotenoids In chapter 29, we discussed the maximum quantum yield of photosyn- thesis in green plants, without paying much attention to the quality of light used for its determination, in the tacit assumption that it is either altogether immaterial or has only a secondary influence. Quantum yield measurements -with Chlorella in monochromatic light were first carried out by Warburg and Negelein (1923). The results are shown in Table 30.1. The quantum yield in blue light is somewhat higher than in red light, if referred to the (estimated) absorption by chlorophyll alone, and somewhat lower, if referred to the total absorption by all pigments. If we assume that the relative yields of Warburg and Negelein in dif- ferent spectral regions are significant (even though their absolute values had been questioned by Emerson et al. ; cf. chapter 29), the figures in Table 30.1 point to a participation of the carotenoids as sensitizers in photosynthe- sis, but with an efficiency inferior to that of the chlorophylls. Table 30.1 Quantum Yields (after Warburg and Negelein 1923) Light X, m/i y e Red 610-690 0.23 0.59 Yellow 578 0.23 0.54 Green 546 0.21 0.44 Blue 436 0.20" 0.34" 0.28* 0.48'' " Referred to all pigments. ' Referred to chlorophyll alone. When Warburg repeated his quantum yields measurements 25 years later (Warburg 1946, 1948), he again found a somewhat lower ;yield (7 = 0.20 and 0.16 in two experiments) in blue light (X = 436 mju) as compared with yellow light. 1148 THE LIGHT FACTOR. ITT. TOLOR CHAP. 30 Briggs (1929) made yield determinations in light of three colors and obtained the results listed in Table 30. lA. Table 30.IA Photosynthesis Yield of Leaves in Colored Light (after Briggs 1929) Yield, ml. O2/5OO cal incident light Species Yellow light Green light Blue light Phaseolus vulgaris 14-17 9-11 7-8 Uhnus, yellow 8.8 6.5 5.3 green — 20 12 Sambucus nigra 8.7 9.3 8.7 Sambiicus nigra — 19.0 15.0 These results obtained at light intensities 5-10 times stronger than those used by Warburg and Negelein, show the expected decline in the energy conversion jaeld with decreasing wave length (the experiment with Samhucus is an exception). In other words, the "equienergetic" action spectrum is askew as expected. The decline from yellow to blue is, how- ever, somewhat stronger than could be explained by the quantum correc- tion (the ratio of the yields in yellow and blue is 1.8 to 2.0, instead of 1.4), again indicating a somewhat lower quantum yield in the region of absorption by carotenoids. Wurmser (1925) found, for Ulva lactuca, a much higher quantum yield in the green than in the red, and a comparatively low yield in the blue; but these results cannot be considered reliable {cf. page 1118). Gabrielsen's data for Sinapis alba (1935), summarized in Table 30. Ill, (p. 1162) are more significant. The relation between the quantum yields in the red and blue is similar to that found by Warburg and Negelein, and thus allows a similar interpretation. The value in the green — which, in contrast to Wurmser's result, is lower than in the red — may perhaps be taken as an indication that the yellow-green filter used by Gabrielsen transmitted much light absorbed by the carotenoids. The only extensive investigation of the quantum yield of photosynthe- sis in a green plant as a function of wave length was carried out by Emerson and Lewis (1941, 1943) with Chlorella pijrenoidosa. They used bands from 5 to 15 m/x wide, obtained by means of a powerful monochromator. Figure 30.1 shows the results. The scattering of the points is indicative of limitations to which the biological "standardization" of cell cultures is sub- ject. Despite this scattering, it appears certain that the yield is approxi- mately constant between 580 and 685 m/x. (The authors believe that the shallow minimum at 660 m/x is real ; for its suggested interpretation, see page 1155.) Below 580 mju, the yield declines considerably, reaches a minimum at 490 mju and then recovers. Roughly, the depression of the quantum yield ACTION RPECTRTTM OF GREEN TTiANTS 1140 curve in the green, blue and violet covers the regions where carotene, luteol nnd other carotenoids con1ril)ute markedl}^ to the Hght al)sorption by ChJorella. An attempt at a quantitative interpretation of the 7 curve meets with some difficulties. Figure 22.44 shows that, if the absorption by the carotenoids is calculated on the basis of extract spectra (by shifting all bands to make their maxima coincide with the absorption peaks of live cells), significant participation of carotenoids in light absorption cannot be GIG 009 008 Q 0.07 z < a G.06 G.G5 004 400 440 480 520 560 600 WAVE LENGTH, m/i 640 680 720 Fig. 30.1. Quantum 3-iold of photo.syii thesis as a function of wave length for Chlorella (after Emerson and Lewis 1943). Points ol)tained on 19 runs are indi- cated by distinct symbols. Band half widths used indicated by horizontal lines of corresponding length. expected above 540 mju. Thus the decline in y, which first begins at 580 mn, cannot be attributed to the carotenoids, unless one assumes that, in the living cell, their absorption bands are not merely shifted, but so strongly broadened as to extend up to 580 m^i. The minimum in the quantum yield curve, at 490 m/x, corresponds satisfactorily to the maximum of c aro- tenoid absorption, according to fig. 22.44 (more exactly, to the center of gravity of the two carotenoid peaks); but one notices that in this region the carotenoids can be expected to account for 60 or 70% of the total al)- sorption; while the depression of the y curve does not exceed 25%. It thus again appears that, even though light quanta absorbed by the caro- tenoids are less efficient than those absorbed by chlorophyll, they are not entirely inefficient. 1150 THE LIGHT FACTOR. III. COLOR CHAP. 30 One may argue that, if the absorption band of the carotenoids in vivo is strongly broadened (as suggested above), the maximum of this band may be much lower than in vitro (since the area of the band — which is pro- portional to the probability of the corresponding electronic transition — is not hkely to be much affected by the state of the pigment) . Consequently the proportion of light absorbed by carotenoids in the region of the 7 mini- mum may be smaller than estimated, thus reducing the discrepancy between this proportion (supposedly, 60%) and the deficiency in 7 (25%). 100 80 O 60 I- Q- q: ° 40 m < 20 0 42 mm' 40 mm 39 mm'' 400 440 480 520 560 600 640 680 720 WAVE LENGTH, m,x Fig. 30.2. Comparison of total absorption of Chlorella with the absorption ac- tive in photosynthesis (after Emerson and Lewis, 1943). It is unlikely, however, that this discrepancy can be completely elimi- nated by this correction, i.e., that the assumption of a strongly broadened and flattened absorption band of the carotenoids will permit the explana- tion of the quantum yield curve on the basis of complete inefficiency of the carotenoids. The average quantum yield deficiency between 400 and 580 mn according to figure 30.1, is about 15%; the average contribution of the carotenoids to light absorption in the same region, according to figure 22.44, is close to 30%. Broadening of the absorption band is more likely to increase than to decrease the latter value. Thus, the most likely conclusion to be drawn from the results of Emerson and Lewis is that the sensitizing efficiency of the carotenoids in Chlorella is not zero, but about one half that of chlorophyll. It need hardly be stressed that all conclusions of this type are predicated on the assumption of a uniform distribution of the pigments, and would have to be revised if it were demonstrated that some pigments form "color screens" between the external light source and the other pigments. CAROTENOIDS AS SENSITIZERS IN GREEN PLANTS 1151 The sharp dechne in quantum yield, which sets in, according to figure 30.1, above 680 mn, will be discussed in section 5. Figure 30.1 was obtained with a dense suspension, which absorbed practically completel}' at all wave lengths. Emerson and Lewis also made measurements with a thinner suspension, which transmitted about one half of incident light, to be able to compare directly the "action spectrum" of photosynthesis in Chlorella with the absorption spectrum of the same specimen. Figure 30.2 shows the results. The two curves were drawn to coincide at 660 m^u. They show a significant divergence above 690 and below 570 m/x. The lower position of the action curve in the green, blue and violet illustrates the relative inefficiency of the carotenoids; but the comparative narrowness of the gap between the two curves confirms the conclusion, reached above, that the carotenoids are not entirely inefficient. The total absorption was measured directly; "active" absorption was calculated from photosynthesis on the assumption that all light used for photosynthesis gives a quantum yield of 0.084. This value was chosen to give agreement between the two curves in the red, where all absorption is assumed to be active. The half widths of the bands used are shown on the figure. The cells used in the run covering the red part of the spectrum were from a separate culture. The relatively low sensitizing efficiency of the carotenoids of green plants in photosynthesis, indicated by these experiments, may be either a uniform property of all pigments of this group, or it may be an average, e. g., some carotenoids may be as efficient as the chlorophylls, while others are entirely inactive. On the basis of the absorption analysis in figure 22.44 it seems unlikely that the inactive fraction of the yellow pigments does not consist of carotenoids at all, but is formed by pigments such as flavones or anthocyanines. It can be argued that the shape of the quantum yield curve below 570 mju could also be explained by assuming that a complete inefficiencj'' of carotenoids is partly compensated by an enhanced efficiency of chlorophyll. The possible difference between the photochemical functions of chlorophyll in the three excited states (corresponding to the blue-violet, orange and red band systems, respectively) is an important problem. The available evidence gives little indication of such a difference. In chapters 21 (page 634) and 23 (page 748) we concluded, from the excitation of the same red fluorescence band by light of all wave lengths, that chlorophyll molecules, excited to the electronic states A or B, are rapidlj^ transferred, by a radia- tionless process, into the lowest electronic excitation state, Y (which is the upper state of the red fluorescence band). However, on page 752, we concluded from Livingston's data, that, in the case of the blue-violet ab- sorption band, the yield of this transformation is far less than 100% — 1152 THE LIGHT FACTOR. III. COLOR CHAP. 30 in other words, that the fate of a large proportion of chlorophyll molecules in state A is different from transfer into state Y {e.g., they may undergo transition into a metastable state; cf. scheme 23.1). Whether this is true not only of chlorophyll in solution but also of chlorophyll in the living cell remains uncertain. In vivo, direct photochemical dissociation of chlorophyll by blue-violet light appears even less likely than in vitro (in consideration of the known photostability of chlorophyll in the living cell) . True, the yield of chloro- phyll fluorescence in vivo, as a function of wave length of the exciting light (represented in fig. 24.5A), shows a slight decline at the violet end of the spectrum; but this decline is most likely due to the presence of carotenoids, and not to a decreased efficiency of light absorbed by chlorophyll itself. As a matter of fact, figure 24. 5A was interpreted on page 814 as an indica- tion that the fluorescence of chlorophyll in Chlorella can be excited — al- though with reduced efficiency — also by the light quanta absorbed by the carotenoids; and this interpretation of the fluoresence curve is in agree- ment with Emerson and Lewis' quantum yield curve of photosynthesis. (However, the fluorescence curve in figure 24. 5A shows no minimum at 490 mn, which is so prominent in the quantum yield curve in figure 30.1.) To sum up, the results of the quantum yield studies of both photosyn- thesis and fluorescence are best explained by the assumption that a con- siderable fraction — of the order of one half — of the quanta absorbed by the carot- enoids in green plants is passed over to chlorophyll, transferring the latter into state Y. Therefore, these quanta can be used in the same way as those absorbed directly by chlorophyll, either for photosynthesis or for fluor- escence. It is furthermore likely that most or all blue-violet quanta absorbed directly by chlorophyll are utilized for conversion into the fluore- scent state Y. Quantum yield measurements of Noddack and Eichhoff (1939) and Eichhoff (1939) covered only the range above 515 m/x and therefore reveal nothing concerning the function of the carotenoids. Their only interesting feature is the high yield in the infrared; therefore they will be discussed in the next section. The action spectrum of the green alga Ulva taeniata, measured polaro- graphically by Haxo and Blinks (1950), is represented in figure 30. 11 A, and is in general agreement with the results of Emerson and Lewis. 3. Photosynthesis of Green Plants in Ultraviolet and Infrared The exten.sion of the photosynthetically active region into the ultra- ^'iolet has not been studied by systematic quantum yield measurements in monochromatic light, although it seems to be generally assumed that the PHOTOSYNTHESIS OF GREEN PLANTS IN ULTRAVIOLET AND INFRARED 1 1 53 yield drops rapidly at, or only a little beyond, the violet end of the visible spectrum (400 m/x)- In vitro, the chlorophyll abso'rption extends throughout the near and medium ultraviolet (c/. fig. 21.3). Whether other common components of the plant cells, which absorb in this region, interfere with the light sup- ply to chlorophyll (by acting as ''screens"), we do not know (this could perhaps be elucidated by fluorescence measurements). Ursprung (1917, 1918) found evidence of photosynthesis in Phascolvs ivJgnris down to 330 him; Hoover (1937) and Burns (1942) ol)scr^•cd it in TrUiciun at 305 niju. Gabrielsen (1940) calculated a yield of about 1 mg. CO2/5O cm.- hr. as the possible contribution of the ultraviolet part of the solar spectrum to the photosynthesis of leaves of Sinapis and Conjlus. Johnson and Levring (1946) found with six marine algae (green and red), a decline in respiration by about 1 mg. O2 per hr. per g. dry weight (deter- mined by Winkler's methods) in near-ultraviolet light (366 mn, intensity 5 X 10"^ cal./cm.- min.). This was interpreted as evidence of photo- synthetic effectiveness of this light. Further in the ultraviolet (260-320 m^, intensity about equal to that of the erytheme-producing radiation in sunlight), no such effect could be observ^ed. Light below 300 m/x is highly injurious to plants. According to Meier (1932, 1934, 1936) an illumination of 1000 erg/cm.^ sec. kills ChloreUa cells in 110 sec. at 260 m/i, and in 10,000 sec. at 302 m/x. Preferential in- hibition of photosynthesis (with the cells still alive and respiring) by the mercury resonance line 253.6 m^ was described by Arnold (1933) (c/. chap- ter 13, Vol. I, page 344); it is not associated with a visible destruction of chlorophyll. The presence of yellow pigments {e. g., flavones, anthocyanines or other hydrophilic compounds, which may be present either in the cell sap or in the cell walls) may cause a decline or complete cessation of photosyn- thesis at comparatively long wave lengths. This is the probable reason why Bums (1933, 1934) found that the photosynthesis of certain conifers — e.g., spruce and white pine — ceases below 450-465 m/x {cf. page 1164). In the spectmm of extracted chlorophyll, the absorption declines rap- idly above 680 m/x; but the absorption spectra of intact cells reveal, in addition to a shift of the red absorption band from 660 to approximately 680 mn, an extension of absorption to much longer waves. This spreading of the red band can be recognized, e. g., in figures 22.10-22.24, 22.44, and 22.48; and it has also been observed in aqueous suspensions of chloroplastic matter (Smith, fig. 21.28). In the latter case, the infrared "tail" of the red band disappeared upon clarification of the suspension by digitonin, and was ascribed by Smith to scattering ("false absorption"). However, in other investigations, e.g., those of Noddack and Eichhoff 1154 THE LIGHT FACTOR. III. COLOR CHAP. 30 (fig. 22.21), the absorption in the far red was found also in measurements purported to represent "tme" absorption. It may be due either to a genuine broadening of the red chlorophyll band, or to the presence of other, infrared-absorbing components (e. g., ferrous salts). The question of the infrared limit of photosynthesis is closely associated with the interpreta- tion of this infrared absorption "tail." The experimental results of Emer- son and Lewis (1943) and Blinks and Haxo (1950), on the one hand, and Eichhoff (1939), on the other hand, are in extreme disagreement. As 0.08 )o 0.06 o" UJ >- P 0.04 < o 0.02 246 mm* 493 mm' 986 mm' S J_ 680 700 720 WAVE LENGTH, m^ 740 Fig. 30.3. Decline in quantum yield in far red (after Emerson and Lewis 1943). Points show apparent quantum yield with three different suspension densities, calculated for equal incident light. Solid curve extrapolated for true yield, (com- plete absorption), assuming that differences between the curves obtained with different suspensions is due to incomplete absorption. shown in figure 30.1, Emerson and Lewis found a sharp drop in quantum yield above 680 m/x. This part of their curve is reproduced in more detail in figure 30.3. To be sure that the decline was not due to incomplete absorption (we recall that the curve in figure 30.1 was obtained by War- burg and Negelein's method which presupposes total absorption), Emerson and Lewis made determinations with several suspensions of increasing den- sity, and extrapolated the results to infinite density; the result is repre- sented by the solid curve in figure 30.3. It shows that, even with a gener- ous allowance for incomplete absorption, there still remains a drop in 7, from about 0.08 at 680 mn, to as little as 0.02 at 730 m/x. PHOTOSYNTHESIS OF GREEN PLANTS IN ULTRAVIOLET AND INFRARED 1155 A similar decline of X above 630 m/n was noted by Blinks and Haxo (1950) in the green alga Ulva taniata (fig. 30.11A), by Emerson and Lewis (1941) with the blue-green alga Chroococcus (fig. 30.10A) and by Tanada (1950) with the diatom Navicula minima (fig. 30. 9A). Ehrmantraut (1950) noted the same phenomenon in the Hill reaction in Chlorella, Livingston (c/. chapter 23, p. 752) noted a drop in the yield of the chlorophyll fluo- rescence (in ether and acetone) in the same spectral region. Emerson and Lewis suggested that the decline in quantum yield on the infrared side of the absorption maximum of chlorophyll a (at 680 m/x) is repeated on the red side of the absorption peak of chlorophyll h, and that this is the explanation for the shallow minimum in the y ciu've that figure 30.1 showed near 660 mn. (The absorption maximum of chlorophyll h in vivo must be situated at about 648 m^u; cf. chapter 22, page 702.) It may be pointed out in passing that the quantum yield curve of Emerson and Lewis provides a direct argument against Seybold's (1941) hypothesis that chlorophyll b does not act as a sensitizer in the reduction of carbon dioxide at all, but is a specific sen- sitizer for the polymerization of sugars to starch. In sharp contrast to the results of Emerson and Lewis are those of Noddack and Eichhoff (1939) and Eichhoff (1939), who found (cf. Table 30.11) that the quantum yield of Chiordla remains high even at 832.5 mju. Table 30.11 Quantum Yields of Chlorella According to Eichhoff van m/i 832.5 0.164 780 0.204 750 0.244 725 0.228 685 0.179 650 0.204 622 0.228 598 0.238 576.5 0.263 558 0.238 542 0.204 527.5 0.232 515 0.222 Figure 30.4 shows a comparison of the action spectrum with the absorp- tion spectrum of Chlorella, according to Eichhoff; the two curves remain closely parallel far above 680 mix. Even if we doubt the correctness of the absolute values of Eichhoff' s quantum yields (cf. page 1098), i\\ewave length dependence of these values could be significant — e. g., it would not be af- fected by an error in calibration (suggested on page 1098). One possible — but not very probable — explanation of the differences between the two y curves in the far red is that only Noddack and Eichhoff have measured true absorption, while others have related the yield to the sum of absorp- tion and scattering. The elucidation of this point is particularly desirable because of the theoretical implications of a decline in y with wave length, as 1156 THE LIGHT FACTOR. III. COLOR CHAP. 30 obsei-ved by Emerson and Lewis, and Haxo and Blinks. The latter (1950) observed, in the action spectmm of Ulva (fig. 30.11), a decline in the far red quite similar to that found by Emerson and Lewis with Chlorella. Altogether, the weight of experimental evidence seems to be against the results of Noddack and Eichhoff . This is remarkable, because from the theoretical point of view one would rather expect the quantum yield to remain constant within the red absorption band of chlorophyll. The general experience in photochemistry is that wave length is not im- portant for the photochemical effect, as long as one remains within a single band system, even if this system extends over all colors of the rainbow. The reason is that, within a single band system, the electronic excitation energy is constant, and all excess energy absorbed by the molecule serves 14 12 10 8 6 4 2 0 70 60 - ^- 50 Q. 40 q: O CD 30 < 20 10 Dense suspension /, . 15 min. V — / \ \ Assimilation \ 1 -^^ Absorption ^ /••~-,^'^'" suspension \ V'v^ \ / / 30min^ ^ \ \ ^ 500 600 700 WAVE LENGTH, m^i ROO I''ig. 30.1. Action spectrum of Chlorella (after Eichhoff 1939). Scale at left represents assimilation. merely to increase its vibrational and rotational energy. If the primary photochemical process is the dissociation of the absorbing molecule, the only effect of variations in wave length is that the dissociation products separate with different relative velocities; usually, this excess energy does not affect the ultimate fate of the dissociation products, because it is lost by collisions before the next reaction step. If the primary photochemical process is electronic excitation, the excess vibrational energy acquired by the absorbing molecules also will usually be lost before the occurrence of the secondary reaction. However, exceptions to this behavior are known. In some cases, the electronic excitation energy i.s too small to bring about dissociation without the assistance of a. certain amount of vibrational energy (an example is the delayed monomolecular photodissociation of large molecules, described by Franck and Ilcrzfcld, 1937, and considered PnOTORYNTHESIS OF GREEN PLANTS IN ULTRAVIOLET AND INFRARED 1 1 57 in chapter 18. page 484, as a possible consequence of the absorption of blue-violet light by rhlorophyll). In other cases, the presence of excess kinetic energy may help the dissociation products, formed in a tightly packed medium, to escape immediate recombination (Franck and Rabino- witch 1934). In still others, the vibrational energy of the excited elec- tronic state may contribute directly to chemical activation; this is par- ticularly likely in cases when the reaction partner forms a complex with the absorbing molecule, so that no energy-dissipating collisions intervene be- tween the primary and the secondary photochemical step. The latter explanation could be suggested for a decline in the 7 values of photosynthesis at long waves, if the experimental 7 curve would indi- cate that a certain minimum vibrational energy of the excited electronic state is required for photosynthesis. However, according to Emerson and Lewis, the decline occurs within the same band (and not upon transition from one band to another, e. g., from the orange to the red band). This is difficult to understand, since, within a single band, the absorption leads everywhere, not only to the same electronic state, but also to the same vi- brational state (at least, in respect to high-frequency band vibrations). One could suggest that the red band of chlorophyll is not a single band, but contains vibrational bands clustering tightly under its wings. How- ever, vibrational bands situated on the infrared side of the main electronic band most likely originate in the vibrating states of the ground state, and lead to the same excited state as the main band; they thus offer no ex- planation for a diminished quantum yield. Another possibility is that the red absorption band of chlorophj-ll conceals a band corresponding to a different electronic transition. In figure 21.20, it was suggested that a comparatively weak band Xo —> Ao is hidden under the strong A'o — > Yo band; one could suggest that absorption in the far red, exhibited by live cells, is caused by a shift of the Xo — > Ao band, rather than by an extension of the Xo — > Yo band. One could also sug- gest that, in the series of transitions Xq -^ Ao, Xo -^ Ai, Xo —* A2, . . ., the latter ones, which lead to the vibrating states Ai, Ai, and give rise to chlorophyll bands in the orange, yellow and green, can be followed by radiationless transitions into state Y, while the first one, which leads to the non vibrating state ^0, cannot produce the same result (because of insufficient energy of the excited state), and is therefore ineffective in bring- ing about photosynthesis. Before considering any of these hj^potheses too seriously, one should ascertain whether the drop in 7 above 680 m^, obsei^ved by Emerson and Lewis, is not due to a more trivial cause, such as scattering (which Emer- son and Lewis considered unlikely); to the presence of ferrous salts (or other infrared-absorbing inorganic components) ; or the presence of a photo- synthetically inactive pigment with a band greater than 680 m/i (perhaps chlorophyll d; see page 1183). Observations of the low yield of photosynthesis in filtered extreme red or infrared light (Urspnmg 1918, Gabrielsen 1940, etc.) without the meas- 1158 THE LIGHT FACTOR. III. COLOR CHAP. 30 urement of absorption have verj^ little significance because of the well- established rapid drop in the absorbing power of leaves in this region. The only real problem is whether the yield of photosynthesis drops 'pro- portionally with absorption, or more rapidly than the latter. Hoover (1937) was able to observe the photosynthesis of wheat up to 750 m/i and Burns (1933, 1934), that of different conifers up to 740 rati. 4. Monochromatic Light Curves, and the Action Spectrum of Photosynthesis in Strong Light When the intensity of monochromatic light is raised, one soon reaches the region in which the shape of the action spectrum becomes variable. The light curves bend earlier or later, and come to saturation more or less suddenly, depending on the value of the absorption coefficient, and on the optical density of the sample, respectively. It was postulated above (page 1145) that, when all curves reach saturation, the rate must become independent of wave length, and the action spectrum must lose all structure. The theoretical and experimental foundations of this postulate will be con- sidered later (pp. 1162, 1165). At present, we will assume it to be valid, and consider only the effect of wave length on the shape of the transition from the linearly ascending part of the light curves (the slope of which at a given wave length is determined by the product of absorption coef- ficient and maximum quantum yield) to the "saturation plateau," the height of which we assume to be independent of wave length. The effect of optical density on light curves was discussed in chapter 28, and the results were illustrated by the schematic figure 28.20. A change in wave length is equivalent to a change in optical density; a cell suspension that is "thin" in green light becomes "dense" in red or violet light. How- ever, as far as the rate of photosynthesis is concerned, transition from green to red light is not in all respects equivalent to an increase in cell concentration, since the saturation level remains unchanged in the first case, but increases proportionately with the number of cells in the second case. Thus, the result of a change from strongly absorbed to weakly ab- sorbed light is likely to be more similar to that of the change from green to aurea leaves {cf. fig. 32.2), where the maximum rate is approximately the same for both varieties. As mentioned before, the comparison of light curves at different wave lengths should be carried out by plotting the rate against N^p (number of incident quanta/cm. ^ sec.,) rather than against the energy flux, 7, in erg (or cal)/cm.^ sec. Otherwise, the light curves for the shorter waves will remain below those for the longer waves, even if the photochemical ef- ficiency of both kinds of quanta are identical (fig. 30. 6C). An example of MONOCHROMATIC LIGHT CURVES 1159 how completely the "quantized" light curves obtained in monochromatic light of different color coincide in their initial sections is given in figure 30.5. Although the two uppermost points in this figure fall into the region of beginning saturation, they still show no difference between the liglit curves in green and red light. However, a divergence of these two curves in the NT E O in 7 - .E 5h E O _» O E (O V) UJ X I- z >- to o I- o X a. Fig. 30.5. 4 - 3 - 2 - - ^o" - ^a > o Red D Green A Blue o / ^ 1 1 1 J. 1 1 1 1 1 J 123456789 10 INTENSITY, einstein/cm^min, x 10' Photos\-nthesis of Chlorella as function of intensity at three wave lengths (after Emerson and Lewis 1943). fAl Q. Green/ /^ //Red Nf,j (absorbed) N^^ (incident) Fig. 30.6. Expected shapes of monochromatic light cm-ves. (A) P vs. Nt,„ for total absorption, A^^^, (absorbed) = A^^^, (incident). (B) Same for incomplete absorption, Nn;, (absorbed) < N^y (incident). (C) For equal absorption, equal quantum yields but different wave lengths; the curves in (C) would coincide if Nh:, were used as abscissa instead of /. In (C), the unbroken curve is for high X. satiu-ation region is theoretically inevitable. The theoretical expectations are illustrated by the schematic figure 30.6. In the case of total absorption of all wave lengths {i. e., conditions under which fig. 30.5 was obtained), the curve for red light must bend earlier than that for green light (because of the more uniform absorption of the latter throughout the cell layer). 1160 THE LIGHT FACTOR. III. COLOR CHAP. 30 The same picture (fig. 30.6A) should be obtained also for partially absorb- ing systems, if P is plotted against the absorbed (rather than against the incident) intensity. If Nn^ (incident) is used as the independent variable and the absorption is comparatively weak, the resulting picture must be that shown in figure 30.6B. If the monochromatic light curves are plotted against light energy (I, or A) instead of number of quanta, Nh;,, the relationships become unneces- sarily obscured {of. fig. 30.6C). (Some investigators, e. g., Montfort, have JO 52 - (A) 1 _ (B) I ^--'■■'^ 48 - /Red - Red y^ - 44 - / - / e 40 o : / - / a. 32 6 - - en- 28 - - 31 (some as I, all en iiJ 24 - / I's multiplied by 2.5) PHOTOSYNTH _ — ro ro CD O - /Green - '^ White 8 - - / /^""'^ 4 i 1 1 1 1 1 1 6 9 1 1 12 15 ISklux 1 1 1 12345 123456 m.c. /(in klux for white light; in "energetic meter candles' for colored light) Fig. 30.7. Light curves of Chlorella in liglit of different color. (A) Green vs. red light (equal energy flux); "dense" suspension (11 X 10^ cells per cc). (B) Red vs. white light (energy flux in klux for white light, in "energetic meter candles" for red light, cf. chapter 29, p. 1098). "Thin" suspension, 2 X 10" cells per cc. used rate measurements in monochromatic light to raise the question whether ])hotosyn thesis "is a quantum process at all" ; a question no photo- cliemi.st would ever ask. The available experimental material to which the above predictions can be applied is very scarce. It is desirable that the precise methods of rate determination, applied as yet almost exclusively to the quantum yield determinations in weak light, should be extended to kinetic studies in strong monochromatic light. As examples of the few available experimental monochromatic light MONOCITEOMATIC LIGHT CURVES 1161 500 600 INCIDENT INTENSITY, cal/50 cm.'' hr. Fig. 30.8A. Light curves of Sinapis alba (after Gabrielson 1935). Top, blue-violet light; center, yellow-green; bottom, orange-red. 7.«— "o X 0-^-r-»- 200 400 600 800 1000 1200 INCIDENT INTENSITY, cal/SOcm'^hr Fig. 30.8B. Light curves of Sinapis alba in light of different color (after Gabrielsen 1940). The same saturation yield is approached at all wave lengths. Circles, n>d- orange light; crosses, yellow-green; triangles, blue-violet. 1162 THE LIGHT FACTOR. III. COLOR CHAP. 30 curves, we consider those in figures 30.7 and 30.8. In figure 30.7A, the relative position of the curves for green and red light is as predicted in figure 30.6B. In figure 30.7B, the curve for white light, lies, as predicted, below that for the more strongly absorbed red light, but the difference is much larger than expected. It was mentioned in chapter 29 (page 1098) that the ratios between photochemically equivalent intensities of red and white light, given by Eichhoff, appeared remarkably small. (Satui'ation in red light was reached at an intensity equivalent to <2000 lux of white light!) If one arbitrarily multiplies the "monochromatic" intensities by a factor of 2.5 (this would reduce quantum yields from 0.25 to 0.10), fig. 29. 7B would be changed as indicated by the dotted line, and acquire a much more plausible appearance. Figure 30.8A and Table 30.III show the results of Gabrielsen (1935). The difference in the quantum yield in the green and in the red is notable Table 30.III Photosynthesis in Colored Light (after Gabrielsen 1935) Max. quantum Max. P, mg. CO2/ Reached at / = Light Av. X, ni^ yield cm.^ hr. 11 cal/cm.^ sec. Red-orange 650 0.100 0.192 1.67 Yellow-green 540 0.083 ca. 0.12 1.85 Blue-violet 430 0.071 ca. 0.05 0.83 (for a discussion of similar results by Emerson and Lewis, cf. page 1148). The maximum rates in the blue-violet, and particularly in the green, also are considerably smaller than in the red; however, figure 30.8A shows that the observed maximum rates may be still far from saturation. Figure 30. 8B, taken from a later publication by the same author (1940), shows the coincidence of the maximum rates in the three spectral regions very clearly. It is obvious from the preceding discussion that the action spectra of photosynthesis, obtained by illuminating plants with light of partly saturat- ing intensity, are difficult to interpret — even if precaution has been taken to use the same incident intensity (or, better still, the same number of incident quanta) in all spectral regions. For example, in an optically thin system, the yield per incident quantum should be smaller in the green than in the red, because of weaker absorption (cf. fig. 30. 6B); while in a dense, completely absorbing system the relation could be reversed, because of the better utilization of the more uniformily absorbed green light (cf. fig. 30.6A). Working with systems that are not too dense, and in light that is not too strong, one may obtain a "quantized" action spectrum resembling more or less closely the absorption spectrum of the pigments. An example is given in figure 30.9, which shows the action spectrum of wheat as ob- served by Hoover (1937) and "quantized" by Burns (1937-38,1942). How- MONOCHROMATIC LIGHT CURVES 1163 ever, no quantitative agreement between action spectrum and absorption spectrum can be expected under these conditions, and conclusions drawn from differences between them, e. g., as to the role of the accessory pig- ments, are in the nature of more or less plausible guesses. Because of the coincidence of the absorjjtion bands of the carol enoids in green plants with the blue-violet bands of the two chlorophylls, correct guessing is in this case much more difficult than in the case of brown or red algae. Engelmann (1887) recognized tliis and based his suggestion that the carotenoids of green plants also act as sensitizers in photosynthesis not on direct experiments with these plants, but on analogy with the results obtained with colored algae. He quoted, as an additional argument in favor of this suggestion, the observation that leaves of the aurea varieties have a com- paratively high yield of photosynthesis, despite their deficiency in chlorophyll. How- ever, he did not consider this argument as conclusive, at least not without x'enewed study; and since then Willstatter and Stoll (1918) have shown that aurea leaves possess a high relative efficiency also in light filtered through a yellow filter. Willstatter and Stoll saw 400 800 500 600 700 WAVE LENGTH, m/i Fig. 30.9. Action spectrum of photosynthesis of wheat "quantized" by Burns (1937, 1938) (after Hoover 1937). in this proof that leaf carotenoids do not contribute to the sensitization of photosynthesis in aurea leaves. However, this conclusion was not convincing because in their experi- ments not only the relative, but also the absolute, yields in both green and aurea leaves were almost unaffected by the interposition of a yellow filter. In other words, the in- tensity of blue-violet light was negligible; therefore their experiments, while proving that aurea leaves are liighly efficient in the light absorbed by chlorophyll alone, proved nothing as to the efficiency or inefficiency of the carotenoids. Wurmser (1921^) found the rate of photosynthesis of Ulva lactuca in the blue-violet to be lower than in the green, but higher than in the red (calculated for absorption by chlorophyll alone), thus indicating a possible active participation of the carotenoids. Schmticker (1930) found, by bubble-counting experiments with Cabomba and Crypto- coryne, that the light intensity required to achieve a certain rate of photosynthesis in- creased from the red to yellow and green inversely proportionately to the wave length — thus indicating a constant quantum yield; in the blue and violet, on the other hand, the increase was about 15% larger than was required by the quantum correction if all pigments were assumed to be active, but somewhat less than could be expected if the carotenoids were entirely inefficient. 1164 THE LIGHT FACTOR. III. COLOR CHAP. 30 After the problem of the role of carotenoids in photosynthesis had been brought to the foreground by experiments with brown algae (cf. section 5), Montfort (1940) made a new attempt to determine whether light absorption by the carotenoids of green plants also contributes to photosynthesis. He compared the rates of photosynthesis of the green alga Ulva lactuca in red and orange light (\ > 550 m/x) with that in blue-green light (X 350-625 ran, maximum at 450-500 m/x) of equal incident intensity. Table 30.IV shows the results. Comparison of the last three figures in the table shows that the Table 30. IV Photosynthesis of Ulva lactuca in Colored Light (after Montfort 1940) P = Rate of Photosynthesis; A = Rate of Absorption P (blue-green)/P (orange-red) 0 . 89 A (blue-green)/A (orange-red)° 1 . 19 A (blue-green)/A (orange-red)'' 1 .00" iP/A) (blue-green)/(PM) (orange-red)" 0.75 (P/A) (blue-green)/(P/A) (orange-red)'' 0.89" Quantum correction: X ( orange-red )/X (blue-green) O.??** " All pigments. '' Chlorophyll alone. " Calculated from data on extracts. '' For wave lengths 625 and 480 mju. quantum yield in the blue-green was only slightly lower than in the red, if referred to the absorption by all pigments, but much higher if referred to chlorophyll alone. This speaks in favor of an almost equal efficiency of chlorophyll and the carotenoids. Thus, the earlier observations of Wurmser, Schmlicker and Montfort can all be quoted in support of the conclusions derived by Emerson and Lewis from the much more con- vincing measurements in weaker and truly monochromatic light, that the carotenoids of green plants do contribute actively, but less efficiently than chlorophyll, to the sensitiza- tion of photosynthesis. In ChlorcUa, the quantum yield deficiency in blue and violet light is not likely to be caused by the presence of a yellow pigment other than the carot- enoids. (The comparison of the absorption spectrum of live cells with that of the extracted pigments, cf. fig. 22.44, does not indicate the presence of such a pigment.) In some higher plants, on the other hand, pigments of the flavone or anthocyanine class often are present in the cell sap or cell walls and compete with the photosynthetically active pigments for blue- violet ({uanta, or even serve es "color screens," particularly when they are located in the* epidermis, or in the cell walls between the chloroplasts and the external light source. The presence of these pigments should leave the saturation yield unaffected, but should depress the quantum yield in the linear range and in the region of partial saturation. Burns (1933, 1942) noted that the quantum yield of photosynthesis of spruce and pine seedlings in the blue-violet (390-470 m/x) was only half as large as in the red (630- 720 mn), or red plus orange (560-720 ran). This can be attributed to the MONOCHROMATIC LIGHT CURVES 1165 presence, in these conifers, of a nonactive yellow pigment. (It was men- tioned on page 1153 that photosynthesis in these plants declines to zero below 450 or 465 m/x.) The same effect should be even more pronounced in leaves of the pur- purea varieties, or other leaves containing large quantities of red antho- cyanine pigments. Engelmann recognized as early as 1887, in an investi- gation entitled ''Leaf Hues and Their Importance for the Decomposition of Carbonic Acid in Light," that the red pigments of land plants do not actively participate in photosynthesis; and Willstatter and Stoll (1918) and Kuilman (1930) confirmed that the presence of these pigments has no influence on the rate of photosynthesis in strong light. Gabrielsen (1940) concluded, from a review of the older work and new experiments with Corylus and Prunus leaves in red-orange, yellow-green and blue-violet light, that for a given amount of incident energy the red varieties have a minimum yield in yellow-green light, where light absorption by the red pigment has its maximum. This minimum was most pronounced in weak light; in strong light, the monochromatic light curves approached the same saturation level, in conformity with the findings of Willstatter and Stoll. Gabrielsen estimated that, in red Corylus leaves, the "screen" absorbed about 37% of incident light in the blue-violet, about 74% in the yellow- green and about 33% in the red-orange region. In Prunus, the absorption was somewhat higher, perhaps because in Corylus the red pigment was pres- ent only in the epidermis cells, while in Prunus it was found also in the mesophyll. These results cause us to give little credence to speculations or observations that relate anthocyanines or flavones to photosynthesis. In 1922 Noack observed a photochemical conversion of flavones into anthocyanines in vivo and interpreted this reaction as a chlorophyll-sensitized oxidation-reduction (cf. chapter 19, page 541). He suggested that flavones and anthocyanines form a reversible oxidation-reduction system, which may play a catalytic role in photosynthesis. Sen (1942) asserted that anthocyanine-carrying leaves have a higher photosynlhetic ef- ficiency than ordinary leaves, despite their lower content of chlorophyll. We now return to the question, raised on page 1158, concerning the third part of the light curves, after the linear range and the transitional region — the saturation plateau. It was })ostulated there that this plateau should have tlu; same height for all wave lengths. .Vs long as tlu; rate is determined only by the kinetic mechanism of pJKjtosyntliesis, theoretical arguments in favor of this postulate appear conclusive. Whether satura- tion is brought about by a limited supply of reactants, or by limited avail- ability of an enzyme, the maximum rate is determined b}^ the velocity of a dark reaction, and should be independent not only of the quantity, but also of the quality of illumination. Light (luality could, however, affect the maxiinuni rote of photosynthe- 1166 THE LIGHT FACTOR. III. COLOR CHAP. 30 sis, if other photochemical processes can interfere with this process. A selective sensitizaton of oxidative processes in the photosynthetic apparatus by the light absorbed by the carotenoids, or by chlorophyll in the blue-violet band, offers one possibility of this type. However, in this case, only light curves of the time-dependent "optimum" type (which have been observed, e. g., in some umbrophilic plants; cf. page 994) should be affected, since the maximum rate in these curves is determined by the (time-dependent) bal- ance of photos:ynthesis and photoxidation. Light curves with true satura- tion plateaus should remain unaffected by an earlier onset of "light inhibi- tion" (except for a shorter extension of the saturation plateau). The effect of wave length on photoxidation has never been investigated systematically. Franck and French (1941) found that photoxidation oc- curs, in carbon dioxide-deprived leaves, in red as well as in blue light, but this was merely a qualitative observation. A selective effect of blue light on the respiration of Chlorella, observed by Emerson and Lewis (1943), was mentioned in chapter 20 (page 568). Another specific function of yellow pigments seems to be well established — the sensitization of photo- tropic movements. As stated on page 681, the changes in the positions of the chloroplasts in light are caused only by light absorbed by yellow pig- ments (Voerkel 1933) ; and the same is true of the phototactic movements of whole cells {of. Castle 1935). These pigments may be the carotenoids, although Galston attributed this function to riboflavin (because the action spectrum showed only a single peak). In purple bacteria, on the other hand, the action spectrum of phototaxis coincides with that of photosynthesis {cf. p. 1188). Very extensive studies of the effects of light of different colors in photo- synthesis and respiration were made by Danilov (1935, 1936), using green, blue-green, and red algae. He reported very complicated results in which the yield was found to depend not only on color (in monochromatic light), but also on the combination of colors (in non-monochromatic light). Fol- lowing the tendency of what Kostychev proclaimed as a new "physiological" approach to photosynthesis (cf. chapter 26, p. 872) he discussed these phe- nomena in vague terms of stimulation and inhibition of different protoplasmic functions by hght of different wave length. Thus, yellow and green rays were credited by him with increasing cell sensitivity to red light, and with making it insensitive to infrared light; blue-violet light was said to assist in the utilization of infrared light (supposedly for activation of the "dark" reaction stages in photosynthesis). Blue-green rays were said to counter- act the stimulation effects of yellow light, and enhance the stimulating effects of blue rays, and, generally, to create in the cells a "regulator of the MONOCHROMATIC LIGHT CURVES 1167 utilization of light energy," and also to determine the reaction of photo- synthesis to changes in temperature. In another paper, Danilov (1938) concluded that the effect of various colors of light on photosynthesis depends on the method of cultivation of the algae. In particular, variations in the sensitivity to different colors of light were caused by changes in hydration of the cells (achieved by culturing Scenedesmus in 1% sodium chloride solu- tion). Ursprung (1917, 19182), working with detached leaves by means of Timiriazev's (1890, 1903) "starch spectrum" method (a spectrum is projected on a starved, destarched leaf, and the "latent starch image" formed by light is "developed" by iodine), found, in light of uniform spectral intensity, a continuous decrease in the production of starch from red to violet, without a second maximum, and attributed this result to the closure of the stomata and consequent quenching of photosynthesis in blue-violet light. Dastur and Samant (1933), Dastur and Mehta (1935) and Dastur and Solomon (1937) described observations that purported to show that pure red light (or pure blue light) is less efficient in photosynthesis than a com- bination of both. Their experiments were criticized by Montfort (1937), who found that the addition of blue-violet light to red light has no effect on the rate of photosynthesis, provided the red light was in itself of saturat- ing intensity (this agrees with the experiments of Button and Manning with diatoms, described on page 1172). Dastur, Kanitkar and Rao (1938) measured the formation of proteins in leaves in light of different color, and found differences which they related to the above-mentioned observations on photosynthesis in colored light. The results of Dastur and co-workers probably are trivial, being caused by the use of optically dense tissues. As explained above, such tissues must (and do) utilize moderately intense green or yellow light better than orange-red (or blue-violet) light of the same intensity, because the latter is absorbed in too thin a layer and causes saturation effects there. White light of partly saturating intensity, since it contains green and yellow, will give, in such tissues, a higher yield of photosj^nthesis than an equally strong red-orange or blue-violet light. The hypothesis of Baly (1935) that photosynthesis requires a quantum of red plus a quantum of blue light, was mentioned before (Vol. I, p. 554) and characterized as entirely without experimental basis. In chapter 28 (page 987) we described the different shape of light curves of shade-adapted and light-adapted plants. Since these plants differ in the composition of their pigment systems, they are likely to show differences also in their response to light of different color. Lubimenko (1923) ob- served that shade-adapted plants often are relatively more efficient in blue- violet light (and relatively easily inhibited by red light). This may be as- 1168 THE LIGHT FACTOR. ITT. COLOR CHAP. 30 sociated with their higher content of chlorophyll 6 (which permits a better utilization of blue light, 450-500 niju), or with a higher content of carot- enoids (which was indicated by Willstatter and Stoll's figures in Table 15. Ill, but was not confirmed, as a general rule, by Seybold and Egle's analysis; cf. Vol. I, p. 414). The composition of the pigment system also depends on the color of the light under which the plants were grown. As described in chapter 15 (page 130), chlorophyll is more efficiently synthesized by green plants in red light, and the carotonoids in blue-violet light (although these assertions have been contested, and may represent over-simplifications). This may explain why plants have often been found to be most efficient in the light in which they were grown. Elodea plants cultured in red light produced more oxygen in the red, while similar plants grown in blue light gave more oxygen in the blue (Harder, Doring and Simonis 1936, Harder and Simonis 1938, and Simonis 1938). Thus, the physiological chromatic adaptation of photosynthesis may be in this case a consequence of the chemical adapta- tion of the pigment system. 5. Quantum Yield and Action Spectrum of Photosynthesis in Brown Algae The study of the relation between wave length and photosynthesis in brown algae and diatoms is of special interest because of the presence in these algae of the carotenoid fucoxanthol, which is not encountered in green plants. The distribution of the light absorption by brown algae and diatoms among the individual pigments was discussed in chapter 22 (page 723) and illustrated by the (very schematic) figures 22.45A, B and Table 22.IX, all taken from Montfort (1940). For diatoms, we also gave the much more adequate figure 22.46 of Button and Manning (1941). The two reasons why even this figure is not too reliable are: first, the un- certain (and undoubtedly to a certain extent incorrect) assumptions made regarding the "red shift" of the absorption bands in vivo, and, second, the neglect of chlorophyll c. According to Tanada's figure 30. 9B, chlorophyll c adds much to light absorption of pigment extracts from bro^vn algae and diatoms between 450 and 500 mju. The brown color of these organisms in VIVO indicates considerable absorption also farther in the green, from 500-550 niM- Part of this absorption may be due to chlorophyll c, but most is probably due to fucoxanthol (to which it has usually been ascribed, cj. page 707). Montfort (1940) discussed the experimental action spectra of photosynthesis in different brown algae and concluded that light ab- sorbed by fucoxanthol is fully utilized for photosjmthesis; but this conclu- sion was not very convincing because of the very primitive experimental approach, which included the use of broad spectral regions, and of light of ACTION SPECTRUM OF BROWN ALGAE IIG'J comparatively high intensit3^ Dutton and Manning (1941) arrived at a similar conclusion by a procedure which was much more satisfactory — at least, in principle — namely, the determination of quantum j-ields in weak and trul}^ monochromatic light. Because the method of Dutton and Man- ning is so much more adequate than that of Montfort (c/. the criticism of Emerson 1937) we will discuss their experiments first. Dutton and Manning used the dropping mercury electrode for the determination of oxygen (c/. chapter 25, page 850). The diatom {Nitzschia closierium) was found to be more sensitive than Chlorella to merciuy; how- ever, its resistance was sufficient to permit measurements of 30 minutes duration without marked poisoning. Chromatographic analysis of the pigments of Nitzschia closierium re- vealed the presence of chlorophyll a, carotene, luteol, fucoxanthol and prob- ably flavoxanthol, but showed no trace of chlorophyll h. No mention was made of chlorophyll c, which was subsequently found by Strain and Man- ning (Vol. I, p. 406) in diatoms and other broMTi algae. Analysis of the absorption spectmm of the extract (c/. fig. 22.46) indicated that in methanol solution at least one half the absorption between 400 and 550 m/i was due to the carotenoids. This is a much larger proportion than in green plants (c/. page 1150); it indicates that the ratio chloroph.yll a/total carotenoids was, in the investigated diatoms, considerably lower than 3/1 (which is the average for brown algae listed in Table 15.III). Monochromatic light from a high-pressure a. c. mercury arc was used for illumination (the effect of intermittency being considered unimpor- tant), as well as bands isolated from the light of an incandescent lamp by appropriate filters. The density of the suspensions was such as to give about 50% absorption in the red, and 75% in the blue. Two portions of the same suspension were placed in two vessels, and a simultaneous deter- mination of the quantum yield was made in both of them, the one vessel being illuminated with violet, blue or green light, and the other with red light. Table 30. V shows the results. This table indicates that the individual y values varied, for each wave length, within wide limits {e.g., in the red, from 0.038 to 0.100); possibly because of various degrees of mercury poisoning, or other factors affecting the vitality of the cells. The conclusions therefore depended entirely on the ratios of the yields obtained in light of different color, in simultaneous experiments with two aliquots of t he same cells. Table 30.V shows that even these ratios varied widely (e. g., from 0.90 to 1.43 in the comparison of yields in violet and red light). The most consistent was the series of measurements at 496 m^u, where seven determinations all gave, for the ratio (quantum yield in blue-green)/ (quan- tum yield in red) values between 0.07 and 0.08. By averaging the ratios, 1 )utton and Manning arrived at the values in the last column of Table 30.V. 1170 THE LIGHT FACTOR. III. COLOR CHAP. 30 Table 30.V Quantum Yields of Diatoms (after Dutton and Manning, 1941) % of total abs. due to caro- 7 tenoids, " From To Av. 38 0.048 0.075 0.075 0 0.048 0.060 0.052 49 0.048 0.079 0.064 0 0.045 0.092 0.063 93 0.039 0.081 0.059 0 0.054 0.100 0.080 48 0.044 0.080 0.065 0 0.038 0.080 0.060 7/7 (red) 0.77 1.20 1.04 ±0.05 0.96 1.27 1.10 ±0.04 X, m^ tenoids," From To Av. From To Av. ^2^L+fJl^^^"^f^ ^n ^Rlf ^-^I^ H-^S 0.90 1.43 1.08 ±0.08 660.0 (red band axis) 435.8 (blue) 665 . 0 (red band axis) 496.0 (blue-green) 93 0.039 0.081 0.059 ^ ^^ 0.80 0.75 ± 0.03 660.0 (red band axis) 546 . 1 (green) 665 . 0 (red band axis) " Calculated by shifting all curves in figure 22.46 toward the red by 20 m/j. (for a criticism of this procedure, see page 726). Chlorophyll c absorption neglected! and concluded that the quantum yields in the violet, blue and green are practically equal to that in the red, despite the fact that in the first two re- gions 40-50% of the light must be absorbed by the carotenoids, whereas in the last one all absorption is due to chlorophjdl. Dutton and Manning therefore suggested that all carotenoids present in diatoms must act as sensitizers of photosynthesis. The value of 7 in the blue-green, at 496 m/x, was definitely lower. Dutton and Manning suggested that this may indi- cate a lesser efficiency of carotene and luteol as compared with fucoxanthol. According to their calculations, the relative contribution of fucoxanthol to the total absorption by carotenoids was less at 496 nifx than at the shorter wave lengths. This assumption is derived from observations in the extract, in which fucoxanthol shows an absorption spectrum terminating sharply at 500 m^ (fig. 2 1.35 A), and a maxi- mum practically coincident with that of luteol. However, this apparently does not apply to fucoxanthol m vivo (p. 706); and some recent observations on e.xtracts also showed absorption extending to 550 m^ (cf. fig. 21.36). The brown color of the algae indicates a considerable spread of the blue-violet ab- sorption region toward the longer waves; and it is plausible — although it cannot be proved — that the pigment responsible for this spread is fucoxanthol. (A possible alter- native is to ascribe part of it to chlorophyll c, although its band lies, in e.xtracts, on the short-wave side of that of chlorophyll a.) As an alternative, Dutton and Manning suggested that an unkno\vn pigment, with absorption restricted to a narrow region around 500 m/x, may be responsible for the minimum in the yield curve at 496 m/x. This result may also be related to Emerson and Lewis' observations of a selec- tive stimulation of respiration in Chlorella by light in the region of 480 mju. Dutton and Manning pointed out that the participation of the carot- enoids in photosynthesis of Nitzschia closierium is indicated, not only by ACTION SPECTRUM OF BROWN ALGAE 1171 the ratios of the quantum yields in the bhie, violet, green and red, but, even more strikingly, by the absolute value of the yield at 496 m/x, where, ac- cording to their estimates, 93% of absorbed light is taken up by the carot- enoids. They argued that, if all photosynthesis observed in this spec- tral region were attributed to chlorophyll, the quantum yield would be 0.059/(1 — 0.93) = 0.84, i. e., much larger than the maximum allowed by thermochemical considerations. However, this estimate was based on the distribution data in figure 22.46, and therefore is subject to possible grave errors. True, at 496 mn, the apportionment of energy is not very sensitive to the postulated specific value of the "red shift"; it would remain almost the same if a shift of 10 or 30 m/i were postulated, instead of 20 m/z (the value used by Button and Manning), or if the shift of the carotenoid bands — particularly^ those of fucoxanthol — were assumed to be twice or three times as large as that of the chlorophyll bands. (A difference of this type is indicated by some data in chapter 22; cf. Table 22. VI and page 706.) It may thus seem as if an extreme and unlikely assumption con- cerning the enhancement of the absorption of blue-green light by chloro- phyll in vivo, or the assumption of a spatial distribution of pigments strongly favoring absorption by chlorophyll would be required to explain the quan- tum yield observed at 496 m^t without recourse to sensitization by carot- enoids. This argument, however, ceased to be quite conclusive, since Strain and Manning (1942) confirmed the presence in blown algae and diatoms, of a pigment with strong absorption in the blue-green, chlorophyll c. According to figure 21.5 this component has an absorption peak at 450 m/x in methanol; in the living cell, its absorption maximum must lie near 470 m/x, if the shift is the same as for chlorophyll a. According to figure 30. 9C, at 470 mix, chlorophyll c in a methanol extract from diatoms accounts for about ten times more absorption than chlorophyll a. The neglect of chlorophyll c in the calculations of Button and Manning thus may have shifted the ratio of the absorptions by the chlorophyll pigments and the carotenoids, from perhaps about 1 to 1, to the extreme value of 9.3 to 0.7. To sum up, the average 7 values found by Button and Manning sup- port the assumption that the carotenoids in diatoms (and fucoxanthol in particular) contribute directly to the sensitization of photosynthesis; but the wide scattering of individual results called for reinvestigation with material and methods giving more consistent results. Furthermore, all results, and particularly the absolute yields at 496 m/x, were in need of re- examination in the light of the possible role of chlorophyll c. This re- examination could have conceivably brought the brown algae in line with green algae — organisms in which a distinctly lower quantum yield of photosynthesis was observed in the regions of the carotenoid absorption, but a yield not sufficiently low to permit the assumption of complete in- 1172 TITE Lir.TTT FACTOR. III. fOT.OTl CTI.\P. 30 acth'ity of the carotenoids. This hypothesis was supported by the action spectrum of the brown alga Coilodesme, determined polarographically by Haxo and BHnks (1950) and reproduced, together with the absorption spectrum, in figure 30.1 IB. Dutton and Manning also performed experiments in stronger light, in which they first measured photosynthesis in saturating red light, and then added violet light. They found — as one would expect — no appreciable effect of this additional illumination on the yield, -.mA interpreted this as a proof that photosynthesis in blue light, although sensitized by both chloroph3ll atul the carotenoids, is limited by the same dark reaction as photosynthesis in red light, which is sensitized by chlorophjdl alone. Wassink and Kersten (1946) studied the diatom Nitzschia dissipata; the spectroscopic results of this study were presented in Chapter 22 (p. 706). These investigators made measiu'ements of the rate of photosynthesis in 012 0.10 0.08 0.06 0.04 0.02 1 1 l " 1 .!=:;•/- -7i ... _ „ 3 / ' \ I 1 — < — \ \ 400 440 480 520 560 600 640 680 720 WAVE LENGTH, rn/i Fig. 30.9A. Quantum yield (ordinate) of photosynthesis as a function of wave length for A^. minima (after Tanada 1951). light of different color, isolated bj^ filter combinations. They reduced the data to a common average intensity (7.3 kerg/cm.^ sec.) in the assumption that they worked in the linear part of the light curve. (This assumption was based on the type of light curves, showing a verj'- extensive linear part, which had been obtained by the Dutch group for green plants, diatoms and purple bacteria, cf. fig. 28.14B and table 28.11; this shape was not con- firmed b}^ other investigators.) Wassink and Kersten estimated that the yield per absorbed quantum is the same in Nitzschia and in ChloreUa, and is constant in red, yellow, and yellow-green light; in blue-green light, it is somewhat lower in both organisms. They concluded that in contrast to other carotenoids, fucoxanthol is fully effective as sensitizer of photo- ACTION SPECTRUM OF BROWN ALGAE 1173 synthesis. They found, similarly to Dutton, ISIanning and Duggar (Chapter 24, p. 814), that chlorophyll fluorescence in li\'ing diatoms can he excited also by light absorbed by fucoxanthol; from this they concluded that the energy absorbed by fucoxanthol is transferred to chlorophyll be- fore it is used for photosynthesis. 440 520 600 WAVE LENGTH, m^ 680 Fig. 30.9B. Absorption spectra of methanol solutions of pigments ex- tracted quant ittitively from cells of \'avic>da minima (after Taiiada, 1951). o o 1.0 0.9 0.8 0.7 0.6 OS'* 0.4 L-, 0.3 0.2 •\ I — I — . Cell suspension — I Fucoxanttiol f 1 1 ♦ — * — * Chlorophyll c 1 1 1 1 \ \ \ .__.. -.Total pigments 1 1 1 1 / ^ \ \ V \ V^'' / \ 1 v ^ ^ h^ / I J \ M L "*" ^ 1 ' ^c ^-o--^ ^ L=„ — 440 520 600 WAVE LENGTH, m^i 680 Fig. 30.9C. Comparison of absori)- tioii spectra of extractefl pigments, and of intact cells of A', minima suspended in glycerol. The whole .spectrum of each pigment shifted to\v;iid the red by an amount equal to that l)y which the blue maximum shifted by extrac- tion. A reinvestigation of the quantum yield of brown algae in light of differ- ent color was undertaken by Tanada (1951) in Emerson's laboratory. The conclusions of Dutton and Manning were confirmed by much more precise measurements, taking into account the presence of chlorophyll c. Tanada worked with the diatom Navicula minima in pure culture. He measured the quantum yield in narrow spectral bands, from 400 to 700 m/i. Fig. 30.9A shows the results: 7 is constant between 520 and 680 mju; as in Chlorella, it drops sharply to almost zero above 710 m/i. The 1174 THE LIGHT FACTOR. III. COLOR CHAP. 30 yield dips by about 20% between 520 and 475 ibm (we recall that a minimum of efficiency was found in this region also for Chlorella and Chroococcus) . The yield rises somewhat further in the violet, but declines again toward the ultraviolet, its value at 400 mju being about 30% below the maximum. This 7 = /(X) curve must be compared with the curve showing the ab- sorption by the several pigments in extract from these diatoms (fig. 30. 9B), and with the curve (fig. 30.9D), derived from it, showing the contribution of each pigment to the total absorption by the cells. The first figure shows chlorophyll a, chlorophyll c, and fucoxanthol, as the major components. 100 o a: o (/) CD < 80 60 u. 40 o z UJ o cc UJ CL /'^\ , / / \ \\ \ / v \ y ^^^ Chlorophyll a & c Y ■ Fucoxanthol A '—^^ Other caretenoids / /\ 1 \ ^ •"^-^^ p^ \ \ ^. 20' 400 450 500 550 600 WAVE LENGTH, m/x 650 700 Fig. 30.9D. Curves showing the estimated distribution of Uglit absorption among pigment groups iii live cells of 'N . minima as a function of wave length. Neofucoxanthol (c/. chapter 37), /3-carotene and an unidentified carotenol (possibly diadinoxanthol) also were identified; their absorption is lumped together under the heading of "other carotenoids." In deriving, from the spectra of pigments in vitro, their contribution to total absorption in vivo, corrections for scattering, band shift, and band broadening, must be applied. The unsatisfactory state of this problem was discussed in chapter 22. Tanada made one improvement in the pro- cedure: he found that most of chlorophyll c and some fucoxanthol can be extracted with 65% methanol, leaving practically all chlorophyll a in the cells; the spectrum of the residue could tlien be used to locate the position of the blue-violet absorption peak of chlorophyll a, and of "other caro- tenoids" in the cell. The red shift of the blue-violet band of chlorophyll a, determined in this way, was 8 mju; that of "other carotenoids," 20 m^i. ACTION SPECTRUM OF BROWN ALGAE 1175 (However, the latter value was derived from a small shoulder on the absorption curve, and is not precise.) The absorption peak of fucoxanthol in vivo was calculated from the difference between the absorption curves of the cells before and after extraction with aqueous methanol ; it indicated a red shift by as much as 40 m/x for this pigment — from 445 mn in methanolic solution, to 485 mju in vivo. The blue peak of chlorophyll c was calculated similarly from the difference between the spectra of cells before and after extraction with 50% methanol; a red shift of 20 m/x was deduced in this way. (The shifts deduced by Tanada for the four blue-violet bands — 8 m/x for chlorophyll a, and 40 mju for fucoxanthol — can be compared with those derived in chapter 22, on pages 705-706, from earlier investigations. The agreement is good for chlorophyll a; for fucoxanthol, the shift found by Tanada — 40 m/x — is about twice that estimated previously by Wassink and Kersten.) When the blue-violet solution bands of all pigments present in the diatoms were shifted as indicated, and superimposed, a composite absorp- tion curve was obtained (see fig. 30.9C). (No attempt was made by Tanada to analyze the region > 620 mju, where all absorption is due to the chlorophylls.) The quantitative agreement is not too good, the composite curve having a higher peak, and a lower valley in the green than the actual absorption cui-ve of the cells. This may be due, at least in part, to scatter- ing— although the cell cur\^e was obtained on Hardy spectrophotometer equipped with an integrating sphere, and with cells immersed in glycerol to reduce scattering. Another likely source of discrepancies is the broadening of the absorption bands in vivo (particular^ of that of fucoxanthol) . The uncertainty implied in the discrepancy between the two curves, had to be accepted in estimating the contribution of the several pigments to total cell absorption at any given wave length. The results of this estimate are shown by fig. 30.9D. It indicates that in the region 500- 550 ran, fucoxanthol takes up most of the quanta absorbed; and yet, fig. 30. 9A shows no drop in yield in this region — except below 520 mju, where absorption by "other carotenoids" sets in. The simplest explanation of the results is that three pigments — chlorophylls a and c and fucoxanthol — are fully effective in photosynthesis (70 — 0.11), while the other carotenoids are either altogether ineffective or have a much smaller efficiency. Tanada found further that 70 of Navicula minima was the same at 1.5, 10, and 20°C., and that 7 was a smooth function of light intensitj' between 0.9 and 6.9 X 10 ~* einstein/cm.^ min., the curvature becoming noticeable >2 X 10~* einstein/cm.2 min. Points obtained in red, orange, blue and green light all fell onto the same slightly curved line, similar to that found for ChloreUa by Emerson and Lewis (fig. 30.5). 117G THE LIGHT FACTOR. III. COLOR CHAP. 30 If fucoxanthol and other carotenoids are to a certain extent active as sensitizers in green plants and brown algae, the mechanism of their par- ticipation is likely to be based on a transfer of energy to chlorophyll, rather than on a direct interaction with the oxidation-reduction system. This hypothesis was suggested by Engelmann over fifty years ago; its first direct confirmation came from the experiments of Button, Manning and Duggar on the excitation of chlorophyll fluorescence by light absorbed by the carotenoids, wliich were described in cluipter 24 (page 814). They were carried out with the same organism {Nitzschia clodcrium) that was also used for the measurement of the quantum yield of photosynthesis. The quantitative results of this study also are in need of re-examination for possible effects of chlorophyll c. We will now describe in brief the experiments in light of indefinite (and probably partly .saturating) intensity, which can be aiiduced in support of the hypothesis that the carotenoids of brown algae actively participate in photosynthesis. Conditions in these organisms appeared somewhat more favorable for coriect guessing than in green plants, because of the absence of chlorophyll b, and consequent enhanced importance of the carotenoids for the absorption in the region between 450 and 500 m/u; liere again the contribution of chlorophyll c requires consideration. As early as 1884, Engelmann found that brown algae (Melosira, Navicula, Pinnul- aria) illuminated by sunlight or gas light, produced the largest amount of oxygen in the green part of the spectrum, and concluded that the "orange pigment" of these algae must participate in sensitization. Fifty years later, Montfort (1934) compared the rate of photosynthesis in white light of a certain standard intensity with its rate in orange-red light. While greeji algae {Ulva lactuca) were equally efficient with both kinds of illuminations, brown algae, Diclyota dichotoma, Alaria and Desmarestia, produced in orange-red light only one half the oxygen liberated in wliite light. Montfort interpreted this as an indication that Phaeophyceae have a liigher relative efficiency in blue and violet light, and attributed this to the presence of fucoxanthol. Later, IMontfort (1936) and his co-worker Schmidt (1937) found that the removal of blue and violet rays from white light depressed the rate of photosynthesis in brown algae {Diclyota and Laminaria) much more strongly than in the green Ulva. They calculated the ratio P/A (photosynthesis per unit ab- sorbed energy) for different colors, using incident light of equal intensity in all spectral regions. An arbitrary selection from the confusing abundance of their material is shown in Table 30. VI. Not all the figures in this table may be strictly comparable, but they show the trend of the results. In green algae, the decrease in P/I from the red to the green and the renewed in- crease in the blue reflected more or less clearly the changes in absorption and in the size of quanta. (These two factors cooperate to make the yield in green light smaller than in the red; but become antagonistic in the blue.) In brown algae — with the exception of Fucus, which Montfort classified as a "xanthophyll alga" (whereas he designated Laminaria and Diclyota as "fucoxanthol algae") — the yields were invariably 30 or 40% higher in the blue than in the red, and would have become twice as high if related to the absorption by chlorophyll alone. In the green, too, the P/I values were relatively higher than they should be according to the absorption of chlorophyll and the size of the ACTION SPECTRUM OF BROWN ALGAE 1177 quanta. Moatfort and Schmidt concluded from these results that, of all the carotenoids, only fucoxanthol is able to assist in the sensitization of photosyntlu^sis, whereas the carotenoids of green algae are inactive (for a criticism of their methods, sec Emerson 1937). Table 30. VI Energy Yield of Algae in Light of Different Colors" (after Schmidt 1937) P/I (green) P/I (blii^ Alga P/I (red) P/I (red) Green ^ „^ Cladophora 0-49 0.80 Ulva laduca 0.46 0.82 Brown Laminaria digitata Strong light (1) 0 1.29 Medium light (Vs) 0.91 1 .32 Weak light Qi) 1-00 1 .23 Phyllilis fascia — 1-40 Dictyota dichotoma 0 . 90 1 . 48 Fucus vesiculosus (brown, but with low fucoxanthol content) 0.38 0.59 " P/I for equal incident intensities. A certain improvement of methods was attempted by Montfort later (1940). The pigments were extracted from the algae and their absorption curves determined (in methanol solution), with results shown in figure 22.45A,B {cf. also Table 22.VIII). These absorption data were applied to the results of Gabrielsen and Steemann-Nielsen (1938), who found that, for equal incident light intensity, the rate of oxygen production by di- atoms is consistently higher in the blue than in the red. (A similar difference was re- ported earlier by Mothes, Baatz and Sagromsky 1939.) The difference is particularly strong in low light, as shown by Table 30. VII. The table describes a perfectly under- Table 30.VII Ratios of Rates of Photosynthesis by Diatoms in Blue and Red Light (after Gabrielsen and Steemann-Nielsen 1938) Incident light, PjblueJ Incident light. ^r^.v cal/cm.2 sec. P (red) cal/cm.2 sec. P (red) 0.25 1.8 1.7 1.4 0.5 l.fi 2.4 1.3 1.0 1.5 3.7 1.1 standable transition from the conditions at low light intensities (where the yield is pro- portional to absorption, and may therefore be higher in the blue than in the red, if the absorption in the blue is so nmch sti'onger as to overbalance the laiger size of the quanta) to the conditions in strong (saturating) light, where the rate must be (and apparently is) independent of wave length. Montfort (1940) preferred, however, to average all the ratios in Table 30.VII and concluded that a ratio of 1.5 between the rates in blue light and red light indicates an active participation of carotenoids in photosynthesis. Using the absorption curviis of methanolic extracts he calculated that the energy conversion rate in blue light is 1.03 times larger than in red light, referred to absorption by all pig- 1178 THE LIGHT FACTOR. III. COLOR CHAP. 30 ments, and 2.46 times larger referred to chlorophyll alone; while, according to the size of the quanta, it should be smaller by a factor of 0.68. In the case of brown algae {Laminaria digitata) a similar calculation (based on Mont- fort's own measurements) gave energy yield ratios (related to the combined absorption by all pigments) of 0.86 and 1.18 (for two different combinations of filters), as compared with the quantum size ratios of 0.64 and 0.77, respectively, again indicating a higher quantum yield in the blue-violet than in the red. At their face value, Montfort's figures appear to indicate that the carotenoids of the diatoms and brown algae are several times more efficient as sensitizers than chlorophyll! Mothes, Baatz, and Sagromsky (1939), Baatz (1941) and Sagromsky (1943) have described observations of the rate of photosynthesis in filtered red and blue light of equal intensity (in energy units). They found for these two rates, a ratio of 1 : 1.2 in the diatom Chaetoceras simplicia centrosperma, as against 1 : 0.7 in two unicellular green algae, and attributed the relatively better utilization of blue light by the brown cells to the presence of fucoxanthol. These experiments were more satisfactory than those of Mont- font in that imicellular algae were used rather than thick thalli; but it was equally un- satisfactory in the use of broad spectral regions, and even less satisfactory in the absence of any absorption measurements, which would permit approximate allocation of ab- sorption to the several pigments. Mothes and co-workers pointed out the difficulty of the latter problem, caused by the difference of the carotenoid spectrum in vivo from that in vitro. (This difference is clearly indicated by the change of color from brown to green caused by placing brown algae in hot water — a treatment which, they assumed, disrupts the molecular association of carotenoids with chlorophyll and proteins.) 6. Quantum Yield and Action Spectrum of Red and Blue Algae. Role of Phycobilins The history of our knowledge of the role of phycobilins in the sensitiza- tion of photosynthesis in red and blue algae is similar to that of the role of carotenoids in brown algae. Here too we find the — we now know, correct — guess by Engelmann, made as early as 1883, that the phycobilins are active sensitizers of photosynthesis, as well as a series of vague and indecisive observations and calculations of several authors, mostly tending to confirm this guess, and finally, quantitative analyses of the quantum yield as a function of wave length, carried out by Emerson and Lewis (1941, 1942), and Haxo and Blinks (1950), which brought convincing confirmation of Engelmann's concept. As in the previous section, we will discuss the more recent and reliable experiments first. Table 29. Ill showed that Emerson and Lewis (1941), in comparing the quantum yield of different plants in yellow sodium light, found no difference between green plants and the blue-green algae Chroococcus, al- though in the latter, far more than half the absorption in the yellow re- gion was due to phycocyanine. This result of a preliminary observation was in itself a more convincing proof of the activity of phycobilins in photo- synthesis than could be derived from all the extensive earlier discussions of this problem. It was confirmed and amplified by the same authors in ACTION SPECTRUM OF RED AND BLUE ALGAE 1179 1943, by a systematic investigation of the relation between wave length and quantum yield in this Cyanophycea. Chroococcus cells are somewhat smaller than Chlorella (2.5 ju in diame- ter) and are surrounded by a gelatinous sheath. They scatter less light 0.10 008 Q _l ill > 006 3 < O 004 0.02 , .X' J Observed Colculoted *-♦-..,. 400 440 480 520 560 600 640 WAVE LENGTH, m^ _L 680 720 Fig. 30.10A. Quantum yield of Chroococcus photosynthesis (after Emerson and Lewis 1942). SoHd line is drawn through experimental points, values obtained in different runs being distinguished by different characters. Broken line shows expected dependence of quantum yield on wave length, on assumption that -yield for light absorbed by chloro- phyll and phycocyanine is 0.08 at all wave lengths, and light absorbed by carotenoids is not available for photosynthesis. 100 Direct From Photosynthesis 10 mm cells/ml 08 mm. ceils/ml. J L 400 700 500 600 WAVE LENGTH, mp. Fig. 30.10B. Comparison of action spectrum (broken curve) and ab- sorption spectrum of Chroococcus (after Emerson and Lewis 1942). than Chlorella cells (probably because of the absence of chloroplasts ; cf. chapter 15, page 355). This makes the determination of the absorption curve easier, and leads to better agreement between the absorption curve of the intact cells and the curve constructed by the combination of the (appropriately shifted) absorption curves of the extracted pigments (illus- 1180 THE LIGTTT FAf'TOU. Til. f'OLOU CHAP. 30 trated by fig. 22.485). The main remaining differences between the "cell spectrum" and the combined extract spectrum is the apparent broadening of the red chlorophyll band in the intact cells, and a somewhat lower absorp- tion of the latter at X < 510 mju. Chroococcus cells were used in carbonate buffer (85% 0.1 Af NaHCOs + 15% 0.1 Af Na2C03). These algae can live without potassium, but not without sodium. The method of determination of y was the same as in the work with Chlorclla. (Bands 6-10 mn ^\'ide were used in the rod and 15-20 m/x wide in the blue-violet; photosynthesis and respiration were measured in alternating 10 minute periods of darkness and light, the value of P hemg derived from the rate of oxygen production in the second half of the illumination period.) Both "dense" (fully absorbing) and "thin" (partially absorbing) suspensions were used. The quantum yield of photosynthesis in Chroococcus as a function of wave length is shown in figure 30.10A. As in Chlorella, y is approximately constant between 570 and 690 nifi (aside from a slight flat maximum at 680 m/x) — despite the fact that in Chlorella all absorption in this region is due to chlorophyll, whereas, in Chroococcus, more than half the total absorption in the region between 560 and 650 m^i must be attributed to phycocyanin. Judging from figure 22.49, the absorption by phycocyanin at 600 m/u should be at least six times as large as that by chlorophyll; but the quan- tum yield is the same as at 660-680 m/x, where chlorophyll accounts for practically all absorption. Thus, the photosynthetic efficiency of phyco- cyanin in Chroococcus must equal that of chlorophyll (the maximum possible difference being of the order of 10-15%). Another way of repre- senting the same results is shown in figure 30.10B. Here, the absorption spectrum of a thin suspension of Chroococcus cells is compared with the quantized action spectrum of photosynthesis. The close parallelism be- tween the two curves in the region above 570 mju shows the approximately equal availability for photosynthesis of the light absorbed by both chloro- phyll and phycocyanin. Particularly convincing are the two separate maxima shown by both curves near 620 and 670 m^, which must be at- tributed to phycocyanin and chlorophjdl, respectively. The large dis- crepancy in the region 420-550 m/x indicates the inefficiency (or relatively low efficiency) of the light absorbed by the carotenoids. The dotted line in figure 30.10A shows, however, that assuming complete inefficiency of the carotenoids leads to an underestimation of the yield between 450 and 550 m/x; the best agreement between measured and calculated yields can be obtained by assuming a quantum yield y = 0.08 for the light absorbed by chlorophyll and phycocyanin, and y = 0.016 for the light absorbed by the carotenoids. The action spectra of several species of red algae were studied by Haxo ACTION SPECTRUM OF RED AND ELITE ALGAE llSl Ulva taeniata — ^ Thollus obsorption Action spectrum c 80 30 40 •^. 20 z o P 0 a. a: o "" in m < E Delesseria decipiens —^ Thollus absorption Action spectrum Aqueous extract UIJ Porphyra naiadum 80 — > \ absorption Absorption of ■ ^•^/ ~v./^^^ r\ aqueous 60 - ■■■ A / \ extract 40 - // V^^.. \ ■ '' \ \ 20 \ J ""~'\\v 0 1 1 -i—l . — 1 1. ..!_., 111. >^--.l , G 100 80 60 40 go- / Porphyra perforata ( Red portion) Absorption Action spectrum -I — I — I — . — I I I . I I I . J Ll^J l_ 400 480 560 640 720 WAVE LENGTH, m/i CO''odesme ^— Thallus absorption Action spectrum B Porphyra nereocystis — — Thallus absorption Action spectrum Extracted phycoeryfhrin D Porphyra perforata (Slate grey portion) Thallus absorption Action spectrum 400 480 560 640 WAVE LENGTH, m/x 720 Fig. 30.11. Action spectra after Haxo and Blinks (1950). 1182 THE LIGHT FACTOR. III. COLOR CHAP. 30 and Blinks (1950) with very striking results. The polarographic method (page 850) was used, with occasional checks by oxygen determination and manometric measurements. Light intensities used were low enough (about twice the compensating intensity) for the results to approximate the maxi- mum quantum yields. In the green alga Ulva taeniata (fig. 30. 11 A) the action spectrum was found to follow closely the absorption spectrum, with the exception of a yield deficiency of up to 100% in the far red (>700 m/x), already noticed by Emerson and Lewis (c/. fig. 30.3), and of a certain yield deficiency around 480 van, also noticed before, and interpreted as evi- dence of partial inactivity of the carotenoid pigments in green cells. It is, however, to be noted that the yield deficiency disappears at 415 m/i, al- though a considerable proportion of light (over 50% in extracts) must be absorbed in this region by carotenoids. In the brown Coilodesme (fig. 30.1 IB), the action spectrum also paral- leled rather closely the absorption spectrum with a moderate deficiency (of up to 20%) in the region of strong carotenoid absorption. In the red algae Delesseria decipiens (fig. 30.11C), Porphyra nereocystis (fig. 30.11D), Porphyra naiadum (fig. 30. HE) (purple, indicating relatively high content of phycocyanin) and Porphyra perforata (fig. 30.1 IF, G) (a slate-green vegetative section containing mainly phycocyanin, and a red carposporic section) the action spectra Avere strikingly different from the absorption spectra. They showed unmistakable maxima corresponding to the ab- sorption peaks of phycoerythrin, at 500 and 565 m/x, and also to those of phycocyanin at 620 m^u (fig. 30. HE, F), but only very little of the chloro- phyll maxima in the red as well as in the violet. The quantum yield was estimated to be the order of 0.06 in the phycoerythrin bands, and as low as 0.02 in the chlorophyll bands. Similar results were obtained with several other Bangiales and Florideae. The absorption by carotenoids appeared to be as little effective as that by chlorophyll. A slight increase in activity shown by some species at X = 440 mju, coinciding as it did with an increase in the absorption of the aqueous phycobilin extract, could not be inter- preted as sign of photosynthetic activity of the chlorophylls or the caro- tenoids. These unexpected results indicate that in contrast to all other plants direct sensitization by chlorophyll plays only a subordinate role in at least some of the red algae, and that photosynthesis in them is sensitized pri- marily by phycobilins. If this is the case, it would seem unlikely that the energy quanta absorbed by the phycobilins are transmitted to chlorophyll, since how could indirectly produced excitation of chlorophyll be more effective than excitation due to energy absorbed directly by chlorophyll? Rather, these results would seem to indicate that the phycobilins are sensitizers of photosynthesis in their own right, and that their presence ACTION SPECTRUM OF RED AND BLUE ALGAE 1183 may perhaps even make that of chlorophyll superfluous, although so far no chlorophyll-free red algae have been found. We Avill see below, however, that a different — and no less striking — interpretation of these results is possible. In one species of Iridophycus, which bleaches to almost green color at high tide, a much greater participation of chlorophyll in photosynthesis was noted in rough experiments. Haxo and Blinks said that, in contrast to Emerson and Lewis's results on Chroococcus, they found only a weak chlorophyll activity also in two blue- green algae, Anaboena and Oscillatoria, whose action spectra were similar to those of Porphyra perphoraia (fig. 36.11F). They suggested that culture conditions may affect the relative activity of different pigments in algae of the same class or even the same species. Haxo and Blinks noted that the saturation rate of photosynthesis of Delesseria is the same in blue light (565 niju) as in red light (672 ran) '• This seems to indicate that the enzymatic mechanism of photosynthesis — or, at least, the rate-limiting enzymatic reaction — is the same whether the quanta are absorbed by a phycobilin or by a chlorophyll. In chapter 24 (p. 815) we desciibed the fluorescence studies of French et at. (1951) and Duysens (1951) that indicated effective transfer of excita- tion energy in red algae from carotenoids and phycobilins to chlorophyll a, and (in certain of them) from chlorophyll a to d (despite the low con- centration of the latter). If one assumes that transfer to chlorophyll d constitutes a "leak" which makes energy unavailable for photosynthesis, the results of Blinks and Haxo become understandable. The question remains why energy transferred to chlorophyll a from phycobilins is not also lost to chlorophyll d but remains available for photosynthesis. It w'as noted on p. 815 that this energy stays with chlorophyll a long enough to cause its fluorescence (Avhile most of the energy absorbed by chlorophyll a itself causes the fluorescence of d). Two suggestions were made there as to the possible reasons for this difference in the fate of excitation energy; but further study is needed for a convincing explanation. Duysens' observations provide the strongest argument at present in favor of assuming that chlorophyll a is the one pigment directly par- ticipating in photosynthesis, and that not only the carotenoids, but also the phycobilins, sensitize photosynthesis by transferring their excitation energy to chlorophyll a. Another argument supporting this view is the observation of French and co-workers (1951) that the yield of fluorescence of the phycobilins in algae shows none of the induction effects and of the peculiar changes with hght intensity which were discussed at length in chapters 24 and 28 (part B) and are indicative of an intimate relationship between chlorophyll and the chemical processes of photosynthesis. 1184 THE LIGHT FACTOR. III. OOLOli CHAP. 30 Beside the measurements of Emerson and Lewis and of Blinks and Haxo, all earlier observations on the role of phycobilins in photosynthesis have only the weight of corroborative evidence. Most of this evidence pertains to red algae, and has been gathered in connection with Engelmann's theory of "complementary chromatic adaptation" of these algae to the blue-green light that prevails under the sea. (Obviously, this color adaptation can only be useful to the algae if the light absorbed by the red pigments can be used for photosynthesis.) Because of the absorption of red and blue- A'iolet light by water, a full utilization of the central part of the visible spectrum — which is only insufficiently absorbed by chlorophyll — is of vital importance for the plants living deep under the sea. This considera- tion was the basis of the theory of chromatic adaptation, developed by Engelmann in 1884. This subject was almost lost sight of in the first quar- ter of the new century, while new methods of quantitative study of photo- synthesis by green leaves were being developed by Blackman and co- workers and by Willstatter and Stoll. Later, Warburg made the green Chlorella cells the favorite subject of photosynthetic studies. In the work on green plants, the presence of accessory yellow pigments was con- sidered to be scarcely more than a nuisance. These pigments were not prominent enough — both in concentration and in the part they took in light absorption — to make them a desirable subject of independent study; but their presence interfered with the quantitative study of chlorophyll- sensitized photosynthesis in the short-wave region of the visible spectrum. The fact that blue and red algae offer a much more promising field for the study of the part played by the "accessory" pigments in photosynthesis was almost forgotten. Engelmann had noticed, however, as early as 1883 (using motile bacteria for the oxygen determination) that the maxi- mum of the photosynthetic efficienc}^ of red algae (CaUithamnion and Cer- amium) lay in the green part of the spectrum, and that of blue algae (Oscil- latoria and Nostoc), in the yellow. As in the case of green plants, the posi- tion of the maximum of photosynthesis coincided roughly with that of the maximum of light absorption. A year later (1884), Engelmann described a "microspectrophotometer," by means of which he was able to show that the parallelism between the absorption spectra and the "photosynthetic action spectra" of the colored algae is quantitative. He concluded that all pigments that contribute to light absorption by the algae also con- tribute to photosynthesis, and expressed this result by the equation -E'abs. = ■E'assim. (£" Standing for energy), w^hich was a direct challenge to the concept of the exclusive sensitizing role of chlorophyll in photosynthesis. One of the developments of Engelmann's theory — ^the concept of "chro- matic adaptation" as a factor determining the composition of the pigment system in plants — has been discussed in chapter 15. Here, we are con- ACTION SPECTRUM OF RED AND BLUE ALGAE 11S5 cemed with the other aspect of the same phenomenon — chromatic adapta- tion as a factor enabhng plants to make better use of available light energy. It was mentioned in chapter 15 that Oltmanns (1893) and others (most recently, von Richter 1912, and Sargent 1934) objected to Engelmann's theory (as well as to its extension by Gaidukov to color changes induced artificially in blue-green algae) and insisted that algae responded only to changes in light intensity, and not color. \m\ Richter (1912) raised a further objection and asserted that "chro- matic adaptation" could not achieve the purpose that was suggested by Engelraann, because phycobilins do not act as sensitizers in jjliotnsynthe- sis. Although, in comparing the ratio of the photosynthetic productions of the green alga Ulva lactuca in red and green light with the corresponding ratios for Plucaniiiim, CallUhamnion, Delesseria and other red algae, von Richter could not help confirming the results of Engelmann and finding that red algae are two or three times as eflicient in green light as the green algae, he nevertheless denied that this proved the photosynthetic activity of the red pigments. He pointed out that similar differences are obtained also by changing the intensity of light, and suggested that the red algae utilize green light better not because of its wave length but because it is only weakly absorbed by chlorophyll — and red algae are adapted to weak light. However, the views of von Richter have not been confirmed by later investigators. Wurmser and Ducleaux (1921) compared the photosyn- thetic eflSciencies of red and green fronds of Rhodymenia palniafa and Chondrus crispus and found that the red varieties give yields two or three times greater than the green ones. Wurmser (1921) compared the photo- synthesis of the green alga Ulva lactuca with that of the red alga Rhodymenia palmaia in red, green and violet light. He found that, if the rate in red is taken as imitj^ that in green is 0.24 in Ulva and 0.49 in Rhodymenia, and that in violet 0.81 and 0.16, respective^ (for equal intensity of incident light). Thus, the red algae are more eflScient in the green, but le::s ef- ficient in the violet than the green ones. Wurmser pointed out that, even if von Richter's intensity effects are real, this does not mean that color ef- fects are only indirect consequences of changes in absorption intensity. Similar results were obtained by Harder (1923), who concluded that both intensity adaptation and color adaptation are real phenomena. He, as well as Ehrke (1932), interpreted the result of the rate measurements with red algae as indicating active participation of the phycobilins in photo- synthesis, and the same conclusions were reached by Montfort (1936) and Schmidt (1937), who found the spectral maximum of the efficiency of phyco- cyanin algae in yellow light, and that of phycoer>'thrin algae in green light. 1186 THE LIGHT FACTOR. III. COLOR CHAP. 30 Levring (1947) determined "action spectra" of photosynthesis of a number of marine algae in filtered sunlight. (Only qualitative results can be expected from such measurements, because of the relatively high light intensity and the relatively strong absorption of the thalli). He found evidence of particularly strong photosynthetic efficiency (high ratio yield/ absorption) in green light in ten species of red algae, and concluded that the red phycobilin pigment is as active (if not more active) than the green chlorophyll. Combining these results with those of his measurements of spectral distribution of light in different depths, he concluded that because of the presence of the red pigment, the Rhodophyceae utilize the blue-green light deep under the sea better than the Chlorophyceae, as postulated by the Engelmann-Gaidukov theory. He agreed, however, that adaptation to low light intensity is an alternative way of adjustment to life in great depths; an important element of it is low respiration. Thus, even more uniformly than in the case of brown algae, the crude observations on the relative efiiciency of photosynthesis of red algae in light of different color support the assumption that the accessory pigments of these organisms are active sensitizers in photosynthesis, and that Engel- mann's theory of chromatic adaptation was fundamentally correct. And one may ask oneself, how could it have been otherwise? Would it not be strange if the appearance of orange or red pigments in deep-water algae would be only a coincidence, and these pigments were helpless in performing the task so obviously set to the plants by the character of the "light field" in which they live — to catch and utilize for their maintenance and propaga- tion radiations in the middle of the visible spectnim, which are the only ones to reach them in some intensity? It may be argued that not all deep-water algae are red, some green algae being encountered in great depth. In other words, algae can sur- vive without phycobilins in the greatest depths where life occurs. How- ever, this in itself is not a convincing argument against Engelmann's theory. Algae could adapt themselves to great depths in two ways: by reducing respiration to a level permitting growth even in extremely Aveak, and weakly absorbed light; and by adjusting their pigment systems to enhance light absorption. The fact that the first adjustment has been sufficient for some green species does not invalidate the hypothesis that red algae have also used chromatic adaptation for the same purpose. Another objection to Engelmann's theory is that many red algae live on or near the surface and that phycobilins are found in blue-green algae, which are surface organisms. It is known, however, that red algae often tend to lose their phycobilin and become green when exposed to sunlight (c/. Vol. I, Chap. 15); and even if many of them (as well as the blue-green algae) apparently find their phycobihn content useful, or at least not harm- I ACTION SPECTRUM OF PURPLE BACTERIA 1187 ful, even on the surface, this does not prove that phycobiHns are not pig- ments primarily intended to permit photosynthesis deep under the sea. One may speculate — particularly in the light of Blinks' experiments — whether photosynthesis with phycobilins may not be an older process than photosynthesis with chlorophyll; perhaps, the development of the green pigment and its substitution for the phycobilins have been the product of a later development, in which plant life, originating in the depths of the ocean, migrated to the surface and finally spread overland. 7. Action Spectrum of Purple Bacteria Engelmann found, in his fundamental work on photosynthetically ac- tive bacteria (1888) that, if a spectrum was thrown on their cultures, they developed only in the absorption bands of the green pigment (which we now call bacteriochlorophyll). Purple bacteria also contain numerous carotenoids, with absorption bands clearly separated from those of bac- teriochlorophyll (c/. fig. 22.21). French (1937) found that the action spec- trum of Streptococcus varians (as determined by the rate of consumption of hydrogen) sho\\Ti in figure 30.12 closely parallels its absorption spectrum in the yellow and red part of the spectrum but does not show maxima in the green or blue that correspond to the absorption bands of the carotenoids (c/. Table 30. VIII). French concluded that the red carotenoid pig- ments of purple bacteria are photosynthetically inactive. It must be noticed that, from the spectroscopic point of view, the conditions in purple o < 1/5 if) < O LJ < ; inhil)ition of photosynthesis by ultraviolet, 1153. See also Emerson, II. and Oppenheimer, R. J.: Energy transfer between pigments, 758-759; enhance- ment of phycobilin fluorescence by cell destruction, 801, 816; fluorescence yield of Chroococcus, 813. Aronoff, F., and Calvin, M.: Spectra of hexaphenjd porphins, 623-625. Asana, R. I. See Dastur, R. H. Aufdemgarten, H.: Thermal conductivity measurements of CO2 exchange in photo- sj'nthesis, 853. Auerbacher, F. See Hagenbach, A. B Baatz, I. See Mothes, K. Baehiach, h]., and Dh^re, C: Fluorescence of diatoms, 806. Baldwin, W. C. G. See Mecke, R. Ballai-d, L. A. T.: Inhibition of photosynthesis by excess CO2, 903. Barker, H. A.: Light curves of diatoms, 969; CO2 curves of purple bacteria, 893, 904. Bazyrina, K., and Chesnokov, V.: CO2 fertilization, 902. See aho Chesnokov, \'.: Kost>Thev, S. P. Beadle, B. W. See Zscheile, F. P. Berthold, G.: No chromatic adaptation?, 995. Biermacher, O.: Absorption spectrum of chlorophyll a and b in different solvents, 637- 639. See also Dhere, C. Blackman, F. F.: Limiting factors, 859-861; light curves and theory of limiting factors, 965. (with Matthaei, G. L. G., and Smith, A. M.): Light curves of photosynthesis, 964. and Smith, A. M.: CO2 concentration as limiting factor, 891 ; CO2 curves of leaves, 892, 894, 904. Blagoveshchenskij, A. V.: High rate of photosynthesis of mountain plants, 997, 1001. Blinks, L. R. See Haxo, F. T. * Complete author index will be found at the end of Volume II, Part 2. 1193 1194 AUTHOR INDEX Bohning, R. H.: Midday depression and CO2 content in air, 875, 876; constant photo- synthesis of land plant under constant conditions, 879; decline of photosynthesis of shade plants in strong light, 989. Bose, J. C: Energy conversion by plants under natural conditions, 1003-1004. Boysen-Jensen, P.: Light curves of Fagus and Sinapis, 966; light curves of heliophilic and umbrophilic plants, 987; maximum rates of photosynthesis, 991; photo- synthesis of leaves under natural conditions, 998. and Miiller, D.: Constant photosynthesis under field conditions, 876; light curves of ferns and lichens, 966; compensation points of trees, mosses, and lichens, 982; light curves of umbrophilic and heliophilic plants, 987, 988, 989; photosynthesis under natural conditions, 997-1001. Brackett, F. S.: Empirical equation for light curves, 1046. See also Hoover, W. H. Brewster, D.: Chlorophyll fluorescence, 740. Briggs, G. E.: Yield measurement with leaves, 1118; yield in blue-violet lower than in yellow and green, 1148. Brilliant, V. A.: Adaptation phenomena, 873. Brown, A. H.: Tracer experiments on respiration in light, 1 117. Brown, H. T., and Escombe, F. : Rate of CO2 uptake by leaves from air, 912; theory of CO2 diffusion through stomata, 913; rate of photosynthesis under natural condi- tions, 1000; energy conversion under natural conditions, 1003-1004. Brown, W. H., and Heise, G. W.: Criticism of Blackman's law of limiting factors, 861, 965. Buder, J.: Fluorescence of bacterioviridin, 811. Buhr, H. See Guttenberg, H. von. Burk, D., and Lineweaver, H.: Calculation of carboxylation equilibrium from CO2 curves of photosynthesis, 935; kinetic analysis of light curves, 1046. .See also Warburg, O. Burns, G. R. : Absorption of light by conifer needles, 684 ; no photosynthesis in conifers below 465 m/x, 1153; photosynthesis up to 740 niyu, 1158, 1164-1165. Burr, G. O. See Miller, E. S. Calvin, M. See Aronoff, F.; Lewis, G. N. and Dorough, G.: Long-lived infrared fluorescence of chlorophyll, 748, 795. Carr. See Loomis, W. Chesnokov, V., and Bazyrina, K.: CO2 concentration threshold of photosynthesis, 907; CO2 exhaustion effects in land plants, 908; gas exchange balance at low CO2 concentrations, 985. See also Bazyrina, K.; Kostychev, S. P. Coblentz, W. W. See Stair, R. Comar, C. L. See Zscheile, F. P. Daniels, F. See Magee, J. L.; Manning, W. M.; Petering, H. G. Danilov, A. N.: Cooperative and antagonistic effects of different colors on photo- synthesis, 1166-1167. Dastur, R. H. (with Asana, R. D., and Gunjikar, L. K.): Has polarization an effect on photosynthesis?, 1147. (with Samant, K. M., Mehta, R. Y., Solomon, S., Kanitkar, U. K., and Rao. M. S): Cooperative and specific effects of different colors in photosynthesis, 1167 AUTHOR INDEX 1195 Decker, J. P. See Kramer, P. J. De Witt, T. W. See Magee, J. L. Dezelic, M. See Stern, A. Dh6re, C. (with Fontaine, M., Raffy, A., and Biermacher, O.): Fluorescence spectrum of chlorophylls a and h, 740-746, 748; of chlorophyll c, 747; of protochlorophyll, 748; of phycobilins, 799, 801; of algae, 806, 807, 808, 809, 811. See also Bach- rach, E. Dorcas, M. J. See Osterhout, W. J. V. Doring, B. See Harder, R. Dorough, G. See Calvin, M. Dorrestein, R. See Wassink, E. C; Katz, E. Drautz, R.:- "Shade" and "sun" regions in a leaf, 873; interpretation of midday de- pression, 874. Ducleaux, J. See Wurmser, R. Duggar, B. M. See Dutton, H. J.; Manning, W. M.; Moore, W. E.; Petermg, H. G. Duntley, S. C.: Absorption and scattering in inhomogeneous systems, 712-713. Dutton, H. J., and Manning, W. M.: Action spectrum of diatoms; sensitizing activity of fucoxanthol, 1169-1172; distribution of light energy among pigments in di- atoms, 725-726; quantum yields in Nitzschia dosterium, 1096. , Manning, W. M., and Duggar, B. M.: Chlorophyll fluorescence sensitized by carotenoids in diatoms, 814-815. Duysens, L. N. M.: Excitation transfer from chlorophyll 6 to a in solution, 790; spectro- photometry of fluorescence of red algae and purple bacteria, 806, 807; fluores- cence of chlorophyll d (?) in Porphyra excited by energy transfer from a, 811-812, 815-816; chlorophj'll a fluorescence excited by energy transfer from 6 in Chlorella, 809; bacteriochlorophyll fluorescence in Chromatium and Rhodospirillum sensi- tized by carotenoids, 810; two kinds of chlorophyll complexes in red algae?, 815- 816; chlorophyll a fluorescence excited bj^ energy transfer from phycobilins, 815- 816; inactivity of chlorophyll a in red algae due to energy transfer to chlorophyll d?, 1183. E Egle, K.: Absorption spectrum of chlorophyll in different solvents, 637-639; absorption and reflection of leaves in infrared, 692. See also Seybold, A. Ehrke, G.: Compensation points of green, brown and red algae, 983; photosynthesis of red algae in colored light, 1185. Ehrmantraut, H. C, and Rabinowitch, E.: Quantum yield of Hill reaction in Chlorella and chloroplasts, 1130-1131. Eichhoff, H. J.: Light curves of Chlorella suspensions of different density, 1008-1009; quantum yield in Chlorella high in near infrared?, 1097-1098. Emerson, R. : Light curves of normal and chlorotic Chlorella, 967. See also Whitting- ham, C. P. and Arnold, W.: CO2 curves of Chlorella, 899. and Green, L.: Photosynthesis in Chlorella insensitive to pH changes, 835; C0» saturation in Chlorella depends on concentration of CO2 molecules only, 891 ; (^02 curves of Chlorella, 893, 896; of Gigarlina, 893, 904, 906, 908; light curves of Gigarlina, 967; maximum rate of GUjartina, 991, 992. and Lewis, C. M.: Transmission spectra of Chlorella and Croococcus, 691, 692, 699, 705; carotenoid bands in algae, 706; phycobilin bands in Chroococcus, 707-708; relative intensities of absorption bands in Chlorella and Chroococcus, 708; en- hanced absorption in far red in vivo, 715; distribution of energy among pigments 1196 AUTHOR INDEX in Chlorella, 722-723; in Chroococcus, 728-729; linear range of liglit curves of Chlorella, 980, 981; CO2 "burst" in quantum yield measurements, 1086, 1087- 1088, 1091-1094; quantum jdelds of photosj'nthesis in different species, 1094- 1095, in the blue-green alga Chroococcus, 1096, in monochromatic light, 1148-1152; carotenoids in Chlorella sensitize photosj'nthesis with lower efficiency than chlorophj'll, 1149-1150, 1151-1152; decline of photosj'nthesis in Chlorella and Chroococcus above 680 mju, 1154; monochromatic light curves of Chlorella, 1159; action specti'um of Chroococcus, full activity of phycobilins, 1178-1180. and Nishimura, M. S. : Criticism of Warburg's quantum yield determinations, 1101-1104. Engelmann, T. W.: Chromatic adaptation, 994-995; a second spectral maximum of photosynthesis, 1143-1144; red pigments of leaves inactive in photosynthesis, 1165; evidence of photosynthetic efficienc}'^ of fucoxanthol in brown algae, 1176; of phycobilins in red algae, 1178, 1183-1184. Escombe, F. See Brovn\, H. T. Evstigneev, V. B., Gavrilova, V. A., and Krasnovsk}', A. A.: Effect of oxygon, alcohol and watci' on absorption spectra of chloroph}!! and pheophytin in nonpolar solvents, 648; effect of solvents on chlorophyll fluorescence, 771-772; activation of fluorescence of chlorophyll, pheophytin and phthalocyanin by oxygen and water, 788. Eymers, J. G., and Wassink, E. C: Effect of thiosulfate concentration on CO- reduction b}' Chromaliutn, 946; linear range in purple bacteria, 980, 981; quantum yield of CO2 reduction by Thiorhodaceae, 1127. Ewart, A. J. : Inhibition of photosynthesis by excess light, 964. Filzer, P. : Periodicity of photosynthesis in detached leaves, 874. See also Harder, R. Fong, J. See Pratt, R. Fontaine, M. See Dhere, C. Forster, T.: Transfer of electronic energy between molecules, 758-759; as mechanism of self-quenching, 759, 774; as mechanism of quenching by admixtures, 785. Franck, J.: "Narcotization" of chlorophyll by metabolites as cause of enhanced fluores- cence, 824-826; effect of CO2 on fluorescence as evidence of CO2 association with chlorophyll, 941 ; effect of CO2 and reductants on fluorescence in plants and bacteria caused by internal narcotization, 942-943, 950; narcotization as kinetic factor in photosjai thesis, 1033, 1034, 1041-1043; interpretation of light curves of fluores- cence, 1070, 1071, 1076, 1077, 1078; mechanism of the CO2 burst, 1086-1087; interpretation of Warburg's and Kok's measurements as indicating partway re- versal of respiration by light, 1117. See also Shiau, Y. G.; Weller, S. and French, C. S.: Photoxidation in CO2 deprived leaves, 1166. , French, C. S., and Puck, T. T.: Fluorescence of leaves and algae, 806; its relation (() photosynthesis, 819, 824; effect of CO2 concentration on fluorescence, 940; liuorescence-light curves of Hydrangea, 1048-1049; effect of CO2 on these curves, 1051; of temperature 1055-1056; of cyanide 1057-1058. and Herzfeld, K. F.: Nondissociable acid as first product of CO2 fixation in photo- synthesis, 917, 918, 919, 927-930; kinetic model of photosynthesis, 1018, 1021- 1022, 1032, 1035, 1036, 1038, 1040; energy requirements and quantum yield of ])liot(isynthesis, 1089-1090; theoretical vs. maximum observable quantum yield, 1139. and Levi, H.: Quenching of chlorophyll fluorescence by oxygen, 778-779; by benzidine and KI, 780. and Livingston, R.: Mechanism of self-quenching, 755-760, 774. AUTHOR INDEX 1197 (with Pringsheim, P., Pollack, M., and Terwood-Lad, D.): Quenching of phos- phorescence a sensitive way of measuring O; production in photosynthesis, 851. Freeland, O. R.: CO2 penetration through cuticle, 911. French, C. S.: Absorption spectrum of bacteriochlorophyll, 616-617; spectra of purple bacteria, 692, 693, 702; carotenoid bands in purple bacteria, 707; effect of H2 pressure on photoreduction of CO2 in bacteria, 944-945; sigmoid shape of light curves in bacteria, 948, 964; quantum yield of purple bacteria, 1124-1125, 1126- 1127, 1332; action spectrum of purple bacteria, inactivity of carotenoids?, 1186- 1188. See also Franck, J.; Rabideau, G. S. and Koski, V. M.: Absorption spectra of phycobilins, 664-665; absorption band of protochlorophj'll in leaves, 705; spectrophotometric study of fluorescence of phycobilins, 799-801; protochlorophyll fluorescence in partially green leaf, 811; light curves of phycobilin fluorescence in red algae linear when those of chlorophyll curved, 1051. and Rabideau, G. S. : Quantum yield of Hill reaction in chloroplasts and Chorella cells, 1094, 1128-1130. (with Van Norman, R. W., Macdowall, F. D., and Koski, V. M.): Spectrophoto- metric study of fluorescence of blue-green and red algae, 806, 807, 809, 811-812; phycoerythrin-sensitized fluorescence of chlorophyll in red algae, 815. Fuller, H. J. : CO2 concentration near ground, 902-903. Gabrielsen, E. K.: CO2 compensation point, 899; re-utilization of respiratory CO2, 900- 901; light curves of Sinapis, 966; of sun and shade leaves of Fraxtnus, 967; of Triticum, 967; high saturating intensity of sun leaves, 987; rate of photos3-n- thesis of Sinapis under natural conditions, 998; quantum yield measurements on leaves, 1188-1189; equal saturation rates in light of different colors, 1145; con- tribution of ultraviolet to photosj^nthesis in sunlight, 1153; monochromatic light curves, 1161-1162; screening effect of red leaf pigments, 1165. and Steemann-Nielsen, E.: Photosynthesis of diatoms in blue and red light, 1177. Galanin, M. D. See Vavilov, S. I. Gavrilova, V. A. See Evstigneev, V. B. Geiger, M.: Closure of stomata as cause of midday depression, 875. Gessner, F.: Absence of midday depression in aquatic plants, 876; photosynthesis of aquatic plants constant if medium effectively renewed, 878, 879; leaf shape and CO2 exhaustion effects, 905; light curves of higher aquatic plants, 967; of shade and sun plants, 988; maximum rates of aquatic plants, 991, 992, 1002; no light inhibition of shade-grown Elodea, 994. Giltay, E.: Rate of photosynthesis of tropical plants, 1001. Green, L. See Emerson, R. Greenfield, S. S.: Light curves of Cu + + and Ni + + inhibited Chlorella, 975. Gunjikar, L. K. See Dastur, R. H. Guttenberg, H. von, and Buhr, H.: Carbohydrate accumulation as cause of midday depression, 875. H Hagenbach, A. : Red shift of chlorophyll bands in leaves, 697. , Auerbacher, F., and Wiedemann, E. : Visible and ultraviolet absorption spectra of chlorins and porphins, 606, 629. Harder, R.: Modified law of limiting factoi-s, 862; adaptation to light intensity and temperature, 873; time course of photosynthesis under constant conditions, 877- 1198 AUTHOR INDEX 878, 994; CO2 curves of aquatic plants, 892, 895, 897-898, 904; light curves of Fontinalis, 967, 971 ; compensation points of algae, 983, 984. and coworkers : Interpretation of midday depression, 874. , Doring, B., and Simonis, W. : Photosynthesis of Elodea most efficient in light of the color in which it grew, 1168. , Filzer, P., and Lorenz, A.: Photosynthesis of desert plants, 997, 1001. Hartel, O.: CO2 supply through roots and midday depression, 910. Harris, D. G., and Zscheile, F. P.: Absorption spectrum of chlorophyll in different solvents, 637-639, 643-645. See also Zschiele, F. P. Haxo, F. T., and Blinks, L. R.: Action spectra of red algae, 1180-1182; of Ulva (green alga), 1152, 1181, 1182; oi Coilodes me (hrown alga), 1172, 1181, 1182; of blue- green algae, 1183; drop of quantum yield of Ulva in far red, 1153, 1156. Heise, G. W. See Brown, W. H. Hendricks, R. H. See Thomas, M. D. Hendricks, S. B. See Warburg, O. Henrici, M. : Photosynthesis of mountain plants, 997. Herzfeld, K. F. See Franck, J. Hill, G. R. See Thomas, M. D. Holt, A. S. See Rabideau, G. S. Honert, T. H. van der: CO2 curves of Hormidium, 893, 895, 904; assimilation numbers of Hormidium, 991. Hoover, W. H.: Photosynthesis in near ultraviolet, 1153; in far red, 1158; action spectrum of photosynthesis in average light, 1162-1163. , Johnston, E. S., and Brackett, F. S. : Constant photosynthesis of land plants under constant conditions, 879; CO2 curves of wheat, 892, 896, 904; light curves of wheat, 966, 970; linear range, 980. Hubert, B. : Absorption spectrum of different chlorophyll colloids, 649. Jaag, O. See Jaccard, P. Jaccard, P., and Jaag, O. : Variations of photosynthesis under constant conditions, 877. James, W. O., Enhancement of photosynthesis by bicarbonate a buffering effect?, 887; CO2 curves of aquatic plants, 892, 903. Johnson, H. G., and Levring, T.: Photosynthesis of algae in near ultraviolet, 1153. Johnston, E. S.: Photosynthesis in polarized light, 1147. See also Hoover, W. H. Juday, C. See Manning, W. M. Kaempfert, W. See Schanderl, H. Kanitkar, U. K. See Dastur, R. H. Karrer, P., and Solmssen, V.: Absorption spectra of carotenoids from purple bacteria, 656, 658. and Wi'irgler, E.: Absorption spectra of fucoxanthol and other carotenoids, 656, 657, 661. Kasha, M. See Lewis, G. N. Katunsky, V. M.: CO2 fertilization, 902. Katsurai, T. See Svedberg, T. Katz, E.: Interpretation of light curves of fluorescence, 1077. See also Wassink, E. C. and Wassink, E. C. : Absorption spectra of chlorophyll in different solvents, 637- 642; of colloidal chlorophyll-protein extracts, 653-656; of colloidal bacterio- AUTHOR INDEX 1199 chlorophj^ll-protein extracts, 654-656; of blue-green algae, 692; of green sulfur bacteria, 694, 704; of Chlorella, 699, 700. , Wassink, E. C, and Dorrestein, R.: Inhomogeneity of light absorption in cell suspensions, 864-866; light curves of purple bacteria, 969, 972; of suspensions of different density, 1009-1011; fluorescence-light curves of C/irowa/nm(, 1049-50. Kautsky, H., and coworkers: Quenching of chlorophyll fluorescence by oxygen, 778; nonquenching by allylthiourea and isoamylamine, 788-789; phosphorescence quenching by oxygen, 793-794; time course of fluorescence in leaves and algae, 806; its relation to photosynthesis, 819, 820, 822; effect of temperature on fluorescence of leaves, 1055; of cyanide, 1057; of narcotics, 1063: of O2, 1063. Kersten, J. A. H. See Wassink, E. C. Kniep, H.: Maximum rates of Ulva and Porphyra, 991, 992. and Minder, I.: Bubble counting study, 847; action spectrum of photosynthesis has two peaks, 1144. Knorr, H. V. See Albers, V. M. Kohl, F. G. : Double-peaked action spectrum of photosynthesis, 1 144. Kok, B. : Chlorella light curves have two linear ranges?, 981 ; quantum yield of Chlorella doubled below compensation?, 1113-1116, 1132; a photosynthesis mechanism involving high energy phosphates, 1115-1116. Komor, J. See Noddack, W. Kopp, C. See Noddack, W. Korzenovsky, M. See Warburg, O. Koski, V. M. See French, C. S. and Smith, J. H. C: Absorption spectrum of protochlorophyll, 618-619. Kostychev, S. P.: "Physiological concept" of photosynthesis, 872-873; variations caused by internal factors make short-time rate measurements meaningless?, 876. and coworkers: Midday depression and accumulation of carbohydrates, 875; no effect of stomatal opening on photosynthesis?, 915; rate of photosynthesis under various climatic conditions, 874, 875, 996-1001. , Bazyrina, K., and Vasiliev, G.: CO2 exhaustion effects in land plants, 908. and Soldatenkov, S. V. : Midday depression in algae, 876. Kramer, P. J., and Decker, J. P.: Deciduous plants umbrophilic, conifers heliophilic, 989. Krasnoselskaja-Maximova, T. A. See Maximov, N. A. Krasnovsky, A. A. See Evstigneev, V. B. Kundt, A. : Effect of solvents on chlorophyll spectrum, 635-637. Kuilman, L. W.: Inactivity of red leaf pigments in photosynthesis, 1165. Kursanov, A. L.: Carbohydrate accumulation and midday depression, 875; midday depression in algae, 876; diurnal course of photosynthesis, 874; rate of photo- synthesis of subtropical plants, 999. Lai, R. N. See Singh, B. N. Levi, H. See Franck, J. Levring, T.: Underwater light fields, 735; action spectra of algae, 1188. See also Johnson, H. G. Lewis, C. M. See Emerson, R. Lewie, G. N. (with Kasha, M., McClure, D. S., and Calvin, M.): Long-lived triplet states, 790-795. Lemberg, R.: Absorption spectra of phycobilins and their proteids, 664-667. 1200 AUTHOR INDEX Liebig, J. : "Law of the minimum," 858-859. Lineweaver, H. See Bm-k, D. Lipmann, F., and Tuttle, L. C: Reductive carboxylation, 937. Livingston, R.: Absorption spectrum of allomerized chlorophyll, 613-614; search for infrared chlorophyll fluorescence, 748, 795-796; drop of fluorescence yield of chlorophyll in far red, 1153. See also Franck, J. and Ke, C.-L.: Concentration quenching of chlorophyll fluorescence, 774-775; quenching of chlorophyll fluorescence in solution by admixtures, 781-788. , Watson, W. F., and McArdle, G.: Effect of alcohols and amines on absorption spectrum of chlorophyll, 646-649; solvent effect on chlorophyll fluorescence, 747; yield of chlorophyll fluorescence in solution as function of wave length of exciting light, 752-753; fluorescence jdeld of allomerized chlorophyll, 754; nonfluores- cence of chlorophyll in nonpolar solvents and its activation by water, alcohols and amines, 766-771, 777. Loewy, A. See Scarth, J. W. Lloyd, F. E.: Chloroplast fluf)rescence under the microscope, 806. Loomis, W., Carr, and Randall, H. M.: Transmission, scattering and reflection of light by leaves, 676, 677, 683; leaf spectra, 688, 691, 699; absorption and reflection of infrared by leaves, 692, 694, 695; spectra of autumnal leaves, 696, 697; relative intensities of absorption bands in leaves, 708. See also Verduin, J. Lorenz, A. See Harder, R. Lubimenko, V. N.: Chlorophyll bands in leaves of different species, 699, 700-702; no carotenoid bands in leaves?, 705; light threshold of photosynthesis?, 964; light curves of umhrophilic and heliophilic plants, 987, 989; shade plants more effective than sun plants in blue light, 1167-1168. Lundegardh, H. : Modification of the law of limiting factors, 862; COo exhaustion effects in land plants, 908; light curves of Oxalis and Stellar ia, 966; of umbrophilic and heliophilic plants, 987; rate of photosynthesis under natural conditions, 999. M Macdowall, F. D. See French, C. S. McArdle, G. See Livingston, R. McAlister, E. D., and Myers, J.: Time course of fluorescence in plants, 806; its relation to photosynthesis, 819, 824; infrared photometry as method for measuring photosynthesis, 852-853; effect of CO2 concentration on fluorescence in leaves, 940; fluorescence-light curves of wheat, 1048; effect of CO2, 1051; effect of oxygen, 1066-1067. McClure, D. S. See Lewis, G. N. McGee, J. M. See Van Rysselberghe, P. McLean, F. T. : Oxygen release during midday depression, 874; photosynthesis of tropical plants, 1001. Mackinney, G.: Spectroscopically pure chlorophyll, 603-604. Magee, J. L., De Witt, T. W., Smith, E. C, and Daniels, F.: Calorimetric measurement of photosynthesis, 854; calorimetric quantum yield determinations, 1123. Manning, W. M., Juday, C, and Wolf, M.: Chemical quantum yield measurements with Chlorella at different depths, 1120. See also Button, H. J.; Strain, H. H. , Stauffer, J. E., Duggar, B. M., and Daniels, F.: Quantum yield measurements with Chlorella by gas analytical methods, 1119-1120. and Strain, H. H.: Absorption spectrum of chlorophyll d and derivatives, 616; light absorption by chlorophyll d in red algae, 720-721; fluorescence spectrum of chlorophyll d, 748. AUTHOR INDEX 1201 Man ten, A. : Action spectrum of phototaxis of purple bacteria shows activity of bacterio- chlorophyll and of carotene ids except spirilloxanthol, 1188. Maskell, E. J.: "Mutual interaction of factors" and the law of limiting factors, 863; periodicity of photosj-n thesis in detached leaves, 874-875; role of stomata in midday depression, 875; stomata as a limiting factor, 915. Matthaei, G. L. G. See Blackman, F. F. Maximov, N. A., and Krasnoselskaja-Maximova, T. A.: Fluctuations of photosynthesis in leaves, 876. Mecke, R., and Baldwin, W. C. G.: Reflection of infrared light by leaves, 094, 696. xMehta, R. Y. See Dastur, R. H. Meier, F. E.: lulling of Chlorella by ultraviolet light, 1153. Menke, W.: Carotenoid bands in chloroplasts, 706; fucoxanthol band in brown algae, 706, 707. Metzner, P. : Fluorescence burst in heated chloroplasts, 817. Meyer, K. P.: Absorption spectra of different chlorophyll colloids, 649-650; leaf spec- tra, 687, 691, 699, 705; carotenoid bands in leaves, 706; relative intensities of bands in leaves, 708; scattering and absorption in leaves, 710; "light sieve" effect in seedlings, 715. Miller, E. S.: Absorption spectra of carotenoids, 656, 660. and Burr, G. O. : CO2 compen.sation point, 898-899; compensation at low CO2, 985. Minder, I. See Kniep, H. Mitchell, J. W.: Constant photosynthesis under constant field conditions, 876, 877. Molvig, H. See Stern, A. Monch, I.: Carbohydrate accumulation and midday depression, 875; rate of photo- sjTithesis of alpine plants, 997, 1001. Montfort, C: Distribution of light energy among pigments in Ulva (green), 719-720, 725; in Fucus and Laminaria (brown), 724-725; rigid and adaptable umbrophiles and heliophiles, 989; light inhibition of deep sea algae, 995-996; photosynthesis of Ulva in colored light indicates efficiency of carotenoids, 1164; that of brown algae shows activity of fucoxanthol, 1168-1169. and Neydel, K.: Midday depression in stomata-free plants, 876. and Schmidt, G.: Efficiency of brown algae in light of different color, 1176-1177; of blue-green and red algae, probable activity of phycobilins, 1185. Moore, W. E., and Duggar, B. M.: Quantum yield of Chlorella in blue, red and mixed light measured by polarographic method, 1121-1122, 1132. Mothes, K., Baatz, I., and Sagromsky, H.: Photosynthesis of green and brown algae in red and blue light, 1178. Miiller, D.: Light curves of arctic plants, 966; rate of photosynthesis of normal and N- deficient leaves, 998; yield of photosynthesis of arctic plants under natural condi- tions, 998. See also Boy sen- Jensen, P. Mulliken, R. S.: Theory of polyene spectra, 663-664. Myers, J. : Photosynthetic characteristics of Chlorella cultures, 834. See also McAlister, E. D. N Nathansohn, A. : CO2 molecules — the only substrate of photosynthesis?, 887. Negelein, E. See Warburg, O. Neuwohner, W.: Interpretation of midday depression, 874, 875; midday depression in algae, 876. Neydel, K. See Montfort, C. Nishimura, M. S. See Emerson, R.; 'VMiittingham, C. P. 1202 AUTHOR INDEX Noack, K.; Fluorescence of colloidal leaf extracts, 776-777; participation of flavones in photosynthesis?, 1165. See also Sierp, H. Noddack, W., and Eichhoff, H. J.: Light scattering by Chlorella suspension, 676-677; absorption spectrum of Chlorella, 690, 692, 699, 700; absorption ratio in red and green in Chlorella, 708; enhanced absorption in far red in vivo, 715-716; light scattering measured by "ellipsoid photometer," 843; constant photosyn- thesis in Chlorella, 877; light curves of Chlorella in red, yellow and white light, 969,1160,1162; linear range, 980; compensation point, 983; quantum yield of Chlorella in monochromatic light, 1155-1156. and Komor, J.: Yields of organic matter in grass plots, 1004-1006. and Kopp, C: Light curves of Chlorella at different temperatures, 969, 973; linear range, 980; assimilation numbers of Chlorella cultures, 991, 993. Ochoa, S.: Reductive carboxylation, 937, 939. Oltmanns, F.: No chromatic adaptation?, 995. Oneto, J. F. See Pratt, R. Oppenheimer, R. J. See Arnold, W. Osterhout, W. J. V., and Dorcas, M. J.: Penetration of carbonic acid into Valonia, 887. Osterlind, S. : Can algae use bicarbonate ions more effectively than CO2 molecules?, 890- 891; CO2 compensation point lowered by bicarbonate, 895; inhibition by excess CO2 in algae, 903. Overkott, O.: CO2 supply through roots, 910. Paauw, P. van der: CO2 curves of Hormidium, 893, 904; adaptation of Hormidium to weak light, 989, 994. Paetzold, I. See Stocker, O. Pantanelli, E.: Photosynthesis as function of light intensity, 964. Pauling, L. : Theory of polyene spectra, 663. Pekerman, F. M. See Vavilov, S. L Petering, H. G., Duggar, B. M., and Daniels, F.: Polarography applied to measurement of photosynthesis, 850-851; quantum yield of Chlorella determined by polaro- graphic method, 1120-1121. Plaetzer, H. : Compensation points of aquatic plants, 982; of algae, 983. Pokrovski, G. L : Reflection, absorption and transmission of infrared light by leaves, 696, Pollack, M. See Franck, J. Pratt, J. See Pratt, R. Pratt, R.: Inhibition of photosynthesis by a metabolite of Chlorella, 833; effect of K and Na bicarbonate on time course of photosynthesis in Chlorella, 835-836, 877. (with Fong, J., Oneto, J. F., and Pratt, J.): A growth-nihibiting substance ("chlo- rellin") formed in Chlorella suspensions, 880; its extraction and inhibiting effect on photosynthesis, 880-881; decline of photosynthesis with age of Chlorella suspen- sions, 881-882. Pringsheim, P. See Franck, J. Prins, J. A.: Life-time of excited chlorophyll, 633-634; yield of chlorophyll fluorescence in solution, 752, 754. Pruckner, F.: Solvent effect on spectra of porphin, chlorophyll and bacteriochlorophyll, 642. See also Stern, A. Puck, T. T. See Franck, J. AUTHOR INDEX 1203 Purevich, K.: Energy conversion by plants under natural conditions, 1003-1004. Putter, A. : Energy conversion by plants over a whole season, 1005-1006. Rabideau, G. S., French, C. S., and Holt, A. S.: Absorption spectra of chloroplast dis- persions, 654-655; of leaves and chloroplast suspension, 688-689, 691. See also French, C. S. Rabinowitch, E.: Interpretation of spectra of bacteriochlorophyll, 617-618; of porphin, 621; of hexaphenyl porphin, 623; of chlorophyll and bacteriochlorophyll, 630-633; absorption and reflection in plane-parallel vessels, 672-674; interaction of scatter- ing and absorption, 711-712; interpretation of fluorescence spectra of chlorophyll and bacteriochlorophyll, 750-752, 753-754; mechanisms of quenching, 755-758; yields of photosynthesis and fluorescence, 820, 821-822, 823-824; three types of saturation curves of photosynthesis, 866-872, 1012-1017; fluorescence-light intensity curves of plants in strong light, 941 ; effect of reductants on fluorescence in bacteria, 950-951; light curves of thin and dense suspension, 1007-1008. See also Ehrmantraut, H. C. Rabinowitch, E., and Epstein, L. F.: Fluorescence quenching by dimer formation, 761- 762, 765. (with Jacobs, E. E.): Kinetic analysis of the CO2 factor in photosjTithesis, 916-939; calculation of carboxylation equilibrium constant, 935-936; analysis of light curves for different kinetic models, 1017-1047; interpretation of light curves of fluorescence, 1067-1078; extrapolation of maximum quantum yield, 1133-1135; theoretical and experimental maximum quantum yield, 1137-1139. Raffy, A. See Dh6r6, C. Randall, H. M. See Loomis, W. Rao, M. S. See Dastur, R. H. Rehm, S. See Stacker, 0. Reinau, E.: CO2 fertilization, 902. Reinke, J.: Discovery of light saturation, 964. Reman, G. H. See Wassink, E. C. Richter, A. von: Action spectrum of photosynthesis has two peaks, 1144; phycobilins inactive in photosynthesis?, 1184-1185. Riecke, F. F.: Quantum yield of photosynthesis — repetition of Warburg and Negelein's experiments, 1090-1091; new measurements, on Scenedesmus and Chlorella, 1095; quantum yield of CO2 reduction by anaerobically adapted Scenedesmus, 1 128. Riley, G. A.: Yield of photosjTithesis in sea, 1006. Roach, J. R. See Zscheile, F. P. Rommel, L. G. : Reaction kinetics and law of limiting factors, 863-864; phase boundary air-cell wall as diffusion barrier, 916. Rudolph, H. : Absorption spectrum of protochlorophyll, 618-619. Ruttner, F.: Photosynthesis of aquatic plants at high pH, 890, 906; CO2 compensation point lowered by bicarbonate, 899. Sachs, J.: Bubble counting as method of measuring photosynthesis, 846; cardinal points, 858; maximum photosynthesis in yellow-green light?, 1142. Samant, K. M. See Dastur, R. J. Sandoval, A. See Zechmeister, L. 1204 AUTHOR INDEX Schanderl, H., and Kaempfert, W.: Changes in light transmission caused by chloroplast movements, 680-681; "light sieve effect" in leaves, 715; inhomogeneity of light absorption in a leaf, 864-865. Scarth, J. W., Loewy, A., and Shaw, H. H.: Fluctations of photosynthesis in detached leaves, 877. Schmidt, G. See Montfort, C. Schocken, V. See Warburg, O. Schoder, A.: Interpretation of midday depression, 874; stomata do not limit photo- synthesis, 915. Schroeder, H. : Diffusion resistance of air channels as limiting factor, 916. Senn, G.: Chloroplast movements in algae, 679. Seybold, A.: Absorption bands of protochlorophylls a and b, 619; absorption of light by leaves, 677, 678, 683; by algae, 683; by single chloroplasts, 683; by nonplastid pigments, 684; leaf spectra, 687, 691, 699; spectra of aquatic plants and algae, 689, 690, 691, 699, 706; relative intensities of absorption bands in leaves, 708; light absorption by carotenoids in Phaseolus leaves, 721-722; "natural light fields," 731-735. See also Sierp, H. and Egle, K.: Nonfluorescence of adsorbed chlorophyll, 775, 776, 777; fluorescence changes in leaves caused by boiling, drying and changes in humidity, 817-818; chlorophyll present in leaves in two forms?, 818-819. and Weissweiler, A.: Absorption spectrum of adsorbed chlorophylls a and b, 651- 652; absorption by leaves with different chlorophyll content, 678; by single chloroplast layers, 683; by aurea and purpurea leaves, 685; leaf spectra, 686, 691 ; spectra of algae, 690, 691; spectra of boiled and ether-filled leaves, 698; of Chlorella, 700; ratios of absorptions in red, green and blue in various leaves, 710; assimilation numbers of aurea leaves, 993. Shaw, H. H. See Scarth, J. W. Shiau, Y. G., and Franck, J.: Fluorescence changes in algae, 806; their relation to photosynthesis, 819-821; effect of quinone on fluorescence of Chlorella, 1067; fluorescence-light curves of Chlorella and Scenedesmus, 1049; effect of oxygen, 1067. Shull, C. A. : Reflection spectra of leaves of different age, 696, 697, 699. Sierp, H., Noack, K., and Seybold, A.: Diffusion through septa, 913. Simonis, W. See Harder, R. Singh, B. N., and Kumar, K.: Sigmoid light curves of leaves?, 964; light curves of Raphanus sativum, 966, 987. , Lai, R. N., and Kumar, K.: CO^ curves of higher plants, 892, 903, 904, 908. Smith, A. M. See Blackman, F. F. Smith, E. C. See Magee, J. L. Smith, E. L.: Absorption spectrum of colloidal leaf extracts, 652-653; CO2 curves of Cabomba, 892, 897, 903, 904; are CO2 curves nonhyperbolic?, 937-938; light curves of Cabomba, 967, 968, 971, 987; higher order equations for light curves, 1044, 1046-1048. Smith, J. H. C. See Koski, V. M. Soldatenkov, S. V. See Kostychev, S. Solmssen, V. See Karrer, P. Solomon, S. See Dastur, R. H. Spoehr, H. A. (and coworkers): Separation of an antibiotic from Chlorella, 882-883; organic matter production by field crops and trees, 1006. AUTHOR INDEX 1205 Stair, R., and Coblentz, W. W. : Infrared absorption spectrum of chlorophyll and phytol, 610-612; of carotenoids, 656. Stalfelt, M. G.: Closure of stomata as cause of midday depression, 875; CO2 diffusion through cuticle, 911; effect of partial closure of stomata on photosynthesis, 915- 916; light curves of mosses, 966; of lichens, 967; compensation points, 982. Stauffer, J. E. See Manning, W. M. Steemann-Nielsen, E.: Utilization of bicarbonate by Myriophyllum, 888-890; CO2 curves of Myriophyllum and Fontinalis, 893, 903; light curves of green and brown algae, 969; induction losses following reduction of light intensity, 1040. See also Gabrielsen, E. K. Stefan, M. J.: Theory of diffusion from point source, 912-913. Stern, A.: Typical porphin and chlorin spectra, 620. and Molvig, H.: Absorption spectrum of porphin, 620-622. , Molvig, H., and Dezelic, M.: Fluorescence of porphyrins and chlorins, 749-750. , Wenderlein, A., Molvig, H., and Pruckner, F.: Effect of substitutions on porphin spectrum, 621-629. Stern, K.: Fluorescence of chlorophyll colloids, 775-776. Stocker, O.: Rate of photosynthesis of tropical plants, 1001. , Rehm, S., and Paetzold, I.: Fluctuation m photosynthesis and in CO2 content of air, 876-877; maximum rates of photosynthesis of leaves, 991. Stokes, G. G.: Composite nature of chlorophyll deduced from fluorescence spectrum, 740; discovery of leaf fluorescence, 805; fluorescence of pigments from red algae, 759. Stoll, A., and Wiedemann, E.: Fluorescence of colloidal leaf extracts, 777. See also Willstatter, R. Strain, H. H.: Absorption spectra of leaf carotenoids, 656, 658, 660. See also Manning, W. M. and Manning, W. M. : Absorption spectrum of chlorophyll c (chlorofucin), 614-615; light absorption by chlorophyll c in diatoms, 720-721. , Manning, W. M., and Hardin, G.: Absorption spectra of carotenoids of brown algae, diatoms and dinofiagellates, 656-657, 659, 661. Svedberg, T., and Katsurai, T.: Absorption spectra of phycobilins, 664-665. Tanada, T.: Absorption spectrum of diatoms, 699; relative band intensities in diatoms, 708, 710; distribution of light energy between pigments in diatoms, 726; quan- tum yield in Navicula minima, 1097; action spectrum of N. minima, full activity of fucoxanthol and chlorophyll c, 1173-1175. Terwood-Lad, D. See Franck, J. Thoday, D.: Midday depression of photosynthesis, 874. Thomas, J. B.: Action spectrum of photosynthesis of purple bacteria shows peaks be- longing to bacteriochlorophyll and carotenoids, other than spirilloxanthol. 1188. Thomas, M. D., Hendricks, R. H., and Hill, G. R.: CO2 compensation point, 899. and Hill, G. R.: CO2 fertilization, 902. Timiriazev, K.: Maximum of photosynthesis coincides with maximum of chlorophyll absorption, 1142-1143; coincides with maximum of solar spectrum?, 1143. Tonnelat, J.: Calorimetric measurement of photosynthesis, 854; calorimetric quantum yield determinations with Chlorella, 1123-1124. 1206 AUTHOR INDEX Tswett, M. : Fluorescence of chloroplasts under the microscope, 806. U Ursprung, A.: Photosynthesis in near ultraviolet, 1153. Van der Honert, T. H. See Honert, T. H. van der. Van der Paauw, P. See Paauw, P. van der. Van der Veen, R. See Veen, R. van der. Van Norman, R. W., French, C. S., and Macdowall, F. D. H.: Absorption spectra of red algae, 692, 699; fluorescence spectra of extracts from red algae, 799. See alto French, C. S. Van Rysselberghe, P., Alkire, G. I., and McGee, J. M.: Electrolytic reduction of COj, 939. Vasiliev, G. See Kostychev, S. P. Vavilov, S. I., Galanin, M. D., and Pekerman, F. M.: Transfer of electronic energy be- tween molecules, 758-759; fluorescence quenching through energy transfer to nonfluorescent pigments, 778. Veen, R. van der: Thermal conductivity as method of measuring photosynthesis, 853; kink in light curve near compensation point, 1116; a mechanism of photosynthesis involving formation of high energy phosphates by back reactions, 1116-1117. Verduin, J.: Mutual interference of pores in diffusion through septum, 913-914. and Loomis, W. E.: CO2 concentration near ground, 902. Vermeulen, D., Wassink, E. C., and Reman, G. H.: Absorption spectra of purple bac- teria, 692, 693; spectrophotometric study of bacteriochlorophyll fluorescence, 747, 748; fluorescence spectrum of Chlorella and Chromatium, 806, 807, 809, 810; quantum yield of fluorescence in Chlorella and Chromatium, 813; its dependence on wave length, 813-814; action spectrum of growth of purple bacteria, inactivity of carotenoids, 1188. See also Wassink, E. C. Voerkel, S. H.: Chloroplast movements in Funaria, 679-682. Volkov, A.: Discovery of proportionality between photosynthesis and light intensity, 964. von Guttenberg, M. See Guttenberg, M. von. von Richter, A. See Richter, A. von. W Wakkie, J. G.: Absorption spectrum of chlorophyll in different solvents, 637, 641; at high concentration, 651; fluorescence of chlorophyll emulsions, 776. Warburg, O.: Use of unicellular algae for the study of photosynthesis, 833-834; mano- metric methods of measuring photosynthesis, 847-850; CO2 curve of Chlorella, 893, 906, 908; effect of O2 pressure on photosynthesis, 960- light curves of Chlorella, 967; second series of quantum yield measurements, 1098-1107. and Burk, D. (also Schocken, V., and Hendricks, S.) : Third series of quantum yield determinations, by the two-vessel method, 1104-1109, 1110, 1132. , Burk, D., Schocken, V., Korzenovsky, M., and Hendricks, S. B.: No light effect on respiration in Chlorella when CO2 is effectively removed, 901. and Negelein, E.: Early curvature of light curves in dense suspension of Chlorella, 980; first determination of quantum yield of photosynthesis, 1085-1086, 1088; quantum yields in colored light, 1 147. AUTHOR INDEX 1207 and Schocken, V.: Chlorophyllide-sensitized autoxidation of thiourea in pyridine as an actinometric reaction, 839-841. Wassink, E. C: CO2 exhaustion effects in leaf discs in buffer, 905; light curves of horticultural plants, 967, 970; linear range, 980, 981; maximum photosynthesis of horticultural plants, 991 ; quantum yields of leaf discs, 1096. (with Katz, E., Dorrestein, R., and Kersten, J. A. H.): Fluorescence-time course in algae, 806; quantum yield of fluorescence in diatoms and purple bacteria, 812; relation of fluorescence to photosynthesis, 819, 820, 824, 825. See also Eymers, J. G.; Katz, E.; and Vermeulen, D. and Katz, E.: Light curves of Chlorella cultures of different age, 978 ■ cyanide stimulation of fluorescence in Chlorella, 1057, 1060, 1061. , Katz, E., and Dorrestein, R.: Absorption spectra of purple bacteria, 692, 702-704; fluorescence of colloidal extracts from purple bacteria, 777; CO2 curves of purple bacteria, 893, 903-904; effect of concentration of CO-, and of reductants on fluo- rescence, 941, 942, 949-950; effect of concentration of thiosulfate, H2S and H2 on rate of C0> reduction, 945-947; CO2 reduction with mixed reductants, 947-948; ^gmoid light curves in bacteria, 948, 964; effects on photoreduction and fluores- cence of bacteria, of pH, 952-954; of cyanide, 954-955; of hydroxylamine, 955- 957; of azide, 958-959; of urethans, 959-960; light curves of purple bacteria at different temperatures, 969, 974; in presence of KCN, NH2OH, NaN, and ure- than, 977; at different pH's, 978; linear range, 980; fluorescence-light curves of Chromatium, 1049-1050; effect of CO2 on them, 1052; of reductants, 1052- 1055; of temperature, 1057, 1059; of cyanide, 1061-1062; of hydroxylamine and azide, 1062-1064; of pH, 1063, 1065; of narcotics, 1063, 1066; interpretation of light curves of fluorescence, 1077; quantum yield determinations with purple bacteria, 1125-1126, 1127-1128, 1132. and Kersten, J. A. H.: Absorption spectra of carotenoids from diatoms, 657, 659; of diatoms, 692, 699; fucoxanthol bands in diatoms, 706; relative intensities of absorption bands in diatoms, 708; effect of CO2 concentration on fluorescence of diatoms, 940-941 ; effect on photosynthesis and fluorescence of diatoms of cya- nide, 945-955; of urethan, 959; light curves of diatoms, 969, 972; in presence of cyanide and urethan, 976; fluorescence-light curves of diatoms, 1050-1051; effect of CO2, 1051; of temperature, 1056-1057, 1059; of cyanide, 1062; quantum yield in Nitzschia disupata, 1097; action spectrum of N. dissipata; efficiency of fucoxanthol, 1172. , Vermeulen, D., Reman, G. H., and Katz, E.: Effect of urethan concentration on photosynthesis in CMoreifc, 959; of O2 pressure, 960 ; photosynthesis-light curves of Chlorella at different temperature, 967, 973; in presence of KCN, 974; in presence of urethan, 975; in O, and N2, 976; fluorescence-light curves of Chlorella, 1048; at different temperatures, 1055; in presence of cyanide, 1057, 1060; of narcotics, 1063, 1065; of O2, 1063. Watson, W. F. See Livingston, R. Waugh, V. G.: Fluctuations of photosynthesis in leaves, 876. Weil-Malherbe, M. -See Weiss, J. Weis, F.: Sun and shade plants, 987. Weiss, J. : Mechanism of fluorescence quenching, 780. and Weil-Malherbe, H.: Concentration quenching of chlorophyll fluorescence, 772- 774, 789-790; quenching by oxygen, 779. Weller, S., and Franck, J.: Light curves of hydroxylamine-inhibited Chlorella, 975. Wenderlein, A. See Stern, A. White. J W. See Zscheile, F. P. 1208 AUTHOR INDEX Whittingham, C. P.: Time course of photosynthesis of Chlorella in carbonate buffers, 908; calculation of carboxylation equilibrium, 935. , Nishimura, M. S., and Emerson, R.: Criticism of Warburg and Burk quantum yield determinations, 1109-1113. Wiedemann, E. See Hagenbach, A.; Stoll, A. Willstatter, R., and Stoll, A.: Constant photosynthesis of detached leaves, 877; light curves of excised leaves, 966; maximum rates of photosynthesis in strong light, assimilation numbers, 990, 992; yield of photosynthesis in strong light in green and yellow leaves, 1136-1137; inactivity of earotenoids?, 1163; of red leaf pig- ments, 1165. WUschke, A.: Fluorescence bands of brown and green algae, 806, 807. Wolf, M. See Manning, W. M. Wiirgler, E. See Karrer, P. Wurmser, R.: Absorption and scattering in inhomogeneous systems, 713-714; absorp- tion ratio red : blue enhanced in vivo, 716; quantum yield measurements with Ulva by Winkler's method, 1118, 1148. and Ducleaux, J.: Relative efficiency of red and green algae in colored light, 1185. Z Zechmeister, L., and Sandoval, A.: Fluorescent carotenoid derivatives, 798-799. Zscheile, F. P. See Harris, D. G. and Comar, C. L.: Absorption spectrum of chlorophylls a and b, 605. , Comar, C. L., and Harris, D. G. : Preparation of spectroscopically pure chlorophj-ll, 603-604. and Harris, D. G.: Absorption spectra of chlorophylls a and b in ultraviolet, 609- 610; spectrophotometric study of the fluorescence of chlorophyll, 741-746, 748; effect of solvent on fluorescence yield of chlorophyll, 763-764, 769. , White, J. W., Beadle, B. W., and Roach, J. R.: Absorption spectra of earotenoids, 656, 659, 660.

C, i. e., the quantum yield of fluorescence may have been correspondingly lower (not reaching 100% at infinite dilu- tion). Comparison of (23.10) with (23.6) shows that by correcting the deriva- tions of Weiss and Weil-Malherbe for the possibility of internal conversion or other mononiolecular deactivation processes, one would obtain, from their value of A", a value of k^ even larger than the one given in (23.9) : (23.12) Ac > 2 X 10'2 The extraordinarily large value of this bimolecular rate constant makes one skeptical about the correctness of Weiss and Weil-Malherbe's experiments, or, at least, of their interpretation. Assuming that Chi* is deactivated by the first encounter with Chi — which is not implausible — the above value of kc gives for the average interval between two encounters (at [Chi] = 1 niole/1.) : (23.13) i = 5 X 10-3 sec. This is at least one whole order of magnitude less than can be estimated from the formula for the frequency of collisions in gases, and two or three orders of magnitude less than the estimate based on the rate of diffusion of dyestuff molecules in solution. The rates of those reactions in solution that occur by the first (or one of the first) molecular collision are determined by the rate of diffusion (which brings the molecules 774 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 together in an "encounter") rather than by the frequency of collisions. The distinction between eiicounters (or "co-ordinations") and collisions, and its importance for reaction kinetics of condensed systems was discussed by Rabinowitch (1937). The unusual magnitude of the constant (23.12) is illustrated by the fact that for nine other dyes, investigated by the same authors, the self- quenching constants were at least 2000 times smaller (and thus of the order of magnitude to be expected from diffusion constants) . It seemed thus that the conclusions of Weiss and Weil-Malherbe were in need of confirmation. Their results might have been vitiated by self- absorption (which the authors dismissed as unimportant "because there was no evidence of surface fluorescence"). In their set-up, the fluores- cent beam was in line with the (ultraviolet) excitmg beam — an arrangement that favors self-absorption. Less mutual overlapping of the fluorescence band and absorption band may explain why self-absorption did not affect equally strongly the results obtained with the other nine dyes. That the observation of Weiss and Weil-Malherbe was due to self- absorption was later confirmed by Livingston and co-workers (1948), who showed experimentally that, if care is taken to avoid self-absorption, no concentration quenching of chlorophyll fluorescence can be noted up to 1.5 X 10~^ mole/1. Weiss (1948) agreed with this, but stated that con- centration quenching does occur at still higher concentrations. Livingston and co-workers (1948), using a cell 1 mm. thick, extended their measure- ments up to 0.10 mole/1, of chlorophyll in butyl ether and found in fact a strong concentration quenching above 2 X 10~^ mole/1, (fig. 23.7); at 7 X 10 ~^ mole/1. , the fluorescence yield was only 7% of that in dilute solu- tion. The insert in fig. 23.7 shows the sigmoid shape of the curve. The absorption spectrum of chlorophyll a in butyl ether, measured at 2 X 10 ~^ mole/1, (a concentration at which self-quenching reduces the fluorescence to about 40% of the maximum), appeared not to differ sig- nificantly from the spectrum of dilute solution. A further pertinent observation was that the partly quenched fluores- cence of a concentrated chlorophyll solution was more strongly depressed by an increase in temperature than the fluorescence of a dilute solution. These observations — the unchanged absorption spectrum, the sigmoid quenching curve and the enhancing effect of temperature on concentration quenching — can all be fitted into the picture of quenching as the result of dissipation of excitation energy in a "weak link" in the energy exchange chain, such as a dimeric molecule (Forster) or a "hot" molecule of the monomer (Franck and Livingston). The concentration at which the quenching becomes noticeable (2 X 10"^ mole/1.) is of the same order of magnitude as that calculated by FSrster for the onset of the energy ex- change (of. chapter 32). CONCENTRATION QUENCHING 775 Numerous observations are known showing that the close packing of chlorophyll molecules in solid chlorophyll, or in chlorophjdl monolayers on water, causes a complete disappearance of fluorescence. Chlorophjdl colloids too are, as a iiile, nonfiuorescent (Willstatter and St oil 1918; Stern 1920, 1921; Albers 1935; Meyer 1939). Meyer described certain fluorescent solutions of chlorophyll in ethanol as "colloidal," but no rea- sons for this description were given (cf. Smith 1941). Meyer's aqueous chlorophyll colloids, in which the density of the particles was similar to that in the chloroplast grana, were nonfiuorescent. Chlorophyll is non- 0.80 0.60 0.40 10.20 OOI 002 0.03 0.04 0.05 O06 0.07 Fig, 23.7. Concentration quenching of fluorescence of chlorophyll a in butyl ether (after Livingston and Ke, 1949). Ordinate, F/Fq. Abscissa, [Chi] in mole/1. fluorescent also in the adsorbed state, e. g., in starch columns used for chromatographic separation (Sej-^bold and Egle 1940). If we attribute the nonfluorescence of solid, colloidal and adsoibed chlorophyll to self-quenching, and consider it a consequence of close pack- ing, the restoration of fluorescence by certain "protective" substances can be attributed either to simple dilution of the pigment, or to effective inter- mption of the interaction between neighboring pigment molecules. Among the compounds reportedly capable of protecting the fluorescence of chloro- phyll, we find first of all lipides and lipophilic solvents. Stern (1920. 776 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 1921) found, for example, that chlorophyll-liiiide emulsions in water are fluorescent, while pure aqueous chlorophyll sols are not. The fluorescence of chlorophyll colloids in the presence of lecithin was confirmed by Bakker (1934). This may be either a true case of protection, or merely a dilution effect, since the concentration of chlorophyll molecules in the lipide drops may be lower than in the particles of the hydrosol. Seybold and E^lc (1940) found that chlorophyll on filter i)aper is fluorescent only if the solution from which it was adsorbed contained lipides, waxes or other lipophilic organic materials. (They even used the ab- sence of a fluorescent rim on a filter paper strip dipped into a chlorophyll solution as a control of the purity of both solution and paper.) Nonfluores- cent chlorophyll adsorbates on starch "light up" if they are wetted by or- ganic solvents, or if ether vapor is blown at them. Wakkie (1935) prepared a series of colloidal chlorophyll solutions (b}' diluting alcoholic solutions with increasing amounts of water), with and without a lipide (sodium oleate) . If no oleate was added, the shift of the absorption band from 660 to 672 m/x (completed within a narrow concen- tration range, and considered indicative of the transition from a monomole- cularly dispersed to an aggregated state) was accompanied by complete disappearance of fluorescence. In the presence of oleate, the absorption band began to shift at the same dilution, but the displacement ceased at 670 instead of 672 m/x; and, at the same time, fluorescence reappeared. By salting out, fluorescent, birefringent chlorophyll oleate "coacervates" could be precipitated from these fluorescent colloidal solutions. From these and similar observations, it appears that lipophilic mole- cules can protect the fluorescence of chlorophyll from self-quenching even without diluting the pigment, and without disrupting the chlorophjdl- protein or chlorophyll-cellulose bond. One can visualize the protecting molecules as enveloping the lipophilic parts of the adsorbed pigment mole- cules (e. g., in the case of chlorophyll, the phytol "tails"), and thus inter- rupting their mutual interaction. The "wrapping up" of flexible parts of the molecule may stiffen the latter and interfere with the internal conver- sion of electronic into vibrational energ3^ This stabilization effect may become manifest in a single pigment molecule, as well as within a complex of several such molecules. In the light of the above-described, more recent experiments by Livingston, one has also to consider the possibihty that tautomeric transformations from a nonfluorescent into a fluorescent form of chlorophyll could be responsible (or coresponsible) for effects of this type. It does not seem that the association of chlorophyll with proteins can in itself protect fluorescence. True, natural "chloroplastin" preparations apparently are fluorescent. Although Smith (1938) called aqueous chloro- phyll-protein extracts from spinach leaves "nonfluorescent," Noack QUENCHING BY ADMIXTURES 777 (1927), StoU and Wiedemann (1938) and Fishman and Moyer (1942) found that they fluoresce weakly, and that fluorescence is preserved also in precipitates prepared from such extracts by salting out. Wassink, Katz and Dorrestem (1942) observed that the yield of fluorescence was about the same (~0.1%) in live purple bacteria and in aqueous, colloidal bacteriochlorophyll-protein suspensions prepared from them. However, chloroplastin fluorescence probably must be attributed to the presence of as much as 30% hpophilic material. Artificial complexes containing only chlorophyll and proteins do not fluoresce. Accordmg to Noack, adsorb- ates of chlorophyll on globin sometimes fluoresce faintly; but Seybold and Egle (1940) suggested — probably with justification — that this must be as- cribed to the presence of impurities of a lipide nature. The bearing of these results on the problem of the state of chlorophyll in the IW'mg cell was discussed in chapter 14 (Vol. I). It was argued there that the nonfluorescence of pure chlorophyll adsorbates on proteins does not prove Seybold's hypothesis that the chlorophyll fluorescence in vivo is caused by a small fraction of chlorophyll dissolved in a lipide phase; more probably, all chlorophyll in the plants is weakly fluorescent {despite its high density and irrespective of its association with proteins) because of its simultaneous association with protective substances such as fats or phospholipides (Hubert, 1935). Livingston's experiments, (p. 766), in- dicate association with "activating" groups (such as OH, NH, or SH) as another possible explanation of fluorescence. 6. Quenching and Activation of Chlorophyll Fluorescence by Admixtures While the limitation of chlorophyll fluorescence in pure solutions may be caused equally well by physical dissipation or by photochemical reac- tions (tautomerization or reaction with the solvent), strong quenching by small amounts (<10~^ mole/1.) of foreign substances must be attributed to chemical interactions, since the rate of physical energy dissipation is not likely to be affected by the comparatively rare encounters of excited dyestuff molecules with the molecules of the "quencher," or by changes in the aver- age properties of the solvent caused by the presence of the latter. (An exception may be the case of resonance — to be discussed below.) As mentioned on page 757, the two most likely mechanisms of chemical quenching are oxidation-reduction (equations 23.1), and complex formation (equation 23.2). The second one is particularly probable when the ([uencher is another dyestuff with overlapjiing bands, so that self-quenching conditions are closely ai)proximated. In this case, permanent association of the ciucncher with the fluorescent molecule also becomes a likely possi- 778 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 bility (like the permanent dimerization discussed on page 761). Study of the effect of the quencher on the absorption spectmm of the fluorescent material, and of the dependence of quenching on the concentration of the quencher, may help to distinguish between "quenching by complex forma- tion" and "quenching by kinetic encounters"; but, at present, very few such data are available for chlorophyll solutions. Finally, in analogy to the above-discussed mechanisms of self-quenching (page 759), still another possibility of quenching should exist when the quencher is "in resonance" with the fluorescent molecule: quenching without complex formation and without kinetic encounters, through transfer of excitation energy over distances considerably larger than the collision diameters. If the molecules of the quencher are nonfiuorescent, they will serve as "traps" in the same way as was postulated by Forster for the dimers. Vavilov and co-workers (Pekerman 1947, Vavilov et at. 1949^'^) found convincing examples of this type in the quencing of dye fluorescence by resonating nonfluorescent dyes. However, strong quenchers in Table 23.IIIC (p. 782) certainly do not owe their effectiveness to a resonance transfer mechanism. They are all oxidants and this points to chemical interaction rather than physical energy transfer. In the second place, they have no absorption bands in the red, and are thus not in resonance with excited chlorophyll molecules. Their lowest excited states must be considerably higher than the fluorescent state A of chlorophyll. Among the substances whose quenching effect on the fluorescence of chlorophyll has been investigated in some detail are the reaction partners in chlorophyll-sensitized autoxidations — molecular oxygen and oxidation substrates such as benzidine, allyl thiourea etc. Because of the importance of these results for the analysis of the mechanism of sensitized autoxidation, they have already been anticipated in part in Volume I (c/. chapter 18, pages 483 and 518, and chapter 19, page 546). The quenching action of oxygen on the fluorescence of different dyestuffs (including chlorophyll) was first investigated by Kautsky and co-workers. Kautsky and Hirsch (1931) had discovered that the fluorescence of certain dyestuffs adsorbed on silica gel is considerably weakened by oxygen at pres- sures of several hundred millimeters, and that their afterglow (phosphores- cence) is completely destroyed even by very much lower pressures of this gas. Later, a similar ciuenching was observed in fluorescent dyestuff solutions, including chlorophyll solutions in acetone. According to Kaut- sky, Hirsch and Flesch (1935), the quenching of chlorophyll fluorescence is proportional to the partial pressure of oxygen, and shows no "saturation effect" even under a partial pressure of one atmosphere. Franck and Levi (1934) and Weil-Malherbe and Weiss (1942) found that the fluorescence of chlorophyll or ethyl chlorophillide solutions in QUENCHING BY OXYGEN 779 ethanol is reduced under one atmosphere oxygen by 30-35% (compared with its intensity in the absence of oxygen). This indicates that 50% quenching must require about two atmospheres oxygen, corresponding to a concentration of about IQ-^ mole/1, (c/. table 23.IIIC). Life-time of the fluorescent state of chlorophyll in ethanol solution was estimated on page 634 as 8 X 10-» sec. The average time that an excited chlorophyll mole- cule has available before its first encounter with a molecule of a solute whose concentration is of the order of 10 -^ mole/1, also is of the order of 10"^ sec. (No exact formulae for the calculation of encounter intervals in solutions are available, but it appears likely that these intervals are some- what— perhaps 10 or 100 times — longer than the collision intervals in gases of the same concentration.) It thus seems that excited chlorophyll mole- cules in the fluorescent state A undergo a quenching reaction by the very first, or one of the first, encounters with an oxygen molecule. The nature of this interaction is not known, but it is most likely to be the autoxidation of chlorophyll. A different hypothesis was suggested by Kautsky and maintained by him despite much criticism. According to this hypothesis, the interaction is a hulk transfer of electronic excitation energy from Chi* to O2. This concept originated in certain obsei-vations made by Kautsky and de Bruijn (1931) and by Kautsky, de Bmijn, Neuwirth and Baumeister (1933) in the study of the autoxidation of leuco malachite green, adsorbed on silica gel. They found that this reaction can be sensi- tized, in an atmosphere of oxygen of very low pressure (10"^ mm.), by the dyestuff trypaflavine, adsorhed on separate particles of the gel. Kautsky ex- plained the "transmission" of the sensitizing action across the air gaps sepa- rating the dyestuff from the acceptor by the assumption that excited dye- stuff molecules transfer their energy to oxygen molecules : (23.14) D* + O2 > D + O2* This process he also made responsible for the quenching of fluorescence. Energy transfer was supposed by him to convert ordinary oxygen into a metastable active form, which Kautsky identified as the state ^S, known from spectroscopic data to be situated 37.3 kcal/mole above the ground state ^n. After Gaffron remarked that infrared excitation of bacterio- chlorophyll provides <37 kcal, Kautsky (1937) suggested that the state ^A (23 kcal/mole) could sei-ve the same purpose. Both states are meta- stable because their multiplicity (singlet) is different from that of the ground state (triplet). The same principle underlies Lewis and Kasha's more recent theory of metastable triplet states in molecules with singlet ground states (c/. p. 730). An alternative chemical explanation of the mechanism of quenching 7S0 FLUORESCENCE OF PlfiMENTS IN VITRO CHAP. 23 by oxj^gen was suggested by Weiss (1935). He postulated an autoxidation (dehydrogenation) of the dye: (23.15) D* + O2 > HO2 + oD (o signifies oxidized). The radical HO2 can diffuse across air gaps and cause the oxidation effects ascribed by Kautsky to metastable oxygen molecules. A reaction of the type (23.15) could be responsible not only for the quenching of the fluorescence of adsorbed dyes, but also for that of the fluorescence of dyestuffs dissolved in organic solvents. Since the quantum jaeld of irreversible photoxidation of chlorophyll in pure organic solvents is very low (Vol. I, p. 496), reaction (23.15), if it is responsible for quenching, must be practically completely reversible (at least as far as the chemical composition of chlorophyll is concerned). The restoration of oxidized chlorophyll may be brought about either by direct reversal of (23.15), or — if the HO2 radicals are partly consumed by dismutation or side reac- tions— by interaction with the solvent. In the latter case, the net result is sensitized autoxidation of the solvent S : (23.16a) Chl*+ O2 > oChl + HO. (23.16b) oChl + S > oS + Chi (23.16c) HO. > ^ H2O + f O2 (23.16) S + i O2 > oS + m.O For the solvent, one may substitute an oxidizable substrate — e. g.' benzidine, or potassium iodide — thus obtaining a mechanism of chloro" phyll-sensitized autoxidation of such substrates. (This mechanism wa^ discussed in Volume I, chapter 18, cf. equations 18.33, and chapter 19, cf. scheme 19.11; there, tautomerization was added as a preliminary step.) An alternative interpretation of ser^sitized autoxidation, also discussed in Volume I, chapter 18 (cf. equations 18.40), envisages a primary reaction between excited chlorophyll inolecules and the oxidation substrate (rather than oxygen). Whenever this mechanism operates, oxidation substrates should quench the' fluorescence of chlorophyll more effectively than does oxygen. Franck and Levi (1934) measured the quenching of chlorophyll fluorescence by benzidine and potassium iodide. Their quenching curves are not labeled and therefore do not permit reading off the half-quenching concentration, but the authors state that, under the conditions of Noack experiments on the chlorophyll-sensitized autoxidation of benzidine (cf. page 528), quenching by benzidine must have been many times more ef- ficient than that by oxygen. If this is true, then the mechanism of this reaction must be different from that of chlorophyll-sensitized autoxidation of substrates such as allylthiourea (cf. below). niYRICAL AND OHEMK'AL QUENCHING 781 Li^•ingston and co-workers (Livingston 1948, Livingston and Ke, 1949) made the first systematic investigation of changes in the intensity of chlorophyll fluorescence, produced by small admixtures. They used various organic compounds, certain salts and several gases. As long as the fluorescence remains the only property measured, all the observed changes can be described as quenching (if fluorescence be- comes weaker) or stimulation (if fluorescence becomes stronger — as it actually does upon addition of traces of an alcohol or amine to a chlorophyll solution in dry hydrocarbon, or upon the addition of iodine to an alcoholic solution of chlorophyll h) . It would be best, however, to restrict the terms "(luenching" and "stimulation" to cases in which the admixture does not affect the composition or state of the light-absorbing molecules in the dark, but acts only on molecules which have been excited by the absorption of hght. We may refer to these phenomena as "true quenching" (or "true stimulation"— if the latter does exist at all, which is doubtful). True quenching can be due to physical or chemical processes. In the first case (physical quenching), kinetic encounters of light-excited, fluorescent molecules with the molecules of the quencher, or mutual proximity of these molecules, lead to accelerated conversion of electronic excitation energy into vibrational energy and, ultimately, into heat. The accelerated dissipa- tion can occur within the excited molecule itself (because its configuration or charge distribution change under the influence of the quencher), or in a complex formed by the excited molecule and the quencher, or even within the quencher molecule alone — which in this case, must first take over the electronic excitation energy "in bulk" and then dissipate it, by internal conversion to vibrational energy. (It is also possible for this energy to be re-emitted by the quencher as sensitized fluorescence.) In the second case (chemical quenching), either the excited molecule or the molecule of the quencher (or both) are changed chemically in the process of quenching. In this case, a kinetic encounter of the two molecules is needed. If the photochemical reaction responsible for quenching is completely reversible by a dark reaction, the net result is the same as in physical quenching— conversion of light energy into heat. Otherwise, a net photochemical change remains; and only if this change does not involve the fluorescent species is a steady yield of fluorescence observable in the presence of the quencher. Contrasted to true quenching can be the changes in the yield of fluores- cence which are caused by alterations in the composition or structure of the light-absorbing molecules produced by the addition of the admixture. (AH cases of "stimulation" probably belong to this class.) These processes can be distinguished from true quenching by the fact that the absorption spectrum of the solution also is changed by the presence of the quencher 782 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 (or stimulant). The absorption changes may be major or minor, depending on the character of the interaction (complexing, tautomerization, dissocia- tion, oxidation, reduction, etc.), but should never be entirely absent. Furthermore, chemical changes in the dark will often (albeit not always) proceed at a measurable rate, thus causing the yield of fluorescence to depend on the length of time between the preparation of the mixture and the illumination. Finally, the dependence of quenching on the concentra- tion of the quencher should be different in the case when the absorbing molecule and the quencher combine (or react) in the dark than when they interact only after the absorption of light. Table 23.IIIC Quenching Data for Chlorophyll a in Methanol, Ethanol and Acetone (after Livingston and Ke, 1949) Quencher Solvent X (excit.), m^i [0] 1/2 " *' '' mole/L Chloranil Me.CO Quinone Me2C0 MeOH Methyl red MeOH Trinitrotoluene MeOH MeOH m-Dinitrobenzene MeOH MeOH Duroquinone MeOH /3-Nitroso-a-naphthol MeOH /3-Nitrostyrene MeOH Nitric oxide EtOH /3-Nitro-i3-methylstyrene MeOH Oxygen EtOH Nitrobenzene MeOH /3-Nitro-)3,7-hexene MeOH o-Aminophenol MeOH Phenylhydrazine EtoO MeOH Dimethylaniline MeOH 2-Phenyl-3-nitrobicyclo- [l,2,2]-heptene-5 MeOH 2,6-Diaminopyridine MeOH "Half-quenching concentration. * Cf. equation 23.16C, p. 785. These general considerations should be kept in mind in the evaluation of the results of quenching experiments. They make it particularly im- portant that measurements of the intensity of fluorescence be combined with measurements of the absorption spectrum (and if possible, also of the fluorescence spectrum) of the light-absorbing species. 645 0.0050 175 645 0.0081 143 645 0.0096 120 645 (0.0088) 101 645 0.0100 90 435.8 0.0098 113 645 0.0122 68 435.8 0.0110 77 645 0.0116 81 645 0.0140 62 435.8 0.0170 59 435.8 (0.0178) 43 435.8 0.022 41 435.8 (0.023) 35 435.8 0.034 36 435.8 0.064 12 435.8 0.138 7.2 435.8 0.151 8.4 435.8 0.31 3.7 435.8 (0.42) 2.4 435.8 (0.61) 1.6 435.8 (0.62) 1.4 QUENCHING EFFICIENCY '83 The experiments of Li\'ingston and Ke (1949) dealt largely with what appears to be true chemical quenching: changes due to reversible reactions of excited chlorophyll molecules with certain organic and a few inorganic molecules. In some cases, at least, this interpretation was confirmed by observations of the constancy of the absorption spectrum, and by the in- stantaneous character of the change. Table 23.IIID NONQUENCHERS OF CHLOROPHYLL FLUORESCENCE IN METHANOL, EtHANOL OR AcETONE" (after Livingston and Ke, 1949) Reagent Solvent [Q], m./I. Fo/F Nitropropane MeOH 0 .09 1 .00 Nitropropane MeOH 0.2 1.02 Butvl nitrate MeOH 0.9 1 .02 Butvl nitrite MeoCO 0. 138 101 Phenvlhvdroxvlamine MeOH 0.021 1 .02 Phenvlhydroxylamine MeOH 0 . 07 1 . 05 Aniline Et.OH 0.16 1 .02 Hydrazine MeoCO 1 .05 1.0 Urethan MejCO 0 .07 1.0 Thiourea Me.CO 0.02 1 .0 2-Aminopvridine MeOH 0 .08 1 .00 Phenvlurea MeOH 0.05 1.01 Urea^ MeOH 0.19 1.00 Guanidine carbonate MeOH (Satd.) 1.0 Phenol MeOH 0.09 1.0 Hvdroqninone* MejCO 0.03 1.0 Phenolphthalein MeOH 0 .04 1.0 Dimethvlo;lvoxime MeOH 0 . 07 1.0 /e/7-Hexvlmercaptan MeOH 0.11 1 . 00 Benzaldehvde MeOH 0.38 1.0 Benzoic acid Me.CO 0.08 1.0 Camphor MeOH 0.15 1.0 Boric acid MeOH (Satd.) 1 .00 Sodium methoxide MeOH 0.05 1.0 Sodium cyanide MeOH 0. 15 1 .01 Sodium oxide MeOH 0.11 1.0 Nitrous oxide EtOH (605 mm.) 1.0 Carbon dioxide EtOH (576 mm.) 1.0 Carbon monoxide MeOH (640 mm.) 1 .00 Potassium thiocyanate MeOH 0 05 1 . 02 Potassium thiocyanate MejCO 0 . 004 0 . 92 " Other nonquenchers: a^^corbic acid; alkali iodide (Evstigneev and Krasnovsky, 1948): allylthiourea(r/. naffe789). ^ The measurements of Evstigneev and Krasnovsky (1948) indicate that hydroquin- one at much higher concentrations has some quenching action, but is less efficient than 2-diaminopyridine, the weakest quencher in Table 23.IIIC. Li\nngston and Ke used a 1.2 X 10~^ molar solution of chlorophyll a in methanol, ethanol, ether or acetone, which they excited by the mercury line 435.8 mju, or bj^ a red band centered at 645 m/z. Tables 23.IIIC and D show that quenching was similar for both types of excitation. The quenchers are listed in the first table in order of declining efficiency; it is 784 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 clear that oxidants (quinones, diazo dyes, nitro and nitroso compounds, and oxygen) are strong quenchers, while redudants (such as amines) are at best only weak quenchers. Among compounds listed in Table 23. HID, only phenylhydroxyamine showed any measurable quenching at all, at concen- trations up to 0.1 mole/1. See also footnote b to Table 23. HID. 1.00 j . ^ 1 / ^' .,< y ^>' /■' ^' /■ \P 2 ^.<.° ^.- .«- ^ ^ M j^' rt ^ ^ ^ . — .-cr' ..^ ^^ .0^-* 4.. «• ^ ..^•^ '- ^"^ Chi* ( > tChl) (23.17b) Chi* (or tChl) + A > oA + rChl (23.17c) rChl + M O2 > Chl*(+ I H2O) (23.17d) Chi* > Chi + hf (23.17) A + O2 )oA( + ^H20) Reaction (23.17c) is an exception from the rule that back reactions in photochemical sensitizations occur unthout chemiluminescence. In chapter 24, we will quote evidence of a chemiluminescence accompanying the back reactions in photosynthesis. LONG-LIVED ACTIVE STATES AND AFTERGLOW 795 Another possible cause of phosphorescence — with a spectrum somewhat different from that of instantaneous fluorescence — can be the direct, radia- tive return of the metastable molecules from state T into the ground state X, with the emission of a quantum. This is an alternative to the above- mentioned, nonradiative return by internal conversion of electronic into vibrational energy. The assumption that this type of phosphorescence alone limits the life-time of the metastable state is the basis of the above- mentioned calculations of Lewis, Kasha and McClure. Long-lived lumi- nescence, with a frequency 2000-20,000 cm.-^ lower than that of direct fluorescence, actually is known for many dyestuffs and other fluorescent organic compounds, particularly at low temperatures in glassy solvents. In the case of chlorophyll, a phosphorescence of this type would have to be sought in the infrared. Calvin and Dorough (1947) reported that in a mixture of chlorophyll a and h, dissolved in a "rigid solvent" (EPA = mixture of ether, pentane and alcohol solidified <-100° C. without crystalHzation), an afterglow lasting 0.2 sec. can be observed after illumination. Spectroscopic obser- vation revealed a band beginning at 780-800 m/z and stretching into the infrared. Similar results were obtained with zinc tetraphenylchlorin, but not with copper tetraphenylchlorin— a difference ascribed by Calvin to weakening, by the paramagnetic Cu'^+ ion, of the metastability of the triplet state. Livingston and co-workers (1948) found no such afterglow, at —180° or -150° C. (2 X 10-^ mole/1, chlorophyll a in EPA). Experiments in other solvents and at other temperatures also gave negative results— with chlorophyll a + & as well as a, and at concentrations from 10 ~- to 5 X 10-^ mole/1., in air or in vacuum. According to Livingston, a personal communication from the Berkeley group confirmed that the luminescence of Chi (a + h) at 800 m^, reported by Calvin and Dorough, probably had been due to impurities. However, Berkeley observers asserted that chloro- phyll b does have a weak infrared afterglow (r = 0.02 sec), starting at 860 m/i. This phosphorescence, if it exists, should be in direct competition with photochemical sensitization by chlorophyll (Weiss, 1948). 8. Summary — A Scheme of Fluorescence and Sensitization In summing up the discussion, we may go back to Volume I (chapter 19) and reproduce again, in a somewhat amplified form, scheme 19.11 796 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 given there to describe the most Ukely mechanism of fluorescence and sen- sitized photoxidation in chlorophyll solutions.* This scheme, amplified to include also internal conversion, self-quenching by collisions and photo- chemical reaction with the oxidation substrate A— c. g., an amine— (but simplified as far as the mechanism of sensitized photoxidation is concerned) , is reproduced in scheme 23.11. To prevent the scheme from becoming too Chl" 6 © tChl ® p, nmst depend on oxygen pressure in this rlnge; and if, at [Oil = 10 ■• to lO^" mole/1.. Aj !<),] i.s«; k, and Af, |(),] is»At', . . ." SCHEME OF FLUORESCENCE AND SENSITIZATION 797 dized to oA (two alternative direct reactions of Chi* and two alternative reactions of the metastable tChl). It neglects all resonance effects. The yield of fluorescence, ^, is according to scheme 23.11: (23.18) f = ^' (Av + A-. + kt) + A-JChI] + A-*[0.] + A-* [A] (This equation is to replace equation 19.5.) Some of the constants in (23.18) can be estimated. kf is 1.2 X 10^ (inverse of the natural life-time of Chi*, as calculated from the intensity of the red absorption band on page 634) . ki may be small compared to A;, — in other words, tautomerization may be a normal intermediate step of internal conversion. This is, however not certain. The quantum yield 7 ^^ 1 was found (c/. Tables 18.11 and 18. Ill) at [A] ^ 5 X 10~^ — i. e., concentrations at which k% [A] may well be high enough to make ki insignificant even if it is not small compared to ki + ki must be about 10^, if one assumes a fluorescence \aeld of ~ 10% in the absence of quenching by O2 or A. fcj should be >10^2 according to Weiss and Weil-Malherbe (c/. page 773), but more probably is about 10^ (the value obtained by Weiss and Weil-AIalherbe for dyes other than chlorophyll). Figure 23.7 indicates half-quenching at [Chi] = 1.5 X 10~2 mole 1., and thus fc, = 67 and kc^^7 X 10^ However, the sigmoid shape of the curve points to resonance transfer rather than collisions as the main quenching mechanism. A;o is about 5 X 10^ calculated from a "half-quenching" concentration of 2 X 10-2 mole/1, (cf. Table 23.IIIC). k\ could be calculated from Franck and Levi's curves for the quench- ing of fluorescence by benzidine, if the concentration units used were known. (The A'l values in Table 23.IIIC are bimolecular rate constants in sec. X 1. X mole"'^ the h, k* and A;a values in eciuation 25.18 are products of these constants and the rate constant of monomolecular deactivation, kf -\- ki -\- k,^ 10«sec.-i.) We refrain from an attempt to deduce from scheme 23.11 an equation for the quantum >aeld of sensitized autoxidation (to replace equation 19.6'), because the result is much too complicated to allow comparison with the experimental data, e. g., with Gaffron's equation (18.32) for the sensitized autoxidation of allylthioui-ea. The "self-quenching" reaction of the tau- tomer tChl: (23.19) tChl + Chi » 2 Chi has been added to account for the decrease in the yield of chlorophyll- sensitized photoxidation with increased pigment concentration [Chi], in 798 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 the [Chi] region where self-quenching of the short-lived state Chi* cannot be significant (cf. Table 18. Ill, page 513, Vol. I). This "deactivation" of tautomeric chlorophyll by normal chlorophyll can perhaps occur via dis- mutation : (23.20) tChl + Chi > oChl + rChl (as suggested, e. g., in equation 18.42b, page 519, Vol. I) followed by back reaction : (23.21) oChl + rChl » 2 Chi B. Fluorescence of Carotenoids and Phycobilins in Vitro* 1. Fluorescence of Carotenoids Carotenoids are usually described as nonfluorescent. According to Willstatter and Stoll (1918), this is true both of the leaf carotenoids and of the carotenoids of brown algae (e. g., fucoxanthol) . However, de Rogovski (1912) asserted that he has observed a fluorescence of carotene in petroleum ether, in the region 505-600 m/x, and Dhere and Castelli {cf. Dhere 1939) stated that, at —180°, three separate fluorescence bands can be observed in carotene solution in xylene. Klein and Linser (1930) mentioned a green fluorescence of carotene solutions in alcohol. Strain (1936) found, in the chromatograms of leaf extracts in petroleum ether, a fluorescent layer situated below that of carotene a, and belonging to an unknown color- less substance, probably a hydrocarbon without sharp absorption bands in the visible or near ultraviolet. Zechmeister and co-workers (see, for ex- ample, Zechmeister and Sandoval 1945, 1946) found a fluorescent, colorless polyene hydrocarbon with sharp absorption bands at 331, 348 and 367 m^u (in petroleum ether) to be present in great abundance in extracts from various plant organs (fruits, stems etc., but not chlorophyllous organs, such as grass, leaves or green needles). This hydrocarbon, called phyto- fluene (probably C40H64), may be a hydrogenation product (or precursor) of the carotenes; it contains seven double bonds, with probably only five of them conjugated. Absence of fluorescence indicates, according to page 799, that the exci- tation energy of carotenoid molecules in solution is dissipated within less than 10~i^ sec. * Bibliography, page 804. FLUORESCENCE OF PHYCOBILINS 799 2. Fluorescence of Phycobilins The phycobilins are usually described as brilliantly fluorescent. How- ever, since no exact determinations of the yield of fluorescence exist, it is impossible to judge whether the fluorescence is so much stronger than that of chlorophyll, or whether its greater brilliancy is due to the fact that the fluorescence bands of the phycobilins lie near the region of the greatest sensitivity of the human eye, whereas those of chlorophyll are in the far red, and partly in the infrared. Association with proteins does not impair the fluorescence of phycobil- ins. On the contrary, according to Lemberg (1930), the isolated pigments fluoresce less strongly than the chromoproteids. The fluorescence of red algae was first described by Stokes, in the same paper in which he also reported the discovery of the fluorescence of green leaves (c/. page 805). Since then,* the fluorescence of both the living algae, and of their aqueous extracts, has repeatedly been observed, e. g., by Schiitt (1888), Hanson (1909), Turner (1916), Lemberg (1928), Dhere, and Fontaine (1931), Roche (1933), Dhere and Raffy (1935), Van Norman. French and Macdowafl (1948), Arnold and Oppenheimer (1949) and French (1951). Van Norman, French and Macdowall (1948) gave a photometric curve for the fluorescence of an extract obtained by grinding a species of the red alga Iridaea under water and centrifuging at high speed. It shows a sharp peak at about 580 m/i, clearly related to the first long-wave absorption band of phycoerythrin at 566 m/x (cf. fig. 23. 9A, from French 1951). A shoulder appears on the long-wave side of the 580 m/x fluorescence band, indicating the presence of a second maximum at about 630 m/x. This probably is the second (0-^1) fluorescence band of phycoerythrin (leading to a vibrational ground state). The first fluorescence band of phycocyanin (correlated with the first absorption band of this chromopro- teid, the peak of which appears, in the same extract, at about 615 m^) lies at 660 m^, according to the curve obtained by French (1951) by subtraction of the phycoerythrin fluorescence from the fluorescence spectrum of the crude aqueous extract from a red alga (fig. 23. 9B) . Earlier, Dhere and Fontaine (1931) gave 578 and 648 m/x, respectively, as the axes of the fluorescence bands of the two phycobilins in aqueous extract. Dh^re and Raffy (1935) noted a second phycocyanin band at about 728.5 m/x. According to French's figures, the fluorescence bands of the phycobilins are shifted to the red of the absorption bands, by 14 m/x in the case of phycoerythrin, and by 45 m/x in tliat of phycocyanin. 800 FLUORESCENCE OF PIGMENTS IN VITRO CHAP. 23 anin). Figure 24.1 shows the fluorescence spectrum of a Pelargonium leaf (No. 7 on panchromatic plate, and Nos. 2 and 3 on Agfa Infrared 730 plate), compared with the absorption spectrum of the same leaf (No. 6) and with the fluorescence spectrum of chlorophyll in ether (No. 4). We note the 808 FLUORESCENCE OF PIGMENTS IN VI VO CHAP. 24 close coincidence of the fluorescence band in vivo (No. 7) with the corre- sponding absorption band (No. 6), and the "red shift" of the fluorescence band in the Uving cell compared to its position in solution (Nos. 2 and 3 compared to No. 4). ... I ii_ o o o O o o Wave lenqth, mu: mo mo m o Spectrum: ^ Reference Fluorescence Reference Absorption Fluorescence Reference o o o o o o o If) O ID O tn O m r^ h- <0 ^O tf> tf> sf o o o o Fig. 24.1. Fluorescence spectrum of living leaves of Pelargonium (Nos. 2, 3, 7) and of chlorophyll in ether (No. 4), compared with the absorption spectrum (No. 6) (after Dhere and Raffy 1935). Dhere and Raffy (1935) suggested that the second fluorescence bands of chlorophylls a and b in vivo, situated in the near infrared, may account for the striking brightness that green vegetation exhibits on landscape photogi'aphs on infrared-sensitive plates (c/. fig. 22. 31 A). However, the fluorescence of living loaves is much too weak to produce such a spectacular effect. Mecke and Baldwin (1937) disproved Dhere's theory by showing that the vegetation remains dark when illuminated with infrared-free light and photographed through a filter that transmits only the infrared. The brightness of green plants in infrared light is thus due to lack of absorption, and not to fluorescence. Dhere and Biermacher (1936) photographed the fluorescence of Pelar- gonium leaves on plates whose sensitivity extended far into the infrared, and foimd a new band, with an axis at 812 m/x, extending to 830 m/i. In Table 24.11, the wave lengths of the peaks of the four known fluorescence bands of the chlorophylls a and h in living cells are compared with the wave lengths of the corresponding bands in ethereal solution. Table 24.11 shows that, in the living cell, the fluorescence bands are shifted by 5-15 m/x toward the infrared from their positions in ethereal solution — i. e., by about the same distance as the corresponding absorption FLUORESCENCE SPECTRA OF PLANTS 809 Table 24.11 Fluorescence Bands of Chlorophyll in Plants and in Ethereal Solution" Axis, m/ifc Maximum, m^ Band Plant Soln. Plant Soln. « Chlorophyll a I . . II. . III. . Chlorophyll b I . . Ill ; .' 680, 685 740 812 656 672'^ 736 801 651'' 713 789 681-685 665 720 649 708 " On Ilford Infrared Plates. * Biermacher (1936). •^ cf. table 23. I. '' The difference between these values and those in Table 23.IB is attributed by Biermacher to the decrease of the sensitivity of the infrared-sensitive plate in the orange. The values in Table 23. IB are supposed to be the correct ones; they are not used here, to preserve the consistency of Table 24.11. bands. The strong shift of peak I is due to self-absorption. French (195/1) found it at 675 m^ in a faintly green, and at 685 m^t in a dark-green leaf. The intensity ratio I: II was 5: 1 in the first, and 1 : 1 in the second, case. Vermeulen, Wassink and Reman (1937) gave spectrophotometric curves of the fluorescence of the alga Chlorella (chlorophyll a + 6) and the bacterium Chromatium (bacteriochlorophyll), which are reproduced in figures 24.2 and 24.3. Chlorella shows a peak at 680 m// and only a shoulder at 740 m^i. Duysens (1951) found that, even in light absorbed by chloro- phyll b, Chlorella shows only the fluorescence of chlorophyll a — indicating efficient energy transfer from h to a. • Illuminated with Hg lomp with UG2 filter o Illuminated with total Hg radiotion 650 700 750 WAVE LENGTH, m/i Fig. 24.2. Fluore.scence spectrum of Chlorella suspension (after Ver- meulen, Wassink and Reman 1937). Illuminated with Hg lamp with UG 2 filter Illuminoted with Hg lamp with yellow-green filter Illuminated with total Hg rodlation ./ y 800 750 800 850 900 950 WAVE LENGTH, m^ Fig. 24.3. Fluorescence spectrum of Chro- matinm suspension (after Vermeulen, Was- sink and Reman 1937). 810 FLUORESCENCE OF PIGMENTS IN VIVO CHAP. 24 Comparison of the figures 24.3 and 23.4 shows that the displacement of the fluorescence band is much stronger in the purple bacteria than in green plants. The fluorescence spectrum of live Chromatium shows only one band, at 926 mju, while that of the extract contains two bands, at 806 and 695 mju, respectively. It was mentioned on page 751 that the first and more intense of these two bands can be correlated with the main ab- sorption band of extracted bacteriochlorophyll, at 770 m/i, but that the correlation of the 695 m^l fluorescence band with the 605 m/z absorption band is doubtful. Absorption spectra of live purple bacteria show two (or three) absorption bands, at 800-870 and 800 mtx, respectively (c/. p. 702); but here again, only the first one can be identified with the main X-^ Z absorption band of dissolved bacteriochlorophyll (at 770 m^), while the identification of the second one— which is comparatively weak and variable in intensity— with the X -^ Y band at 605 mfx is uncertain (cf. page 702). The relation between the fluorescence and the absorption bands of bacteriochlorophyll is illustrated by Table 24.IIA. The table shows that the fluorescence band I in vivo is shifted toward the infrared by as much as 120 m^, compared to the position of the fluorescence band in vitro, and by about 60 mn compared to the position of the corresponding absorption band in vivo. Table 24.IIA Spectra of Bacteriochlorophyll in Vitro and in Vivo I, mti II, ni/i III, mil Band {X -» Z) (X -* Y) ? Absorption Cell 860-870 — 800 Extract 770 605 — Fluorescence Cell 926 — — Extract 806 695 (?) — This tabulation indicates that, in the living cell, bacteriochlorophyll fails to show one absorption band and one fluorescence band that are found in extracts, but shows one (or two) extra absorption bands, without corresponding fluorescence bands, which have no counterpart in the solution spectrum. Duysens (1951) confirmed that Chromatium and Rhodospirillum show only one fluorescence band, correlated with the absorption band at 890 m^. Absorption in the 800 and 850 m^ bands contributes to the excitation of this fluorescence band; so does — with a 50-70% lower efficiency— the absorption by cartenoids. Excitation energy is thus transferred from all pigments of the bacteria to the bacterio- chlorophyll form having the lowest excitation energy. FLUORESCENCE SPECTRA OF PLANTS 811 The fluorescence of green bacteria (Chlorohium mirahle), probably due to "bacterioviridin" (cf. Vol. I, p. 407), was observed by Buder (1913). A fluorescence band at 635 m^, noted by French (1951) in a partially green leaf could be due to photochlorophyll, whose absorption band lies at 620-630 ni/x in ether. French estimated its position in vivo as 650 niju (from the action spectrum of chIoroph\'ll forma- tion), but noted that this is incompatible with the location of the fluorescence at 635 m/x. The fluorescence of the phycohilins in red and blue-green algae is of great interest, because it permits a study of the interaction of two different fluorescent pigments in one cell. The first photograph of the fluorescence spectrum of Rhodymenia, made by Dhere and Fontaine (1931), showed one band in the orange (phycoerythrin) and one in the red (chlorophyll and phycocj^anin). Van Norman, French and Macdowall (1948) determined fluorescence curves of two red algae — Gigartina and Iridaea; both showed three peaks, at 575 myu (phycoerythrin, cj. p. 799), 055 m^ (phycocyanin, cj. p. 800) and 700 m/x (chlorophyll, probably a -^ d, cf. beloAv). French (1951) gave fig. 24.4 for Porphyridium: Only chlorophyll fluoresces when cells are excited with X 436 m^ or 450 m/x, i.e., by light absorbed by chlorophyll and caro- tenoids only. Phycohilin bands develop with excitation by 470, 490, and 546 m/x but, even though most of the incident light is now absorbed by phycoerythrin, chlorophyll fluorescence remains strong. u o z UJ o CO UJ cr o 3 T~l I I I I I I I I I I I I I I I 436 m;i (a) 476 m/i (b) I I I I I I I I I J__L I I I I I I I I I I I I I I I I I I I I I I I I I 546mjLi (c) 1 I I I I I 1 I I I I I I I I I I I I 600 650 700 750 600 650 700 750 Fig. 24.4. Fluorescence spectra of a red alga when illuminated with equal energies of different wavelengths. Courtesy L. N. M. Duysens. 812 FLUORESCENCE OF PIGMENTS I IV VIVO CHAP. 24 The chlorophyll a band in fig. 24.4 is in the usual position — at 685 mix. However, an additional band is indicated at 730 m/u; it can be attributed to chlorophyll d. The weakness of chlorophyll d absorption (mere ripple in fig. 22.20!) makes one suspect that chlorophyll d fluorescence is excited mostly by energy transfer from other pigments. This is strikingly con- firmed by the observation, reported by Duysens (1951), that an "uniden- tified pigment" — presumably chlorophyll d — whose absorption, in Por- 420rTyi r\ ^°' 150 - r\ ■ >% /■ K (A /! \\ C J \ £=100 - o> i \ u c » *v 4> \\ O pi cJ i «. ' A S 50 3 •? ^ _> O / ■' • / 1 -o— 2 650 700 750 wave length in m>i algae —° ° ° ^^~ chlorophyll —^ — phycocyanin -■ — sum of chlor. a and phycocyanin - unknown pigment - 546nn;i (b) 150 A >s wn ' /; i « \ >^ /• ' * >/>k •i=IOO / 1 II x ' o c 0) // -A \ ative fluoresc en o 1 rff 1 L 9> 650 700 750 wave length in m>i algae — o o o o o chlorophyll — ^> •<> ° ° phycocyanin — ■ • ■ ■• sum of chlor. a. and phycocyanin ■*■ -^ ■* unknown pigment — • • • Figure 24.5 yhyra lacineata, is < 0.1% of that of chlorophyll a, emits, when 420 m/x is used for excitation, ten times more fluorescence than chlorophyll a (fig. 24.5). With excitation by 546 mfx, the chlorophyll a band at 685 m/i is slightly higher than the "chlorophyll d" band at 730 m/jL. In agreement with French (fig. 24.4) the phycocyanin fluorescence band at 665 m^u is very weak in violet exciting light but becomes strong in green light. 2. Fluorescence Yield and Sensitized Fluorescence in Vivo All observers concur that the fluorescence of chlorophyll in the living cells is "weak," but absolute measurements of its yield are very few. Ver- meulen, Wassink and Reman (1937) determined the quantum yields of fluorescence in four Chlorella suspensions, and found values between 0.15 and 0.30%; Wassink and Kersten (1944) found a yield of about 0.15% FLUORESCENCE YIELD 7A^ VIVO 813 in a suspension of diatoms. In Chromatium (a purple bacterium) Vermeu- len, Wassink and Reman (1937) at first found a much smaller yield — between 0.005 and 0.01%. These values are incredibly low for a "fluores- cent" material. (Theoretically, the lowest yield to which fluorescence could sink even in a "nonfluorescent" pigment is about 0.001%, since the ratio of the period of a molecular vibration and the life-time of electronic excitation is lO^^^-yiO"^-^ = 10~^) A redetermination of the yield of fluorescence of Chromatium by Wassink, Katz and Dorrestein (1942) in fact gave considerably larger figures — of the order of 0.1%, i. e., similar to those found in algae. Self-absorption may have been the cause — or at least, one cause — ^of the error of the earlier determinations. 500 550 SCO WAVE LENGTH, m/x 650 Fig. 24. 5A. Yield of fluorescence of Chlorella suspensions in relation to wave length of exciting light (after Vermeulen, Wassink and Reman 1937). 4 sets of measurements. lOOp 450 700 750 500 550 600 650 WAVE LENGTH, m/i Fig. 24. 5B. Yield of fluorescence of Chromatium suspensions in relation to wave length of exciting light (after Vermeulen, Wassink and Reman 1937). 4 sets of measurements. The yield of fluorescence in live blue-green algae (Chroococcus) was esti- mated by Arnold and Oppenheimer (1950) by a rather crude method (visual comparison with the intensity of light scattered by a block of magnesium) ; they found it to be of the order of 1.5% — about ten times higher than the yield of fluorescence in green cells. Presumably, this fluorescence origi- nates predominantly in phycocyanin, although chlorophyll, too, may con- tribute to it. Interesting results were obtained in the study of the effect of wave 814 FLUORESCENCE OF PIGMENTS 7A^ VIVO CHAP. 24 length of the exciting hght on the fluorescence of live cells. Vermeiilen, Wassink and Reman (1937) found that the spectral distribution of the fluor- escent light of Chlorella and Chromatium is independent of the wave length of exciting radiation. The quantum yield of fluorescence (<^) also was ap- proximately constant, between 442 and 624 mju in Chlorella, and between 450 and 750 ray. in Chromatium. However, a slow systematic decrease of fp was observed in Chlorella at the shorter waves — a trend that became accelerated below 424 m^u (fig. 24. 5A). In Chromatium., maxima and minima of (p were observed in two or three places in the visible spectrum (fig. 24. 5B). The Dutch investigators concluded from these observations that the quantum yield of chlorophyll fluorescence in vivo does not depend on wave length except when carotenoids interfere with the light absorption by chlorophyll or bacteriochlorophyll. In green plants, this occurs only below 520 mju ; the carotenoids of purple bacteria, on the other hand, have absorption bands in the green, yellow and orange {cf. Table 21. IX and fig. 22.27), and these bands could perhaps account at least for the first minimum of the fluorescence yield noticeable in figure 24.5B at about 550 mju. While qualitatively the conclusions of Vermeulen and co-workers ap- pear plausible, quantitative considerations lead to some interesting compli- cations. In Chlorella, for example, ^ declined, in the violet, by only 10 or. 20%, while figure 22.43 indicates that the carotenoids must account for at least one third the total absorption in this region! It thus appears as if the chlorophyll fluorescence can be excited, with considerable probability, also by the light absorbed by carotenoids! This hypothesis has been strikingly confirmed by experiments with fucoxanthol-containing diatoms. The results of these experiments, carried out by Dutton, Manning and Duggar (1943), are shown in Table 24. III. They indicate that the yield of chlorophyll fluorescence is the same, whether it is excited by red light, ab- sorbed exclusively by chlorophyll, or by blue-green light (470 m^u), three quarters of which probably is absorbed by carotenoids, mainly fucoxan- thol (see fig. 22.46 and figs. 30.9B and C. Table 24.III also contains new results with Chlorella, which confirm the conclusions drawn above from the earlier work of the Dutch observers. These results indicate that the light absorbed by carotene and luteol is almost — but not quite — as ef- ficient in the excitation of chlorophyll fluorescence in green algae as the light absorbed by fucoxanthol in diatoms. In striking contrast is the result of the experiment with an acetonic extract from Nitzschia ; here, light absorbed by the carotenoids is completely lost for fluorescence. The ex- periment with acetonic solutions of chlorophylls a and h shows that the quantum yield of chlorophyll fluorescence in solution does not increase with SENSITIZED FLUORESCENCE IN VIVO 815 the wave length of exciting Hght between 436 and 578 m^u. This proves that an increased yield of nonsensitized fluorescence of chlorophyll in violet light cannot be offered as alternative explanation of the results of the first two experiments. Excitation transfer from carotenoids to chlorophyll might become possible in sufficiently concentrated solutions; we men- tioned on p. 790 that Duysens (1951) reported excitation transfer from chlorophyll b to chlorophyll a in 10 ~^ M solution in acetone. Table 24. III. Direct and Carotenoid-Sensitized Fluorescence of Chlorophyll IN Diatoms (after Dutton, Manning and Duggar 1943) Ratios of fluorescence yields excited by the two wave ! lengths Fraction of absorbed energy Excitation absorbed Expected (with- wave length.? by clilorophyll" out energy iSIaterial compared, ni/i % transfer) Observed Nitzsckia closterium 470 vs. 578 or 600 26 ys. 95 or 99 0.27 1.2 ± 0.2 436 vs. 578 or 51 vs. 0.53 1.1 =i= 0.2 600 95 or 99 Chlorella pyrenoidosa . . . 470 vs. 578 or 600 52 vs. 100 0.52 1.05 ± 0.04 436 vs. 578 or 81 vs. 0.81 0.93 =fc 0.18 600 100 A^. closterium, acetone extract 470 vs. 600 19 vs. 100 0.19 0.22 ± 0.02 436 vs. 578 40 vs. 99 0.40 0.49 ± 0.07 Acetone soln. of Chi a and b 436 vs. 578 100 1.0'' 0.97 zfc 0.06'' ° These estimates are rather crude approximations {cf. chapter 22, page 726), but it seems improbable that errors could be large enough to account for all the differences between the calculated and observed ratios in the two last columns. '' The proportion of light absorbed by these two chlorophyll components is nearly equal at 436 and 578 m/j.. Hence the yield ratio should be approximately 1.0 despite differences in the fluorescence spectra of chlorophylls a and b. The experiments of Dutton and Manning are of importance for the theory of photosynthesis. They indicate that the energy absorbed by some carotenoids may become available to chlorophjdl in almost the same measure as that absorbed directly by the green pigment. Thus, chloro- phyll can play the part of a photocatalyst in photosynthesis, even when another pigment acts as a "primary" sensitizer. Results similar to those in Table 24. Ill have also been obtained by Wassink and Kersten (1946) with Nitzschia dissipata. Even more interesting are the results obtained with red algae. Van Norman, French, and Macdowall (1948) first found indications that chlorophyll fluorescence in Gigartina and Iridaea can be excited with equal (if not higher) intensity by light absorbed by phycoerythrin (at 560 mju) as by light absorbed by chlorophyll (at 650 m^u). French's (1951) fig, 24.4 clearly shows that light absorbed by chlorophyll (and, partly, by 816 FLUORESCENCE OF PIGMENTS IN VIVO CHAP. 24 carotenoids) at 436 mju, does not excite the fluorescence phycoerythrin or phycocyanin (cui-ve a). Light absorbed mainly by phycoerythrin, on the other hand, excites the fluorescence not only of both phycobilins, but also of chlorophyll curves a, h, c. The hump on the red side of the chlorophyll a band probably is due to chlorophyll d {cf. below). Duysens' (1951) fluorescence spectra of Porphyra lacineata (fig. 24.5) indicate that after excitation with 420 ra/x (absorbed by chlorophyll a and the carotenoids) over 90% of excitation energy is transferred to an "un- known pigment" (probably chlorophyll d); less than 10% of total fluo- rescence is emitted by chlorophyll a, and only a negligible proportion by phycocyanin. Excitation with 546 mju (light absorbed mainly by phyco- erythrin) causes strong fluorescence of both phycocyanin and chlorophyll a, and a comparatively weak fluorescence of "chlorophyll d." Duysens interpreted these results as indicating the existence, in red algae, of two kinds of pigment complexes: the largest part of chlorophyll a he suggested, must be coupled with chlorophyll d, and transfer practically all excitation energy to the latter pigment, although it is present in such a small amount as to be hardly noticeable in the absorption spectrum at all {cf. p. 812). A small part of chlorophyll a, not coupled with chlorophyll d, appears to be associated with the phycobilins, and serves as ultimate recipient of the major part of quanta absorbed by them. French's results indicate that the extent of energy "leak" into chlorophyll d must vary Avidely from species to species. In chapter 29, this hypothesis will be tied up with the results of quantum yield determinations by Haxo and Blinks, who noted a low photosynthetic efficiency of light absorbed by chlorophyll in some red algae. In chapter 32, we shall explore whether the paradoxical fact that fluorescence of chlorophyll a can be excited more strongly by light absorbed by phycobilins than by light absorbed by chlorophyll a itself, could be explained without Duysens' assumption of tAvo different pigment complexes in red algae. The general rule, indicated by the above-described fluorescence experi- ments, is that plant cells contain one (sometimes, perhaps, two) pigment complex in Avhich light energy absorbed by any one component tends to flow into the component with the lowest excitation level, and is therefore remitted mainly as fluorescence of the latter — even if it is present in a very low relative concentration. This picture is supported by Duysens' observations (p. 810) that all energy absorbed, in purple bacteria, by some carotenoids or by different forms of bacteriochlorophyll, flows into the form of bacteriochlorophyll that has the lowest excited level. On page 801, we mentioned the increase in the intensity of the phyco- bilin fluorescence observed by Arnold and Oppenheimer (1950) upon EFFECTS OF HEAT AND HUMIDITY 817 breaking Croococcus cells under water, and their interpretation of this effect as a consequence of suppression of energy transfer from excited phycobilin molecules to chlorophyll molecules, caused by dilution. If we assume that all chloroph^dl in the cells is present in the same form, then a fluorescence yield ip means the shortening of the normal life-time of the excited state, to, to r = ^ro. In the case of chlorophyll in ChloreUa, assuming ro = 8 X 10~^ sec. (page 634), we obtain: (24.1) T = (1.5 to 3 X 10-3) X 8 X IQ-^ = (1.2 to 2.4) X lO-" sec. and in the case of bacteriochlorophyll in Chromatiuni, assuming the same value of To : (24.2) r = 7 X 10-3 X 8 X 10-8 = 0.6 X iq-io sec. Two factors may determine the life-time of excited chlorophyll mole- cules in live cells, and thus account for the above-calculated small values of t: "normal" energy dissipation in the pigment-protein-lipide complex (chloroplastin) and quenching (or stimulation) of fluorescence by metabolic processes. The latter phenomena may themselves be of two kinds : direct "photochemical quenching" by competition between sensitized photo- chemical reaction and fluorescence, and indirect quenching (or stimulation) of fluorescence due to the metabolic formation of substances that diminish (or enhance) the fluorescence of chlorophyll. Observations described in section 3 can be interpreted as revealing changes in the general structure of the chloroplastin complex, while in section 4 — and, in more detail, in chapters 27, 28 and 33 — we will discuss variations in intensity of fluores- cence closely associated with participation of chlorophyll in photosynthesis. 3. Effects of Heat and Humidity on Chlorophyll Fluorescence in Vivo It has been found that, when chloroplast sediments (Noack 1927) or live leaves (Seybold and Egle 1940) are placed in hot water, their fluorescence vanishes almost immediately; at the same time, the red absorption band is shifted toward the shorter waves. This transformation occurs at a tem- perature of 64-72° C. If the leaves are kept in hot water for several minutes, fluorescence reappears, but the absorption band remains in the siiifted position. Metzner (1937) probablj^ dealt with the same phenom- (4ion when he described a "burst" of fluorescence caused by heating the chloroplasts under the fluorescence microscope. According to Seybold and Egle (1940), drying extinguishes the fluores- cence of fresh leaves, but not that of leaves killed l)y boiling. The fluores- 818 FLUORESCENCE OF PIGMENTS IN VIVO CHAP. 24 cence of some plants is highly sensitive even to minor changes in humidity : for example the fluorescence of Pleurococcus colonies on wood bark van- ished after one or two hours in an atmosphere of less than 80% relative humidity (at 25° C.) ; while the fluorescence of Mnium pundatum disap- peared when the humidity declined below 85%. The fluorescence of the leaves of Adiatum and Paretaria was found to be somewhat less sensitive, but it, too, ceased to be visible after one or two days in an atmosphere of 75% relative humidity. The fluorescence of sharply dried leaves cannot be restored by simple wetting, but returns upon immersion into boiling water. A similar trans- formation of the "sensitive" fluorescence of live cells into the "stable" fluorescence of dead cells can be achieved by freezing or immersion into ether. In the latter case, the fluorescence after the treatment is consider- ably stronger than it was in the living state. Seybold and Egle interpreted these results as indication that practically all chlorophyll in leaves is present in a nonfluorescent (probably, protein- bound) state, but that a small fraction of the pigment is dissolved in a lipide phase, and therefore capable of fluorescence. They suggested that, upon drying, the fraction of chlorophyll normally present in the lipide phase is transferred into the colloidal aqueous phase, while, upon heating, chlorophyll is first extracted from the lipide phase into the colloidal pro- teinaceous phase (thus causing the fluorescence to disappear), but later returns into the lipophilic material (concomitantly with the denatura- tion of the proteins and melting of lipides) , and thus again becomes fluores- cent. (Metzner 1937 also had attributed the "burst" of fluorescence caused by heating to the melting of the lipides.) Underlying this "two- phase" hypothesis of Seybold and Egle was the conviction that all chloro- phyll-protein complexes are nonfluorescent. However, while this seems to be true enough of pure chlorophyll-protein precipitates (cf. page 775), it does not apply to complexes which contain both proteins and lipides (e. g., to "coacervates" of the type described by Hubert and Frey-Wyssling ; cf. chapter 23, page 777). Seybold and Egle's argument is therefore not convincing. The effects of heating and drying on chlorophyll fluorescence in vivo can be explained in a much simpler way than suggested by Seybold and Egle : by assuming that the pigments normally contained in a weakly fluorescent protein-chlorophyll-lipide complex lose the protection against self-quenching (and therefore become nonfluorescent), when the lipides melt in the heat and form a separate phase, but diffuse into this new phase if the pigment-protein link is broken by denaturation {e. g., by somewhat more prolonged heating) and thus again become fluorescent. The dis- placement of the fluorescence bands of chlorophyll in living cells (by 5-15 mjLi toward longer waves from their position in organic solvents) agrees with the assumption that fluorescence is emitted by the same chlorophyll molecules responsible for the (similarly displaced) absorption bands. VARIATIONS OF CHLOROPHYLL FLUORESCENCE 819 Seybold and Egle, on the other hand, had to attribute the fluorescence bands to the fraction of chlorophyll dissolved in a lipide, and the absorp- tion bands to the bulk of chlorophyll present in a protein-bound colloidal state. Therefore their theory was predicated on the contention that the fluorescence band of chlorophyll is shifted in lipides toward the longer waves much more strongly than the corresponding absorption band. This hypothesis was characterized as implausible on page 746. To sum up, there seems to be no reason to attribute the fluorescence of living plants to a small fraction of chlorophyll molecules, present in a strongly fluorescent solution, rather than to the whole mass of the pigment forming a weakly fluorescent complex with proteins and lipides (including the carotenoids) . The close relationship between fluorescence intensity and rate of photosynthesis, which will be discussed in the next section, also indicates that fluorescence is a property not of a small fraction but of the bulk of chlorophyll in the cell. 4. Variations of Chlorophyll Fluorescence Related to Photosynthesis In the preceding section, we discussed chlorophyll fluorescence in vivo in relation to what may be called the gross state of the green pigment in the living cell — its high concentration and its simultaneous association, in the "chloroplastin," \\ith proteins and lipides. The effects of drying, heating, boiling or immersion in ether, described in that section, can be assumed to be indicative of a partial or complete disintegration of the chloroplastin. In the present section, we will deal with reversible changes in the yield of fluorescence that are more or less closely associated with photosensitizing activity and can be assumed to occur without essential changes in the com- position and structure of chloroplastin. Kautsky discovered in 1931 that rapid changes in the intensity of fluorescence of leaves occur during the first seconds and minutes of illum- ination after a period of darkness, and bear definite relation to the pre- viously known changes of the rate of photosynthesis during this "induc- tion period." Subsequent investigations by Kautsky and co-workers (1931-1948), Franck and co-workers (1934-1949), McAhster and Myers (1940) and of the Dutch group of investigators (Ornstein, Wassink, Katz, Dorrestein et al. (1937-1949) have revealed many striking examples of close interrelation between the intensity of fluorescence and the momentary rate of photosynthesis. This relationship can be observed not only dur- ing the induction period, but also in the steady state. Factors such as light intensity, temperature, concentration of reactants that take part in photosynthesis, presence of oxygen and various poisons and narcotics are found to affect significantly the yields of both fluorescence and photo- S3m thesis. The close interrelation of the fluorescence of chlorophyll and its photo- 820 FLUORESCENCE OF PIGMENTS IN VIVO CHAP. 24 sensitizing activity (revealed through parallel measurements of the yields of photosynthesis and fluorescence) has made fluorescence measurements an important tool in the kinetic analysis of photosynthesis. We will therefore restrict ourselves in the present chapter to some general considera- tions of this relationship, postponing more detailed description of experi- mental results and their interpretation to the several chapters in part IV dealing with the effects of light intensity, temperature, carl)on dioxide and other external factors, on the kinetics of photos^mthesis. Fluorescence is one of the several ways in which excited chlorophjdl molecules can dispose of their energy. Others include energy dissipation (conversion into vibrational energy and ultimately into heat), and photo- chemical reactions (either involving the chlorophyll molecule itself, or sen- sitized by it). The intrinsic capacity of the excited chlorophyll molecule to fluoresce (the monomolecular fluorescence constant kf, or its reciprocal, the ''natural life time" of the excited state, t/) can be considered as con- stant as long as the absorption spectrum of the chlorophyll molecule re- mains essentially unchanged. The intrinsic capacity for energy dissipa- tion (the monomolecular dissipation constant kt) and the rate of energy loss through chemical reactions (rate constants ki, k^. ■ ■, which can be mono- molecular or bimolecular) are, on the other hand, subject to changes de- pending on the association of the chlorophyll molecule with other molecules before excitation, and on its encounters with other molecules during excita- tion (as discussed in the sections of chapter 23 dealing with the quenching and self-quenching of chlorophyll fluorescence in vitro). The variations of chlorophyll fluorescence in vivo associated with variations in the rate of photosynthesis must therefore be attributed to changes in the composition or structure of the chlorophyll-bearing molecular complex (and consequent alterations in the values of monomolecular constants of dissipation and chemical quenching), and to changes in the probability of the chlorophyll complex encountering, during the excitation time, molecules capable of serving as effective "physical" or "chemical" quenchers (and consequent alterations in the values of bimolecular constants of quenching) . The several more or less detailed interpretations of fluorescence changes in photosynthesizing plants, which have been suggested, all are based on these general ideas but differ in emphasis laid on one or the other specific mechanism of quenching. Some (Kautsky; Wassink and Katz) attribute the main function to "chemical quenching" by the reactants taking part in photosynthesis, and consider each increase in fluorescence as evidence of a decrease in the efficiency of the sensitized photochemical process (and consequent decline of chemical quenching), and each decrease in fluores- cence as evidence of increased efficiency of utilization of excitation energy for the sensitized photochemical reactions (and consequent increase of chemical quenching). Others (Franck) see the most important cause of CHLOROPHYLL FLUORESCENCE AND PHOTOSYNTHESIS 821 changes in fluorescence intensity in the formation of chlorophyll complexes with surface-active substances ("narcotics") which slow down energy dis- sipation, and at the same time inhibit photochemical sensitization by pre- venting photosensitive substrates from reaching the chlorophyll. This amounts to a weakening of both processes (sensitization and dissipation) which compete with fluorescence ; whereas in theories of the first-mentioned type, onl}' one competing process (sensitization) is affected, while the other two (dissipation and fluorescence), profit equally by the elimination of a common competitor. We will now describe more specifically the several suggested mechanisms of interrelation of fluorescence and photosynthesis beginning with the pic- ture used in Volume 1 (chapter 19). In scheme 19. Ill (Vol. I, page 547) we attempted to represent the prob- able relationship between sensitization and fluorescence of chlorophyll in vivo. This scheme was formulated primarily for the interpretation of sensitized photoxidation, but essentially similar conditions may be as- sumed to prevail in photosjaithesis as well. The primary process was as- sumed in chapter 19 to be a "tautomerization" of the complex X. Chi. HZ (formed bj- association of chlorophyll with oxidant X and reductant HZ) : (24.3) X-Chl-HZ ^X-Chl*-HZ > HXChl-Z In photosynthesis, this primary process must be followed by secondary, catalytic reactions, in which HX is oxidized back to X (directh' or indi- rectly) by the carbon dioxide-acceptor compound, {CO2} *, and Z is reduced back to HZ (directly or indirectly) either by water (in ordinary photosyn- thesis of green plants) or by reductants such as H2, H2S or thiosulfate (in the photosynthesis of purple bacteria). In this picture, variations in fluorescence can be related to those in photosynthesis in both the above-mentioned ways — by means of primary changes in the probabihty of sensitized chemical reaction, and by means of primary change in the rate of dissipation of energy in the chlorophyll-bearing complex. If the dissipation rate is constant, fluorescence is an indicator of the efficiency with which the excitation en- ergy of chlorophyll is used for the primary photochemical process (equation 24.3) : Whenever the latter process is retarded for one reason or another, the sum of the probabilities of the two competing processes — fluorescence and internal dissipation of the excitation energy — increases correspond- ingly. Since fluorescence and internal energy dissipation are two alterna- tive monomolecular processes, the yield of both will be changed in the same proportion. Fluorescence thus becomes an index of the yield of the pri- mary photochemical process, even though the absolute yield of fluorescence * ir X = { CO2 , the secondary reaction is the replacement of a reduced by a fresh molecule i CO2 ! • 822 FLUORESCENCE OF PIGMENTS IN VIVO CHAP. 24 ( < 1%) is much too small to make it a significant competitor of this process. For example, if the yield of the primary photochemical process drops from 80 to 40%, and the sum of the yields of dissipation and fluorescence there- fore increases from 20 to 60%, the yield of each of these two processes will increase by a factor of 3. If the yield of fluorescence was