THE IDENTIFICATION OF NAVAL FUELS AND NATURAL FLUOROPHORS IN SEA WATER BY "FLUORESCENCE SPECTROMETRY" Hugh Wyman Howard L \n ATE SG Monterey, California The Identification of Naval Fuels and Natural Fluorophors in Sea Water by "Fluorescence Spectrometry" by Hugh Wyman Howard, Jr. The s is Advis or E. D. Traganza March 1972 Approved &oa pubtic fttteMse.; da>.t'Ubution uiiLunLtcd. The Identification of Naval Fuels and Natural Fluorophors in Sea Water by "Fluorescence Spectrometry" by Hugh Wyman Howard, Jr. Lieutenant, United States Navy B.S., United States Naval Academy, 1965 Submitted in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE IN OCEANOGRAPHY from the NAVAL POSTGRADUATE SCHOOL March 1972 ABSTRACT Fluorescence and Excitation spectra of Navy Standard Fuel Oil (NSFO), Navy Distillate Fuel (ND) , Diesel Fuel and Navy Aircraft fuels (JP-4 and JP-5) were obtained with the Turner 210 Absolute Spec trof luor ome ter . Excitation spectra (peaks at 320 ni, 240 nm, 250 nm, 270 nm, 250 nm respectively) and Fluorescence spectra (peaks at 350, 410 nm, 330 nm, 340 nm, 240, 325 nm , 240 nm respectively) are characteristic and may allow selective identification of these fuels. Quantitative determinations by fluorescence analysis of ND fuel oil extracted from sea water samples, with cyclohexane, showed saturation values of approximately 11 ppm. An all glass, in-situ vacuum filtering water samp- ler was designed and built for collection of filtered (.45y glass) noncontaminated sea water samples for the fluores- cence analysis determination of the natural background fluorescence of the Monterey Bay region. Fluorescence spectra of sea water from Monterey Bay, obtained on board R/V ACANIA, and samples from the Arctic Ocean, showed broad banded emission in the region of 450 nm. TABLE OF CONTENTS I. INTRODUCTION -.--- 10 i II. BACKGROUND 13 A. HISTORY OF FLUORESCENCE ANALYSIS OF FUEL OIL 13 III. METHODS 15 A. SEA WATER SAMPLING AND HANDLING PROCEDURES 15 B. FILTRATION PROCEDURES 25 C. CONTAMINATION PREVENTION 27 D. FLUORESCENCE ANALYSIS OF MONTEREY BAY AND ARCTIC OCEAN SEA WATER SAMPLES 29 E. SELECTIVITY EXPERIMENT FOR FLUORESCENCE TRACING OF COMMON NAVAL FUELS 29 F. QUANTITATIVE DETERMINATION OF OIL CONTENT IN SEA WATER 30 G. RELIABILITY 31 1. Sensitivity 31 2. Precision 32 IV. RESULTS 35 A. BACKGROUND FLUORESCENCE OF MONTEREY BAY AND THE ARCTIC OCEAN 35 B. CHARACTERIZATION OF NAVAL FUEL OIL SAMPLES-- 4 3. C. QUANTITATIVE DETERMINATION OF NAVY DISTILLATE FUEL OIL AT KNOWN PERCENT SATURATION IN SEA WATER 53 D. PRECISION 60 1. Instrumental 60 2. Extraction 63 E. SENSITIVITY . 63 V. DISCUSSION OF RESULTS 67 VI. CONCLUSIONS 78 BIBLIOGRAPHY 80 INITIAL DISTRIBUTION LIST 82 DD FORM 1473 84 LIST OF TABLES TABLE 1. WATER SAMPLE STATIONS IN MONTEREY BAY- 17 TABLE 2. CHARACTERISTIC PEAK EXCITATION FOR FIVE CYCLOHEXANE FLUORESCENCE AND NAVAL FUELS IN 52 TABLE 3. STANDARD DEVIATION RESULTS FOR PRECISION EQUIPMENT USING REPLICATE FLUORESCENCE SPECTRA FOR 10 PPM (104ng/ml) QUININE BIFULFATE IN 0 . IN SULFURIC ACID ■ 61 TABLE 4. FLUORESCENCE OF 50% and 5% SATURATED ND SEA WATER BEFORE AND AFTER EXTRACTION WITH CYCLOHEXANE 62 TABLE 5. FLUORESCENCE EXCITATION AND EMISSION DATA FOR PETROLEUM AND PETROLEUM PRODUCTS 72 LIST OF FIGURES 1. Map of Sampling Stations 16 2. Surface 125 ml ground glass sampling device 18 3. Components of surface sea water sampling device for fluorescence analysis 19 4. All-glass, in-situ vacuum filtering water sampler 20 5. All glass, in-situ vacuum filtering sea water sampler 21 6. Valve and actuator system for an all glass, in-situ vacuum filtering sea water sampler 22 7. Components of all glass, in-situ vacuum filtering sea water sampler 23 8. Assembled sampling flask and filtration system for all glass, in-situ vacuum filtering sea-water sampler 24 9. "Bell Jar" filtration unit with 125 ml, brown glass, sampling bottle positioned inside the vacuum dome 26 10. 4 ml, fluorescence analysis, cuvette washing apparatus 28 11. Turner 210 Absolute Spec trof luorome ter 33 12. Front control panel, recorder group and open sample compartment of Turner 210 Absolute Spec trof luorome ter 34 13. Fluorescence spectra of surface sample of nonfiltered sea water from station one 36 14. Fluorescence spectra of a nonfiltered bottom sample of sea water from station one 37 15. Fluorescence spectra of a filtered bottom sample of sea water from station one 38 16. Fluorescence spectra of a bottom sample of nonfiltered sea water from station two 39 17. Fluorescence spectra of a filtered surface sample of sea water from Arctic Ocean 40 18. Fluorescence spectra of a filtered surface sample of sea water from station two 41 19. Excitation spectra of filtered surface sample of sea water from station two 42 20. Fluorescence spectrum of Navy Distillate Fuel in cyclohexane 44 21. Excitation spectra of Navy Distillate Fuel in cyclohexane with analyzing monochromator set at 340 nm 45 22. Fluorescence spectra for Navy Standard Fuel Oil in cyclohexane 46 23. Excitation spectrum of Navy Standard Fuel Oil in cyclohexane with analyzing mono- chromator set at 450 nm 47 24. Excitation spectra of JP-4 , JP-5 and Diesel Fuel in cyclohexane with analyzing mono- chromator set at 310 nm 49 25. Fluorescence spectra of JP-4, JP-5 and Diesel Fuel in cyclohexane 49 26. Fluorescence spectra of JP-4, JP-5 and Diesel Fuel in cyclohexane 50 27. Fluorescence spectra of JP-4, JP-5 and Diesel Fuel in cyclohexane 51 28. Fluorescence spectra of Navy Distillate Fuel in sea water before extraction with cyclohexane 54 29. Fluorescence spectra of Navy Distillate fuel in sea water after extraction with cyclohexane 55 30. Fluorescence intensity of ND in sea water vs concentration 56 31. Fluorescence of ND in cyclohexane extract 57 32. Standardization fluorescence spectra for ND in cyclohexane 58 33. Standard curve for known concentrations of ND in Spec troquality cyclohexane 59 34. Fluorescence spectra of ND in cyclohexane 64 35. Fluorescence spectra of ND in cyclohexane 65 36. Fluorescence spectra of ND in sea water 66 37. Fluorescence spectra of Kuwait crude oil in sea water, extract from phy toplankton presumably contaminated with Kuwait crude oil 70 38. Fluorescence and excitation spectra of a class of fuel oils 75 39. Fluorescence spectra of various fuel oils 75 ACKNOWLEDGEMENTS The author expresses his sincere appreciation to: Dr. Eugene D. Traganza for his advice, inspiration, and attention to all details of thesis production. Dr. Charles F. Rowell for his advice, assistance and critical review of this work. Dr. Carl Heller of the China Lake Research Laboratory who made this research possible through the lending of the Turner 210. The Research Department at the Naval Postgraduate School and associated facilities, especially Bob Scheile, for his expert technical assistance during all phases of the design of water sampling devices and other miscellaneous equipment . The Material Science and Chemistry Department at the Naval Postgraduate School and associated facilities for use of equipment for this work, especially Roy Edwards for his pleasant disposition and helpfulness, also Ken Graham for his helpful suggestions. Mr. Leon E. Hiam from the San Francisco branch of Per- kins-Elmer Corporation, for his willingness to teach me how to use the MPF-2A Fluorescence Spectrophotometer and make it available for analysis of sea water samples from the Arctic Ocean. I. INTRODUCTION Fluorescence in sea water is not new [Kalle 1937, Shteg- man 1966, Traganza 1967, Monizikoff 1969]; however, the practical application of its measurement in sea water is only beginning to be appreciated or utilized. The increasing frequency of oil spills both reported and not, is reaching alarming proportions. Fluorescence offers both active and passive identification of oil pollution and may be an answer to problems of monitoring, detecting and evaluating the source of common contaminating fuel oils. Oil discharged into the ocean will form slicks which eventually spread and cover large areas. The danger to marine life is mostly due to oil dissolved in the bulk of the sea water [Parker 1967], One of the first attempts to investigate the natural fluorescence of marine waters was Kalle's study of Gelb- stoff as early as 1937 [see Kalle 1966], Interest in the study of fluorescence in sea water continued with Shuleikins (1953) investigation of sea water coloration [Shtegman 1966]. Traganza (1967) studied fluorescence and excitation spectra of dissolved organic matter in sea water. Traganza's spectral analysis of sea water shows a con- trasting picture of broad band background spectra against transient characteristic spectra of natural biological origins. The suggestion is that the natural fluorescence 10 of the ocean provides a broad banded low level background against which the fluorescence from recent biological or other events such as trace oil contamination from a distant spill can be detected. The determination of organics, which are responsible for the natural background fluorescence, will involve the difficult task of isolating and concentrating minute quanti- ties of organic matter in sea water. Monizikoff (1969) used fluorescence and certain concentrating and extraction pro- cesses to selectively identify several fluorescent sub- stances present in sea water. In the past, a major limiting factor to the study of dissolved organics in the ocean, was the lack of suitable instrumentation, capable of detecting the trace quantities of organics in "salt water." "On the average, every gram of organic matter is dwarfed by 36,000 grams of salts in 900,000 grams of water" [Diehl 1971]. Recent development of sophisticated and sensitive fluorescence spectrophotometers with better resolution may substantially improve the capability for analyzing trace amounts of organics. In some cases, direct analysis may be possible, but in any case, where a fluorescence technique is applicable, one will not be faced with the problem of pro- cessing inordinately large volume of sea water samples. This in itself should improve the probability of getting real time synoptic data on organics in the ocean. 11 An excellent discussion of various instruments may be found in Udenfriend (1962). Turner (1964) adds to this with a good description of one of the most sophisticated instruments available today. Aminco Bowman Corporation is advertising an instrument with a sensitivity of parts per trillion. This paper is directed toward the development of rou- tine methods for fuel oil contamination analysis, yielding type, source and concentration of the pollutant. The results of this work may provide a foundation from which the Navy can develop a system capable of reliable, simple, economical and continuous, monitoring and preven- tion of potentially critical pollution levels in sea water. Detection and positive identification of the fuel oil pol- lution sources is the final objective. 12 II. BACKGROUND A. HISTORY OF FLUORESCENCE ANALYSIS OF FUEL OIL The cumulative literature indicates that "passive fluor- escent tracing" is a promising though little utilized method for the detection of many organic compounds in the ocean including petroleum products. "Active methods" in which water or oil is spiked with a known fluorescent material also offers a variety of possible applications. [See. Rieker 1962], Bentz and Strobel (1933) examined fluorescence of oil resulting from excitation with a mercury vapor lamp. It was noted that refined oils fluoresced in the blue region while crude oil fluoresced in the brown and yellow region. Mel- hase (1936) found that no two samples of California crude oils fluoresced with the same shape spectra or intensity unless they were from the same parent oil sand. Work done by Shuldiner (1951) reported that oil spills in harbors could be matched with standards by comparing the shape and color of paper chroma tograms under ultra violet excitation. His method was successful in collecting identification in- formation sufficient for prosecution and conviction of oil polluters on the East Coast of the United States of America. [See Parker 1962]. Kats and Sederov (1954) have analyzed fluorescence spectra of various crude oils and their frac- tions. They proposed that fluorescence spectra be utilized as a method for identifying crude oil and various petroleum 13 products. Mihul, Ruscior and Pop (1956) studied the fluor- escence spectra of various oils and suggested that fluores- cence be used to identify petroleum products from various sources . C. A. Parker, et al . (1967) utilized fluorescence anal- ysis to determine the concentration of oil in sea water and plankton, resulting from the Torrey Canyon oil spill. Thruston and Knight (1971) were able to comparatively iden- tify several fuel oils by two methods of characterizing each fuel oil; first the ratio of shoulder to peak inten- sity for each undiluted fuel and second, the shoulder to peak ratio for various dilutions of the same fuel (Fig. 37). A. W. Hornig (1971) illustrated a number of materials in the ocean which may be detected and identified by fluores- cence. One of these is a number-two fuel oil. He suggests that other crude oil samples will have signature variations which may allow identification of the oil. The fact that various petroleum products can be made to fluoresce charac- teristically is significant and common to all of these efforts . 14 III. METHODS A. SEA WATER SAMPLING AND HANDLING PROCEDURES Marine water samples were taken on 2 November 1971, aboard R/V ACANIA at stations established by Lewis (1970). The stations are described in Table 1 and Fig. 1. Station selection was based upon the probability that, where water samples would most exhibit fluorescence. The samples were taken at three depths in order to establish a general spatial pattern for the distribution of organic matter. Samples were also collected in the Arctic at Point Barrow Research Station in 50 feet of water. Surface samples were obtained by lowering a 125 ml brown bottle sampler (Figs. 2 and 3) which was opened from the surface by a lanyard attached to the ground glass stopper . Midwater samples were collected utilizing a specially designed, all glass, in-situ vacuum filtering device (Fig. 4 through 8). The sampler is supported in a stainless steel housing on which the trip release is attached. The glass portion of the sampler consists of the components shown in Fig . 7 . A precombusted 0.45y filter was placed in the sampler and the rig was lowered to the desired depth. At this point a weight was released on the hydro wire to actuate the 15 PACIFIC OCEAN POINT LOflOS ) 1 Statute miles iT-_B_B Fig. 1. Chart of sampling stations utilized for collection of sea water samples for Fluorescence Analysis. 16 TABLE 1. WATER SAMPLE STATIONS IN MONTEREY BAY Station Number Position 126 56.0'W 36° 3 3.0'N 121 55.9'W 36° 38.3'N 121" 54.8'W 36° 37.6'N 121 47.5'W 36° 48.3'N 121 46.7'W 36° 44.8'N 121 50.6'W 36° 37.7'N 121 53.3'W 36° 36.3'W lX££ Sewage outfall Carmel Bay, Calif. Sample Depth Surface 50 ft 100 ft Surface 30 ft Sewage outfall Pacific Grove, Calif. 60 ft Kelp beds Surface 50 ft 100 ft Surface Industrial discharge 30 ft 60 ft Salinas River Surface 30 ft 60 ft Surface Rip current 50 ft Del Monte Beach, CA 100 ft Monterey Harbor Surface 20 ft 17 HH Fig. 2. Surface 125 ml ground glass stoppered sampling device for collection of sea water samples for Fluores- cence Analysis. 18 19 Fig. 4. An all glass, in-situ vacuum filtering (0.45m glass filter) water sampler for fluorescence analysis of sea water . 20 Fig. 5. An all glass, in-situ vacuum filtering sea water sampler . 21 Fig. 6. Valve and actuator system for an all glass, in-situ vacuum filtering sea water sampler. 22 23 24 spring tension, rotating valve system, which breaks a glass tip. This allowed the vacuum to withdraw a sample into the two liter sample collection flask. A battery of pre-evacua- ted flasks was taken on the cruise to collect individual samples. Nonfiltered mid-water samples were collected utilizing evacuated glass tubes of two liter capacity. They were lowered to the desired depth on an aluminum support frame and actuated by releasing a weight attached to the hydro wire to break the tips extending from the evacuated glass tube samplers. ' Bottom water samples were collected by pouring off the water recovered along with bottom sediments in a "Shipek" sampler . The surface and bottom samplers were separated into filtered and unfiltered samples. B. FILTRATION PROCEDURES Samples were contained in specially cleaned brown glass, 125 ml, sample bottles. The filtration unit was constructed of a 300 ml Millipore funnel, with a Teflon seal, mounted on a fritted glass base attached to a vacuum cell (Fig. 9). The sample bottle was placed inside this cell to receive the filtered sample. This one step filtration minimizes possible contamination. The samples were stored at 10 C in the specially cleaned and combusted, 125 ml sample bottles. Prior to each filtration run, a glass filter (0.45u) was inserted after being combusted at 400 C to oxidize organic 25 Fig. 9. "Bell Jar" filtration unit with 125 ml, brown glass, sampling bottle positioned inside tin- vacuum dome . 26 matter in the filter. The filter support was thoroughly cleaned by repeated flushing with distilled water. An improvement was made to this filtration system by replacing the fritted glass filter support with a ground glass base and stainless support screen with Teflon gasket. This unit seals better and is easier to clean and allowed further minimizing of contamination problems (Fig. 7). C. CONTAMINATION PREVENTION Five general rules were adopted to prevent contamina- tion. No polyethylene, rubber, cork or organic compound was used as a bottle, stopper, sampler or allowed to come in contact with the sampler [Lewis 1971] . All glassware was thoroughly scrubbed with biodegradable soap, prefer- ably Calgon due to its low residual fluorescence blank after rinsing [Traganza 1969], and followed by a distilled water rinse. The glass was then cleaned in chromic acid and sulfuric acid followed by rinsing with acetone. This was followed by several rinses in organic free distilled water. This water was treated with 5 grams per liter of potassium persulfate and allowed to stand overnight. All exposed sample storage vessels were covered with aluminum foil after cleaning. Promptly after sample collection and filtration, the sea water was refrigerated to cut down on degradation. Filter contamination was minimized by precom- bustion at 450°C for 4 to 6 hours. Filters were stored in aluminum foil [Diehl 1971]. Instrument cuvettes were 27 Fig. 10. 4 ml, fluorescence analysis, cuvette washing apparatus. 28 washed with soap and rinsed with water followed by acetone and organic free distilled water. The glass cuvette clean- ing apparatus (Fig. 10) was used. D. FLUORESCENCE ANALYSIS OF MONTEREY BAY AND ARCTIC OCEAN SEA WATER SAMPLES Each 125 ml sea water sample was analyzed in turn on the Turner 210 Absolute Spec trof luorometer (Fig. 11 & 12) utiliz- ing the uncorrected fluorescence mode of operation. The samples were excited at selected wavelengths from 200 nm to 500 nm. Particular care was taken to ensure that the sample cuvette was specially cleaned prior to each run. After ex- ploratory excitation and emission spectra were completed, the instrumental settings were selected to provide the best possible characteristic trace. E. SELECTIVITY EXPERIMENT FOR FLUORESCENT TRACING OF COMMON NAVAL FUELS Five Naval fuels were investigated; Navy Distillate (ND) , Navy Standard Fuel Oil (NSFO) , Diesel Fuel (DF), and two Aircract fuels (JP-4 and JP-5). Each fuel was diluted successively in Spectro quality cyclohexane to lof lof 5x10? 2x10? 10? 10? 10? loj 10_1 and 10~2ng/ml(i.e. 1000, 100, 50, 20, 10, 1.0, 0.1, 0.01 ppm and 0.1, 0.01 ppb) . The samples were placed in specially cleaned glassware to avoid contamination (see above). Each fuel type was excited from 200 nm to 500 nm in 50 nm steps. After pre- 29 liminary searching, characteristic traces of each fuel were obtained (see Fig. 19 through 26). F. QUANTITATIVE DETERMINATION OF OIL CONTENT IN A KNOWN PERCENT SATURATED, NAVY DISTILLATE IN SEA WATER, SAMPLE A saturated solution of Navy Distillate in sea water was diluted with sea water from 100% to 10% saturation in steps of ten. The saturated solution was obtained by vigorous shaking of Navy Distillate and sea water and allowing to stand for several weeks. Each sample was excited at 310 nm to determine its ini- tial fluorescence spectrum. Six milliliters of "Spectro- quality" cyclohexane was added to each sample in a separa- tory funnel. After shaking vigorously 3 times, the samples were left standing for 4 hours. The sea water was then drawn off and examined on the fluorometer in the uncorrected mode. The extract was also examined. Curves were obtained of fluorescence intensity of sea water before and after extracting against percent saturation. The samples and extract were then returned to the sep- aratory funnels and an additional milliliter of cyclohexane was added and shaken. After standing overnight, the fluor- escence intensity was again recorded. It was suspected after observing the results for the above that pre-cen tr if ugat ion of the sea water would elimi- nate any inconsistencies in fluorescence intensity associa- ted with oil globules in the water samples. 30 Ten milliliters of 10, 30, 50, 70, 90 and 100% saturated ND in sea water was placed on the centrifuge for 5 minutes at medium speed. Five milliliters was withdrawn at the bottom of the aliquotes with a pipette and examined on the fluorometer. Excitation was at 310 nm. The samples were placed in separatory funnels and 10 ml cyclohexane was added, After standing for a period of 4 hours, the fluorescence spectra were recorded separately for sea water samples and extract . The data collected was plotted on semi log paper to de- termine the relationship between the concentration of oil and the fluorescence intensity. Known amounts of Navy Distillate were dissolved in cyclohexane to establish standard curves of fluorescence versus concentration (Fig. 32). The samples were excited at 290 nm . The extract from above was compared to these curves to determine the amount of oil in the extract and by extrapolation to determine the concentration of oil in each oil-sea water sample. G. RELIABILITY 1 . Sensitivity In order to determine the lowest concentration of fuel oil detectable with the Turner instrument, the follow- ing procedure was conducted. A serial dilution was made of Navy distillate fuel in "Spec tr oquali ty " cyclohexane and in sea-water. Each 31 serial dilution was excited at the characteristic wavelength for Navy Distillate (290 nm) and the fluorescence spectrum was recorded. The object was to determine the lowest concentration detectable with the instrument at the proper settings for maximum sensitivity. 2 . Precis ion Instrumental precision was established by replicate excitation of a known standard over a 4 hour, 45 minute 4 period. The standard was quinine bisulfate in 10 ng/ml (10 ppm) solution, with H SO, . The fluorescence spectra were recorded for the characteristic peak excitation wavelengths of 250 nm and 350 nm. The fluorescence maximum was observed at 460 nm. Precision was also established for the extraction experiment by fluorescence analysis of 3 replicate samples of Navy Distillate fuel at 50% and 5% saturation in sea water, before and after extraction. Comparison of the fluorescence maximum intensities yielded a standard deviation for the process. 32 u (U 4-1 Q) e o u o rH 4-1 O M ■U O 1) & 00 0) 4-1 3 rH o w U 0) H •H 33 34 IV. RESULTS A. BACKGROUND FLUORESCENCE OF MONTEREY BAY AND THE ARCTIC OCEAN Sea water samples fluoresced in response to ultraviolet excitation. Common mineral salts and end products from the decomposition of organic matter were assumed to have no in- fluence on this mission [Hornig 1971, Shtegman 1966]. Fluorescence spectra of marine waters were broad essen- tially featureless bands in the region of 450 nm (Fig. 14 through 17). The maximum excitations for this fluorescence ranged from 310 nm to 350 nm. An attempt was made to separate the fluorescent organic structures present in the sampled regions, by utilizing high instrumental sensitivity and excitation energy with various optimum slit widths, to provide good resolution. The re- sults were disappointing and agreed with work done by Sinel- 'nikov and Ryzhekov [Shtegman 1966] who concluded that fluorescence spectra did not selectively identify the com- position of natural waters; however, they are capable of detecting recent events in the water column. Traganza (1969) showed recent biological events are detectable and that individual background fluorescent compounds may be select- ively identified when fluorescence techniques are properly combined with suitable concentration and isolation procedures 35 H H CO !3 W H M w > M H M H H H M H ►J W Pi 500 550 600 650 Wavelength nm Fig. 18. Fluorescence spectra of a filtered surface sample of sea water from station two (Fig. 1). Exci- tations are: A, 300 nm; B, 250 nm; C, 400 nm . 41 H M w S3 w H 53 > M H W OH 250 350 300 Wavelength nm Fig. 19. Excitation spectra of a filtered surface sample of sea water from station two (Fig. 1). Ex- citation peaks are at: A, 230 nm; B, 280 nm ; C, 320 nm . 400 42 The broad banded fluorescence centered at 450 nm may indicate the presence of a group of compounds in the samples . The only exception observed to the general fluorescence emission observed at 450 nm was the fluorescence spectra of a filtered surface sample (Fig. 11) taken at station two (Fig. 1). The characteristic emission was at 520 nm when the sample was excited at 250 nm, 300 nm and 400 nm . The excitation spectrum showed maximum peaks at 230 nm, 280 nm and 320 nm (Fig. 19) . B. CHARACTERIZATION OF NAVAL FUEL OIL SAMPLES Excitation and fluorescence spectra for the 5 Naval fuels in cyclohexane are shown in Fig. 20 and Fig. 27. The maximum excitation and fluorescence peaks for each fuel are summarized in Table 2. The traces are characteristic to a degree which may per- mit direct identification, by passive fluorescence techniques against the natural fluorescence background. Navy Distillate fuel has characteristic excitation peaks at 290 nm and 240 nm when the analyzing monochromator is set at 350 nm (Fig. 21). The fluorescence spectrum shows a maximum at 330 nm with a slight shoulder at 315 nm and a more prominent one at 340 nm (Fig. 20). Navy Standard Fuel Oil has characteristic excitation peaks at 315 nm, 350 nm, and 400 nm when the analyzing mono- chromator is set at 450 nm (Fig. 33). When excited at 320 nm, 43 80 . 70 . 60 H M W S3 W H !S M W > M H <« W Pi 300 350 400 450 Wavelength nm Fig. 20. Fluorescence spectrum of Navy Distillate Fuel (ND) in cy clohexane (1 O^ng/ml) with excitation of 290 nm. 44 80 70 60 50 40 30 20 10 .250 300 350 400 Wavelength nm Fig. 21. Excitation spectra of Navy Distillate Fuel in cyclohexane (10 ng/ml) with analyzing monochroma t or set at 34 0 nm . 45 0) C CO X -W o 0) CO .a O 0) in cyclo , exciation o fn u o w o Fuel Oil (N at 34 0 nm; o c lO T3 O V U -H cfl -U T3 cfl S C -P C CO tH •U O JZ co X ■u % a > * 0) cfl M o o rH is t ID • n > K 0 CO O C 3 pectra f at 320 o en C vn O r> luorescence A, excitati o tn o ,-^ co rH • B > O c •H O W S3 o 6 e iH o o S 0) O C T3 4-1 U ccj jd n) 4-) •a 4J M C a) C ca to a) 4-1 H C/l J-l o 0) O o > >> 4J o cfl > ca 13 trum of Na monochrom o a m o 0) C n o o M 0) a P. t4 to N >> a iH o CO ■H C 4-1 ca cfl 4J x; •H 4-1 O •H X !5 w •~^ rH • 0 n ^. CN M c • M H < ►J W DIESEL 250 300 Wavelength nm 350 400 Fig. 24. Excitation spectra of JP-4, JP-5 and Diesel Fuel in cyclohexane (106ng/ml) with analyzing mono- chromator set at 310 nm. 48 300 350 400 450 Wavelength nm Fig. 25. Fluorescence spectra of JP-4, JP-5 and Diesel Fuel in cyclohexane (10 ng/ml) with excitation at 250 nm . 49 H M W W H !a M w > M H M H 300 350 400 450 Wavelength nm Fig. 27. Fluorescence spectra of JP-4 , JP-5 and Diesel Fuel in cyclohexane (106ng/ml) with excitation at 280 nm . 51 TABLE 2 Characteristic Peak Fluorescence and Excitation for Five Naval Fuels in Cyclohexane EXCITATION SPECTRA Navy Peak and/Shoulder Fuel Fl uqrescence Ob 450 served Excitation NSFO 320, 340, 400 Distillate 350 - 235, 290 Diesel 350/310 300/305 JP-4 310 275 JP-5 310 275 EMISSION SPECTRA Peak and/Shoulder Sample Excitation Emission NSFO 320 350, 410 340 410 400 450 Distillate 310 340/350 Diesel 250 340 340 340 JP-4 250 290, 325 290, 325 340 ii-- l, :•'.(> 290 290 290, 340 250 270 280 250 270 280 250 270 280 52 peak fluorescence was observed at 350 nm and 410 nm. When excited at 340 nm, the peak emission occurred at 410 nm with a slight shoulder at 385 nm. When excited at 400 nm, NSFO fluorescence was recorded at 450 nm (Fig. 22). The excitation spectra for JP-4 and JP-5 shows peaks at 275 nm; however, their fluorescence spectra vary with exci- tation-wavelength. Selectivity for these two fuels has been achieved by observing these variations for several excitation wavelengths (Fig. 25 through 27). Figure 25 shows fluorescence of JP-4 and JP-5 when excited at 250 nm. JP-5 has a prominent peak at 290 nm and a slight shoulder at 320 nm. JP-4 on the other hand, has two prominent peaks, one at 290 nm and one at 325 nm. When excited at 270 nm , intensity shifts were observed; however, JP-4 maintained two prominent peaks to JP-5's one (Fig. 26). When excited at 280 nm , a complete shift in peak intensities was observed with JP-5 now showing two prominent peaks at 290 nm and 340 nm against a single distinct peak for JP-4 at 340 nm with shoulders at 290 nm and 310 nm (Fig. 27). Diesel fuel exhibits fluorescence at 340 nm when excited at all three of the wavelengths above (Fig. 25 through 27). C. QUANTITATIVE DETERMINATION OF NAVY DISTILLATE FUEL OIL AT KNOWN PERCENT SATURATION IN SEA WATER Figure 27 and 29 are fluorescence spectra of six samples of sea water contaminated with Navy Distillate Fuel. Figure 28 is before extraction and Figure 29 is after extraction with spectro quality cyclohexane. 53 350 450 500 400 Wavelength nm Fig. 28. Fluorescence spectra of Navy Distillate Fuel in sea water before extraction with cyclohexane. Ex- citation is at 310 nm. A, 100% Sat.; B, 90% Sat.; C, 70% Sat.; D, 50% Sat.; E, 30% Sat.; F, 10% Sat. 54 H H w H > M H < w 350 450 500 400 Wavelength nm Fig. 29. Fluorescence spectra of Navy Distillate fuel in sea water after extraction with cyclohexane. Ex- citation is at 310 nm. A, 100% Saturation; B, 90% Sat. C, 70% Sat.; D, 50% Sat.; E, 30% Sat.; F, 10% Sat. 55 100 % sal. 53 O M H . > cd 3 «4-l •"■* O iH a o 0) ~~- o C t)0 *~ o c •HCO 4J O M * W « " 2 C W CD rH tj > ° a u = T3 >, o t-l 4-J * cd -H T3 M H M H < W 50 40 30 20 10 350 400 450 500 550 Wavelength nm Fig. 36. Fluorescence spectra of Navy Distillate Fuel in sea water at excitation 310 nm. A, 5% Saturated; B, 0.5% Sat.; C, 0.05% Sat.; D, 0.005% Sat. 66 V. DISCUSSION OF RESULTS The background natural fluorescence established for the Monterey Bay region suggests that the natural fluorescence found by others [Traganza 1967; Kalle 1966; Shtegman 1966] is "typical" of sea water in general, i.e. low concentration in the wave band 350 nm to 500 nm. The fluorescence of Arc- tic water (emission 400 nm to 450 nm) gives further evidence in support of this observation. It appears that on the average, the natural fluorescence background should not in- terfere with the fluorescence analysis of sea water for fuel oil contamination. This conclusion is supported by the relatively low instrumental sensitivity settings required for fuel oil determination in this study. Interference may occur if concentrations of some fuel oils are as low as 1.0 ng/ml or during biological events which can cause a trans- ient "blackout" with blank values so high that fuels fluores- cing in the same wavelength region will be masked. Naval fuels were selectively identified at very low con- centrations using passive fluorescence analysis of oil "fin- gerprints." Each Navy fuel was found to have characteristic fluorescence spectra (Fig. 20 through 27). For example, JP-4, JP-5 and Diesel were identifiable on the fluorometer but not separable by gas chromatography in another study in progress at this laboratory. 67 Table 5 is a summary of fluorescence characteristics re- ported for petroleum and various petroleum products. Riecker 1962, Shtegman 1966 and Smith 1968 have determined that var- ious fuel oils fluoresce in the region from 440 to 630 nm. The only naval fuel which fluoresced in this region was NSFO which had a characteristic peak at 450 nm when excited at 400 nm (Fig. 2 2) . Parker and Barnes (1960) pointed out that fluorescence of hydrocarbons present in a sample, represents the cumula- tive fluorescence of a complex mixture and the lighter the fraction, the shorter the wavelength of fluorescence. This may explain the decreasing wavelengths observed for fluores- cence of Diesel Fuel, JP-4 and JP-5 (Fig. 25 through 27). Diesel Fuel was found to be well separated from JP-4 and JP-5; however, JP-4 and JP-5 fluoresce in the same band of wavelengths (Fig. 25 through 27). Selectivity was achieved between JP-4 and JP-5 by obtaining several fluores- cence spectra at a variety of exciting wavelengths and com- paring them. It is clear that the emission spectra are not similar for both JP-4 and JP-5 whereas their excitation spectra are (Fig. 25). Thus, we may assume the presence of more than one molecular species or reactive groups, which are fluorescent. With this variable excitation technique which can produce a separation in characteristic spectra, fluorescence spectra and excitation spectra either singly or together will provide selective identification of the Navy fuels examined. 68 Figure 37 shows the characteristic curve obtained for Kuwait Crude Oil and phy toplankton after the Torrey Canyon oil spill [Parker, et al . 1970], The emission peak at 360 nm and shoulder at 400 nm are in the general region of NSFO emissions (Fig. 22). The fluorescence of the plankton sug- gests an uptake of crude oil by phy toplankton ; however, the fluorescence- was not unambiguously identified as due to crude oil. Good instrumental sensitivity was achieved allowing de- tection down to 0.1 ng/ml of sea water (0.1 ppb). Parker and Barnes (1960) reported sensitivities from 0.2 ng/ml to 0.3 ng/ml (0.2 to 0.3 ppb) and suggest that the fluorescence blank from the cyclohexane solvent used may have been the limiting factor. For this reason, spec troscopically pure cyclohexane was used in this study with a slight increase in sensitivity to 0.1 ppb. There was no observed Raman emission interference from the cyclohexane solvent in this study. This may be explained by the relatively high concentration level of the fluorescent material after extracting, which allowed fluorescence analy- sis to be done at instrumental sensitivities too low to be affected by Raman emissions. Due to optimizing instrumental slit width and other sensitivity settings, to achieve high resolution and charac- teristic spectra for each fuel, this author feels that the overall method sensitivity, while superior to most tech- niques is "instrument limited" [Parker 1968] only to the 69 RELATIVE INTENSITY 70 extent of instrument sophistication. Increased sensitivities are possible with newly developed instruments. The instrumental precision and extraction precision had average standard deviations of + 1.54 and + 2.18 intensity units respectively (Table 3 and 4) . These may even be im- proved upon once standard methods are established and stream- lined for the fluorescence analysis of fuel oil on a routine basis. In order to ensure accuracy , standardization and cali- bration procedures are necessary on each run. Ideally, fluorescent standards for each fuel or wavelength of interest should be selected and the instrument should be calibrated for this wavelength region which is a function of the sus- pected fuel type in an unknown sample. In this study, it was assumed that evaporation loss of low boiling compounds or fractions from fuel oils, produced no detectable change in fluorescence characteristics or intensity in agreement with a study by Thruston and Knight (1971). Parker (1970) also supports this idea since accord- ing to his paper, the fluorescent components are almost com- pletely nonvolatile and, therefore, total fluorescence is constant regardless of evaporation, assuming no concentra- tion effect. On the other hand, this hypothesis perhaps should be tested in future research. Thruston and Knight (1970) and Parker (1970) also con- sidered the effects of solar decomposition, leaching, oxi- dation and microbial action on fuel oil spills. There appears to be degradation in fluorescence characteristics 71 TABLE 5 Fluorescence Excitation and Emission Data for Petroleum and Petroleum Products Investigator Sample Excitation Fluorescence Shtegman Petroleum 365 nm 450-500 nm (1966) products . Parker and Compressor 248 nm 333-384 nm Barnes (1960) lubricant. 286 nm 333-384 nm Crude oil. 250 nm 350-400 nm Auto engine lubricant. 250 nm 360 nm Auto Diesel fuel. 250 nm 300-350 nm Auto Petrol. 250 nm 300-320 nm Hornig (1971) Fuel Oil #2 322 nm 351 nm Parker, et al . Kuwait Crude 250 nm 360 nm (1967) Smith (1968 Crude Oil 400 nm 450 nm Riecker (1962) Crude Oil unknown 440-630 nm Thruston and Crude and 340 nm 386 nm Knight (1970) semi refined fuel oil . Howard (1972) NSFO 320 nm 350, 410 nm 34 0 nm 410 nm 400 nm 450 nm JP-4 250 nm 290, 325 nm 270 nm 290, 325 nm 280 nm 340 nm JP-5 250 nm 290 nm 270 nm 290 nm 280 nm 290, 340 nm Diesel 250 nm 340 nm 270 nm 340 nm 2 80 nm 34 0 nm Distillate 290 nm 330nm,315nm, 340 nm 240 nm 330 nm 72 and intensity with time of exposure, so for this reason, it would be advisable to sample a new oil spill quickly to sup- port positive identification of the pollutant and pollutor. It is possible that water sampled at some depth may escape these effects should it reach a protective depth before any significant degradation occurs. Further support for the application of natural fluores- cence identification of fuels is the work done by Thruston and Knight (1971). Different samples of the same general fuel type were identified on the basis of two parameters for each fuel; the ratio of shoulder to peak fluorescence in- tensity for each undiluted fuel and the shoulder to peak ratio for various dilutions of the same fuel (Fig. 38 and 39). Variations within the same fuel were detected. Sam- ples from an actual oil spill were examined with this method and four other standard identification methods. All were in agreement and the fuel was unambiguously identified. The application of this technique to the problem of detection of fuels as diverse as the naval fuels examined in this study should be a source of additional corroborating data. The ability to identify the source and type of oil in an oil spill on a beach, in a river, a harbor or in the open ocean is fundamental to the effective enforcement of water pollution control procedures and statutes already in effect. There will be occasions when the only requirement is to monitor a process, such as bilge water discharge or ship ballast discharges or for early warning of fuel oil contam- 73 ination in a regional environment. At times, only the pres- ence of fuel and its possible source will be required. The results of this study can be applied to all of these situations. The characteristic excitation and fluorescence wavelengths determined for the five naval fuels will allow proper selection of filters for operation of low cost, sim- ple filter fluorometers such as the Turner 111. Quantitative information is required for base data studies on fuel contamination levels for environmental quali- ty control or for cryptic long term effects on biological processes which may be toxic or produce subtle unwanted alteration of natural phenomena. Quantities as low as 0.1 ng/ml (0.1 ppb) may be detected and the potential exists for sensitivities as high as parts per trillion. The extraction techniques developed in this study may yield a workable method for the determination of absolute concentration levels. More work in this area will be re- quired to establish standard curves from which a concentra- tion value can be obtained. Work by Drushel and Sommers (1966) indicated that phos- phorimetry can significantly enhance pho toluminiscent iden- tification of various compounds in petroleum. This author agrees in principle; however, fluorescence is a direct, inexpensive technique which can in itself be applied to at sea fuel analysis problems. The results of this study show that passive detection and identification of naval fuels is possible by fluorescence 74 320 400 4gOn« 220 ■ioo 3fl0«« Wavelength nm Fluorescence and excitation spectra fuel oils which have a peak emission 386 nm and a shoulder at 440 nm when excited at 340 nm [from Thruston and Knight 1970], Fig. 38. class of of at >-< H M CO w H M w > M H < W $iO 4iO 4^0 nm. tA 340 4oO 4<"nm *40 400 4?&rtUHtEa.C s\jlf fuel GULF f JCL FIELD 8 Wavelength nm Fig. 39. Fluorescence spectra of various fuel oil showing peak intensities and shoulder to peak ratios for each fuel and dilutions of each fuel [from Thrus- ton and Knight 1970] . 75 spectrometry. The application of this method calls for some practical considerations. If bilge water or ballast water is to be monitored for contamination levels, it is quite within the possibilities of passive fluorescence techniques, without interfering processes such as contamination or en- vironmental weathering. The problem of oil spill analysis becomes more difficult with the later added effects. Efforts are now underway to legally trace fuel contamination to the vessel responsible. The method for accomplishing this must be cheap and unambi- guous. Since at least four mutually supporting pieces of evidence are the minimum requirements for successful prose- cution and conviction of offenders, the application of pas- sive fluorescence tracing may not be enough alone. For example, it is not known how much environmental weathering makes an oil spill unidentifiable. This means that supporting evidence must accompany excitation and fluorescence spectra of the oil spill. A promising alter- nate or supplement may well be "active tagging" or the addi- tion of some known and readily identifiable fluorophor to the oil. This may be thought of as an identification "license plate." "The added material must satisfy several criteria: it must be soluble or dispersible in oil and insoluble and non- dispersible in water; it must be easily detectible in ex- treme dilution; it must be chemically and physically stable in both spilled and unspilled oil; it must not interfere 76 with the commercial uses of petroleum; and it must not be too expensive." [Horowitz, et al. 1971]. In "passive tagging," the oil slick must provide all of the information for its own identification by fluorescence; in other words, it must show its "fingerprint." The addition of well-characterized fluorophors to fuel oil is possible today. ■• Small measured quantities can be added to the oil as it is loaded on a ship with a metering device. One or more distinct fluorescent "license plates" could be added for positive identification of the fuel oil and the carrier by fluorescence analysis. By this method, the required, separate, supporting pieces of evidence could be quickly and reliably collected for use in any oil spill situation . 77 VI. CONCLUSIONS The value of this technique for fuel oil contamination control and detection presents some interesting possibili- ties. Fluorescence techniques for active ("license plates") or passive ("fingerprints") could be used either separately or in support of one another in such areas as the quality control of overboard discharge of ballasting water from naval ships. An early warning system of high oil concentra- tion in this water could actuate a quick acting automatic control valve to prevent discharge of potentially dangerous fuel oil contamination. This could be accomplished with realtively inexpensive filter instruments set at characteris- tic wavelengths for the fuel oil it would monitor. "Passive fluorescence" techniques would be sufficient for this appli- cation. For oil spill detection, simple passive techniques are all that would be required; however, for the positive iden- tification of the pollutant and pollutor, a combination of "active" and "passive tagging" would be appropriate. "Pas- sive tagging" would be utilized to establish background fluorescence for all regions of interest. "Active and pas- sive tagging" would be utilized for the gathering of suffi- cient identification information to make a strong legal case for the positive identification of the responsible parties . 78 The Navy would profit from establishing base data for the present levels of fuel oil contamination from which to measure the effectiveness of newly instituted preventive methods and cleanup programs. 79 BIBLIOGRAPHY 1. Diehl, G. F., Jr., Systems Analysis of Methods for Me as u ling Trace Dissolved Organic Matter in Sea Water, Masters Thesis, Naval Postgraduate School, Monterey, 1971. 2. Guilbault, G. G., Fluorescence - Theory, Instrumenta- tion and Practice, Decker, 1967. 3. Hercules, D. M., Fluorescence and Phosphorescence An a 1 y s i s , Wiley, 1966. 4. Hornig, A. W., Remote Sensing of Marine and Fisheries Resources by Fluorescence Methods, paper presented at Symposium on Remote Sensing in Marine Biology and Fishery Resources, Texas A & M University, College Station Texas, January 1971. 5. Horowitz, J. and others, "Identification of Oil Spills: Comparison of Several Methods," American Petroleum Institute Publication No. 4040, p. 283-296, 1969. 6. Kullbom, S. D., Smith, H. F., Flandreau, P. S., Fluorescence Spectroscopy in the Study and Control of Water Polution," Paper No. 288 presented at the Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, 1970. 7. Lewis, L. W., Jr., Chemical Characterization of Dis- solved Humic Substances in Monterey Bay by Nylon Adsorption and Paper Chromatography , M asters Thesis, Naval Postgraduate School, Monterey, 1972. 8. Parker, C. A. and Barnes, W. J., Spec trofluorime try of Lubricating Oils: Determination of Oil Mist in Air, Analy s t , v. 85, p. 3-8, January 1960. 9. Parker, C. A., Pho tolumines cence of Solutions, Elsevier, 1968. 10. Parker, C. A., Oil Contents of the Sea Following the Torrey Canyon Incident, Admiralty Materials Labora- tory Report A/70(m), November 1967. 11. Pringsheim, P., Luminescence of Liquids and Solids and its Practical Applications, In ters cience , 1946. 80 12. Riecker, R. E., "Hydrocarbon Fluorescence and Migration of Petroleum," Bulletin of the American Association of Petroleum Geologists, v. 46, No. 1, p. 60-75, January 1962. 13. Shtegman, B. K. , Production and Circulation of Organic Matter in Inland Waters, Translated from Russian, Israel Program for Scientific Translations, Jerusalem, 1966. 14. Shuldiner, J. A., "Identification of Petroleum Products by Chromatographic Fluorescence Methods," Analytical Chemis try , v. 23, No. 11, November 1951. 15. Smith, H. F., "Luminescence Spectroscopy - A Versatile Analytical Tool," Research/Development, p. 20-27, July 1968. 16. Thruston, A. D., Jr. and Knight, R. W., "Characteriza- tion of Crude and Residual - Type Oils by Fluores- cence Spectroscopy," Environmental Science and Technology , v. 5, No. 1, January 1971. 17. Traganza, E. D., "Fluorescence Excitation and Emission Spectra of Dissolved Organic Matter in Sea Water," Bulletin of Marine Science, v. 19, No. 4, p. 897- 904, December 1969. 18. Turner, G. K., "An Absolute Spec tr f luorome ter , " Science, v. 146, No. 3641, p. 183-189, October 1964. 19. Udenfriend, S,, Fluorescence Assay in Biology and Medicine , Academic Press, 1962. 81 INITIAL DISTRIBUTION LIST No. Copies 1. Defense Documentation Center 2 Cameron Station Alexandria, Virginia 22314 2. Library, Code 0212 2 Naval Postgraduate School Monterey, California 93940 3. Assistant Professor E. D. Traganza, Code 58Tg 3 Department of Oceanography Naval Postgraduate School Monterey, California 93940 4. Lieutenant Hugh W. Howard, Jr. 4 91 Phillips Brooks Road Westwood, Massachusetts 02090 5. Oceanographer of the Navy 1 The Madison Building 732 N. Washington Street Alexandria, Virginia 22217 6. Commander Navy Ship Systems Command 1 Code 901 Department of the Navy Washington, D. C. 20305 7. Department of Oceanography 3 Naval Postgraduate School Monterey, California 93940 8. Assistant Professor C. F. Rowell, Code 5413 1 Department of Chemistry Naval Postgraduate School Monterey, California 93940 9. Dr. Carl A. Heller, Code 6059 1 Naval Weapons Center China Lake, California 93555 10. Mr. Jack Marsh 1 G. K. Turner Associates Pulgas Avenue East Palo Alto, California 82 11. Mr. Leon E. Riam 851 Hinckley Road Burlingame, California 94010 12. Mr. Earl K. Hinson 1451 Main Street El Centro, California 92243 13. Dr. Ned A. Ostenso, Code 480D Office of Naval Research Arlington, Virginia 22217 83 Security Classification DOCUMENT CONTROL DATA -R&D (Security classification of title, body of abstract and indexing annotation must be entered when the overall report is classified) DONATING ACTIVITY (Corporate author) Naval Postgraduate School Monterey, California 93940 2«. REPORT SECURI TY CLASSIFICATION Unclassified 2b. GROUP EPORT TITLE The Identification of Naval Fuels and Natural Fluorophors in Sea Water by Fluorescence Spectrometry ESCRIPTIVE NOTES (Type ol report and. inclusive dates) Masters Thesis, March 1972 UTHORISI (First name, middle initial, last name) Hugh W. Howard, Jr. EPORT DATE March 1972 CONTRACT OR GRANT NO. PROJECT NO. 7«. TOTAL NO. OF PAGES 85 7b. NO. OF REFS 19 9a. ORIGINATOR'S REPORT NUMBER(S) 9b. OTHER REPORT NOISI (Any other numbers that may be assigned this report) DISTRIBUTION STATEMENT Approved for public release; distribution unlimited. SUPPLEMENTARY NOTES 12. SPONSORING MILITARY ACTIVITY Naval Postgraduate School Monterey, California 93940 ABSTRAC T Fluorescence and Excitation spectra of Navy Standard Fuel Oil (NSFO) , Navy Distillate Fuel (ND) , Diesel Fuel and Navy Aircraft fuels (JP-4 and JP-5) were obtained with the Turner 210 Absolute Spectrof luorometer . Excitation spectra (peaks at 320 nm, 240 nm, 250 nm, 270 nm, 250 nm respectively) and Fluores- cence spectra (peaks at 350, 410 nm , 330 nm , 340 nm , 240, 325 nm 240 nm respectively) are characteristic and may allow selective identification of these fuels. Quantitative determinations by fluorescence analysis of ND fuel oil extracted from sea water samples, with cyclohexane, showed saturation values of approxi- mately 11 ppm. An all glass, in-situ vacuum filtering water sampler was designed and built for collection of filtered (.45u glass) noncontaminated sea water samples for the Fluorescence analysis determination of the natural background fluorescence of fhe Monterey Bay region. Fluorescence spectra of sea water from Monterey Bay, obtained on board R/V ACANIA, and samples from the Arctic Ocean, showed broad banded emission in the region of 450 nm . FORM I47O N 0101 -807-681 I D (PAGE I ) 84 Security Classification 1-31408 Security Classification KEY WO ROS marine fluorescence naval fuels marine oil pollution dissolved organics fluorescence spectrometry sea water sampler marine fluorophors )D,F.r.91473 'BACK, /N 0101-807-6821 85 Security CM.issifiriilion 26 FES 73 I4MAY75 ■iiifS Howa rd Tne 'dentif.Va*.- nava' fueJs a„? '°n of by "Fluorescln a Water trometryM 6nce SPec- ?6 FIB 73 '4 WAV75 HIM' Thesis HS213 c.l 134278 Howa rd The identification of nava) fuels and n-tturol fluorophors in sea water by "Fluorescence Spec- trometry". Ul ' • OUM.EY KNOX UBftAAY