Skip to main content

Full text of "Analysis of returned comet nucleus samples : proceedings of a workshop held at Milpitas, California, January 16-18, 1989"

See other formats

NASA Conference Publication 10152 

Analysis of Returned 
Comet Nucleus Samples 

Compiled by Sherwood Chang, Ames Research Center, Moffett Field, California 

Proceedings of a workshop held at 

Milpitas, California 

January 16-18, 1989 


NASA Conference Publication 10152 

Analysis of Returned 
Comet Nucleus Samples 

Compiled by Sherwood Chang 

Proceedings of a workshop held at 

Milpitas, California 

January 16-18, 1989 

National Aeronautics and 
Space Administration 

Ames Research Center 

Moffett Field, California 94035-1000 


Go VMVU:t\^ UV^'oiv vcMoriA 







Nucleosynthesis and the Isotopic Composition of Stardust' 1 

A. G. M. Tielens 

Interstellar and Cometary Dust' 29 

J. S. Mathis 

Refractory Solids in Chondrites and Comets: How Similar?* 59 

J. A. Wood 

Disequilibrium Chemistry in the Solar Nebula and Early Solar System: 

Implications for the Chemistry of Comets* 73 

B. Fegley 

Sulfur Compounds in Comets 93 

5. J. Kim and M. F. A'Heam 

Spectrophotometric Observations of Comet P/Giacobini-Zinner 103 

/. Konno, S. Wyckoff and P. A. Wehinger 

Physical Processing of Cometary Nuclei* 1 19 

P. R. Weissman and S. A. Stem 

The Organic Matter of Comet P/Halley as Inferred by Joint Gas 

and Solid Phase Analysis* 167 

J. Kissel , F. R. Krueger and A. Korth 

On the Measurement of Cosmogenic Radionuclides in Cometary Materials 183 

G. F. Herzog. P. A. J. Englert, R. C. Reedy, K. Nishiizumi, C. P. Kohland 
and J. R. Arnold 

Morphology suid Compositional Differentiation of the Surface of Comets 199 

W. F. Huebner and D. C. Boice 

The In-Situ Particulate Size Distribution Measured for One Comet: P/Halley 205 

J. A. M. McDonnell, G. S. Pankiewicz, P- N. W. Birchley, S. F. Green and 

C. H. Perry 

Organic Chemistry in Interstellar Ices: Connection to the 

Comet Halley Results 217 

W. A. Schutte, V. K. Aganval, M. S. de Groot, J. M. Greenberg, 

P. McCain, J. P. Ferris and R. Briggs 

Evolution of Carbonaceous Chondrite Parent Bodies: Insights 

into Cometary Nuclei?* 225 

H. Y. McSween, Jr. 

Indicates invited papers 

Interplanetary Dust Particles Optical Properties: A Clue to 

Cometary Dust Structure? 239 

A. C. Levasseur-Regourd 

Cometary Evolution: Clues on Physical Properties from Chondritic 

Interplanetary Dust Particles 247 

F. J. M. Rietmeijer 

Laboratory Simulations: The Primordial Comet Mantle" 255 

R. E. Johnson 

Metamorphism of Cosmic Dust: Processing from Circumstellar 

Outflows to the Cometary Regolith* 277 

J. A. Nuth III 

Experimental Studies of Gas Trapping in Amorphous Ice and Thermal 

Modelling of Comets — Implications for Rosetta 293 

A. Bar-Nun, D. Prialnik, I. Kleinfeld and D. Laufer 

Modifications of Comet Materials by the Sublimation Process: Results 

from Simulation Experiments* 315 

E. Griin, J. Benkhojf, A. Bischoff, H. Diiren, H. Hellmann, P. Hesselbarth, 
P. Hsiung, H.U. Keller, J. Klinger, J. Knolker, H. Kochan, G. Neukum, 

A. Oehler, K. Roessler , T. Spohn, D. Stoffler and K. Theil 

The Nature of Comet Materials and Attachment to Them 333 

J. Stephens 

Mechanical and SEM Analysis of Artificial Comet Nucleus Samples 341 

K. Thiel, H. Kochan, K. Roessler, E. Griin, G. Schwehm, H. Hellmann, 
P. Hsiung and G. Kolzer 

Ion Bombardment Experiments Suggesting an Origin for Organic 

Particles in Pre-Cometary and Cometary Ices 353 

T. J. Wdowiak, E. L. Robinson, G. C. Flickinger and D. A. Boyd 

On the Isotope Analysis of Cometary Dust* 365 

F. Begemann 

Analysis of Organic Compounds in Retumed Comet Nucleus Samples 377 

J. R. Cronin 

Concepts for the Curation, Primary Examination, and Petrographic 

Analysis of Comet Nucleus Samples Retumed to Earth' 399 

D. Stoffler, H. Diiren and J. Knolker 

Candidate Sample Acquisition Systems for the Rosetta Mission 417 

P. G. Magnani, C. Gerli, G. Colombina and P. Vielmo 

Description and Analysis of Core Samples: The Lunar Experience* 433 

David S. McKay and Judith H. Allton 

Indicates invited papers 


The Origin, Composition and History of Comets from Spectroscopic 

Studies' 455 

L. J. Allamandola 

Laboratory Analyses of Micron-Sized Solid Grains: Experimental Techniques 

and Recent Results' 473 

L. Colangeli, and E. Bussoletti 

Handling and Analysis of Ices in Cryostats and Glove Boxes in View of 

Cometary Samples 491 

K. Roessler, P. Hsiung, M. Heyl, G. Neukum, A. Oehler and H. Kochan 

Electron Spin Resonance (ESR) Studies of Returned Comet Nucleus Samples 505 

F. D. Tsay, S. S. Kim. and R. H. Liang 

Indicates invited papers 


This conference publication contains many of the invited and contributed papers presented at the 
workshop on Analysis of Returned Comet Nucleus Samples. The workshop was held on January 16- 
18, 1989, in Milpitas, California, under the sponsorship of the Lunar and. Planetary Institute and 
NASA Ames Research Center. After the fly-by missions to comet Halley in 1986, the prospect of 
launching the joint ESA/NASA Rosetta/Comet-Nucleus-Sample-Retum (CNSR) mission in 2002 with 
a sample return from a comet in 2010 raised considerable scientific interest and provided the 
stimulus for the meeting. The papers assembled here provide a partial historical record of the 
scientific issues and measurement objectives articulated in 1989 for the analysis of returned comet 
nucleus samples. My plan for a timely publication of these papers after the workshop was frustrated 
by a lengthy illness, and by the time I returned to active work, the momentum had been lost. To 
colleagues whose papers did not get into print until now, I offer my deep apologies. 

During the intervening years, comet missions suffered setbacks and advances. In the early 1990's the 
Comet Rendezvous Asteroid Fly-by Mission (CRAF, work on which was formally begun in 
November, 1989) was canceled by NASA, and the Rosetta/CNSR mission underwent a substantial 
transformation, becoming a comet rendezvous mission like CRAF and including an instrumented 
soft lander, but no longer a sample return. Rosetta will be launched in 2003 with arrival at comet 
Wirtanen in 2011 and landing in 2012. In the mid-1990's interest in samples remmed from comets 
was stimulated by the success of the Stardust proposal in competing for a place in the Discovery 
Program. Stardust — a mission to fly through a cometary coma and capture dust particles intact in 
aerogel, and then return them to Earth — will be launched in 1999 to a fly-by of Comet Wild 2, with 
sample return anticipated in 2006. As these missions developed, astronomical detection of Kuiper 
Belt comets and the extraordinary apparitions of comets Hyakutake and Hale-Bopp continue to fuel 
comet science. In the same timeframe, the technical challenge of a CNSR mission was taken up by 
NASA in the form of the Deep Space 4/Champollion mission now taking shape as part of NASA's 
New Millennium Program, and including participation by the French space agency, CNES. DS- 
4/Champollion will be launched in 2003 to a rendezvous with periodic comet Tempel 1 in late 2005, 
with a landing in 2006, and cryogenic, sub-surface sample return to Earth in 2010. Another 
anticipated comet mission is Contour which was selected through the Discovery Program and which 
will fly by at least 3 short-period comets. After some substantial setbacks, the fortunes of cometary 
missions are once again on the upswing, though future implementation is never a sure thing. 

Certainly, scientific interest in comets will persist. Astronomical observations alone leave wide gaps in 
our knowledge of the nature and origin of cometary volatiles and dust. Continuing studies of 
interplanetary dust particles will reveal more about the connections between th«ir components and 
their possible origins from comets, asteroids or interstellar dust. Instrumental methods of analyses 
will continue to advance across a broad front toward applicability to smaller and smaller sample 
sizes. Synergies will arise in the advancement of analytical techniques as the prospects for return of 
samples from Mars in 2006 develop in parallel with those for comet nucleus sample return. Samples 
of comet dust will be returned from the Stardust mission in 2006, and samples of cometary ices and 
dust will be returned by DS-4/Champollion in 2010. Therefore, laboratory studies will be conducted 
on samples of known comets. It is only a matter of time. 

My belated thanks go to workshop organizers and participants and to authors who supplied 
manuscripts. I am grateful to Paul Weissman for providing an update on comet missions. 

Sherwood Chang 

NASA Ames Research Center 




Monday, January 16th. 

7:00-8:00 a.m. Registration 

8:00 a.m. Welcome and Introduction to Workshop 

Sherwood Chang, NASA, Ames Research Center 

8:10 a.m. Rosetta — Comet Nucleus Sample Return 

Mission: Status Report 
Geoffrey Briggs, NASA Headquarters 
Mr. Marcelo Coradini, European Space Agency 

Chairman: Sherwood Chang 

8:30 a.m.-12:00 noon 


Invited Speaker Presentations 

Nucleosynthesis and the Isotopic Composition of Stardust 
Alexander Tielens 

Interstellar and Cometary Dust 
John Mathis 

Refractory Solids in Chondrite and Comets: How Similar? 
John Wood 

10:30 a.m. BREAK 

Solar Nebular Condensates and the Composition of Comets 
Jonathan Lunine 

Dis-equilibrium Chemistry in the Solar Nebula and Early Solar System: Implications for the Chemistry of 


Bruce Fegley 

Poster Presentations 

Thermal and Chemical Processing of the Outermost Layer of Cometary Nuclei 
Campins H. Krider E.P. 

Sulfur Compounds in Comets 
Kim S. A'Heam M. 

Spectrophotometric Observations of Comet P/Giacobini-Zinner 
Konno I. Wyckoff S. Wehinger P.A. 

Computer Simulation of Dust Grain Evolution 
Liffinan K. 



Chairman: Thomas Ahrens 

1:30-5:30 p.m. 


Invited Speaker Presentations 

Comments on Comet Shapes and Aggregation Processes 
William Hartmann 

Physical Processing of Cometary Nuclei 
Paul R. Weissman S. Alan Stem 


Gas and Ice Composition of Comet P/Halley 
Peter Eberhardt 

3:30 p.m. BREAK 

Composition of Dust from Comet P/Halley: The Mineral Fraction 
Yves Langevin 

The Organic Matter of Comet P/Halley as Inferred by Joint Gas and Solid Phase Analysis 
J. Kissel F.R. Knieger A. Korth 

Poster Presentations 

On the Measurement of Cosmogenic Radionuclides in Cometary Materials 
Herzog G.F. Englert P.A.J. Reedy R.C. et al. 

Measurements of Long-Lived Cosmogenic Nuclides in Returned Comet Nucleus Samples 
Nishiizumi K. Kohl C.P. Arnold J.R. 

Morphology and Compositional Differentiation of the Surface of Comets 
Huebner W.F. Boice D.C. 

Role of Dust to Gas Production Rate Ratio in Cometary Physics 
Ibadov S. 

The In-Situ Particulate Size Distribution Measured for One Comet: P/Halley 
McDonnell J.A.M. Pankiewicz G.S. et al. 

Organic Chemistry in Interstellar Ices: Connection to the Comet Halley Results 
Schutte W.A. Agarwal V.K. et al. 


(cocktails and hors d'oeuvres buffet) 


7:30-9:30 p.m. 

Chairmen: Douglas Blanchard, NASA Johnson Space Center 

Dieter Stoffler, Inst, fiir Planetologie, Universitiit Munster 

Panel Members: Benton Clark, Martin Marietta 

Peter Eberhardt, Physikalisches Inst., University of Bern 

John Ot6, University of Houston 

Edward Whalley, National Research Council of Canada 


Tuesday, January 17th. 

Chairman: Yves Langevin 

8:30-12:15 p.m. 



Invited Speaker Presentations 

The Comet Rendezvous Asteroid Flyby Mission 

David Morrison Marcia Neugebauer Paul Weissman 

Electron Beam Analysis of Particulate Cometary Material 
John Bradley 

Evolution of Carbonaceous Chondrite Parent Bodies: Insights into Cometary Nuclei? 
Harry Y. McSween, Jr. 

10:20 a.m. BREAK 

Isotopic Microanalysis of Returned Comet Nuclei Samples 
Ernst Zinner 

The Carbon Chemistry of Meteorites: Relationships to Comets 
Sherwood Chang 

Poster Presentation 

Interplanetary Dust Particles Optical Properties: A Clue to Cometary Dust Structure? 
A.C. Levasseur-Regourd 

Identification of Solar Nebula Condensates in Interplanetary Dust Particles and Unequilibrated Ordinary 


Klock W. Thomas K.L. McKay D.S. 

Cometary Evolution: Clues on Physical Properties from Chondritic Interplanetary Dust Particles 
F.J.M. Rietmeijer 

Trajectory- Capture Cell Instrumentation for Measurement of Dust Particle Mass, Velocity and Trajectory, 
and Particle Capture 
Simpson J.A. Tuzzolino A.J. 

The Measurement of Trace Elements in Interplanetary Dust and Cometary Particles by Ultra-high 

Sensitivity INAA 

Zolensky M.E. Lindstrom D.J. Lindstrom R.M. Lindstrom M.M. 

Chairman: Ezio Bussoletti 

1:45-5:00 p.m. 


Invited Speaker Presentations 

Laboratory Simulations: The Primordial Comet Mantle 
R.E. Johnson 

Metamorphism of Cosmic Dust: Processing from Circumstellar Outflows to the Cometary Regolith 
Joseph A. Nuth III 

From Interstellar Dust to Comets 
J. Mayo Greenberg 

3:25 p.m. BREAK 


Experimental Studies of Gas Trapping in Amorphous Ice and Thermal Modelling of Comets — 
Implications for Rosetla 
Akiva Bar-Nun 

Modifications of Comet Materials by the Sublimation Process: Results from Simulation Experiments 
Eberhardt Grun and KOSI team 

4:45 p.m. 

The Return of Comet Samples and the Issues of Planetary Protection 
John Rummel, NASA Planetary Protection Officer 


Chairman: Larry Nyquist 

5:00-5:30 p.m. 

Poster Presentations-IIB 

Direct Determination of the Morphology, Structure and Composition of Cometary and Interstellar Ice 
Analogs in the Laboratory 
Blake D.F. Allamandola L.J. 

Thermal Histories of the Samples of Two KOSI Comet Nucleus Simulation Experiments 
Spohn T. Benkhoff J. Klinger J. Griin E. Kochan H. 

The Nature of Cometary Materiids and Attachment to Them 
Stephens J. 

Mechanical and SEM Analysis of Artificial Comet Nucleus Samples 
Thiel K. Kochan H. Roessler K. Grun E. Schwehm G. 
Hellmann H. Hsiung P. Kolzer G. 

Ion Bombardment Experiments Suggesting an Origin for Organic Particles in Pre-Cometary and 

Cometary Ices 

Wdowiak T.J. Robinson E.L. Flickinger G.C. Boyd D.A. 


(cocktails and hors d'oeuvres buffet) 


Professor Linus Pauling 

Inorganic Analogs to Biological Specificity 

Dr. Linus Pauling was awarded the Nobel Prize in Chemistry in 1954 and the Nobel Peace Prize in 1963. 
He has received 30 honorary degrees from United States and foreign universities. In addition, he has 
received the following awards and honors: the Langmuir Prize, American Chemical Society, 1931; the 
Nichols Medal, 1941; the Linus Pauling Medal, 1966; the Davy Medal, Royal Society, 1947; Medal for 
Merit, 1948; Pasteur Medal, Biochemistry Society, France, 1952; Addis Medal, National Nephrosis 
Federation, 1955; Phillips Memorial Award, American College of Physicians, 1956; Avogardro Medal, 
Italian Academy of Science, 1956; Fermat Medal, Sabatier Medal and International Grotius Medal, 
11957; Order of Merit, Republic of Italy; Medal, Academy Rumanian People's Republic, 1965; President 
of Honor, International Society of Research Nutrition and Vital Substances, 1965; Silver Model, Institute 
France, 1966; and, the Supreme Peace Sponsor, Worid Fellowship of Religions, 1966. His work includes: 
determination of structure of crystals and molecules; application of quantum mechanics to chemistry; 
rotation of molecules in crystals; sizes of ions; theory of stability of complex crystals; chemical bond; 
line spectra; immunochemistry; structure of proteins; molecular abnormality in relation to disease; sickle 
cell anemia; orthomolecular medicine, vitamin C and cancer; metals and alloys; and ferromagnetism. Dr. 
Pauling now leads the Linus Pauling Institute of Science and Medicine in Palo Alto, California. 

Wednesday, January 18 

Chairman: Kurt Marti 

8:30-12:00 noon 


Invited Speaker Presentations 

Rosetta Mission Description 
Gerhard Schwehm 

On The Isotope Analysis Of Cometary Dust 
F. Begemann 

Isotopic Compositions Of Hydrogen, Carbon, Nitrogen and Oxygen 
Robert Clayton 

10:20-10:40 BREAK 

Analysis of Organic Compounds in Returned Comet Nucleus Samples 
John R. Cronin 

Concepts for the Curation, Primary Examination cmd Petrographic Analysis of Comet Nucleus Samples 

Returned to Earth 

Dieter Stoffler H. Diiren J. Knolker 

Poster Presentations 

Trace Element Abundance Determination by Synchrotron X-ray Fluorescence (SXRF) on Returned 
Comet Nucleus Mineral Grains 
Flynn G.J. Sutton S.R. 

Prompt Gamma Activation Analysis (PGAA): Technique of Choice for Nondestructive Bulk Analysis of 
Returned Comet Samples? 
Lindstrom D.J. Lindstrom R.M. 

Candidate Sample Acquisition System for the Rosetta Mission 
Magnani E.G. Gerli C. Colombina G. Vielmo P. 

Nondestructive Trace Element Microanalysis of As-Received Cometary Nucleus Samples Using 
Synchrotron X-ray Fluorescence 
Sutton S.R. 

12:00-1:30 p.m. LUNCH 

Chairman: Dave Stevenson 

1:30-5:00 p.m. 


Invited Speaker Presentations 

Analytical Study of Comet Nucleus Samples 
Arden Aibee 

Description and Analysis of Core Samples: The Lunar Experience 
David McKay 


2:50 p.m. BREAK 

The Origin, Composition (ind History of Comets from Spectroscopic Studies 
Louis J. Allamandola 

Laboratory Analyses of Micron-sized Solid Grains: Experimental Techniques and Recent Results 
L. Colangeli Ezio Bussoletti A. Blanco A. Borghesi S. Fonti V. Orofmo G. Schwehm 

CLOSING REMARKS— Sherwood Chang 

Poster Presentations 

Microanalytical Characterization of Biogenic Components of Interplanetary Dust 
di Brozolo F.R. Meeker G.P. Fleming R.H. 

Handling and Analysis of Ices in Cryostats and Glove Boxes in View of Cometary Samples 
Roessler K. Hsiung P.. Hey M. Neukum G. Oehler A. Kochan H. 

Analysis of Particulates of Comet Nucleus Samples: Possible Use of Olivine as Indicator Phase 
Steele I.M. 

Electron Spin Resonance (ESR) Studies of Returned Comet Nucleus Samples 
Tsay F.D. 




Thomas Ahrens 

California Institute of Technology 

Lou Allamandola 

NASA Ames Research Center 

David Blake 

NASA Ames Research Center 

Donald Brownlee 
University of Washington 

Theodore E. Bunch 

NASA Ames Research Center 

Humberto Campins 
Planetary Science Institute 

Sherwood Chang, Convener 
NASA Ames Research Center 

Jeff Cuzzi 

NASA Ames Research Center 

Eberfiard Griin 
Max-Planck-Institut fur Kemphysik 

Martha Manner 

Jet Propulsion Laboratory 

Alan Harris (Ex Officio) 
Jet Propulsion Laboratory 

John Kerridge 

University of California, Los Angeles 

Yves Langevin 
Universite de Paris, Sud 

Larry Nyquist, Convener 
NASA Johnson Space Center 

Gerhard Schwehm 

European Space Agency, ESTEC 

Paul Weissman 

Jet Propulsion Laboratory 




A. G. M. Tielens 

Space Sciences Division 

Ames Research Center 

Nucleosynthesis and the Isotopic Composition of Stardust 

A.G.G.M. Tielens 
Space Sciences Division 
MS 245-3, Moffett Field 
CA 94035 


Various components have been isolated from carbonaceous meteorites with 
an isotopically anomalous elemental composition. Several of these are generally 
thought to represent Stardust containing a nucleosynthetic record of their 
birthsites. This paper discusses the expected isotopic composition of Stardust 
based upon astronomical observations and theoretical studies of their birthsites: 
red giants and supergiants, planetary nebulae, C-rich Wolf-Rayet stars, novae and 
supernovae. Analyzing the Stardust budget, it is concluded that about 15% of the 
elements will be locked up in Stardust components in the interstellar medium. This 
Stardust will be isotopically heterogenous on an individual grain basis by factors 
ranging from 2 to several orders of magnitude. Since comets may have preserved a 
relatively unprocessed record of the Stardust entering the solar nebula, isotopic 
studies of returned comet samples may provide valuable information on the 
nucleosynthetic processes taking place in the interiors of stars and the elemental 
evolution of the Milky Way. 


One of the most interesting developments within the field of interstellar 
dust in recent years is the realization that some interstellar and circumstellar 
grains have been incorporated into meteorites and interplanetary dust particles 
without totally losing their identity (see the reviews by Kerridge 1986 and by 
Anders et al. 1989). Evidence for this rest on the measurement of isotopic 
composition of such materials. Although the meteoritic composition is in gross 
sense remarkedly homogeneous, non-mass-dependent isotopic anomalies do exist 
in many elements. These include the noble gasses, the light elements (H, C, N, and 

O), and the heavy elements (e.g., Ca, Ti, Cr, Ni, Nd, Sm and others). Although some 
unusual process in the solar nebula might have produced non-mass-dependent 
isotopic fractionation in some elements, it is unlikely that it could account for all 
of them. Moreover, the measured isotopic anomalies are very characteristic for a 
presoiar origin of the material. In particular, some Xe and Kr isotopic anomalies 
associated with a carbon phase carry the signature of the nucleosynthetic s- 
process in red giants, suggesting the presence of largely unmodified carbon 
Stardust in meteorites (Kerridge and Chang 1985). Other components isolated from 
meteorites seem to carry the nucleosynthetic record of novae and supernovae 
(Anders et al. 1989). The best Stardust characterizations have been made for C- 
dust. To a large extent this reflects merely the importance of vaporization and 
recondensation processes in the solar nebula for the mineral component of 
meteorites leading to a large degree of dilution of Stardust oxides. Since the solar 
nebula was 0-rich, contamination of the C-stardust record by solar nebula 
condensates is probably much less severe. 

Comets are among the most pristine objects in the solar system and their 
study may reveal much about starformation and solar nebulae processes. One 
popular model of comets envisions them as dirty snowbals formed by the 
aglomoration of ices and interstellar dust in the outer reaches of the 
protoplanetary disk (Greenberg 1989). In that case, processes in the solar nebula 
as well as on the parent body are unimportant and a large fraction of refractory 
Stardust, including the oxides, may be readily identifiable. Thus even more than for 
meteorites and interplanetary dust particles, analysis of the isotopic composition 
of comet samples may reveal important information on the nuclear processes 
taking place in Stardust birthsites. Such knowledge is expected to revolutionize 
our understanding of the composition and evolution of stars and the Milky Way. 

This paper considers the isotopic composition of Stardust inferred from 
observational and theoretical studies of their birthsites. Section II discusses the 
galactic Stardust budget for the three main dust components identified in stellar 
ejecta, C-dust, silicates and SiC. The ecological impact of Stardust on the total 
interstellar dust budget is also briefly discussed. There are several processes that 
can lead to the incorporation of (isotopic) trace species in a condensing solid. 
These are reviewed in section III. Section IV summarizes the nuclear processes 
taking place in the interiors of stars and their resulting evolution. The emphasis is 
on known or suspected sources of Stardust: 0-rich and C-rich red giants, red 
supergiants, planetary nebulae, C-rich Wolf-Rayet stars, novae and supernovae. 
Astronomical observations of the isotopic composition of their ejecta is 
discussed. Section V briefly discusses the isotopic composition of various 
suspected carbon-stardust components isolated from carbonaceous meteorites, 
their likely birthsites and the processes by which these anomalies were included 

in the condensing grains. Finally, the implications for the Rosetta mission are 
briefly discussed in section VI. 

I I Stardust Characteristics 

Many stars go through a pronounced phase of mass loss at the end of their 
life. Upon expansion, the ejected material cools down and conditions often become 
favorable for dust nucleation and condensation. This socalled Stardust manifests 
itself then by extincting the stellar light and reradiating the absorbed energy at 
infrared wavelengths. Among the stellar objects known to be associated with 
newly formed dust are red giants and supergiants, novae, planetary nebulae, and WC 
Wolf Rayet stars, infrared observations show evidence for the presence of 
amorphous silicates, hydrogenated amorphous carbon (ie., soot), polycyclic 
aromatic hydrocarbon molecules (ie., PAHs), silicon carbide, and magnesium sulfide 
grains around these objects (cf., Tielens and Allamandola 1987a). Until recently 
there was no direct astronomical evidence for dust formation in supernova ejecta. 
However, various independent lines of evidence have revealed that dust has formed 
in the ejecta of SN 1987A in the Large Magelanic Cloud (Wooden 1989). Moreover, 
small diamonds have been isolated from carbonaceous meteorites with an isotopic 
composition suggesting an origin in SN ejecta (Anders et al. 1989). Thus, it is 
likely that all SNe are important sources of Stardust as well. The chemical make- 
up of these supernova condensates is, however, unknown. 

A) The Stardust Budget 

Table 1 summarizes the carbon, silicate, and SiC Stardust budget of the galaxy (see 
Gehrz 1989; Tielens 1989). Carbon Stardust predominantly, but not exclusively, 
forms in C-rich ejecta (ie., C/0 > 1). C-giants dominate the carbon Stardust budget, 
with a small contribution from WC stars (for detail see Tielens 1989). The novae 
contribution is somewhat uncertain since the composition of the ejecta is 
sensitive to the specific conditions and is observed to vary (Truran 1985; Wiescher 
et al. 1986). The value quoted in table 1 assumes a factor ten enrichment in 
elemental C compared to solar, which increases it slightly from the previous study 
(Tielens 1989). We have included the contribution from the central stars of 
planetary nebulae (PN), based on a typical C-mass loss rate of 10"^ Mq yr"! and a 

surface density of 40 kpc"2 PN in the galaxy (Pottasch 1984). Although there is no 
direct evidence for dust formation in these high velocity winds (=1000 km/s), the 
UV/visual line spectrum of some PN nuclei resemble those of late type WC stars 

which are known to condense C-dust. Note that PNe are surrounded by extensive 
dust shells resulting from mass loss in the preceding red giant phase, which is 
seperately listed and not included in the PN value given in table 1. Since the ashes 
from the He-burning zone are C-rich (Woosley and Weaver 1986a), type II 
supernovae might be an important source of C-dust (Clayton 1981). However, 
various observations of SN 1987a in the LMC suggest that turbulence is very 
important in a supernova explosion (Arnett et al. 1989). This will mix the He ashes 
with products of the more fiercely burned core and, for completely mixed ejecta, C 
is only a minor elemental component (e.g., C/O=0.1 for a 25Mq progenitor; Woosley 
and Weaver 1986a). Little if any C-dust formation is then expected. This also holds 
for type la supernovae, but their overall contribution to the C-budget of the galaxy 
is never expected to be large (Thielemann et al. 1986; Woosley and Weaver 1986a). 
The supernova values given in table 1 correspond to the maximum amount possible 
and the actual contribution might be much less. The H-rich conditions in C-giant 
ejecta will lead to the formation of highly hydrogenated amorphous carbon dust 
(i.e., soot). In contrast, C-dust formed from He-ashes (eg., WC stars, SNe) will not 
be hydrogenated and, based on extensive laboratory studies (Curl and Smailey 
1989), might have a disordered, fullerene structure (Tielens 1989) 

When the elemental C/0 ratio is less than unity, silicate or metal grains are 
expected to condense out (Salpeter 1977). M-giants and supergiants show 10 and 
20)im emission or absorption features generally ascribed to tlie SiO stretching and 
bending vibrations in amorphous silicates. They are important contributors to the 
silicate Stardust budget. The estimate for giants in table 1 is based upon solar Si 
abundances, an estimated H mass loss rate of 5x10"'* Mq kpc"2 yr"! for M-giants 
(Jura 1987), and assumes that all Si condense out as (Mg,Fe)Si04. Supergiants 
typically have a mass loss rate of 10*5 Mq yr^. With a galactic supergiant surface 
density of 5 kpc"2, this results in 2x10"^ Mg kpc'2 yr"" . Supernovae (type II as 
well as type la) are the dominant source of newly synthesized silicon and can thus 
be important silicate producers as well. The estimates quoted in table 1 are based 
on nucleosynthesis SN calculations (0.2 and 1 .2 Mq per type la and type II SN 
respectively; Thielemann et al. 1986; Woosley and Weaver 1986a) and a SN rate of 
10-5 kpc-2 yr"! for each (Tammann 1982). Since these numbers assume that all 
the Si condenses as silicates, they should be considered as upper limits. Some 
novae show evidence for silicate dust condensation, besides the more abundant C- 
dust (Gehrz 1988). The estimate in table 1 is based on a mass ejection of lO^^ Mq, 

a nova rate of 40 yr^, and a solar Si abundance (Truran 1985). Note that the 
composition of the condensing silicates is not well known. Observations suggest a 
highly disordered structure, probably the result of the low condensation 
temperature (Tielens and Allamandola 1987b). 

Table 1: The Stardust budget of the galaxy 

Source contribution (10-^ Mg kpc-2 yri)a 






- - 



- - 






planetary nebulae 


- - 

- - 

Red Supergiants 

- - 


WC stars 


- - 

type II supernovae 



type la supernovae 



notes: a) Uncertainties are typically a factor 3. b) Mass loss during the red giant phase Is not included (see 
text lor details, c) Stardust fonnation has been observed in supernovae 1987a, but the composition of the 
dust is not known, d) No astronomical evidence for SiC formation in novae (see §11). 

The 11.3)im feature due to SIC grains is observed in circumstellar shells 
around C-rich giants (and planetary nebulae) whenever the eiennental C/0 ratio is 
in the range 1 to 1.5 (Roche 1989). Assuming that all the Si condenses out, this 
corresponds to a production rate of 7x10''' Mq kpc-2 ypl. However, analysis of IR 
observations suggests that only 10% of the Si has condensed out in SiC in the 
"prototypical" C-giant IRC 10216 (Martin and Rogers 1987) and we have adopted 
this conservative value in table 1. Isotopic anomalous SiC grains isolated from 
carbonaceous meteorites probably result from nova ejection (Anders et al. 1989) 
but, since Si is not greatly enhanced in novae explosions, their contribution is 
small. Moreover, the astronomical evidence for SiC formation in nova is 
controversial. A lOjim feature, resembling those observed in 0-rich giants and 
attributed to silicates, has been observed in nova V1370 Aql. However, the absence 
of the corresponding 20(im silicate emission feature in this nova has led to the 
suggestion that the carrier is a diatomic grain material, presumably SiC, despite 
the poor spectral match (Gehrz 1988). But, in contrast to dust shells around 0-rich 

giants which have dust at a range of temperatures, nova shells contain dust at a 
single (high) temperature and the 20M,m silicate feature might be much weaker 
than normally expected (Roche, private comm.). Thus, the absence of this feature 
might not be decisive. Formation of SiC grains is not expected in SN ejecta, since 
either the C and Si zone are mutually exclusive or mixing produces 0-rich ejecta 
and probably leads to oxide dust. 

B) Stardust and Interstellar Dnst 

Studies of interstellar extinction and polarization have shown that about 
half of the availlable, condensible elements (i.e., heavier than He) has to be in the 
form of interstellar dust (Helens and Allamandola 1987b). The ubiquity of the 
10|im silicate feature, corresponding to =30% of the total dust mass (Cohen et al. 
1989), attests to the importance of silicates in the interstellar dust budget. In 
general, it is assumed that this refractory grain component is formed in stellar 
ejecta. However, recent laboratory studies have shown that under some conditions 
silicate formation may also occur in the interstellar medium itself (Nuth and 
Moore 1988). Because of abundance constraints, the remainder of the interstellar 
dust mass is probably in the form of a carbonaceous dust component. Both models 
based upon graphite (or perhaps amorphous carbon) formed in C-rich stellar ejecta 
(Mathis 1989) as well as on hydrocarbon grains formed by accretion and UV 
photolysis in the interstellar medium (Greenberg 1989) have been proposed. Either 
of these models can provide excellent fits to the observed interstellar extinction 
and polarization given an appropriate grain size distribution. In fact, the derived 
grain size distributions are exceedingly similar and little additional information 
can be gleaned from such observations. The total elemental Si injection rate is 
about 4x10-6 Mq kpc-2 yr"! and, thus, about 2% of all Si is ejected in the form of 
SiC, of which at most =0.1% originates in novae. Thus, SiC Stardust seems 
unimportant on a galactic scale and models based on it (Gilra 1972) have been 
largely abandoned. Observational ly, a conservative upper limit on the ratio of the 
11.3 )j,m SiC band to the 10)im silicate band of 0.1 in the ISM translates to at 
least 10 times more Si in silicates than in SiC. 

Studies of elemental depletion patterns in the interstellar medium also 
indicate that silicates and carbonaceous grains are important components of the 
interstellar dust (Jenkins 1989). The gas phase abundance of Si, as well as the 
metals, Fe, Ca, Al, and Ti, is observed to be highly depleted (by a factor 10 or 
more) compared to the cosmic abundance value. Presumably, the missing fraction 
is locked up in silicate (or metal) grains (Field 1974). Although quite uncertain, C 
seems to be quite depleted in the interstellar medium (by about 50%). However, 
elements such as O and N, which predominantly form volatile condensates (ie., H2O 

and NH3), are much less depleted (<20%). Such elemental abundance studies have 
also revealed that high velocity clouds have nearly solar abundances of elements 
such Fe and Si (Jenkins 1989). These clouds have probably recently been shocked - 
hence the high velocity - which has presumably lead to destruction of most of the 
dust mass. Based upon expected shock frequencies in the ISM, the dust destruction 
rate in the ecology budget of the galaxy is then estimated to be about 1.4x10"*^ Mq 
kpc'2 yr"!, corresponding to a Stardust lifetime of about 4x1 0^ yr, a number 
uncertain by perhaps a factor two (McKee 1989). A further 5x1 0"5 Mg kpc-2 yr"! of 
dust and its associated gas is lost due to starformation. 

These numbers can be directly compared to the Stardust injection rates in 
table 1. The maximum total dust injection rate, including SN, is 2x1 0'5 Mq kpc'2 
yr-l. For a total ISM gas mass of 5x1 0^ Mq (Scoville and Sanders 1987) and a dust 
abundance of 1% by mass, this corresponds to an injection timescale of about 
2.5x1 o9 yr. Therefore, in equilibrium only about 15% of the Stardust survives in an 
average patch of the ISM. If we exclude SN for which the dust fraction is not well 
known, the average fraction of the elements locked up in Stardust is only 3%. Thus, 
interstellar dust models based on Stardust alone, which require 70 and 90% of the 
C and Si in the form of Stardust in the interstellar medium (Mathis 1989), face 
serious problems. In fact, it is difficult to explain elemental depletions of more 
than 75% by Stardust formation alone, since that would correspond to a dust 
destruction rate of less than 7x10-6 Mq kpc-2 yr"! - a factor 20 lower than the 
theoretical expected value (McKee 1989). This problem is further compounded for 
the most heavily depleted elements, Fe (>90%) and Ca (>99%), which require dust 
destruction rates of less than 2x10^6 and 2x10*'' Mq kpc"2 yr"!. This is even more 
so since mass loss from O and B stars is known to contribute about 10% of the 
elements exclusively in gaseous form (Jura 1987). Obviously, accretion processes 
in the interstellar medium have to be very important in determining the gas phase 
abundance of these heavily depleted metals (Snow 1975). Given this conclusion, 
models based upon a dust component formed by accretion (and photolysis) 
processes in the interstellar medium, containing perhaps 25% of the condensible 
elements (Greenberg 1989), have to be taken very serious. Indeed, observations 
along three different lines of sight show the presence of hydrocarbon grains 
containing between 5 and 25% of the elemental C (Tielens and Allamandola 1987b; 
Cohen et al. 1989). Such hydrocarbon grains have not been observed in 
circumstellar dust sources and presumably reflect a dust component formed by 
local interstellar medium processes. 

ill Impurities in Stardust 

Several distinctly different processes can lead to the incorporation of 
impurities in Stardust. First, during condensation chemical substitution may take 
place. This is particularly true for isotopes such as ""Sc. Second, Stardust 
formation generally occurs under highly supercooled conditions and, even though 
thermochemistry might indicate the condensation of one (or more) well defined 
mineral (ie., Mg2Si04; Grossman and Larimer 1974), dust formation may be 
expected to be highly heterogeneous leading to a highly mixed elemental 
composition. Thus, metal cations such as Ca and Al are expected to be readily 
substituted for Mg in circumstellar silicates, despite the difference in 
coordination and binding. Likewise, metals such as Fe and Ti can replace Si in SiC 
and N can replace C in diamond grains. The presence of such impurities at the few 
percent level or higher is well known for terrestrial specimens, indeed, the colors 
produced by dissolved metal ions were already well appreciated by glass artisans 
in 14<f^ century Venice. 

Third, impurities can form chemical bonds with dangling bonds at internal or 
external grain surfaces during the grain condensation process. Peripheral groups 
such as OH are well known both for silicate as well as for carbonaceous grains. 
Highly disordered carbon soot with its many internal "surfaces" and edges contains 
in general a large complement of peripheral groups. The planar aromatic structure 
of soot also lends itself well to intercalation of impurities. The k electron system 
can accept as well as donate electrons with a typical binding energies of 0.5-1 eV 
and thus a variety of species can be incorporated, including metals such as Na and 
K. In contrast, noble gasses such as Ar and Xe have only a binding energy of 0.1- 
0.15 eV on a graphitic (or silicate) surface (Jaycock and Parfitt 1986), which 
corresponds to a residence time of =10"''"' sec at a dust condensation temperature 
of 800 K. Thus, noble gasses are not expected to be easily incorporated into 
condensing grains. Nevertheless, experimental studies of Xe adsorption on carbon 
black samples have revealed an unexpected, tightly bound Xe component which 
remains up to temperatures of 1300 K (Wacker et al. 1985). Note that most of the 
Xe was bound on external surfaces and readily evaporated upon warm up. At the low 
dust temperatures of the ISM (=15K) evaporation of such a noble gas component 
will be of little importance, but in view of the low binding energy and the 
importance of gas-grain interactions in the ISM, it is unlikely that they will be 
preserved into the solar nebula. The tightly bound component may represents a long 
random walk in the extensive pore network of carbon black with release possibly 
inhibited by bottleneck pores. Alternatively, it is known that He implanted into 
graphite remains trapped at much higher temperatures (=450K) than expected from 
simple physisorption (E=0.02eV; Moller et al. 1982). Presumably, this involves 



2000 - 


Fig. 1: The mean projected range for implantation of ions with an energy of 5 keV/amu, corresponding to a 
grain-gas drift velocity of 1000 km/s. 

vacancies, a well known effect for metals where binding energies of a few eV have 
been measured (Scherzer 1983). Although the literature is much sparser, heavier 
noble gasses seem also to be trapped at such vacancies. Thus, Stardust could 
contain traces of tightly bound noble gasses. 

Fourth, high velocity ion implantation can lead to a large concentration of 
impurities, but the high velocities required limit this process to supernova and 
nova ejecta and the fast winds of the central stars of planetary nebulae. Figure 1 
shows the mean projected range for ion-implantation in various solids (Burenkov 
et al. 1986). A gas-grain drift velocity of 1000 km/sec has been assumed, 
corresponding to =5 keV/amu. This picture is appropriate for (super)nova ejecta 
overtaking previously ejected material (Clayton 1981). In this case, the typical 
depth of implantation is about 1500 A fairly independent of grain material. Since 
this is a physical process, all elements can be retained and this process may be in 
particularly important for the noble gasses which are difficult to adsorb 
otherwise. The penetration depth is insensitive to the mass of the impacting ion 


(i.e., energy). For example, from Ar (40 amu) to Xe (130 amu) the penetration depth 
in graphite increases only from =1750 to 2200 A. Thus, only minor fractionation 
will occur. Larger differences in the retention of ions of different mass may occur 
on a timescale associated with the diffusion of the lightest ion, but at ISM dust 
temperatures (=15 K) this will be unimportant, except perhaps for H and He. When 
the dynamics of the outflow are dominated by radiation pressure on dust balanced 
by gas drag (i.e., red giants), the typical gas-grain drift velocities are quite small 
(<20 km/sec; Helens 1983). Since penetration depends strongly on the ion energy, 
much smaller implantation depths are obtained. For example, a Kr atom impacting 
at 20 km/sec (=0.2 keV) will only penetrate 5 A (< two graphite layers). Such 
shallowly implanted ions are probably lost by subsequent interstellar medium 
processing (i.e., sputtering in low velocity shocks) and this is unlikely to be 
important for the isotopic composition of Stardust. 

I V Nucleosynthesis and Stellar Evolution 

Following the seminal work by Burbride et al. (1957) and Cameron (1957), it 
is now generally accepted that almost all elements heavier than He have been 
synthesized in the interiors of stars by nuclear reactions. This transmutation of 
the elements plays a key role in the evolution of stars. Also the energy released 
stabilizes stars against gravitational collapse and, of course, provide the photons 
obsen/ed by us. The text books by Clayton (1968) and Rolfs and Rodney (1988) 
provide a thorough discussion of stellar nucleosynthesis. A recent review of the 
state of the art in nucleosynthesis is provided by the volume edited by Arnett and 
Truran (1985). 

The key nuclear processes that play a role in the synthesis of the elements 
include the major energy burning cycles: H- burning into He; He-burning into C and 
O; and C-, 0-, and Si-burning producing the intermediate peak (or silicon peak) and 
iron peak elements (16<A<60). Since the binding energy per nucleon decreases for 
high (A>60) mass number, the elements more massive than Fe have to be produced 
by other means than static nuclear burning. Essentially, these elements are thought 
to result from the combined effects of capture of neutrons onto Si-peak and Fe- 
peak elements and |3-decay. Depending on whether neutron capture is slower (s- 
process) or more rapid (r-process) than p-decay different elements are 
synthesized. The s-process will occur under neutron-rich, "static" conditions and 
red giants are considered the likely origin of most of the solar system s-process 
elements with some contribution from type II SN. The r-process characteristically 
occurs under "explosive" conditions, when a large number of neutrons are rapidly 






















Fig. 2: A schematic Hertzsprung-Russell Diagram indicating the location of various types of stars and their 
internal nuclear energy source. Theoretical evolutionary tracks for three different mass stars (1, 25, and 
120 Mg) are indicated by solid and dashed lines. Dust formation has been observed in red giants and 
supergiants, planetary nebulae, WC Wolf-Rayet stars, and novae. The recent type 11 supernova, SN1987A, 
also shows evidence for dust formation. See text for details. 

in the released. Generally, the solar system r-process elements are attributed to 
supernova explosions. The origin of a few heavy elements is not well understood, 


but might involve proton capture (p-process). Their abundance is however low. 
Finally, for completeness spallation due to the cosmic ray bombardment in the ISM 
should be mentioned. This process is, however, only expected to be of importance 
for the synthesis of the lighter elements. 

Traditionally, a discussion of stellar evolution starts with the Hertzsprung- 
Russell diagram (HR diagram), in which a star's luminosity, L, is plotted versus its 
effective ( ie., surface) temperature, Teff, or equivalently its color (Fig 2). The 
overwhelming majority of the stars fall into a few well defined regions (ie., main 
sequence, giants, supergiants) in such a diagram. To a large extent the position of a 
star in this diagram is determined by its internal nuclear energy source and its 
mass. Due to the nucleosynthesis in their interiors, stars will (slowly) evolve. In 
particular, when a particular energy source is exhausted and a new one is turned 
on, completely different internal conditions (ie., higher density and temperature) 
are required. Often, the photosphere will "rapidly" adjust itself to these changed 
conditions and move to a new location in the HR diagram. Figure 2 shows 
schematically where H-, He-core burning and H/He shell burning is of importance 
in the HR diagram. Thus, stars on the main sequence burn H into He in their core, 
while red supergiants burn He into C. Giants on the other hand have exhausted their 
central energy supply and burn H and/or He in a shell surrounding the core. 

The evolution of stars is different for different stellar masses. Stellar 
evolution can also be influenced by the mass loss associated with a strong stellar 
wind (i.e.. Wolf Rayet stars). Several theoretical evolution tracks are superimposed 
upon the Hertzsprung-Russell diagram in figure 2 (Iben and Renzini 1983; Maeder 
and Meynet 1987). During several distinct evolutionary phases material can be 
ejected by the star either in the form of a steady wind (ie., giants, supergiants, 
Wolf Rayet stars) or explosively (ie., novae, supernovae). Dust nucleation and 
condensation has been observed to take place in the outflows from red giants, 
supergiants and C-rich Wolf Rayet stars (WC stars). Boundaries for dust formation 
in these phases are schematically indicated in fig. 2. Dust formation is also 
observed in some novae and, recently, in supernova 1987a. 

Mixing of freshly synthesized material from the core or regions surrounding 
it to the surface is a common phenomena in late stages of stellar evolution. This is 
in particular true for the Stardust forming objects: giants, supergiants and WC 
stars. As a result the surface composition of such objects can change drastically. 
Indeed, such systematic variations have been observed and form one of the most 
important tests of stellar evolution theory (Lambert 1988; Willis 1982; Maeder 
1987). Likewise, material ejected by novae and supernovae has a distinctly non- 
solar composition (Truran 1985; Woosley and Weaver 1986a). Table 2 summarizes 
some of these variations for phases of stellar evolution associated with Stardust 
formation. The emphasis is on the isotopes of C and O, which because of their 


Table 2: Isotopic Anomalies 

Object Abundance ratios^ 

12C/13C 160/170 160/18Q s-process other 






- - 






= 10 





- - 

- - 

= 20 


SiC, C-dust 




- - 

- - 

26 Al (26Mg) 
22Na (22Ne) 

C-dust, PAHs 





- - 




WC stars 





12C, 22Ne 


notes: a) Observed abundance ratios relative to solar ratios in various Stardust forming objects, b) Observed 
dust materials (see section II for details), c) Predicted theoretical values. 

abundance have been the easiest to measure in astronomical spectra. Observed 
enhancements in the abundance of s-process elements, as well as some other 
relevant elements, are also indicated. The type of solid materials observed to 
condens out around these objects are also summarized in table 2. These have been 
discussed in more detail in §11 and in Tielens and Allamandoia (1987a). In the 
remainder of this section, these isotopic abundances and their nucieosynthetic 
origin will be discussed. 

A) Evolution of low mass stars 

Low mass stars (M < 8 Mq) like the Sun spend most of their lifetime (=6x1 o9 
yr for the Sun) on the main sequence burning H into He (fig. 2). After H in the core 
is exhausted, H-burning in a shell surrounding the He core takes over as the energy 
source and the star moves over onto the giant branch. The star moves slowly up on 
the giant branch (ie., higher L) until the core is massive enough to ignite He burning 
into C and O. At this point the star will rapidly move down the giant branch to the 
core-He burning region of the HR diagram (fig. 2). For stars somewhat more 


massive than the sun (i.e., >2.3 Mg) core He-burning occurs in a region distinct from 
the giant branch (at higher Teff). When the He fuel is exhausted in the core, the C-0 
core will contract until it is supported by electron degeneracy against further 
gravitational collapse. In these low mass stars, central pressures and 
temperatures never get high enough to iginite further stable burning. The star is 
now in essence a C-0 white dwarf surrounded by an extensive envelope in which 
alternating H- and He-shell burning occurs. This adds mass to the core and again 
the star moves up in luminosity on the (asymptotic) giant branch. When the 
envelope mass becomes very small mainly due to the strong stellar wind, the white 
dwarf core becomes visible and ionizes the previously ejected material forming a 
planetary nebula. At this point, nuclear energy generation has ceased and the white 
dwarf will slowly darken into oblivion. For the most massive stars in this regime 
(=8 Mq) static nuclear burning may proceed slightly further producing an 0-Ne-Mg 
white dwarf (Nomoto 1984). 

M-, S-. and C-Giants : During its evolution on the giant branch, the chemical 
composition of a low mass star will change. First, starting out as an O-rich M- 
giant, material is dredged up from the H-burning shell surrounding the He-core, 
resulting in an enrichment of "•^C, '•'^N and I^q and a small depletion in ■'2c and 
possibly "ISo (ie., CNO cycle). Once He burning is ignited, convective instabilities 
may mix He ashes to the surface. As a result the star may change from an O-rich 
M-giant to a C-rich C-giant, possibly passing through the S-giant phase in which 
C/0=1. Due to the pronounced effect of the C/0 ratio on the molecular composition 
of these cool stars, this enrichment in C can readily be observed in the molecular 
composition of the stellar photosphere. During the He shell burning phases 
(socalled thermal pulses) conditions are favorable for the formation of s-process 
elements. Thus, simultaneously with the increase in "1 2c, the abundance of s- 
process elements in the photosphere will increase relative to that of Fe. 
Observations are in good agreement with these expectations (Lambert 1988; Smith 
1989). Note that this change in photospheric composition also has a pronounced 
influence on the chemical make up of the Stardust: silicates in M-giants versus 
hydrogenated amorphous carbon and SiC in C-giants. This merely reflects the high 
stability of the CO molecule, which locks up either all of the C or all of the O, 
depending on which has the lesser abundance. The excess O or C is then available 
for oxide or carbonaceous dust formation (Salpeter 1977). 

B) Low mass binaries 

For members of close binary system, the last stages of evolution can be 
slightly different due to accretion of material from a companion on to the white 
dwarf surface. Slow accretion of H-rich material may reignite nuclear burning at 


the white dwarfs surface in a thermal nuclear runaway, leading to a nova explosion 
and the ejection of about lO^'* Mq (cf., Truran 1985). Such a system may experience 
many nova "puffs". In contrast, when accretion is rapid, the white dwarf is 
compressed and heats up. When enough material has accreted, nuclear burning is 
ignited again. A subsonic C-def lag ration wave propagates outwards and the 
released nuclear energy completely disrupts the white dwarf in a type la supernova 
explosion (Nomoto 1985). 

Novae outburst are thought to result from a thermonuclear runaway on the 
surface of an accreting white dwarf in a close binary system. Nuclear burning 
probably proceeds through the CNO cycle and enhancements in '•'^N, ISn, "ISc and 
"I^O and depletion of H are expected. Indeed, significant enhancements in "I^C and N 
have been observed for many novae ejecta (Truran 1985; Wiescher et al. 1986). 
However, enhancements in the total C, N, and abundances as well as unexpected 
enhancements of Ne, Na, Mg, and Al have also been observed. It seems that hot 
explosive H-burning and/or mixing of white dwarf material into the ejecta can be 
very important. Finally, explosive H burning in novae may also lead to the 
formation of the radioactive elements 22Na and 26ai, which decay to 22^0 and 
26Mg respectively. 

The white dwarf material in tvoe la supernova explosions undergoes 
explosive C, Ne, O and Si burning at the passage of the C deflagration wave, 
ultimately resulting in the production of iron- peak elements (Nomoto 1985). The 
outer layers burn only partially or not at all and significant amounts of 
intermediate mass elements such as ''^c, ISq, 24Mg, 28si, 32s, 36Ar, 40ca are 
ejected as well. The presence of such elements in the ejecta has been well 
established observationally (Branch et al. 1982; Branch 1984) and forms one of the 
major points in favor of C deflagration explosions over the C detonation models for 
type la SN explosions. If such a SN result from the accretion of material on to a 
white dwarf surface, rather than from the merging of two white dwarfs, than He 
flashes may produce some s-process elements. The conditions during the SN 
explosion (ie., in the precursor shock wave) may then favorable for the production 
of r-process elements through neutron capture on these s-process elements 
(Nomoto 1985). 

C) Evolution of massive stars 

After H is exhausted in their cores, a massive star will evolve to the right in 
the HR diagram and will burn He in its core as a red supergiant (Fig. 2). In contrast 
to low mass stars, the resulting C, O core of massive stars never becomes 
degenerate and, after He is exhausted, further "quiet" nuclear burning can occur, 
transforming C and O through a number of intermediate steps into the Fe peak 


elements. Since the Fe peak elements are the most tightly bound nuclei, further 
burning will not release energy and the iron core will rapidly contract to nuclear 
densities. This core may bounce like a rubber ball driving an outward propagating 
Shockwave. This, coupled with energy transfer between the core and the envelope 
through neutrinos, may lead to violent ejection of most of the star - a socalled 
type II supernova explosion - leaving a compact remnant behind (ie., neutron star or 
black hole; Woosley and Weaver 1986a). Initially, the temperature behind the 
outward propagating shock is high enough for further (explosive) nucleosynthesis, 
but a large fraction of the ejecta contains the ashes of previous burning cycles 
(ie., the elements C through Al). For example, detailed numerical calculations show 
that a 25 M^ star leaves a 2 Mq compact remnant behind and ejects about 4.3 Mq of 
freshly synthesized heavy elements, resulting in a typical elemental enrichment of 
a factor 10 over solar (Woosley and Weaver 1986a). Since the later stages of 
evolution are very rapid, the stellar photosphere does not have time to readjust 
itself and they will all occur in the same part of the HR diagram, the red 
supergiant regime (Fig. 2). 

The detailed evolution of very massive stars is heavily influenced by their 
mass loss rate and these stars spend only a small fraction of their (He-core 
burning) life as red supergiants. Instead, they mainly burn He in their core as blue 
or yellow supergiants (Chiosi and Maeder 1986). As a result of the mass loss, the H 
envelope is progressively lost. Moreover, convection will carry freshly synthesized 
material to the surface and the surface composition will evolve, showing 
progressively the H-burning (CNO cycle) and He-burning products. Thus these 
massive stars will evolve through the WN (N-rich) and WC (C-rich) Wolf Rayet 
phases. Finally, these stars will explode as supernovae, but because by that time 
they are H-poor, their spectrum differs from that of typical type II supernovae. 
Presumably, this is the origin of the type lb supernovae (Woosley and Weaver 

Red Supergiants : Mixing during this phase will bring the products of H- 
burning via the CNO cycle to the photosphere. Enhancements in the """^N and "I^C 
abundance are expected and have been observed (Maeder 1987). Changes in the 
isotope ratios are also expected, but a detailed comparison is somewhat hampered 
by uncertainties in the relevant reaction rates. Such (anomalous) O isotope ratios 
have been observed for the M supergiants a Sco and a Ori (Lambert et al 1984; 
Harris and Lambert 1984) and these are summarized in table 2. These anomalous 
isotopic ratios will be preserved in the silicate dust observed to condens around 
these objects. 

C-rich Wolf-Ravet stars : Due to extensive mass loss, very massive stars go 
through a Wolf Rayet phase in which the photospheric composition has changed 
drastically. In the C-rich phases (ie., WC stars), the photosphere consist of He 


burning products (ie., '•^c, 22fv|e) and H, "ISc, "l^o and "f^o essentially disappear. WC 
stars will also show overabundances of 23^3, 25,26Mg and 29,30si (Prantzos et al. 
1986). The overabundance of 26ai, produced during the preceding N-rich Wolf- 
Rayet phase, will however disappear on a short timescale (=10^ yr). The dust 
observed to condense out around WC stars has an amorphous cartDon character, but 
will obviously not contain H. 

Type II SN explosions will eject C-, 0-, Si-burning products into the ISM. 
This includes the iron peak elements, as well as intermediate mass elements 
associated with the Si peak (Woosley and Weaver 1986a). Supernovae are probably 
the origin of the r- and p-process elements and may also contribute to the solar 
abundance of some s-process elements (Woosley and Weaver 1986a). An 
enhancement of 26^1 is produced and may be incorporated into any condensing dust 

Only a very limited study of supernova explosions of Wolf Rayet stars ( type 
lb SN) have been performed, mainly concentrating on the explosion mechanism and 
the (possible) black hole remnant (Woosley and Weaver 1986b). It is likely that 
some of the iron core and the surrounding partially burned envelope will be 
ejected. If, as is likely, the Filipenko-Sargent object represents a type lb SN, then 
O, Na and Mg are ejected by the SN explosion of such a massive star (Filipenko and 
Sargent 1985). 

The measured isotopic composition of Stardust 

In recent years various components have been isolated from carbonaceous 
meteorites which because of their anomalous isotopic composition are generally 
identified with Stardust which has been incorporated into the meteorite parent 
body without major modification in the ISM or solar nebula (ie., melting, 
vaporization). Since these studies provide some clues to the Stardust components 
that may expected in cometary bodies, it is of some relevance to briefly 
summarize the results on these components and their origin. Excelent reviews of 
this field can be found elsewhere (Anders 1988; Anders et al. 1989). We will 
concentrate on some of the carbonaceous components, Ca, p, 5, and e, which 
correspond to amorphous carbon, diamond, and two types of SIC grains. 

CP : This component consist of SiC grains with Xe isotopic abundances 
characteristic for the s-process. This anomaly is accompagnied by a similar 
enrichment in s-process Kr isotopes as well as by the presence of a substantial 


enrichment in "l^c (''2c/"'3c=40; Anders 1988). These isotopic anomalies point 
towards C-rich red giants as the birth site of this Stardust component (see table 
2). The low "ISc/ISc reflects the interplay of first dredge up during the 0-rich red 
giant phase, leading to "• 20/130=10, and the dredge up of pure "l^c during He 
flashes in the asymptotic red giant phase which transform the 0-rich giant into a 
C-rich one. Further astronomical support for this identification results from the 
presence of the 11.3|im SiC stretching vibration in the IR spectra of many C-rlch 
giants. Since such stars also produce copious amounts of amorphous carbon, a 
similar isotopic anomaly in the amorphous carbon phase of carbonaceous 
meteorites might also be expected. Its isolation might, however, be more difficult. 

The nature of the process that lead to the trapping of the noble gas 
impurities is unclear. Ion-implantation is, a priori, the most likely origin for 
trapped noble gasses. However, the gas-grain drift velocities expected in the 
outflows from red giants are less than 20 km/sec (Tielens 1983), resulting in an 
implantation depth of less than sA. It is unlikely that this outermost layer would 
survive the frequent shocks in the ISM. The central stars of planetary nebulae, the 
descendant of red giants, do have high velocity winds (=1000 km/sec) which could 
lead to deeper implantation, but they will interact with only a small fraction of 
the dust ejected in the red-giant phase. This holds even more for a possible type la 
supernova explosion, which can occur in the white dwarf descendant if it is part of 
a double system. Furthermore whether these ejecta would contain s-process 
elements remains to be seen. 

The alternative explanation for the origin of the trapped noble gasses, 
adsorption during condensation, has its share of problems. The high release 
temperatures (>1400K) and high noble gas content (Xe=10"'' cm"3 STP /g) are hard 
to reconcile with adsorption. Although amorphous carbon when exposed to Xe at a 
partial pressure of =10-'' atm, does lead to "tightly bound" Xe (=10-9 cm-3 STP /g) 
which is released only at high temperatures (=1000K), this probably is a particular 
property of the disordered structure of amorphous carbon with its large porous 
network (Wacker et al. 1985) and may not apply to SiC grains. Moreover, although 
the Xe partial pressure in these experiments was 10 orders of magnitude larger 
than expected for C-rich giants, the tightly bound Xe content was two orders of 
magnitude less than in Cp. Further experimental studies particularly on highly 
disordered samples are required to address this quantitative aspect. 

Surprisingly, only 4x10*5 of the Si in carbonaceous meteorites is contained 
within SiC (Anders et al. 1988). The SiC Stardust budget analysis in section IIA 
concluded that about 2% of all the silicon is injected in into the ISM in the form of 
SiC. Taking into account destruction of Stardust by strong SN shocks, the average 
volume of the ISM should contain about 0.3% of all the Si in the form of SiC 
Stardust - about two orders of magnitude more than measured. Tang and Anders 


(1988) have measured the 21 Ne content of SiC grains, presumably a cosmic ray 
spallation product, and concluded that their average ISM lifetime was only 50 Myr. 
This is decidedly less than the theoretical estimate (4x1 o8 yr; McKee 1989) and 
may imply that the sun formed in an non-average region of the galaxy (i.e., OB 
association) which has been heavily affected by SN shocks. Assuming this lifetime, 
results in an average fraction of Si contained in SiC of 4x10"^, still an order of 
magnitude more than measured. Of course, such an average will not mean much if 
the local ISM was not in a steady state situation. An alternative explanation for 
the low abundance of SiC grains may be that most of the Stardust (=99%) was 
destroyed upon entering the solar nebula. Recent theoretical studies of the early 
solar nebula predict mid-plane temperatures of 1500K at the asteroid belt (Boss 
1988), close to the evaporation temperature of SiC grains (=1600K; Larimer and 
Bartholomay 1979). Chrondules and Ca-AI rich inclusions show abundant evidence 
for such high temperatures in the asteroid belt, although this might be related to 
transient heating events (i.e., lightning; see, Grossman et al. 1988; Hewins 1988; 
MacPherson et al. 1988). Likewise, the coarser matrix material, which is abundant 
in carbonaceous meteorites, may result from nebular and/or parent body processes 
(Scott et al. 1988). The Stardust components identified in carbonaceous meteorites 
may then refer to a relatively late addition just prior to the formation of the 
meteoritic parent body, possibly when most of the dust has settled and 
temperatures are lower (i.e., reduced opacity). 

Ce : This is another SiC dust component, often associated with spinel and thus 
separable from C|3. It is highly enriched in ^^Ne and '^^C O^Cf^'^C < 10) and shows 
small enrichments in the heavy isotopes of Si (Anders et al. 1989). Although most 
of the 22Ne in the galaxy may result from He burning (in WC stars; Maeder 1983), 
the high isotopic fractionation suggests an origin in 22^3 which decays to 22^0 
with a half time of 2.5 yr (Clayton and Hoyle 1976). Since, in contrast to Ne, Na 
binds strongly, large fractionations are possible for this radioactive daugther 
product. The high ^^C enrichments are also characteristic for novae (Truran 1985). 
Finally, the variable enrichment in the Si isotopes may also be consistent with a 
novae origin, particularly if the progenitor is an Ne-O-Mg white dwarf (Wiescher et 
al. 1986). 

Surprisingly, there is no evidence for SiC grains in the IR spectra of novae 
(see section II). This is somewhat discomforting since it should be detectable if 
all the available Si condenses out as SiC. This might imply that most of the Si 
forms silicates and only a minute fraction SiC. However, this poses a second 
problem with the nova interpretation. The meteoritic abundance of SiC containing 
22Ne relative to that containing s-process is about unity (Anders et al. 1989). This 


should reflect the relative Importance of these two sources of SIC dust. However, 
assuming very conservatively that all the Si in nova ejecta and only 10% in C-giant 
ejecta forms SiC we estimate that C-rich giants produce ten times more SiC (table 
1). Of course, as for Cp, the measured absolute amount of Si in SiC from novae is 
lower than predicted (see above). Another puzzling aspect is the close association 
of Ce, but not Cp, with spinel. It suggests perhaps an interrelationship dating back 
to the nova ejecta. Now, novae are the only Stardust source suspected to condens 
oxides (silicates) as well as C-dust (see section II), so a mixed bag of dust is 
perhaps feasible. It would suggest, though, that this spinel phase has to be heavily 
enriched in '•^O as well as in the daugther product of 26^1, 26Mg (table 2). 

Here it is worth noting that another carbonaceous component ( Ca ) with a high 

22Ne content has been isolated from carbonaceous meteorites. Its structure is 
poorly defined but is probably some form of amorphous carbon (Anders 1988). The 
22Ne and "I^N again point towards an origin in novae. Since amorphous carbon is a 
suspected nova condensate, this seems entirely reasonable. The 22Ne was probably 
trapped as intercalated 22Na, while N substituted for C. The measured C and N 
isotopes are somewhat different from those of Cp and might suggest a slightly 
different set of novae contributed to these two carbonaceous components in 
meteorites or. a slight contamination by Stardust with a different origin (i.e., C- 
giants ?). 

Q b: This is the most abundant form of elemental C in carbonaceous 
meteorites (=0.1% of elemental C). It consists of small (25A) diamond 
microcrystals containing a Xe anomaly known as Xe H-L, which is enriched in the 
light as well as the heavy Xe isotopes (Anders et al. 1989). Curiously, the C 
isotopes are solar but "I^N is anomalous (depleted). The measured Xe anomalies 
resemble the results of the p- and r-process in various zones of type II SNe 
(Anders 1988). The r-process might actually be driven by neutrons generated by the 
interaction of neutrinos (released by the central neutron star) with He atoms in 
the He/''2c-rich zone in the ejecta (Epstein et al. 1988). 

The carbonaceous carrier of the Xe anomaly may have condensed in this C- 
rich zone of the supernova ejecta, or may represent grains formed during an earlier 
mass-loss phase of the SN progenitor (Clayton 1981). In either case, the trapping 
of the noble gasses is probably due to ion implantation. The ejected material, 
coasting at 1000-5000 km/sec, will drive a strong shock wave into the 
surrounding gas, sweeping it up, while a reverse shock will propagate backwards 
(in mass coordinates) slowing the ejecta down. Initially, the ejecta and swept up 
circumstellar material will interact only over a thickness comparable to the 


stopping length (=10'' ^ cm). Thus, only a very minor fraction of the previously 
ejected dust would be bombarded by the ejecta. However, the contact discontinuity 
"separating" the ejecta and swept up matter will be Rayleigh-Taylor unstable, 
driving global mixing at a (subsonic) velocity of =1000 km/sec. Because of their 
inertia, large grains in this turbulent velocity field will develop drift velocities of 
this same order. Thus, both dust formed in the ejecta as well as "old" dust will be 
bombarded by ions with energies of =5keV/amu and implantation to a depth of 
=1500A will occur. Note that very small grains (<100A) will not stop very many 
impacting ions (<0.1%). Moreover, they would show very large fractionation effects 
between for example Kr and Xe. Thus, in such a model the small microcrystais 
observed in carbonaceous meteorites initially had to be part of a much larger 
polycrystalline grain (=1500A; Blake et al. 1988). 

Some implantation of isotopicaily anomalous species can also result from 
the thermal bombardment in the non-radiative (post) shocked gas. A shock velocity 
of 1000 km/sec corresponds to a thermal energy of =5 keV/ion and will lead to 
implantation at a depth of about 75 A. Again the smallest grains (now < 25 A) will 
stop only few (< 1%) impacting ions. Of course, this bombarment will also lead to 
sputtering from the surface and irrespective of size the outer =300 A of a grain 
will be removed over the expansion timescale of the shock (Seab 1987). Thus, 
again, an initial grain size much larger (>300A) than measured for the carrier of 
the Xe anomaly is implied. 

Clayton (1981) has suggested that implantation occurred when the fast SN 
ejecta swept up C-dust ejected in a previous evolutionary phase. Since C-dust is 
required, a C-giant progenitor is implicated but such stars will not explode as type 
II SN. They can explode as a type la SN, if a member of a binary system, and a 
variant based on this premisses has been developed by Jorgensen (1988). However, 
it is unclear whether such SN will indeed lead to a large enrichment in r-process 
elements (Nomoto 1985). Moreover, the SN explosion may occur much later than the 
red-giant phase and little of the latters grains may be affected by the high 
velocity SN ejecta. Thus, it would be difficult to explain the large trapping 
efficiency and the large fraction of elemental C in C5 in carbonaceous meteorites. 
Finally, the C isotope ratio measured in the diamonds is unlike that for C-giants 
and in fact, unlike any single source of C-dust (table 2). The almost solar C isotope 
ratio is indeed very curious, since "l^c and "I^C are made by quite different 
nucleosynthetic reactions (He burning versus CNO cycle) and their galactic budget 
is dominated by different types objects (C-giants and WC stars versus novae, and 
C- and 0-giants). It seems unavoidable that multiple birth sites for the C5 phase 
are involved, which represent an average cross section of the injection of all 
elemental C. Since the dominant elemental carbon sources show C-stardust 
formation, such a model is perhaps possible. The decreased '•^N/'I'^N ratio in C5 


may then just reflect the imortance of C-giants in the C-stardust budget, since 
such stars will have been enriched in '''^N via the CNO cycle. 

Two distinct models for the formation of interstellar diamond dust have been 
advocated, metastable chemical vapor deposition in stellar ejecta (Anders et al. 
1989), and high pressure transformation of amorphous carbon dust due to grain- 
grain collisions in interstellar shocks (Tielens et ai. 1987). Some laboratory 
techniques have been very succesful in depositing C-films with properties very 
similar to diamond. In general, however, amorphous carbon with a predominantly 
aromatic bonding character is formed. For example, burning a hydrocarbon flame 
will result in copious amounts of soot and (unfortunately) not diamonds. This 
merely reflects the thermodynamic preference for graphite at low temperatures 
and pressures. Although soot contains tetrahedrally bonded C, its structure - 
aromatic platelets connected by diamond like hydrocarbon chains - is quite unlike 
that of the C5 phase (Blake et al. 1988). The conditions in the outflow from C- 
giants is quite similar to those in hydrocarbon flames and soot formation is 
expected (Tielens 1989). Indeed, C-rich PN - the descendants of C-rich giants - 
show abundant evidence for aromatic hydrocarbon molecules, the building blocks of 
soot. Likewise, red fluorescence observed in the ISM indicates a considerable 
amount of hydrogenated amorphous carbon (a:C-H) grains which has a structure 
similar to soot (Duley 1985). The IR spectra of WC stars also show evidence for 
aromatic rather than diamond binding of the condensing grains (Cohen et al. 1989). 
Nevertheless, it is perhaps possible that the very specific conditions required for 
diamond formation by CVD techniques are satisfied in some celestial objects. 
However, in the eyes of this (perhaps biased) reviewer, whether such an object can 
actually provide sufficient amounts of diamond and particularly a solar C isotope 
ratio is doubtful. 

Grain-grain collisions (v> 10 km/sec) behind strong shocks can provide the 
high pressures (>400 kbar) required to convert graphitic C into diamond (Tielens et 
al. 1987). Numerous laboratory studies have shown that this process is very 
efficient (Bundy et ai. 1973). Several of the measured properties of the meteoritic 
diamond phase are readily explained by this high pressure process. These include 
the microcrystalline grain size (very similar to those measured in laboratory 
experiments), the structure (small crystals surrounded by an amorphous rim), and 
the quite normal '^^C/'^^C ratio (many C-stardust birthsites of which only one 
contributed Xe H-L; Tielens 1989; Blake et al. 1988). The high pressure method has 
one major drawback: the expected leftover amorphous carbon Stardust component 
with a similar isotopic composition has not been detected in carbonaceous 
meteorites to an upper limit of perhaps 10% of that of C5 (Anders et al. 1989). 
Theoretical models of shock processing of interstellar dust imply that about 15% 
of the C-stardust survives into the solar nebula (§ II), about 10% is transformed 


into diamond (Tieiens et al. 1987), and the remainder is destroyed. If, however, the 
Stardust lifetime of 50 Myr derived from SiC measurements is adopted (Tang and 
Anders 1988; see above), then only 3% of the C-stardust survived into the solar 
nebula. The fraction converted into diamonds (10%) is, however, not very affected 
by the absolute shock frequency since they are formed as well as destroyed by 
shocks (Tieiens et al. 1987). Nevertheless, the predicted surviving Stardust 
component is still somewat uncormfortably large. Of course, if dust sources and 
sinks in the local neighbourhood from which the sun formed were not in 
equilibrium (ie., nearby SNe destroyed most of the local dust; Anders et al. 1988), 
then essentially all of the surviving C-stardust could have been transformed into 
diamonds. Thus, at this point, it seems that this is not lethal argument against a 
shock origin for these diamonds. 


As table 2 demonstrates the isotopic composition of the major elements, C 
and O, varies by factors two to several orders of magnitude from one class of 
sources to another and even within classes. Similar variations have been observed 
for s-process elements such as Yttrium, Zirconium, and Neodymium. The Stardust 
condensates are expected to preserve this nucleosynthetic record on an individual 
grain basis. Conversely, this implies that any bulk^ sample will contain a mixture 
of Stardust from many different birthsites and with a very heterogeneous isotopic 
composition. Since the Stardust birthsites also dominate the return of all elements 
to the ISM, the average of such a bulk sample might not differ much from solar. 
Indeed, the almost solar "1 2c/'' 3c ratio of C5, despite the Xe signature indicating a 
SN origin, may be a good example of such a mixture. For a proper assesment of the 
nucleosynthetic record contained within comets, analysis on an (interstellar) grain 
by grain basis is therefore almost imperative. 

Theoretical studies indicate that about 15% of the Stardust originally 
ejected will survive the rigors of the ISM. Yet, meteorites seem to contain only 
about 0.1% of the cosmic elemental C in the form of Stardust components. In fact, 
while about 50% of the interstellar dust must be made up of carbonaceous 
material, the abundance of carbonaceous solids in meteorites is much more limited 
(=2% by mass). Likewise SiC Stardust is much less abundant in meteorites than 
expected from Stardust budget considerations (§V). There are two possible 
explanations for this discrepancy (Anders et al. 1989): 1) Solar nebula processes 

1 bulk relative to Stardust masses (=10"1^ g) 


have lead to the vaporization of most of the interstellar dust mass in the asteroid 
belt. 2) The sun formed out of an anomalous region of interstellar space (e.g., an OB 
association) where most of the preexisting dust was destroyed by nearby SNe. 
Since the refractory component of cometary materials is expected to be much less 
affected by solar nebula and parent body processes, a comet sample return mission 
is expected to be very valuable for the study of particularly the more fragile 
Stardust components. In particularly, since the oxide Stardust preserved by 
meteorites has been heavily diluted by solar nebula condensates - probably only 
0.1% of the Si is in surviving Stardust and 99.9% in solar nebula condensates - such 
a mission is indispensable for an analysis of silicate Stardust. 

Since present day techniques require sample sizes larger than a micron, new 
powerful analytical techniques have to be developed to investigate samples on an 
individual interstellar/circumstellar grain basis. Given that a typical Stardust 
grain contains only 10^ atoms and that the accurate determination of, for example, 
the C isotope ratio requires "counting" to an accuracy of 105 atoms, the challenge 
is clear. In meteoritic studies this challenge has been circumvented by sorting on 
chemical and physical properties. This has resulted in larger samples, but some 
information may be lost by this averaging process. This can be particularly 
misleading if different birthsites lead to Stardust with similar properties (ie., M- 
giants and supergiants) or if an isotopic ratio varies from source to source (ie., O- 
isotopes in M giants). Moreover, presently such sorting is primarily based on 
chemical properties (i.e., resistance against acid treatment, pyrolysis, 
combustion) and other sorting techniques - perhaps based on structural or 
mineralogical characteristics, may lead to hitherto unsuspected (fragile) Stardust 
components. Now, some data can only be obtained on "bulk" samples. The anomalous 
isotopic composition of the noble gasses is a case in point. With a Xe abundance of 
=10"8 per C atom in C5, an individual 1000 A grain would contain only 10 ''36xe 

atoms and only about one in every 10^ individual 25A diamond microcrystallites 
would contain such an atom. Clearly "bulk" analysis has proven its value already. Of 
course, both approaches will be valuable and to some extent complimentary. 

Despite these difficulties associated with the analysis of returned samples, 
it is expected that the Rosetta mission will provide an important testing ground 
for theories of nucieosynthetic origin of the elements, as well as the formation 
and evolution of interstellar dust, and the formation of solid bodies in the solar 



Anders, E., 1988, in Meteorites and the Early Solar System, eds. J. Kerridge and M. Mathews, (Univ. Arizona 

Press, Tucson), p. 927. 
Anders, E., Lewis, R.S., Tang, M., and Zinner, E., 1989, in Interstellar Dust, eds. LJ. Allamandola and A.G.G.M. 

Helens, (Dordrecht, Reidel), p.389. 
Arnett, D., Fryxell, B., and Muiler, E., 1989, Ap. J. Letters, 34A, L63. 
Amett, W. D., and Truran, J.W., 1985, Nucleosynthesis: Challenges and New Developments, (Univ. Chicago Press, 

Blake, D.F., Freund, F.. Krishnan, K.F.M., Echer, C.J., Shipp, R., Bunch, T.E., Tielens, A.G., Lipari, R.J., 

Hetherington, C.J.D., and Shang, S., 1988, Nature, 332, 611. 
Boss, A.P., 1988, Science, 2A-[, 565. 
Branch. D., 1984, Ann. N.Y. Acad. ScL, 422, 186. 
Branch, D., et al., 1982, Ap. J. Letters. 252, L61. 
Burenkov, A.F., Komarov, F.F., Kumakhov, and Temkin, M.M., 1986, Tables of Ion Implantation Spatial 

Distributions, (Gordon&Breach, New York). 
Burbridge, E.M., Burbridge, G.R., Fowler, W.A., and Hoyle, F., 1957, Rev. Mod. Phys., 29. 547. 
Cameron, A.G.W., 1957, Chalk River Report, (Atomic Energy Canada, Ltd), CRL-41. 
Chiosi, C. and Maeder, A., 1986, Ann. Rev. Astr. Ap., 24, 329. 
Clayton, D.D., 1981, Proc. Lunar Planet Sci.,^2B, 1781. 

Clayton, D.D., 1968, Principles of Stellar Evolution and Nucleosynthesis, [McGraw Hill, New York). 
Clayton, D.D., and Hoyle, F., 1976, Ap. J., 203. 390. 

Cohen, M., Tielens, A.G.G.M., and Bregman, J.D., 1989, Ap. J. Letters. 344, in press. 
Curl, R.F., and Smalley, R.E., 1988, Science, 2A2, 1017. 
Duley, W.W., 1985, M.N.R.A.S., 215, 259. 
Field, G.B., 1974, Ap. J., 187, 453. 

Fiiipenko, A.V., and Sargent, W.LW., 1985, Nature, 3^6, 407. 

Gehrz, R.D., 1989, in Interstellar Dust, eds. LJ. Allamandola and A.G.G.M. Tielens, (Dordrecht, Reidel), p.445. 
Gehrz, R.D., 1988 Ann. Rev. Astr. Ap., 26, 377. 
Gilra, D.P., ???? 
Greenberg, J.M., 1989, in Interstellar Dust, eds. LJ. Allamandola and A.G.G.M. Tielens, (Dordrecht, Reidel), 

Grossman, L, and Larimer, J.W., 1974, Rev. Geophys. Space Phys., 12, 71. 
Grossman, J.N., Rubin, A.E., Nagahara, H., and King, E.A., 1988, in Meteorites and the Early Solar System, eds. J. 

Kerridge and M. Mathews, (Univ. Arizona Press, Tucson), p. 619. 
Harris, M.J., and Lambert, D.L, 1984, Ap. J., 281, 739. 
Hewins, R.H.,1988, in Meteorites and the Early Solar System, eds. J. Kerridge and M. Mathews, (Univ. Arizona 

Press, Tucson), p. 660. 
Iben, L, and Renzini, A., 1983, Ann. Rev. Astr. Ap.,2^, 271. 

Jaycock, M.J., and Parfitt, G.D., 1 986, Chemistry of Interfaces, (Wiley and Sons, New York). 
Jenkins, E.B., 1989 in Interstellar Dust eds. L.J. Allamandola and A.G.G.M. Tielens, (Dordrecht, Reidel), p. 23. 
Jura, M., 1987, in Interstellar Processes, eds. D. Hollenbach and H. Thronson, (Reidel, Dordrecht), p. 3. 
Kerridge, J.F., 1986, in Interrelationships among Circumstellar, Interstellar and Interplanetary Dust, eds. J. Nuth 

and R. Stencel, NASA CP 2403. p.71. 
Kemdge, J.F., and Chang, S., 1985, in Protostars and Planets II, eds. J. Black and M. Mathews, (Univ. Arizona 

Press, Tucson), p. 738. 
Lambert, D.L, Brown, J.A., Hinkle, K.H.. and Johnson, H.R., 1984, Ap. J.. 284, 223. 
Lambert, D.L, 1988, in The evolution of Peculiar Red Giants, (Reidel, Dordrecht), in press. 
Larimer. J. W., and Bartholomay, M., 1979, Geochim. Cosmochim. Acta, 43, 1455. 
Iben, I., and Renzini, A., 1983, Ann. Rev. AsU. Ap..2^, 271. 


Jorgenson, U.G., 1988, Nature, 3Z2, 702. 

MacPherson. G.J., Wark, D.A., and Armstrong, J. T., 1988, in Meteorites and the Early Solar System, eds. J. 

Kerridge and M. Mathews, (Univ. Anzona Press, Tucson), p. 746. 
Maeder, A., 1983, Astr. Ap.,MQ, 130. 
Maeder, A., 1987, Astr. Ap..^73, 247. 
Maeder, A., and Meynet, G., 1987,y4sfr. /4p., 182. 243. 
Martin, P.G., and Rogers, C, 1987, Ap. J., 322, 374. 

Mathis, J. S., 1989, in Interstellar Dust. eds. LJ. Allamandola and A.G.G.M. Tielens, (Dordrecht, Reidel), p.357. 
McKee, C.F.. 1989, in Interstellar Dust, eds. LJ. Allamandola and A.G.G.M. Tielens, (Dordrecht, Reidel), p.431. 
Molier. W., Scherzer, B.M.U., and Ehrenberg, J., 1982, J. Nuci. Mat., 111&112, 669. 
Nomoto, K., 1984, Ap. J., 277, 791. 

Nomoto, K., 1985, in Nucleosynthesis: Challenges and New Developments, (Univ. Chicago Press, Chicago), p. 202. 
Nuth, J.A., and Moore, M.H.. 1988. Ap. J. Letters, 329, L113. 
Pottasch, R.N., 1984, Planetary Nebulae, (Reidel, Dordrecht). 
Prantzos, N., Doom, C. Arnould, M., and de Loore, C. 1986, Ap. J., 304, 695. 
Roche, P.F., 1989, in Planetary Nebulae, eds. S. Torres-Peimbert, (Kluwer, Dordrecht), p. 11 7. 
Rolfs, C.E., and Rodney, W.S.. 1988, Cauldrons in the Cosmos, nuclear Astrophysics, (Univ. Chicago Press, 

Salpeter, E.E., 1979, Ann. Rev. Astr. Ap., 15, 267. 
Scherzer, B.M.U., 1983, in Sputtering by Particle Bombardment II, ed. R. Behrisch, (Springer Verlag, Berlin), 

Scotl, E.D.R., Barber, D.J., Alexander. CM., Hutchison. R., and Peck, J.A., 1988, in Meteorites and the Early 

Solar System, eds. J. Kerridge and M. Mathews, (Univ. Arizona Press, Tucson), p. 718. 
Scoville, N.Z., and Sanders, D.B.,1987, in interstellar Processes, eds. D. Hollenbach and H. Thronson, (Reidel, 

Dordrecht), p. 21. 
Seab, G.. 1987, in Interstellar Processes, eds. D. Hollenbach and H. Thronson, (Reidel, Dordrecht), p. 491. 
Smith, V.V., 1989, in Cosmic Abundances of Matter, ed. C.J. Waddington, A.I.P. Conference Publication 113. 
Snow, T.P., 1975, Ap. J. Letters, 202, L87. 
Tammann, G.A., 1 982, in Supemovae: A Survey of Current Research, eds. M.J. Rees and R.J. Stoneham, (Reidel, 

Dordrecht), p. 371. 
Tang, M., and Anders, E., 1988, Ap. J. Letters. 235, L31. 
Thielemann. F.K.. Nomoto. K., and Yokoi. K.. 1986. in Nucleosynthesis and its Implications on Nuclear and Particle 

Physics, eds. J. Audouze and N. Mathieu, (Reidel, Dordrecht), p. 131. 
Tielens, A.G.G.M., 1983. Ap. J., 271, 702. 

Tielens, A.G.G.M.. 1989. in Carbon in the Galaxy: Studies from Earth and Space, ed. J. Tarter, NASA CP, in press. 
Tielens, A.G.G.M.. and Allamandola. L.J.. 1987a, in Physical Processes In Interstellar Clouds, eds. G.Morfil and M. 

Scholer, (Dordrecht, Reidel). p.333. 
Tielens. A.G.G.M.. and Allamandola, LJ.. 1987b. in Interstellar Processes, eds. D. Hollenbach and H. Thronson, 

(Reidel, Dordrecht), p. 397. 
Tielens, A.G.G.M., Seab, C.G., Hollenbach, D.J., McKee, C.F., 1987, Ap. J. Leffers, 319, L103. 
Truran. J.W., 1985. in Nucleosynthesis: Challenges and New Developments, (Univ. Chicago Press. Chicago), p. 

Wacker, J. F., Zadnik, and Anders, E., 1985, Geochim. Cosmochim. Acta, 49, 1035. 
Wiescher, M., Gorres, J., Thielemann. F.-K., Ritter, H., 1986, Astr. Ap., 160, 56. 
Willis, A.J., 1982. in Wolf-Rayet Stars: Observations. Physics. Evolution, eds. C.W.H. de Loore and A.J. Willis. 

(Dordrecht. Reidel). p.87. 
Wooden, D.. 1989, Ph. D. Thesis, UCSC, in preparation. 
Woosley. S.E.. and Weaver. T.A.. 1986a. Ann. Rev. Astr. Ap..2A. 205. 

Woosley, S.E.. and Weaver, T.A.. 1986b, in Radiation Hydrodynamics in Stars and Compact Objects, eds. D. Mihalas 
and K.H. Winkler. (Dordrecht. Reidel), p.91. 



John S. Mathis 

Washburn Observatory 

University of Wisconsin-Madison 



John S. Mathis 
Washburn Observatory, University of Wisconsin-Madison 


"Interstellar dust" forms a continuum of materials with differing properties which I divide into 
three classes on the basis of observations: (a) Diffuse dust, in the low-density interstellar 
medium; (b) outer-cloud dust, observed in stars close enough to the outer edges of molecular 
clouds to be observed in the optical and ultraviolet regions of the spectrum, and (c) inner-cloud 
dust, deep within the cores of molecular clouds, and observed only in the infrared by means of 
absorption bands of C-H, C=0, 0-H, C^N, etc. 

There is a surprising regularity of the extinction laws between diffuse- and outer-cloud dust. 
The entire mean extinction law from infrared through the observable ultraviolet spectrum can be 
characterized by a single parameter. There are real deviations from this mean law, larger than 
observational uncertainties, but they are much smaller than differences of the mean laws in diffuse- 
and outer-cloud dust This fact shows that there are processes which operate over the entire 
distribution of grain sizes, and which change size distributions extremely efficiently. 

There is no evidence for mantles on grains in local diffuse and outer-cloud dust. The only 
published spectra of the star VI Cyg 12, the best candidate for showing mantles, does not show 
the 3.4 |im band which appreciable mandes would produce. Grains are larger in outer-cloud dust 
than diffuse dust because of coagulation, not accretion of extensive mantles. 

Various theories of grains are included in Table 1 . Core-mande grains favored by J. M. 
Greenberg and collaborators, and composite grains of Mathis and Whiffen (1989), are discussed 
more extensively (naturally, I prefer the latter). The composite grains are fluffy and consist of 
silicates, amorphous carbon, ans some graphite in the same grain. 

Grains deep within molecular clouds but before any processing within the solar system are 
presumably formed from the accretion of icy mandes on and within the coagulated outer-cloud 
grains. They should contain a mineral/carbonaceous matrix, without organic refractory mantles, in 
between the ices. Unfortunately, they may be significandy processed by chemical processes 
accompanying the warming (over the 10 K of the dark cloud cores) which occurs in the outer solar 
system. Evidence of this processing is the chemical anomalies present in interplanetary dust 
particles collected in the stratosphere, which may be the most primitive materials we have obtained 
to date. The comet return mission would greatly clarify the situation, and probably provide 
samples of genuine interstellar grains. 


Mathis, John S. 


One of the most interesting results of a comet-retum mission will be the recovery of relatively 
pristine grains which will, hopefully, tell us a great deal about the nature of interstellar dusL The 
main problems in interpreting the results will be in determining the amount of processing which 
has taken place at two separate stages in the evolution of the returned grains: (a) since the grains 
left the low-density ("diffuse") interstellar medium (ISM) entered the inner regions of a relatively 
dense, cold, and dark molecular cloud, and formed the protocometary grains which later clumped 
together to become the comet's nucleus, and (b) modifications in the nature of the originally 
interstellar grains after they were clumped together into the comet nucleus. The chemical species 
present in the returned material will provide information about the reactions which took place as 
the nucleus was accreting. 

In this paper I wiU try to explain what we know about interstellar grains from observations, 
what theories have been suggested to explain these observations, and speculate upon the evolution 
of the grains throughout their histories. The very term "interstellar grain" encompasses a variety 
of materials. 

It is misleading for us to refer to the solid materials seen along various lines of sight as 
"interstellar dust," unless we keep clearly in mind that these materials vary considerably from place 
to place within interstellar space. We are not dealing with any homogeneous, well-defined 
substance, but rather a collection of particles whose composition and size distribution surely 
changes considerably from the diffuse ISM to deep within the cold, dark, dense clouds in which 
comets are formed. The physical conditions from these two regions of space are vastly different 
In parts of the diffuse ISM, the density of atomic H is about 0.1 cm-3 and the temperature about 
10^ K, while in the diffuse "clouds," the density is about 30 cm-3 and the temperature 100 K. The 
radiation density from starlight is about 0.5 eV cnr^. In contrast, within the deepest parts of 
molecular clouds the density is 10^ cm^^ and the temperature about 10 K, with a sky which shows 
no visible stars. It is small wonder that the grains are also very different in the two regions. We 
shall have to keep the various regions firmly in mind as we discuss the dust. 

As the Galaxy rotates, its interstellar matter passes through the spiral arms, through shocks 
from supemovae or the violent winds of OB stars, and changes its form back-and-forth from the 
diffuse ISM to molecular clouds. Roughly half of the mass of the gas is in clouds or in the diffuse 
ISM at any given time, but the volume of the diffiise ISM is vastly larger because of its very low 


Mathis, John S. 

Our knowledge of grains comes from observations, and we can conveniently divide the whole 
continuum of possible sites into just three different principal regions on the basis of these 
observations. These three kinds of dust will be called (a) the "diffuse" dust, meaning the grains in 
the low-density ISM which occupies most of the volume near the plane in our Galaxy; (b) the 
"outer-cloud dust," close enough of the edge of the cloud for us to observe the dust in the optical 
and even ultraviolet parts of the spectrum; and (c) the "inner-cloud dust," which is located within 
the inner regions of very cold, dark, and dense molecular clouds. All of these observations are 
made by comparing a star of a known spectral type seen through the dust to one of the same 
spectral type which is relatively unobscured. In practice, hot stars are far better than cool because 
their intrinsic spectrum in not so sensitive to their precise temperature (or, equivalently, spectral 

Observationally, inner-cloud dust is distinguished by showing absorption bands of molecular 
ice (starting with water and ammonia, but in some cases including CO, methanol, and many 
others). The column density of dust is so large that at present observations can only be made in 
the near-infrared (NIK) region of the spectrum, and we have no information regarding the 
extinction law for wavelengths shorter than 1 |im or so. Outer-cloud and diffuse dust are smdied 
down to wavelengths as short as 0.10 ^m (with the Copernicus satellite) and commonly to 0.12 
|im (with the International Ultraviolet Explorer). 

It is convenient, although not ideal, to describe the wavelength dependence of the extinction 
by the ratio relative to the visual extinction, A(X)/A(V). The extinction laws of diffuse and outer- 
cloud dust form a continuous progression which can be characterized by a single parameter. 
Because of tradition, this parameter is taken to be the so-called "total-to-selective extinction ratio," 
Rv [= A(V)/E(B-V)]. The average value of Ry in the diffuse ISM is 3.1. The largest values of 
Rv are about 5.5, found in the outer parts of the Orion and Ophiucus molecular clouds. I will 
rather arbitrarily define diffuse dust as having Ry ^ 3.4, and outer-cloud dust as Ry >3.4. 

B. Summary of Observations of Interstellar Dust 

The observational properties of interstellar dust are well summarized by papers at the lAU 
Symposium 135, "Interstellar Dust," held at Santa Clara University in August, 1988 (Allamandola 
and Tielens, 1989). There are many papers discussing various aspects of the ISM in general in the 
volume Interstellar Processes (Hollenbach and Thronson 1987). In particular, there is an 
excellent summary of the composition and forms of interstellar dust by Tielens and Allamandola 

Continuous extinction. The continuous extinction, observed from about 5 |im to 0.1 |im, is the 
most valuable information available regarding the properties of interstellar dust The extinction is 


Mathis, John S. 

remarkably free from spectral features, which is one of the largest problems in determining the 
nature of the constituents and size distributions. As we will see, the extinction laws for outer- 
cloud and diffuse dust form a regular progression, and those of inner-cloud dust are presentiy 
unknown because of the large extinction of the starlight in those clouds. 

Spectral features in extinction: There are spectral absorption features which provide insight to the 
nature of grains. For outer-cloud and diffuse dust, the strongest feature in the entire spectrum (in 
the sense of requiring the largest oscillator strength for a given column density of carriers) is the 
A2175 A "bump," which reaches an equivalent width of 50 times Lyman-a. 

Some 29 other "diffuse interstellar bands," all in the optical part of the spectrum (Herbig 
1988), are quite variable in strength relative to A(V). They can be placed in at least three groups, 
among which there is a good correlation (Krelowski and Walker 1987, Chlewicki et al. 1987, 
Josafatsson and Snow 1987). Apparently the bands at 4430, 5780, and 6284 A are not produced 
by coatings on the surface of aligned grains, since there is no polarization in these bands in the 
spectra of two polarized stars (Martin and Angel 1975). They fall in two of the three groups into 
which the bands have been placed. This exclusion is important, since the strength of the 
polarization is so great in some cases that most or all large grains must be aligned. 

There are strong bands at 9.7 |im and 20 ^im, both ascribed to the Si=0 stretch in silicates. 
This feature is found in comets as well, with a somewhat different profile which indicates that 
cometary silicates have been partially annealed. There is a weak feature at 3.4 jim ascribed to 
aliphatic hydrocarbons in the spectrum of the supergiant IRS 7 near the galactic center (Butchart et 
al. 1986). It might arise from mantles on grains, as we shall discuss at some length later. 

For inner-cloud dust, there are many spectral features of molecular ices. These have been 
recently reviewed by Teilens and Allamandola (1989). The lack of these features in diffuse and 
outer-cloud dust suggests that there is a fundamental difference in the properties we expect of the 
grains for the respective regions. 

Emission spectral features: Much of research on dust in the past few years has been concerned 
with the "Unidentified Infrared Bands" (UIBs) which occur at 3.3, 3.4, 6.2, 7.7, and 11.4 ^m, 
with weaker ones at several other wavelengths. In the diffuse ISM they account for 10 - 20% of 
the energy radiated by warm dust. They are found in a wide variety of objects, with the common 
feature that all such objects have carbon-rich or ordinary interstellar dust present, as opposed to the 
circumsteUar dust surrounding oxygen-rich stars. Their origin is not completely clear, but both 
their spectra, relative strengths, and to some extent their variations from place to place are well 
explained by the hypothesis that they are produced by polycyclic aromatic hydrocarbons (PAHs) 
(Leger and Puget 1984, Allamandola, Tielens, and Barker 1985), probably ionized in the diffuse 


Mathis, John S. 

Polarization: Interstellar polarization provides another integral of cross- sections of grains over the 
size distribution (see Hildebrand 1988). Unfortunately, the integral also involves the degree of 
alignment of grains of various sizes as well as their cross-sections, and there is no very definitive 
theory of alignment (Hildebrand 1987). The principal features pertaining to interstellar 
polarization are: (a) Only the larger grains are aligned. Groimd-based observations of intersteUar 
polarization reach a maxiTnnm at a mean wavelength of 0.55 ^m, while the extinction law is much 
larger in the ultraviolet (especially at 0.22 |im and for X, < 0.17 |im). (b) All present theories of 
interstellar grains can explain the observed polarization law, if large grains are aligned but small 
ones are not, because the polarization cross-sections for cylinders, believed to pertain to elongated 
grains, mimic the observed polarization law (Mathis 1986). (c) The polarization of the 9.7 and 19 
pm silicate bands in the heavily obscured object "BN" in the Orion molecular cloud might provide 
a powerful diagnostic for grains (Hildebrand 1988); see below in § IV. (d) Grains are aligned 
quite efficiently, even within dense molecular clouds. The polarization of the emitted radiation 
from deep within the Orion molecular cloud or near the galactic center (Dragovan 1986; Werner et 
al. 1988; Hildebrand 1989, Hildebrand et al. 1989) shows that the grains are very well aligned, 
but the physical conditions in these regions should strongly militate against alignment. 
Depletions of the elements in the ISM: Observations show that several elements (Fe, Mg, Si, Al, 
and others) are heavily depleted in the gas phase of the ISM (see Jenkins 1987, 1988 for reviews). 
Roughly 90% of these elements are depleted in the diffuse ISM and 99% in the very dense regions 
(see Joseph 1988 for a discussion of correlations). Unfortunately, the depletions of the abundant 
elements C, N, and O are especially difficult to determine because their abundant ions do not 
happen to have resonance lines with oscillator strengths suitable for abundance analyses, and are 
consistent with the mean depletion of these elements ranging all of the way from almost none up to 
90%. Carbon is an especially important element for interstellar dust, and there is only one 
determination of the column density of O, its most important ion within regions where hydrogen 
is mainly atomic. Towards the star 5 Sco, Hobbs, York, and Oegerle (1982) found that about 
30% of the cosmic carbon is in the gas phase. This number is probably consistent with the 
uncertain abundance of gaseous CO/H2 in molecular clouds (Watson et al. 1985). Models of 
grains are, therefore, constrained to use no more than about 70% of the cosmic C. 
Other diagnostics: The far-infrared emission fixim warm grains serves as some diagnostic of 
grains, but not as a very direct one because it also involves the heating of the grains and 
temperature excursions of small grains heated by a single photon. The recent theory of Chlewicki 
and Laureijs (1988) has used the lE^S measurements of clouds to make the very interesting 
suggestion that metallic Fe might be present in grains. There are also measurements of the X-ray 
haloes around grains (Mauche and Gorenstein 1986, and ref. therein) which provide some 
information regarding the spatial distribution of heavy elements in the grains. 


Mathis, John S. 

We next consider at some length the information provided by the best-determined of these 
diagnostics, the interstellar extinction law. 

n. Interstellar Extinction and Its Relevance to Cometary Dust 
We must recall that, by definition, we can only observe the interstellar extinction law for 
diffuse and outer-cloud dust, while comets are undoubtedly foimed deep within molecular clouds. 
However, the observed extinction laws have direct relevance to the processes by which dust is 
modified within dense regions. 

A. Systematics of Observed Extinction Laws 
Very recently, it has recently become clear that there is a very surprising degree of regularity 
among the extinction laws in various lines of sight [Cardelli, Clayton, and Mathis 1988, 1989 
(CC!M)]. Figure 1 shows the observed extinction laws of many lines of sight, expressed as 
A(X)/A(V), plotted against 1/Rv, for several values of X ranging from the red to almost the limit of 
the lUE spacecraft (1200 A). There are good linear relationships in each case, so that clearly there 
is an excellent relation between the optical extinction law (as expressed by Ry) and the other 
wavelengths (including ultraviolet). The "mean" interstellar extinction laws of Savage and Mathis 
(1979) or of Seaton (1979) refer to the diffuse ISM, where Ry = 3.1, but Fig. 1 shows that they 
are not very appropriate for values of Ry which differ greatly from that value. 

CCM has fitted the slopes of the various A(X)/A(V) - Ryl relations, examples of which are 
shown in Figure 1, by an analytic formula which represents the mean extinction law as a function 

2 - 

1 1 — 







£^ NCC2Z44 


O MD 29647 



+ SM79 



V* A 


• • 






• f^« 



• • 



• • 



** •• — 

• • •»•• 





, 1 

. " 




Fig. 1. — The observations of A{X)IA{'V), 
plotted against IfRy, where Ry = 
A(V)/E(B-V) (from Cardelli, Clayton, and 
Mathis 1989). An refers to 1200 A, A22 
to 2175 A, A28 to 2800 A, andAjo to 
7000 A (the standard R filter). The 
observational values of black dots are from 
Fitzpatrick and Massa( 1988). The 
regularity of the observations, and the 
scatter about the mean relationship, is 
shown. The scatter is greater than the 
observational errors, and is therefore real. 


Mathis, John S. 

of Rv- Fig. 2, from CCM, shows the results for three values of Ry: (a) from using the formula, 
and (b) observations of actual stars with those values. The central panel is about as discrepant as 
actual observations are from the mean relationship. The dispersion of individual extinction laws 
around that mean law is shown in Rg. 1 (from the spread in the individual observed points) and in 
Fig. 2 (as error bars). The lowest set of curves plotted in Fig. 2 are for Herschel 36, an exciting 
star of M8 and considered to have very "peculiar" extinction. Rather, it has a rather peculiar value 
of Rv (= 5.3), but a "normal" extinction law for that value of Ry. Note, however, that there are 
real deviations from the mean extinction law for any particular values of Rv- These deviations are 
especially large at 1200 A, where the mean deviation of A(X)/A(V) from the mean extinction law is 
about 0.3. 

Fig. 2. — Three cases of a mean 
extinction law, obtained by fitting the 
slopes of the A(X)IA(V) - R\rJ 
relationship (see Fig. l)byan analytic 
formula, and actual extinction laws of 
stars with the appropriate values ofRy. 
The middle panel shows about as large of 
a discrepancy as exists in the data of 
Fitzpatrick and Massa (1988). 

The implications of the information contained in Figures 1 and 2, as regards the nature and 
evolution of grains, are considerable. It seems that the processes which modify the size 
distributions and possibly the compositions of interstellar grains operate on all sizes 
simultaneouslv . It is very conceivable that some lines of sight the small grains would be heavily 
modified, while the larger grains, and therefore the optical extinction, were relatively untouched. 
However, in practice such is not the case (at least among the lines of sight we have investigated); 
if the short- wavelength extinction law is rising relatively steeply with V^, the value of Rv is 
always relatively small, and so on. How can this fact be understood? Apparentiy the processes 
which modify the grains act over the entire size range. These processes must be quite efficient as 
well in order to produce such regular extinction laws for diverse regions and conditions (the stars 
plotted in Figures 1 and 2 are found in various directions aU over the Galaxy). The overall 


Mathis, John S. 

regularity of the extinction laws is that correlations of one extinction or color ratio with another are 
probably not indicative of any causal relation between the quantities; they are probably both 
responding to general processes. 

The most efficient ways of changing grain distributions are by grain-grain collisions 
(shattering and coagulation) or by shocks (which can destroy grains in a variety of ways: see Seab 
1987 for a review). It is "easy" to show that such processes as grain-grain interactions are quite 
inefficient in the rather low densities (nn = 10^ cm-3?) in these outer regions of clouds, but the 
regularity of extinction laws suggests otherwise. The implications for cometary grains are that 
these interactions will probably continue to be efficient at even higher densities than can be 
observed directly, so that there might well be an unexpected uniformity of conditions of cometary 
grains. The grains seem to be modified quite efficiently as they progress from the outer to the 
inner parts of the clouds. 

B. The X2175 Bump 
One of the most mysterious features of interstellar dust is the bimip (see Draine 1989 for a 
review). It has recently been discussed in some detail (Fitzpatrick and Massa 1986; Cardelli and 
Savage 1988, and refs. therein). Its puzzling properties are that its central wavelength, Xo, is 
almost the same for various lines of sight, but there definitely is some variation (see especially 
Cardelli and Savage 1988). Its width varies considerably, in a seemingly random pattem. Its 
strength is rather well correlated with Ry'^ (CCM). It is so strong that its carrier must be very 
abundant, almost surely C, N, or O (in my opinion, but see Duley and Williams 1988). It is one 
feature that most theories of grains do not try to model in great detail, although Hecht (1987) has a 
rather plausible explanation for many of the observed phenomena. 

in. Present Theories of Grains 
There are at present several theories of dust which attempt to explain the observational facts 
outlined above (plus a great deal which are not mention there). Several which have been recently 
discussed are: (a) the silicate-core/organic refractory mantle theory of Greenberg (Hong and 
Greenberg 1980; Greenberg 1989); (b) the bare (no mantles) silicate/graphite theory of Mathis, 
Rumpl, and Nordsieck (1977), often called "MRN," but considerably improved by Draine and Lee 
(1984; DL); (c) a recent theory of Mathis and Whiffen (1989; MW) of composite grains, in which 
bare grains containing silicates, amorphous carbon, and graphite are proposed; (d) silicate core 
grains, with hydrogenated amorphous carbon clusters upon their surfaces (Duley and Williams 
1988, Duley, Williams, and Jones 1989); and (e) a mixture of silicate-core/organic- refractory- 
mantle particles, similar to Greenberg/Hong, but with metallic iron to provide the observed IRAS 
intensities (Chlewicki and Laureijs 1988). In addition, there are other theories similar to the 


Mathis, John S. 

above, but not derived from them. A theory quite different from the others is that of biota (Hoyle 
and Wickramasinghe 1984; see Jabbir et al 1986 for refs. and a late version of the theory). 

Table 1 summarizes some of the salient features of most of the "complete" models of grains, 
meaning that the entire spectral range is addressed by the model. There are models specific for the 
bump which I have not included. I note that the most popular explanation of the bump, graphite 
(shown by Table 1 to be almost unanimous), was originated by Hoyle and Wickramasinghe 
(1962) in the form of a prediction. 

Common features of grain theories: All of the theories listed in Table 1 claim to be able to fit the 
observed extinction and polarization laws for the diffuse ISM. This fact illustrates the fundamental 
problem of understanding grains: a lack of uniqueness, imposed by the fact that only integral 
properties of the size distributions are observed. At some level we have to be content to judge 
theories on the basis of their plausibilities rather than from completely objective criteria, such as 
predictions which are absolutely incompatible with the observations. However, detailed 
predictions of the extinction law, ranging from 1000 \ixn to 0.1 |im, might also be good 
discriminants of grain theories, but have only been applied to DL, MW, and DuleyAVilliams. 
While the theories in general can explain the mean extinction law for the diffuse ISM, some of 
them might have difficulty making the differences in the extinction laws between diffuse- and 
outer-cloud dust plausible. There are rather powerful constraints imposed upon the models if one 
requires both extinction laws to be fitted by similar amounts of materials, including carbon in the 
various specific forms (graphitic and amorphous). There are possibly other tests (see § IV C 
below) which will discriminate among the various models, or even eliminate aU of them. 

One reason I am not emphasizing the biological theory of grains is that it requires more 
phosphoms than the solar abundance (Whittet 1984, Duley 1984), although Hoyle and 
Wickramasinghe (1984) claim that their theory can accommodate the P abundance within a factor 
of 1.5 if favorable assumptions are made. Observations of depletions show that phosphorus is 
about half in the gas phase in the diffuse ISM (Jenkins, Savage, and Spitzer 1986), making the 
problem even more acute. 

Fitting the entire range of interstellar extinction requires a very broad distribution of sizes, 
which is one property shared by all theories. The observations of the silicate bands at both 9.7 ^m 
and 20 pm almost require the presence of sihcates. The breadth and lack of structure in these 
bands in both the Orion Nebula emission and the absorption in various objects, as well as 
emission firom cool, oxygen-rich stars producing and ejecting circumstellar dust, indicate that the 
silicates are amorphous in nature. It is widely assumed that amorphous silicates present in grains 
contain essentially all of the silicon. Tielens and Allamandola (1987) have pointed out that 


Mathis, John S. 


Brief tide of theory 
(and authors') 


Core-mantle -i- iron 
(Chlewicki, Laureijs) 

Draine-Lee (or MRN) 

Composite grains 
(Mathis, Whiffen) 

Fractals (Wright) 

IRAS-compatible dust 


(Hoyle, Wickramasinghe) 

of Grains 

Silicate cores/organic 
refractory mandes; also 
small silicates, graphite 

Similar as Greenberg, with 
small metallic iron particles, 

Bare silicate, graphite 

Silicate cores, hydrogenated 
amor. C clusters, small silicates 

Silicates, amor. C, small graphite 
in the same grains (with free 
small graphite, possibly silicates) 

Grains with fractional 
dimensions (from growth) 

Silicates, amorphous C 
Graphite, bacteria, diatoms 

Size distr. 
Qf Grain? 

of Bump 












OH" ions 
near Si 















a References: 1, Hong and Greenberg (1980); 2, Chlewicki and Greenberg (1989); 3, Chlewicki 
and Laureijs (1988); 4, Madiis, Rumpl, and Nordsieck (1977; MRN); 5, Draine and Lee (1984; 
DL); 6, Duley and Williams (1988), Duley, Wdliams, and Jones (1989); 7, Madiis and Whiffen 
(1989); 8, Wright (1987); 9, Rowan-Robinson (1986); 10, Hoyle and Wickramasinghe (1984), 
Jabbiretal. (1986). 

^ "Flat" is really an exponential-type distribution which is plotted in Fig. 3 of this paper, 
c "Small" graphite means that the radii of all graphite particles are < 0.22/(27c) |im, or < 0.005 
}im. Below this size, the extinction properties do not change with size. 


Mathis, John S. 

the strength of the 9.7 \im feature is such that 1.5 times the solar silicon abundance is required to 
produce the 9.7 jim feature from amorphous silicates alone. However, crystalline sihcates have 
band strengths over twice as large, and it is perhaps plausible that the silicates in grains are 
somewhat ordered. Theories assume that all of the silicon is contained in the silicates. 

Another feature which all present grain theories share is leaving at least some observations 
unexplained, partly because the authors feel that our present knowledge of optical constants or 
other physical properties does now warrant detailed modelling. These observations include the 
bump, which is likely to be produced by a material something like graphite or hydrogenated 
amorphous carbon (Hecht 1987), but which has not been positively identified. The origin of the 
UIBs is by no means settled; although PAHs seem to be very plausible candidates, the bands have 
other possible origins, and have been referred to as the "overidentified infrared bands." Except for 
Chlewicki/Laureijs, the theories in Table 1 do not try to make detailed predictions of the origins of 
the carrier(s) of these emission bands. 

There are also surely small amounts of other minerals, such as silicon carbide, present in 
grains, since carbon stars show an emission feature which arises from this material and it is quite 
refractory. Other unexplained observations include the origin of the "diffuse interstellar bands," 
which may (or may not) have interesting implications as to the nature of grains. 
Differences among the grain models: Obviously, there are differences among each of the models. 
Possibly the broadest difference, as regards the prediction of the nature of cometary grains, is 
whether or not the silicates in diffuse and outer-cloud dust are coated with tough organic refractory 
mantles, as envisioned in theories of Greenberg and Chlewicki/Laureijs, but not the others. The 
size distributions are also quite different, as shown by Figure 3. I shall discuss these differences 
in § TV below. 

Fig. 3. — The size distributions in the 
theories of corelmantle grains (Hong and 
Greenberg 1980; the actual size distribution 
shown here was taken from a preprint by 
Chlewicki and Greenberg) and the MRN- 
Draine and Lee (1984) theories. Small 
graphite and silicate particles, of unspecified 
size distributions except that the sizes are < 
0.05 /Jm or so have been omitted from the 
corelmantle theory. 


T ' 












*v ^^ 

= 103 




\ x. 












10- ■« 

1 1 





Mathis, John S. 

Another apparent difference between the core/mantle theory and bare-grain theories might be 
the presence of elemental carbon in forms other than the small graphite needed for providing the 
bump, which is only about 25% of the solar abundance. There have been small diamonds found 
in meteorites (Lewis et al. 1987), but theories do not include them explicidy because their 
abundance is too low. Besides the graphite needed to produce the bump, the MW theory has most 
of the carbon in amorphous solid phase and Greenberg's theory has it is the organic refractory 
mandes. However, the situation is not so clear, since there may have been enough processing of 
the organic refraaory mantles by cosmic rays and ultraviolet radiation to convert them essentially 
to amorphous carbon. 

A. Are there manties on diffuse- and outer-cloud dust grains? 

Should we expea interstellar silicate grains in a cometary nucleus to have manties of organic 
refractory materials? In this section I examine the evidence for those manties on grains outside of 
the dense cores of molecular clouds. It is clear that the cometary grains will acquire manties of 
molecular ices deep within the molecular cloud, and it is not those manties we are debating. 

The extinction law for the outer-cloud dust strongly indicates that the grains there are larger 
than the average for the diffuse ISM. There are two aspeas of the mantie question I would like to 
address in turn: (a) Do the grains in outer-cloud dust have their increased sizes because of having 
acquired mandes? (b) Is there spectroscopic evidence for manties in the diffuse ISM? 

Do the grains in outer-cloud dust have their increased sizes because of having acquired 
manties? If grains in the diffuse ISM have manties, one might expect any increases in the mean 
grain sizes to come from accretion of more manties. However, there is good evidence (it seems to 
me) that grains are larger in the outer-cloud regions mainly because of coagulation rather than 
accretion of manties. The extra depletion of the refractory elements which is observed in dense 
regions (the gas-phase Fe going from 90% depleted to 99% in comparing diffuse ISM dust with 
outer-cloud dust) will increase the grain volume by only about 5%. Any substantial increase in 
grain volume must come from accretion of the abundant and relatively undepleted elements N or 
O, in combination with others, since C is already mostiy depleted in the diffuse ISM. 

The answer to the question comes from considering the absolute amount of extinction, 
A(k)/N(H), rather than the extinction law. If the grains in the outer regions of dense clouds are 
large, relative to those in the diffuse ISM, because they accrete materials, the extinction per H 
nucleus [i.e., N(H I)h-2N(H2)] must increase. If the grains grow by coagulation, A(X)/N(H) car. 
either increase (because there is an increase of extinction efficiency by increasing the sizes of 
certain grains because of collective effects, mainly with the increase of scattering) or decrease 
(because the material in the center of the grain is somewhat shielded from absorbing the radiation). 


Mathis, John S. 

The results of observations are discussed briefly in CCM. There are two steps in the 

(a) Lines of sight with a large value of Ry have a smaller total value of the extinction, 
integrated over the wavenmnber x (= \fk). One can see from Figure 2 that there will be a 
considerably smaller value for J A(k)/A(y) dx for large Ry (the lower curves) than for small Ry- 
In CCM we found that 

J A(X)/A(V) dx = (-4.3 + 79.1 Ryl) ^im-1 , (1) 

where the integral extends fix)m the infrared (0.3 \ixa.-^) out to the largest wavenumbers observable 
with the Copernicus satellite, x = 10 |im'l. The second term is dominant, and the total integral is 
roughly inversely proportional to Ry^ 

(b) To determine the total integrated extinction, we must relate the visual extinction in 
equation (1), to the total H column density, which is not easily determined observationally. The 
problem is that the N(H I) must be found from the highly saturated Ly-a profile, and the N(H2) 
from the Copernicus observations of the Lyman and Werner bands. However, for two weU- 
observed lines of sight (towards NU Ori and p Oph) in the outer regions of dense clouds, the 
results are unambiguous: the quantity A(V)/N(H) is smaller than in the diffuse ISM. There is no 
measurement of the H2 column density towards NU Ori, so its value of A(V)/N(H) is an upper 
limit These stars have large values of Ry. The total integrated extinction values of p Oph and 
NU Ori are 0.73 and < 0.43 of the value in the diffuse ISM (Bohlin, Savage, and Drake 1978), 

There are other stars for which estimates of N(H I) and N(H2) are available, but not so 
accurately determined as for p Oph (de Boer et al. 1986), or so extreme an A(V)/N(H I) value as 
for NU Ori (Shull and van Steenberg 1985). We reserve judgement on the other stars. 

From this data on extinction per H nucleus one concludes that grains are larger in outer-cloud 
dust because of coagulation rather than accretion. Similar results were derived by Jura (1980) and 
Mathis and WaUenhorst (1981) on the basis of A(V)/N(H) alone. The integrated value of the 
extinction makes the case stronger. It is difficult to have even coagulation produce enough 
decrease in extinction cross-section per H (see MW), and any accretion of mantles makes this task 
even worse. It is very difficult to imagine that there is extensive accretion onto grains until they 
are deeper in clouds than can be observed at present in the ultraviolet part of the spectrum. 

Is there spectroscopic evidence for mandes in the diffuse ISM? If there are mantles on grains 
in the diffuse ISM, one would expect to find some spectroscopic evidence for them. In fact, there 
is such evidence: the 3.4 ^m absorption feature about 30% deep seen in the spectrum of IRS 7, a 
supergiant seen near the galactic center (Butchart et al. 1986) with an extinction, A(V), of = 25 


Mathis, John S. 

mag. The presence of that feature has been used as a justification for the existence of organic 
refractory mantles, since the 3.4 |im feature is an indication of an aliphatic C-H stretch absorption 
and the 3.4 |im feature is found in the spectrum of organic refractory mandes (Schutte 1988). 
Figure 4 (from Schutte 1988) shows the 2.8 - 3.8 |im spectrum of IRS 7 from Butchart et al. 
(1986). It also shows a spectrum of photolyzed ice, which fits the 3. 1 |im portion of the IRS 7 
spectrum well, and some "char," or a mixed mixture of paraffins, olefins, and other carbonaceous 
materials, which fit the 3.4 pm spectrum well. These might be present as mantles on grains, 
although their presence along the line of sight is no indication that they are coated onto grains, and 
the 3.0 |im "ice" band found in the spectrum is surely not present in even outer-cloud dust. Let us 
examine the evidence for materials of this type, manties or not, in local dusL 







» .*^ 

\ / J^^'*' 

\v » ♦ 

\ /-'" 


\ '' / * 


-'\ / • 



M • - 


,V-^ .M,** 




/ HjO/CO/KH-ZO, . ■ 



^ ^— ^ C.32:0.23;0.J3C.32.' 


* 'f'hv, and vx 



' , , . 

X O/m) 

Fig. 4. — The spectrum of IRS 7 near the 
galactic center (spectrum from Butchart et 
al. 1986; figure reproduced from Schutte 
1988). The 3 jjon band is compared to the 
ultraviolet-irradiated ice sample 
H20/CO/NH3/02 = 032/023/0.13/032. 
The solid line is the spectrum of "charred" 
carbonaceous residue. Compare with the 
spectrum of absorptions through "local" 
dust, towards Vi Cyg 12 (Fig. 5, below). 

The most likely local candidate for having the 3.4 pn feature is VI Cyg 12, a B5 Ia+ star with 
E(B-V) = 3.31 (Hunqjhreys 1978; Serkowski 1965). It is the best object for studying local 
extinction because it is relatively nearby, is about as luminous as any other star in the Galaxy (My 
= -10!), and is very heavily reddened. Its spectrum in the 3-Mm region (Gillett et al. 1975) is 
shown in Figure 5. The A(V) of VI Cyg 12 is 10 mag; tiiat of IRS 7 is 25 mag. Figure 5 
indicates the strengths of the absorptions at 3.1 and 3.4 ^rni which one would expect in VI Cyg 
12, if the dust were of the same nature as towards IRS 7. To the resolution of the Gillett et al. 
spectrum, there is no trace of the 3.4 |im absorption feature in VI Cyg 12. However, at the 
discussion following a presentation at the lAU Symposium 135, Interstellar Dust, Dr. D. C. M. 
Whittet stated that VI Cyg 12 has x(3.4 nm)/A(V) of about 1/4 that of IRS 7. Thus, it seems that 
the absorption by any organic refractory manties are locally considerably less, per A(V), than 
towards the galactic center. 


Mathis, John S. 




Fig. 5. — The observed spectrum of 
the local star VI Cyg No. 12 (Gillett et 
al.l975). This is a blow-up of the 
portion 2.9 - 3J fJm, with the 
wavelength and intensity scales 
superimposed. The slanted, irregular 
solid line shows the observations. The 
expected depth of the absorption is 
shown as the dashed line, on the 
assumption that the absorption per 
A(V) is the same as towards IRS 7 near 
the galactic center (Fig. 4). 

But, how typical is the line of sight towards VI Cyg 12 as regards the diffuse ISM? 
According to one criterion, that of Ry, the answer is "very typical"; Ry = 2.99 ± 0.15 (Rieke 
1974), as compared to the mean of 3.1 for the diffuse ISM. But according to another criterion, 
that of A(V) per kpc, the answer is "very atypical." The distance of the VI Cyg association is 1.5 
kpc (Schulte 1958), and the other stars in the association are reddened considerably less than No. 
12. Stars No. 5 and 12 have E(B-V) = 2.00 and 3.31, respectively, which represents a difference 
of A(V) of 4 mag. The two stars are separated by an angle of 4.5'. This corresponds to a 
projected linear distance of 2 pc if d = 1 .5 kpc, or an actual separation of perhaps 4 pc. The 
excess extinction of No. 12, relative to the other stars in the same association, is presumably 
caused by dust in the vicinity of the cluster, so A(V)/d > 1 mag pc^l (!). Thus, the local density of 
the dust seen to No. 12 must be = 500 times the typical value for the diffuse ISM (about 2 visual 
mag per kpc). It would seem reasonable that such an extreme value of density would promote 
mantle growth, so I would suspect that the mean absorption strength of the 3.4 ^im band in the 
diffuse ISM is considerablv below even the low value found in VI Cyg 12. Such a low value 
makes the volume of the organic refractory manties rather unimportant as regards the extinction 
properties of the grains. There might be enough mantie on grains to influence their surface 
properties considerably, in particular how well they stick together during collisions or how well 
they form molecules on their surface. 

It is perhaps not too surprising that dust along the entire path length to the galactic center 
would be somewhat different, perhaps very different, from local dust. There is a definite 
composition gradient in the Galaxy (Shaver et al. 1983) which indicates an increase in N(0)/N(H) 
of about a factor of two. Probably the abundance of other elements is higher by similar factors. 
Furthermore, the mean density of the ISM also increases by about a factor of four (Giisten and 


Mathis, John S. 

Mezger 1982). Both factors favor the growth of mantles. The peculiarity of the dust towards IRC 
7 is emphasized by the 3.1 \mi absorption, which is not of the same shape as the usual 3.07 |im 
ice band, but is somewhat stronger than the 3.4 ^m band which might show the existence of 

Perhaps it is merely semantics to discuss whether or not there are "organic refractory" mantles 
in space, meaning materials similar to those which have been produced in the laboratory. As 
organic refractory material is processed by radiation in the laboratory (Schutte 1988), it steadily 
loses O and N relative to C. If the processing by the cosmic rays and ultraviolet radiation is severe 
enough, the mantles might become amorphous carbon for all practical purposes, and just about all 
of the authors listed in Table 1 can declare a victory to their respective funding agencies. 

The size distribution of grains: Another major difference in the present core/mantle theories 
and most bare-grain theories is the size distribution, at least the grains responsible for the visual 
extinction. This distribution is almost a truncated step-function for the core/mantle theory, with a 
rather indefinite number of very small grains to provide the ultraviolet extinction, and a power-law 
for most others. It seems to me that the power-law size distribution is the most plausible part of 
the MRN theory, for both theoretical and observational reasons. Biermann and Harwit (1979) 
showed that collisional processes should establish a power law with the exponent in the range 
considered by MRN and DL (n(a) proportional to a-3-5). Radar studies of particles in Saturn's 
rings (Cuzzi et al. 1984), shattered terrestrial rocks and gravel beds (Hartmann 1969), and of 
cometary dust (MacDonnell et al., this volume) are all heavily peaked in numbers towards small 
sizes, but with much or most of the mass in the larger grains. These distributions are all produced 
by collisional processes among the particles. Thus, if there are collisional processes among the 
grains, it would seem difficult to avoid an approximate power-law distribution. 

The fact that grains in outer-cloud dust have apparentiy grown from those in the diffuse ISM 
by coagulation rather than by accretion implies to me that they have, in fact, grown by collisional 
processes. While I doubt that it is precisely a power-law size distribution which is established, I 
would find it difficult to believe that the size distribution is very flat. 

The coagulation of grains in outer-cloud dust is a strong motivation for the composite-grain 
theory of MW. Grains are originally ejected from stellar atmospheres, novae, or supemovae with 
a certain specific composition which small particles can retain (and do, as shown by isotopic 
anomalies in meteorites). However, there are "large" (a = 0.1 jim) grains which provide the 
extinction in the optical part of the spectrum, and which observations suggest are rather heavily 
modified by shattering and coagulation as they go into and out of dense clouds. It is difficult to 
imagine that the component small particles fit together without voids or holes within them. The 
interplanetary dust particles (IDPs) collected by Brownlee and his coworkers (see Brownlee 1987 
for a review of the solar system-interstellar dust connection) show the kind of open structure 



Mathis, John S. 

which partially suggested these ideas. However, there has also been chemical processing of the 
IDPs, as shown by their occasionally homogeneous chemical compositions within grains, but 
varying compositions from grain to grain (Brownlee, this workshop; Mackdnnon and Reitmeijer 

In fact, I strongly suspect that the composite theory is too simple. There is a constant transfer 
of refractory elements back-and-forth from grains to the gas phase. Grains are destroyed in 
shocks, as shown by the low depletions in high- velocity gas (see Jenkins 1987 for several 
references) relative to the diffuse ISM, and there is a more severe depletion of refractory elements 
in dense clouds. For computational purposes, it was assumed in the composite model that the 
individual particles in a grain are either silicates or elemental carbon. Although this assumption is 
necessary to compute indices of refraction, it really seems dubious: the refractory elements, 
carbon as well as sihcon, etc., are deposited onto grains together. The individual particles in the 
composite model should probably consist of a core of perhaps pure material, produced in a star, 
and a mantie of refractory elements (not organic refractories) surrounding the core. There might 
well be a coating of organic refractory on the surfaces of grains, but (it seems to me) much less 
than envisioned by Greenberg. The existence of the "Stardust" core to grains- very refractory 
minerals which have survived from their formation in stellar atmospheres or supernova explosions 
throughout their entire lifetime in the ISM- is heavily supported by the isotopic anomalies which 
are found in meteorites (e.g., Clayton 1988, Tang and Anders 1988). However, I beUeve that 
most of the grain volume must have been heavily processed by processes in the diffuse ISM. 

If there are few grain-grain collisions within space, after the grains are formed in cool stellar 
atmospheres, then the Greenberg core/mantie theory size distribution becomes much more 
plausible, although Biermann and Harwit (1979) considered collisions within a stellar atmosphere 
to be the dominant mechanism for grain evolution. If grain-grain collisions are negligible, 
chemically homogeneous grains are plausible if the organic refractory mandes produced in dense 
clouds are cleaned off by shocks, as theory (e.g., Draine and Salpeter 1979a,b) suggests. Bare 
grains of homogeneous compositions, as DL (or a similar silicate/amorphous-carbon theory, to 
avoid the problem of making large graphite grains) become plausible as well. 

A good test of the various theories: While there is too much ambiguity in explaining the 
observations of extinction and polarization alone to discriminate among the various theories, there 
is an excellent test which has not yet been applied to the extent that it should be: explaining the 
observations of the 3.1 |im ice band, the 9.7 [ixa and 19 ^m silicate features in the Beckhn- 
Neugebauer (BN) object in the Orion molecular cloud (Lee and Draine 1985; Hildebrand 1988), 
not only as regards the relative extinctions and polarizations, but also the contrast in the 
polarization within and between the bands. Weak absorption features show polarization with the 
same wavelength dependence as extinction. As a resonance becomes stronger, the polarization 


Mathis, John S. 

reaches a maximum at longer wavelengths than extinction, with a shift which is dependent upon 
grain shapes and coatings. The silicates features are strong enough to provide such a diagnostic 
for the environment of the silicates, and such the shift is actually observed in BN. Predictions are 
quite different on the various theories (coated on the core/mantle theory; juxtaposed with carbon on 
the MW theory; bare and alone on the DL theory). Lee and Draine (1985) found a fair fit of MRN 
with the observations, provided that oblate (pancake) grains of about 2:1 size ratio were assumed. 
Martin (1989) showed that his calculations with core/mantie grains produced a negative 
polarization which is not observed. I doubt that the composite grains of MW will be able to pass 
this test either, but one must be confident of the indices of refraction, which are more difficult to 
estimate in the case of composite grains than for homogeneous materials. It has a poorer chance 
than DL, in which case the silicates are bare, but probably a better chance than any theory in which 
the silicates are coated with mantles. There is one potential problem with this test: there is 
evidence for scattering in the surrounding K-L nebula (Werner, Dinerstein, and Capps 1983, and 
refs therein) in the 2 - 5 ^im spectral region, with the scattered light being reflected fixsm both BN 
and fix)m IRc 2 nearby. However, the polarization of BN itself is dominated by the aligned grains 
rather than scattering; the total polarized intensity of the 2 - 5 \ixn radiation on BN itself is very 
much larger than that from the polarization from scattering of the surrounding KL nebula. 

V. The Evolution and the Nature of Interstellar Dust 

In this section I shall summarize my views regarding the evolution of interstellar dust, starting 
with its formation and ending with it being incorporated into a cometary nucleus. 

The formation of grains, and the fact that stellar eiecta are not much like interstellar grains: 
There are many sources of grains: planetary nebulae, novae, supemovae, and, especially, cool 
giant and supergiant stars (see Gehrz 1989 for a review). Just which of these processes are the 
dominant ones need not concern us here — we just need to recognize that in each particular 
source, the grains are formed with a composition which reflects the local gaseous chemical 
composition, and, especially, whether C > O, or the reverse. The reason, of course, is that the 
element which has the lesser abundance by number is almost completely converted into the very 
stable molecule CO, which cannot participate in grain foimation. Observations (Merrill 1977) 
show that O-rich stars form grains of a silicate nature, as shown by the spectra in the 10 and 20 
|J.m region. Grains formed in a carbon-rich enviromnent show a rather featureless optical 
spectrum, except for a band at 1 1.2 |im which is attributed to SiC. Most carbon stars are too cool 
and dusty to be observed in the ultraviolet part of the spectrum, but a feature resembling the bump, 
but displaced to 2300 A, has been found in very carbon-rich, H-poor objects (Greenstein 1981; 


Mathis, John S. 

Hecht et al. 1984). This bump is similar to that expected from amorphous carbon, so it is widely 
assumed that carbon stars produce grains of this form (see also Martin and Rogers 1987). 

Roughly half of the grains come firom each type of source (C- or 0-rich). The materials that 
are injected into the diffuse ISM from the primary sources of grains, then, are not mostly grains 
themselves. Rather, all of the lesser of the elements carbon and oxygen is tied up in gaseous CO, 
which in turn is photodissociated and then photoionized into O and O in most of the ISM (and in 
more highly ionized forms in the H+ regions found in some 10% of the mass of the diffuse ISM). 
Some or most of the remaining C or O is in the form of amorphous C or silicates, respectively. 
Some of the silicon, but apparently not a large fraction (Martin and Rogers 1987), is tied up in the 
refractory SiC from the carbon-star ejecta. 

Processing of the materials within the ISM: Within the ISM, grains must be processed 
extensively by grain-grain collisions, shocks from supemovae and stellar winds, sputtering by the 
hot phase of the ISM, and accretion of atoms and molecules from the gas phase. The lifetime of a 
particular parcel of gas in the Galaxy, as regards being incorporated into a star, is Mgas/Mgas = 
5E9 M© /(3 M© yr 1) > 1 Gyr. The mass contained in the very hot ISM, or in H n regions, is 
quite small as compared to the atomic or molecular phases, each of which contains about half of 
the mass. Since molecular clouds are preferentially found in the spiral arms encountered twice 
each galactic rotation period, or about every 10^ yr, a typical parcel of gas should have been and 
out of a molecular cloud several times during its lifetime. 

Present theories of the effectiveness of various processes, especially of grain destruction 
rates, seem to me to be in conflict with observations (see also Seab 1987). According to theory 
(Draine and Salpeter 1979a,b; Snow and Seab 1983), shocks from supemovae are quite 
destructive of grains, and it is not easy to understand why grains have not been rather completely 
destroyed by these shocks in times of the order of 10^ yr. Observations provide a rather different 
view: except for high-velocity gas (see refs. in Jenkins 1987, p. 550), presumably being affected 
by one or more shocks at the present time, the refractory elements are always highly depleted. 
Near the galactic plane, the mass of the high-velocity gas at any instant is very small in comparison 
to the bulk of the ISM. The almost ubiquitous depletion of refractory elements is especially 
significant in view of the faa that there is a continuous return of gaseous refractory elements into 
the diffuse ISM fir>m stars, especially those of early type and planetary nebulae. Grains are so far 
apart in the diffuse ISM that it is difficult to have the gaseous refractory elements condensed onto 
grains efficientiy. One finds by integrating the theoretical size distributions, or just by the fact that 
the mean free path of radiation is about 1 kpc, that the area of grains is roughly 10"21 cm^ 
(H atom)-^. For the diffuse ISM clouds (n(H) = 30 cm-3) and relative speeds of refractory atoms 
and grains of 100 m/sec (which represents a nonthermal grain-gas relative speed), the time for a 
grain to collide with a grain is > lO^ yr. Within this time, current theories of grain destruction 


Mathis, John S. 

suggest that the grains should be destroyed, and their refractory elements returned to the gas 
phase. Unless we imagine that each individual parcel of gas becomes part of a dense molecular 
cloud several times within this 10^ yr, a significant portion of the refractory elements should be in 
the gas phase. There is an apparent conflict between theory and observation. 

The easiest way for me to understand the observed high depletions is by assuming that grains 
survive the rigors of the ISM far better than the theory would suggest, so they do not readily lose 
the Fe, Al, and other refractory elements into the gas phase. I am not only considering the 
survival of the silicates, but also of the carbon which is invoked to explain the bump and the total 
amount of extinction per H. Having carbonaceous mantles around silicate cores wovild not help 
this situation; if the mantles were destroyed, they would have to re-form in the diffuse ISM to 
provide enough extinction to match observations even if they protected the silicate core. 

There may not be a significant ercoT in theory of destruction of the grains after they are hit by 
a shock of a given speed. The estimate of the mean lifetime of the grains also depends upon the 
frequency with which a parcel of gas is hit by the shock. That time, in turn, is dependent upon the 
interstellar magnetic field and supernova frequency. A larger interstellar magnetic field than is 
commonly assumed (about 5 |J.G instead of 2 |J.G) greatly reduces the volume of space over which 
a hard shock from a supernova propagates (Cox 1988). 

Even if grains are not destroyed by shocks as efficiently as theories suggest, the regularity of 
the observed extinction law is surprising. Draine (1985; see also Volk et al. 1980) has discussed 
the evolution of grains within clouds. The time required for a moderate-sized grain (0.1 \isn) to 
increase its mass significantly is = 7E10 (nn)'^ V3-I yr, where V3 is the relative speed of grains, in 
units of 10^ cm sec"^ For grains in clouds of nn =■ 10^ cm-3 and v = 10 - 100 m sec^^ the time is 
about 10^ - 108 yr. However, there is a high degree of uniformity of the extinction laws in outer- 
cloud dust (CCM). Some of these clouds, such as near the Orion Nebula, presumably have the 
grains emerging from the dense regions. One would expect that in other regions the molecular 
cloud is growing, and the grains near the cloud reflect the nature of the diffuse ISM. If the 
processes by which grains are HKxiified by interstellar processes were not surprisingly efficient, 
there would be a large dispersion in the relative numbers of the smaU to the large particles. 

It is also surprising that the interstellar polarization near dense molecular clouds is fairly 
strong and very regular in outer-cloud dust, so the grains are clearly very well aligned. For 
instance, the polarization around the Cot Aus cloud (Vrba, Coyne, and Tapia 1981) is very regular 
and follows the contours very closely, so either the cloud is quite old or the alignment is very 
efficient The extinction in the cloud shows the expected large values of Ry, so the grains have 
presumably coagulated while spinning. These considerations show that either my theoretical 
notions, or interpretation of the observations, or both, are wrong. I suspect that it is the theoretical 


Mathis, John S. 

Possibly there is sinprisingly good circulation between the outer and inner parts of the cloud. 
If so, either the motions do not carry the grains to the mantle-forming regions (since mantles are 
counterindicated by the extinction/N(H) ratio in the Orion and Ophicius clouds) or the mantles are 
cleaned off very efficiently in the outer parts of the clouds. The circulation might be deep enough 
to reach local densities of 10^ - 106 H2 molecules cm-3, with a correspondingly large grain 
density, so that grain-grain collisions occur rapidly. Such a circulation would suggest that the 
relative speeds of the grains relative to the gas, and to each other, is nonthermal, but instead is 
governed by hydromagnetic waves, in which the relative speeds depend on the charge/mass ratios 
of the various constituents. In this case, the grains and gas will all move rather gentiy with respect 
to each other, and the coagulation can occur with a drastically reduced time scale over estimates for 
the outer regions of the clouds. Since the ISM is never allowed to come very close to mechanical 
equilibrium because of supemovae and violent winds from early-type stars, the presence of 
hydromagnetic waves seems plausible. These conditions will lead to the observed reduced 
extinction cross- section per H nucleus. 

At some point, grains penetrate the regions of the molecular clouds where molecular mantles 
form, and the icy coatings change the grains' properties considerably. Mandes of water, 
ammonia, methanol, formaldehyde, and several other molecules appear in the spectra of stars 
embedded within the clouds. Tielens and Allamandola (1987) give an excellent discussion and 
summary of this aspect of the ISM, with references to many of the claims I make below. 

In the cores of dense molecular clouds, the time scale for molecules to stick to the cold grains 
becomes very short, and on timescales of 10^ yr or less the heavy molecules should freeze out It 
is obvious that the many microwave observations of all sorts of gas-phase molecules from 
molecular clouds show that this theoretical idea is quite wrong. I believe that the correct 
explanation is that of Greenberg and his co-wo±ers — that explosions of the free-radical-rich icy 
mandes, triggered by cosmic ray heating events or grain-grain collisions, return much of the 
frozen ices back into the gas phase, with molecular reactions that contribute to a complex 
chemistry. The free radicals in the mandes must be produced by damage by ultraviolet photolysis 
or by cosmic rays to the otherwise amorphous stable ices. The radiation might arise from the 
excitation of H2 by cosmic rays. The icy reactive mandes may well form the "yellow stuff' which 
is observed in the laboratory, although the efficiency of its formation is low enough so that it may 
not be important even within these dense cloulds. 

Formation of pre-cometarv dust: In the inner parts of molecular clouds, there are apparentiy 
enough "turbulent" motions (the hydromagnetic waves required to maintain relative grain motions 
in the outer parts of the molecular clouds?) to usually prevent the self-gravitation of the cloud to 
produce a collapse. Occasionally such a collapse occurs, possibly triggered by the winds or 
supernova explosions of previously formed early-type stars. There are no measurements of any H 


Mathis, John S. 

column densities deep within molecular clouds, so it is pure speculation to guess that the grains 
continue to coalesce as they accrete icy mantles. The coalescence would imply that there are voids 
or spaces within the grains, since the particles cannot fit together very welt The organic cometary 
material found in P/Halley , such as "CHON," is condensed out of the cloud at this stage. The 
refi:actory solids, the former interstellar grains, would form a loose, weak matrix spread 
throughout the ices, and the evaporation of the ices when the cometary dust is warmed by sunlight 
would disrupt the weak grains rather efficiently. 

There would have to be some additional processing of the frozen grains and ices before they 
accumulated into a full-fledged comet. The energy released in the impacts of the protocometary 
lumps as they coagulated into larger and larger groupings would probably cause some of the ices 
to evaporate or change into crystalline phases, and reactions between the chemically active ices 
would produce orgaiuc molecules as well as heat. These reactions would probably convert a 
considerable fraction of the elementary carbon into organic compounds which are observed in 
P/Halley as "CHONS" (including sulfur). 

Processing of the grains would continue in the outer solar nebulae, where the temperature is 
far higher (> 100 K) than in the cores of dense clouds (= 10 K). The returned cometary materail 
will, of curse, reflect this processing, however, it would be far better at providing insight into the 
nature of the original interstellar grains than any other material in the solar system. 


1. Observations show rather clearly that whatever processes modify the size distributions of 
grains operate efficiently on all sizes. Extinction laws follow a rather regular pattern over the 
entire wavelength range which can be observed (down to X = 0.1 nm). There are deviations of 
individual extinction laws from the mean shape, but those deviations are smaller than the 
deviations of the mean shapes of the extinction laws from dust in the diffuse ISM as compared to 
that in the outer parts of dense clouds (the densest regions in which we can observe the extinction 

2. The solid materials ejected by stars are heavily modified, but not completely destroyed, by 
processing in the ISM. Isotopic anomalies in meteorites show that at least some grains survive 
intact ("Stardust"). The size distributions of interstellar grains are changed many times, though, by 
coagulation within the outer regions of dense clouds. 

3. There are several theories (see Table 1 for references) which can explain at least the gross 
features of the mean extinction and polarization law for the diffuse ISM. Some have been applied 
to the details of the extinction law with fairly good success. None can explain the X2175 "bump" 
completely well. Not many have tested as regards differences in the extinction law from the 


Mathis, John S. 

diffuse ISM to outer-cloud dust. It seems likely to me (Mathis and Whiffen 1989) that grains in the 
observable outer parts of dense clouds are probably fluffy and composed of both carbon and 
silicates within the same grain. They should contain voids or vacuum as they coalesce. 

4. Grains grow in the outer parts of dense clouds mainly by coalescing rather than accreting. 
The small extra fraction of refraaory elements (Fe, Mg, etc.) which they accrete is not a large 
fraction of their volume in the diffuse ISM. The evidence for coalescence rather than accretion is 
that their total extinction cross-section per H atom decreases for the well-observed stars in the 
outer parts of dense clouds. Accretion adds to the grain volume, contrary to observations. The 
decrease in extinction arises from the reduced extinction of the inner parts of the grains. 

5. It seems to me (but not to others!) that there is rather good evidence against the presence of 
considerable amounts of organic refractory manties on diffuse and outer-cloud dust, mainly 
because of the weakness of the 3.4 pm absorption feature in VI C^g 12, the most heavily 
reddened nearby star. The dust cloud seen in front of it is much denser than average (by a factor 
of = 500), so grain manties would be much more likely to form and survive than in the diffuse 
ISM. Evidently organic refractory manties, if they are formed deep within molecular clouds, are 
destroyed by the processes which operate in the outer parts of those clouds. The line of sight 
towards IRS 7 near the galactic center, where organic refractory manties along with ices seem to 
be present, differs from local dust in having a much larger fraction of heavy elements (at least 
oxygen) and gas density, both of which favor the formation of manties. 

6. Deep within molecular clouds, icy manties form. I feel the scenario espoused by Prof 
Greenberg is correct in this case — there is probably processing by cosmic rays, both directiy and 
by the ultraviolet radiation they produce. This processing forms the free radicals in the icy 
manties. Explosions in the chemically reactive manties release heavy molecules into the dense, 
cold gas in the cloud cores. 

7. When the cloud cores coUapse into stars, planets, and comets, the grains, with rather thick 
icy manties, form icy lumps which in turn coalesce into cometaiy nuclei. 

This work has been partially supported by Contract K 957996 from the Jet Propulsion 
Laboratory, for the space Astrophysics Data Analysis Program and by the Graduate School of the 
University of Wisconsin and the Lunar and Planetary Institute, for which I am grateful. I have 
benefitted from preprints and conversations with J. A. CardeUi, B. T. Draine, R. H. Hildebrand, 
P. G. Martin, B. D. Savage, and many others. 


Mathis, John S. 


Allamandola, L. J., and Tielens, A. G. G. M. (eds.) 1989, Interstellar Dust, proceedings of 

Symposium 135 of the International Astronomical Union, held at Santa Clara University, July 

26-30, 1988 (Reidel: Dordrecht). 
AUamandola, L. J., Tielens, A.G.G.M., and Barker, J. R. 1985, Ap. J. (Lett.), 290, L25. 
Biennann, P., and Harwit, M. 1979, Ap. J. (Lett.), 241, L33. 
Bohlin, R. C, Savage, B. D., and Drake, J. F. 1978, Ap. J., 224, 132. 
Bregman, J. D., Campins, H., Witebom, F. C, Wooden, D. H., Rank, D. M., Allamandola, L. 

J., Cohen, M., and Tielens, A. G. G. M. 1987, Astr. Ap., 187, 616. 
Brownlee, D. E. 1987, in in Interstellar Processes, D. J. Hollenbach and H. A. Thronson, eds. 

(Reidel: Dordrecht), p. 513. 
Butchart, L, McFadzean, A. D., Whittet, D. C. B., GebaUe, T. R., and Greenberg, J.M. 1986, 

Astr. Ap., 154, L5. 
Campins, H., and Ryan, E. V. 1989. Ap. J., June 15, in press. 
CardeUi, J. A., and Savage, B. D. 1988, Ap. J., 325, 864. 
CardeUi, J. A., Clayton, G. C, and Mathis, J. S. 1988, Ap. J. (Lett.), 329, L33 . 
Cardelli, J. A., Clayton, G. C, and Mathis, J. S. 1989, Ap. J., submitted. 
Chlewicki, G., and Greenberg, J. M. 1989, Ap. J., in press. 
Chlewicki, G., and Laureijs, R. J. 1988, Astr. Ap., 207, Lll. 
Chlewicki, G., de Groot, M. S., van der Zwet, G. P., Greenberg, J. M., Alvarez, P.P., and 

Mampaso, A. 1987, Astr. Ap., 173, 131. 
Clayton, D. D. 1988, Ap. J., 334, 191. 
Cox, D. P. 1988, in Supernova Remnants and the Interstellar Medium, R. S. Roger and T. L. 

Landecker, eds. (Cambridge University Press: Cambridge), p. 73. 
Cuzzi, J. N., Lissauer, J.J., Esposito, L.W., Holberg, J. B., Marouf, E. A., Tyler, G. L., and 

Boischot, A. 1984, in Planetary Rings, ed. R. G. Greenberg and A. Brahic (Univ. of Arizona 

Press; Tucson), p. 73. 
de Boer, K. S., Lenhart, H., van der Hucht, K. A., Kamperman, T. M., Kondo, Y., and 

Bruhweiler, F. C. 1986, Astr. Ap., 157, 119. 
Dragovan, M. 1986, Ap. J., 308, 270. 
Draine, B. D. 1985, in Protostars and Planets II, D. C. Black and M.S. Matthews, eds., (Univ. 

of Arizona Press: Tucson), p. 621. 
Draine, B. T. 1989, in Interstellar Dust, proceedings of Symposium 1 35 of the Intemational 

Astronomical Union, held at Santa Clara University, July 26-30, 1988 (Reidel: Dordrecht). 
Draine, B. T., and Lee, H. M. 1984, Ap. J., 285, 89 (DL). 


Mathis, John S. 

Draine, B. T., and Salpeter, E. E. 1979a, Ap. J., 231, 77. 

Draine, B. T., and Salpeter, E.E. 1979b, Ap. J., 231, 438. 

Duley, W. W. 1984, Quar. J. Roy. Astr. Soc, 25, 109. 

Duley, W. W., and WiUiams, D. A. 1988, M. N. R. A. S., 231, 969. 

Duley, W. W., WUUams, D. A., and Jones, A. P. 1989, M. N. R. A. S., in press. 

Fitzpatrick, E. L., and Massa, D. 1986, Ap. J., 307, 286. 

Fitzpatrick, E. L., and Massa, D. 1988, Ap. J., 328, 734. 

Gehrz, R. D. 1989, in Interstellar Dust, proceedings of Symposium 135 of the International 

Astronomical Union, held at Santa Clara University, July 26-30, 1988 (Reidel: Dordrecht). 
Gillett, F. C, Jones, T. W., Merrill, K. M., and Stein, W. A. 1975, Astr. Ap., 45, 77. 
Greenberg, J. M. 1989, in Interstellar Dust, proceedings of Symposium 135 of the International 

Astronomical Union, held at Santa Clara University, July 26-30, 1988 (Reidel: Drodrecht). 
Greenstein, J. L. 1981, Ap. J., 245, 124. 

Glisten, R., and and Mezger, P. G. 1982, Vistas in Astronomy, 26, 159. 
Hartmann, W. K. 1969, Icarus, 10, 201. 
Hecht, J. H. 1987, Ap. J., 314, 429. 

Hecht, J. H., Holm, A. V., Donn, B., and Wu, C.-C. . 1984, Ap. J., 280, 228. 
Herbig, G. H. 1988, Ap. J., 331, 999. 
Hildebrand, R. H. 1987, Ap. Lett, and Comm., 26, 263. 

Hildebrand, R. H. 1988, Q. J. Roy. Astr. Soc, 29, 327. 

Hildebrand, R. H., Davidson, J. A., Gonatas, D., Novak, G., Piatt, S. R., and Wu, X. 1989, 
Ap. J., in press. 

Hobbs, L. M., York, D. G., and Oegerle, W. 1982, Ap. J. (Lett.), 252, L21. 

Hong, S. S., and Greenberg, J. M. 1980, Astr. Ap., 88, 194. 

Hoyle, F., and Wickramasinghe, N. C. 1962, M. N. R. A. S., 124, 417. 

Hoyle, F., and Wickramasinghe, N. C. 1984, From Grains to Bacteria (University College 
Cardiff Press). 

Humphreys, R. M. 1978, Ap. J. Suppl., 38, 309. 

Jabbir, N. L., Jabbar, S. R., Salih, S. A. H., and Majeed, Q. S. 1986, Ap. Sp. Sci., 123, 351. 

Jenkins, E. B. 1987, in Interstellar Processes, D. J. Hollenbach and H. A. Thronson, eds. 
(Reidel: Dordrecht), p. 533. 

Jenkins, E. B. 1989, in Interstellar Dust, proc. of lAU Symposium 135, held at Santa Clara 
University, July 26-30, 1988 (Reidel: Drodrecht). 

Jenkins, E. B., Savage, B. D., and Spitzer, L., Jr. 1986, Ap. J., 301, 355. 

Josafatsson, K., and Snow, T. P., Jr. 1987, Ap. J., 319, 436. 

Joseph, C. L. 1988, Ap. J., 335, 157. 


Mathis, John S. 

Jura, M. 1980, Ap. J., 235, 63. 

Krelowski, J., and Walker, G. A. H., Ap. J., 312, 860. 

Lee, H. M., and Draine, B. T. 1985, Ap. J., 290, 85. 

Lewis, R. S., Tang, M., Wacker, J. F., Anders, E., and Steel, E. 1987, Nature, 326, 160. 

Leger, A., and Puget, J. L. 1984, Astr. Ap., 137, L5. 

Mackinnon, I. D.R., and Reitmeijer, F. J. M. 1987, Rev. of Geophy., 25, 1527. 

Martin, P. G. 1989, in Interstellar Dust, prcx:. of lAU Symposium 135, held at Santa Clara 

University, July 26-30, 1988 (Reidel: Drodrecht). 
Martin, P. G., and Angel, J. R. P. 1974, Ap. J., 188, 517. 
Martin, P. G., and Angel, J. R. P. 1975, Ap. J., 195, 379. 
Martin, P. G., and Rogers, C. 1987, Ap. J., 322, 374. 
Mathis, L S. 1986, Ap. J., 308, 281. 

Mathis, J. S., and Wallenhorst, S. G. 1981, Ap. J., 244, 483. 
Mathis, J. S., and Whiffen, G. 1989, Ap. J., 341, in press (MW). 
Mathis, J. S., Rumpl. W., and Nordsieck, K.H. 1977, Ap. J., 217, 425 (MRN). 
Mauche, C. W., and Gorenstein, P. 1986, Ap. J., 302, 371. 
Merrill, K. M. 1977, in The Interaction of Variable Stars with Their Environment, lAU Colloq. 

42, ed. R. Kippenhahn, J. Rahe, and W. Strohmeier (Bamberg: Remeis-Stemwarte), p. 446. 
Nuth, J. A. 1985, Nature, 318, 166. 
Rieke, G. H. 1974, Ap. J. (Lett.), 193, L81. 
Rowan-Robinson, M. 1986, M. N. R. A. S., 219, 737. 
Savage, B. D., and Mathis, J S. 1979, Ann. Rev. Astron. Astrophys., 17, 73. 
Savage, B. D., Bohlin, R. C., Drake, J. F., and Budich, W. 1977, Ap. J., 216, 291. 
Schulte, D. H. 1959, Ap. /., 124, 41. 
Schutte, W. 1988, PhD. Thesis, University of Leiden. 
Seab, C. G. 1987, in Interstellar Processes, ed. D. J. Hollenbach and H. A. Thronson (Reidel: 

Dordrecht), p. 491. 
Seab, C. G., and Shull, J. M. 1983,Ap. J., 275, 652. 
Seaton, M. J. 1979, M. N. R. A. S. 187, 73p. 
Serkowski, K. 1965, Ap. J., 141, 1340. 
Shaver, P. A., McGee, R. X., Newton, L. M., Danks, A. C, and Pottasch,, S. R. 1983, M. 

N. R. A. S., 204, 53. 
Shull, J. M., and van Steenberg, M. E. 1985, Ap. J., 294, 599. 
Tang, M., and Anders, E. 1988, Ap. J. (Utt.), 335, L31. 
Tielens, A. G. G. M., and Allamandola, L. J. 1987, in Interstellar Processes, ed. D. J. 

Hollenback and H. A. Thronson (Reidel: Dordrecht), p. 397. 


Mathis, John S. 

Vrba, F. J., Coyne, G. V., and Tapia, S. 1981, Ap. J., 243, 489. 
Volk, H. J., Jones, F. C, MorfiU, G.E., and Roser, S. 1980, Astr. Ap., 85, 316. 
Watson, D. L., Genzel, R., Townes, C. H., and Storey, J. W. V. 1985^ Ap. J., 298, 316. 
Werner, M. W., Davidson, J. A., Hildebrand, R. H., Morris, M. R., Novak, G., and Piatt, W. 

R. 1988, Ap. J., 333, 729. 
Whittet, D. C. B. 1984, M. N. R. A. S., 210, 479. 
Wright, E. L. 1987, Ap. J., 320, 818. 




John A. Wood 
Harvard-Smithsonian Center for Astrophysics 


Refractory solids in chondrites and comets: how similar? 

John A. Wood 
Harvard-Smithsonian Center for Astrophysics 

The raw materials of the solar system were interstellar gas, grains of ice, refi^ctory dust, and 
organic material. Gravitational collapse caused these ingredients to fall together into a protosun and 
accretion disk (the solar nebula), out of which the planetary system grew. The raw materials have 
been preserved to differing degrees in the most primitive solar system bodies, asteroids and comets. 

The present paper deals mostly with the state of preservation of one of the primordial in- 
gredients, the refractory dust. The study of samples of asteroids, in the form of chondritic meteor- 
ites, reveals that the dust component in the inner solar system was extensively altered by high- 
temperature events and processes before it was aggregated into chondritic planetesimals. The 
chondritic material was further altered by metamorphic heating in its parent planetesimals after 


The nature of these high-temperature events and processes is not known, but the evidence of 
their operation is pervasive and unequivocal. Chondritic meteorites are aggregations of particulate 
matter. There are three principal categories of particles: Ca,Al-rich inclusions, chondrules, and 
matrix dust grains. All were transformed by high temperatures prior to their aggregation, presumab- 
ly while dispersed in the solar nebula. 

CaAl-Rich Inclusions 

Ca,Al-rich inclusions (CAI's) are objects ranging is size from several cm (very rare) down to 
the limit of visibility. They make up as much as 10% of the volume of some chondrites, but are 
absent from others. Some show clear textural evidence of having been melted; others have textures 
that are harder to interpret. The clearest evidence for high- temperature processing of CAI's lies in 
their chemical compositions: they are systematically depleted in the more volatile elements. Mg, Si, 
and Fe (cosmically abundant elements of intermediate volatility) are depleted by a factor of ten or 
more relative to the most refractory major elements (Ca, Al, Ti), compared with the cosmic abun- 
dances. The resulting enhancement of Ca, AI, and Ti in these objects explains their name; they are 
also referred to as refractory inclusions. The minerals in them tend to be rich in Ca and Al, and 
relatively poor in Si. They include spinel, MgAl^O*; melilite, Ca2(Al2,MgSi)Si07; perovskite. 


CaTiO,; and hibonite, CaAljjO,,. CAI's also contain elevated abundances of involatile trace ele- 
ments (Ba, Sr, Sc, Y, rare earth elements, Zr, Hf, Th, V, Nb, Ta, Mo, W, U, Re, Ru, Os, Rh, Ir, 

The significance of these element-abundance patterns is that CAI's must have become depleted 
in relatively volatile elements by high-temperature events that either incorripletely vaporized precur- 
sor material; or totally vaporized it, after which the system cooled and recondensed selectively, such 
that the CAI's incorporated early-condensing high- temperature minerals, but not later, less-refractory 
compounds. Isotopic mass- fractionation effects measured in Mg, Si, Ca, and Ti in CAI's indicate 
that the latter experienced a complex history of both partial vaporization and condensation (Fig. 1; 
Clayton et al., 1985). 

Figure 1. Isoiopic compositions of Si in individual CAI's 
from the Allende chondrite. Mass fractionation of isotopes 
between the CAI's and their environment has dispersed 
compositions of the former along a line of slope 1:2. Points 
at the heavy end of the line are chiefly coarse-grained 
CAI's; evaporation residues tend to be isotopically heavy. 
Points at the light end of the hne, where condensates should 
plot, are mostly fme-grained CAI's. Figure from Clayton et 
al. (1985), who point out that processes more complex than 
simple evaporation or condensation are required to explain 
the data. 



\ I 

-I — 

1 ■— I 1 

T 1— 

1 1 




7 ^ 

11 ^ 

J /• 



. 5^^ 















~ ^ 




1 . 1 


1 1 

1 1 

1 1 

-5 -4 -3-2-1 1 2 


On the other hand, some CAI's contain O, Mg, Si, Ca, Sr, Ba, Nd, and Sm with anomalous 
isotopic compositions not attributable to mass firactionation or radioactive decay, which are inter- 
preted to be the surviving signatures of particular presolar nucleosynthetic sources (Fig. 2; see, e.g., 
Lee, 1988). These show that their host CAI's must contain a component that was never vaporized 
and mixed with other solar system material. (If all the interstellar dust had been totally vaporized 
in the nebula, its nuclides would have mingled in the nebular gas and ±ese anomalous isotopic 
signatures would have been lost) 


Members of the most abundant category of particles in chondrites are called chondrules. These 
are the order of a millimeter in diameter, tend to be spheroidal in shape, and make up as much as 
75% of the volume of some chondrites. Most of them show clear textural evidence of having been 
melted. There is no chemical evidence for vaporization or condensation, however, the major 
element abundances in chondrules are close to the cosmic abundance pattern. Since the dominant 
metallic elements in the cosmic abundance table are Si, Mg, and Fe, the most abundant minerals in 
chondrules are olivine, (Mg,Fe)2Si04, and low-Ca pyroxene, (Mg,Fe)Si03. 

Many chondrules contain relic grains of an earlier generation of mineral matter, which was 
melted to form the dispersed chondrules (Nagahara, 1983; Kracher et al, 1984). These grains have 


Figure 2. Isotopic anomalies in two CAI's 
(EK-1-4-1 and C-1) that cannot be explained 
by mass fractionation or any other solar 
system process. Mass numbers appear on 
the abscissa; deviations of isotopic abundan- 
ces from a solar system standard (dashed 
line), in parts per l& (epsilon units), on the 
ordinate. These anomalies must be presolar 
in origin: r and s along the ordinate denote 
the r- and s- nucleosynthetic processes in 
stellar interiors, which give rise to these 
nuclides. Figure from Lee (1988). 

resorbed margins (i.e., shapes reminiscent of partly-melted ice cubes), and compositions that differ 
from those of minerals surrounding them, which crystallized from the chondrule melt. The relic 
grains are often quite large (tens or even hundreds of microns), so they are unlikely to be surviving 
presolar interstellar grains; they appear to have been formed in earlier cycles of nebular activity, 
either by crystallization from a melt or by condensation. 

Laboratory measurements in the last decade have established the time scale on which chon- 
dniles cooled (Hewins, 1988). The morphologies of crystals growing from melts, and also the 
degree to which chemical zonation in the forming crystals is preserved, is a function of the rate at 
which the melts cooled. Mek-droplets having the compositions of chondmles have been cooled at 
various rates under carefully controlled laboratory conditions, and it has been found that cooling 
rates of 100-2000 K/hr are needed to reproduce the properties of real chondmles. 

The isotopic compositions of chondmles have not been smdied as extensively as those of 
CAI's. Chondrules have been found to show the same kinds of isotopic mass fractionation effects 
as CAI's (Clayton et al., 1985; Fig. 3), but the degree of fractionation- the range of isotopic 

Figure 3. Isotopic composition of Si in chondrules from 
the Allende chondrite. Like the CAI's in Fig. 1, these 
display mass fractionation. Figure from Clayton et al. 





• Dork Ifldusionj 

• Bulk 

-1.0 -0.5 

8^ Si [%c. 



compositions observed-- is only about 1/6 that displayed by CAI's. It is not known whether this 
fractionation occurred when the chondrules were melted, or if it already existed in the chondrule 
precursor material. 

Few studies of chondrriles have been made that were capable of revealing nucleosynthetic 
isotopic anomalies of the type observed in the Mg, Si, Ca, etc. of CAI's. -The isotopic composition 
of O in chondrules has been found to vary in a way that is not attributable to mass fractionation 
(Fig. 4; Clayton ei al., 1985), and this has been widely ascribed to the incomplete mixing of two 
reservoirs (dust and nebular gas) in which presolar differences in O isotope composition were 
preserved, but it is also possible that a non-mass-dependent fractionation process operated in the 
nebula that could have produced the effect from an initially well-homogenized system (Thiemens 
and Heidenreich, 1983). 




Figure 4. Isotopic composition of oxygen in chondrules from 
the Allende chondrite. The trend cannot be attributed to mass 
fractionation, or it would parallel the "Terrestrial fractionation" 
(slqje 1:2) curve. It must reflect mixing in the chondrules of 
O fi'om two (or more) resMvoirs, one of which is richer in "O 
than the other. Figure from Clayton et al. (1985). 

8'°0 %o rel. SMOW 


The chondrules and CAI's in chondrites are embedded in a matrix of fine mineral grains, 
mosdy in the 1-10 |im size range. In some chondrites the matrix consists of hydrated clay miner- 
als, and it is clear that this material is secondary, having been formed in the parent meteorite planet 
by aqueous alteration of an earlier generation of matrix grains of unknown character (Zolensky and 
McSween, 1988; see McSween's paper in this volume). Where the microscopically visible grains 
are not obviously alteration products, they have been variously interpreted as nebular condensates 
and as comminuted debris from the collisions of larger objects (Scott et al, 1988). 

The only known bona fide presolar interstellar grains in chondrites occur in the matrix. These 
are submicron grains of carbonaceous matter: organic carbon, diamond, graphitic or amorphous 
carbon, and SiC (Anders et al., 1988; see the contribution by Chang to this volume). These exist 


Figure 5. Isotopically anomalous Xe released upon heating a C-rich 
residue which was obtained by dissolving a sample of the Murray 
chondrite in acids. This pattern is characteristic of the Xe isotopic 
abundances produced by the s-process in stellar interiors. Figure 
from Tang ei al. (1988). 



■■■I 1 I 





(r ' 



1 1 1 1 

138 130 H2 


between the larger grains described above, and make up <1% of the matrix. They exhibit distinc- 
tive anomalous isotopic signatures for C, included noble gases (e.g.. Fig. 5), and Si (where present), 
presumably impressed upon them in their different nucleosynthetic sites. 

The matrix may also contain a minor component of other interstellar phases, such as silicates, 
but these have not been identified. 

Postaccretional Metamorphism 

Mineralogical and textural evidence show that after this particulate matter accumulated into 
chondritic planetesimals, internal temperatures in tiie planetesimals increased, and the chondritic 
material was metamorphically altered. Some of the chondrites we smdy today have been metamor- 
phosed less severely than others, but none are completely unmetamorphosed. CI and C2 car- 
bonaceous chondrites were metamorphosed in a watery environment (Zolensky and McSween, 1988); 
other chondrite subtypes experienced anhydrous metamorphism (McSween et al., 1988). The source 
of planetary heat is not known. Conjectural possibilities that are often discussed are the decay of 
short-lived ^Al (half-life, 0.73 my), and electromagnetic induction driven by a dense solar wind 
emitted by the pre-main-sequence sun (Wood and Fellas, 1989). 


Are the refractory particles in comets likely to be similar to these chondrite components? We 
have had two glimpses of the nature of cometary particles, in the form of stratospheric inter- 
planetary dust particles and the interception of Halley's Comet dust particles by the VEGA and 
GIOTTO spacecraft in 1986. Interplanetary dust particles of the chondritic porous variety, which 
are distinctly different in character from the matrix or any other component of meteorites, are 
presumed to be of cometary origin (e.g., Brownlee, 1985); certainly, some component of the inter- 
planetary dust must be. These particles can be studied in exquisite detail using microbeam analyti- 
cal techniques, but many types of study are precluded by the minuscule aggregate amount of 
material that has been or ever can be collected by this method. Further, stratospheric collection 


techniques cannot provide samples of larger (order of Ig) refractory objects like those that were 
found in the Halley coma by GIOTTO. Chemical compositions and the size distribution of Halley 
particles were measured by the flyby spacecraft, but the latter were not equipped to study the many 
other propenies of cometary refractory panicles. 

It is widely believed that chondritic and cometary refractory grains had very different histories 
in the solar nebula, so they must have ended up having dissimilar physical properties. Tempera- 
tures were high in the inner solar nebula, and there (as noted above) the interstellar grains that 
joined the disk were thermally processed into chondrules, CAI's, and matrix. The temperature in 
solar nebula models decreases with radial distance, however, and the outer nebula (where the giant 
planets formed, and presumably also the comets) was probably never hot enough to melt or vapor- 
ize silicates. There interstellar grains would have survived unchanged. Some became incorporated 
in accumulating comet nuclei, and today, after long residence in the Con cloud, representatives of 
these panicles are being released by subliming comets. 

According to this picture, which is probably correct in broad oudine, chondritic and cometary 
particles have nothing in common except their ancestry. Yet the picture rests on three over- 
simplified assumptions, which may not be as solidly founded as they are commonly thought to be: 
(1) nebular temperatures decreased outward in the nebula, (2) comets formed much farther out in 
the nebula (and lower on the thermal gradient) than chondrites, and (3) there was no opportunity for 
diffusive mixing of particles between these two widely-separated zones. These assumptions are 
examined in the following sections. 


The simplistic concept that "temperature decreased outward in die nebula" dates back to 
Cameron's (1962) model of the nebula, which formed by the simultaneous adiabatic compression of 
all the mass of a rotating, gravitationally-collapsing volume of interstellar gas and dust when it fell 
together. Gas arriving near the center of gravity of the system fell faster, was compressed to higher 
pressures, and grew hotter than gas tiiat joined the periphery of the disk. Cameron's 1962 model 
was abandoned when it was found that gas in a collapsing system of this type would not all fall 
together simultaneously; the nebula came to be seen to be an example of a protostellar accretion 
disk, in which gas falls onto the disk and is redistributed within it by viscous processes over a 
period of time (Wood and Morfill, 1988). Viscous dissipation of energy in the disk would have 
heated it, and the profile of temperature in it would have decreased outward, but in most nebula 
models the gas temperature achieved by this mechanism at 3 AU (representative of the position of 
the asteroid belt, the source of chondritic meteorites, in the present solar system) is only a few 
hundred K, far too low to melt or vaporize refractory dust. 

It has also been realized that the time scale of cooling of the chondrules in meteorites (100- 
2000K/hr, noted above) is too rapid to be consistent with the behavior of a hot nebula. If immer- 
sion in a hot nebula had been what melted the chondrules, there is no way such a vast structure 
could have lost heat fast enough, nor could the chondrules have been moved from a hot zone to a 
cooler one fast enough, to produce this cooling rate. The thermal processing of chondrules and 
other refractory components must be attributed to transient, localized energy releases in the nebula, 
not to high ambient nebular temperatures. The nature of these transient energetic events is not 
known. Many proposals have been put forward, but all made to date have serious weaknesses. 


[Hypotheses have included nebular lightning (Whipple, 1966), impacts (e.g., Fredriksson, 1963), 
reconnection of magnetic field lines (Sonett, 1979), drag heating of solids falling into the nebula 
(Wood, 1984), and magnetic flares above the nebular surface (Levy, 1988)]. 

The idea that refractory particles were thermally processed in the inner solar system has to be 
looked at in the context of transient high-energy events rather than that of a hot nebular environ- 
ment. It might seem that the effect of radial distance would be the same in both cases, since the 
frequency and/or intensity of all the mechanisms listed above increase with nearness to the sun and 
the center of gravity of the solar nebula. However, it is probably unwise to make this generali2a- 
tion, since the particular mechanism that melted chondrules has yet to be identified. Even if the 
generalization is correct, the fact that discrete events seem to have been responsible is crucial: a 
falling-off of effectiveness of the mechanism with radial distance might mean that the energetic 
events occurred less frequendy, but not necessarily with less intensity. In this case, some fraction 
of the refractory grains far out in the nebula might have been processed in the same way the 
chondrite components were. 


Oort's (1950) concept that the comet nuclei accreted in the solar system, after which gravita- 
tional encounters with the giant planets flimg them out of the planetary system (some becoming 
gravitationaUy imbound, others left coasting slowly in huge (Dort-cloud orbits), has gained wide 
acceptance. Few remember that in Oort's (1950) classic paper the comets accreted in the asteroid 
belt (relatively near the zone where the refractory particles in chondrites were thermally processed) 
and were ejected by Jupiter. It has been recognized since that Jupiter is not an efficient supplier 
for the Oort cloud: Jupiter's throw is too powerful, and the great majority of objects it ejected 
would have left at greater than the escape velocity. Only a small fraction would remain in the Oort 
cloud. Neptune and Uranus are gentler ejectors; most of the comets in the Oort cloud probably 
formed in die Neptune-Uranus zone of the nebula. From this fact the generalization has taken hold 
that comets formed at much greater radial distances than chondrites, where conditions were far 

However, while in the current picture most of the comets reentering the solar system from the 
Oort cloud formed in the Neptune-Uranus zone of the nebula, all of them did not. According to 
Safronov's (1972) analysis, approximately 8% of comets in the Oort cloud were ejected by Jupiter. 
Though Jupiter does waste comets by ejecting them too powerfully, this is partly offset by the fact 
that nuclei probably formed much more abundandy in the Jupiter zone than farther out Once 
more, it is important to recognize that we are dealing with discrete entities (in this case comets 
rather than transient heating events), not generalized qualities or populations. A mission sent to 
sample the nucleus of a single comet and return material to Earth would have an 8% chance (if 
Safronov is right) of collecting refractory grains that experienced the nebular environment close to 
the asteroid belt rather than in the Neptune-Uranus zone. 

But surely all the comet nuclei must have formed at much greater radial distances than aste- 
roids, since the former incorporated ices and the latter did not? This is another generalization that 
cannot be made with confidence. It has become increasingly apparent that there is no sharp line of 
demarcation between comets and asteroids (Hartmann et al., 1987). The asteroids parental to the 
carbonaceous chondrites may have incorporated ice when they accreted, which subsequently melted 


and altered accompanying refractory dust to the assemblage of clay minerals observed today in Cl 
and C2 chondrites [e.g., Grimm and McSween (1988); see the article by McSween in this volume]. 
Some asteroids may still contain subsurface ice at the present time (Lebofsky et al., 1981). It is 
likely that planetesimals containing the whole spectrum of possible ice contents accreted in the solar 
nebula. Objects that formed inside the present orbit of Jupiter, close to the zone where the car- 
bonaceous chondrite parent bodies accumulated, might have incorporated^ less ice and a different 
blend of ices than objects accreting in the Neptune zone, yet the Jupiter objects would display bona 
fide cometary activity if orbital changes brought them closer to the sun; they are comet nuclei. 

If there is a sharp dichotomy that can be drawn among early solar system planetesimals, it is 
based on whether they grew warm enough internally to melt any water ice that was present. If 
not- if because of their small size, large orbital mean distance, or some other factor, they were not 
heated to the extent that the parent meteorite planetesimals were- then presolar interstellar grains 
can have siuvived in them. The porous interplanetary dust particles captured in the stratosphere 
have been spared hydrothermal alteration in their parent bodies. In planetesimals where the ice did 
melt, primordial presolar grains will have been destroyed, altered into clay minerals and other 
secondary products of the type found now in Cl and C2 carbonaceous chondrites. (Some of the 
stratospheric dust panicles also have this character.) 

The previous section discussed the likelihood that the amount of preaccretionary thermal 
processing experienced by refractory grains in the nebula diminished with radial distance. There is 
no assurance that the curve of degree of thermal processing (tapering outward) and the curve of 
condensation and accretion of ices (tapering inward) did not overlap. C2 chondrites, which were 
altered by planetary fluids that may have come from melting ices, also contain once-molten chon- 
drules (though relatively few of them). 


Would refractory particles necessarily accrete at the same radial distance in the nebula where 
they were thermally processed? Morfill (1985) has put forward a model of chondrite formation in 
which CAI's were thermally processed at a much snialler distance (where the nebular gas is hotter) 
than chondrules, after which radial diffusive transport of the two ingredients mixed them before they 
accreted. Such an effect, if it operated, also could have diffused chondrules outward to the zone 
where the giant planets formed. Stevenson (1989) has examined the tendency of refractory particles 
to diffuse in the nebula as a result of turbulent mixing; he finds the effect to be small, but not 
zero. In the steady state, particulate matter at radial distance R will contain a component of grains 
derived from R^ (<R) whose fractional abundance is -R/R. 

The steady state presupposes, however, that particles remain unaccreted for a long time, long 
enough for turbulent motions of the gas to diffuse them for distances of many AU. Little is known 
about the mechanics of accretion in the nebula or its time scale. The petrographic properties of 
chondrites suggest to me that accretion occurred relatively rapidly. Some chondrite subtypes differ 
in little more than their petrographic textures. CV3 and C03 chondrites, for example, have almost 
identical major element compositions, but the chondrules in CV3 chondrites are conspicuously larger 
and more irregular in shape than those in COS chondrites (see Fig. 4 of Wood, 1985). The chon- 
drule-forming process, whatever it was, must have operated differently on the same type of raw 
material at two different times and/or places in the solar nebula, producing the dissimilar popula- 


tions of chondrules in these two chondrite subtypes. I have argued that a corollary conclusion can 
be drawn, that accretion of chondrules into chondrites must have occurred prompdy, before the 
dispersed chondrules had a chance to diffuse into other zones of the nebula and become mixed with 
the products of other transient thermal events; or else these textural differences could not have been 
retained (Wood, 1985). Once the chondrite planetesimals achieved some size, motions of the thin 
nebular gas would be ineffective in diffusing them to greater radial distances. Even if this qualifi- 
cation Morfill's and Stevenson's conclusion is correct, however, it is unlikely that rapid accretion in 
the early nebula could sweep up 100% of the chondrules, CAI's, and matrix dust immediately after 
thermal processing created them. Some fraction of this material would remain available for disper- 
sal to other zones of the nebula. 


All of the refractory grains in comets may not be totally dissimilar to the refractory grains in 
chondritic meteorites. The great bulk of cometary grains is likely to consist of more or less well- 
preserved interstellar grains, probably analogous to the porous interplanetary dust particles collected 
in the terrestrial stratosphere. However, there may also be a component of solids that is related to 
particulate ingredients of chondrites, because (1) the thermal events that processed chondritic par- 
ticles may also have occurred (though less frequentiy) in the zone of the giant planets; (2) some 
comets probably formed in the Jupiter zone, maybe even inside the present orbit of Jupiter, in or 
near the same region where chondrites formed; and (3) mechanical mixing in the turbulent nebula 
may have contaminated the particulate matter at large radial distances with a small component of 
refractory particles that were thermally processed at much smaller radial distances in the nebula. 

Such a chondritic ingredient would not be without scientific interest, even though we already 
have access to copious amounts of these particles in the chondritic meteorites. If the chondrule- 
and CAI-forming events operated less energetically at the greater radial distances where comets 
formed, they might yield only-partly-transformed products, the study of which could shed light on 
the nature of the nebular high-energy events that so profoundly affected refractory material in the 
inner solar system. Our efforts to infer the nature of these events from the study of chondrites 
have been singularly improductive to date. 



Anders, E., Lewis, R. S., and Tang, M. (1988) Interstellar grains in meteorites: diamond and silicon 
carbide. In Interstellar Dust, Proc. of lAU Symposium 135, Santa Clara CA, in press. 

Brownlee, D. E. (1985) Cosmic dust: collection and research. Ann. Rev. Eurth Planet. Sci. 13, 147- 

Cameron, A. G. W. (1962) The formation of the sun and planets. Icarus 1, 13-69. 

Clayton, R. N., Mayeda, T. K., and Molini-Velsko (1985) Isotopic variations in solar system mater- 
ial: evaporation and condensation of silicates. In Protostars and Planets II (Eds. D. C. Black 
and M. S. Matthews). Univ. Arizona Press, Tucson, pp. 755-771. 

Fredriksson, K. (1963) Chondrules and the meteorite parent bodies. Trans. N. Y. Acad. Sci. 25, 756- 

Grimm, R. E. and McSween, H. Y., Jr. (1988) Water and the thermal history of the CM car- 
bonaceous chondrite parent body. Lunar Planet. Sci. XIX, 427-428. 

Hartmann, W. K., Tholen, D. J., and Cruikshank, D. P. (1987) The relationship of active comets, 
"extinct" comets, and dark asteroids. Icarus 69, 33-50. 

Hewins, R. H. (1988) Experimental studies of chondrules. In Meteorites and the Early Solar System 
(Eds. J. F. Kerridge and M. S. Matthews). Univ. Arizona Press, Tucson, pp. 660-679. 

Kracher, A., Scott, E. R. D., and Keil, K. (1984) Relict and other anomalous grains in chondrules: 
implications for chondrule formation. Proc. Lunar Planet. Sci. Conf. 14th, B559-B566. 

Lebofsky, L., Feierberg, M., Tokunaga, A., Larson, H., and Johnson, J. (1981). The 1.7- to A.l-yjn 
spectrum of Asteroid 1 Ceres: evidence for structural water in clay minerals. Icarus 48, 453- 

Lee, T. (1988) Implications of isotopic anomalies for nucleosynthesis. In Meteorites and the Early 
Solar System (Eds. J. F. Kerridge and M. S. Matthews). Univ. Arizona Press, Tucson, pp. 

Levy, E. H. (1988) Energetics of chondrule formation. Meteorites and the Early Solar System (Eds. 
J. F. Kerridge and M. S. Matthews). Univ. Arizona Press, Tucson, pp. 697-711. 

McSween, H. Y. Jr., Sears, D. W. G., and Dodd, R. T. (1988) Thermal metamoiphism. In Meteor- 
ites and the Early Solar System (Eds. J. F. Kerridge and M. S. Matthews). Univ. Arizona 
Press, Tucson, pp. 102-113. 

Morfill, G. E. (1985) Physics and chemistry in the primitive solar nebula. In Birth and Infancy of 
Stars (Eds. R. A. Lucas, A. Omont, and R. Stora). North-Holland, Amsterdam, pp. 693-792. 

Nagahara, H. (1983) Chondrules formed through incomplete melting of the pre-existing mineral 
clusters and the origin of chondrules. In Chondrules and their Origins (Ed. E. A. King). 
Lunar Planet. Inst., Houston, pp. 211-222. 


Safronov, V. S. (1972) Ejection of bodies from the solar system in the course of the accumulation 
of the giant planets and the formation of the cometary cloud. In The Motion, Evolution of 
Orbits, and Origin of Comets, lAU Symp. No. 45 (Eds. G. A. Chebotarev and E. I. Kazimir- 
chak-Polonskaya). D. Reidel, Dordrecht, pp. 329-334. 

Scott, E. R. D., Barber, D. J., Alexander, C. M., Hutchison, R., and Peck, J. A. (1988) Primitive 
material surviving in chondrites: matrix. In Meteorites and the Early Solar System (Eds. J. F. 
Kerridge and M. S. Matthews). Univ. Arizona Press, Tucson, pp. 718-745. 

Sonett, C. P. (1979) On the origin of chondrules. Geophys. Res. Lett. 6, 611'6%Q. 

Stevenson, D. J. (1989) Chemical heterogeneity and imperfect mixing in the solar nebula. Astro- 
phys. J., in press. 

Tang, M., Lewis, R. S., Anders, E., Grady, M. M., Wright, I. P., and PilUnger, C. T. (1988) Isoto- 
pic anomalies of Ne, Xe, and C in meteorites. I. Separation of carriers by density and chemi- 
cal resistance. Geochim. Cosmochim. Acta 52, 1221-1234. 

Thiemens, M. H. and Heidenreich, J. E. (1983) The mass independent fractionation of oxygen: a 
novel isotope effect and its possible cosmochemical implications. Science 219, 1073-1075. 

Whipple, F. L. (1966) Chondrules: suggestion concerning the origin. Science 153, 54-56. 

Wood, J. A. (1984) On the formation of meteoritic chondrules by aerodynamic drag heating in the 
solar nebula. Earth Planet. Sci. Lett. 70, 11-26. ■ 

Wood, J. A. (1985) Meteoritic constraints on processes in the solar nebula. In Protostars and 
Planets II (Eds. D. C. Black and M. S. Matthews). Univ. Arizona Press, Tucson, 687-702. 

Wood, J. A. and Morfill, G. E. (1988) A review of solar nebula models. In Meteorites and the 
Early Solar System (Eds. J. F. Kerridge and M. S. Matthews). Univ. Arizona Press, Tucson, 
pp. 329-347. 

Wood, J. A. and Pellas, P. (1989) What heated the parent meteorite planetesimals? In The Sun in 
Time (Eds. C. P. Sonett, M. S. Giampapa, and M. S. Matthews). Univ. Arizona Press, Tucson, 
in press. 

Zolensky, M. and McSween, H. Y., Jr. (1988) Aqueous alteration. In Meteorites and the Early 
Solar System (Eds. J. F. Kerridge and M. S. Matthews). Uruv. Arizona Press, Tucson, pp. 114- 



Bruce Fegley, Jr. 

Abteilung Kosmochemie 

Max-Planck-Institut fiir Chemie 

F.R. Germany 



Bruce Fegley, Jr., Abteilung Kosmochemie, Max-Planck-Institut fur Chemie, 
Saarstrasse 23, D6500 Mainz, F.R. Germany 


A growing body of observations demonstrates that comets, like the chondritic meteorites, are disequi- 
librium assemblages, whose chemistry and molecular composition cannot be explained solely on the basis of 
models of equilibrium condensation in the solar nebula. These observations include: 

1. The coexistence of reduced (e.g., CH4 and organics) and oxidized (e.g., CO, CO2, and H2CO) carbon 
compounds observed in the gas and dust emitted by comet P/Halley (Allen ei al. 1987;Combes ei al. 
1988; Eberhardt ei al. 1987a; Kawara et al. 1988; Kissel and Krueger 1987; Krankowsky ei al. 1986; 
Woods ei al. 1986). 

2. The coexistence of reduced (e.g., NH3) and oxidized (e.g., A^2) nitrogen compounds in the gas emitted 
by comet P/Halley (Allen ei al. 1987; Tegler and Wyckoff 1989; Wyckoff and Theobald 1989). 

3. The observation of large amounts of formaldehyde in the gas emitted by comet P/Halley {H2CO/H2O 
~ 1.5-4 %) and by comet Machholz (1988J) (Combes ei al. 1988; de Pater ei al. 1990; Mumma 
and Renter 1989). Formaldehyde would be rapidly destroyed by thermal processing in the solar nebula 
and must be formed by some disequilibrating process either in the solar nebula or in some presolar 

4. The observation of large amounts of the oxidized carbon gases CO and CO2 in comet P/Halley at levels 
fair exceeding those predicted by chemical equilibrium models of solar nebula carbon chemistry. In fact, 
oxidized carbon gases {CO + CO2 + H2CO) are the most abundant volatile (after water vapor) emitted 
by comet P/Halley. 

5. The observation of HON, which is not a predicted low temperature condensate in the solar nebula (e.g., 
Lewis 1972), in comet P/Halley (e.g., Schloerb ei al. 1987) and in comet Kohoutek (Huebner ei al. 

6. The observation of S2, which is argued to be a parent molecule vaporized from the nucleus, in comet 
IRAS-Araki-Alcock (1983d) by A'Hearn ei al. (1983) and Feldman ei al. (1984). This molecule is not 
an equilibrium condensate in the solar nebula and must result from disequilibrium chemistry. 

7. The deduction that organic grains (C-H-O-N particles) comprise about 30% of the dust emitted by comet 
P/Halley and that about 75% of the total carbon inventory of Halley is in these grains (Delsemme 1988; 
Jessberger el al. 1989) also implies substantial disequilibrium chemistry. 

8. The deductions that polyoxymethylene (polymerized formaldehyde or POM) is a constituent of the C- 
H-O-N particles emitted from comet P/Halley (e.g., Huebner 1987; Huebner ei al. 1987; Mitchell ei al. 
1987). If actually present in the C-H-O-N particles, POM is also a product of disequilibrating processes 
which took place in the solar nebula and/or in a presolar environment. 

Taken together, the observations listed above indicate that a variety of disequilibrating processes such as 
the kinetic inhibition of thermochemical reactions, grain catalyzed chemistry, lightning induced shock chem- 
istry, and photochemistry played an important role in establishing the chemistry and molecular composition 
of comet P/Halley in particular and presumably cometary material in general. However, the observational 
data do not by themselves constrain the timing and/or location of these various processes. 

This paper reviews the relevant observational data and attempts to quantify as far as possible by using 


current theoretical models and experimental data the relative importance of equilibrium and disequilibrium 
processes for the chemistry of comets. "Key" experimental and observational measurements which are 
important for better constraints on cometary origins are proposed. Finally, important measurements to be 
made by a comet nucleus sample return mission such as Rosetta are also suggested. 


Although observations of other comets are rapidly increasing our knowledge of their chemistry and 
molecular composition, the data for comet P/Halley are the most extensive and will form the basis for most 
of the discussion in this paper. 

Water Vapor 

As Weaver (1989) has noted, prior to the return of Halley observers had constructed a strong circum- 
stantial case for the dominance of water vapor in the volatiies emitted by comets (e.g., see Delsemme 1982). 
However, water vapor was first observed directly in comet P/Halley using the Fourier Transform Infrared 
Spectrometer (FTIR) on the Kuiper Airborne Observatory (Mumma et al. 1986) and was later observed 
in comet Wilson (19861) with the same apparatus (Larson et al. 1989). Subsequent measurements by the 
Neutral Mass Spectrometer (NMS) on Giotto showed that H2O comprised > 80% of the volatiies emitted by 
Halley (Krankowsky et al. 1986) and that the water vapor has a D/H ratio in the range of 0.6 — 4.8 x 10"'' 
(Eberhardt et al. 1987b). For reference, the terrestrial D/H ratio is about 1.6 x 10"'' and the primordial 
cosmic value is estimated to be 0.3 x 10"'' (Anders and Grevesse, 1989). Finally, more recent measurements 
by Mumma et al. (1987, 1988) have provided data on the ortho-lo-para ratio of water vapor emitted by 
Halley and by Wilson( 19861). For Halley, the derived ortho/para ratio is 2.3 ± 0.2 while for Wilson( 19861) 
the derived ortho/para ratio is 3.2 ± 0.2. The Halley ortho/para ratio implies a nuclear spin temperature of 
25 K while the Wilson( 19861) ortho/para ratio implies statistical equilibrium (T > 50 K). 

Carbon Compounds 

As pointed out in the introduction, the second most abundant group of volatiies (after water vapor) 
emitted from Halley are the oxidized carbon gases CO, CO2, and H2CO. Five different measurements 
provide information on the abundance and distribution of CO emitted from Halley. The Giotto NMS data 
for mass 28, which is probably dominated by CO (see the discussion for A'^2 below), has been interpreted 
as indicating a comet nucleus source for CO having CO/H2O < 0.07 and an extended source in the inner 
coma for CO having CO/H2O < 0.15 (Eberhardt et al. 1987a). Infrared measurements from the IKS 
experiment on the Vega space probes yield CO/H^O ~ 0.05 for a comet nucleus source (Combes et at. 
1988). Pioneer Venus Orbiter Ultraviolet Spectrometer (PVOUS) measurements of resonance UV emission 
from atomic hydrogen, oxygen, and carbon in the coma of Halley yield nominal H;0:C atomic ratios of 
1 : 0.7 : 0.07, which are consistent with CO/H2O ~ 0.25 or with CO2/H2O ~ 0.14 (Stewart 1987). Rocket- 
borne ultraviolet spectrometer measurements of resonance UV emission from atomic oxygen and carbon in 
the coma of Halley also yield similar CO/H2O ratios of 0.20 ± 0.05 for the one flight and 0.17 ± 0.04 for a 
second flight (Woods et al. 1986). Finally, measurements from the International Ultraviolet Explorer (lUE) 
satellite abo yield a rough estimate for the CO/H2O ratio of 0.1 -0.2 for Halley (Festou et al. 1986). This set 
of observations is generally interpreted as indicating a nucleus source for CO having CO/H2O ~ 0.02 — 0.07 
(by number) and a dispersed source which accounts for the balance of the observed CO (e.g., see Eberhardt 
et al. 1987a; Weaver 1989). 

Carbon monoxide has also been detected in comet West (1976 VI) with CO/H2O ~ 0.3 and in comet 
Bradfieid (1979 X) with CO/H2O ~ 0.02 (see Weaver 1989, and references therein). As Weaver notes, the 
"high" CO abundance in Halley, the "high" CO abundance in comet West (1976 VI), and the "low" CO 
abundance in comet Bradfieid (1979 X) are consistent and are plausibly explained by differences in the types 
of observations made and in the cometary dust production rates. Thus observations made with large fields of 
view of comets with large dust production rates (e.g., the UV observations of Halley and West) yield larger 
apparent CO abundances because CO production from the nucleus and also from evaporating organic grains 
is being observed, while the observations made with smaller fields of view (e.g., the IKS experiment) or 
observations made of comets with low dust production rates yield smaller apparent CO abundances because 


only CO production from the nucleus is being observed. 

Carbon dioxide was observed with both the NMS experiment on Giotto and the IKS experiment on 
Vega. The NMS data yield CO2IH2O ~ 0.04 (Krankowsky ei al. 1986) while the IKS data yield a ratio 
of about 0.03 (Combes ei al. 1988). Both ratios are appropriate for CQ2 emitted from the nucleus and 
together with the adopted values of ~ 0.02 - 0.07 for CO/H^O from the nucleus yield CO/CO2 ~ 0.50 - 2.3. 
In other words, roughly equal amounts of CO and CO2 are being emitted from the nucleus of Halley. As 
discussed below, this rough equsility h3is important implications for the origin of carbon-bearing gases and 
grains in Halley. 

Interestingly, the amount of formaldehyde H2CO emitted by Halley is of the same magnitude. The 
H2CO/H2O ratio obtained from the IKS measurements is ~ 0.04 (Combes et at. 1988; Mumma and Renter 
1989) while a slightly lower value of ~ 0.02 was derived from radio wavelength observations by Snyder {ei al. 
1988). Both the radio wavelength observations and the Giotto NMS data apparently indicate a distributed 
source for at least some of the H2CO (Weaver et al. 1990). However, the IKS measurements supposedly 
refer to a nucleus source of H2CO and not to formaldehyde released from the decomposition of POM in dust 
grains (Combes et al. 1988). FormaJdehyde has also been observed at radio wavelengths in comet Machholz 
(1988J) with a production rate an order of magnitude larger than that in Halley (de Pater ei al. 1990). 

The CH4/H2O ratio in the volatiles emitted by Halley is also in the range of a few percent. Modeling 
of the Ion Mass Spectrometer data of Balsiger et al. (1986) by Allen et al. (1987) yields a CH^/ H2O 
ratio of ~ 0.02. Infrared observations from the Kuiper Airborne Observatory by Drapatz ei al. (1987) gave 
an upper limit for CH4/ H2O < 0.04 while IR observations at Cerro Tololo by Kawara ei al. (1988) gave 
CH4/H2O ~ 0.002- 0.01 for assumed rotational temperatures of 50 to 200 K. The value adopted here for 
the CH4IH2O ratio is ~ 0.01-0.05 from the review by Weaver (1989). Together with the adopted COjH^O 
ratio of ~ 0.02 - 0.07 and the adopted CO2I H2O ratio of ~ 0.03 - 0.04, this leads to CO/CH4 ~ 0.4 - 7.0 

Finally, semiquantitative estimates indicate that ~ 30% (by mass) of the dust emitted by Halley is 
organic material (e.g., see Jessberger et al. 1989) and that ~ 2% (by mass) of the dust is polymerized 
formaldehyde (H2C0)n (Mitchell et al. 1987). Kissel and Krueger (1987) have inferred that highly unsatu- 
rated organic compounds are abundant in the organic grains, but analysis of the dust particle mass spectra 
is still continuing and further information on the composition of the organic fraction may become available 
in the future. 

Nitrogen Compounds 

In contrast to the carbon compounds discussed above, neither A^2 nor NHz has been observed in Halley. 
In both cases the inferred abundances of the parent molecules are deduced from observations of daughter 
molecules presumably produced by photolysis of the parents. 

Allen ei al. (1987) originally derived a 7^/^3/^/20 ratio of ~ 0.01-0.02 from their analysis of the Giotto 
IMS data. However, a subsequent reandysis of the same data by Marconi and Mendis (1988), who unlike 
Allen ei al. (1987) assumed a highly elevated solar UV flux at the time of the Halley spacecraft encounters, 
led to the conclusion that NH3/H2O < 0.01 and indeed may even be zero. However, the total absence 
of NH3 in Halley is extremely unlikely given the Earth-based observations of NH2 (Tegler and Wyckoff 
1989; Wyckoff ei al. 1988, 1989a) which is most plausibly produced by the photodissociation of NH3. In 
fact Tegler and Wyckoff (1989) have derived NH3/H2O = 0.005 ± 0.002 in comet P/Halley. In the absence 
of any compelling evidence for favoring either the analysis of the Giotto IMS data by Allen ei al. (1987) 
or the Earth-based observations of Tegler and Wyckoff (1989), the value adopted here for the NH3/H2O 
ratio is ~ 0.005 - 0.02. A similar range of values has also been adopted by Weaver (1989). Wyckoff et al. 
(1989a) have also observed N H2 emission from comet P/Borrelly, comet Hartley-Good, and comet Thiele 
and derived NH3/H2O ratios of ~ 0.002 for Borrelly, ~ 0.0002 for Hartley-Good, and ~ 0.001 for Thiele. 
However, as Weaver et al. (1990) note, the only direct observation of NH3 in a comet is a marginal detection 
of a radio line in comet IRAS-Araki-Alcock (1983 VII) by Altenhoff et al. (1983). 

Until recently only upper limits were available for the N2/H2O ratio in Halley. However, Wyckoff and 


Theobald (1989) observed A^^ in Halley and calculated a N2/CO ratio ~ 2 x 10"^. Using the adopted value 
of ~ 0.02 - 0.07 for the CO/H^O ratio leads to Ni/H^O ~ 4 x 10"* to 1 x lO"''. A higher JV2///2O ratio 
of ~4 X IQ-'' was derived by WyckofT and Theobald (1989), but their calculation assumed COjH-iO ~ 0.2. 
In any case, the low N2ICO ratio derived by Wyckoff and Theobald (1989) indicates that most of the mass 
28 peak observed in the Giotto NMS is due to CO rather tlian to A'^2- The adopted values for the N^l H^O 
and NH3JH2O ratios correspond to N2/NH3 ~ 0002 - 0.025 while the value for N2/NH3 calculated by 
Wyckoff and colleagues on the basis of their own observational data is ~ 0.1. 

Finally, HCN has also been detected at radio wavelengths in comet Kohoutek (Huebner ei al. 1974) and 
in Halley (Schloerb ei ai 1987; Despois et al. 1986). The derived HCN/H2O ratio is ~ 0.001 in Halley. 
Wyckoff et al. (1989b) used high resolution spectra of CA^ emitted from Halley to derive a ^'^C/^^C ratio of 
65 ± 9, which is significantly lower than the terrestrial value of 89. 

Sulfur Compounds 

Although CS and 5 have been observed frequently in comets (e.g., see Weaver et al. 1981), including 
Halley (Feldman et al. 1986a; Opal ei al. 1986), the only "parent" sulfur-bearing molecule observed in 
comets to-date is S2 (A'Hearn ei al. 1983; Feldman et al. 1984). Weaver (1989) and Weaver ei al. (1990) 
reviewed obserrvations of sulfur-bearing molecules in comets and noted that if all the observed CS in comets 
comes from CS2, then CS2/H2O ~ 0.001 is implied. Similarly, the S2 abundance in comet IRAS-Araki- 
Alcock (1983d) corresponds to S2/H2O ~ 0.001. For reference, the solar abundance of sulfur (Anders and 
Grevesse, 1989) corresponds to S/0 ~ 0.02, which implies that neither CS2 nor 52 is the dominant reservoir 
of sulfur in comets. Finally, Weaver ei al. (1990) note that there is marginal evidence for OCS in the IKS 
spectra of Halley and give a conservative upper limit for OCS in comets corresponding to OCS/H2O ~ 0.01, 
or about 50% of the sulfur solar abundance. 


In principle, the observed molecular abundances and isotopic ratios in Halley can be used to constrain the 
origin of this comet in particular and by inference comets in general. However, in practice, the observational 
data lend themselves to a variety of interpretations which preclude any unambiguous conclusions from being 
reached regarding the origin of Halley. For example, Fegley and Prinn (1989) recently interpreted the 
abundances of CO, CH4, N2, and N H3 in Halley in terms of a two component mixing model involving both 
"oxidized" material from the solar nebula and or the interstellar medium plus smaller amounts of "reduced" 
material from outer planet subnebulae. The rationde for this model is as follows. Kinetic inhibition of the 
CO — ► CH^ and the N2 — ♦ N Hz conversions in the solar nebula leads to a CO, N2 bearing solar nebula. 
However in the higher pressure environments of the giant planet subnebulae (e.g., around Jupiter, Saturn, 
Neptune, etc.) these conversions are both kinetically and thermodynamically favorable so CH4 and NH3 are 
the dominant carbon and nitrogen bearing gases. The early analyses of spacecraft observations of volatiles 
emanating from Halley indicated intermediate CO/CH4 and N2IN Hz ratios which are not representative 
of either the solau- nebula or of giant plzinet subnebulae (or in fact of pristine interstellar material). However 
the intermediate ratios can be obtained by the physical mixing of condensate grains formed in the two 
environments. This mixing could occur during the accretion of cometary nuclei and is analogous to the 
extensive mixing that has occurred among the different chondrite types (e.g., see Wilkening 1977). 

On the other hand, Lunine (1989) and Engei ei al. (1989) interpreted the same data solely in terms 
of a solar nebula origin, except for N Hz- In this model, the intermediate CO/CH4 ratio is ascribed to the 
conversion of CO to C//4 on the surfaces of Fe metal grains in the soleu- nebula. However, not enough N Hz 
can be made to explain the intermediate N2/ N H3 ratio by the same process and a giant molecular cloud 
source is postulated for NH3. It should also be noted that the efficiency of the grain catalyzed CO — CH4 
conversion is controversial (Fegley 1988; Fegley and Prinn 1989) and in fact may not be possible to any 
significant extent in the solar nebula. Finally, Stevenson (1990) has recently interpreted cometatry chemistry 
solely in terms of an interstellar origin "except possibly for a small contamination which is due to catalyzed 
hydrogenation of CO to CH4 and other hydrocarbons". 

Rather than simply recapitulating these diverse proposals, the remainder of this paper will instead 


attempt to quantify as far as possible, by using the observational data base reviewed above, in concert with 
current theoretical models and experimental data, several "key" issues relevant to the origin of comet Halley 
and of comets in general. In particular, the relative importance of equilibrium and disequilibrium processes 
in presolar and solar nebula environments for cometary chemistry will be addressed. "Key" experimental 
and observational measurements which are important for better constraints on cometary origins, including 
important measurements to be made by a comet nucleus sample return mission such as Rosetia, are also 

Mechanisms for Trapping Cometary Volatiles 

One important issue in cometary chemistry is the mechanism for retaining volatiles such as CO, CO2, 
CH4, N2, NH3, etc. in cometary nuclei. These volatiles could have been retained as ices such as CO (solid), 
C02(solid), etc. which would have formed if the nebular temperature was low enough, or they could have 
been retained by absorption/adsorption on the surfaces of water ice grains, or they could have been trapped 
in the water ice crystal structure (clathration). It is important to understand the mechanism for trapping 
volatiles in cometary nuclei because it provides constraints on the formation conditions of comets and also 
has important implications for spacecraft analyses of cometary nuclei. 

For example, condensation of CO and N2 ices, requires temperatures of ~ 20 — 25 K at the typical 
pressures (P ~ lO"*^ bars) expected in the outer solcir nebula (e.g., see Yamamoto 1985). Although some 
investigators have argued against such low temperatures in the comet formation region, it is important 
to note that the observed ortho-to-para ratio in water in Halley was frozen in (or quenched) at a similar 
temperature of about 25 K. On the other hand, trapping of CO and N2 by absorption/adsorption on water 
ice or by clathration in water ice does not require as low a temperature (e.g., see Lunine 1989). Thus 
identifying the form of trapped volatiles in cometary nuclei will provide information about the temperatures 
in the outer so\ai nebula during cometeury formation. 

Similarly, if volatiles were retained as pure ices, the different condensation temperatures of these phases 
may span a considerable range. Again, referring to Yamamoto 's (1985) condensation calculations, water ice 
will condense at 152 K, HCN ice at 95 K, NH3 ice at 78 K, CO2 ice at 72 K, H2CO ice at 64 K, CH^ 
ice at 31 K, CO ice at 25 K, and N2 ice at 22 K for a typical outer solar nebula gas density of ~ 10^^ 
molecules cm~^. Qne consequence of volatile trapping as pure ices may be considerable heterogeneity (both 
radial and spatial) in the cometary nucleus. This could result simply from clumps of more or less volatile ice 
grains accreting together so that one portion of the comet nucleus is then composed of more volatile material 
than another portion or from accretion proceeding more rapidly than nebular cooling so that an onion-skin 
structure with concentric layers of successively more volatile ices results. The observation of outbursts and 
increased emissions of CO2 from Halley (Feldman et al. 1986b) is circumstantial evidence for heterogeneity 
in the nucleus of this comet. On the other hand, volatile retention by sorption on or clathration in water ice 
grains implies a more homogeneous cometary nucleus (both radially and spatially). Circumstantial evidence 
reviewed by Delsemme (1982) indicates that mamy (but not all) cometary nuclei are homogeneous. 

Unfortunately the data from the spacecraft encounters with Halley do not provide firm constraints on 
the volatile retention mechanbms. Again, the circumstantial evidence can be interpreted in several ways. 
For example, the total abundance of CO + CO2 + CH4 apparently emitted from the nucleus of Halley is 
~ 0.06 — 0.16 relative to water. This is consistent with volatile retention as a clathrate which would yield 
a trapped gas/water ratio of ~ 0.17 for the ideal formula of G« 6H2O. (Formaldehyde was excluded from 
this comparison because Davidson (1973) does not list it as forming a clathrate with water ice. However if it 
were included, the abundance ratio relative to water would increase to ~ 0.08— 0.20, which is still partially 
consistent with volatile retention as a clathrate.) But as mentioned above, the CO/CH^ and N2/NH3 
ratios observed for Halley are not representative of the solar nebula and argue against a simple clathrate 
condensation model. Also, the observed outbursts (Feldman ei al. 1986b) argue for a heterogeneous nucleus, 
which is not expected from a clathrate condensation model, but which is expected from a two component 
mixing model involving both ice and clathrate components. 

Since the observational evidence for volatile trapping mechanisms is circumstantial and ambiguous, 
it is instructive to examine volatile trapping from the perspective of the constraints imposed by physical 


chemistry. What caii this tell us about the possibility of volatile retention by clathrate formation? 

Constraints on Clathrate Formation 

Delsemme and Swings (1952) originally proposed that volatiles were retained as clathrates such as 
CH^ • &H2O in order to resolve the problem of the large differences in the vapor pressures of H2O, CH^, 
etc. and the more or less simultaneous appearance of bands of OH, CH , etc. in cometary spectra. During 
the past 4 decades, several other investigators have reiterated the proposal that clathrates are present in 
cometary nuclei (e.g., Miller 1961; Delsemme and Miller 1970; Delsemme and Wenger 1970; Delsemme 1976; 
Sill and Wilkening 1978; and most recently Lunine 1989). However as noted by Lunine and Stevenson 
(1985) "there are no compelling observational data strongly for or against a primordial clathrate component 
to cometary volatiles and (2) no physical cometary phenomena require (or rule out) the presence of clathrates 
in comets." 

The presence of clathrates in comets depends both on the thermodynamic stability of the relevant 
clathrate molecules and on the kinetic favorability of the clathrate formation process. The net thermochem- 
ical reaction responsible for clathrate formation can be schematically represented by: 

G(gas) + 6y/20(ice) = C • 6i/20(solid) (1) 

where G is a gas such as CO, CO1, CH4, N2, H2S, but not N H3 which can form several different hydrates 
(e.g., NH3 • H2O) with water. Assuming that both the stzirting water ice and the final clathrate are pure 
crystalline solids, their thermodynamic activities will be unity and the equilibrium constant expression for 
reaction (1) is given by: 

where the fugacity {fa) can be taken as the partial pressure (Pc) for sufficiently low pressures such that the 
gas behaves ideally. Measured (or estimated) dissociation pressures for different clathrates can then be used 
to calculate their condensation curves as a function of the total assumed nebular pressure and of the partial 
pressures of the enclathrated gas molecules. An example of such a calculation is given in Figure 1 of Fegley 

Although similar calculations have been presented repeatedly in the literature (e.g., see Lewis 1972; Sill 
and Wilkening 1978; Lunine and Stevenson 1985), have been used to argue for the presence of clathrates 
in comets, and to constrain details such as the COJCH^ ratio of the solar nebula gas assumed to be in 
equilibrium with an assumed cometary clathrate (e.g., Lunine 1989), there has been very little (if any) dis- 
cussion of the reliability of the thermodynamic data used to make the calculations. However, the conclusions 
drawn from these calculations, especially detailed inferences about the CO/CH4 ratio of solar nebula gas, 
are dependent on the accuracy and precision of the available thermodynamic data. Just how reliable are the 
clathrate thermodynamic data in the literature? 

The main clathrates of interest for discussions of cometary chemistry are the clathrates of carbon gases 
such as CO, CO2, CH^, and ^2 clathrate. (As previously mentioned, N H^ forms hydrates with water 
instead of clathrate compounds.) For discussions of sulfur chemistry, clathrates formed by H2S and OCS 
would also be of interest. However, the present discussion will focus on clathrates of carbon gases and N2- 

A review of the literature shows that the available data for these clathrates are very limited and are of 
uncertain reliability. Most of the published data for CH^ clathrate are for the reaction 

CHAig^) + 6//20(Iiquid) = CH4 • 6//20(solid) (3) 

and were obtained 40 — 50 years ago when clathrate formation in natural gas pipelines was being studied 
(e.g., Deaton and Frost 1946). Although these data can in fact be used to calculate the thermodynamic 
properties of CH4 clathrate, this has not yet been done. Instead, literature calculations have relied on the 
dissociation pressure measurements of Miller (1961) and Delsemme and Wenger (1970) which are stated to 
be for reaction (l)-formation of C//4 clathrate from CH4 gas and water ice However, neither the existence 


of a clathrate having the composition CH^ • 6H2O nor the establishment of equilibrium was demonstrated 
in either set of measurements. Thus it is uncertain if the measured dissociation pressures are for a clathrate 
or for adsorption/absorption of gas on ice/clathrate mixtures and it is also uncertain if the data obtained 
are relevant to a clathrate having the composition CH4 • 6//2O or some other composition. Furthermore, 
no attempt was made to show that the low temperature data are thermodynamically consistent with the 
high temperature data obtained by other groups (i.e., for clathrate formation from liquid water). In the 
case of CO clathrate, no data were available until the dissociation pressure for the three phase equilibrium 
(clathrate-liquid water-gas) at CC was reported by Davidson ei al. (1987). No data are available on the 
temperature dependence of the dissociation pressure for temperatures of interest for CO clathrate formation 
in the solar nebula (~ 40 - 80/\ ). Although dissociation pressures as a function of temperature have been 
reported for CO2 clathrate (e.g., Miller 1961; Miller and Smythe 1970), the composition of the clathrate 
prepared was not determined and the attainment of equilibrium was not demonstrated in these studies. 
Finally, two different vapor pressure equations (Miller 1961, 1969) have been reported for A^2 clathrate and 
neither is based on dissociation pressure measurements at temperatures relevant for N'2 clathrate formation 
in the solar nebula (~ 40 - SOK). 

A critical assessment of the available data base, including attempts to determine the consistency of the 
low temperature and high temperature dissociation pressure data for several clathrates is underway and will 
quantify the uncertainties in the clathrate condensation calculations. However, laboratory measurements 
of the thermodynamic properties of the cosmochemically important CO, CH4, CO2, and A^2 clathrates at 
temperatures relevant to their formation in the soleir nebula and in giant planet subnebulae are required 
not only to assess the applicability of solar nebula condensation calculations but ako to interpret data 
from the CRAF mission and from a planned comet nucleus sample return mission. The thermodynamic 
properties of interest include the dissociation pressure, enthalpy of formation, and the heat capacity of the 
clathrates. Furthermore, these measurements should (1) adequately characterize the clathrate produced, 
(2) demonstrate the attainment of equilibrium, and (3) consider mixed clathrates as well as one component 
clathrates in order to determine the thermodynamic solid solution properties. In addition to thermodynamic 
measurements, rheological and transport properties such as the viscosity and thermal conductivity should 
also be measured in order to help interpret data from CRAF and a comet nucleus sample return mission. 

However, as noted above, the presence of clathrates in comets also depends on the kinetic favorabihty 
of clathrate formation. Several groups have expressed concerns about the kinetic feasibility of clathrate 
formation in the low temperature, low pressure environment of the outer solar nebula where comets are 
generally believed to have been formed. For example, Lunine and Stevenson (1985) noted that their "work 
adds a theoretical argument against primordial clathrate being a primary component of cometary nuclei, 
if indeed the formation region of these bodies was in the outer (trans-Neptunian) solar nebula: kinetic 
inhibition of clathrate formation would be expected under conditions in the outer solar nebula." (However, 
it is interesting to note that despite this cautionary statement, Lunine (1989) and Engel ei al. (1989) argue 
for clathrate formation in the solar nebula and the presence of solar nebula clathrates in comet P/Halley.) 

The most recent study (Fegley, 1988) of the kinetics of clathrate formation clearly demonstrates the 
inherent difBculties. As Figure 1 in Fegley (1988) illustrates, CO clathrate CO»6H20 does not become stable 
until the temperature in the solar nebula drops to ~ 60A'. At this temperature, Fegley (1988) found that the 
time for 6% of all CO (which is the maximum amount of CO that can be clathrated before exhausting the 
supply of H2O ice) to collide with r = Ifxm spherical ice grains is ~ 4 x lO"* seconds for the solar nebula (P,T) 
profile used in his cedculations. If every collision of a CO molecule with an ice grain led to the formation 
of CO clathrate, this collision time would also be the time required for CO clathrate formation in the solar 
nebula. However, only a small fraction of collisions that possess the necessary activation energy Ea lead to 
chemical reaction. Following the treatment for gas-grain reactions given by Fegley (1988), this fraction is 
given by 

/.■ = u,exp{-EJRT) (4) 

where j/,- is the total number of collisions of CO molecules with all ice grains in each cm^ of the nebula and 
is given by the equation 


u, = 2.635 X 10"(P,/(A/,T)'/2]^ (5) 

where P, is the CO partial pressure in this case, M, is the CO molecular weight, T is tiie temperature 
(Kelvins), and A is the total surface area of all ice grains per each cm'' of the. solar nebula. 

Now in order for the CO clathrate formation time to be < 10'^ seconds, which is the estimated lifetime 
of the solar nebula (e.g., see Cameron 1985), the corresponding activation energy for clathrate formation 
Ea must be < 8 kJ mole "'. Higher activation energies will lead to longer clathrate formation times and 
thus to the kinetic inhibition of clathrate formation in the solar nebula. Fegley (1988) pointed out that 8 kJ 
mole ~' is a low activation energy even by comparison with a facile process such as H F diffusion through 
ice, which has an activation energy of ~ 19 kJ mole"' (Haltenorth and Klinger 1969). For reference, if 
CO clathrate formation has a similar activation energy, the corresponding formation time for clathration of 
r = 1/im spherical, monodisperse ice grains would be ~ 10^' seconds, or about 10"* times longer than the 
age of the solar system. In fact, Miller and Smythe (1970) have derived Ea ~ 24.7 kJ mole ~' for formation 
of CO2 clathrate; if a similar activation energy is required for CO clathrate formation it will certainly be 
kinetically inhibited in the solar nebula. Similar conclusions hold for the formation of A'2 clathrate, which 
becomes thermodynamically feasible at similar temperatures. Thus, unless the formation of CO and A'2 
clathrates is dissimilar to the formation of CO7 clathrate and is a process with essentially no Ea barrier, 
it will probably be kinetically inhibited at the low solar nebula temperatures and pressures where clathrate 
formation is thermodynamically feasible. 

However, as pointed out by Fegley and Prinn (1989), clathrate formation is predicted to be kinetically 
favorable in the higher pressure environments of giant plamet subnebulae. In this case, CH^ is the dominant 
carbon gas and CH^ clathrate formation becomes thermodynamically feasible at higher temperatures. For 
the specific giant planet subnebula (P,T) profile considered by Fegley and Prinn (1989), CH^uQH^O becomes 
stcible at T ~ 95 K and P ~ 10"^ bars. This pressure is approximately 5 orders of magnitude higher than 
the corresponding solar nebula pressure at the CO • 6H2O formation temperature of ~ 60 K and thus leads 
to higher CH4 gas collision rates with water ice grains. For example, the time for 22% of all CH4 (which 
is the maximum amount that can be clathrated before using up all the available water ice) to collide with 
r = Ifim spherical, monodisperse ice grains is only ~ 10~' seconds. In this case, the activation energy 
for formation of CH4 • 6H2O can be as large as 25 kJ mole ~' to have the process take < 10'^ seconds. 
This activation energy is virtually identical to that derived by Miller and Smythe (1970) for CO2 clathrate 
formation. Therefore, the results of these first order calculations predict (in accord with intuition) that 
CO and N2 clathrate formation will be kinetically inhibited in the solar nebula but that CH4 clathrate 
formation will be kinetically feasible in giant plcinet subnebulae. However, as previously stressed by Fegley 
(1988, 1990) and Fegley and Prinn (1989), experimental studies of the kinetics of clathrate formation are 
required for a comprehensive understanding of the kinetic constraints on clathrate formation in the solar 
nebula and in giant planet subnebulae. These experiments should focus on CH4, CO, and N2 clathrates and 
should be suitably designed so that the dependence of the rate on gas partial pressures, ice particle sizes, 
and temperature can be quantitatively measured. 

Origins of Coinetary C-H-O-N-S Compounds 

A closely related issue to the trapping of cometary volatiles is the origin of the different C—H — — N — S 
compounds - both volatiles such as CO, CO2, H2CO, CH4, N2, N H3, HCN, and S2 as well as less volatile 
compounds such as poly oxy methylene POM and the CHON particles. However, once again the available 
observational data can be interpreted in several different ways depending on one's preconceptions and basic 
assumptions. For example, a carbon budget for Halley which takes into account the volatile carbon-bearing 
gases and the carbon contained in the CHON particles indicates that carbon is probably present at its 
solar abundance with the majority of the carbon being in the CHON particles (Delsemme, 1988). This 
carbon mass balance could be interpreted as indicating that both the volatile carbon molecules and the 
less volatile CHON particles originated from the same reservoir which was fractionated into volatile carbon 
gases and involatile organic compounds Alternatively, the carbon mass balance could be regarded as merely 
a coincidence and the volatile species and the CHON particles could be regarded as having separate and 
decoupled origins (e.g., Lunine 1989). 


Prior work in this area (e.g., Fegley and Prinn 1989; Lunine 1989) has focused more on volatile carbon 
and nitrogen gases such as CO, CH^, N2, and NH3 than on other C-H-0-N-S compounds. A 
major result of this work is the conclusion that volatile species in comet Halley have undergone at least some 
chemical reprocessing in environments such as the solar nebula and the subnebulae of the giant planets and 
are not pristine interstellar molecules. The focus in the remainder of this- section will be on the important 
C— H — O— N — S compounds not considered in detail by these investigators and how models for their 
origin and abundance help to constrain the formation conditions for Halley. An underlying assumption in 
this work is that an interstellar origin for different species should not be proposed simply on the basis of 
insufficient knowledge about the diversity and complexity of chemical processes which could have operated 
in nebular environments (solar nebula and giant planet subnebulae) and in the early solar system. 

Abundances of Carbon Compounds 

The observational data reviewed earlier indicate that COjCO^ ~ 0.5 - 2.3, COjCH^ ~ 0.4 - 7.0, and 
COiJCEa ~ 0.6 - 4.0, or in other words roughly equal abundances of these three carbon gases within the 
uncertainties of the observational data. Furthermore, these gases combined account for ~ 25 — 30% of the 
total carbon in Halley with the remaining ~ 70 - 75% being found in the CHON particles (Delsemme 
1988). In this regcird it is interesting to note that calculations by Lewis ei al. (1979) predict that at chemical 
equilibrium in a solar composition gas Pco ~ PcOj ~ PcH, ~ 5 A^r , (A^r is the graphite abundance) at T 
~ 400 K and P ~ 10"^ bars. While it is easy to show that gas phase thermochemistry cannot possibly proceed 
at these low temperatures and pressures, the possibility remains that grain catalyzed reactions may allow 
reactions to proceed at sufficiently rapid rates to approach chemical equilibrium. If one regards the graphite 
predicted by the calculations of Lewis ei al. (1979) as a proxy for orgzmic matter, then grain catalyzed 
thermochemistry is a possible explanation for the observed abundances of carbon gases and involatile carbon 
compounds in Halley. Indeed, on the basis of calculations by Fegley (1988), Lunine (1989) and Engel ei al. 
(1989) have suggested that the abundances of CO, CO2, and CH4 in Halley are the result of grain catalyzed 
thermochemical reactions, although they regard the involatile organic material as having a separate and 
decoupled origin from the carbon gases and have not discussed its origin in the light of the work by Lewis 
ei al. (1979). 

Several different experimental and observational results are in favor of a grain catalyzed origin for boih 
the carbon gases and the involatile organic material in Halley. Extensive industrial experience with the 
production of synthetic fuels from CO + H2 via Fischer-Tropsch reactions (e.g., see Dry 1981; Biloen and 
Sachtler 1981) indicates that Fe-based materiails are good catalysts for these reactions. Since Fe is the 
third most abundaint rock-forming element (after Mg and Si) in solar composition material (Anders and 
Grevesse 1989) f e-bearing grains are expected to be the most abundant and active catalyst present in the 
solar nebula. Also, depending on the exact experimental conditions, CO2 and CH4 may also be produced in 
addition to more complex organic compounds as a result of Fe-catalyzed Fischer-Tropsch reactions (Vannice 
1982; Krebs ei al. 1979; Dry et al. 1972). Furthermore, laboratory studies of Fischer-Tropsch-type reactions 
by Anders and coworkers (e.g., see Studier ei al. 1968; Hayatsu and Anders 1981) have produced organic 
compounds similar to those observed in meteorites. FinzJly, involatile organic matter has been observed in 
association with Fe-beeiring phases such as Fe — Ni alloy, carbides, and oxides in chondritic interplanetary 
dust particles (Bradley ei al. 1984,1989), at least some of which are believed to be samples of cometary dust 
(Bradley ei al. 1989, and references therein). 

On the other hand, there are also significant arguments against efficient Fe-grain catalysis of carbon 
gas interconversions and organic compound synthesis at low temperatures in the solar nebula. As Fegley 
(1988) has pointed out, the temperature range over which Fe-grain catalysis is possible is limited at high 
temperatures by evaporation to Fe(gas) and at low temperatures either by the formation of FeS coatings 
at ~ 680 K or by "rusting" to form magnetite Fe304 at ~ 370 - 400 K. In fact, petrographic studies of the 
unequilibrated ordinary chondrites, which are generally believed to preserve a record of nebular processes, 
reveal Fe5-rimmed metal grains, which may be nebular condensates (e.g., Rambaldi ei al. 1980; Rambaldi 
and Wasson 1981,1984). Also, as Fegley (1988) and Fegley and Prinn (1989) have noted, the available 
kinetic data for the Fe-catalyzed CO — + CH4 conversion are for ultra-clean, high-purity Fe grains which 
very probably do not apply to Fe grains in the solar nebula which will be contaminated by several elements 


such as phosphorus, sulfur, carbon, hydrogen, nitrogen, and oxygen at T > 1000 K in the solar nebula 
(Kozasa and Hasegawa 1988; Fegley and Lewis 1980). Furthermore, the studies of Fischer-Ttopsch reactions 
which have indicated similarities between the laboratory products and the organic compounds found in 
meteorites have generally been done under conditions such as high COjHi ratios which may not be relevant 
to the solar nebula where the CO/H2 ratio is ~ 7 x 10"'' if all carbon is present as CO. 

In other words, the situation is ambiguous and again experimental work is needed to resolve the un- 
certainties. Of particular interest are quantitative experiments in which well characterized Fe catalysts and 
near solar CO/H2 ratios are used to study the product yields, distribution, and formation rate as a function 
of parameters such as temperature, total pressure, CO/H2 ratio, catalyst type and treatment, etc. As Fegley 
(1990) has noted, these studies have potential applications in many areas of planetary science such as the 
origin of the dark material, presumably organic matter on some outer planet satellites and asteroids. 

Hydrogen Cyanide 

As mentioned earlier, HCN has been detected in comet Kohoutek (Huebner ei al. 1974) and in Halley 
(Schloerb ei al. 1987; Despois ei al. 1986) where the derived HCN/H2O ratio is ~ 0.001. This ratio is many 
orders of magnitude larger than the predicted HCN/H2O ratio al chemical equilibrium in the outer solar 
nebula {HCN /H2O ~ 10"^^ at T ~ 100 K) and is direct evidence for a potent disequilibrating mechanism 
in the outer solar nebula. Hydrogen cyanide is an important precursor for the abiotic synthesis of complex 
organic molecules (Oro and Kimball 1961; Abelson 1966) and it is of interest to determine how this much 
HCN could have been produced. 

One possibility is the quenching of thermochemical reactions involving HCN in the inner solar nebula 
and the outward radieJ mixing of this HCN-beai'ing gas to the comet formation region. Calculations by 
Fegley presented in Prinn and Fegley (1989) predict quenching of the homogeneous gas-phase HCN —> N2 
conversion in the inner solar nebula at T ~ 1460 K where HCN/N2 ~ 10"^. This corresponds to a 
HCN/H2O ratio of ~ 10"^ *, or about a factor of 6300 too low to explain the Halley observations. Grain 
catalysis of the HCN — ♦ A^2 conversion may lead to a lower quench temperature, however; because the HCN 
abundance is decreasing rapidly with temperature an even lower HCN/H2O ratio would then result. Another 
possibility is the quenching of the HCN — > NH3 conversion in the giant planet subnebulae followed by HCN 
condensation onto grains and the mixing of these grains into the comet formation region after the nebular 
gas has dissipated. Again, calculations by Fegley presented in Prinn and Fegley (1989) predict quenching 
of the HCN -» NH3 conversion at T ~ 1220 K where HCN/NH3 ~ 10"^ for a model Jovian subnebula 
(P,T) profile. However, the corresponding HCN/H2O ratio of ~ 10"®'' is also about a factor of 2500 too 
low. Again, grain catalysis of the HCN — > NH3 conversion will result in a lower quench temperature and a 
lower HCN abundance. Therefore, the results of thermochemical kinetic calculations show that the HCN 
abundance in Halley cannot be produced by quenching either the HCN — ♦ N2 conversion in the inner solar 
nebula or the HCN — • NH3 conversion in a giant planet subnebula. 

Another possibility for explaining the HCN abundance is to look at the effect of lightning induced shock 
chemistry in the solar nebula. Nebular lightning has been suggested as a mechanism for chondrule formation 
(e.g., Cameron 1966; Whipple 1966) but may also have distinctive chemical consequences. As Fegley has 
noted in Prinn and Fegley (1989), the high temperatures (several times 10^ K) reached in lightning discharges 
lead to increasing degrees of molecular dissociation, atomization, and ionization with increasing temperatures. 
The recombination of these simple fragments during the rapid cooling of the shocked gas leads initially to 
the production of more complex fragments, then to thermally stable molecules such as HCN . Sufficiently 
rapid cooling quenches these stable molecules at their high temperature abundances, which are generally 
enhanced over their equilibrium abundances at much lower temperatures. Thus, lightning is a potentially 
significant source of disequilibrium products, especially in the cool, thermochemically inactive regions of the 
nebula near and beyond the water ice condensation point (T ~ 150 - 200 K). 

Calculations by Fegley in Prinn and Fegley (1989) model lightning induced shock chemistry in the 
solar nebula and in giant planet subnebulae as adiabatic shock heating and predict that the maximum 
HCN concentrations in these two different environments occur near temperatures of 3000 — 4000 K where 
~ (0.2 — 6) X lO^^HCN molecules are formed per mole of shocked gas. This corresponds to maximum 


conversions of ~ 0.3% (for the solar nebula) and ~ 6.5% (for the Jovian subnebula) of the total nitrogen 
abundance into HCN . This compares favorably with Halley where the adopted N H3/H2O ratio of ~ 
0.005 - 0.02 and the HCN/H2O ratio of ~ 0.001 correspond to a HCN/NH3 ratio of ~ 0.05 - 0.2. Thus, 
the HCN observed in Halley can be explained on the basis of lightning induced shock chemistry in either 
the solar nebula or in a giant planet subnebula, but only a small dilution of the shocked gas by unshocked 
gas (by < a factor of 20 for a solar nebula source and by < a factor of 70 for a subnebula source) is implied. 

A potential problem with a lightning induced shock chemistry source for the HCN in Halley is the 
^^C/^^C isotopic ratio of 65 ± 9 derived for CN observed in Halley (Wyckoff et al. 1989b). A similarly low 
ratio is implied for HCN which is the probable, although possibly not the only, parent for the observed CN. 
At the high temperatures where the HCN abundance will be quenched in the rapidly cooling shocked gas, 
isotopic exchange with the major carbon-bearing gases via reactions exemplified by 

"C/f4 + H"CN = i2ci/4 + H'^^CN (6) 

in giant planet subnebulae and via reactions exemplified by 

^^CO + H^^CN = ^^C0 + H'^CN (7) 

in the solar nebula will lead to isotopic equilibration because the thermochemical isotopic fractionation factors 
are unity at high temperatures (e.g., see the tabulation by Richet et al. 1977). However, the ^^C/^^C ratio 
implied for the HCN in Halley is significantly lower than the terrestrial value of 89. But, is it also different 
from outer solar system values? 

Unfortunately, this question can only be partially answered. A ^^C/^^C ratio of 89li| was derived for 
CH4 on Saturn from Earth-based observations by Combes et al. (1977) but three different carbon isotopic 
ratios have been reported for Jupiter. Voyager 1 infrared observations oiCH^ gave a ^^C/^^C ratio of 160^55 
(Courtin et al. (1983), Earth-based observations oiCH^ gave 110 ± 35 (Fox et al. 1972), 70^^ (DeBergh 
ei al. 1976), and 891}^ (Combes et al. 1977), and finally Earth-based observations of C2H2 gave 20tfo 
(Drossart et al. 1985). No carbon isotopic ratios are presently available for either Uranus or Neptune. 

The only data available for "icy" bodies in the outer solar system are values for four other comets 
because no carbon isotopic ratios are available for Titan or other outer planet satellites. As summarized by 
Wyckoff et al. (1989b), the cometary values are 70± 15 for Ikeya 1963 I (Stawikowski and Greenstein 1964), 
100 ± 20 for Tago-Sato-Kosaka 1969 IX from Owen (1973), llStfg and 1351" for Kohoutek 1973 XII from 
Banks et al. (1974), and 100^1° for Kobayashi-Berger-Milon 1975 IX from Vanysek (1977). All of these 
other cometary carbon isotopic ratios are derived from observations of C2- 

Taken at face value, the reported ^"^C/^^C ratios for outer so\zi system bodies range from ~ 20 to ~ 160 
and the value for Halley is not inconsistent with this wide range. Although Wyckoff et al. (1989b) conclude 
that the low carbon isotopic ratio of ~ 65 for CN from Halley appears to exclude the Uranus- Neptune 
region as a condensation site for this comet, the limited data which are available-none of which are in fact 
for Uranus or Neptune themselves-do not strongly support this conclusion. Indeed, the observation of Jovian 
C2H2 having a ^^C/'^^C ratio of ~ 20 (Drossart et al. 1985) illustrates the danger of using an apparently 
anomalous isotopic ratio as the only basis for assigning an interstellar origin to a molecule. 


The D/H ratio of 0.6 - 4.8 x 10"'' determined for water vapor in Halley (Eberhardt ei al. 1987b) is 
also of interest for several reasons. First, this range is comparable to several other solar system D/// ratios 
including the terrestrial value of ~ 1.6 x lO"'', the value of 1.51J | x 10"" for Titan (Coustenis et al. 1989), 
the value of 9.01^ 5 x lO"^ for Uranus (DeBergh et al. 1986), and the value of ~ 1.5 x lO"'' for Neptune 
(Lutz ei al. 1990). Second, the Halley value is at least twice as large as the estimated primordial value of 
(0.3 ± 0.1) x 10"^ (Anders and Grevesse 1989). Last, the D/H value, like the ortho-to-para ratio in the 
water, is a potential cosmothermometer. However, in this case the thermometer applies to the temperature 
where isotopic exchange last occured rather than the temperature where nuclear spins last equilibrated. 


This isotopic exchange temperature can be calculated by assuming exchange with H2 having the pri- 
mordial D/ H ratio via the net reaction 

H2 + HDO-HD+H2O (8) 

and using the thermochemical isotopic fractionation factors tabulated by Richet el al. (1977). The lower end 
of the range of D/H values derived for Halley {D/H ~ 0.6 x 10"'') corresponds to a temperature of ~ 465 
K, while the upper end of the range of values {D/H ~ 4.8 x lO""*) corresponds to a temperature of ~ 148 
K. Although the classical picture of isotopic exchange envisions increasing deuterium enrichment in hydrides 
such as H2O, CH^, etc. with decreasing temperature, the rates of both gas-phase and grain catalyzed D/H 
exchange reactions are so slow that they take longer than the estimated lifetime of the solar nebula and 
are therfore kinetically inhibited(e.g., see Grinspoon and Lewis 1987). Instead it is better to regard these 
isotopic exchange temperatures as the maximam temperatures at which the water in Halley last exchanged 
deuterium with nebular H^- If this view is taken, then the exchange process is viewed as a back-reaction 
in which Z?-rich water is losing deuterium to the surrounding nebular //j. This process may occur as a 
consequence of reactions driven by thermochemistry (e.g., in the subnebulae surrounding the giant planets) 
or as a consequence of reactions driven by the interstellar radiation field impinging on the outer layers of the 
primitive solar nebula. The latter possibility is essentially the reverse of the scheme proposed by Yung ei al. 
(1988). However, if the Halley D/H and ortho-to para data are taken at face value, the isotopic exchange 
process either cannot affect nuclear spin exchange or must occur prior to nucleju- spin equilibration at a 
significantly lower temperature. Again, this is an area where both further theoretical work and experimental 
studies are required to resolve the existing uncertainties. 

Nebular Pkotocheimstry and H2CO and S-2 

Finally, the presence of H2CO in Halley and other comets and the presence of S2 in comet IRAS-Araki- 
Alcock (1983d) may also provide evidence for nebular photochemistry driven by the interstellar radiation 
field. Several authors (e.g., Wood and Chang 1985; Prinn and Fegley 1989; Yung et al. 1988) have speculated 
on the importance of solar nebula photochemical processes, but aside from the deuterium enrichment model 
of Yung et al. (1988), no quantitative photochemical calculations are available. However Grim and Greenberg 
(1986) have shown that Sj can be produced from the ultraviolet irradiation of sulfur containing ices and it 
is interesting to ask if this process-which they suggest took place in the interstellar medium-could also have 
occurred in the outer regions of the solar nebula where the only important gaseous opacity sources are H2 
and CO (because all other gases are frozen out, absorb at shorter wavelengths, or have smaller abundances). 
Although H2CO may be produced by the grain catalyzed Fischer-Tropsch-type reactions discussed earlier, it 
is also a photochemical product of CO -1- H2O ice irradiation (e.g., Tielens, 1983). Both theoretical models 
of nebular photochemistry and experimental studies of ultraviolet irradiation of appropriate "icy" grains 
would appear to be fruitful areas for further research. 


The data obtained from the recent Earth-based, Earth-orbital, and spacecraft studies of comet P/Halley 
have expanded greatly our knowledge of the chemistry of comets. However, many of the observational data 
which have been obtained lend themselves to a variety of interpretations which preclude any unambiguous 
conclusions from being reached regarding the origin of Halley in particular and of comets in general. Prior 
work (e.g., Fegley and Prinn 1989; Lunine 1989) which has focused on more volatile carbon and nitrogen 
gases such as CO, CH4, N2, and NH3 has led to the conclusion that volatile species in Halley have undergone 
at least some chemciai reprocessing in environments such as the solar nebula and the subnebulae of the giant 
planets and are not pristine interstellar molecules. However, a comprehensive overview of both these volatiles 
as well as the other species observed in Halley indicates that the picture may not be as simple as initially 
believed and the experimental and theoretical investigations suggested in this paper will probably contribute 
to a more complete understanding of the chemical processes involved in cometary formation. 



My cosmochemistry research is presently supported by the Max-PIanck-Institut fiir Chemie. I want to 
thank the students and staff of the MPI for their help and support. 


Abelson, P.H., Chemical events on the primitive Earth. Proc. Nail Acad. Sci. (USA) 55, 1365-1372 

A'Hearn, M.F., Feldmain, P.D., and Schleicher, D.G., The discovery of 52 in comet IRAS-Araki-Alcock 

1983d. Asirophys. J. Lett. 274, L99-L103 (1983). 
Anders, E. eind Grevesse, N., Abundances of the elements: Meteoritic and solar. Geochim. Cosmochim. 

Acta 53, 197-214. (1989). 
Allen, M., Delitsky, M., Huntress, W., Yung, Y., Ip, W.-H., Schwenn, R., Rosenbauer, H., Shelley, E., 

Balsiger, H., and Geiss, J., Evidence for methane and ammonia in the coma of comet P/Halley. 

Asiron. Asirophys. 187, 502-512 (1987). 
Altenhoff, W.J. et ai, Radio observations of comet 1983d. Asiron. Asirophys. 187, 502-512 (1987). 
Balsiger, H. et ai. Ion composition and dynamics at comet Halley. Nature 321, 330-334 (1986). 
Biloen, P. and Sachtler, W.M.H., Mechanism of hydrocarbon synthesis over Fischer-Tropsch catalysts. 

in Advances in Catalysis, ed. D.D. Eley, H. Pines, and P.B. Weisz, Academic Press, NY, pp. 

165-216 (1981). 
Bradley, J. P., Brownlee, D.E., and Fraundorf, P., Carbon compounds in interplanetary dust: evidence 

for formation by heterogeneous catalysis. Science 233, 56-58 (1984). 
Bradley, J. P., Sandford, S.A., and Walker, R.M., Interplanetary dust particles, in Meteorites and the 

Early Solar System, ed. J.F. Kerridge and M.S. Matthews, University of Arizona Press, Tucson, 

AZ, pp. 861-895 (1989). 
Cameron, A. G.W., The accumulation of chondritic material. Earth Planet. Sci. Lett. 1,93-96(1966). 
Cameron, A.G.W., Formation and evolution of the primitive solar nebula, in Proiostars and Planets 

II, ed. D.C. Black and M.S. Matthews, University of Arizona Press, Tucson, AZ, pp. 1073-1099 

Combes, M., Maillard, J. P., and DeBergh, C, Evidence for a telluric value of the ^^C/^^C ratio in the 

atmosphere of Jupiter and Saturn. Asiron. Asirophys. 61, 531-537 (1977). 
Combes, M. et ai. The 2.5 — 12 //m spectrum of comet Halley from the IKS- Vega experiment. Icarus 

76, 404-436 (1988). 
Courtin, R., Gautier, D., Marten, A., and Kunde, V., The ^^C/^^C ratio in Jupiter from the Voyager 

infrared investigation. Icarus 53, 121-132 (1983). 
Coustenis, A., Bezard, B., and Gautier, D., Titan's atmosphere from Voyager infrared observations. 

II. The CH3D abundance and the D/H ratio from the 900 — 1200 cm"' spectral region. Icarus 

82, 67-80 (1989). 
Danks, A.C., Lambert, D.L., and Arpigny, C, The ^^C/^^C ratio in comet Kohoutek (1973f). Asiro- 
phys. J. 194, 745-751 (1974). 
Davidson, D.W., Clathrate hydrates, in Water: A Comprehensive Treatise, ed. F. Franks, Plenum 

Press, NY, voL2, pp. 115-234 (1973). 
Davidson, D.W. ei ai, A clathrate hydrate of carbon monoxide. Nature 328, 418-419 (1987). 
Deaton, W.M. and Frost, E.M., Jr., Gas Hydrates and Their Relation to the Operation of Natural- Gas 

Pipe Lines, U.S. Bureau Mines Monograph 8 (1946). 
DeBergh, C, Maillard, J. P., Lecacheux, J., and Combes, M., A study of the Zt/z — CHji region in a high 

resolution spectrum of Jupiter recorded by Fourier transform spectroscopy. Icarus 29, 307-310 

DeBergh, C, Lutz, B.L., Owen, T., Brault, J., and Chauville, J., Monodeuterated methane in the 

outer solar system. II. Its detection on Uranus at \.Qfim. Asirophys. J. 311, 501-510 (1986). 
Delsemme, A.H., Chemical nature of the cometaxy snows. Mem. Soc. Roy. Sci. Liege IX, 135-145 



Delsemme, A.H., Chemical composition of cometary nuclei, in Comets, ed. L. L. Wilkening, University 

of Arizona Press, Tucson, AZ, pp. 85-130 (1982). 
Delsemme, A.H., The chemistry of comets. Phil. Trans. R. Soc. Land. 325 A, 509-523 (1988). 
Delsemme, A.H. and Miller, D.C., Physico-chemical phenomena in comets-II. Gas adsorption in the 

snows of the nucleus. Planet. Space Sci. 18,717-730(1970). 
Delsemme, A.H. and Wenger, A., Physico-chemical phenomena in comets-I. Experimental study of 

snows in a cometary environment. Planet. Space Sci. 18, 709-715 (1970). 
Delsemme, A.H. and Swings, P., Hydrates de gaz dans les Noyaux Cometaires et les Grains Interstel- 

laires. Annates d'Asirophys. 15, 1-6 (1952). 
dePater, I., Palmer, P., and Snyder, L.E., A review of radio interfermetric imaging of comets, preprint 

Despois, D. et al., Observations of hydrogen cyanide in comet Halley. Astron. Astrophys. 160, L11-L12 

Drapatz, S., Larson, H.P., and Davis, D.S., Search for methane in comet P/Halley. Astron. Astrophys. 

187, 497-501 (1987). 
Drossart, P., Lacy, J., Serabyn, E., Tokunaga, A., Bezard, B., and Encrenaz, T., Detection of '^C^^C/fj 

on Jupiter at 13//m. Astron. Astrophys. 149, L10-L12 (1985). 
Dry. M.E., The Fischer-Tropsch synthesis, in Catalysis Science and Technology, ed. J.R. Anderson 

and M. Boudart, Springer-Verlag, Berlin, pp. 159-255 (1981). 
Dry. M.E., Shingles, T., and Boshoff, L.J., Rate of the Fischer-Tropsch reaction over iron catalysts. 

J. Catalysis 25, 99-104 (1972). 
Eberhardt, P. et al.. The CO and A^2 abundance in comet P/Halley. Astron. Astrophys. 187, 481-484 

Eberhardt, P. et al.. The D/H ratio in water from comet P/Halley. Astron. Astrophys. 187, 435-437 

Engel, S., Lunine, J.L, and Lewis, J.S., Solar nebula origin for volatile gases in Halley's comet. Icarus, 

in press (1989). 
Fegley, B., Jr., Cosmochemical trends of volatile elements in the solar system, in Workshop on the 

Origins of Solar Systems, ed. J.A. Nuth and P. Sylvester, LPI Technical Report No. 88-04, pp. 

51-60 (1988). • 
Fegley, B., Jr., The applications of chemical thermodynamics and chemical kinetics to planetary at- 
mospheres research, in Proceedings of the First International Conference on Laboratory Research 

for Planetary Atmospheres, NASA CP, in press (1990). 
Fegley, B., Jr. and Lewis, J.S., Volatile element chemistry in the solar nebula: Na, K, F, CI, Br, and 

P. Icarus 41 439-^55 (1980). 
Fegley, B., Jr. and Prinn, R.G., Solar nebula chemistry. Implications for volatiles in the soleir system. 

in The Formation and Evolution of Planetary Systems ed. H.A. Weaver and L. Danly, Cambridge 

University Press, Cambridge, England, pp. 171-211 (1989). 
Feldman, P.D., A'Hearn, M.F., and Millis, R.L., Temporal and spatial behavior of the ultraviolet 

emissions of comet IRAS-Araki-Alcock 1983d. Astrophys. J. 282, 799-802 (1984). 
Feldman, P.D. et al., lUE observations of comet Halley: Evolution of the UV spectrum between 

September 1985 and July 1986. in Proc. 20th ESLAB Symposium on the Exploration of Halley's 

Comet, ESA SP-250, pp. 325-328 (1986a). 
Feldman, P.D. et al.. Is CO2 responsible for the outbursts of comet Halley? Nature 324, 433-436 

Festou, M.C. et al., lUE observations of comet Halley during the Vega and Giotto encounters. Nature 

321, 361-363 (1986). 
Fox, K., Owen, T., Mantz, A.W., and Rao, K., A tentative identification of ^^CHa and an estimate of 

^^C/^^C in the atmosphere of Jupiter. Astrophys. J. 176, L81-L84 (1972). 
Grim, R.J. A. and Greenberg, J.M., Photochemical studies of 52 formation and their implications on the 

source and evolution of comets, in The Comet Nucleus Sample Return Mission Proc. Workshop 

Canterbury UK, ESA SP-249, pp. 143-151 (1986). 


Grinspoon, D.H. and Lewis, J.S., Deuterium fractionation in the presolar nebula: Kinetic limitations 

on surface catalysis, /cants 72, 430-436 (1987). 
Haltenorth, H. and Klinger, J., Diffusion of hydrogen fluoride in ice. in Physics of Ice, ed. N. Riehl, 

B. Bullemer, and H. Engelhardt, Plenum Press, NY, pp. 579-584 (1969). 
Hayatsu, R. and Anders, E., Organic compounds in meteorites and theif origins. Topics in Current 

Chemistry 99, 1-39 (1981). 
Huebner, W.F., First polymer in space identified in comet Halley. Science 237, 628-630 (1987). 
Huebner, W.F., Boice, D.C., and Sharp, CM., Polyoxymethylene in comet Halley. Asirophys. J. Lett. 

320, L149-L152 (1987). 
Huebner, W.F., Snyder, L.E., and Buhl, D., HCN radio emission from comet Kohoutek (1973f). Icarus 

23, 580-584 (1974). 
Jessberger, E.K., Kissel, J., and Rahe, J., The composition of comets, in Origin and Evolution of 

Planetary and Satellite Atmospheres ed. S.K. Atreya, J.B. Pollack, and M.S. Matthews, University 

of Arizona Press, Tucson, AZ, pp. 167-191 (1989). 
Kawaxa, K., Gregory, B., Yamamoto,T., and Shibai, H., Infrared spectroscopic observation of methane 

in comet Halley. Asiron. Asirophys. 207, 174-181 (1988). 
Kissel, J. and Krueger, F.R., The organic component in dust from comet Halley as measured by the 

PUMA mass spectrometer on board Vega 1. Nature 326, 755-760 (1987). 
Kozasa, T. and Hasegawa, H., Formation of iron-bearing materials in a cooling gas of solar composition. 

Icarus 73, 180-190(1988). 
Krankowsky, D. et al., In situ gas and ion measurements at comet Halley. Nature 321, 326-329 (1986). 
Krebs, H.J., Bonzel, H.P., and Gafner, G., A model study of the hydrogenation of CO over polycrys- 

talline iron. Surface Sci. 88, 269-283 (1979). 
Larson, H.P., Weaver, H.A., Mumma, M.J., and Drapatz, S., Airborne infrared spectroscopy of comet 

Wilson (19861) and comparisons with comet Halley. Astrophys. J. 338, 1106-1114 (1989). 
Lewis, J.S., Low temperature condensation from the solar nebula. Ica7~us 16, 241-252 (1972). 
Lewis, J.S., Bajshay, S.S., and Noyes, B., Primordial retention of carbon by the terrestrial plauiets. 

Icarus 37, 190-206(1979). 
Lunine, J.L, Primitive bodies: Molecular abundances in comet Halley as probes of cometary formation 

environments, in The Formation and Evolution of Planetary Systems ed. H.A. Weaver and L. 

Danly, Cambridge University Press, Cambridge, England, pp. 213-242 (1989). 
Lunine, J.L and Stevenson, D.J., Thermodynamics of clathrate hydrate at low and high pressures with 

application to the outer solar system. Astrophys. J. Suppl 58, 493-531 (1985). 
Lutz, B.L., Owen, T., and DeBergh, C, Deuterium enrichment in the primitive ices of the protosolar 

nebula. Icarus, in press (1990). 
Marconi, M.L. and Mendis, D.A., On the ammonia abundance in the coma of Halley's comet. Astro- 
phys. J. 330, 513-517 (1988). 
Miller, S.L., The occurrence of gas hydrates in the solar system. Proc. Nat. Acad. Sci. USA 47, 

1798-1808 (1961). 
Miller, S.L., Clathrate hydrates of air in antarctic ice. Science 165, 489-490 (1969). 
Miller, S.L. and Smythe, W.D., Carbon dioxide clathrate in the Martian ice cap. Science 170, 531-533 

Mitchell, D.L. et al., Evidence for chain molecules enriched in carbon, hydrogen, and oxygen in comet 

Halley. Science 237, 626-628 (1987). 
Mumma, M.J. and Reuter, D., On the identification of formaldehyde in Halley's comet. Astrophys. J., 

in press (1989). 
Mumma, M.J., Blass, W.E., Weaver, H.A., and Larson, H.P., Measurements of the ortho-para ratio 

and nuclear spin temperature of water vapor in comets Halley and Wilson (19861) and implications 

for their origin and evolution, in The Formation and Evolution of Planetary Systems: A Collection 

of Poster Papers, ed. H.A. Weaver, F. Paresce, and L. Danly, STScI publication, pp. 157-168 

Mumma, M.J., Weaver, H.A., and Larson, H.P., The ortho-para ratio of water vapor in comet P/Halley. 

Astron. Astrophys. 187, 419-424 (1987). 


Mumma, M.J., Weaver, H.A., Larson, H.P., Davis, D.S., and Williams, M., Detection of water vapor 

in Halley's comet. Science 232, 1523-1528 (1986). 
Opal, C.B., McCoy, R.P., and Carruthers, G.R., Far ultraviolet objective spectra of comet P/Halley 

from sounding rockets, in Proc. 20ih ESLAB Symposiitm on ike Exploration of Halley's Comet 

ESA SP-250, pp. 425-430 (1986). 
Oro, J. and Kimball, A. P., Synthesis of purines under possible primitive Earth conditions. I. Adenine 

from hydrogen cyanide. Arch. Biockem. Biophys. 94, 217-227 (1961). 
Owen, T., The isotope ratio ^"^Cj^^C in comet Tago-Sato-Kosaka 1969g. Astrophys. J. 184, 33^3 

Prinn, R.G. and Fegley, B., Jr., Solar nebula chemistry: Origin of planetary, satellite, and cometciry 

volatiles. in Origin and Evolution of Planetary and Satellite Atmospheres, ed. S.K. Atreya, J.B. 

Pollack, and M.S. Matthews, University of Arizona Press, Tucson, AZ, pp. 78-136 (1989). 
Rambaldi, E.R. and Wasson, J.T., Metal and associated phases in Bishunpur, a highly unequilibrated 

ordinary chondrite. Geochim. Cosmochim. AciaAb, 1001-1015(1981). 
Rambaldi, E.R. and Wasson, J.T., Metal and associated phases in Krymka and Chainpur: Nebular 

formational processes. Geochim. Cosmochim. >lc<a 48, 1885-1897 (1984). 
Rambaldi, E.R., Sears, D.W., and Wasson, J.T., 5j-rich Fe - Ni grains in highly unequilibrated 

chondrites. A'aiure 287, 817-820 (1980). 
Richet, P., Bottinga, Y., and Javoy, M., A review of hydrogen, carbon, nitrogen, oxygen, sulphur, and 

chlorine stable isotope fractionation among gaseous molecules. Ann. Rev. Earth Planet. Sci. 5, 

Schloerb, F.P., Kinzel, W.M., Swade, D.A., and Irvine, W.M., Observations of HCN in comet P/Halley. 

Astron. Astrophys. 187, 475-480 (1987). 
Sill, G.T. and Wilkening, L.L., Ice clathrate as a possible source of the atmospheres of the terrestrial 

planets. Icarus 33, 13-22 (1978). 
Stawikowski, A. and Greenstein, J.L., The isotope ratio C^^/C^"^ in a comet. Astrophys. J. 140, 

1280-1291 (1964). 
Stevenson, D.J., Chemical heterogeneity and imperfect mixing in the solar nebula. Astrophys. J. 348, 

Stewart, A.I.F., Pioneer Venus measurements of H, O, and C production in comet P/Halley near 

perihelion. Astron. Astrophys. 187, 369-374 (1987). 
Studier, M.H., Hayatsu, R., and Anders, E., Origins of organic matter in early solar system-I. Hydro- 
carbons. Geochim. Cosmochim. Acta32, 151-173(1968). 
Tegler, S. and Wyckoff, S., NH2 fluorescence efficiencies and the NH3 abundance in comet Halley. 

Astrophys. J. 343, 445-449 (1989). 
Tielens, A.G.G.M., Surface chemistry of deuterated molecules. Astron. Astrophys. 119, 177-184 

Vannice, M.A., Catalytic activation of carbon monoxide on metal surfaces, in Catalysis Science and 

Technology, ed. J.R. Anderson and M. Boudart, Springer-Verlag, Berlin, pp. 139-198 (1982). 
Vanysek, V., Carbon isotope ratio in comets and interstellar medium, in Comets, Asteroids, and 

Meteorites: Interrelations, Evolution, and Origins, ed. A.H. Delsemme, University of Toledo 

Press, Toledo, OH, pp. 499-503 (1977). 
Weaver, H.A., The volatile composition of comets, in Highlights of Astronomy 8, 387-393 (1989). 
Weaver, H.A., Feldman, P.O., Festou, M.C., A'Hearn, M.F., and Keller, H.U., lUE observations of 

faint comets. Icarus 47, 449-463 (1981). 
Weaver, H.A., Mumma, M.J., and Larson, H.P., Infrared spectroscopy of cometary parent molecules. 

in Comets in the Post-Halley Era ed. R. Newburn and J. Rahe, Kluwer Academic Publishers, in 

press (1990). 
Whipple, F.L., Chondrules: suggestions concerning their origin. Science 153, 54-56 (1966). 
Wilkening, L.L., Meteorites in meteorites: evidence for mixing among the asteroids, in Comets, 

Asteroids, Meteorites: Interrelations, Evolution, and Origins, ed. A.H. Delsemme, University of 

Toledo Press, Toledo, OH, pp. 389-396 (1977). 


Wood, J. A. and Chang, S. eds.. The Cosmic History of ike Biogenic Elements and Compounds, NASA 

SP-476 (1985). 
Woods, T.N., Feldman, P.D., Dymond, K.F., and Sahnow, D.J., Rocket ultraviolet spectroscopy of 

comet Halley and abundance of carbon monoxide and carbon. Nature 324, 436-438 (1986). 
Wyckoff, S., Lindholm, E., Wehinger, P.A., Peterson, B.A., Zucconi, J.M., and Festou, M.C., The 

i^C/i^C abundance ratio in comet Halley. Astrophys. J. 339, 488-500 (1989b). 
Wyckoff, S. and Theobald, J., Molecular ions in comets. Adv. Space Res. 9(3), 157-161 (1989). 
Wyckoff, S., Tegler, S., and Engel, L., Ammonia abundances in comets. Adv. Space Res. 9(3), 169-176 

Wyckoff, S., Tegler, S., Wehinger, P. A., Spinrad, H., and Belton, M.J.S., Abundances in comet Halley 

at the time of the spacecraft encounters. Astrophys. J. 325, 927-938 (1988). 
Yamamoto, T., Formation environment of cometary nuclei in the primordial solar nebula. Asiron. 

Astrophys. 142, 31-36 (1985). 
Yung, Y.L., Fried!, R.R., Pinto, J. P., Bayes, K.D., and Wen, J.S., Kinetic isotopic fractionation and 

the origin of HDO and CH3D in the solar system. Icarus 74, 121-132 (1988). 



Sang J. Kim 

Michael F. AHeam 

University of Maryland 



Sang J. Kim and Michael F. A'Hearn (University of Maryland) 

Cometary comae exhibit abundant sulfur and sulfur compounds, most of 
which are absent in planetary atmospheres. Sulfur compounds have also 
been detected in the interstellar medium, including SO, SO2, CS, etc., 
but excluding S2 which was identified only in comet IRAS-Araki-Alcock 
(lAA) 1983d. The study of the origin and parent molecules of these 
compounds, therefore, may yield a clue to the question of the formation 
and evolution of comets from the interstellar medium. Our work, is aimed 
at determining abundances of the various sulfur compounds in comets. 

We have found new evidence of S- in the ground-based spectra of comet 
lAA (Fig. 1) observed by Stephen Larson (Univ. of Arizona); there was at 
least one S^ outburst before the one detected by A'Hearn, Feldman and 
Schleicher (!)• The observations indicate that S2 is often present in 
comet lAA. We undertook fluorescence calculations to analyze the B-X 
system of Sj which appeared both in lUE spectra (Fig. 2) and in the 
ground-based spectra of comet lAA. Single- (Fig. 3) and multiple-cycle 
(Fig. 4) fluorescence calculations indicate that fluorescent equilibrium 
accounts for the observed spectra despite the fact that the S^ lifetime 
against solar ultraviolet radiation is relatively short. This analysis 
confirms unambiguously that emission peaks in the 3000 - 4000 A spectral 
range of the ground-based data are due to the B-X bands of S2 (Fig. 1) 
implying that S2 should be sought in archival ground-based spectra of 
comets. The fluorescence calculation indicates that the previous S2 
production rates should be reduced at least a factor of two. 

Most models of comets suggest that SO and/or SOj should be abundant in 
the coma both because of reactions between the observed species S and OH 
and because of irradiation of other sulfur compounds in icy grains prior 
to accretion. If the Sj in lAA was produced by irradiation of other 
sulfur compounds in icy-mantles of grains as proposed by A'Hearn and 
Feldman (2), then SO under most circumstances should be much more 
abundant than S^- A tentative identification of SO has been proposed by 
Wallis and Krishna-Swamy (3). We have calculated S3mthetic spectra of 
SO and compared them with spectra of various comets observed with the 
lUE (Fig. 5). We find no evidence for the presence of SO and set upper 
limits on the relative production rate of SO in comets (Table 1) 
assuming that the SO is a parent. 

Since the abundance of S^ in lAA was of order 10~ that of water, it 
appears that only the upper limit of SO in lAA is approaching 
interesting values. Although observations with EST will be more 
sensitive than with lUE, it appears that in situ measurements may be 
required to detect SO and test the irradiation hypothesis for formation 
of S2. 

It has been difficult to explain the high resolution lUE spectra of the 
0-0 band of CS at 2577 A, because CS radicals are formed near the 
nucleus where collisions may affect the rotational structure of this 
band. Since calculation of abundances requires an accurate knowledge of 
the emission process, we were motivated by the imcomplete analysis by 






3000 3200 3400 3600 3800 




Figure 1. Comparison between a ground-based spectrum (solid line) 

obtained with the Catalina telescope and the B-X model of Sj (dashed 
line). The identified bands of the B-X system are marked by *. 


Figure 2. The B-X system of $2 appeared in lUE spectra of comet IRAS- 
Araki-Alcock 1983d. 


2800 3000 3200 


Figure 3. Single-cycle fluorescence model for the B-X system of S-. 


2800 3000 



2800 3000 3200 


Figure 4. Fourth- and fifth-cycle fluorescence models for the B-X 
system of S2. The fifth-cycle model is very similar to the 
fluorescent equilibrium model. 





















*— « 



















r— « 














































• l-H 




_ I- 
t f 

I-A-A UWR1590S 
So 6-1 noOEL 


m M 

k i 




3M0 2300 

: lue 

: So 8-». Mooeu 


CERMIS v.vr;S75: 




So J-;k 



Figure 5. Comparison between SO models and lUE spectra of various 


C ome t H a I I e y 





— MODEl- 



Figure 6. Comparison between the 0-0 band model of CS and an lUE 
spectrum of comet Halley. 

Co-.« Vltts., 

0.0 t— — -I 1- I I-- 

• • I... J ..... 1 1 ... ..I. J J. I 

351^.0 rS73 S 7376,0 717«,J 7S77 7577 1 7i78 7478 i »79 



--1 1 1- 4 I 

... L. ... L. , . . I , . L.. I >..!..,. 

»7S.O 357*. S 7S7«.0 7574 5 7S77 O 7S77 1 3578 3578 i 337» ( 

Co—* vJ;ii.^ 

— -ftt '* «>o7tJ 

r _ I I _ ..»_ 

. I ... .1. ... I ... . 1... . . I . . . . I .. .". I 

7i75 7575 5 757ft 757S 5 7577 7577 5 75 78 7578 5 7579 

Figure 7. Comparison between an lUE spectrum of comet Wilson and the CS 
band models with various electron densities. The best fit indicates 
that the electron density is about 2 x 10 cm in the region where 
CS forms. 


Prisant and Jackson (4) to construct a band model which Includes 
fluorescence processes initiated by solar ultraviolet radiation, and 
collisional excitation by electrons and neutrals. In Fig. 6 and 7, we 
present models of the 0-0 band, which give satisfactory fits to the high 
resolution lUE spectra of comets Halley and Wilson, respectively. We 
found that the rotational excitation by electrons is a dominant process 
in determining the ground §tate rotational population. We derived an 
electron density of 2 x 10 era in the region several thousand 
kilometers from comet Wilson's nucleus. In our future works, CS 
abundance will be calculated using the new g-factor and be compared with 
data from the interstellar medium. 

Ultimately the abundances of the sulfur bearing molecules must be 
compared to assess the complete sulfure chemistry in comets. The atomic 
sulfur, also observed from lUE, is curerntly being studied by Feldman 
and collaborators since there are indications (5) that the g-f actors for 
some lines of this species may be incorrect. 


1. A'Hearn, M.F., Feldman, P.D., and Schleicher D.G. (1983) Astrophys. 
J. v. 274, pp. L99-L103. 

2. A'Hearn, M.F. and Feldman, P.D. (1985) in Ices in the Solar System, 
ed. J. Klinger et al., 463-471. 

3. Wallis, M.K. and Krishna-Swamy , K.S. (1987) Astron. & Astrophys. v. 
197, pp. 329-332. 

4. Prisant, M.G. and Jackson, M.W. (1987) Astron. & Astrophys. v. 187, 
pp. 487-496. 

5. Roettger, E.E., Feldman, P.D., A'Hearn, M.F., Festou, M.C. , 
McFadden, L.A., and Gilmozzi, R., (1989) Icarus, in press. 



Ichishiro Konno 
Southwest Research Institute 

Susan Wyckoff 

Peter A. Wehinger 

Arizona State University 



Ichishiro Konno^ Susan WyckofF", and Peter A. Wehinger- 
^Southwest Research Institute, ^Arizona State University 


Spectrophotometric observations of Comet P/Giacobini-Zinner were obtained in March, June, September, and 
October 1985. The September observations were obtained at perihelion, exactly at the time of the International 
Cometary Explorer (ICE) encounter with the comet. Spatial profiles extracted from the long-slit spectra were analyzed 
by using a Monte Carlo method to determine scale lengths and lifetimes for the observed radicals, C2 and NH2, and 
their respective parent molecules. The scale length for the parent of C2 was found to be (7.5 ± 1.5) x lO"* km and for 
the parent of NH2 (2.4 ± 0.4) x lO"* km. The brightness profile of C2 and the lifetime of the parent of C2 indicate that 
C2 probably comes from many different sources which may include C2H4, C2H2, and dust particles, C2 and NH2 were 
found to be depleted in Giacobini-Zinner relative to an average comet by factors of 10 and 5, respectively. The water 
production rate was obtained for June, September, and October observations from the measurements of the [0 I] 6300 
A line. The water production rate at the time of the ICE encounter was found to be 2.4x10"^ molecules s"-', in good 
agreement with spacecraft results. 


The primary goals of the observations of Comet Giacobini-Zinner were to 1) identify the emission lines in the 
spectra, 2) obtain spatial distributions of various species to identify the possible parent molecules of observed species, 
and 3) obtain the production rates and abundances of observed species and water. A summary of the telescopes and 
instruments used for observations of Comet Giacobini-Zinner are shown in Table 1. All the observations were made 
at the National Optical Astronomical Observatory (NOAO) at Kitt Peak. The comet was observed from March to 
October 1985, covering the heliocentric distances of 2.29 AU to 1.03 AU including the precise time of the ICE encounter, 
11 September 1985 at 11:00 UT, giving direct comparisons of ground-based and spacecraft observations. 


Observations on 20 March 

In the spectrum obtained on 19 March 1985 only CN(At; = 0) emission was marginally present with a strength 
■~ 1(7 above the noise level (which was determined by measurements of the continuum). No NH2 bands were found 
at the 1 a significance level. The results of the measurement of the emission band flux and the column density are 
presented in Table 2. The p-factors which were used to calculate column densities for various species are shown in 
Table 3. 

Observations on 20 and 21 June 

Figures 1 and 2 show the spectra of Giacobini-Zinner at and off the nucleus on the night of 20 June 1985. The 
aperture was set at the nucleus and at 10" from the nucleus along the plasma tail on 20 June and was set at the nucleus 
and 8" from the nucleus along the plasma tail on 21 June. Since the nuclear spectra on the nights of 20 and 21 June 
are almost identical in terms of emission features and intensities, only the spectra on 20 June are shown. The emission 
band fluxes and the column densities are shown in Tables 4 and 5. 

















Kitt Peak 

20 Mar 








20 Jun 








21 Jun 








11 Sep 








19 Oct 








r: heliocentric distance; A: geocentric distance; PA: position angle; IIDS; intensified image dissector scanner; CC 
charge-coupled device 


r=2.29AU A = 2.28 AU r = -14.98 km s'^ 




Integrated flux 
(erg cm~- s~^) 

Column density 




(2.8±1.2) xlO* 

f: heliocentric velocitv 



^-factor at 1 AU 
(erg s-^) 
















[0 I] (6300) 


















Feldman and Brune (1976) 
Schleicher (1987b) 
Schleicher (1983) 
Schleicher (1983) 
Cooper and Jones (1979) 
Schleicher (1983) 
A'Hearn et al.. (1980) 
Magnani and A'Hearn (1986) 
Magnani and A'Hearn (1986) 
A'Hearn et al.. (1985) 
A'Hearn et al.. (1985) 
A'Hearn (1982) 
A'Hearn (1982) 
Lutz (1987) 

Festou and Feldman (1981) 
A'Hearn (1982) 

value for r(heliocentric velocity)=0. **: value for r = —15 km s 



O '2 










J L 


O -r> 









o<J ^ 












„_Olxt tT-0^x2 

(,_y ,_s 3_ino 9j8) ^^ 


— -^ 

2 - - 


(-T -' -..-' 

\ •* _ 













r - 



'1-5=:- ^i:--^ 

(14=:: :: 

.'5.0=4.5; X 10— * 


'7.3=: :- = ::-^ 


• 5-7=1 : • 11"-^ 



(94-HJ-5 xli 

(3 1-!2-.x-^--^ 

(e 9-- f xi: 

I 5 ^~'~ -^ X _ . 

(s.i=i : x: 

' 1 1=: i X ii"""^ 

(7 1-3 7. X 

I ~—\ i ■■ V.~ ~ 

? :=: " : 1 : 

r=1.4" .A.r A=':.9^ AU r = -15-27 kn s"- 



- 5--) 












(1.2=0^:. xi: 

(4.9^:0.4:. xlQ-" 

.'2.2=: I = ::~-^ 

(2-6=1-0) x 10-" 

7 :=c.5"ixio-^3 

c 9=1-4; X 10--= 



{2^=; - ':':: 

fl !=. ; 1 . " 

(i-5=; i = i;^ 


(2:=:; - 

(l-SdrG-l'- xl«v= 

Bands of Or. and CN are srrisiigiy do-nira::: in liisse ^lectra. Tlrree baEC seqasics cf CN. Ar = I. a=.c 
Ar = —1. are sesn. The C; S^an c-anis are she nexi sct:nges; bands in ^e speiEs. br: as ias resi renins:! :-i rj 
several observiers. ttese bands ■"€;« nc: as strong in Giaccrmi-Zmner as m mos* C5n€r cjsxsss- A'Eearn an; -'■ ' 's 
(15S0;. Coch-an and Barker (l&S"! and Scnieicier e: al- ■ 19S7. hv -::--: ir "i- :: --7"--^- i i-Tt-: i Tt 
and C3 relative to GN- They ha's's independenkly found isjaZ ;ie pr;i_;:-:~ ri.i; :: -: i^i -; i^^ -:~^ -. ^ "i:.:r 
of abou: 5 relative to CX compared wii averages of other cc~eis- Our sre-rira snc-s- ma: XH. CH. aac Nj±: 
tntensiiies comparable kO C^- 

In kbe spectra obtained wiih kbe sli: o£ iie -_:-e"_i- '5.zZ~ features are mrre zr;ni.nen: tn^n rn tne n.i-t-i 
especiallv on 20 Jtme. Tiie relaiive intensiiies of the H^Q— banns rerrnie strc-nger tri%ii tlK distance xrcen tne nnt-ecs- 
Tiiis behavior is due tc the accsleratio-ns of ions cacsed by the ;- - ^ - - - ; - - - : -s-itn tn^ - - 

the plasaia tail .A. -Krealv conrrib-ticn by OR— to tie feature ?. . - : , t~ t-; m :n-r . 



Observations on 11 September A 

In Figure 3 a spectral cut covering a spatial extent 2.0 x 10^ km at a projected distance 8,000 km tailward from 
the nucleus is shown. This spectrum obtained at 11:00 UT on 11 September represents the column containing the 
ICE spacecraft at mid-encounter. Various ion and neutral species have been identified and are labeld in the figure. 
The solar continuum reflected from the cometary dust has been subtracted from the spectrum. In Tables 6 and 7 the 
integrated fluxes of the observed features at the nucleus and in the tail (Figure 4) are given together with the column 
densities. We searched for CO"*" features, expecially the (1-0) and (1-1) bands at 4555 A and 5055 A but no CO''' 
features were seen at the Zcr significance level. Thus, we can set only an upper limit on the CO'^'/HoO"'" ratio from the 
flux measurements. 


r=1.03 AU A=0.47AU f =1.96 km s-^ 

Species Wavelength Integrated flux Column density 

(A) (erg cm~* s"-') (cni~") 

C2(Av=0) 5101 (6.6±l.S)xl0-" (1.1±0.3) xlO^ 

NH2(10-0) 5732 (3.7±2.4)xl0-i^ (9.0±5.9)xl0^ 

NH2(9-0) 5996 (8.9±3.3)xl0-^' (2.1±0.8)xl0i° 

[0 1] 63O0 (3.5±0.2)xl0-" (1.5±0.07)xl0i° 

Integrated flux 
(erg cm~- s"-') 

Column densitv 





(4,0±0.7)xl0-i = 






Species Wavelength 


CO+(1-0) 4550 

C0+(1-1) 5050 

C2(Av=0) 5078 

NH2(10-0) 5720 

NH2(9-0) 5994 

H26+(8-0) 6189 

[O" I] 6300 

Observations on 19 October 

Figure 4 shows the spectrum of the nucleus with an extraction sum of 10 pixels which gives 8" (5,000 km) onto 
the comet in the tail-axis direction. The main differences with the .lune and September spectra are the vi-eaker Cj 
band features. The results of measurements of fluxes and column densities are shown in Table 8. 


r=1.20AU A=0.65AU f =11.94 km s'^ 

Species Wavelength Integrated flux Column density 

(A) (erg cm~- s~^) (cm~-) 

NH2(9-0) 5995 (6.4± 4.9)xl0-^5 (1.2± 0.9)xl0i° 

[0 1] 6300 (2.7±0.4)xl0-" (9.5± 1.2)xlOi° 

[0 1] 6364 (8.9± 1.3)xl0-''' (9.3± 1..3)xl0" 









.= o -^ 

S o r- 

i ?^ 

O « i« 

-2 t--" 

" ° g 

_o = c; - 

-i tyj o — 

X « o 

o — ' ::!. c: 

_ ~ M = 

•S ^ i^ o 

— ^ o s 

CO o .:: - 

c^ r_^ .= 3> 

2 O •= t-" 

5) -J = i 

■-r — c r 





(x_Y T-^ Z-"^"" ^•'^) ""^ 










- a 





>; — 

? s 


n " 


w ^ 


^ -^ 



"^ s 


5 i^ 


Jl y 


_ w 


- r: 


- «; 




■0 = 


2 2P 


— _o 

^ J 
_• ° 

9 ;: 

— ~ 









(,.Y ,-S 3.^3 2-19) "^i 



The observed brightness profiles for NH2(9-0) and C2(Ai' = 0) extracted from the CCD spectrum on 11 September 
are shown in Figures 5 and 6 with profiles calculated using a Monte Carlo method.^The parameters for the fits are shown 
in Table 9. The production rate for NHo was found to be (5.1 ±2.5) x 10^" molecules s~^ and for Co (5.3 ± 1.0) x IQ-" 
molecules s~^. 



lifetime at 
1.03 AU (s) 

lifetime at 
1.00 AU (s) 

scale length at 
1.00 AU (km) 

NHo parent 




(2.4 ± 0.7) X 10" 
(2.4 ± 0.8) X 10" 

(2.4 ± 0.7) X 10" 
(3.6 ± 1.2) X 10" 

C2 parent 



(7.5 ± 1.5) X 10" 
(9.4 it 1.9) X 10" 

(7.5 ± 1.5) X 10" 
(1.1 ± 0.2) X 10^ 

The lifetime of the NH2 parent was found to be (2.4 ± 0.7) x 10" s. The calculated photodissociation lifetime of 
NH3 is 5.9x10^ s at 1 AU from the sun (Huebner and Carpenter, 1979). Thus the calculated NH3 lifetime is a factor of 
4 smaller than the observed lifetime of the NH2 parent. This difference may indicate that NH3 can be excluded from 
the candidates for the parent of NH2. Delsemme and Combi (1983) have obtained the lower limit for the lifetime of 
the NH2 parent to be 2.7x10" s from observations of comet Kohoutek by reinterpreting the Haser scalelengths using 
the average random walk model (Combi and Delsemme 1980a) assuming the parent speed v = 0.58 r~^^- km s~^. If, 
instead, v = 1.0 r~'^/- is used, their lower limit reduces to 1.6 xlO" s, which is still a factor of about 3 larger than the 
NH3 photodissociation lifetime. It has been suggested by Huebner et al. (1989) that NH2 may come from polymers 
embedded in grains. 

We obtained a lifetime for the C2 parent at 1 AU of (7.5 ±1.5) x 10" s. Huebner and Carpenter (1979) have 
calculated the photochemical lifetimes of C0H4 and C2H2. Possible reactions which may ultimately produce Co are: 

(l)CoH4 + hiy — C2H2 + H2(orH + H); - = 2.1 x 10" s (98%) 

(2)C2H2 + hr/ — CoH -f H; r = 1 .0 x 10^ s (74%) 

(3)C2H2 + hz/ — C2 + Ho; r = 3.7 x 10^ s (20%) 

(4)C2H + hi/ — C2 + H; 7- = 3.3 X 10" s (estimate, Schmidt et al., 1987) 
Here the values in the parentheses are the branching ratios. If C2 is produced by the reactions that take the path (1) 
followed by (3), the process takes at least 3.9 x 10^ s. This time scale is 5 times larger than the lifetime observed for the 
Co parent(s). Therefore, we can probably exclude this process as a dominant production mechanism for Co. In the case 
that C2 is produced by the reactions (1), then (2), then (4), 1.5x10^ s is required, which is two times longer than the 
lifetime we observed for the parent of C2. The shortest process to make Co is process (4), but CoH is very unlikely to 
be present in the comet nucleus, therefore, the most Hkely path is the process (2) followed by (4), which takes 1.3x10^ 
s. We see then that both C2H4 and C2H0 take times about twice as long as the observed Co parent lifetime. We must 
remember, however, that the model fit to the observation is not very good near the nucleus which suggests that either 
C2 may be the third generation or it has multiple parents, or both. Yamamoto (1981) and O'Dell et al. (1988) have 
shown that a Haser model for C2 as the third generation fits very well with observations and therefore Co is likely the 
third generation of the original comet nucleus composition. C2 may also come from dust grains. It is very likely that 
Co comes from several parent or grandparent molecules mostly from the nucleus and partly from dust. 


Co and C3 emissions have been shown to be weak with respect to CN in Giacobini-Zinner by Herbig (1976), and 
tlie.S"' result? liave hecn confirmed during the recent apparition by Schleichrr (1985), Cochran and Barker (1987). and 








i— I 











uo Ei; 


^ ^ 



ii "r 


"c "" 

z ^ 


~ ff 


K — 

C - — 





• z? = 


— « 


1' ^ 

ft K 


t-' -li 


~ — 

n — 

<j *"" 


1 = 




> — 

D ?j 

^ s 

5 :; 


o " 

V _u 

3: "c 


(AHYHlia^') A1ISN31NI 3AI1V13H DOT 









































I 1 



^ u 









^— I 










"5 i2 

^ = 

a >> 


7 ID 

I 1 — I 




o = 


Schleicher et al. (1987). The production rates obtained in this work show that in Giacobini-Zinner, C2 and NHt are 
depleted by a factor of 10 and 5, respectively, relative to H2O compared with the values for an average comet obtained 
by Schleicher ei ai. (1987). Water production rates were obtained (see Figure 7) from the measurements of the [O I] 
6300 A line using the method developed by Spinrad (1985,1987). The water production rate for 11 September 1985 
was found to be (2.4 ± 0.2) x 10"* molecules s~\ in agreement with the spacecraft results of 2-5x10"'^ molecules s"' 
by lUE and Pioneer Venus Orbiter (Stewart ei al. 1985). The depletion of C2 and C3 in Giacobini-Zinner may be 
related to the low dust production rate. Wagner et al. (1987) have estimated the dust-to-gas mass ratio to be ~ 0.07, 
which is a factor of 3 lower than the average of the 17 comets observed by Newburn and Spinrad (1985). If dust 
grains are the source of observed C2 and C3 molecules, there may be a correlation between the dust-to-gas mass ratio 
and the relative abundances of carbon species. The data obtained by Newburn and Spinrad (1985) do not show an 
obvious correlation between the dust-to-gas mass ratio and the C2/OH abundance ratio among the observed comets. 
In several comets, however, the abundance of C2 does vary with the dust-to-gas mass ratio. Therefore dust may be 
responsible for at least part of the C2 and C3 production in some comets. But, since the method used to determine the 
dust-to-gas niass ratio in comets introduces an error of about a factor of three, it is difficult to derive the correlation 
between the dust-to-gas mass ratio and the production rates of any species. Comet Halley, however, has much higher 
C2 abundance and its dust-to-gas mass ratio is in the range of 0.1 to 0.25 (Sagdeev et al. (1986), which is 3 to 8 times 
higher than Giacobini-Zinner. If dust grains are responsible for most of the production of C2 and C3 molecules, then 
the low production rates of these molecules in Giacobini-Zinner may be explained by the low dust production rate in 
this comet. 

It was found that NH2 is also depleted in Giacobini-Zinner. WyckofTet al. (1988) have shown that NH2 abundance 
is very low in comets they observed, suggesting the low NH3 abundance in comets in general. As it was shown in the 
previous section, however, the lifetime for the NH2 parent was found to be too long for NH3 to be tlie parent. Therefore 
NH2 seen in spectra of comets probably does not come from NH3, which may be more abundant than found from the 
NHo abundance, but polymers embedded in dust grains, may be responsible for NH2 in comet spectra. The low 
abundance of NHo in Giacobini-Zinner might also be related to the low dust-to-gas ratio in this comet. 




CO u 























2 8 












»— H 






^ J 






(^_S S3inD310I\) (0^H)6 001 



A'Hearn, M.F.: Spectrophotometry of Comets at Optical Wavelengths, in Comets, ed. Laurel L. Wilkening, The Uni- 
versity of Arizona Press, 1982, pp. 433-460. 

A'Hearn. M.F.; and Millis. R.L.: Abundance Correlations among Comets. A. J., vol. 85, 1980, pp. 1528-1537. 

Cochran, A.I..: and Barker E.: Comet Giacobini-Zinner: A Normal Comet? A. ./., vol 92, 1987, pp. 239-243. 

Combi, M.R.; and Delsemnie, A.H.: Neutral Cometary Atmospheres I. An average Random Walk Model for Photodis- 
sociation in Comets. Ai>. ./., vol. 237, 1980, pp. 633-640. 

Cooper, D.M.; and Jones, J..].: An Experimental Detection of the Section the Swings Band System of C3. •/. 
Quant. Sped. Had. Trans., vol. 22, 1979, pp. 201-208. 

Delsemnie, A.H.; and Combi, M.R..: Neutral Cometary Atmospheres IV. Briglitness Profiles in the Inner Coma of 
Comet Kohoutek 1973 XIII. Ap. .J., vol. 271, Aug. 1, 1983, pp. 388-397. 

Feklman, P.D.; and Brune, W.H.: Carbon Production in Comel West (1975n). Ap. .}., vol. 209, 1976, iip. L4.5-L48. 

Festou, M.C.; and Feldman. P.D.: The Forbidden Oxygen Lines in Comets. .Asiroi). Asirnphys, vol. 103, 1981, pp. 

Horbig. G.: Review of Cometary S|)ectra. In The Slucly of Comets, eds. B. Donn, M. Miimma, VV. .lackson, M. A'Hearn, 

and R. Harrington (Washington; NASA SP-393), 1976, pp. 136-158. 
Huebner, W.F.; and Carpenter, C.W.: Solar Photo Rate Coefficients. Los Alamos Sctcviific Lab Report, LyV-8085-MS, 

Huebner, W'.F.; Boice, D.C; and Korth, A: Halley's Polymeric Organic Molecules. Advances in Space Research, in print. 
Luts, Barry L.: Fluorescence Efficiency Factors for Ionized Water Vapor. Ap. J., vol. 315, no. 2, April 15, 1987, pp. 

Magnani, L.; and A'Hearn, M.F.: CO"*" Fluorescence in Comets. Ap. ./., vol. 302, 1986, pp. 177-487, 

Newburn, R.L.Jr.; and Spinrad. II.: Spectrophotometry of Seventeen Comets. II. The Continuum. A. ■]., vol. 90, no. 

12, Dec. 1985, pp. 2591-2608. 
O'Dell, C.R.; Robinson, Ronald R.; Swamy, K.S. Krishna; McCarthy, Patrick J.; Spinrad, II.; C2 in Comet Ilalley; 

Evidence for Its Being Third Generation and Resolution of the Variational Population Discrepancy. Ap. ./., vol. 

334, Nov. 1, I9S8, pp. 476-488. 
Sagdeev, R.Z.; Blainont, .1.; (laleev, k.i\.\ Moroz, V.I.: Shapiro, V.D.; Shevchenko, V.I.: and Szcgo, K.: Vega Spacecraft 

Encounters with Comet Ilalley. Nature, vol. 321, no. 6067, 15-21 May. 19S6a, pp. 259-202. 

Schleicher, D.G.; Ph.D. IDissertation, University of Maryland, 1983. 

Schleicher, D.(.'.: Bull. Am. Astron. Soc, vol. 17, 1985, p. 686. 

Schleicher, D.G.; and A'Hearn, M.F.: Fluorescence of OH in Comets. Ap. /.. vol. 258, 1981, p. 864. 

Schleicher, DC; Millis, R.L.; Birch, P.V.: Photometric Observations of Comet P/Gincobini-Ziiiner, A.iirun. .Aslrophys., 

vol. 187, 1987. pp. 531-536. 
Schmidt, II. U.: Wegmanii, II.: Hin-hner, \V.F.;and Boicc, D.; Cometary Ga.s ami Plasma Flow uitli I)(-t;ul<-d ( 'hoiiii.stry. 

Computer Physics Comvnnncatiovs, vol. 49. 1988, pp. 17-59. 
Spinrad. II.: Observation.s of tlie Rod Aiuoral O.xygen Lines in Nin<' (.■oniots. Puhi .islrov. Soc. Pacific, vol. 94. Dec. 

1982, |>p. 1008-1016. 
Spinrad, II.: C-omets anil Their Composition. Ann. Rev. Asir. Ap.. vol. 24. 1987. 
Stewart, A. IF.; Combi, MR.; and Smyth, W.H.: Bxdl. Am. Astron. Soc, vol. 17, 1985, p. 6Sfi. 
Wagner, R.M.; Lutz, Barry L.; Wyckofl', S.: Groundbased Constraints on th.' lIjO+ZCO''" Abundance Ratio and Dust 

Impact Rate in Comet P/Giacobini-Zinner: T'onifiarison with the Spacecraft Results Ap. .J., vol. 322. Mac. 1, 

1987, i>|>. .')44-548. 
Wyckoir, S,; 'IVgh-r, S.; ami Lngel. L.: Ammonia Ainnulances in ( 'oiim-is. I'i iicc<dnii)s of 29th COSl'A li. 19tS9. m print. 
Vanmiiiotf). Tclsiio: On I Ur IMiotO( heniu :i| Format 1011 of CN. ("•_.. and ( ':j Radicals in ( ■\ C'oina<', 7'/u Uwni anil 

till Plan, I-., vol. 24. MiM. pp -15:;- l(i3. 



Paul R. Weissman 

Earth and Space Sciences Division 

Jet Propulsion Laboratory 

S. Alan Stem 

Laboratory for Atmospheric and Space Physics, and 

Department of Astrophysical, Planetary, and Atmospheric Sciences 

University of Colorado 



Cometary nuclei preserve a cosmo-chemical record of conditions and processes in the primordial 
solar nebula, and possibly even the interstellar medium. However, that record is not perfectly 
preserved over the age of the solar system due to a variety of physical processes which act to 
modify cometary surfaces and interiors. Possible structural and/or internal processes include: 
collisional accretion, disruption, and reassembly during formation; internal heating by long and 
short-lived radionuclides; amorphous to crystalline phase transitions, and thermal stresses. 
Identified surface modification processes include: irradiation by galactic cosmic rays, solar 
protons, UV photons, and the Sun's T Tauri stage mass outflow; heating by passing stars and 
nearby supemovae; gardening by debris impacts; the accretion of interstellar dust and gas and 
accompanying erosion by hypervelocity dust impacts and sputtering; and solar heating with 
accompanying crust formation. These modification processes must be taken into account in both 
the planning and the interpretation of the results of a Comet Nucleus Sample Return Mission. 
Sampling of nuclei should be done at as great a depth below the surface crust as technically 
feasible, and at vents or fissures leading to exposed volatiles at depth. Samples of the expected 
cometary crust and near-surface layers also need to be returned for analysis to achieve a better 
understanding of the effects of these physical processes. We stress that comets are still likely 
less modified than any other solar system bodies, but the degree of modification can vary greatly 
from one comet to the next. 



It has become almost a matter of faith among solar system astronomers that "comets are 
the best obtainable source of original solar nebula material. " There will be a temptation to 
quickly apply many of the new findings from a comet sample return mission to a description of 
the primordial solar nebula at the time the comets were forming. But, the comets, like every 
other solar system body we have studied with flyby and orbiter spacecraft, are evolved bodies. 
Over the history of the solar system comets have been subjected to a variety of physical 
processes which have modified them, admittedly much less than the larger planets and satellites, 
but stiU in very significant ways. 

To fully understand and interpret the results of a comet sample return mission it is 
necessary to consider the target comet's complete physical and dynamical history, and the 
processes which have likely acted to modify it from its original "pristine" state. This is 
necessarily a statistical exercise, considering those "likely" processes which we can conjecture 
from our current understanding of the solar system, the Oort cloud, and the galaxy. One cannot 
foresee unique, low probability events in the history of any individual comet that might have 
modified it in some significant fashion, different from that generally experienced by the majority 
of other comets. Nor can we rule out that the comet itself is a unique body, formed (or 
captured) in some unique way that again does not represent the majority of the cometary 

The situation has been additionally complicated recently by the suggestion that the source 
of the short-period (SP) comets is dynamically distinct from that of the long-period (LP) comets. 
Duncan et al. (1988) showed that SP comets dynamically evolved from the randomly oriented 
orbits of LP comets would preserve their random inclination distribution, contrary to the 


relatively modest inclination, direct orbits which are observed. They have also shown that a 
more likely source for the SP comets is a flattened ring of ~ 10* to 10' comets beyond the 
orbit of Neptune, a remnant of the solar system's nebula accretion disc.~ They have named this 
the Kuiper belt, for Kuiper's (1951) suggestion that such an extended accretion disk might exist. 
Because the Kuiper belt is much closer to the Sun, the processing history of comets in the belt, 
which is partially if not totally within the Sun's heliosphere may be somewhat different than 
comets in the more distant Oort cloud. 

A comet's evolutionary history can be broken into four distinct periods: 1) formation, 
presumably coincident with the formation of the Sun and planetary system; 2) initial processing, 
prior to ejection to the Oort cloud; 3) dynamical storage in the Kuiper belt or the Oort cloud at 
moderate to large solar distances for most of the history of the solar system; and 4) processing 
upon re-entry into the planetary system and the observable region, r < 5 AU. Each of these 
periods will be discussed below. In addition, the uncertainty associated with the various 
dynamical paths by which comets may evolve from the Kuiper belt or the Oort cloud to SP 
orbits, and the possible dynamical history of SP comets will be briefly examined. 

Exactiy what constitutes a "pristine" cometary nucleus has different meanings to different 
investigators. Even the interstellar medium prior to the comet's formation might not be 
considered truly "pristine" because of the complex chemistry that occurs when volatile ices 
condense on interstellar grains and are irradiated by UV photons and galactic cosmic rays 
(Greenberg and D'Hendecourt, 1985). Material in interstellar clouds has been processed in and 
by stars, shocked in supernova explosions, irradiated, etc. , so it is meaningless to speak of truly 
pristine material. Given that the interstellar medium we observe today has already undergone 
similar processing, we will assume that condition as our starting point. It will be seen that the 
degree of modification as the comet forms and evolves can usually only be defined in a relative 


sense. The object of this paper is to survey the possible physical processes, to rank them in 
relative importance, and to make recommendations regarding sampling strategies, leaving the 
detailed quantitative evaluation to future work. 


Hypotheses of cometary origin fall into two major categories (Weissman, 1985a): 
primordial hypotheses in which comets formed at the same time, and as part of the formation 
of the Sun and planets, and hypotheses in which comets are formed or captured episodically at 
essentially random times, often as a result of catastrophic processes, and possibly repeatedly 
over the history of the solar system. In general, the episodic hypotheses have not gained very 
wide acceptance, and will not be discussed further here (c.f. Bailey et al., 1986). 

All primordial hypotheses agree on one basic fact: comets formed far from the Sun in 
the cooler regions of the solar nebula. The high volatile content of comets, largely in the form 
of ices, and the relatively small sizes of cometary nuclei, provide proof of that fact. 
Disagreement exists in just how far away the formation zone actually was. The suggestions 
range from the Uranus-Neptune zone at 20 to 30 AU from the Sun, to neighboring fragments 
of the proto-solar nebula, at distances of 10* AU or more. 

Formation of comets among the outer planets was first suggested by Kuiper (1951) who 
noted that water ice would not condense any closer to the Sun than about Jupiter's orbit. But 
dynamical studies (Safronov, 1969; Fernandez and Ip, 1981) have shown that Jupiter and Satiun 
tend to eject the majority of the icy planetesimals in their zones to interstellar space, whereas 
Uranus and Neptune with their smaller masses and larger semimajor axes are more likely to 
place a sizeable fraction of planetesimals into distant bound orbits with dimensions — 10^ to 


10^ AU or more, the region of the Oort cloud. Temperatures in the Uranus-Neptune zone are 
expected to have been < 100 K during cometary formation, allowing volatile ices such as HjO, 
CO, CO2, NH3, and CH4 to condense and/or to be trapped as clathrate hydrates in the icy matrix 
(Delsemme, 1982). 

According to the standard planetary formation scenario (Greenberg et al. , 1984) ice and 
silicate grains in the infalling solar nebula descend to the equatorial plane of the nebula due to 
gas drag, forming an accretion disc. Agglomeration of grains leads to the growth of larger 
particles, both during the fall towards the nebula plane, and while circulating in the accretion 
disc (Weidenschilling, 1980), leading to growth of initial icy-conglomerate objects as large as 
10 meters. When the density of material in the plane reaches a sufficient value, gravitational 
instabilities (Goldreich and Ward, 1973) lead to fragmentation of the disc and collapse into 
planetesimals several kilometers in dimension, within about 10'' to 10^ years after the start of 
nebula collapse. These objects were presumably the proto-comets. The total initial mass of 
planetesimals in the Uranus-Neptune zone is estimated to be on the order of 10 times the 
combined present masses of those two planets (Safronov, 1969; Lissauer, 1987). The proto- 
comets were ejected to the Oort cloud or to interstellar space by the growing proto-planetary 

The degree to which interstellar material is modified as it is brought together by the 
processes described above is highly uncertain. Infalling nebula material in the Uranus-Neptune 
zone may be moving inward with a velocity of a few km s', and is likely slowed and shocked 
as it reaches the denser parts of the nebula around the accretion disc. This could raise the 
temperature of the material significantly, vaporizing icy particles and driving volatiles off of 
refractory grains. The degree of heating is also affected by the opacity of the accretion disc: 
how well it traps the heat generated at the shock boundary. 


The grains cool rapidly as they begin to agglomerate and settle towards the nebula plane. 
The interparticle velocities at this point are very low, and material is brought together with 
relatively little compaction (Goldreich and Ward, 1973). This is expected to lead to a highly 
porous, low density, loosely bound structure composed of a heterogeneous mixture of ice and 
dust grains covering a wide range of particle sizes (Donn and Rahe, 1982; Greenberg, 1986). 

The breakup of the accretion disc and collapse into planetesimals due to gravitational 
instabilities is also expected to be a relatively gentle process, with interparticle velocities on the 
order of only a few m s"' at most. The total gravitational potential energy of a 10 km diameter 
icy conglomerate planetesimal is not enough to raise the average temperature of the material by 
even one Kelvin, when it is brought together with a mean density of 1.0 g cm"^. Some local 
compaction and heating might be expected at interfaces where larger particles come together, 
probably resulting in a "welding" of the particles into a loosely bound agglomerate of icy chunks 
covering a fairly wide size range. 

The precise structure of the cometary nucleus created in this fashion is still a matter of 
considerable debate. Figure 1 shows four suggested models for cometary nuclei (Whipple, 1950; 
Donn et al., 1985; Weissman, 1986; Gombosi and Houpis, 1986). Common features of aU the 
models are the irregular shape of the nucleus, the heterogeneous mixture of icy and nonvolatile 
materials, and the existence of substantial voids within the nucleus. Differences exist over the 
exact degree and scale of the heterogeneity, and the degree to which the initial agglomeration 
of particles has been modified into a single, well compacted body. The initial nucleus structure 
will be important to its subsequent evolution, in particular to the way energy is deposited in, or 
liberated from the nucleus interior, and to its survivability against collisional destruction. Mixed 
in with the icy conglomerate material of the Uranus-Neptune zone will be some material formed 
closer to the Sun and then dynamically ejected by the growing proto-Jupiter and proto-Satum. 



'■ h^ 


Figure 1. Four suggested models for the structure of cometary nuclei: a) the icy 
conglomerate model (Whipple, 1950); b) the ftactal model (Donn et al., 1985); c) the 
primordial rubble pile (Weissman, 1986); and d) the icy-glue model (Gombosi and 
Houpis, 1986). All but d) were suggested prior to the HaUey spacecraft encounters in 


Some contamination might even be possible with material from the terrestrial planets zone, but 
it is likely that Jupiter will be a strong dynamic filter that ejects most of that material 
hyperbolically, rather than placing it in the Uranus-Neptune zone (material from the terrestrial 
planets zone approaches Jupiter with a high relative velocity as compared with the low velocity 
of material already in the Jupiter-Satum zone, and thus has an even higher probability of 
dynamic ejection). Thus, any initial chemical differentiation in the solar nebula due to the radial 
temperature gradient will be blurred somewhat but will not be erased by dynamic exchange 
between different proto-planetary accretion zones. 

Once the nuclei form they will continue to circulate in their orbits with low relative 
velocities. Collisions should initially be highly efficient in growing larger bodies. But as the 
larger accretion cores grow they will also serve to perturb the orbits of the remaining small icy 
planetesimals, increasing their relative velocities and decreasing the accretion efficiency. Higher 
velocity collisions will result in more destructive collisions, creating large amounts of debris 
which will quickly be swept up again by the planetesimals. Collisions which are not totally 
disruptive will result in crushing and compaction of the initial nucleus structure. The nuclei will 
thus likely go through a period of competing accretionary and destructive forces. 

This process will slow when the runaway growth of one of the planetary embryos 
becomes sufficient to start ejecting comets to inner Oort cloud distances, on the order of 5 x Itf 
AU. At those distances galactic and stellar perturbations will be able to detach the proto-comets' 
perihelia from the Uranus-Neptune zone and make them semi-permanent residents of the inner 
Oort cloud. As material was either ejected or incoiporated into the proto-planetary cores, the 
density of planetesimals remaining in the planetary zone would drop dramatically and collisional 
evolution would decrease accordingly. 

Alternative hypotheses place the formation zone for the cometary nuclei farther from the 


Sun. Among the attractive features of such hypotheses are the facts that the nebula material will 
undergo less heating prior to being incorporated into the icy planetesimals, and that subsequent 
evolution will also likely be milder with lower collision rates and collision velocities between 
planetesimals, and less acceleration of velocities since no large proto-planets grew (as far as we 
know) beyond the Uranus-Neptune zone. The principal drawback of these hypotheses involves 
the difficulty of bringing 10 km sized icy planetesimals together at large solar distances in a 
reasonable period of time. 

One suggestion (Cameron, 1962, 1978) is that the nebula accretion disc did not stop at 
the orbit of Neptune, but rather extended out several hundred AU or more. The lack of giant 
planets beyond Neptune (Pluto is presumed to be an example of the largest icy conglomerate 
body to grow in its zone) was not a result of the nebula disc running out of material, but rather 
of it running out of sufficient time to build a giant planet. Accretionary times are dependent on 
the orbital period as well as the density of material, both of which act against planet buUding 
processes in the outer solar system. 

Observational support for this concept has come from the IRAS discovery of dust shells 
aroimd some young main sequence stars in the solar neighborhood (Aumann et al., 1984). An 
optical photograph of the /3 Pictoris disc edge-on (Smith et al., 1988) shows a flat disc extending 
up to lO' AU from the star, with a maximum thickness of about 50 AU. Estimates of the mass 
of these discs range ft-om a minimum of - 15 Earth masses (M® ), to possibly 200 to 300 M® 
if an asteroidal size distribution can be assumed. 

This extended accretion disc is expected to become the Kuiper belt of comets beyond 
Neptune, as proposed by Duncan et al. (1988). However, some comets in the belt will be 
perturbed to Neptune and Uranus crossing orbits, and along with the comets formed in that zone 
will be ejected on long-period orbits. Duncan et al. (1987) showed that once the semimajor axes 


of orbits reached about 5 x lO' AU, galactic and stellar perturbations would raise the perihelia 
of the orbits and gravitationally detach them from the major planets. Comets initially ejected 
to distances greater than 2 x 10* AU would form the primordial Oort cloud. However, most of 
those comets will be lost over the history of the solar system due to major perturbations from 
stars and giant molecular clouds in the galaxy (Hut and Tremaine, 1985). The outer Oort cloud 
will be replenished by comets from the inner Oort cloud reservoir, those comets initially ejected 
to orbits between 5 x 10^ and 2 x 10* AU. They will be pumped up to the outer cloud by the 
same major perturbations that strip comets from the outer Oort cloud (Fernandez, 1985; 
Weissman, 1985b; Shoemaker and Wolfe, 1986). 

An entirely different formation process involves differential radiation pressure on distant 
nebula fragments, perhaps 5 x 10^ AU from the forming proto-Sun (Whipple and Lecar, 1976; 
Hills, 1982; Cameron, 1985). Because of the opacity of the nebula fragments, they will feel a 
net radiation pressure from the proto-Sun, forcing material together. At some point the density 
of the fragment will rise sufficiently for self-gravity to take over and for the fragment to collapse 
to form the proto-comet. In this manner comets could be formed in isolation and very cold, 
with virtually no processing of material from its primitive interstellar state. 

The suggestion has also been made that distant nebula fragments may form their own, 
less massive accretion discs, in which comets would form within a relatively benign environment 
(Cameron, 1973). Or perhaps the comets formed in neighboring nebula fragments around other 
stars in the same open cluster in which the Sun formed. Initial relative velocities between the 
stars might be < 1 km s"^ low enough to allow capture of comets from the other stars (Donn, 

There is no good way at present to discriminate between these various hypotheses. Each 
has its particular strong points and advocates, and each its weak areas and detractors. Comets 


may be a direct consequence of planetary formation in the outer solar system, or they may be 
a largely unrelated by-product of far more distant nebula processes. Most likely, a fraction of 
the present day comets may have formed through each of the suggested processes, and what we 
observe is a truly heterogeneous mixture of material from different formation sites and with 
different degrees of physical processing. 

Some attempts have been made to look for "cosmic thermometers" in comets that might 
indicate the heliocentric distance of their formation zones, or isotopic anomalies that could be 
interpreted as evidence for formation elsewhere in the galaxy. In the case of the former, 
observations of Sj in comets have been put forward as evidence for a low temperature formation, 
T < 25 K (A'Heam and Feldman, 1985). Also, measurement of the ortho-para ratio for water 
vapor in Comet Halley by Mumma et al. (1988) found values of 2.2 - 2.3, interpreted as 
implying a nuclear spin temperature of 25 K. This, in turn, was interpreted as implying the 
temperature in the Halley formation zone, placing it beyond the Uranus-Neptune zone, possibly 
in the Kuiper belt. Similar measurements for Comet Wilson, a dynamically new comet from 
the Oort cloud found a higher ortho-para ratio of 3.2, consistent with statistical equilibrium. 
Mumma et al. interpret this higher value as a resetting of the ortho-para ratio in the comet's 
outermost layers by cosmic ray bombardment while the object was resident in the Oort cloud. 
However, this is still a relatively new technique and both the Halley and Wilson interpretations 
need to be checked against the statistics of a larger number of objects. 

As for isotopic ratios, Eberhardt et al. (1987) measured the deuterium-to-hydrogen ratio 
in Comet Halley from the Giotto spacecraft and found a value of 0.6 x KT* < D/H < 4.8 x 10" 
*. This range of values, shown in Figure 2 along with D/H measurements for other solar system 
and galactic reservoirs, is comparable to that for the Earth's oceans and for primitive meteorites, 
possibly indicating a related origin. Eberhardt et al. also measured the ^*0/'^0 ratio in Halley 



(— \ 











^— ' 










































1 — ' 









— ' 




N ' 


V— ^ 

> — ' 









> — ' 










•— ■ 









V— / 























>— i 

























•— ■ 





















4: ^ K-f-^ 

^ ^ e 4> ^ ^ 




















































re 2P 


and found a value of 0.0023 ± 0.0006, in agreement with the terrestrial value of 0.00205. Both 
the D/H and **0/'*0 ratios suggest that comets formed out of material that was isotopically 
similar (if not identical) to the solar system. Similarly, early measurements of the '^C/'^C ratio 
in comets yielded values around 100, close to the terrestrial value of 89 (Vanysek and Rahe, 
1978). However, spectral measurements of the '^C/'^C ratio in Comet Halley (Wyckoff et al., 
1989) found a somewhat lower value of 63 -65, less than the terrestrial value but comparable 
to that for interstellar clouds near the Sun. At the same time Wyckoff et al. found nitrogen 
isotope ratios > 200, consistent with the bulk solar system value of 250. Thus, most isotopic 
evidence gives a good match to terrestrial and/or solar system values, but there are still 
unexplained differences. 


Following their formation, however it might occur, the cometary nuclei will undergo 
warming due to long and short-lived radionuclides mixed in with the other nonvolatiJes. A study 
of the internal temperatures of icy satellites and planetesimals in the outer solar system by Lewis 
(1971), assuming a chondritic composition diluted 3:1 by ices and with K, U, and Th as 
radioactive heat sources, and no solar heating, found that the internal temperatures of objects 
< 300 km in radius would be < 25 K. That study assumed a thermal conductivity typical of 
solid water ice, considerably higher than the conductivities suggested for cometary interiors. 
However, assuming more reasonable conductivities would still result in very little internal 
heating for bodies < 30 km in radius (about five times the mean radius of the nucleus of Comet 

However, if short-lived radionuclides are present, then the results can be very different. 


^*A1 with a half-life of 7.4 x lO' yr has been suggested as an early solar system heat source, and 
evidence for it has been found in inclusions in carbonaceous chondrites. If similar 
concentrations were incorporated into the comets, then, assuming a low thermal conductivity, 
it would be possible to melt the interiors of nuclei greater than 3 to 6 km in radius (Wallis, 
1980; Prialnik et al., 1987). This could lead to an interesting nucleus structure, melted and 
refrozen at the center, and least modified in the near surface layers. Such a comet may display 
interesting changes in behavior as it aged and the outer layers sublimed away. 

Or, it may be that the cometary formation process took long enough that short-lived 
radionuclides were exhausted before they could be buried in large (kilometer sized) bodies. 
However, planetesimal formation times in the outer solar system are estimated to be -~ 10^ to 
10^ years, and so ^*A1 likely has played some role in the early internal heating of the cometary 
nuclei, at least those formed in or near the planetary region. 

One stage in the cometary formation process that has received very little attention to date 
is the effect of the Sun's T Tauri phase on the comets. This period is characterized by the 
development of a very strong stellar wind and substantial mass outflow, presumably clearing 
away the remnants of the solar nebula. Already formed planetesimals and planetary embryos 
will likely survive this period, but all gas and fine dust will be swept away. The T Taiui stage 
is expected to begin within — 10^ years of the initial nebula collapse, and thus favors 
hypotheses which lead to rapid giant planet formation. That time scale is much less than the 
very long dynamical clearing times for planetesimals in the Uranus-Neptune zone. Thus, most 
comets will likely still be relatively close to the Sun and will suffer considerable near-surface 
modification as a result of the intense solar wind. The modifying processes will be similar to 
the irradiation by galactic cosmic rays in the Oort cloud (see section 4 below) which causes 
sputtering of volatiles and polymerization of hydrocarbons in the upper several meters of the 


nucleus surface. Even if comets formed farther from the sun, the T Tauri stage may still result 
in significant physical modification. As noted above, this is an area that deserves further study. 
A second area that has received relatively little consideration is the collisional evolution 
of the cometary nuclei in the Uranus-Neptune zone prior to their ejection to the Oort cloud. As 
described above, the growth of planetary embryos in that zone will accelerate the remaining 
cometesimals, raising encounter velocities and causing a transition from accretionary to 
disruptive collisions. Low velocity collisions can be expected to cause compaction at contact 
interfaces and some local heating (Donn and Meakin, 1988), while high velocities will result in 
shock, intense heating, and fragmentation. By the time the nuclei are finally ejected to the Oort 
cloud they will likely have been disrupted and reassembled several times, resulting in a chaotic 
collection of collisionally processed and unprocessed materials. Clearly, this area merits further 
study as well. 


At first glance, the Oort cloud would appear to present a relatively benign environment. 
The average spacing between comets is on the order of 15 AU or more in the outer, dynamically 
active cloud, and about 1 AU in the inner Oort cloud. The maximum surface temperature on 
cometary nuclei ICP AU from the Sun is only 13 K; comets at greater distances would be even 

However, four physical processes have been identified which modify the outer layers of 
cometary nuclei while they are in the Oort cloud. These are: irradiation by galactic cosmic rays 
and solar protons, heating by luminous passing stars and supemovae, gardening by debris 
impacts within the Oort cloud, and accretionary and erosive interactions with the interstellar 


medium. As described below, each of these processes acts on the outermost 0.2 - 20 meters of 
all comets in the Oort cloud. 

The study of radiation bombardment of icy surfaces has been undertaken in a number of 
laboratories (Shulman, 1972; Lanzerotti et al., 1978; Greenberg, 1982; Moore et al., 1983; 
Lanzerotti, 1983; Draganic and Draganic, 1984; Calcagno et al., 1985; Johnson et al. 1987; and 
Thompson et al., 1988). Energetic particles penetrate the nucleus surface to a depth which 
depends on both the energy of the particles and the density of the nucleus surface layers. 
Assuming a unit density material, low energy solar protons of a few, to a few hundred KeV 
penetrate only the first millimeter of surface or less; cosmic ray protons with typical energies 
- 2 MeV penetrate about one meter. Since penetration depth goes inversely as density, lower 
density ices suffer damage to proportionately greater depth. The compositional and structural 
evolution occurs in the uppermost ~ 100 g cm"^ of the cometary surface layers. 

The effect of the irradiation is to break chemical bonds, producing volatile free radicals, 
some of which recombine to form a dense, dark polymer which is far less volatile than the 
original material, and some of which provide a latent energy source when the ices are warmed. 
Polymerization occurs particularly if hydrocarbons (e.g., CH4, NH3) are present. At the nucleus 
surface the energetic particle bombardment results in a net erosion by sputtering, and likely 
escape of the more volatile species, leaving behind a low volatility residue. At depth the 
volatiles are retained, but changed. One important result is the development of nonvolatile 
crusts (also sometimes referred to as mantles) on comets while they are still in the Oort cloud 
and before they ever enter the planetary region. This crust may begin the process of sealing off 
the nucleus surface against sublimation, well before the expected cnistal development from solar 
heating during perihelion passage (see section 6, below). 

Comets in the Kuiper belt likely experience a less intense cosmic ray processing history 


than those in the inner and outer Oort cloud. Because the expected location of the Kuiper belt 
is partially, if not totally, within the Sun's heliosphere, the comets are likely shielded from the 
full effect of galactic cosmic rays. Thus, the degree of processing at depth will be somewhat 
less than that for the more distant Oort cloud comets. 

The second process which acts to modify cometary surfaces in the Oort cloud is the 
passage of luminous stars through or near the cloud (Stem and ShuU, 1988). It has long been 
recognized that random passing stars regularly penetrate the Oort cloud (Weissman, 1980). This 
process results in the ejection of comets close to the path of the passing star (~ 500 AU for one 
Mo stars), diffusion of comets from the inner cloud to the outer cloud, and the randomization 
of orbits in the outer cloud. Stem and Shull (1988) calculated the importance of such encounters 
for the heating of comets in the cloud. They found that although the number of stellar 
encounters with the Oort cloud (several times 10^) is dominated by solar type stars and white 
dwarfs, these stars are not sufficiently luminous to heat a significant fraction of the comets. 
Also, the effective ejection radius for solar mass stars exceeds the heating radius. 

However, luminous O and supergiant stars were found to be capable of heating the entire 
Oort cloud to temperatures sufficient to induce selective mobilization of species such as CO, Ar, 
Ne, and CEj in the outermost layers of the nuclei. For example, a typical star with 
luminosity 5 x 10^ L© can heat all the comets in the Oort cloud to - 20 K from about a parsec 
away. At such distances, its gravitational perturbations on the cloud are modest. Such 
encounters typically last a few times 10^ years and can raise the temperature of cometary 
surfaces to between 16 and 24 K. Because vapor pressures are exponential in their temperature 
dependence, this temperature regime contrasts sharply with the ambient background temperature 
of 5 - 6 K in the Oort cloud between luminous star encounters. Stem and Shull estimated that 
there has been a 100% probability of an encounter with a luminous star which raised the Oort 


cloud to 19 K, and a 10% chance of an encounter which heated the cloud to 34 K. 

Stem and Shull also estimated the effects of nearby supemovae on Oort cloud heating. 
They found that a Type I supernova exploding at 19 parsecs can raise comet surface 
temperatures to 30 K. Using supemovae luminosity and formation rate statistics (Tammann, 
1982; Narayan, 1987), it was found that aU comets in the Oort cloud have been heated by 
supemovae to 50 K, and that there is a 50% probability that they have been heated to 60 K. 
Stem and Shull' s results are summarized in Figure 3. Although supemovae can heat comets to 
higher temperatures than O stars, supernova events last only a few weeks. The thermal skin 
depth during a typical 3 x 10^ year stellar encounter with a luminous O star is 20 to 140 meters, 
depending on the porosity and the amorphous/crystalline nature of the surface ices. Supemova 
encounters, while hotter, are briefer (about 10' sec); therefore, their heating pulse is estimated 
to propagate only 0. 1 to 1 meter into the comets. Heating of Kuiper belt comets by passing stars 
and supemovae will likely not be significant since their closer distance to the Sun already warms 
them to comparable temperatures. 

Collisions between comets, and between comets and small debris in the Oort cloud, is 
the third process affecting their evolution (Stem, 1988). A characteristic collision velocity in 
the inner Oort cloud is ~ 1 km s '^ Although individual comet-comet collisions were found 
to occur every few months in the inner Oort cloud and every few centuries in the outer cloud, 
less than one comet in 10* suffers such a collision over the age of the solar system. However, 
using constant area and constant mass per bin power-law size distributions, bracketing the size 
distribution for main belt asteroids, Stem showed that collisions between comets and small debris 
in the Oort cloud occur quite frequently. Such collisions provide a strong feedback mechanism 
for the production of additional small debris in the Oort cloud. For comets in the inner Oort 
cloud the frequency of collisions is sufficient to have overtumed the entire surface several times 


























10 20 30 40 50 60 70 



80 90 100 

Figure 3. Fraction of comets heated by all types of passing stars as a function of processing 
temperature (heavy curve); all comets have been heated to - 16 K, and there is a 10% 
chance that the cloud may have been heated to > 34 K. Fraction of comets heated by 
supemovae as a function of T during encounters (light curve); all comets have been 
heated to > 45 K, and there is a 10% chance that a supernova has occurred close 
enough, < 8 parsecs, to heat the Oort cloud to ~ 60 K. A surface emissivity of 0.8 
and an albedo of 0.05 were assumed. Adapted from Stem and Shull (1988). 


to a depth of 0.5 to 5 meters (assuming a surface density of 1 g cm"^), thereby promoting 
regolith development during residence in the Oort cloud. Collision rates in the Kuiper belt may 
even be higher. 

The fourth and final processes known to cause cometary evolution in the Oort cloud is 
the interaction of comets with the interstellar medium (ISM). O'Dell (1971) investigated the 
accretion of interstellar dust and gas and found that a layer of 10 - 100 micrometers thickness 
would be built up over a period of 4.5 x 10* years. It had been speculated (Whipple, 1978) that 
this layer of interstellar volatiles accounted for the anomalous brightness of dynamically new LP 
comets on their first pass through the planetary region. However, Stem (1986) showed that 
erosion due to hypervelocity dust impacts is some 7(X) to 10(X) times more efficient than gas 
accretion, causing comets to actually loose material due to the ISM interaction. The rate and 
energy of the impacts is controlled by the Sun's peculiar velocity through the ISM, ~ 20 km 

A more recent and comprehensive study which includes grain erosion, accretion due to 
molecular sticking, gas sputtering, and thermal evaporation due to grain impacts, and which 
employs a realistic model of the various gas and cloud phases of the ISM, including giant 
molecular clouds, has been completed (Stem 1989). Taking into account the results of 
laboratory studies of micro-cratering in ices (Cintala, 1981; Croft, 1982; Lange and Ahrens, 
1987), Stem found that erosion causes comets in the Oort cloud to lose 1.0 to 5.0 meters of 
surface material (for a density of 1.0 g cm'^) over 4.5 x 10' years (Figure 4). Thus, erosion 
may remove some or all of the radiation and heat processed crust on the cometary surfaces. 

On the other hand, recent laboratory experiments simulating capture of cometary dust 
grains during high speed spacecraft flybys have suggested that impacts into a very low density, 
"fairy castle" surface structure may lead to intact capture of the particles, with little net mass 












li 0.001 







-I — I — I — I I I I 

I I 1— I 1 — TT- 

' 1111 L. 

' ■ I ] I I 

10^ 10^ 

TIME (years) 



Figure 4. Integrated surface mass loss due to hypervelocity dust impacts experienced by 
comets and other debris in the Oort cloud, as described by Stem (1989). This model 
assumes an icy cometary surface and a four-phase ISM consisting of coronal gas, warm 
H I gas, and atomic and molecular clouds distributed in proportion to their galactic 
volume filling factors is assumed. The greatest erosion results from passages through 
giant molecular clouds. 


loss (Tsou et al., 1984). The experiments have generally been done for relatively modest 
velocities, on the order of a few km s'K Whether or not the same effect would be true for the 
20 km s'* or more velocity expected for interstellar grains is open to-some doubt. But it is 
possible that continued experiments might show that the erosion is not as significant as described 
above, and that there is, in fact, a net mass gain, as well as a slow compacting of materials, in 
the surface layers of cometary nuclei (Ostro et al., 1986). 

It is clear that the Oort cloud is not just a storage bin in which comets reside in 
suspended animation. Instead, a number of important effects operate in an interrelated manner 
to evolve cometary surfaces. In addition to the four modifying processes discussed here, there 
may be other as yet undiscovered processes which affect the comets during their long residence 
in the Oort cloud. For example, creep of icy materials in the weak cometary gravitational field 
may lead to a compacting and densification of the nucleus. Thus, although comets may be 
ejected to the Oort cloud as low density, fractal-structured objects, they may return after 4.5 Gyr 
as single, well compacted nuclei with a density more typical of ordinary water ice, or at least 
a compacted snow bank. 


There are a variety of dynamical paths that a LP comet might take from the Kuiper belt, 
or the inner or outer Oort clouds, to become a visible SP comet. The relative efficiencies of the 
different paths vary by orders of magnitude, but because of the very large number of comets in 
the Oort cloud reservoir, any of the major paths discussed below have a finite probability of 
producing a particular SP comet. 

Put another way, it is not possible to integrate the orbit of a known SP comet backwards 


in time to learn how it evolved into the planetary system. Small errors in the initial orbit 
determination or in the masses of the perturbing planets, unpredictable nongravitational forces 
from jetting of volatiles on the nucleus surface, and other unmodeled forces lead to a monotonic 
growth of uncertainty as the orbit is integrated backward. 

Thus, the problem is limited to probabilistic arguments based on the relative efficiency 
of the various dynamical paths. In addition to evolution inward from the Kuiper belt, two other 
dynamical routes have been identified for the origin of the SP comets: 1) repeated perturbation 
of dynamically new LP comets from the Oort cloud passing through the observable region 
(Everhart, 1969); or, 2) repeated perturbation of dynamically new LP comets with perihelia in 
the Jupiter-Saturn zone, with a final dumping of cometary perihelia into the terrestrial planets 
zone (Everhart, 1972). The advantage of the second path over the first is that it allows comets 
to spend most of their dynamical evolution at solar distances where water ice sublimation is 
negligible, preserving the comets physically so that they are still active when they become SP 
comets. However, both paths are estimated to be a factor of Itf to 10^ less efficient than the 
Kuiper belt for producing SP comets (Fernandez, 1980; Bailey, 1983). 

Once deposited in a SP orbit, through whatever dynamical path, a comet will continue 
to random walk in both perihelion distance and semimajor axis as a result of planetary 
perturbations. For example, Monte Carlo studies of the evolution of meteoroids in the planetary 
region (Arnold, 1965) showed that 82% of all meteors which struck the Earth had at one time 
or another been within Mercury's orbit. To first order, those studies are equally applicable to 
the random walk of SP comet orbits. Similar evidence can be found among the observed SP 
comets. Since its discovery in 1902, periodic comet Grigg-Skjellerup has evolved from an initial 
perihelion distance of 0.753 AU, to 1.003 AU in 1967, and 0.989 AU currenUy (Marsden, 
1986). Even more dramatic is the case of comet P/Lexell. It was discovered in 1770 when it 


approached to within 0.015 AU of the Earth, with an orbit perihelion of 0.67 AU. Integration 
of the orbit backwards in time (Kazimirchak-Polonskaya, 1972) showed that Lexell had passed 
within 0.02 AU of Jupiter in 1767, and had previously had a perihelion of about 3 AU. After 
two passages around the Sun in its new orbit, the comet re-encountered Jupiter in 1779 at a 
distance of 0.0015 AU (half the radius of Jo's orbit) and was perturbed to a new orbit with a 
perihelion beyond Jupiter's orbit. 

Thus, for any observed SP comet, it is impossible to say where in the solar nebula it 
formed, and where it has been since the time of formation. A comet could have spent a part 
of its past in the inner solar system, then have been returned to the Oort cloud, and subsequently 
evolved back into the planetary region a second time. The fact that a comet's current perihelion 
is relatively large is no proof that the comet has not been much closer to the Sun in the not too 
distant past. Similarly, comets which presently have small perihelia may be recently arrived at 
those close solar distances, and may not yet have been heated very deep within the nucleus, 

A possible example of the latter is P/Wild 2. The comet was deposited in its present 
orbit with a perihelion of 1.49 AU by a 0.0061 AU encounter with Jupiter in 1974 (Carusi et 
al. , 1985). Prior to that time the comet had a perihelion near Jupiter's orbit. Integration of the 
orbit further backwards shows it being captured from a near-parabolic orbit about 400 years ago. 
However, as described above, such integrations are not reliable. They provide a representative 
example of possible cometary motion, rather than a detailed reconstruction of it. 


When cometary nuclei return to the planetary region as LP and SP comets, their physical 


evolution is dominated by the heating they receive from direct solar radiation. Other processes 
such as irradiation by solar wind protons and impacts by interplanetary dust particles will also 
intensify with decreasing heliocentric distance, but they do not compare in either mass removal 
rate or depth of penetration with the changes brought about by solar heating. The effect of the 
heating will manifest itself in a number of ways. 

First will be conversion of amorphous water ice to the crystalline form. Prialnik and 
Bar-Nun (1987) showed that the slow heating of the surface of an amorphous ice nucleus, 
presumably formed at low temperature, < 100 K, would cause a transition to crystalline ice at 
about 5 AU inbound on its first perihelion passage. The amorphous to crystalline phase 
transition occurs at — 140 K. Since this is an exothermic reaction, an additional heat pulse 
would push inward converting a layer 10 to 15 meters thick to crystalline ice. Prialnik and Bar- 
Nun showed that a chain reaction converting the entire nucleus to crystalline ice does not occur: 
the pulse is eventually dissipated as it reaches colder ice layers at greater depths, and by 
warming of nonvolatile dust mixed with the ice. The amorphous to crystalline ice transition may 
supply sufficient energy to blow off pieces of the primitive irradiated crust, resulting in the 
anomalously bright behavior at large solar distance often displayed by dynamically "new" LP 
comets. Subsequently, the amorphous-crystalline transition does not repeat for several orbits, 
until a sufficiently thick layer of the overlying crystalline ice is sublimated away, and the orbital 
heat pulse can penetrate to the buried amorphous core. A more detailed model by Prialnik and 
Bar-Nun (1988) which includes the effect of a nonvolatile dust crust on the nucleus surface (see 
below) found that the minimum thickness of the crystalline ice layer overlying the amorphous 
ice was ~ 15 meters, and was often much thicker, ~ 25 to 40 meters. 

The second process is the sublimation of volatile ices at the nucleus surface, which results 
in the development of the extended cometary atmosphere, the coma. The evolving gases will 


cany with them solid grains of dust and ice, creating the dust coma and (perhaps) an icy- 
hydrocarbon halo around the nucleus, respectively. Larger grains of nonvolatiles that are not 
carried off will begin to form a lag deposit on the nucleus surface, accumulating to form a crust 
that will begin to seal the nucleus off against further mass loss, and thermally insulate the 
nucleus ices interior to it (Brin and Mendis, 1979; Prialnik and Bar-Nun, 1988). 

There are thus, two possible scenarios for crust formation. The comet may retain some 
or all of its original cosmic ray irradiated crust, and this serves as the foundation for additional 
crust growth. Or, debris gardening and the energy release from the amorphous-crystalline ice 
phase transition may blow away all of the primitive crust, and the comet then grows a new lag 
deposit of heavy nonvolatile grains, possibly glued together by complex organics. 

Violent rupture of the crust due to the pressure of evolving gases below it may result in 
visible outbursts or even disruption of the nucleus. Further heating of amorphous ice within the 
nucleus will result in sporadic transitions to crystalline ice, also possibly resulting in visible 
outbursts. Thermal stresses on the nucleus caused by substantial temperature gradients within 
the ice may also result in cracking and exposure of "fresh" ices, and possibly outburst or 
disruption phenomena. 

More subtle effects will include the migration of highly volatile molecules, both outward 
through the still frozen water ice matrix, and inward towards cooler regions of the nucleus where 
they will recondense. Also, as mass is lost from the rotating nucleus, its moments of inertia will 
change and the nucleus will precess, changing the orientation of the rotation pole and the balance 
of solar insolation across the nucleus surface, with perhaps additional interesting implications. 
For example, the changing axis orientation will change the centrifugal stresses on weakly bonded 
nucleus fragments, perhaps resulting in their breaking off and appearing as secondary nuclei. 

For the purpose of analyzing the relative pristinity of cometary nuclei for a sample retum 


mission, the important factor will be the heating of the cometary nucleus at depth. The "thermal 
skin depth," 5, is defined as 

5 = (KP/xpC)''^ ^ (1) 

where K is the conductivity, P is the period (length of day for diurnal skin depth, orbital period 
for orbital skin depth), p is the density, and C is the specific heat. The thermal skin depth is 
the distance over which a temperature perturbation at the surface will decrease by a factor of 
1/e. For conductivities typical of solid, crystalline water ice, 5 = 20 cm for a rotation period 
of 24 hours, or 9.2 meters for an orbital period of 6 years. However, conductivity decreases 
sharply for porous, low density structures as are suspected for cometary nuclei, and actual values 
of 5 may be considerably smaller. Most measured values of surface conductivity in the solar 
system, including the icy Galilean satellites, are extremely low. If comets are similar, then a 
more realistic estimate of the orbital thermal skin depth may be ~ 1 meter. This is a key 
measurement that can be made by the CRAF Penetrator experiment. 

Internal temperature profiles for cometary nuclei are a complex function of the comets' 
orbital parameters, rotation pole obliquity, and the thermal properties of the comet surface 
materials. For example, the variation of temperature with depth for a typical 1 km radius SP 
comet nucleus, ignoring the diurnal temperature cycle and sublimation, and assuming solid 
crystalline ice, is shown in Figure 5 for four points around the comet's orbit (Herman and 
Weissman, 1987). Although the surface layer undergoes extreme temperature variations, the 
temperature several orbital thermal skin depths below the surface is virtually constant. Below 
this depth the nucleus has been heated to its average orbital temperature, but not appreciably 

Temperature profiles in a real case will not be as smooth as shown in Figure 5. Because 
of the typically low values of conductivity, many returns will be required for a comet to be 





220 - 

200 - 

180 - 


140 - 


1 1 




"" APHELION V / / 

'3AU ~ 


I 1 1 





980 1 000 


RADIUS - meters 

Figure 5. Equilibrium temperature profiles in the near surface layers of a hypothetical 1 
km radius crystalline ice comet in a short-period orbit, for four points around the orbit. 
From Herman and Weissman (1987). 


heated to its equilibrium internal temperature. But the comet's orbit is likely to change 
significantly over that same period of time. Since the changes are essentially random, some will 
result in additional internal heating, while others will result in a net cooling of the cometary 
nucleus. The final result might be a fairly complex interweaving of warm and cooler layers. 
Nevertheless, the equilibrium internal temperature of cometary nuclei can be approximated to 
within + 10% by the mean temperature for a nucleus in a circular orbit with the same semimajor 

T„ = 280 (1 - A)^"* a-^'^ e""' K (2) 

where A is the bond albedo of the nucleus surface, a is the semimajor axis in AU, and e is the 
surface emissivity. For comets in non-circular orbits the problem cannot be solved analytically, 
so numerical techniques must be employed. 

The intense solar heating near perihelion and increased time for cooling near aphelion, 
as well as the temperature dependence of the thermal conductivity, lead to a complex behavior 
of the central temperature as a function of semimajor axis and eccentricity. That behavior is 
illustrated in Figure 6 for the case of a 1 km radius crystalline ice nucleus (Herman and 
Weissman, 1987) where the "normalized temperature" is the calculated central temperature 
divided by the mean temperature from Equation 2. 

The final central temperature which the nucleus reaches is also a function of the 
magnitude of the thermal conductivity, but is not a function of the nucleus radius (Herman and 
Weissman, 1987). However, larger nuclei could take hundreds or even thousands of orbits to 
reach equilibrium, longer than the time constant for appreciable changes in the comet's orbit due 
to Jupiter perturbations. 

As a result of this internal warming, volatile ices with sublimation points below the 
central temperature will diffuse through the icy matrix and possibly escape the nucleus, unless 



0.4 0.6 


Figure 6. Variation of central temperature for short-period comets as a function of orbital 
semimajor axis and eccentricity. The normalized temperature is tliat for a comet in a 
circular orbit: see text. From Herman and Weissman (1987). 


they are trapped as clathrate hydrates. The near surface layers will certainly be depleted in 
many of the more volatile ices. As a result, gas production rates for some volatile species may 
show a temporal dependence: as the comet approaches perihelion, the. more volatile ices may 
not appear until the thermal wave has penetrated deep enough into the nucleus to reach 
undepleted layers. In addition, some deeper layers in the nucleus may be enriched in certain 
volatiles which were sublimated from the warm near-surface layers, and were able to diffuse to, 
and recondense in the cooler interior. 

Calculated central temperatures for eight well known SP comets, most of which are under 
consideration as targets for comet rendezvous and/or sample return missions, are given in Table 
1. The quantity T^ is the equilibrium central temperature for the comet's present orbit, assuming 
normal water ice sublimation over the entire nucleus surface. T/ is the same but assuming no 
sublimation, a condition the comet would evolve to as it built up a nonvolatile crust covering 
most of the nucleus surface. 

Note that with the exception of P/Halley, all the SP comets in Table 1 have a central 
temperature above 140 K, the transition temperature for amorphous to crystalline water ice. 
Thus, it is highly unlikely that any comet sample return mission will bring back a sample of 
amorphous ice; this certainly represents a major departure from the state in which the comet 
formed. For a low albedo nucleus, Equation 2 gives a central temperature of 140 K at a 
semimajor axis of 4.0 AU, or an orbital period of 8.0 years. Of 135 known periodic comets 
in the most recent catalog (Marsden, 1986), 82 have periods less than that value. Thus, those 
comets can be expected to have completely converted to crystalline ice in their interiors. 

For those comets with longer orbital periods, or those comets recently arrived in the 
inner planets region, a sizeable fraction of the nucleus interior can still be expected to be 
amorphous ice (provided that the comet has not been significantly closer to the Sun in the past). 


Table 1. Estimated central temperatures for several 
short-period comets.* 

'From Herman and Weissman (1987). 


q- AU 


T, -K 

T,' -K 











Tempel 2 





Wild 2 


























However, according to Prialnik and Barn-Nun (1988), the amorphous ice is located at least 15 
meters, and possibly 40 meters, beneath the surface. 

A variety of models have been developed to study crust growth on cometary nuclei. 
Calculations have shown that a layer only one or two centimeters thick would probably be 
sufficient to insulate the ices below and to greatly reduce the sublimation rate (Brin and Mendis, 
1979; Fanale and Salvail, 1984; Horanyi et al., 1984). Estimates of the number of returns 
required to form the crust range from — 1 to 20. Various theories have proposed that the crust 
is either porous, allowing continued gas diffusion through it to the surface, or sealed, resulting 
in a buildup of pressure beneath it that might cause violent rupture events. The results of the 
spacecraft flybys of Comet Halley have produced strong evidence for the existence of an 
insulating crust on the nucleus surface (Sagdeev et al., 1986, Keller et al., 1986), but at the 
same time have presented us with several apparent paradoxes. The fraction of active area on 
the nucleus of the comet has been estimated at 20 - 30% of the sunlit surface, based on 
spacecraft imaging and predictions of the gas production rate (Weissman, 1987). However, it 
is not clear why such a small fraction of the Halley nucleus is active. Virtually all predictions 
for Halley were that the nucleus surface would be crust free, the crust having been blown away 
by the high gas production. 

Given that 70-80% of the nucleus surface was crusted over, why was it not 100%? Did 
the few active areas serve as pressure release points for the entire nucleus, implying a highly 
porous nucleus structure? Did the active areas change as the comet moved along its orbit, or 
from one perihelion passage to another? Earth-based observations of the comet suggested that 
the active areas did turn on and off irregularly, and Giotto imaging suggested at least one surface 
structure that may be a crusted over, former active area. On the other hand, some observers 
alleged a link between active areas observed in 1985-86, and locations of active areas derived 


from 1910 observations (Sekanina and Larson, 1986), a very tenuous possibility which is 
exceedingly difficult to prove. 

A further question is whether or not the allegedly inactive areas on the Halley nucleus 
were really inactive? Clearly, the sources of the dust jets appear to be confined to relatively 
small areas. But is it possible that gas from sublimating ices is still diffusing through the dark 
porous crust in other areas, while being unable to carry entirained dust with it? The relatively 
poor resolution of the Halley images does not permit an answer to this question. 

Another piece of this complex puzzle comes from estimates of nongravitational forces 
on the motion of Comet Halley. The highly elongated nucleus seen in the spacecraft imaging, 
with its irregular and asymmetrically distributed active areas, would be expected to precess 
rapidly. But the fitted nongravitational forces on the comet have been virtually constant over 
the past 2,200 years (Yeomans and Kiang, 1981). How then can the nongravitational forces be 
so constant when the nucleus appears to be so dynamic and ever-changing? 

Perhaps the best illustration of the complexity of this problem is to cite a real example. 
In Alaska, a volcanic dqx)sition of ash over a snowbank at the base of a glacier resulted in the 
unusual cone shaped sublimation features shown in Figure 7 (photograph courtesy of M. Malin). 
Note that this strange topography is local to one area on the glacier, suggesting a unique 
combination of both location and lighting that led to it. It is highly unlikely that any scientist 
would have predicted such a topography from first principles, even knowing far more about the 
detailed conditions in Alaska versus our current, relatively poor knowledge of cometary nuclei. 

What this all must lead to is a major rethinking of tiie physics and chemistiy of the crust 
formation and removal process. Present models simply cannot explain the observations of the 
Halley nucleus. Further analyses of the Halley spacecraft data will yield a refined definition of 
the problem. But extensive laboratory work and theoretical modeling, as well as in situ 


Figure 7. Sublimation induced features due to deposition of volcanic ash over a snowbank 
at the snout of one of the Knife Creek Glaciers in the Valley of Ten Thousand Smokes, 
Katmai, Alaska. Local melting may also have played a role in formation of these 
unusual cones. It is highly unlikely that such features could be predicted from an 
examination of the theory of ice sublimation. What then will sublimation features on 
cometary nuclei look like? This illustrates the complexity of the processes and 
interactions that may occur, and the reason why in situ measurements are imperative. 
Photograph courtesy of M. Malin (VTTS83-7-18). 


observations of an evolving cometary nucleus over a sizeable fraction of its orbit, as will be 
provided by the Comet Rendezvous Asteroid Flyby (CRAF) mission, are needed to completely 
understand crust formation. 

Another area of considerable interest is that of thermo-mechanical stresses on cometary 
nuclei (Kuhrt, 1984; Green, 1986; Tauber and Kuhrt, 1987). The dust jets in the Halley 
spacecraft images appear to be highly collimated, and appear to occur along linear features 
(Sagdeev et al., 1987). This suggests deep crevices opened up by stresses on the nuclei, or 
perhaps freshly exposed voids in a fractal structure. These crevices may penetrate sufficientiy 
deep in the nucleus to reach relatively unmodified volatile rich layers, several thermal skin 
depths below the crusted surface. They also may be the best sites for obtaining samples of the 
cometary nucleus. But little is known yet about thermal stresses and the two groups working 
on the problem can not even agree on whether the stresses are tensional or compressional. More 
work in this area is clearly needed. 


The sections above have described a number of physical processes, summarized in Figure 

8, which have modified cometary nuclei since their formation in the primordial solar nebula, 4.5 
Gyr ago. Many of these processes are restricted to the nucleus surface or to near-surface layers. 
But others affect the entire nucleus, and in at least one case, short-lived radionuclides, affect the 
center of the nucleus the most. The evidence appears to be that comets have only received 
modest heating over their histories, so as to drive off and/or mobilize a fraction of the more 
volatile ices, but leaving the nonvolatile constituents, in particular those at depth, relatively 
unmodified. The two qualifications on that statement are the possible early melting of cometary 












Figure 8. Conceptual diagram of the many different processes which act to modify 

cometary nuclei over their history. The important question is to understand how these 

various individual processes interact. For a nucleus sample return mission, the most 

primitive samples are likely to be found at least 5 to 10 meters beneath the surface, or 

even deeper, but well away from the core. 


interiors by short-lived radionuclides, though the effects of this are hidden deep below the 
nucleus surface, and the polymerization of hydrocarbons and other materials in the upper 1 to 
10 meters of the nucleus surface, though the polymerized layer may have since been lost. 

The greatest degree of processing has almost certainly occurred at and near the nucleus 
surface, as a result primarily of solar heating, gardening from small debris impacts, and cosmic 
ray bombardment. To find relatively unprocessed materials requires penetrating beneath those 
processed layers. In the case of solar heating that would likely require a depth of several to tens 
of meters. For gardening the expected depth is 5 - 10 meters, but that layer may have already 
sublimated away. For cosmic ray bombardment the required depth would again be several 
meters, but it is again likely that the modified layer may have already been completely lost due 
to sublimation and crust blow-off. Or, it may have provided the basis for later crust growth and 
thus, could still be there. 

The important question is how do all these various processes interact? For example, 
debris impacts in the Oort cloud would repeatedly break up the cosmic ray irradiated crust, with 
most of the ejecta escaping in the weak cometary gravity field. Hypervelocity dust impacts into 
an underdense regolith may lead to dust capture, crust compaction, and densification, which in 
turn may then lead to erosion by subsequent dust impacts, and so on. More complex models 
of these processes seem necessary, but our current knowledge of cometary materials is likely 
too poor to create adequately constrained models. 

Observational evidence provides considerable support for the belief that comets still retain 
a very large fraction of their original volatile and nonvolatile constituents. The most abundant 
molecules detected in interstellar clouds have also been detected in cometary spectra (Irvine et 
al., 1985). Compositional measurements of solid grains during the Halley flybys show a mix 
of high temperature refractory grains and pure hydrocarbon grains, as well as more complex 


grains with a heterogeneous composition (Kissel and Krueger, 1987; Jessberger and Kissel, 
1987). Recovered interplanetary dust particles, which are believed to come from comets, have 
a botryoidal or fractal structure of sub-micron grains, much like Figure~lb, with a composition 
similar to the most primitive, undifferentiated meteorites (Fraundorf et al., 1982). In some cases 
the observations cited here are for "fresh" LP comets from the Oort cloud, but in other cases 
the evidence has also been found in SP comets, particularly when observers have looked hard 

An understanding of the modification processes discussed above is important to the 
planning of targeting and sampling strategies for a comet nucleus sample return mission. 
Although the ultimate choice of target comet will be made with numerous operational 
considerations in mind, processing considerations should also be factored into the selection. The 
optimum choice is a comet new to the inner planetary region, though there is still the problem 
of drilling through the meters-thick cosmic ray, thermal, and collisionally modified surface 
layers. If the target comet has been resident in a SP orbit for hundreds or even thousands of 
returns, it will have built up a substantial lag deposit crust and solar heated near-surface layers 
that may be several to tens of meters thick (Fanale and Salvail, 1984; Herman and Weissman, 
1987). Unfortunately, most SP comets reachable by planetary exploration spacecraft fall into 
this latter category. P/Wild 2 may be a notable exception, because of its recent perturbation into 
a small perihelion orbit. 

Concerning sampling strategy itself, any sample returned from the uppermost few meters 
of a short-period comet will reflect the damage of a variety of modifying processes; it will likely 
be quite far from pristine. To obtain the most primitive sample possible, two strategies are 
suggested. First, the maximum technically feasible drilling depth should be obtained. Because 
most processing scale lengths are on the order of a few meters and because damage should 


generally decrease exponentially with depth, one needs to sample many meters below the surface 
to have a reasonable chance of obtaining primitive materials. Indeed, without a deep drill 
sample return, one cannot even know how damaged the surface samples are. A particular 
advantage of deep drilling is that the bore hole can be fitted with instrumentation to measure heat 
transport and radiation penetration at depth as the comet orbits the Sun, continuing to monitor 
the nucleus after the sample return vehicle has left. 

Second, it is recommended to sample material excavated naturally at cometary vents or 
fissures. Vents may reach material at depths greater than can be reached by feasible spacecraft 
drilling techniques. Vent-excavated material might be acquired by sampling the outflow gas and 
dust before it contacts sunlight, i.e., from within the vent or during outflow after sunset. 
Although both deep drilling and vent sampling are technologically challenging, they represent 
the best chance for obtaining primitive material with a comet nucleus sample return mission. 

Finally, it is clear that collected samples should be preserved at temperatures at or below 
the temperatures at which they are collected, to minimize any ftirther processing during the 
return voyage to Earth. 

Acknowledgments: We thank M. Malin for the use of his photograph of unusual ice sublimation 
features in Alaska. S. A. Stem acknowledges support under NGT-50236 and wishes to thank 
Sherwood Chang for supporting his travel to this Workshop. This work was supported, in part, 
by the NASA Planetary Geology and Geophysics Program, and was performed, in part, at the 
Jet Propulsion Laboratory under contract with the National Aeronautics and Space 



A'Heam, M. P., and Feldman, P. D. 1985. S2: a clue to the origin of cometary ice. In Ices in 
the Solar System , eds. J. Klinger, D. Benest, A. Dollfus, and R. Smoluchowski, D. 
Reidel, Dordrecht, pp. 463-472. 

Arnold, J. R. 1965. The origin of meteorites as small bodies. Astrophys. J. 141 . 1536-1556. 

Aumann, H. H., Gillett, F. C, Beichman, C. A., de Jong, T., Houck, J. R., Low, F., 
Neugebauer, G., Walker, R. G., and Wesselius, P. 1984. Discovery of a shell around 
alpha Lyrae. Astrophys. J. 278 . L23-L27. 

Bailey, M. E. 1983. Comets, planet X and the orbit of Neptune. Nature 302 . 399-400. 

Bailey, M. E., Clube, S. V. M., and Napier, W. M. 1986. The origin of comets. Vistas in 
Astronomy 29, 53-112. 

Brin, G. D., and Mendis, D. A. 1979. Dust release and mantle development in comets. 
Astrophys. J. 229, 402-408. 

Calcagno, L., Foti, G., Torrisi, L., and StrazzuUa, G. 1985. Fluffy layers obtained by ion 
bombardment of frozen methane: Experiments and applications to Satumian and 
Uranian satellites. Icarus 63. 31-38. 

Cameron, A. G. W. 1962. The formation of the sun and planets. Icarus i, 13-69. 

Cameron, A. G. W. 1973. Accumulation processes in the primitive solar nebula. Icarus 18 . 


Cameron, A. G. W. 1978. The primitive solar accretion disc and the formation of the planets. 
In The Origin of the Solar System , ed. S. F. Dermott, John Wiley & Sons, New York, 
pp. 49-75. 

Cameron, A. G. W. 1985. Formation and evolution of the primitive solar nebula. In 
Protostars and Planets n . eds. D. C. Black and M. S. Matthews, Univ. Arizona Press, 
Tucson, pp. 1073-1099. 

Carusi, A., Kresak, L., Perozzi, E., and Valsecchi, G. B. 1985. Long-Term Evolution of Short- 
Period Comets . Adam Hilger, Bristol, 350 pp. 

Cintala, M. J. 1981. Meteoroid impact into short-period comet nuclei. Nature 291 . 134-136. 
Croft, S. W. 1981. Hypervelocity impact cratering in icy media. Lunar and Planet. Sci. Conf. 
n, 190-192 (abstract). 

Delsemme, A. H. 1982. Chemical composition of cometary nuclei. In Comets , ed. L. L. 
Wilkening, Univ. Arizona Press, Tucson, pp. 85-130. 

Donn, B. 1976. Comets, interstellar clouds, and star clusters. In The Study of Comets . 
NASA SP-393, pp. 663-672. 


Donn, B., and Rahe, J. 1982. Structure and origin of cometary nuclei. In Comets , ed. L. L. 
Wilkening, Univ. Arizona Press, Tucson, pp. 203-226. 

Donn, B., Daniels, P. A., and Hughes, D. W. 1985. On the structure of the cometary nucleus. 
Bull. Amer. Astron. Soc. H, 520 (abstract). 

Donn, B., and Meakin, P. 1988. The accumulation and structure of the cometary nucleus: the 
fluffy aggregate model. Bull. Amer. Astron. Soc. 20, 840 (abstract). 

Draganic, I. G., and Draganic, E. D. 1984. Some radiation-chemical aspects of chemistry in 
cometary nuclei. Icarus 60, 464-475. 

Duncan, M., Quinn, T., and Tremaine, S. 1987. The formation and extent of the solar system 
comet cloud. Astron. J. 94, 1330-1338. 

Duncan, M., Quinn, T., and Tremaine, S. 1988. The origin of short-period comets. Astrophys. 
J. 328, L69-L73. 

Eberhardt, P., Dolder, U., Schulte, W., Krankowsky, D., Lammerzahl, P., Hoffman, J. H., 
Hodges, R. R., Berthelier, J. J., and niiano, J. M. 1987. The D/H ratio in water from 
comet Halley. Astron. Astrophys. 187 , 435-437. 

Everhart, E. 1969. Close encounters of comets and planets. Astron. J. 74, 735-750. 

Everhart, E. 1972. The origin of short-period comets. Astrophys. Let. 10, 131-135. 

Fanale, F. P., and Salvail, J. R. 1984. An idealized short-period comet model: surface 
insolation, H2O flux, dust flux, and mantle evolution. Icarus 60, 76-51 1. 

Fernandez, J. A. 1980. On the existence of a comet belt beyond Neptune. Mon. Not. Roy. 
Astron. Soc. 192, 481-491. 

Fernandez, J. A., 1985. The formation and dynamical survival of the comet cloud. In 
Dynamics of Comets: Their Origin and Evolution , eds. A. Carusi and G. B. Valsecchi, 
D. Reidel, Dordrecht, pp. 45-70. 

Fernandez, J. A., and Ip, W.-H. 1981. Dynamical evolution of a cometary sw^arm in the outer 
planetary region. Icarus 47, 470-479. 

Fraundorf, P., Brownlee, D. E., and Walker, R. M. 1982. Laboratory studies of interplanetary 
dust. In Comets , ed. L. L. Wilkening, Univ. Arizona Press, Tucson, pp. 383-409. 

Goldreich, P., and Ward, W. R. 1973. The formation of planetesimals. Astrophys. J. 183, 

Gombosi, T. I., and Houpis, H. L. 1986. The icy-glue model of the cometary nucleus. Nature 
324 . 43-44. 

Green, J. R. 1986. Stress, fracture, and outburst in cometary nuclei. Bull. Amer. Astron. 
Soc. 18, 800 (abstract). 


Greenberg, J. M. 1982. What are comets made of? A model based on interstellar grains. In 
Comets , ed. L. L. Wilkening, Univ. Arizona Press, Tucson, pp. 131-164. 

Greenberg, J. M. 1986. Fluffy comets. In Asteroids, Comets, Meteors II . eds. C.-I. 
Lagerkvist, B. A. Lindblad, H. Lundstedt, and H. Rickman, Uppsala Univ. pp. 221-223. 

Greenberg, J. M. and D'Hendecourt, L. B. 1985. Evolution of ices from interstellar space to 
the solar system. In Ices in the Solar System , eds. J. Kiinger, D. Benest, A. Dollfus, 
and R. Smoluchowski, D. Reidel, Dordrecht, pp. 185-204. 

Greenberg, R., Weidenschilling, S. J., Chapman, C. R., and Davis, D. R. 1984. From icy 
planetesimals to outer planets and comets. Icarus 59, 87-113. 

Herman, G., and Weissman, P. R. 1987. Internal temperatures of cometary nuclei. Icarus 69 . 

HiUs, J. G. 1982. The formation of comets by radiation pressure in the outer protosun. Astron. 
J. 87, 906-910. 

Horanyi, M., Gombosi, T. I., Cravens, T. E., Korosmezey, A., Kecskemety, K., Nagy, A., and 
Szego, K. 1984. The friable sponge model of a cometary nucleus. Astrophys. J. 

Hut, P., and Tremaine, S. 1985. Have interstellar clouds disrupted the Oort comet cloud? 
Astron. J. 90, 1548-1557. 

Irvine, W. M., Schloerb, F. P., Hjalmarson, A., and Herbst, E. 1985. The chemical state of 
dense interstellar clouds: an overview. In Protostars and Planets II . eds. D. C. Black 
and M. S. Matthew^s, Univ. Arizona Press, Tucson, pp. 579-620. 

Jessberger, E. K., and Kissel, J. 1987. Bits and pieces from Halley's comet. In Lunar and 
Planet. Sci. Conf. 18, 466-467 (abstract). 

Johnson, R. E., Cooper, J. F., Lanzerotti, L. J., and Strazzula, G. 1987. Radiation formation of 
a non-volatile comet crust. Astron. Astrophys. 187 , 889-892. 

Kazimirchak-Polonskaya, E. I. 1972. The major planets as powerful transformers of cometary 
orbits. In The Motion. Evolution of Orbits, and Origin of Comets , eds. G. A. 
Chebotarev, E. I. Kazimirchak-Polonskaya, and B. G. Marsden, D. Reidel, Dordrecht, 
pp. 373-397. 

Keller, H. U., Arpigny, C, Barbieri, C, Boimet, R. M., Cazes, S., Coradini, M., Csomovici, C. 
B., Delamere, W. A., Huebner, W. F., Hughes, D. W., Jamar, C, Malaise, D., 
Reitsema, H. J., Schmidt, H. U., Schmidt, W. K. H., Seige, P., Whipple, F. L., and 
Wilhelm, K. 1986. First Halley multicolour camera imaging results from Giotto. 
Nature 321, 320-326. 

Kissel, J., and Krueger, F. R. 1987. The organic component in dust from comet Halley as 
measured by the PUMA mass spectrometer on board Vega 1. Namre 326 . 755-760. 


Kuhrt, E. 1984. Temperatiires profiles and thermal stress on cometary nuclei. Icarus 60 , 

Kuiper, G. P. 1951. On the origin of the solar system. In Astrophysics , ed. J. A. Hynek, 
McGraw Hill, New York, pp. 357-424. 

Lange, M. A., and Ahrens, T. J. 1987. Impact experiments in low-temperature ice. Icarus 69 . 

Lanzerotti, L. J., Brown, W. L., Poate, J. M., and Augustyniak, W. M. 1978. Low energy 
cosmic ray erosion of ice grains in interplanetary and interstellar media. Nature 272, 

Lanzerotti, L. J., Brown, W. L., and Johnson, R. E. 1983. Implications of Voyager data for 
energetic ion erosion of the icy satellites of Saturn. J. Geophys. Res. 68, 8765-8770. 

Lewis, J. S. 1971. Satellites of the outer planets: their physical and chemical nature. Icarus 
15, 174-185. 

Lissauer, J. J. 1987. Timescales for planetary accretion and the structure of the proto- 
planetary disk. Icarus 69, 249-265. 

Marsden, B. G. 1986. Catalogue of Cometary Orbits . 5th Edition, Smithsonian Astrophys. 
Observatory, Cambridge, 102 pp. 

Moore, M. H., Donn, B., Khanna, R., and A'Heam, M. F. 1983. Studies of proton irradiated 
cometary type ice mixtures. Icarus 54, 388-405. 

Mumma, M. J., Blass, W. E., Weaver, H. A., and Larson, H. P. 1988. Measurements of the 
ortho-para ratio and nuclear spin temperature of water vapor in comets Halley and 
Wilson (19861) and implications for their origin and evolution. Bull. Amer. Astron. 
Soc. 20, 826 (abstract). 

Narayan, R. 1987. Supemovae rates in external galaxies. Astrophys. J. 319 . 162-179. 

O'Dell, C. R. 1971. A new model for cometary nuclei. Icarus 19, 137-146. 

Ostro, S. J., Tsou, P., and Stephens, J. B. 1986. Impact cavities in underdense regoliths? 
Meteoritics 21, 476 (abstract). 

Prialnik, D., and Bar-Nun, A. 1987. On the evolution and activity of cometary nuclei. 
Astrophys. J. 313, 893-905. 

Prialnik, D., Bar-Nun, A., and Podolak, M. 1987. Radiogenic heating of comets by 26 a1 and 
implications for their time of formation. Astrophys. J. 319 . 993-1002. 

Prialnik, D., and Bar-Nun, A. 1988. The formation of a permanent dust mantle and its effect 
on cometary activity. Icarus 74, 272-283. 

Safronov, V. S. 1969. Evolution of the Protoplanetary Cloud and Formation of the Earth and 


Planets . Moscow, Nauka Press, (NASA TT-F-667, 1972). 

Sagdeev, R. Z., Szabo, F., Avanesov, G. A., Cruveilier, P., Szabo, L., Szego, K., Abergel, A., 
Balazs, A., Barinov, I. V., Bertaux, J.-L., Blamont, J., Detaille, M., Demarelis, E., 
Dul'nev, G. N., Endroczy, G., Gardos, M., Kanyo, M., Kostenko^ V. I., Krasikov, V. A., 
Nguyen-Trong, T., Nyitrai, Z., Reny, I., Rusznyak, P., Shamis, V. A., Smith, B., 
Sukhanov, K. G., Szabo, F., Szalai, S., Tamopolsky, V. I., Toth, I., Tsukanova, G., 
Valnicek, B. I., Varhalmi, L., Zaiko, Yu. K., Zatsepin, S. I., Ziman, Ya. L., Zsenei, M., 
Zhukov, B. S. 1986. Television observations of comet Halley from Vega spacecraft. 
Nature 321, 262-266. 

Sagdeev, R. Z., Smith, B., Szego, K., Larson, S., Toth, I., Merenyi, E., Avanesov, G. A., 
Krasikov, V. A., Shamis, V. A., Tamapolski, V. I. 1987. The spatial distribution of 
dust jets seen during the Vega 2 flyby. Astron. Astrophys. 187, 835-838. 

Sekanina, Z., and Larson, S. M. 1986. Dust jets in comet Halley observed by Giotto and from 
the ground. Nature 321, 357-361. 

Shoemaker, E. M., and Wolfe, R. F. 1986. Mass extinctions, crater ages, and comet 
showers. In The Galaxy and the Solar System , eds. R. Smoluchowski, J. N. Bahcall, 
and M. S. Matthews, Univ. Arizona Press, Tucson, pp. 338-386. 

Shulman, L. M. 1972. The chemical composition of cometary nuclei. In The Motion. Evolution 
of Orbits, and Origin of Comets , eds. G. A. Chebotarev, E. I. Kazimirchak-Polonskaya, 
and B. G. Marsden, D. Reidel, Dordrecht, pp. 263-270. 

Smith, B. A., Fountain, J. W., and Terrile, R. J. 1988. The beta Pictoris disk. Bull. Amer. 
Astron. Soc. 20, 875 (abstract). 

Stem, S. A. 1986. The effects of mechanical interaction between the interstellar medium 
and comets. Icarus 68, 276-283. 

Stem, S. A. 1988. Collisions in the Oort cloud. Icarus 73, 499-507. 

Stem, S. A. 1989. ISM induced erosion and gas dynamical drag in the Oort cloud. Icarus, 

Stem, S. A., and Shull, J. M. 1988. The thermal evolution of comets in the Oort cloud by 
passing stars and stochastic supemovae. Nature 332, 407-411. 

Tammann, G. A. 1982. Supernova statistics and related problems. In Supemovae: A 
Survey of Current Research , eds. M. J. Rees and R. J. Stoneham, D. Reidel, Dordrecht, 
pp. 371-403. 

Tauber, F., and Kuhrt, E. 1987. Thermal stresses in cometary nuclei. Icarus 69, 83-??. 

Thompson, W. R., Sagan, C, and Khare, B. N. 1987. Coloration and darkening of methane 
clathrate and other ices by charged particle radiation: Applications to the outer solar 
system. J. Geophys. Res. 92, 14933-14947. 


Tsou, P. D., Brownlee, D., and Albee A. 1984. Experiments on intact capture of 
hypervelocity particles. Lunar & Planet. Sci. Conf. 15, 866-867 (abstract). 

Vanysek, V., and Rahe, J. 1978. The ^2c/13c isotope ratio in comets, stars, and interstellar 
matter. Moon & Planets 18, 441-446. 

Wallis, M. K. 1980. Radiogenic melting of primordial comet interiors. Namre 284 . 431-433. 

Weidenschilling, S. J. 1980. Dust to planetesimals: settling and coagulation in the solar 
nebula. Icarus 44, 172-189. 

Weissman, P. R. 1980. Stellar permrbations of the cometary cloud. Nature 288, 242-243. 

Weissman, P. R. 1985a. The origin of comets: implications for planetary formation. In 
Protostars and Planets 11 . eds. D. C. Black and M. S. Matthews, Univ. Arizona Press, 
Tucson, pp. 895-919. 

Weissman, P. R. 1985b. Cometary dynamics. Space Sci. Revw. 41, 299-349. 

Weissman, P. R. 1986. Are cometary nuclei primordial rubble piles? Nature 320 . 242-244. 

Weissman, P. R. 1987. Post-perihelion brightening of Halley's comet: Springtime for Halley. 
Astron. Astrophys. 187, 873-878. 

Whipple, F. L. 1950. A comet model I: The acceleration of comet Encke. Astrophys. J. 111 . 

Whipple, F. L. 1978. Cometary brightness variation and nucleus structure. Moon & Planets 
18, 343-359. 

Whipple, F. L., and Lecar, M. 1976. Comet formation induced by the solar wind. In The 
Smdv of Comets . NASA SP-393, pp. 660-662. 

Wyckoff, S., Lindhokn, E., Wehinger, P. A., Peterson, B. A., Zucconi, J.-M., and Festou, M. C. 
1989. The 12c/13c abundance ratio in comet Halley. Astrophys. J. 339, 488-500. 

Yeomans, D. K., and Kiang, T. 1981. The long term motion of comet Halley. Mon. Not. Roy. 
Astron. Soc. 197, 633-643. 




F. R. Krueger 

Ing. -Buero Dr. Krueger, Messelerstr. 24, D-6100 Darmstadt 12 

Frankfurt, Germany 

A. Korth 
Max-Planck-Institut fuer Aeronomie, D-341 1 Katlenburg-Lindau 3 

Frankfurt, Germany 

J. Kissel 

Max-Planck-Instimt fuer Kemphysik, D-6900 Heidelberg 1 

Frankfurt, Germany 




F. R. Krueger" , A. Korth*, and J. Kissel + 

" Ing. -Buero Dr. Krueger, Messelerstr. 24, D-6100 Darmstadt 12 / FRG 

* Max-Planck-Institut fuer Aeronomie, D-3411 Katlenburg-Lindau 3/ FRG 

+ Max-Planck-Institut fuer Kernphysik, D-6900 Heidelberg 1/ FRG 


During the encounters with comet Halley, PICCA on GIOTTO measured the 
gas phase organic ion composition of the coma, and PUMA on VEGAl 
measured the dust composition. Joining those results a consistent 
picture of the parent organic matter from which dust and gas is 
produced can be obtained. One recognizes a complex unsaturated poly- 
condensate, which splits during coma-formation into the more refrac- 
tory C=C, C-N-containing dust part, and the more volatile C=C, C-0- 
containing gas part. The responsible exothermal chemical reactions, 
triggered by the sun light may play a major role in the dynamics of 
coma formation. The latent heat and reactivity may cause problems 
regarding a sample return mission. 

1. Introduction 

A wealth of organic molecular information has been obtained during the 
encounters of the GIOTTO and VEGA spacecrafts with comet p/Halley by 
means of mass spectrometry. Analyzing the refractory organic dust 
composition PUMA on board VEGAl was the most successful one of the 
three impact mass spectrometers flown (1), so that a gross characteri- 
zation of the organic residue was possible (2). Due to the breakdown of 
an amplifier PIA onboard GIOTTO was not sensitive enough to measure 
molecular ions. However, when now comparing the VEGA solid phase fwith 
the GIOTTO gas phase measurements, in both cases data were used 
refering to distances of several 10,000 km from the nucleus as well. 

The organic gas composition was measured, i. a. , by the PICCA ion mass 
spectrometer on board GIOTTO (3). The latter one succeeded especially 
well in analyzing higher mass gas phase ions up to m/2=100 (4), which 
ion data are now interpreted in detail with respect to the possible and 
compatible parent neutrals in this paper. It is obvious that both 
instruments not only analyze different phases (condensed and gas, 
respectively) but thus also, at least in part, different components of 
the parent cometary organic matter, separated due to different regimes 
of vapor pressure. However, we are able to show that under this respect 
both mass spectrometric results are well compatible with each other 
inferring a common parent matter. As an additional result we will show 
that a major part of the organic matter may have been produced in an 
exothermic polycondensation reaction under the elimination of water, 
thus forming the coma during the passage of the inner solar system. 


2. The PICCA coma ion analysis 

2. 1 Instrumentation 

The Positive Ion Cluster Composition Analyzer (PICCA) on board the 
spacecraft GIOTTO, which is part of the RPA-Copernic plasma instrument, 
was designed to determine the composition and the energy distribution 
of positively charged ions in the direction of the spacecraft motion 
(ram direction) in the coma of comet Halley. PICCA measures E/z 
(energy/charge) instead of the mass (8); however, ions in the inner 
coma are cold, so their motion with respect to the spacecraft is highly 
collimated along the ram direction at the relative flyby velocity of 
68. 4 km/s. 

Most of these ions are apparently singly charged, as there are no 
traces of half-integer lines found in the high-resolution mass regime 
(see below). So the measurements of E/z are, to a very good 
approximation, proportional to the ion mass. 

The mass resolution of the instrument is Am = 0. 4 amu (atomic mass 
units) in the mass range 10 - 50 amu, and flm = 1 amu in the mass range 
51 - 210 amu. Outside the ionopause, which is the outer boundary of 
the magnetic cavity region detected at a cometocentric distance of 
about 4500 km, the thermal energy spread of the incoming ions is larger 
than the instrumental energy resolution. The effective mass resolution 
amounts there to about 3 amu. 

?0P ?cs::ivE 'c\s 

.or. rass ar.£-yze; 


Fig. 2 : 

PICCA ion mass 
spectrum in the well 
resolved mass regime. 


40 45 


2.2 Ion observations in the gas phase' 

As mentioned above the ion resolution outside the ionopause (>4600 km) 
is poor due to the thermal spread of the measured ions. Between 25,000 
km cometocentric distance and the ionopause the spectrum shows indeed 
groups but with clear peaks at 31, 45, 61, 75, 90 and 105 

20, 000 km and the closest approach of the spacecraft 
peaks from 30 - 33 amu are saturated whereas the mass 
only saturated at about 11,000 km close to the 
of all ion species (9). 

broad mass 
amu (4). Between 
(605 km) the mass 
peak at 45 amu is 
intensity maximum 

Inside the ionopause where 
comparible to the instrume 
range between 2 3 amu and 5 
30-33 (saturated), 35, 37, 
instrumental mass resoluti 
50 and 70 amu were determi 
The fits show peaks at 57, 
peaks could not be resolve 
which increased rapidly ne 

the thermal spread of the incoming ions is 
ntal energy resolution we observe in the mass 
amu the following mass peaks (10): 26, 28, 

39, 42, 43, 45, 47, 48. Above 50 amu the 
on decreases and the dominent masses between 
ned by a least-square fitting procedure (4). 

61, and 6 3 amu. Above 70 amu the broad mass 
d due to the presence of a hot ion background 
ar the ionopause. 

The unambiguous identification of the given peaks in the spectra with a 
particular ion is not possible, however we will list the dominant ion 
species that are expected from a C-H-O-N dominated medium. The measur- 
ed peaks show a strong preference for odd mass numbers over even mass 
numbers. The ratio is difficult to derive because of the limited mass 
resolution. However, an estimation in the mass range from 35 - 49 amu 


is possible and amounts to about 0. 3 or less. The odd mass peaks give 
an indication for a nitrogen-poor and non-radical ion mixture: Namely, 
if radical and non-radical species would both contribute significantly, 
no such an overrepresentation of one even-odd character would occur. As 
it is consistent with gas phase measurements of other authors, nitrogen 
is depleted in the gas phase. If it were not, then again a clear 
preference in even-odd character could not be expected, due to the odd 
number of valences of an odd number of N atoms, and an even numer of 
valences of an even number of N atoms in the molecule. As a consequence 
also the PICCA-ions apparently contain little N. However, as they are 
preferably odd mass numbers, they are non-radical species when 
containing no other elements than H, C, O, and S. 

In the following we will try to identify the different m/z peaks and 
will take, as pointed out earlier, only singly charged ions (z=l). We 
start with the less abundant even mass numbers, which may of course 
contain nitrogen, if not being accounted for a minor radical contribu- 
tion which, however, is ruled out, as major radical species frequently 
seen in any electron impact spectrum are missing here. As therefore 
only non-radical ions are taken into account for analysis, most of them 
can at least formally described as protonation products of stable 
neutral molecules. This does in turn not necessarily mean, that proton- 
ation in the gas phase is the dominant mechanism of ion production. 
Another mechanism is proton exchange in the solid state and further 
disintegration of that solid state, which is known (9) to always 
produce such species. As we will see, this mechanism is also accounted 
for in Halley' s coma formation, because further chemical ionization in 
the gas phase is by far not possible due to kinetic reasons. 


26, 28 non-radical ions which belong to the hydrocyanic 

acid such as CN+ and H2CN+ 
42 acetonitrile ion CH3CNH+ 

4 8 unsure, possibly non-radical aminomethanol, 

30, 32 reduction products of hydrocyanic acid like 

methylimine CH3NH+, and methylamine CH3NH3+ ; 

(also sulphur, S, may contribute) 

Mass peaks are included in the broad mass peak 

30-33 amu where the instrument shows saturation. 

The maximum is at 31 amu. 

The following odd mass numbers will not contain any nitrogen 

31,45 These are the highest maxima in the surroundings. 

Formaldehyde, H2C0H+, and acetaldehyde, H3CHC0H+ or 
in principle it could also be protonated alkanes 
like ethane, C2H7+ and propane, C3H9+. However, due 
to its non coordinative binding state these ions 
are very labile and will never show up as the 
largest peak. 

35, 37 water compounds like H20. 0H+ and H20. H30+ 

39 aromatic cyclopropenyl C3H3+ (10) 


43, 45, 47, 57, 61, 

63,75, (89,91) the abundance of these ions indicate that the parents 

are unsaturated hydrocarbons with none or one 
oxygen-heteroatom. If the parents were carbon-homo- 
atomic hydrocarbons than we will not find the 
peaks at m/z 45 and 47 amu. The pattern m/z 61 and 
63 missing 59 amu and 75, 89 and" 91 point to 
hydrocarbons CnHm+ with a comparible low m like 
C5H+, C5H3+, C6H3+, C7H5+, and C7H7+ . 

The measured peaks in the mass spectra indicate that the ions were not 
generated by photoionization from neutral molecules in the gas phase. 
Collisions, however, in the gas phase could produce the observed ions, 
but the probability from gas kinetic reasons are low. Therefore we 
conclude that the formation of these ions could only be from 
decomposition of the dust. 

It should be pointed out already here that the above ions certainly are 
fragment ions produced from larger molecules in the solid state of the 
comet. Namely, speaking about a "molecular ion" in that space analysis 
only means distinguishing against "atomic ions". Moreover, in terms of 
a mass spectrometrist we would even not expect molecular ions, as those 
are generally radical. Protonation would only produce quasi-molecular 
ions (M+H)+. These we can only assign for species from which we know by 
other analyses that they really exist as neutral molecules, too, namely 
water (H30+), Ammonia (NH4+), Hydrocyanic acid (H2CN+) and so on. 

3. The PUMA dust particle analysis 

3. 1 Measurement - method and objectives 

The dust impact mass analyzer PUMA was flown i. a. on the Soviet 
spacecraft VEGAl and obtained a wealth of data, especially also 
molecular ones. The detailed description of the instrument and the 
analysis strategy has been already published (2). Consequently a short 
overview may be sufficient here. 

With the PUMA instrument ions are generated by the impacting dust 
particle on a silver target. Due to the high relative velocity between 
the dusty coma and the spacecraft of v = 78 km/s, any impacting dust 
particle is nearly completely vaporized and ionized, and so is some 
target material, too. From the number of atomic projectile ions (due to 
each element) one can infer (5) the elemental composition of that very 
particle - together with the number of atomic target (Ag) ions one can 
also determine mass and density of that particle. 

However, there is a slight probability in the very first moment when 
the impact shock wave reaches the outermost layers that from those 
molecules and molecular ions are ejected with only minor decomposition. 
It is known that all the ionization mechanisms from solid surfaces in 
such rapid dissipation processes are very much comparable with each 


other (9), i. e. yield similar results. Namely, one generally obtains 
preferably non- radical stable ion-species, under which those carrying 
local charges (due to N-sites, or to a lesser extent, O-sites) are very 
much overrepresented. Molecular lines evidently showed up in the PUMA 
mass spectra, and consequently an analysis was performed, using these 
ion formation rules as a basis. 









Fig. 3 : The PUMA dust particle impact time-of- flight mass spectrometer 


Fig. 4 : A typicaJ converted PUMA mass spectrum showing atomic and molecular ions 




Moles ll 

crcof Ml sub-fr; 

iclions [val] 




Unsaluralcd liydrocarbo 

I'i Unsaturalei) niirogc 

n- Oxygcn-comainmg Sulfur-coni. 


wiilioul hcicroaloiii.s 

conlaining species 


species (unsure) 




























Kalions (val) 













60 40 

20 2 



2Na* (2) 



70Mg=* (140) 



5A1^* (15) 



















0.1 K* (0.1) 




CO?- so^- 

4Ca2* (8) 

OH- POi" 



0.5 Ti^* (1) 



ICr"* (3) 



1 Mn^* (2) 








0.5 Co'* (1) 



3 Ni'* (6) 



0.1 Zn'* (0.2) 

val: 200 40 20 20 2.5 (282.31 

Fig. 5 : The most probable mean conposition of the dust particles as measured by PUMA. 

3. 2 Results 

Ion correlations 

When all the apparent atomic ion signals had been subtracted, those 
residual signals were assigned to mass lines, which time-of-flight were 
consistant with integer mass and which amplitude was above noise level. 
These resultant mass spectra have been compared with a random mass dis- 

Several m/z numbers apparently showed up to be overrepresented in the 
PUMA mass spectra (after subtraction of the atomic ions) when compared 
with random, namely 

m/z = 18, 25, 26, 28, 29, 42, 43, 44, 46, 60, 63, 65, 67, 68, 71, 
78, 79, 81, 87, 89, 118, 122, 

other underrepresented, namely 

m/z 45, 47, 49, 51, 52, 59, 61, 76, 80, 82, 83, 90, 94, 100, 101 

- both significant when comparing with Poisson statistics. 

However, this is not sufficient to determine the class of substances, 
although it gives at least some hints. A much more powerful technique 
is looking for ion-ion-correlations in the mass spectra, i. e. , how 
often appears a certain mass line in one mass spectrum together with 
another certain mass line - and this far more frequently than 
independent chance would suggest. 


The following m/z ion-ion-correlations have been found to be 
statistically strongly overrepresented: 

m/z( 1 ) -m/z (2) = 26-28, 26-70, 28-62, 28-64, 28-78, 28-89, 42-66, 

42-67, 42-71, 42-77, 42-78, 48-68, 48-89, 64-81, 

66-81, 68-71, 68-97, 71-77, 71-78, 78-88, 78-97, 

78-98, 85-91, 89-91, 89-98, 91-93, 95-97. 

From this list some key mass numbers are recognized to be very often 
members of such correlations, and thus are to be considered as key 
decomposition (fragment) ions of the impacting (certainly larger) 
molecules, namely: 

m/z 28, 42, 71, 78, 91, and, to a minor extent, m/z 48, 77, 89. 

It was the task to reveal the information, certainly accumulated in 
those mass numbers, about the parent molecule class yielding these 
ionic decomposition products. 

3. 3 Refractory organic molecule species 

Yielding the ion formation processes valid as known from other rapid 
dissipation processes near solid surfaces, a first glance especially to 
the above key ions show that saturated hydrocarbons without heteroatoms 
cannot be responsible for this ion decomposition pattern. Namely, such 
ions would prefer odd mass numbers, which is not seen. Also unsaturated 
hydrocarbons without heteroatoms cannot account for, because they also 
strongly prefer odd mass numbers; but they could be distinguished from 
saturated ones by the exact m/z values: saturated prefer CnH2n+m ions 
with m= + 3,+l,-l, whereas unsaturated prefer those with m=- 1, -3, -5, . . 

The importance of m/z 28 and 42 in any reaction chains is always 
recognized if nitrogen plays an important role in the parent molecules 
being decomposed down to the nitrile ions (m/z 30, 44, and 58, less 
important with the PUMA spectra, would show a decomposition only down 
to the imine ions). Whereas amines would yield the latter ones, 
iminies, enamines or nitriles would yield the, observed, former ones. 
Consequently, many of the molecular ions seen can be accounted to 
unsaturated N-containing hydrocarbons. 

0-containing hydrocarbons generally yield key ions with m/z 31, 45, 59, 
which were not found; moreover, they occured even statistically 
underrepresented (31 is however totally accounted as P atomic ion 
within the cosmic abundance). 

If one thus assumes the correlational ions mostly as unsaturated ions 
containing one or (with larger m/z) sometimes two N heteroatoms, nearly 
all. correlations are easily explained as members of reaction chains 
with, at least formally, hydro-dehydro and HCN or C2H2 addition-elimin- 
ation exchange reactions. This formal result is at least consistent 
with a general chemical matter being built up from acetylene and hydro- 
cyanic monomers ionized during impact by proton transfer reactions. 


However, we must strongly point out here again, that N-containing ion 
species are much more probable to be seen in such mass spectra than 
ions from hydrocarbons without heteroatoms. According to the low N/C 
atomic ratio we found in the dust (4%), we conclude that there must 
have been a lot of precursor fragments not containing N and thus not 
showing up as ions. 

So, we cannot recognize the refractory organic as a hydrocyanic (HCN) 
polymer or a acetonitrile (CH3CN) polymer, though both being unsatura- 
ted. The low N/C ratio forbids this. However, it is completely within 
the results of the above analysis to say, that the refractory organic 
consists of an unsaturated hydrocarbon polymer with (not too much) 
nitrogen heteroatoms. In contrary it cannot consist of an aldehyde or 
keto-polymer, whether saturated or not, as O seems not be a major part 
of the molecular ions, and other analysis shows (2), that the oxygen in 
the organic part may mainly be due to water ice or some acids, at least 
in the large distances of VEGAl from the nucleus in the outer coma 
(GIOTTO measurements give hints of an elimination of CO and other 
0-containing species in the inner coma nearer than 10, 000 km from the 
nucleus ) . 

Before we combine the results of gas phase and solid phase 
measurements, respectively, let us first make some remarks about 
polymer vapor pressures. 

4. Polymers and their vapor pressures 

It takes some hours time for the dust and gas after ejection to reach a 
several thousand km distance from the nucleus. It is obvious that the 
vapor pressure of the organic material determines whether it will find 
itself in the coma gas phase or in the dust particles. Materials of 
medium vapor pressure will give rise to an extended source of gas, 
because its vapor pressure may be too low to be ejected directly into 
the gas phase, and too high to survive in the condensed phase of the 

Let us now discuss more quantitatively the kinetics of vaporization of 
polymers as found with PI CCA and PUMA. For simplicity we can regard the 
isothermal (T) behaviour of the vapor pressure as follows: The 
probability P per unit time that a molecule leaves the surface, where 
it is bound to, is proportional to the product of the conditional 
probabilities w^- that any structural element of the polymer possess in 
that very time interval a vibrational energy higher than a certain 
limit value dependent on the nature of that element, (i: structural 
element number). For large molecules the w are independent from each 
other and Boltzmann' s theory yields 

w^- = c^ exp(-E£-/kT) 

with Ej- the limit vibrational energy. The total probability P is then 
given as 

C TT exp(-E./kT) or log P = log C + J* (-E. 



Due to kinetic gas theory that probability P is directly proportional 
to the vapor pressure p via the Hertz -Knudsen-equation we will need in 
a moment, so that with constant temperature T we can simply rewrite as 

- log p = C + ^ a^. 

with a; = E^/kT and C =-4 if p is measured in atm (log is natural 
logarithm). It is now the task to determine at least roughly the mean 
contributions a of any structural element. This is done by comparing 
the vapor pressures at T=300 K (which is a reasonable dust temperature) 
of two molecules differing in that one structural element (7). (With 
other temperatures slight changes are due to a recalculation using 
Clausius-Clapeyron' s equation for dp/dT. ) 

For each additional CH2-group in an alkane chain it is approximately 

a=l; an exchange of an -ane into an -ene or an -yne group (introduction 

of a double or triple bond into the chain) an additional a=o. 5 has to 

be applied. 

Nitrogen bearing species are very potent in reducing the vapor 

pressure. An additional N-atom bound as nitrile yields approximately 

a = 5, as amine or imine a = 4. 

Oxygen bearing species are less potent in reducing the vapor pressure; 

only O bound as an -ole (alcohol) or -ale (aldehyde) group yields. 

approximately a=4, as keto-groups a=3, and as ethers only a=l. 5 

Introducing now explicitly the Hertz-Knudsen-equation we have a 
relation between the number of desorbing molecules per unit time and 
unit surface d Z/(dt dO), the vapor pressure at temperature T expressed 
in numbers n of molecules per unit volume, and the mass m of the 


r = d Z/(dt dO) = n v/kT/(2Trm) 


If that rate density r is about 10 /cm /s, a lot of the material of a 
1 yun diameter dust particle is evaporated in the above discussed coma 
development times. This corresponds to vapor pressures of several 10 " 
atm. This means that materials with vapor pressures lower than 10 atm 
(jra>27) will preferably be found in the condensed (dust) phase, whereas 
those higher than 10~^atm (^a^ 25) will preferably be found in the gas 

For instance, a pure CH20-polymer (polyoxymethylene) must possess at 
least 11 chain elements (molecular mass M=330 amu) to survive in the 
dust particles. Such large pure C/0 molecules are, however, very 
unlikely due to elemental abundances. Smaller ones are possible in the 
gas phase, and are found, although only to a minor extent when 
comparing with pure CH-polymers, in the PICCA mass spectra (in contrary 
to an earlier finding of other authors, 13) 

In contrary, a pure HCN-polymer (polymethylenimine) must possess at 
least only 5 chain elements to survive in the dust, and there are a lot 
of indications for decomposition products of such species in the PUMA 
ion-correlation mass spectra found. Within the gas phase species 
containing N-atoms cannot be very large if abundant. Consequently 
besides the abundant HCN (hydrocyanic acid) and aceto- nitrile not very 
many N-bearing species are found. 


5. The organic parent substance of gas and dust 

5. 1 Inferred polymeric parent species 

The above results together with the measured elemeiit ratios in the 
comet (which are actually solar, 5) now allow a better characterization 
of the cometary matter before (! ) coma formation. A remarkable point is 
the low 0/C-ratio in the molecular ions found (about 0. 6 in the gas, 
probably below 0.3 in the dust), although the total 0/C-ratio is above 
1. , also in the organic component alone. Apparently most of the oxygen 
is bound in water molecules and not in polymeric molecules. From the 
above measurements, however, one can not deduce, whether the huge 
amount of water seen in the coma is already chemically present as water 
in the comet nucleus, or whether it has been produced by any reaction 
out of the pristine matter. 

Nevertheless, at least shortly after evaporation of the comet's nucleus 
we observe a fairly unsaturated polymeric organic matter with about 35% 
C-atoms, 18% 0-atoms, 2% N-atoms, and 45% H-atoms (the latter deduced 
from the mean degree of saturation). The mean chain length (or ring 
magnitude) may be about 10 nuclear (CNO) atoms, so that molecular 
weights of 100 - 200 amu may be typical. This can be concluded from the 
organic matter dust to gas ratio, together with the vapor pressures, 
and the mean travel time from the nucleus to the loci of detection. 

Together with this polymeric material there is a lot of small molecular 

matter containing much oxygen, especially water, HOH, but possibly also 

formic acid, HCOOH and/or other oxidized species forming CO under 
decomposition processes. 

5. 2 Possible polycondensate formation mechanism 

If one looks to a fairly unsaturated polymer accompanied by water, one 
may think that it was polycondensation rather than polymerization how 
such a species may have been formed. It is an open question up to now 
(11), whether these organics have been produced by irradiation pro- 
cesses of frozen gases in the inner solar system, or whether the comet 
consists of organics already in the Oort' s cloud. If the latter is the 
case, water (and to a lesser extent CO) may be driven out at higher 
temperatures by polycondensation mechanisms, especially if latent heat 
is present in the organic matter (e.g. due to free radicals). Anyway, 
the ultimate result of such reactions in laboratory simulation are 
apparently unsaturated products, and water is driven out. 


6. Consequences and recommendations for ROSETTA or any CNSR 

There are a lot of precautions to be taken with any comet nucleus 

sample return (CNSR) mission, e.g. ROSETTA, which results from the 

chemical properties of that matter. 

First of all, at least the organic fraction (if not both) are not yet 

in an equilibrated state, so that temperature and pressure rise is 

expected to trigger chemical reactions. By no means a liquid state 

should be achieved neither during transport nor during subsequent 

analysis, because thence a series of chemical reactions would take 

place completely altering the primordial matter. 

Another problem may cause the material of the containers, as catalytic 

action with the cometary matter has to be strictly avoided. 

Sealing and flanging may not be trivial as well, because there is a 

wide range of size distributions of the matter. 

It seems that there may be a rather hard crust like a compact polymeric 
plastics with minerals embedded, covering the comet' s surface, which 
thus may be hard to overcome. In contrast to it, the primordial matter 
below will be very soft and fluffy. As neither the thickness nor the 
elastic and inelastic modules of that crust is known, such a layered 
structure may cause severe problems when penetrating it: Namely, if the 
force acting is too small the penetrator may not penetrate at all, if 
it is too large the penetrator may vanish in the comet' s interior. 

7. Conclusion 

It has been clearly shown that the molecular gas phase ions are not 
arising from small parent molecules, but from larger polymers in the 
cometary matter the more volatile of which forming the gas phase, the 
less volatile the dust. 

The volatile component contains an appreciable amount of oxygen. If 
nitrogen is present only small molecular ions are formed in the gas 
phase. The volatile matter is apparently not totally saturated, but 
bears generally some double or triple bonds. 

The involatile component apparently contains less oxygen in the polymer 
species rather than in smaller molecules. It consists mainly of 
unsaturated hydrocarbons with some nitrogen present. 

Both components may be highly reactive under normal conditions, i. e. , 
in a liquid water environment, under which it can form a lot of 
prebiotic chemical species (2, 11) with large free energy, especially 
with the help of the catalytic action of the large specific area doped 
sil'icatic backbone of the dust grains. Consequently, a series of 
special precautions have to be taken with a comet nucleus sample 
















Kissel, J., Sagdeev, R. Z. & al. , Nature 321, 280 (1986) 

Kissel, J. & Krueger, F. R. , Nature 326, 755 (1987) 

Korth, A. , Richter, A. K. , & al. , Nature 321, 335 (1986) 

Mitchell, D. L. et al., to appear in Adv. Space Res., 1988 

Jessberger, E. K. , Kissel, J. , Fechtig, H. & Krueger, F. R. , In: Comet 

Nucleus Sample Return Workshop Proc. ESA SP-249, 27 (1986) 

Krueger, F. R. , Z. f. Naturforschung 38a, 385 (1983) 

CRC-Handbook of Chemistry and Physics, 56th Ed. (1976) 

Korth, A. et al. , J. Phys. E. Sci. Instr. ,20, 787, 1987 

Korth, A. et al. , Eur. Space Ag. , SP-250, 199, 1986 

Korth, A. et al. , to appear in Nature, 1989 

Ferris, J. P., Origins of Life and Evolution of the Biosphere 18, 

161 (1988) 

Strazzulla, G. , In: Laboratory Simulation of Organic Cometary 

Material, Catania (1987) 

Huebner, W. F. , Boice,D. C. & Sharp, C. M. , The Astrophysical J. 

320 (1987), L149; Huebner, W. F. , Science 237 (1987), 628 




G. F. Herzog 
Rutgers University 

P. A. J. Englert 
San Jose State University 

R. C. Reedy 
Los Alamos National Laboratory 

K. Nishiizumi, C. P. Kohl, and J. R. Arnold 
University of California, San Diego 



G. F. Herzog 

Rutgers University 

P. A. J. Englert 
San Jose State University 

R. C. Reedy 
Los Alamos National Laboratory 

K. Nishiizumi, C. P. Kohl, and J. R. Arnold 
University of California, San Diego 


Determinations of the cosmogenic nuclide concentrations in cometary material will help to define 
the recent surface history of the comet and its exposure to cosmic rays. In particular, the rates for the 
removal or mixing of surface material could be studied, and any variations in cosmic-ray intensity 
implied by the data could be used to infer orbital changes during the last few million years. The 
measurement of the shorter-lived isotopes poses technical challenges that should be addressed now. The 
measurement of longer-lived isotopes will be straightforward provided that rates of mass loss are not too 


An important goal of solar system exploration is to obtain for laboratory study cometary material 
that has been subjected to a minimum of post-accretional processing. The proposed Rosetta mission 
(Ahrens et al., 1987), for example, would achieve this goal by recovering core, volatile, and non-volatile 
surface samples. Close to the comet's surface, solar and cosmic-ray irradiation will have induced nuclear 
reactions that produce a wide variety of cosmogenic nuclides, both radioactive and stable. We consider 
here some strategies for the measurement and interpretation of the cosmogenic nuclide record in 
returned cometary material. Our main purpose is to point out how cosmogenic radionuclide data may 
help in reconstructing the recent history of the cometary surface. Few other approaches would provide 
comparable information. 

In the sections to follow we 1) list some isotopes of interest and the ways in which they are 
produced; 2) review current analytical capabilities for these isotopes; 3) describe some semi-empirical 
calculations of production rates in a quiet comet, calculations that may serve as a convenient framework 
for discussion; 4) consider some processes that would perturb the comet's surface and how they would 
affect the cosmogenic nuclide record; 5) point out the kinds of samples and measurements we think best 
suited for the reconstruction of the irradiation history and 6) identify those gaps in present, relevant 
knowledge that pre-launch research should fill. 



The cosmogenic radionuclides of particular interest for unravelling the histories of cometary 
surfaces are listed in Table 1 along with half-lives, major target elements, current detection limits and 
methods of measurement. As shown in Table 1, these nuclides may be produced by nuclear interactions 
with galactic cosmic rays (GCR) or solar cosmic rays (SCR) or by the capture of thermal neutrons 
(Reedy and Arnold, 1972; Reedy et al., 1983). Because of the variety of the cosmic-ray particles and of 
their modes of interactions, the effective depths of the interactions and their products vary considerably. 
This diversity in the types of products and their depths will be very useful in studying the cosmic-ray 
record of the comet (cf.. Reedy et al., 1983). In particular the concentrations of SCR products furnish a 
most sensitive measure of surface disturbance (see below). An examination of the target elements listed 
in Table 1 shows further that most of the nuclides are produced primarily in dust particles: i.e., silicate 
or possibly metal grains. Production of ^^C and i^Be may also occur in icy and/or organic material. 


Half-life Target Detection E>etection Principle 

Elements Limit Method Means of 

(10^ atom) Production* 

7Be 53.3 d C, O, Mg, Si 0.05* Counting GCR 

66Co 77.7 d Fe 0.01* Counting SCR 

MMn 312 d Fe 0.05* Counting GCR+SCR 

22Na 2,61 y Mg, Si 1.* Counting SCR+GCR 


5.27 y 






12.3 y 

0, Mg, Si, Fe 


Mass spec. 



5730 y 

O, N 





0.10 My 

Ca, Fe 





0.30 My 

CI, K, Ca, Fe 





0.71 My 

Al, Si 





1.5 My 

C, O, Mg, Si 





3.7 My 

Fe. Ni 




* These numbers are based on laboratory measurements; in situ detection 
limits may be an order of magnitude higher. # GCR=Galactic cosmic rays, 
SCR=Solar cosmic rays, Nth=thermal neutron capture. 


The isotopes listed in Table 1 have a wide range of half-lives. This range is advantageous because 
it opens observational windows on a correspondingly broad range of rates for cometary processes. From 
a technical standpoint, it implies a need to employ a number of different experimental methods. In the 
text below, we will take 5 y as the half-life that operationally separates long-lived species which, as we 
will see, are readily amenable to terrestrial measurement, from short-lived species. 

Short-lived isotopes - In practice it will be difficult to measure the nuclides with shorter half-lives for 
several reasons. A major difficulty is the long transportation time (-3 years) of the sample return. The 
original radioactivity produced in the comet will decay appreciably during transit from comet to 
laboratory, while at the same time measurable amounts of the nuclides will be produced by cosmic ray 
bombardment of the sample in the space probe. Two possible alternatives to laboratory measurement are 
discussed in section 5. 


The precision attainable for the short-lived nuclides would not match that for the longer-lived 
species. The detection limits for the short-lived nuclides listed in Table 1 are based on laboratory 
measurements after chemical separation. The detection limits for in-situ measurements or nondestructive 
counting in the space probe may be more than an order of magnitude higher. Nonetheless, the 
importance of obtaining the activities of these short-lived nuclides warrants further investigation of the 
feasibility of the two approaches described in section 5. 

Long-lived isotopes - Determinations of longer lived radionuclides should present few technological 
challenges either in sample preparation or measurement. The detection limits for nuclides measured by 
accelerator mass spectrometry (AMS; see Table 1) are typically 10^ atoms, 2-4 orders of magnitude lower 
than achievable by decay counting (Elmore and Phillips, 1987). The detection limit of ss^n by neutron 
activation is about 10* atoms (Nishiizumi, 1983) but AMS may lower this value to 10® atoms within a 
few years. The nuclides ^°Be, ^^A\, and S3\in have been measured in 100 /ig size samples of deep sea 
spherules and in mg size samples of lunar rocklets by AMS and by neutron activation (Raisbeck et al., 
1985; Nishiizumi, 1983). The size of the cometary sample required for our study will depend on the rate 
of surface mass loss and on the depth of the material below the cometary surface. If the rate of mass 
loss is less than 10"' g/cm^-y, all of the long-lived nuclides proposed could be measured with an error 
of less than 10% in a sample of 10-50 mg. The chemical procedures needed to separate these nuclides 
from silicate and ice are well established and have been widely applied. Means for the separation of 
^°Be from CO2 and organic material will have to be tested before actual sample preparation. A second 
developmental question concerns •*iCa. Fink et al. (1988) have reported a minimum detectable •*iCa/'*0Ca 
ratio of 5 X 10"^® (see also Kubik et al., 1986). This nuclide is expected to be measurable at the level 
given in Table 1 within a year. Small amounts of ^H can be detected by mass spectrometric 
measurements but only if ^He is allowed to accumulate for suitable times in the sample and if most of 
the ^He initially present can be removed (Clarke et al., 1976). 


Nature of the interaction - Cosmogenic nuclides are normally not present in appreciable quantities in 
material buried beneath many meters of overburden. When a piece of comet is exposed to cosmic-ray 
particles, concentrations of cosmogenic nuclides steadily increase until, after several half-lives, 
radionuclide activities approach an equilibrium value where the rate of decay equals the production rate. 
In contrast, stable isotopes such as ^He and 2iNe and tracks made by heavy (Z > 20) cosmic-ray nuclei 
may continue to accumulate indefinitely and can be used to determine the total length of time that a 
sample was exposed to cosmic rays. 

Two types of energetic particles can induce nuclear reactions in matter, the galactic cosmic rays 
(GCR), which have low fluxes but high (E - GeV) energies, and the solar cosmic rays (SCR) (often 
called solar energetic particles), which are emitted irregularly from the Sun with low (E - 10-100 MeV) 
energies but high fluxes (cf.. Reedy and Arnold, 1972; Reedy et al., 1983). The nuclei in both types of 
cosmic rays are mainly protons and a particles (with a proton/a- particle ratio of about 10-20), with 
about 1% heavier nuclei. 

The energy, charge, and mass of a cosmic-ray particle and the composition of the target material 
determine which interaction processes are important and which cosmogenic products are formed. 
Energetic nuclear particles interact with matter mainly in two ways: ionization energy losses and 
reactions with nuclei. All charged particles continuously lose energy by ionizing the matter through 
which they pass. A nuclear reaction between an incident particle and a target nucleus generally involves 
the formation of new, secondary particles (such as neutrons, protons, pions, and photons) and of a 
residual nucleus that is usually different from the initial one. One-GeV protons have a range of about 
400 g/cm2 and a nuclear-reaction mean free path of ~100 g/cm^, so only a few percent go their entire 
range without inducing a nuclear reaction. 

SCR Products - The depth profiles of cosmogenic nuclides made by SCR particles are very different 
from those made by the galactic cosmic rays. The relatively low-energy solar protons and q particles are 
usually stopped by ionization energy losses near (top few millimeters) the surface. The SCR particles that 
induce nuclear reactions as a rule produce few secondary particles and the masses of the reaction 


products are close to those of the target nuclei: e.g., ^^o from iron (Reedy and Arnold, 1972). The 
fluxes of SCR particles as a function of depth can be calculated accurately from ionization-energy-loss 
relations, so a nuclide's production rates can be predicted well if the cross sections for its formation are 
known. Like the densities of heavy-nuclei tracks, the activities of SCR-produced nuclides decrease 
rapidly with depth, most being made within a centimeter of the surface (Reedy and Arnold, 1972; Reedy 
et al., 1983). 

GCR Products - The GCR particles that produce nuclear reactions can roughly be divided into four 
components: high-energy (E > 1 GeV) primary particles, medium-energy (about 0.1 to 1 GeV) particles 
produced partially from the first component, a low-energy group (E < 100 MeV) consisting mainly of 
energetic secondary neutrons, and slow neutrons with energies below about a keV. The fluxes of the 
high-energy primary GCR particles decrease exponentially with depth as they are removed by nuclear 
reactions. The numbers of secondary particles as a function of depth build up near the surface, where 
most of them are made, but eventually decrease roughly exponentially. In the moon, the rate of neutron 
production is about 13/s-cm2 (Lingenfelter et al., 1972). This neutron production rate may be compared 
to an average omnidirectional flux of about S/s-cm' for the GCR primaries in space (Reedy et al., 
1983). The difference between the two fluxes shows the importance of the large cascade of nuclear 
reactions that produces numerous secondary particles. 

Neutron products - Neutrons slowed to energies of keV or eV can produce nuclides by neutron-capture 
reactions. Some reactions of interest are s®Co(n,7)60Co, i<N(n,p)i<C, and 68j<[i(n,'y)59Ni and others 
leading to stable isotopes such as i^osm and isscd (Lingenfelter et al., 1972). Most GCR-induced 
reactions involve incident particles with energies of 1 MeV or more, emit one or more nucleons either 
individually or in clusters like a particles, and are called spallation reactions. Such products are referred 
to as "spallogenic" nuclides. Sometimes the spallation reactions are divided into low-energy and high- 
energy groups. High-energy spallation reactions involve particles with energies above about 100 MeV, 
produce numerous secondaries, and can make, in relatively low yields, many different product nuclides. 
Examples of high-energy spallation reactions are ^®0(p,X)3He or 2'«Mg(p,X)iOBe, where X can be any 
one of a large number of possible outgoing particle combinations. Low-energy spallation reactions 
usually involve particles with energies below 100 MeV and can produce certain nuclides in high yields 
because both the fluxes of particles and the cross sections are relatively large. The reaction 
2'*Mg(n,Q)2iNe is such a low-energy reaction and is the major source of ^iNe in most stony objects. 
Because the distributions of low-energy and high-energy GCR particles are not the same for all 
locations, the production rate-versus-depth profiles for different cosmogenic nuclides can vary. A high- 
energy product like 'He has a depth-versus-concentration profile that is fairly flat near the surface, 
whereas a low-energy product, like ^iNe, builds up in concentration considerably with increasing depth 
near the surface (Reedy and Arnold, 1972; Reedy et al., 1983). 

Model calculations - Reedy (1987a) reviews several approaches that have been used to predict the rates, 
ratios, or profiles for the production of cosmogenic nuclides. Simulations using thick targets bombarded 
by a beam of high-energy protons have been used to model the production rates and profiles of several 
cosmogenic nuclides (Englert et al., 1987). Several research groups have developed semi-empirical 
models from measured concentrations of cosmogenic nuclides. Several models for the production rates of 
cosmogenic nuclides are based on particle fluxes estimated inside the object and cross sections for the 
nuclear reactions of interest. The most important parts of these models are the expressions adopted for 
the fluxes of cosmic-ray particles as a function of particle energy and sample depth. For SCR particles, 
ionization-energy-loss relations can be used to calculate the particle fluxes (Reedy and Arnold, 1972). 
For GCR particles in the moon. Reedy and Arnold (1972) derived particle spectra that varied with 
depth. Cross sections used for the production of a nuclide are measured ones, if available; if not they 
can be estimated from nuclear models or other systematics, such as spallation formulae (Reedy, 1987a). 

In a few cases, production rates and profiles have been calculated from theoretical expressions for 
nuclear interactions. Armstrong and Alsmiller (1971) used a Monte Carlo code for intranuclear cascades 
to calculate the distribution of cosmic-ray particles and product nuclides in the moon. Drake et al. 
(1988) have used a similar code to calculate particle fluxes in Mars. Rates for neutron-capture reactions 


have been calculated for the moon, meteorites, and Mars using several neutron-transport codes. Given 
an initial distribution of secondary neutrons, which are made with energies of the order of 1 MeV, these 
calculations transport the neutrons through the object and follow the scattering reactions that slow the 
neutrons to thermal energies (E < 1 eV) (Lingenfelter et al., 1972; Drake et al., 1988). The production 
rates for neutron-capture reactions vary much more with sample depth and peak more deeply than do 
those for spallation reactions (Reedy et al., 1983). This large variability with radius and depth makes 
neutron-capture products, like SCR products, useful in determining sample histories. However, the 
composition of the medium, especially its hydrogen and carbon content, can strongly affect neutron 
transport (Drake et al., 1988). 

Comparison of models with experiments - The absolute rates calculated with the models are sometimes 
not very reliable and are generally normalized to measured ones. When, as is often the case, the activity 
of a radionuclide in an extraterrestrial sample is in equilibrium with its production rate, measured 
activities can directly be used as production rates, assuming no complications in the sample's history. A 
measured activity of a cosmogenic radionuclide and a production ratio can also be used to infer the 
production rate for a stable nuclide, especially if the pair of nuclides are made by similar reactions or if 
the stable nuclide is the decay product of the radionuclide. Radioactive/stable pairs used in this way 
include ^H/^He, 22Na/22Ne (with care to correct for variations in 22isja activity over an 11 -year solar 
cycle), 36Cl/S6Ar, a^Ar/ssAr, and 8iKr/»3Kr. 

To a good first approximation, the cosmogenic nuclide depth profiles in the cometary surface should be 
analogous to those observed in the lunar surface because the irradiation geometry is Ir in both cases. 
Fig 1 shows i°Be, 26a1, 36q^ and ssMn depth profiles measured for the Apollo 15 drill core (Nishiizumi 
et al., 1984a,b) along with model calculations (Reedy and Arnold, 1972) for both SCR and GCR 
production of these nuclides. 

200 300 

Depth ( 3/em^ ) 

200 300 

Oeptn (g/cm*) 





10 - 

■ 1 ■ ■ ■ T r- 'J ■ 1 


1 ' r 

IS Drill Core 


Reedy- Arnold 
GCR profile 



> : 

Normalizefl to average 
chemical composition 


I 1 


200 300 

Deoth (S/c™^ ) 

9 2° 

^*Cl APOLLO IS Drill 



■D 10 


^ — ,^_^ Reedy -Arnold 
y^ • •^~~-»^^^ theoretical 

^~^^ GCR profile : 

^ 5 




Normalized to average 


-^ 2 

chemical composition 







200 300 

Deoth (g/cm^) 


Figure 1. Measured and calculated cosmogenic nuclide profiles in the lunar surface. For references, see 

text. .„„ 



We can expect three sources of perturbations to the baseline production rates for cosmogenic 
radionuclides: 1) mass loss, which may vary from phase to phase; 2) mixing as a result of cratering, 
gas release, or other processes; 3) variation in the cosmic ray flux that follows either orbital evolution or 
compositional change. Adjustments to model calculations can be made to allow for these effects. Such 
adjustments have worked well in the lunar case. 

Mass Loss - The loss of matter from the surface of the comet will reduce the radionuclide inventory 
from its baseline value. The key parameter governing the size of the reduction is the ratio er/p where e. 
is a nominal erosion rate, r the half-life of an isotope and p the thickness of matter that will reduce the 
relevant production rate by a factor of two. Put another way, significant radioisotope losses will occur 
whenever the thickness of matter eroded during one half-life is comparable to the half- thickness. With 
the aid of the cosmogenic nuclide measurements one can in principle determine the sizes of the 
reductions and from them estimate the rates of mass loss. Is this goal achievable in practice? 

The answer to this question will depend on the sampling strategy adopted. Weissman (1989) has 
proposed blasting surface material away prior to coring. Such an approach would almost surely 
obliterate the cosmic-ray record. A second strategy for retrieving deeper-lying material is to select as 
active a region as possible (Y. Langevin, pvt. comm.) on the theory that the least altered and most 
volatile-rich areas will have the greatest activity. While it is difficult to predict what the rate of mass 
loss will be, we do have some limiting cases to consider. 

In meteorites, micrometeorite milling proceeds at rates corresponding to average values of e of 
about 1 mm/My (Schaeffer et al., 1981). Lunar rates may be somewhat higher but in either case only 
isotopes with the longest half-lives and/or smallest half- thicknesses (e.g., SCR-produced ^SNi, 26aI, or 
"Mn in lunar surface rocks) show signs of much loss (Lanzerotti et al., 1973; Fruchter et al., 1981; Kohl 
et al., 1978). For a comet with a radius of 10 km, the global rates of mass loss given by, e.g.. Glass 
(1982) and Ahrens et al. (1987), translate into average erosion rates between 0.001 and 2 g/cm^-y. 




EROSION RATE (g/cm2-y) 

0.001 1 

Depth (g/cm2) | 

200 200 1 

Activity (atom/g) or (dpm/kg) | 

22Na 2.61 y 
"C 5730 y 
S6C1 0.30 My 
26AI 0.71 My 
">Be 1.5 My 
"Mn 3.7 My 
"Mn (dpm/kg Fe) 

4.8x10* 24 15 
e.lxW 14 11 
3.4x10^ 1.5 0.8 
8.6x10^ 16 12 
1.9x10^0 17 13 
3.4x10^° 12 10 
4.5x10^^ 161 136 

24 15 
13 10 
0.4 0.2 
2.4 1.7 
1.3 1.0 
0.4 0.3 
5.1 4.4 

To assess the impact of 
erosion in a comet nucleus, we 
have calculated cosmogenic 
nuclide concentrations at the 
surface and at a depth of 200 
g/cm2 (Table 2). The 
calculations are based on the 
computational model of Reedy 
and Arnold (1972) and the 
chemical composition reported 
for comet P/Halley by 
Jessberger et al. (1988). Two 
cases are presented here, the 
first with no mass loss and the 
second with a mass loss of lO"^ 
g/cm2 year. The results show 
that the short-lived species are 
relatively insensitive to erosion, 
while long-lived species 
produced by cosmic rays are 
almost completely removed. 


Even though the effects of solar cosmic rays may not be detected in the long-lived nuclides, it is of 
critical importance that the depths below cometary surface be known as exactly as possible for all 

Unfortunately from our point of view, imaging of comet P/Halley and other evidence indicate that 
only a small fraction - perhaps 10% - of the comet's surface is active at any time. This observation 
implies local erosion rates much higher (-20 g/cm^-y) than the average considered above. With erosion 
rates so rapid, only the shortest-lived isotopes hold any promise of providing useful information. All 
SCR products, for which production is confined to the uppermost centimeter or so, and the bulk of the 
longer-lived isotopes will have vanished within a few apparitions. 

There may be some mitigating factors. On the centimeter scale relevant to the behavior of a 
sampling site we do not know, for example, whether a specific location becomes active and remains so; 
or whether - as seems unlikely - activity occurs uniformly throughout. Moreover, depending on the 
comet selected, we may know little a priori about the total duration of the activity. Finally, and perhaps 
most important, the nature and durability of the crust widely believed to cover the comet's surface is 
uncertain. The crust may retain for long periods constituents tough enough to resist volatilization but 
porous enough to permit the escape of gases (Brin and Mendis, 1979; Horanyi et al., 1984). Simulation 
experiments and direct observation suggest that the crust includes silicates, polymeric organic material, 
and water ice (Sagdeev et al., 1986; Klinger et al., 1988; Wdowiak et al., 1989). Among these phases, the 
silicates certainly and the others possibly would retain cosmogenic nuclides. 

Mixing Processes - Stern (1988) argues that comets undergo regolith formation, albeit at slow rates; 
Keller et al. (1986) present possible evidence for cratering on comet P/Halley. Cratering may lead to 
both vertical and horizontal mixing of material and thereby scramble the record of cosmogenic nuclides. 
Studies of the lunar regolith demonstrate that cosmogenic nuclide data can unscramble the record. 
Nishiizumi et al. (1983), for example, have shown that certain depth profiles on the lunar surface differ 
from profiles calculated for undisturbed material (e.g., Nishiizumi et al., 1983) but can be reconciled 
with them by allowing for the effect of meteoritic mixing with the model of Langevin at al., (1982). 
The effects extend to about 50 g/cm^ over a 1-10 My time scale. 

The devolatilization of comets may also lead to mixing of surface material. Dust lifted from one 
portion of the surface - perhaps by electrostatic processes (Flammer et al., 1986) - could return to 
another. Material may slide into depressions created by evaporation. Gas may recondense as the comet 
moves away from perihelion (Whipple, 1987). The explosive release of gas pockets in the interior (Bar- 
Nun, 1989) could lead to the extensive relocation of material. 

Variations of the cosmic-ray flux - The galactic cosmic-ray flux has a positive radial gradient of about 
2%/AU with little dependence on latitude (Venkatesan et al., 1984; Van Allen and Randall, 1985; 
Webber and Lockwood, 1986). The flux of solar cosmic rays decreases as l/R^. For isotopes with half- 
lives longer than a few periods (15-20 y), the variation in production rates due to changing heliocentric 
distance will be minimal for the range of orbital parameters considered. On the other hand, longer-term 
orbit variations may be significant. 

Mission planning assumes that as the comet orbit evolved over the last few million years its average 
heliocentric distance decreased in a stochastic but more or less monotonic way (Ahrens et al., 1987). At 
distances beyond the heliopause (> 50 AU), cosmic-ray fluxes may increase to ~4 times the values 
typical of the inner solar system (Reedy, 1987b). Such fluxes could raise production rates high enough 
to create observable effects in the longest-lived and stable isotopes provided that the comet achieved its 
rendezvous orbit or one much like it within a few half-lives of capture from the Oort cloud. Figure 2 
illustrates the effect schematically for a case in which the comet went from an orbit with high fluxes to 
one with lower, local fluxes about 1,000 years ago. Figure 2 suggests that high concentrations of longer- 
lived isotopes alongside of normal inventories of shorter-lived isotopes in the same sample would 
constitute a signal for high fluxes beyond the heliopause. With careful analysis of radioisotope 
inventories as a function of half-life, we may be able to separate the effects of mass loss from those due 
to changes in the cosmic-ray flux. 


Fractional inventory 

Figure 2. A schematic illustration of the expected inventories of cosmogenic radionuclides vs. their 
respective half-lives. The calculation assumes that the comet arrived in the inner solar system 1000 y 
ago from a region where the cosmic ray flux had a value L times its present one. e is the erosion rate; p 
is the thickness of matter that reduces the nuclide production rate by a factor of two. 


Another time-dependent effect of comparable magnitude but subtler origin may affect cosmogenic 
nuclide contents. Begemann and Schultz (1988) have argued that the development of the flux of 
secondary cosmic ray particles, which produce more than half the inventories of most isotopes, depends 
strongly on the composition of the matrix. Accordingly, if the composition of the matrix changes as ices 
vaporize or with regolith development (Cintala and Horz, 1987), so will the cosmogenic nuclide 
production rates. Fortunately the production rates are amenable to laboratory and computational study 
and it should be possible to infer initial and final compositions of the core from direct observation. 


To decipher the cosmogenic record we will need to know the cosmogenic nuclide profiles for each 
of the separate phases and the petrography and stratigraphy of the core as a whole. We next consider 
sampling and measurement protocols that will allow us to obtain this information. 

The total amount of cometary material returned to terrestrial laboratories may be very small. 
Magnani et al. (1989) list as the samples likely to be recovered: one deep segmented drill core; a 'pristine' 
sample from the bottom of the drill hole; and possibly a surface sample from a different location 
obtained with a "harpoon." B. Clark (pvt. comm.) believes that we may be able to recover a larger and 
more diverse set of samples. To be conservative, our suggestions for sampling sites, sampling 
procedures, and sample transportation and handling requirements are based on the sampling scenario of 
Magnani et al. (1989). 

Sampling locations - If the core is to come from a location of high activity, then only the shorter-lived 
isotopes will cast light on the history of the material. The measurements are possible but challenging 
(see below). Accordingly, unless a reliable, alternative approach becomes available, we would urge that 
an effort be made to recover additional material from a less active and presumably older site. With such 
material we could more effectively address questions related to surface loss mechanisms, to orbital 
evolution, and to the variability of the cosmic ray flux. 

For our purposes a narrower but longer core would be preferable to a shorter, thicker one. The 
reason is that if coring penetrates deeply enough to reach the regime where production rates decrease 
monotonically with depth then estimates of total inventories and hence of mass loss (vaporization) or 
gain (lateral transport) will be possible. 

Documentation - The production rates of cosmogenic nuclides reflect the geometric conditions under 
which irradiation took place. In a column cored from a semi-infinite slab, production rates depend only 
on depth and chemical composition. In more complex topographic settings - valleys or mounds - 
production rates will also reflect the sample's "view" of the sky (Russ and Emerson, 1980). Therefore, in 
order to interpret the cosmogenic nuclide contents we will have to know the locations of all samples with 
respect to each other and to the surface. Toward this end, the neighborhood of each sampling site 
should be documented and the sampling sequence recorded by stereometric photography or other means 
(see also: Ahrens et al., 1987; Englert, 1988). In order to insure the fidelity of the stratigraphic record, 
coring procedures should be designed so as to minimize the physical disturbance. Maintenance of the 
core's integrity during transport and subsequent handling procedures is also important. If sample 
deterioration and the consequent loss of geometrical information seem likely due to acceleration, 
vibration, or heating, then it would be desirable to store separately subsamples for cosmogenic nuclide 

Extraterrestrial measurement of short-lived isotopes - The 2-3 year return to earth will occur with the 
samples exposed to altered irradiation conditions. As a result, cosmogenic nuclide contents will change 
as will other radiation-related properties of the material such as thermoluminescence and free radical 
concentrations. The effects will be largest for the cosmogenic nuclides with half-lives of a few years or 
less and in samples from greater depth where cometary production rates were low. Two kinds of 
measures could help to conserve the information potential of the isotopes with half-lives between 2 and 
5 years, namely, optimization of shielding and monitoring of the nuclear-active flux en route. We 
consider each of these measures in turn. 


Weight constraints eliminate the possibility of screening out cosmic rays entirely by the use of 
massive shielding. The obvious fallback, a moderately shielded location in, say, a heavy cargo 
compartment, would unfortunately also promote the development of a substantial flux of secondaries. 
As it turns out, the best practical way to limit cosmogenic nuclide production in transit is to place the 
samples in the least shielded location (Englert, 1988) consistent with temperature control requirements. 
At least part of the residual secondary radiation from the spacecraft - the thermal neutrons - could be 
eliminated by the judicious use of Cd, B, or Li shielding. 

To monitor the remaining flux, the spacecraft should be equipped with active detectors for charged 
particles and neutrons. When the return trip begins the spacecraft should also have the capability to 
deploy passive detectors, i.e., foils with various activation thresholds that will accumulate short-lived 
(ti/2 < 3 y) nuclear-reaction products that can be conveniently determined in the terrestrial laboratory. 
Track detectors that could be positioned over fissionable or alpha-emitting material at the time when 
samples are transferred to the spacecraft would also be useful (Woolum et al., 1975). 

Even these measures will probably not suffice for the shortest-lived isotopes (Table 1) for their 
concentrations will have been too greatly altered by the time the mission returns with samples. Each 
isotope has some characteristic utility - ®0Co as a thermal neutron monitor and '^Be and ^^Nz as possible 
indicators of exposure age. In the case of ^Be and, with some restrictions, 22^3 extraterrestrial 
measurements, either on the comet or in the spacecraft, offer the only way to determine activities. The 
radiation environment in the spacecraft will be inhospitable to the measurement of short-lived 
radioactivity; large guard counters would be needed to reduce background rates. Nonetheless the 
technical feasibility of such measurements deserves further study. An alternative may be the 
measurement of cosmic-ray particle fluxes on the comet, at the sample site itself, by means of neutron-, 
gamma-, and/or charged particle spectrometry. Such an approach has proved useful on the moon (15) 
but will require further study for application to a comet. 

Material requirements for non-destructive measurements - Upon return, the entire sample should be 
analyzed by non-destructive, low-level counting of t radiation. As a bonus, the spectra could provide 
information about the naturally occurring radioactive species *°K, Th, and U (provided they are 
abundant enough). The construction of low-level, low-temperature equipment suitable for this 
application should be carried out in advance of the mission. 

Material requirements for destructive measurements - Most long-lived cosmogenic nuclides can be 
determined in small samples - less than 1 mg in favorable cases. For the analysis of a large suite of 
nuclides, samples of 20-40 mg would seem appropriate. The siliceous phases are the most likely to retain 
a coherent record of cosmic-ray effects and deserve the most intense study. Significant production of 
'H, i<C and ^^Be will also occur in the icy and organic components of a comet. Therefore 40-mg 
samples of these phases, too, should be reserved for cosmogenic nuclide analysis. 

We would request perhaps 20 samples from the core. The mass required would be a very small 
percentage of the total planned recovery. As the primary aim of the measurements is to understand 
surface processes, the sampling density close to the surface should be greater than in the presumably 
undisturbed depths of the comet. The distances between these samples should probably increase 
exponentially as depth along the core increases. Should older material be available, from the harpoon 
perhaps, then especially dense sampling of the topmost few centimeters would be useful for tracking the 
products of solar cosmic rays. 

Information from related measurements - As noted in Table 1, many cosmogenic nuclides can be made 
from more than one of the elements that are present, so the chemical composition of a sample is needed 
to interpret a cosmogenic-nuclide measurement. Other related information of interest would include 
nuclear track densities, thermoluminescence results, free radical concentrations, and the concentrations of 
stable cosmogenic nuclides. 



We know little about the production and retention of cosmogenic nuclides in volatile materials like 
those observed in comets. To lay the groundwork for the interpretation of sample data, cosmogenic 
nuclide production in material rich in hydrogen and carbon should be studied by using both theoretical 
and experimental techniques. Accelerator bombardments (e.g., Englert et al., 1987) of volatile-rich 
material can simulate the cosmic-ray irradiation of comets. The computer codes developed to model 
high energy interactions and neutron transport should be tested against the results. The information now 
lacking includes the numbers of secondary particles, especially neutrons, made in cometary-like material 
and the energy spectra of these particles from the eV to the GeV range. Also missing are cross sections 
for many nuclear reactions, such as that for the production of i°Be from carbon. 

Related work is already under way. There are gamma- ray and neutron spectrometers scheduled to 
go to Mars in 1992 on the Mars Observer. Measurements by these instruments over the Martian poles 
and perhaps elsewhere on the planet will broaden our experience with planetary surfaces that are rich in 
hydrogen and carbon (Drake et al., 1988). 

The behavior of the cosmogenic nuclides in material subject to volatilization deserves study. 
Refractory isotopes (such as i°Be) may be left behind if the surrounding material is slowly lost by 
volatilization. By using material with the composition of a comet, we may be able to carry out the 
appropriate simulation experiments. 

Finally, we should assess carefully the kinds of spacecraft instrumentation that might allow the 
determination of the shorter-lived nuclides. Even if direct counting proves impractical for them, 
appropriate monitoring of the cosmic ray flux would enhance the information conveyed by isotopes with 
somewhat longer half-lives. 


The utility of cosmogenic nuclide measurements in returned cometary material will depend, finally, 
on the rate of mass loss in the area selected for sampling. If that rate turns out to have been low to 
moderate, then the return of information available from the analyses will amply justify the necessary 
investment of material and time. In anticipation of a favorable outcome, we would urge that the 
definition of curatorial responsibilities and the equipping of the curatorial facility allow specifically for 
the needs associated with the measurement of the cosmogenic radionuclides. A modest program of 
terrestrial research in preparation for the mission would yield results applicable in a variety of planetary 



Ahrens, T. J.; Atzei, A.; Begemann, F.; Brownlee, D. E.; Campins, H.; Chang, S.; Coradini, A.; 
Eberhardt, P.; Festou, M. C; Grun, E.; Harris, A. W.; Hechler, M.; Kerridge, S. J.; Langevin, Y.; 
McDonnell, J. A. M.; Pillinger, C. T., Schwehm, G.; Stoffler, D.; Wanke, H.; Wasserburg, G. J.; West, R. 
M.; and Wood, J. A.: Rosetta - The comet nucleus sample return mission. SCI(87)3, Plan. European 
Space Agency, 1987. 

Armstrong, T. W.; and Alsmiller, Jr., R. G.: Calculation of cosmogenic radionuclides in the Moon and 
comparison with Apollo measurements. Proc. Lunar Sci. Conf. 2nd, 1971, pp. 1729-1745. 

Bar-Nun, A.: Experimental studies of gas trapping in amorphous ice and thermal modelling of comets - 
implications for Rosetta. Abstract in Workshop on analysis of returned comet nucleus samples. Lunar 
Planet. Inst. Contrib. 691, 1989, pp. 5-7. 

Begemann, F..; and Schultz L.: The influence of bulk chemical composition on the production rate of 
cosmogenic nuclides in meteorites. Lunar Planet. Sci., vol. 19, 1988, 51-2. 

Brin, G. D.; and Mendis, B. O.: Dust release and mantle development in comets. Astrophys. J., vol. 
229, 1979, 402-408. 

Cintala, M. J.; and Horz. F.: The effects of impact velocity on the evolution of experimental regoliths. 
Proc. 18th Lunar Planet. Sci. Conf., Cambridge Univ. Press, 1987, pp. 409-422. 

Clarke, W. B.; Jenkins, W. J.; and Top, Z.: Spectrometric measurement of ^He. Int. J. Appl. Rad. Iso., 
vol. 27, 1976, 515- 522. 

Drake, D. M.; Feldman, W. C; and Jakosky, B. M.: Martian neutron leakage spectra. J. Geophys. Res., 
vol. 93, 1988, 6353-6358. 

Elmore, D.; and Phillips, F. M.: Accelerator mass spectrometry for measurement of long-lived 
radioisotopes. Science, vol. 236, 1987, 543-550. 

Englert, P. A. J.: Cosmogenic nuclides in the Martian surface: constraints for sample recovery and 
transport. Lunar Planet. Inst. Tech. Rep. 88-07, 1988, 75-76. 

Englert, P.; Reedy, R. C; and Arnold, J. R.: Thick-target bombardments with high- energy charged 
particles: Experimental improvements and spatial distribution of low-energy secondary neutrons. Nucl. 
Instrum. & Methods, vol A262, 1987, 496-502. 

Fink, D.; Middleton, R.; Sharma, P.; and Klein, J.: AMS measurements of ^^Ca in terrestrial samples 
without pre-enrichment. Abstract presented at V.M. Goldschmidt Conf., Baltimore, MD, May 11-13, 
1988 (Geochem. Soc). 

Flammer, K. R.; Jackson, B.; and Mendis, D. A.: On the brightness variation of comet Halley at large 
heliocentric distance. Earth Moon Planet., vol. 35, 1986, 203-212. 

Fruchter, J. S.; Reeves, J. H.; Evans, J. C; and Perkins, R. W.: Studies of lunar regolith dynamics using 
measurements of cosmogenic radionuclides in lunar rocks, soils and cores. Proc. Lunar Planet. Sci Conf., 
12th, 1981, 567-575. 

Glass, B. P.: Introduction to Planetary Geology, Cambridge Univ. Press, 1982, p. 346. 

Horanyi, M.; Gombosi, T. I.; Cravens, T. E.; Korosmezey, A.; Nagy, A. F.; and Szego, K.: (1984) The 
friable sponge model of a cometary nucleus. Astrophys. J., vol. 278, 1984, 449-455. 

Jessberger, E. K.; Christoforidis, A.; and Kissel, J.: Aspects of the major element composition of Halley's 
dust. Nature, vol. 332, 1988, 691-695. 


Keller, H. U.; Arpigny, C; Barbieri, C; Bonnet, R. M.; Cazes, S.; Coradini, M.; Cosmwici, C. B.; 
Delamere, D. A.; Heubner, W. F.; Hughes, D. W.; Jamar, C; Malaise, D.; Reitsema, H. J.; Schmidt, H. 
U.; Schmidt, W. K. H.; Seige, P.; Whipple, F. L.; and Wilhelm K.: First Halley multicolour imaging from 
Giotto. Nature, vol. 321, 1986, 320-326. 

Klinger, J.; Benkhoff, J.; Espinasse, S.; Grun, E.; Ip, W.; Joo, F.; Keller, H. v.; Kochan, H.; Kohl, H.; 
Roessler, K.; Sebold, W.; Spohn, T.; and Thiel, K.: (1988) How far do results of recent simulation 
experiments fit with current models of cometary nuclei? Lunar Planet. Sci., vol. 19, 1988, 611-612. 

Kohl, C. P.; Murrell, M. T.; Russ, G. P. Ill; and Arnold, J. R.: Evidence for the constancy of the solar 
cosmic ray flux over the past ten million years: ^^Mn and 26a1 measurements. Proc. Lunar Sci. Conf. 9th, 
1978, 2299-2310. 

Kubik, P. W.; Elmore, D.; Conard, N.; Nishiizumi, K.; and Arnold, J. R.: Determination of cosmogenic 
^^Ca in a meteorite with tandem accelerator mass spectrometry. Nature, vol. 319, 1986, 568-570. 

Langevin, Y.; Arnold, J. R.; and Nishiizumi, K.: Transport processes on the lunar surface: comparison of 
model calculations with radionuclides data. J. Geophys. Res., vol. 87, 1982, 6681-6691. 

Lanzerotti, L. J.; Reedy, R. C; and Arnold, J. R.: Alpha particles in solar cosmic rays over the last 
80,000 years. Science, vol. 179, 1973, 1232-1234. 

Lingenfelter, R. E.; Canfield, E. H.; and Hampel V. E.: The lunar neutron flux revisited. Earth Planet. 
Sci. Lett., vol. 16, 1972, 355-369. 

Magnani, P. G.; Gerli C; and Colombina, G.: Candidate sample acquisition systems for the Rosetta 
mission. Abstract in: Workshop on analysis of returned comet nucleus samples. Lunar Planet. Inst. 
Contrib. 691, p. 47. 

Nishiizumi, K.: Measurement of 53Mn in deep-sea iron and stony spherules. Earth Planet. Sci. Lett., vol. 
63, 1983, 223-228. 

Nishiizumi, K.; Murrell, M. T.; and Arnold, J. R.: ^SMn profiles in four Apollo surface cores. Proc. 
Lunar Planet Sci. Conf., 14th, J. Geophys. Res., Suppl., vol. 88, 1983, B211-B219. 

Nishiizumi, K.; Elmore, D.; Ma, X. Z.; and Arnold, J. R.: i°Be and ^eci depth profiles in an Apollo 15 
drill core. Earth Planet. Sci. Lett., vol. 70, 1984a, 157-163. 

Nishiizumi, K.; Klein, J.; Middleton, R.; and Arnold, J. R.: z^aI depth profile in Apollo 15 drill core. 
Earth Planet. Sci. Lett., 70, 1984b, 164-168. 

Raisbeck, G. M.; Yiou, P.; Klein, J; Middleton, R; and Brownlee, D.: 26Al/iOBe in deep sea spherules as 
evidence of cometary origin. In "Properties and Interactions of Interplanetary Dust" R. H. Giese and P. 
Lamy, eds., D. Reidel, 1985, 169-174. 

Reedy R. C: Predicting the production rates of cosmogenic nuclides in extraterrestrial matter. Nucl. 
Instrum. & Methods, vol. B29, 1987a, 251-261. 

Reedy, R.C.: Nuclide production by primary cosmic-ray protons. Proc. 17th Lunar Planet. Sci. Conf, 
Part 2, J. Geophys. Res., vol. 92, 1987b, E697-E702. 

Reedy, R. C; and Arnold J. R.: Interaction of solar and galactic cosmic-ray particles with the moon. J. 
Geophys. Res., vol. 77, 1972, 537-555. 

Reedy, R. C; Arnold J. R.; and Lai, D.: Cosmic-ray record in solar system matter. Annu. Rev. Nucl. 
Part. Sci., vol. 33, 1983, 505-537: and Science, 219,127-135. 


Russ, G. P.; and Emerson, M. T.: ^^Mn and 26a1 evidence for solar cosmic ray constancy - an improved 
model for interpretation. Proc. Conf. Ancient Sun, Pergamon, 1980, pp. 387-399. 

Sagdeev, R. Z.; Blamont, J.; Galeev, A. A.; Moroz, V. I.; Shapiro, V. D.; Shevchenko, V. I.; and Szego, 
K.: Vega spacecraft encounters with comet Halley. Nature, vol. 321, 1986, 259-273. 

Schaeffer, O. A.; Nagel, K.; Fechtig, H.; and Neukum, G.: Space erosion of meteorites and secular 
variation of cosmic rays (over 10^ y). Planet. Space Sci., vol 29, 1981, 1109-1118. 

Stern, S. A.: (1988) Collisions in the Oort cloud. Icarus, vol. 73, 499-507. 

Van Allen, J. A.; and Randall B. A.: Interplanetary cosmic ray intensity: 1972-1984 and out to 32 AU. 
J. Geophys. Res., vol. 90, 1985, 1399-1412. 

Venkatesan, D.; Decker R. B.; and Krimigis S. M.: Radial gradient of cosmic ray intensity from a 
comparative study of data from Voyager 1 and 2 and IMP 8. J. Geophys. Res., vol. 89, 1984, 

Wdowiak, T. J.; Robinson, E. L.; Flickinger, G. C; and Boyd, D. A.:(1989) Ion bombardment 
experiments suggesting an origin for organic particles in pre-cometary and cometary ices. Abstract in: 
Workshop on analysis of returned comet nucleus samples. Lunar Planet. Inst. Contrib. 691, pp. 79-80. 

Webber, W. R.; and Lockwood J. A.: Interplanetary cosmic-ray radial and latitudinal gradients derived 
in 1984 using IMP 8, Voyager, and Pioneer data. Astrophys. J., vol. 302, 1986, 511-516. 

Weissman, P. R.: Physical processing of cometary nuclei. Abstract in: Workshop on analysis of returned 
comet nucleus samples. Lunar Planet. Inst. Contrib. 691, pp. 81-82. 

Whipple, F. L.: The cometary nucleus: current concepts. Astron. Astrophys., vol. 187, 1987, 852-858. 

Woolum D. S.; Burnett, D. S.; Furst, M.; and Weiss, J. R.: Measurement of the lunar neutron density 
profile. The Moon, vol. 12, 1975, 231-250. 




W. F. Huebner 

D. C. Boice 

Southwest Research Institute 

San Antonio, Texas 


Morphology and Compositional Differentiation of the Surface of Comets 

W. F. Huebner and D. C. Boice, Southwest Research Institute, San Antonio, TX 78284 

Giotto images reveal many features on the nucleus of Comet Halley, including gas- and dust- 
producing sources surrounded by an inactive region. In the inactive region, crater-Uke structures 
can be seen that may be extinct sources. These structures may develop by surface erosion of an 
active area and deposition of some excavated material on the periphery, creating crater-like rims. 
These rims are formed from "clumps" of comet regolith that can be lifted by the escaping gas. The 
lack of lift caused by the divergence of the gas flow near the boundary of cin active region lets them 
fall back on the nucleus and create a rim. This may be a continuous process during perihehon 

Supplementing the original concept of investigating the active and inactive regions, we con- 
clude that three compositionally distinct areas should be sampled during the Rosetta mission: (1) 
The active regions rich in frozen gases and unprocessed dust. (2) The inactive region covered by a 
thin layer of fine dust enriched in organics that may be sintered. (3) The crater-like rims containing 
"clumps" of processed organics, siHcates, and trapped frozen gases. 

To assess the concepts of gas production we consider Comet Halley for comparison. The 
measured rate of gas production was (5 — 7 lO'^^ molecules s"-' during the Giotto encounter at a 
heliocentric distance of r = 0.9 AU. About 80 to 85% of the gas was water. For an albedo of 0.03 
and a 30° average angle of sunlight incidence, the specific gas production rate is Z^ ~ 1.4 lO'^'^ 
molecules m~'^ s~-^. The active area is therefore about 45 km". Since the total stirface area of 
P/Halley is approximately 400 km", the active area is about 20% of the illimiinated surface. On 
the Halley Multicolour Camera (HMC) images, no dust production is evident from the inactive 
area (see Fig. 1). Nevertheless, if we asstmie that as much as 10% of the gas is produced on the 
inactive area, then the specific gas production on the inactive area, which covers about 80% of the 
illuminated svirface, is only 3% of that of the active area. A reasonable result is Z\ = (0.03 ±0.03) 
Za- Such a low gas production rate cannot entrain large dust particles. It is therefore reasonable to 
assume that all the dust comes from the active area. Figure 1 is consistent with this interpretation. 

For other short-period comets the total gas production rate is smaller than that of P/Halley. 
If comet compositions are similar, this suggests that the total active area is smaller than that on 
P/Halley. Nothing definitive can be said about the specific gas production rates, but it appears 
reasonable that both Za and Z\ are similar to the values of P/Halley. 

Since the gas production rate is higher for the active areas than for the inactive areas, an 
overpressure tends to develop over the active regions. Pressure equilibration then causes a surface 
wind to develop. This wind carries fine dust rich on CHON with it. The smallest dust particles 
can be moved laterally out of the dust "jet" by a few collisions with the gas, while the large 
dust particles require many colHsions and therefore remain in the dust "jet" . In support of this 
argument, the HMC images show an intensity gradient across the nucleus. As can be seen in Fig. 1, 
the intensity decreases from the active regions in the antisolax direction. Some of the fine dust may 
settle on the surface, particularly on the night side. When this side faces the Sun again during the 
comet's rotation, the organic dust may get sintered. 

The gas over the active areas entrains the dust. The entrainment can be approximated in 
two limiting cases (Huebner, 1970): If the mean free path, .i. of the gas is large with respect to 






' ' -r ":'^^H 


*,■;'■);', ^H 


, 1 ■. . ^^ 

C MPAE, 1986 ^H 

1 UPAE - PK8S0C7-K 

Fig. 1. A composite of six images of Comet Halley obtained with the Halley Multicolour Camera on the Giotto 
spacecraft (Keller et al., 1988). The resolution increases toward the brightest part. Illumination by the Sun is 
from the left at about 28° above the horizontal and 12" behind the image plane. Most of the visible surface is not 
illuminated. The light intensity gradient from the top left on the nucleus to the bottom right is apparent. The 
dust emission is concentrated in jet-like features emanating from the subsolar hemisphere. Structural details of the 
surface are visible down to the resolution limit of about 100 m. A crater-like structure is visible on the nucleus 
between the two brightest dust emissions. 


the dust particle size of effective radius a, i.e.. A = v/{Zcr) > a . free molecular flow is a good 
approximation. In the other limit, when A < a, fluid dynamics is a good approximation. 
The equation of motion for a dust particle in the free molecular flow approximation is 

47r 3 (PR ^^, dR^^ ,Za/i?N\" (47r/3)-aVd-RxPNC? JL 

yapd-^ = m.M(.--)-.a--(^— J ^^ , ® 

where R is the radial position of the particle in the coma, i^N is the effective radius of the nucleuS; 
a the radius of the particle, v the gas velocity, M the molecular weight of the gas, m^ the unit 
atomic mass, pd the density of the dust particle, pN the density of the nucleus, Za. the specific 
gas production rate of an active area, and G the universal gravitational constant. The term on 
the left side of Eq. (1) is mass times acceleration of the dust particle. The first term on the 
right represents the time rate of change of momentum of the dust particle caused by collisions 
with gas molecules and the second term on the right represents the weak gravitational pull of the 
nucleus. The centripetal acceleration can be neglected since it contributes only a few percent at 



the equator. To solve for the mcLximxim particle size, an,, that can be entrained by the gas at the 
surface, R = i?N, we set d'R/dt'^ = dR/dt = and obtain 


dm = 



For i?N ^ km, u ~ 100 m s-^ M ~ 18, ^a ^ Pn - 500 kg m"^, Z ~ 10^2 m-^ s'^ at 1 AU, 
we find Om ~ 0.1 m. 

A similar equation can be solved in the fluid dynamic limit. In that case 


_ 1.8510-^r^/- ,^. 

"^ ~ 1 + 680/T ' ^ ' 

is the viscosity. For the same parameters as above and with T ~ 200 K, yl ~ 0.1 m and a^ — 0.1 
m, i. e., free molecular flow does not apply, but the fluid dynamic hmit has not yet been reached. 

It is interesting to note that the fluid dynamic Hmit also gives an upper hmit to the dust size 
distribution. Even for large gas production rates at smaller hehocentric distances, the maximum 
size of particles that can be lifted from the siirface by gas entrainment remains nearly constant 
since the temperature of the gas and dust does not significantly change. 

Centimeter- and decimeter-sized "clumps" of dust are easily entrained by the gas in the center 
of an active region. Gas production will fluctuate. During a small decrease in the gas production 
the Hft by the gas will be reduced and a "climip" , under the action of the weak gravity, falls back 
to the nucleus. Similarly, the divergent flow of the gas over an active region results in a reduction 
of Hft and the "clumps" faU back. This process has also been suggested by Sekanina (1983). The 
tendency will be to fall outside of the active area, where they will accumulate into a rim structure. 
Even though centimeter- sized particles are less sensitive to these two actions then decimeter- sized 
particles, they move only slowly to some height above the nucleus. As the nucletis rotates under 
them the hft from the gas "jet" is removed and they will fall back to the nucleus on the evening 
side of the active area. Thus any rim that may form will be imperfect. Figure 1 shows several 
crater-like structures. One structure is clearly visible on the nucleus between the two brightest 
jet-Hke features. It may be an extinct active area or one that has not yet turned on in the early 
morning on the nucleus. It is very shallow and about 2 km in diameter. The rim is not perfectly 
circular and may be open on two sides. Figure 2 presents an intensity trace perpendicular through 
the northern jet-Hke feature about 50 to 100 m above the nucleus. The feature is opticaUy thin, 
so that the intensity is directly proportional to the column density of the dust through the "jet". 
A Gaussian curve has been drawn through the intensity profile. The deviations to this Gaussian 
are filaments that persist to large distances firom the nucleus. The details in the intensity profile 
are not sensitive to the height of the intensity trace above the surface. 

The "clumps" in the rim will be rich on unprocessed dust, i.e., dust that contains sihcates as 
well as CHON. The "clumps" may still contain some trapped frozen gases, while their surface may 
be depleted of the ices. 

We conclude that large particles are Hfted from the surface near the center of an active region. 
A small reduction in the gas production causes the largest particles to fall back on the nucleus, 
while the smallest particles are carried by the surface wind and may fall back to the s\irface over 


20000 _j 





18000 _ 



16000 _ 





14000 _ 




12000 _ 





10000 _ 
8000 _ 
6000 _ 
4000 _ 
2000 _ 





3 1.5 


1 1 

4.5 6.0 

Fig. 2. An intensity trace about 50 to 100 m above the nucleus perpendicular ihrough the top jet-like feature seen m 
Fig. 1 (Reitsema, 1989). A Gaussian curve has been fitted to the intensity data. The structural deviation from riie 
Gaussians are filaments of dust in the jet-like feature; they are reproducable at different heights above the surface. 

the inactive region. The intermediate dust with particle size of the order of 1 /iin stays in the 
dust-hke features and is carried into the coma and eventually into the dust tail. Centimeter-sized 
particles can be Hfted to some height above the nucleus. However, since the nucleus rotates under 
the particles, they lose their lift from the entraining gas and fall back on the nucleus toward the 
evening side. 

Erosion of an active region is of the order of tens of meters per apparition. On the same time 
scale the rim formation may be of the order of a few meters. At least three different regions on a 
nucleus should be sampled. The original plan for the Rosetta mission was to sample two different 
regions on the nucleus: One area was an active region which is expected to be rich on frozen gases 
and unprocessed dust. A core sample from such a region is most desirable. The second sample is 
from an inactive region which is rich on inert material that possibly is covered Idv a thin layer of 
fine, organic dust. This dust may be sintered on the surface. A thin surface sample would be most 
desirable. .A. third region that should be sampled is the rim around an active area. It may contain 
coarse material with some trapped ices. It will give information about the formation of "clumps 
in or on a comet nucleus. 

This work was supported by funds from the NASA Planetary .\tmosplieres Program. 

Huebner, W. F., Astron. Astophys. 5, 2S6 (1970). 

Keller. H. U.. Kramm, R., and Thomas, N., Nature 331. 227 (19SS). 

Reitsema. H. J., private communication (1989). 

Sekanina. Z.. Adv. Space Res. 2. 121 (1DS3). 




J. A. M. McDonnell 

G.S. Pankiewicz 

P. N. W. Birchley 

S. F. Green 

C. H. Perry 

Unit for Space Sciences 

University of Kent 
Cantebury, Kent, U. K. 




J.A.M. McDonneU, G.S. Pankiewicz, P.N.W. Birchley, S.F. Green, C.H. Perry. 
Unit for Space Sciences, University of Kent, Canterbury, Kent CT2 7NR, U.K. 


The comet Halley dust mass distribution measured by the Giotto DIDSY and PIA 
experiments is used to derive the dust to gas mass ratio \i for the nucleus material. The 
excess of grains observed for masses > 10-9 kg places }i in the range 1-200 if the observed 
size distribution is representative of the average properties of the coma. The lower bound 
corresponds to integration up to the largest particle (-Ig) impacting Giotto. The mass and 
area distributions at the nuclear surface for distributions with and without this large 
particle excess are compared. 


The close approach of Giotto to comet P/Halley during its 1986 apparition offered a 
unique opportunity to study the distribution of particulates of masses up to one gram. Data 
acquired by the dust shield detector system, DIDSY [1] and the front end channels of the 
highly sensitive mass spectrometer PIA [2], provide definition of the detected distribution 
as close as 1000km to Ac nucleus. Measured particles extend from lO-l^kg (~ 0.02nm) to 
some 30mg (~ 2mm) and can also be inferred for masses up to the region of Ig (~ 5mm) by 
virtue of the spacecraft deceleration of 23.05 cm s'^ [3]. 

This work examines implications of the measured data with respect to: 

1 . The flux and size distribution of particles leaving the nucleus surface. 

2. The dust to gas mass ratio in the nucleus matrix as a function of mass, |i(m). 

3. The mass and area distributions of grains in the nucleus matrix. 

The results may be applied to the task of remote sensing of a cometary nucleus to 
locate active areas. In a Comet Nucleus Sample Return Mission such as CNSR - 
ROSBTTA, the identification of a fresh production surface by reflectance or emission 
properties at wavelengths from optical to radar depends upon the scale depth of absorbing 
material and the detectability of ice in the matrix. 


Calculation of dust fluxes from Giotto DIDSY and PIA impact data is described by 
McDonneU et al [4]. Figure la shows the measured cumulative flux Oc(m) (number of 
particles of mass > m impacting the spacecraft per m^ per second) of particles observed by 
DIDSY and PIA in the coma at a mean distance from the nucleus of 5240km determined 
from post encounter measurements in the period +60 to +120s. The soHd line (IN-SITU) 
represents the distribution as measured, with the same slope assumed for m>10-5kg (~ 
largest measured particle derived from DIDSY data, [4]). The dotted line (MODEL) 
indicates the distribution assuming that the observed excess of large masses is not 
representative of the average properties of the coma. Distributions of this form have 



radius (m 







-3 -2 






•? 2 














X -2 

.1 -^ 














-18. -15 -12 -9 -6 -3 

Log mass (kg) 

Log radius (m) 

-8 -7 -6 -5 -U -3 -2 -1 



Z 6 




■V\^ IN-srru 


1 -2 


• N 
• \ 

E -A 




MODEL ■■•.. 


■ b) 

— 1. — 1. . I . . 1 ■ 

Figure 1 

Cumulative mass 

a) in the coma, 

b) at the nucleus. 
The solid curve is for 
Giotto data (nuclear 
distance of 5240km post- 
encounter) and the 
dotted line for a model 
v/ith uniform mass index 
at large masses. 

-18 -15 -12 -9 -6 -3 

Log mass (kg) 


generally been assumed from interpretation of remote sensing data and were used for pre- 
encounter modelling [5] with a constant size index at large masses of u = 3.7 (a = 0.9) 
The dust cumulative flux On(m) at the nucleus surface is given by 

Xkl III 

0^(m) = Jn^(m)dlogm = Jn^(m)— — (— ) dlogm 

^ " 

where the grain velocity v(a) is taken from figure 2, Rn=5.2km is the effective nucleus 
radius (derived from model of Sagdeev et al, [6]) and Vs is the spacecraft velocity = 68.4 
km s"l, nn(m) and nc(m) are the differential fluxes in the nucleus and coma respectively. 
We derive the cumulative flux distribution at the nucleus (figure lb) assuming radial 
trajectories and a velocity distribution of the form shown in figure 2. Dust velocities are 
calculated using the approximation of Divine [7] with a maximum liftable mass of radius 
15cm (assuming a nucleus density = 800 kg m-3 and an active fraction of the nucleus of 
10% of the total surface area, [8]). The grain density is assumed to take the form 

p(a) = 3000 - 2200 (-^) kg m'^ 


[5] where ao=2xl0-6m. 


The resultant fluxes are then integrated over all relevant masses and used to obtain a 
dust production rate that may be directly compared to the measured gas production rate of 
2.55 X 10^ kg s'l [9] to yield a dust to gas mass ratio as a function of mass (figure 3). The 
dust to gas mass ratios p.(m) are shown here as a function of the mass of the largest grains 
included. The shaded region indicates the limits of the likely value of |j.(m) where 

- in 

|i(m) = n (m') m' dlogm' 

whereas the dotted line (MODEL distribution) may be compared with previously derived 
results from remote sensing. Qg is the gas production rate = 6.9x1029 mol s-1 [9] and Tl is 
the mean molecular mass = 3.7xlO-26kg [5]. Although the nucleus is known to comprise 
localised active areas on a predominantly inactive surface the gas production rate is an 
average value for the data measured during the whole encounter. Likewise, since DIDSY 
and PIA data show relatively small flux variations they can be treated as an average value 
for the period +60 to +120 seconds. During this period the sub-sateUite track was over the 
sunlit hemisphere. 

The MODEL distribution produces a bulk dust to gas ratio of ~ 0.2 but figure 3 
clearly shows how the measured large mass distribution enhances the total mass of dust to 
produce a bulk ratio in the range 1-200. The lower limit to the shaded region is based on 
the IN-SnU distribution but with no grains more massive than the largest inferred from 
the total spacecraft deceleration (~lg). The largest particles measured by DIDSY indicate 


2 ■ 





Log radius (m) 
-6 -5 -A -3 


Gas velocity 

Largest liftable 

-18 . -15 -12 -9 -6 -3 

Log mass (kg) 


^10 - 




Figure 2 
Grain velocity 
distribution from [7]. 

Log radius (m) 

-5 -i. -3 -2 -T 


-12 -9 -6 
Log mass (kg) 



Figure 3. 

Dust to gas mass ratios |J. 
as a function of the 
largest grains included. 


a lower limit of 1.0 to the bulk ratio, with expected values in the shaded region, but only if 
the observed large mass excess is representative of the coma as a whole. There is 
considerable evidence for an abundance of mm - cm sized grains in cometary comae 
[10,1 1], but ground-based infrared observations of sihcate emission (e.g. [12]) indicate that 
for much of the time grains <20jim in size dominate the coma. 


Using the above distributions the numbers of dust particles embedded in any volume 
of interest may be calculated. Figures 4 and 5 display the number, area and mass 
distributions in Im^ in the coma (5240km from the nucleus) and in the nucleus matrix 
itself. The differential distributions nc(m), Adm) and m.c(ni) are shown for the IN-SITU 
distribution (solid line), and the MODEL distribution (dotted line) in figure 4. The 
number of particles per m3 per log mass interval 

nc(m) dlogm = (nc(m)/vs) dlogm, 
the area of particles per m^ per log mass interval 

s^c(ni) dlogm = (nc(m)7ia2/vs) dlogm 
and the mass of particles per m^ per log mass interval 

mc(m) dlogm = (nc(m)m/vs) dlogm. 
Although the majority of grains are of mass < 10-9 kg, the total grain mass is dominated by 
the largest grains. Remote sensing observations, which are dependent on the cross- 
sectional area, indicate dominant grain masses -10' ^"^kg. The DIDSY data indicate that a 
significant contribution to the cross-sectional area could come from large grains. Figure 5 
shows similar differential distributions for the nucleus material where the number of 
particles per m3 per log mass interval 

nn(m) dlogm = (nn(m)A^) dlogm, 
the area of particles per m^ per log mass interval 

^n(ni) dlogm = (nn(m)7ca2/V) dlogm 
and the mass of particles per m^ per log mass interval 

tnn(m) dlogm = (nn(m)mA^) dlogm, 
where V is the total volume of material ejected per m^ per second from the nucleus (see fig 
6). The nucleus mass distribution is dominated by the largest grains. The maximum mass 
plotted is the calculated largest liftable mass ~10kg. In reality, the tme dust to gas mass 
ratio will depend on the largest grains present in the nucleus which may be larger or 
smaller than this value of maximum Hftable mass. The total cross-sectional area of grains 
at the nucleus for the MODEL distribution is dominated by grains -lO-l^kg. 


The large masses measured by DIDSY significantly enhance the nucleus area and mass 
distributions above -lO-^kg and reduce the effective areas and masses from smaller 
particles because of the lower volume available for such particles. This results in an 
optically thinner nucleus matrix on the scale of several microns, as is evident from the 
boxes of nucleus material represented schematically in figure 6. Each window depicts the 



Log radius (ml 

-8 -7 

6 -5 -U -3 -2 -1 





\ /'' 



\ / 

^ -4 

\ -g- 

/.- ■■ 

•S -6 


mass / 


^ -"" 

Log no. per ht 







if \ ■■■■.,., 




? 1 

-12 -9 -6 

Log mass (kg) 


-12 e 
















• -20 

- -22 

Figure 4. 

Differential number, 
cross-sectional area and 
mass distributions of 
dust grains in the coma. 
Solid line - IN-SITU 
distribution; dotted line 
MODEL distribution. 

15 ■ 

13 ■ 


- 9 

^ 7 

Log radius (m) 
-5 -U -3 



\ ■■. mass ,-'"*, 



V :■;,.. 5 1 




• ■ 



H. ,.- : 




-18 -15 -12 -9 -5 -3 

Log mass (kg) 

• 2 









Figure 5. 

Differential number, 
cross-sectional area and 
mass distributions of 
dust grains on the 
nucleus. Solid line - IN- 
SITU distribution; 
dotted line - MODEL 


number of grains in a 1mm slice of the surface at scales of Im, 1cm, 100|jin and l|im. The 
number of grains of mass m per m^ of the surface material 

n (m) = 





where the denominator is the volume occupied by gas and dust. The dust to gas mass ratio 
|j. is taken from figure 3 at m=10kg and pgas assumed to be 200 kg m-3. The grains 
represented in each case have a range of radii from 0.5 to 10-3 of the respective scale sizes. 
The model distribution is optically thick (optical depth t=1 implying 63% attenuation) at a 
depth of 0.1mm with grains of size <50jim dominating. The presence of large mass grains 
in the IN-SITU distribution (figure 5) reduces the optical depth (t=1 at a depth of 4mm), 
allowing incoming radiation to penetrate further into the nucleus. The MODEL 
distribution gives a very dark sooty appearance for a box 100|im on each side and 1mm 
deep as it is dominated by micron sized particles (the scale depth at which the optical depth 
T = 1.0 in this case is 98|jm). The measured DIDSY number distribution however produces 
a clearer matrix (the scale depth calculated is 3.3mm), even though the distribution is 
dominated by larger grains. 


The appearance of the nuclear surface depends critically on the true distribution of 
large mass grains at the nucleus. Giotto data indicate an excess up to m~lg along its 
trajectory but give no information for larger grains or for the coma as a whole. 


[1] McDonnell, J.A.M., 1987. /. Phys. E.: Sci. Instrum., 20, 741-758. 

[2] Kissel, J., 1986. ESA SP-1077, 67-83. 

[3] Edenhofer, P., Bird, M.K., Brenkle, J.P., Buschert, H., Kursinsiki, E.R., Mottinger, 
N.A., Porsche, H. & SteLzried, C.T., 1987. Astron. Astrophys., 187, 712-718. 

[4] McDonnell, J.A.M., Alexander, W.M., Burton, W.M., Bussoletti, E., Evans, G.C., 
Evans, S.T., Firth, J.G., Grard, R.J.L., Green, S.F., Griin, E., Hanner, M.S., Hughes, 
D.W., Igenburgs, E., Kissel, J., Kuczera, H., Lindblad, B.A., Langevin, Y., 
Mandeville, J.C, Nappo, S., Pankiewicz, G.S.A, Perry, C.H., Schwehm, G.H., 
Sekanina, Z., Stevenson, TJ., Turner, R.F., Weishaupt, U., Wallis, M.K. & 
Zamecki, J.C, 1987. Astrophys. J., 187, 719-741. 

[5] Divine, N., Fechtig, H., Gombosi, T.L, Hanner, M.S., Keller, H.U., Larson, S.M., 
Mendis, D.A., Newbiun, R.L., Rheinhard, R., Sekanina, Z. & Yeomans, D.A., 1986. 
Space Sci. Rev., 43, 1-104 

[6] Sagdeev, R.Z., Krasikov, V.A., Shamis, V.A., Tamopolski, V.I., Szego. K., Toth, I., 
Smith, B., Larson, S. & Merenyi, E., 1986. ESA SP-250, Vol II, 335-338. 

[7] Divine, N., 1981. ESA SP-174, 25-30. 








1— { 

























Figure 6. Schematic of the appearance of dust grains on the nucleus surface on 
different size scales. See text for details. 


[8] Keller, H.U., Delamere, W.A., Huebner, W.F., Reitsema, HJ., Kranun, R., Thomas, 

N., Arpigny, C, Barbieri, C, Bonnet, R.M., Cazes, S., Coradini, M., Cosmovici, 

C.B., Hughes, D.W., Jamar, C, Malaise, D., Schmidt, K., Schmidt, W.K.H. & Siege, 

P., 1987. Astron. Astrophys. 187, 807-823. 
[9] Krankowsky, D. Lammerzahl, P., Herrwerth, I., Woweries, J., Eberhardt, P., Dolder, 

U., Herrmann, U., Schulte, W., Berthellier, J.J., Iliano, J.M., Hodges, R.R. & 

Hof&nan, J.H., 1986. Nature, 321, 326-330. 
[10] Eaton, N., Davies, J.K. & Green, S.F., 1984. Mon. Not. R. astr. Soc. 211 15p-19p. 
[11] Harmon, J.K., Campbell, D.B., Hine, A.A., Shapiro, LI. & Marsden, B.G., 1989. 

Astrophys. J., in press. 
[12] Hanner, M.S., Tokunaga, A.T., Golisch, W.F., Griep, D.M. & Kaminski, CD., 1987. 

Astron. Astrophys., 187, 653-660. 



W. A. Schutte 

Laboratory Astrophysics, Leiden, the Netherlands 

NASA Ames Research Center, Moffet Field, California 

V. K. Agarwal 

Rensselaer Polytechnic Institute 

Troy, New York 

M. S. de Groot 

J. M. Greenberg 

Laboratory Astrophysics 

Leiden, the Netherlands 

P. McCain 

J. P. Ferris 

Rensselaer Polytechnic Institute 

Troy, New York 

R. Briggs 

Center for Laboratories and Research 

New State Department of Health, Albany, New York 


Organic Chemistry in Interstellar Ices; 
Connection to the Comet Halley Results. 

W.A. Schutte*#, V.K. Agarwal@, M.S. de Groot*, 
J.M. Greenberg*, P. McCain@, J.P. Ferris@ and R. Briggs- 

Laboratory Astrophysics, Leiden, the Netherlands. 
#^ NASA Ames Research Center, Moffett Field, Ca.. 
Rensselaer Polytechnic Institute, Troy, N.Y.. 
Center for Laboratories and Research, N.Y. State Department of 
Health, Albany, N.Y. 



Experiments simulating the photolysis of ice mantles on grains in 
dense clouds produce organic molecules similar to the ones observed by 
Giotto's Picca heavy ion analyzer near Comet Halley. 

1. Introduction. 

Mass spectroscopic measurements on the gas and dust in the coma of 
Comet Halley revealed the presence of considerable amounts of organic 
species (Mitchell et al. 1987, 1988, Kissel et al. 1987). Greenberg (1973) 
proposed that prior to the formation of the comet UV processing of the ice 
mantles on grains in dense clouds could lead to the formation of complex 
organic molecules. Theoretical predictions of the internal UV field in dense 
clouds (Prasad et al. 1983) as well as the discovery in interstellar ices of 
species like OCS and OCN- which have been formed in simulation 
experiments by photoprocessing of interstellar ice analogues (Geballe et 
al. 1985, Grim and Greenberg 1987, see also fig. 1) point to the importance 
of such processing. We undertook a laboratory simulation study of the 
formation of organic molecules in interstellar ices and their possible 
relevance to the Comet Halley results. A detailed review is given in 







1 1 


-1 1 

1 1 > 1 1 1 1 ^ 

- "''' -^ 

\ \ 

^^.-^-'^ ^_x*^-''""\/\/'' 

\ \ 






J J 
I 1 
/ / 



/ /■ 








II 1111 — 1 — 




\ \ 


\ \ 

y^ o 

o \ 

/ / 


\ \ 

/o o 

O W33 A vs. 

\ 1 


/ / 






O H2O/CO/NH3 = 5:2:1- 



1 1 


1 I 

Q with/without hu 

I , , . , 1 , 

A (Mm) 


Figure 1 . Lower part. A comparison of the 6 |im interstellar ice feature observed towards the embedded 
source W33 A (circles; Tielens et al. 1988) with the ice mixture H20/CO/NH3 = 5:2:1, before and after UV 
irradiation (solid and dashed line respectively). The growth of features from photoproducts substantially 
increases the match with the interstellar band. The laboratory spectra were smoothed to the resolution of 
the observations. Upper part. The laboratory spectra at their original resolution. 

2. Experimental 

Mixtures of astrophysically relevant molecules (H20/CO/NH3 = 5:2:1) were deposited 
on a cold finger (12 K) with simultaneous UV irradiation. About 0.2 UV photons were 
absorbed in the ice per molecule in this experiment. In the interior of molecular clouds this 
corresponds to about 106 years of irradiation (Prasad et al. 1983). Subsequently, the 
sample was slowly warmed up (0.5 K min-1), the changes being monitored In situ with an 
I.R. spectrometer. The refractory yellow residue remaining on the substrate at room 
temperature was investigated with Gas Chromatography / Mass Spectrometry. 


2. H2O/CO/NH3 = 0.63:0.25:0.12, + hu 







r ■■■ !■ 

1 ' 


1 . 1 . 

1 r—_ 











1 \ 



m \\ 





\\ 0.- — ~~^ 















' : 



, . 1 , 


1 ] - 

100 200 

T (K) 

Figure 2. The appearance and disappearance of species during warm-up of the photolysed sample 
H20/C0/NH3 = 5:2:1, monitored by following the depth of a characteristic I.R. feature (assignments from 
Hagen 1982, Schutte 1988 and Grim et ai. 1989; R denotes carbon chain or H atom), a. H20 (3300 cm- 
1); b. CO (2138 cm-1); c. 13C02 (2278 cm-1); d. H2CO (1499 cm-1); e. HCO (1848 cm-1); f. CH30H 
(1020 cm-1); g. HC00H/HC0NH2 (1686 cm-1); h. HOCH2CH20H (1030 cm-1); i. unidentified (1580 
cm-1); j. unidentified (2400 - 3350 cm-1); k. NH3 (1060 - 1200 cm-1); I. OCN- (2167 cm-1); m. NH4+ 
(1410 - 1520 cm-1); n. N03- (1388 cm-1); 0. NH4N03 (1330 cm-1); p. R-N02 (1540 cm-1); q. R3-C-H 
(2900 cm-1). 

3. Results. 

A large number of molecules are fornied in the laboratory sample during the photolysis 
and from reactive photoproducts during warm-up. The appearance and sublimation 
during the warm-up of the species detected by I.R. was monitored by following the 
intensity of a characteristic feature (fig. 2). Table 1 lists the organic species that were 
observed in the ice and in the non-volatile residue. The are composed of 1 - 3 C atoms 
with -NH2, -OH and -(C=0)- functional groups. A large number of volatile organic 
species can however have escaped detection due to the confusion of the numerous 
features in the I.R. spectrum. About 3% of the available carbon was converted to organic 
species that are volatile at room temperature and about 2.5% to residue material. 


Table 1. Photochemically produced organic species from deposited H20, 
CO and NHS (from Agarwal et ai. 1985, Schutte 1988). 

Molecule Mass (amu) volatile (at 293 K) 











































4. Comparison of the Experimental Results with Comet Halley 

Giotto's PICCA heavy-ion analyzer detected molecules rich in C, H, O 
and N in the gas near comet Halley (Mitchell et al. 1987, 1988). The 
spectra showed masses in the range 30-120 amu with a total 
abundance of the order of a few percent of that of the water ions. 
Furthermore, Vega's PUMA dust impact mass spectrometer tentatively 
identified species like formic acid (HCOOH) and oxalic acid (HOCOCOOH) 
in the cometary grain mantles (Kissel et al. 1987). 

There is a good correspondence between the elemental composition 
of the laboratory residue material and the elemental composition 
inferred from the PICCA mass spectra (table 2). Although the residue 
species are not expected to sublimate at the dust temperatures during 
the Giotto and Vega encounters (= 300 K; e.g., Tokunaga et al. 1989), the 
more volatile organic component produced in the experiment should 
have similar elemental composition since it is formed by similar 
chemical processes (Schutte 1988). The detected abundances are in 
good agreement with the amount of organic material that could form in 
interstellar ice mantles over astrophysical time scales. 


Table 2. Comparison between the elemental composition of the 
laboratory organic residue and the organic species detected by PICCA 
near Comet Halley. 

rel. abundance 










1 .2 - 1 .4 





We simulated the photoprocessing of interstellar ices in dense 
clouds in connection with the detection of organic species by PICCA in 
the coma of Comet Halley. The following conclusions can be drawn: 

The simulation experiments produce organic molecules of mass 

between 30 and 1 10 amu rich in -(C=0), -OH, and -NH2 molecular 


The elemental composition of the organic species detected by 

PICCA is similar to that of the experimentally produced 


The production efficiency of organics in molecular clouds could be 

sufficient to explain the observed abundance of organic species in 

the coma of Comet Halley. 


Agarwal, V. K., Schutte, W., Greenberg, J. M., Ferris, J. P., Briggs, R., 

Connor, S., van de Bult, C. P. E. M., and Baas, F. 1985, Origins of Life, 

Butchart, I., McFadzean, A. D., Whittet, D. C. B., Geballe, T. R., and 

Geballe, T. R., Baas, F., Greenberg, J. M., and Schutte, W. 1985, Astr. 

Ap. (Letters)A46, L6. 
Greenberg, J. M. 1986, Astr. Ap. (letters), 154, L5. 
Greenberg, J. M. 1973, in Molecules in the Galactic Environment, ed. 

Gordon and Snyder (Wiley Interscience), p. 94. 
Grim, R. J. A., and Greenberg, J. M. 1987, Ap. J., 321, L91. 


Grim, R. J. A., Schutte, W. A., Greenberg, J. M., Baas, F., and Schmitt, B. 

1989, Astr. Ap., in press. 
Hagen 1982, Ph.D. thesis, University of Leiden, the Netherlands. 
Kissel, J., and Krueger, F. R. 1987, Nature, 326, 755. 
Lacy, J. H., Baas, F., Allamandola, L. J., Persson, S. E., McG/egor, P. J., 

Lonsdale, C. J., Geballe, T. R., and van de Bult, C. E. P. 1984, Ap. J., 

276, 533. 
Mitchell, D. L, Lin, R. P., Anderson, K. A., Carlson, C. W., Curtis, D. W., 

Korth, A., Reme, H., Sauvaud, J. A., d'Uston, C, and Mendis, D. A. 1987, 

Science, 237, 626. 
Mitchell etal. 1988, preprint. 

Prasad, S. S., and Tarafdar, S. P. 1983, Ap. J., 267, 603. 
Schutte 1988, Ph.D. thesis, University of Leiden, the Netherlands. 
Tielens et al., 1989, Ap. J., in preperation. 
Tokunaga, A. T., Golisch, W. F., Griep, D. M., Kaminski, C. D., and Manner, 

M. S. 1988, Astron. J., 96, 1971 . 



Harry Y. McSween, Jr. 

University of Tennessee, Knoxville 

Knoxville, Tennessee 



Harry Y. McSween, Jr. 
University of Tennessee, Knoxville 


"More than ever, comets appear to be made of a pristine material older 
than the planets, preserved in its primitive state by the very deep cold 
of interstellar space and able to give us information about the chemistry 
of the early solar nebula." (Delsemme, 1988) 

Much of the excitement about obtaining cometary samples accrues from 
the conventional view, expressed eloquently in the statement above, that 
they comprise the most primitive materials that we are likely to get our 
hands on. But is this true? Although "parent body" alteration of such 
samples would not necessarily detract from this interest, we should keep 
in mind the possibility that certain kinds of secondary processes may have 
affected cometary nuclei. Weissraan (1986) has proposed some mechanisms by 
which comet nuclei might be altered, but observational evidence supporting 
the physical processing of comets is not yet generally available. This 
paper will take another approach: inferences about the kinds of 
modifications that might be encountered can be drawn from data on the 
evolution of carbonaceous chondrite parent bodies. The following 
observations suggest that, of all the classes of chondrites, these 
meteorites are most applicable to the study of comets: 

(1) Carbonaceous chondrites are chemically the most primitive 
meteorites. The elemental abundances of CI chondrites, normalized to 
silicon, provide the closest match with the composition of the solar 
photosphere (Holweger, 1977; Anders and Grevesse, 1989). 

(2) Spectral reflectivity surveys of asteroids suggest that 
carbonaceous chondrite- like bodies reside primarily in the more distal 
portions of the asteroid belt (Gradie and Tadesco, 1982). Their formation 
locations thus lie at greater solar distances than those of other 
meteorite types, closer to inferred sites for comet accretion (Weissman, 
1985; Hartmann et al., 1987). 

(3) Petrographic studies of carbonaceous chondrites indicate that they 
formed in volatile-rich environments (McSween, 1979; Kerridge and Bxinch, 
1979), and H2O and other volatile components may have been incorporated 


initially as ices (Bunch and Chang, 1980; Prinn and Fegley, 1988). 

(4) Some types of chondritic porous interplanetary dust particles 
(IDPs), which may be solid debris from short period comets, are 
mineralogically similar to carbonaceous chondrites (Bradley and Brownlee, 
1986; Tomeoka and Buseck, 1988), although some compositional distinctions 
occur (Schramm et al., 1987). Moreover, both of these materials appear to 
be broadly similar in composition to Comet Halley dust (Rietmeijer, 1987; 
Blanford et al., 1988). * 

It seems increasingly unlikely that carbonaceous chondrites are comet 
nucleus samples. However, these meteorites were probably derived from 
planetesimals that originally contained ices, though possibly in lesser 
proportions than comets, so the compositional distinction between 
carbonaceous chondrite parent bodies and comets may be one of degree. The 
possibility of an orbital evolution of cometary bodies into asteroidal 
orbits has also been suggested (Wetherill, 1979). For these reasons, it 
seems prudent to examine the processes which have affected carbonaceous 
chondrite parent bodies as possible analogs for the evolutionary history 
of comets . 


Most carbonaceous chondrites show evidence of parent body heating, 
either in the form of thermal metamorphism or, more commonly, aqueous 
alteration (Zolensky and McSween, 1988). Although aqueous alteration 
clearly took place at low temperatures near the freezing point of water 
(Clayton and Mayeda, 1984), heat was necessary to produce water from ice. 
Aqueous alteration at low temperatures may also have affected some IDPs 
(Reitmeijer and Mackinnon, 1985a). 

The source of heat for chondrite parent bodies is still 
controversial. Decay of short-lived radionuclides like "^"Al is one 
plausible mechanism (Lee et al, 1977; Hutcheon et al. , 1987) that could 
presumably affect asteroids and comets. External heat sources such as 
electromagnetic induction by a massive solar wind (Herbert and Sonett, 
1979) have also been suggested, but the decrease in effectiveness of this 
mechanism with solar distance renders this heat source unlikely for 
cometary bodies. The time scale for aqueous alteration in carbonaceous 
chondrites (Macdougall et al., 1984) was fairly short and commenced soon 
after accretion, as would be appropriate for either heating mechanism. 

Melting of ice in carbonaceous chondrite parent bodies has resulted 
in profoxind mineralogical changes. The original (presumed anyhydrous) 
chondrite assemblage has been altered to nonequilibrixom mixtures of 
fine-grained phyllosilicate minerals like serpentine, smectite, and 
chlorite, as well as poorly crystallized oxides, hydroxides, sulfides, 
carbonates, and carbonaceous phases (e.g. Barber, 1981; Mackinnon, 1982; 


Tomeoka and Buseck, 1985, 1988). High-resolution transmission electron 
microscopy has revealed intimate intergrowths of complex phases (Fig. 1) 
that are extremely difficult to characterize. CI chondrites are cut by 
veins containing sulfate and carbonates (Richardson, 1978) precipitated 
from fluids of differing compositions (Fig. 2). 

Figure 1. HRTEM photomicrograph of tubular phyllosilicate phase with 5 
A lattice spacing in the Mighei CM carbonaceous chondrite. Two distinct 
core regions are indicated by A and B. This phase formed by aqueous 
alteration. From Tomeoka and Buseck (1983). 

Figure 2. Sulfate and carbonate veins crosscutting a thin section of 
the Orgueil CI carbonaceous chondrite. The section is approximately 10 nm 
wide. From McSween (1979). 

It is generally believed that chemical changes accompanying aqueous 
alteration were minor (McSween and Richardson, 1977), although this 
conclusion is based primarily on similarities to solar elemental 
abundances which are not precisely measured. However, observed 
modifications of the isotopic composition of oxygen (Clayton and Mayeda, 


198A), which constitutes nearly half of mass of these meteorites, suggests 
the likelihood of other chemical changes. Bunch and Chang (1980) noted 
that the strongest reported enrichments of the heavy isotopes of 0, N, and 
C occur in chondrites which have been exposed to aqueous alteration. 

Grimm and McSween (1987) constructed thermal models for ice-bearing 
planetesimals. Not surprisingly, the presence of ice was found to act as 
a thermal buffer for such bodies, possibly accounting for the difference 
in metamorphic history between parent bodies for ordinary and carbonaceous 
chondrites. More recent work suggests that exothermic hydration reactions 
may have significant thermal effects which should be included in thermal 
modeling. Grimm and McSween (1989) have developed an interior-alteration 
model driven by "^"Al decay and a regolith-alteration model powered by 
either "^^Al or impacts. Both models are capable of producing 
homogeneously altered materials in relatively short time spans. 


Because carbonaceous chondrites contain carbon in carbonates and 
organic compounds extractable by water and solvents , Bunch and Chang 
(1980) raised the possibility of organic synthesis during aqueous 
alteration. Fischer-Tropsch type reactions in the nebula or in 
interstellar clouds have been suggested to have produced many hydrocarbons 
in carbonaceous chondrites (Hayatsu and Anders, 1981), but other, 
secondary mechanisms are worthy of consideration. 

Organic compovmds are much more abundant in altered carbonaceous 
chondrites than in unaltered ones (Cronin et al., 1988). Pelzer et al. 
(1984) noted that the presence of both amino acids and hydroxy acids in 
these meteorites suggests formation by a Strecker-cyanohydrin synthesis in 
an aqueous, ammonia-containing medixom, which might be reasonably expected 
to accompany aqueous alteration. Thermodynamic calculations by Shock 
(1988) indicate that carboxylic and amino acids should be produced by 
reactions with water and ammonia at 25°C. The Miller-Urey synthesis, 
which is capable of producing many other organic compounds, is postulated 
to have taken place on parent body surfaces, although its link with 
aqueous altertion is tenuous. 

Poorly graphitized carbon is also present in carbonaceous chondrites 
and IDPs (Mackinnon and Rietmeijer, 1987). This material may either have 
been derived from, or alternatively, be the remnants of interstellar 
organic phases. The carbonization reaction involves removal of and N 
and the subsequent graphitization of hydrocarbons. Rietmeijer and 
Mackinnon (1985b) showed that this process could occur at temperatures of 
several hundred degrees or less. 



Virtually all classes of meteorites show the effects of impact 
processes, so it seems possible that cometary materials may also show 
shock effects. The results of impact phenomena in carbonaceous chondrites 
take many forms . 

Shock metamorphic effects include partial destruction of the crystal 
structures of many minerals, easily recognizable from their optical and 
X-ray diffraction properties. The breakup of chondrules to produce 
isolated mineral grains (Richardson and McSween, 1978) may have been 
facilitated by both shock and aqueous alteration. Uniaxial compaction and 
deformation of chondrules in a few carbonaceous chondrites have also been 
recognized (Cain et al., 1986). Although this has been attributed to 
overburden on the parent asteroid, similar strain effects in ordinary 
chondrites are clearly related to shock (Sneyd et al., 1988). 

Most, if not all, carbonaceous chondrites are breccias, containing 
clasts with variable alteration histories (Nagy, 1975; McSween and 
Richardson, 1977; Olsen et al., 1988). Man3^ of these formed in regoliths 
although some, such as that shown in Figure 3, may be accretional breccias 
(Kracher et al., 1985). Mixing of different kinds of materials within the 
outer parts of comets might also have occurred, even if impacts were 
infrequent . 

Figure 3. Photomicrograph of the Leoville CV carbonaceous chondrite, 
illustrating dark carbonaceous clast in lighter colored host. This 
meteorite is thought to be an accretionary breccia. The clast is 
approximately 2 mm across. From McSween (1977). 

Lange and Ahrens (1982) showed experimentally that shock can cause 
dehydration of serpentine and suggested that impacts into carbonaceous 
chondrite bodies may result in devolatilization. Loss of structural water 
from minerals in carbonaceous chondrites has been documented in at least 
one meteorite (Akai, 1988), although this was attributed to heating. 



Many chondrites have been irradiated by solar-wind, solar-flare, and 
cosmic-ray particles. The penetration depths for these particles vary 
with their energies (galactic cosmic rays 1 m, solar cosmic rays 1 mm, 
solar wind <1 lun), but none can penetrate to appreciable depths. For this 
reason, most chondrite irradiation occurred in regoliths on parent body 
surfaces or during exposure as small meteoroids in space. 

A high proportion of carbonaceous chondrites contain solar-wind 
implanted noble gases, as well as significant amounts of cosmogenic 
nuclides and solar flare tracks in mineral grains (Fig. A). All of these 
features are indicative of irradiation. More information on the 
multi-stage exposure histories of these meteorites can be obtained if 
compaction ages are available. For the few carbonaceous chondrites for 
which exposure histories have been measured, the solar-wind and 
solar-flare irradiations occurred before 4.2-4.4 b.y. ago (Macdougall and 
Kothari, 1976). Exposure may have occurred in small bodies that were 
precursors to the carbonaceous chondrite parent asteroids, in small 
fragments of disrupted parent bodies that subsequently reaccreted, or in 
asteroid megaregoliths (Goswami et al., 1984). 

Figure 4. Olivine crystal with tracks (enlarged by etching) on some 
faces, caused by irradiation. Arrow points to an included chromite grain. 
From Macdougall and Kothari (1976). 


How applicable, if at all, are these processes to comets? Based on 


our present lack of data on and detailed \inderstanding of cometary nuclei, 
we probably cannot afford to rule any of them out. 

Calculations of the thermal effect of decay of long-lived 
radionuclides in comets suggest that interior temperatures could reach 
above 50 K for 10-km-diameter objects (Yabushita and Wada, 1988). If the 
proportion of possible short-lived radionuclides such as Al in cometary 
materials were similar to those in chondrites, and if the time scale of 
comet accretion was fast enough to permit incorporation of "live" 
radionuclides, comets might have reached significantly higher 
temperatures. Rietmeijer (1985) has considered the possibility of 
cryogenic (<0°C) alteration in comet nuclei, owing to the presence of thin 
interfacial water layers. Even modest temperature excursions in cometary 
nuclei might result in thermal histories somewhat like those of 
carbonaceous chondrite parent bodies. We might then predict that cometary 
dust should contain some phyllosilicate minerals and other phases formed 
by aqueous alteration at low temperatures. At least some chondritic IDPs 
contain phyllosilicates and carbonates (Rietmeijer and Mackinnon, 1987), 
and aggregate IDPs may closely approach CI chondrites in bulk composition 
(Blanford et al., 1988). The occurrence of graphitized carbon in some 
IDPs suggests that these particles may have experienced temperatures of up 
to several hundred degrees (Rietmeijer and Mackinnon, 1985b). Based on a 
chemical comparison between Comet Halley dust and cosmic abundances, 
Anders and Grevesse (1989) suggested that non-solar Fe/Si and Mg/Si ratios 
indicate that the Halley dust component cannot be pristine interstellar 
matter. On the other hand, similarities between the chemical compositions 
of anhydrous IDPs and Comet Halley dust (Rietmeijer, 1987) may suggest 
that aqueous alteration processes have not appreciably affected this 
cometary nucleus. Detailed mineralogical characterization of returned 
comet samples may be required to resolve this question. 

Thermal effects in comet nuclei may also result from repeated close 
passage to the sun (Weissman, 1986), or near passing stars in the Oort 
cloud (Stem, 1986). Spectral observations from the Giotto mission 
indicate dust temperatures at the surface of comet Halley may have reached 
80°C at 1 A.U. (Gombosi and Houpis, 1986). These temperatures are 
certainly sufficient for aqueous alteration processes, provided that the 
kinetics of such reactions are reasonably rapid. The chemical and 
mineralogical diversity of chondritic IDPs suggests that their sources are 
heterogeneous or differ in degree of thermal processing (Mackinnon and 
Rietmeijer, 1987). If all IDPs are derived from comets, they may have 
different thermal histories because of their orbits or other factors. 

It is difficult to say anything about the origin of carbonaceous 
compounds in comets. Comet Halley shows the C-H stretch band at 3.4 
microns in emission, indicating the presence of organic matter on its 
nucleus (Combes et al., 1986), but specific compounds have not been 
identified. IDPs also show this C-H stretching feature (Sandford and 
Walker, 1985). Organic constituents in IDPs are still poorly 
characterized, though evidence of some thermal processing in the form of 
graphitized carbon is present (Rietmeijer and Mackinnon, 19885b). 
Understanding of carbon compounds in comets has been greatly improved by 


the recognition that much of this material resides in the dust fraction 
(Delsemme.. 1988), and return of a condensed sample may allow chemical and 
isotopic studies which address this problem. 

Impact processes probably affected cometary materials during their 
initial accretion. Low number densities of comets in the Oort cloud may 
have minimized subsequent collisions, but their ejection from the outer 
solar system into the Oort cloud may have been catastrophic. We should be 
prepared to find that comets are heterogeneous, consisting of rock 
fractions with different thermal and shock histories. In fact, the comet 
nucleus itself may ultimately be viewed as a megabreccia, comprised of 
rock and ice blocks and clasts. 

Suggestions that dusty regoliths may persist on cometary surfaces 
suggest that irradiation of cometary materials is likely. Conceptions of 
comet surfaces that envision refractory lag deposits or icy pedestals 
capped by dust also offer opportunities for sample irradiation, and the 
shielding characteristics of ice are less than that of rock. Beyond the 
heliopause, comets may be exposed to increased amounts of galactic cosmic 
rays. Moreover, progressive devolatilization of comets during passages 
close to the sun should expose increasing amounts of rocky material to 
solar radiation, and "extinct" comets might have irradiation histories 
similar to asteroids. It is noteworthy that cosmic-ray tracks have been 
observed recently in chondritic IDPs (Bradley and Brownlee, 1986), 
although these tracks were probably implanted during exposure to solar 
flares while the particles were in interplanetary orbits rather than on 
the parent objects. 

Comets may indeed turn out to be pristine materials, as stated in the 
quotation that opened this paper, but we are not yet positive of that. 
Their formation and storage at great solar distances does not 
automatically guarantee them immunity from the kinds of processes that 
have affected other solar system bodies, and repeated approaches to the 
inner solar system may offer other opportunities for alteration processes 
to affect those comets that we will be able to sample. 


Akai J. (1988) Incompletely transformed serpentine-type phyllosilicates in 
the matrix of Antarctic CM chondrites. Geochim. Cosmochim. Acta 52, 

Anders E. and Grevesse N. (1989) Abundances of the elements: Meteor itic 
and solar. Geochim. Cosmochim. Acta , in press. 

Barber D. (1981) Matrix phyllosilicates and associated minerals in C2M 
carbonaceous chondrites. Geochim. Cosmochim. Acta 45, 945-970. 


Blanford G.E., Thomas K.L, and McKay D.S. (1988) Microbeam analysis of 
four chondritic interplanetary dust particles for major elements, 
carbon and oxygen. Meteoritics 23, 113-121. 

Bradley J. P. and Brownlee D.E. (1986) Cometary particles: Thin sectioning 
and electron beam analysis. Science 231, 15A2-1544. 

Bunch T.E. and Chang S. (1980) Carbonaceous chondrites-II. Carbonaceous 
chondrite phyllosilicates and light element geochemistry and indicators 
of parent body processes and surface conditions. Geochim. Cosmochim. 
Acta 4A, 1543-1577. 

Cain P.M., McSween H.Y. , and Woodward N.B. (1986) Structural deformation 
of the Leoville chondrite. Earth Planet. Sci. Lett. 77, 165-177. 

Clayton R.N. and Mayeda T.K. (1984) The oxygen isotope record in Murchison 
and other carbonaceous chondrites. Earth Planet. Sci. Lett. 67, 

Combes M. and 15 coauthors (1986) Detection of parent molecules in comet 
Hal ley from the VEGA experiment. In Proc. 20th ESLAB Symp. on the 
Exploration of Halley's Comet , vol. 1, E3A SP-250, pp. 353-358. 

Cronin J.R., Pizzarello S., and Cruikshank D.P. (1988) Organic matter in 
carbonaceous chondrites, planetary satellites, asteroids and comets. In 
Meteorites and the Early Solar System , ed. J.F. Kerridge and M.S. 
Matthews, Univ. of Arizona Press, Tucson, pp. 819-857. 

Delsemme A.H. (1988) The chemistry of comets. Phil. Trans. R. Soc. Lond. 
A 325, 509-523. 

Gombosi T.I. and Houpis H.L.F. (1986) An icy-glue model of cometary 
nuclei. Nature 324, 43-44. 

Goswami J.N. , Lai D., and Wilkening L.L. (1984) Gas-rich meteorites: 
Probes for particle environment and dynamical processes in the inner 
solar system. Space Sci. Rev. 37, 111-159. 

Gradie J.C. and Tedesco E.F. (1982) The compositional structure of the 
asteroid belt. Science 216, 1405-1407. 

Grimm R.E. and McSween H.Y. (1987) Water and the thermal history of the CM 
carbonaceous chondrite parent body. Lunar Planet. Sci. XIX , 427-428. 

Grimm R.E. and McSween H.Y. (1989) Water and the thermal evolution of 
carbonaceous chondrite parent bodies. Icarus , submitted. 

Hartmann W.K. , Tholen D.J., and Cruikshank D.P. (1987) The relationship 
between active comets, "extinct" comets, and dark asteroids. Icarus 69, 

Hayatsu R. and Anders E. (1981) Organic compounds in meteorites and their 


origins. In Topics in Current Chemistry Vol. 99, ed. F.L. Boschke, 
Springer, 1-37. 

Herbert F. and Sonett C.P. (1979) Electromagnetic heating of minor planets 
in the early solar system. Icarus AO, 48A-496. 

Kolweger H. (1977) The solar Na/Ca and S/Ca ratios: A close comparison 
with carbonaceous chondrites. Earth Planet. Sci. Lett. 34, 152-154. 

Hutcheon I.D., Hutchison R., and G.J. Wasserburg (1987) Evidence of the 

in-situ decay of '^"Al in a Semarkona chondrule. Lunar Planet. Sci. XIX, 

Kerridge J.F. and Bunch T.E. (1979) Aqueous activity on asteroids: 

Evidence from carbonaceous meteorites. In Asteroids , ed. T. Gehrels, 
Univ. of Arizona Press, Tucson, pp. 745-764. 

Kracher A., Keil K. , Kallemeyn G.W., Wasson J.T., Clayton R.N., and Huss 
G.I. (1985) The Leoville (CV3) accretionary breccia. J . Geophys . Res . 
90 Suppl., D123-126. 

Lange M.A. and Ahrens T.J. (1982) Shock release adiabat measurements on 
volatile bearing minerals and implications for an impact generated 
atmosphere. Lunar Planet. Sci. XIII , 421-422. 

Lee T. , Papanastasiou D.A., and Wasserburg G.J. (1977) Aluminum-26 in the 
early solar system: Fossil or fuel? Astrophys. J. 211, L107-110, 

Macdougall J.D. and Kothari B.K. (1976) Formation chronology for C2 
meteorites. Earth Planet. Sci. Lett. 33, 36-44. 

Macdougall J.D., Lugmair G.W., and Kerridge J.F. (1984) Early solar system 
aqueous activity: Sr isotope evidence from the Orgueil CI meteorite. 
Nature 307, 249-251. 

Mackinnon I.D.R. (1982) Ordered mixed- layer structures in the Mighei 
carbonaceous chondrite matrix. Geochim. Cosmochim. Acta 46, 479-489. 

Mackinnon I.D.R. and Rietmeijer F.J.M. (1987) Mineralogy of chondritic 
interplanetary dust particles. Rev . Geophys . 25, 1527-1553. 

McSween H.Y. (1977) Petrographic variations among carbonaceous chondrites 
of the Vigarano type. Geochim. Cosmochim. Acta 41, 1777-1790. 

McSween H.Y. (1979) Are carbonaceous chondrites primitive or processed? A 
review. Rev. Geophys. Space Phys. 17, 1059-1078. 

McSween H.Y. and Richardson S.M. (1977) The composition of carbonaceous 
chondrite matrix. Geochim. Cosmochim. Acta 41, 1145-1161. 

Nagy B. (1975) Carbonaceous Meteorites. Elsevier Pub. Co., Amsterdam. 


Olsen E.J., Davis A.M., Hutcheon I.D., Clayton R.N., Mayeda T.K., and 
Grossman L. (1988) Murchison xenoliths. Geochim. Cosmpchim. Acta 52, 

Pelzer E.T., Bada J.L., Schlesinger G. , and Miller S.L. (1984) The 

chemical conditions on the parent body of the Murchison meteorite: Some 
conclusions based on amino, hydroxy and dicaroxylic acids. Adv. Space 
Res. A, 69-74. 

Prinn R.G. and Fegley B. (1988) Solar nebula chemistry: Origin of 

planetary, satellite, and cometary volatiles. In Origin and Evolution 
of Planetary and Satellite Atmospheres , ed. 3. Atreya, J. Pollack, and 
M. Matthews, Univ. of Arizona Press, Tucson, in press. 

Richardson S.M. (1978) Vein formation in the CI carbonaceous chondrites. 
Meteoritics 13, 141-159. 

Richardson S.M. and McSween H.Y. (1978) Textural evidence bearing on the 
origin of isolated olivine crystals in C2 carbonaceous chondrites. 
Earth Planet. Sci. Lett. 37, 485-491. 

Rietmeijer F.J.M. (1985) A model for diagenesis in proto-planetary bodies. 
Nature 313, 293-294. 

Rietmeijer F.J.M. (1987) A quantitative comparison of fine-grained 

chondritic interplanetary dust and Comet Halley dust. Liinar Planet. 
Sci. XIX , 980-981. 

Rietmeijer F.J.M. and Mackinnon I.D.R. (1985a) Layer silicates in a 
chondritic porous interplanetary dust particle. J. Geophys. Res. 90 
Suppl., D149-155. 

Rietmeijer F.J.M. and Mackinnon I.D.R. (1985b) Poorly graphitized carbon 
as a new cosmothermometer for primitive extraterrestrial materials. 
Nature 315, 733-736. 

Sandford S.A. and Walker R.M. (1985) Laboratory infrared transmission 
spectra of individual interplanetary dust particles from 2.5 to 25 
microns. Astrophys . J . 291, 838-851. 

Schramm L.S., Brownlee D.E., and Wheelock M.M. (1987) The elemental 

composition of interplanetary dust. Lunar Planet. Sci. XIX , 1033-1034. 

Shock E.L. (1988) Relative abundances of amino acids in the Murchison 
meteorite: Clues to synthesis pathways or sampling bias? EOS 59, No. 
44, 1287. 

Sneyd D.S., McSween H.Y. , Sugiura N. , Strangway D.W., and Nord G.L. (1988) 
Origin of petrofabrics and magnetic anisotropy in ordinary chondrites. 
Meteoritics 23, 139-149. 

Stern S.A. (1986) Cometary capture rates and extra-solar Oort cloud 


encounters. Lunar Planet. Sci. XVIII , 950. 

Tomeoka K. and Buseck P.R. (1983) A new layered mineral from the Mighei 
carbonaceous chondrite. Nature 306, 354-356. 

Tomeoka K. and Buseck P.R. (1985) Indicators of aqueous alteration in CM 
carbonaceous chondrites: Microtextures of a layered mineral containing 
Fe, S, and Ni. Geochim. Cosmochim. Acta 49, 2149-2163. 

Tomeoka K. and Buseck P.R. (1988) Matrix mineralogy of the Orgueil CI 
carbonaceous chondrite. Geochim. Cosmochim. Acta 52, 1627-1640. 

Weissman, P.R. (1985) The origin of comets: Implications for planetary 
formation. In Protostars and Planets II , ed. D.C. Black and M.S. 
Matthews, Univ. of Arizona Press, Tucson, 895-919. 

Weissman, P.R. (1986) How pristine are cometary nuclei? Proc . Comet 
Nucleus Sample Return Mission Workshop, ESA SP-249 , pp. 15-25. 

Wetherill, G.W. (1979) Steady-state populations of Apollo-Amor objects- 
Icarus 37, 96-112. 

Yabushita S. and Wada K. (1988) Radioactive heating and layered structure 
of cometary nuclei. Earth Moon Planet. 40, 303-313. 

Zolensky M. and McSween H.Y. (1988) Aqueous alteration. In Meteorites and 
the Early Solar System , ed. J.F. Kerridge and M.S. Matthews, Univ. of 
Arizona Press, Tucson, pp. 114-143. 



A. C. Levasseur-Regourd 

Service d'Aeronomie 
Vairieres, France 

R. Dumont 

Observatoire de Bordeaux 

Hoirac, France 

J. B. Renard 

Service d'Aeronomie 

Varrieres, France 




1 . Service d'A6ronomie, Verri§res, France 2. Observatoire de Bordeaux, Floirac, France 

Both the solar light scattered by 
interplanetary dust particles and the thermal 
emission from these particles have been 
extensively observed. Techniques of inversion of 
the line-of-sight brightness allow to derive local 
optical properties of the interplanetary dust. 

In the ecliptic plane near 1 au, the 
heliocentric gradients of local polarization, 
temperature and albedo are found to be of the 
order, respectively, of 0.8, - 0.3, and - 0.7. 
Towards the ecliptic pole, the local polarization is 
found to be much smaller than in the ecliptic at 
the same solar distance. Also the temperature 
decreases faster with increasing distance than it 
does in the ecliptic, whence an increase in bulk 

Some optical properties of the dust particles 
may be inferred from these results. They are 
consistent with a scenario of evolution of dark 
fluffy cometary grains evaporating as they spiral 
towards the Sun. 


Zodiacal light 

Our present knowledge of the distribution of 
brightness and polarization of light scattered by 
interplanetary dust particles (= zodiacal light), 
as seen from the Earth's orbit, comes from 
extended ground based programmes, together with 
space observations. Zodiacal light is a faint 
extended source, which is collected along the line 
of sight together with atmospheric airglow 
(negligible for space observations), starlight, 
galactic light and diffuse galactic light 
(negligeable at high galactic latitudes). The 
results are usually given as a function of ecliptic 
latitude p and helioecliptic longitude X - XqoI the 


Line c' sight 




Edipbe (or 
symmetry) plane 

Fig. 1. Geometry of ID P observations 

line of sight (fig. 1). There is a fairly good 
agreement between the observations published 
since the mid seventies (see for instance 
compilations by Levasseur and Dumont, 1980 or 
Fechtig et al., 1981). 

The main feature for the zodiacal light 
brightness in an increase towards the Sun, 
together with a secondary increase due to some 
backscattering in the antisolar region 
(gegenschein). The zodiacal light is found to be 
rather stable with time, but for some annual 
temporal variations at high and medium ecliptic 
latitudes ; these oscillations originate in the 
slight inclination (about 1.5°) of the symmetry 
surface of the zodiacal cloud upon the ecliptic 

The polarization distribution is characterized 
by a region of strong linear polarization near 60° 
elongation. The Fresnel vector is in the scattering 
plane (observer-Sun, observer-line of sight), 
but for the antisolar region which exhibits a 
negative polarization. 


Thermal emission 

With a spectrum that is almost solar like, 
zodiacal light prevails below 5 nm. In the 5 to 50 
^im range, the thermal emission of the 
interplanetary dust cloud turns out to be the most 
prominent component of the sky, at least for high 
and medium galactic latitudes. 

Thermal emission has been observed at 
various wavelengths since the beginning of the 
eighties from balloons or rockets and from IRAS 
satellite (Hauser et al., 1984 ; Murdock and 
Price, 1985 ; Salama et al., 1987). 

Due possibly to calibration problems, there is 
a discrepancy by a factor of about 2 between the 
satellite and rocket emissivities, while the 
thermal emission is most likely to be stable with 
time. The survey performed by IRAS, even though 
limited to elongation angles e in a 60° to 120° 
range, clearly confirms that the zodiacal cloud 
symmetry surface is slightly inclined upon the 
ecliptic {Dumont and Levasseur-Regourd, 1978 ; 
Hauser et al., 1985) and that it is not entirely 
smooth, due to nan-ow dust trails ahead or behind 
cometary nuclei and to dust bands of asteroidal 
origin (Levasseur and Blamont, 1976 ; Sykes et 
al., 1986 ; Dermott et al.. 1986). 

Need for an inversion 

The measurements of both zodiacal light and 
thermal emission provide integrals along a line of 
sight of local brightnesses. To interpret these 
measurements in terms of local properties, it is 
necessary to develop model fitting or, even 
better, inversion methods. 

We present here new results for the "nodes of 
lesser uncertainty" inversion method, which may 
be of interest to derive some physical properties 
of the interplanetary dust and of its cometary or 
asteroidal sources. 


Nodes of lesser uncertainty on a secant 
to the Earth's orbit 

Once surveys of zodiacal light or of thermal 
emission are available in the ecliptic plane, 



nodaa 1 

Sun omn 



r ^-^>x/ 



: y' Ml 



Fig. 2. Nodes in the ecliptic plane 

integrated brightnesses can be obtained along the 
same line of sight (that intersects the Earth's 
orbit in Mi and M2) for a given helioediptic 
longitude and for the supplementary angle (fig. 
2). The knowledge of integrated brightnesses 
along Mi «> and M2 ~ offers some possibility 
of partly inverting the brightness integral under 
limited assumptions. 

The inversion is indeed feasible without any 
assumption only for the first section (from 
where the observation is performed) of a line of 
sight tangent to the orbit of the Earth (or of a 
moving spacecraft). For any other location and 
direction, various assumptions are required. 

We assume here a steady state and a rotational 
symmetry of the interplanetary dust cloud, as 
strongly suggested by the absence of seasonal 
dependence of optical and infrared brightnesses 
after correction for the oscillations of the Earth 
on either side of the symmetry plane. The local 
optical brightness can therefore be written as F x 
Cvis (R®. ©y R©^. where F is the solar intensity 
and where Cvis (directional scattering cross- 
section of the unit volume) is only a function of 
the heliocentric distance R© and the scattering 
angle 6. Also the local infrared brightness can be 
written as F x C|r(R©,X)/R©2 where C| r 
(monochromatic thermal cross-section of the 
unit volume) is only a function of the distance R© 
and the wavelength X. 

We also assume the functions that 
characterize the scattering or thermal cross- 
sections to be positive (optically thin medium) 
and rather monotonous (relative smoothness of 
the cloud, at least to the first order), and to 
decrease asymptotically to zero with increasing 
R© (absence of interplanetary dust far away from 
the Sun). From the whole of set of constraints 
(integrals along Mi «= and M2 °° lines of sight, 
plus values and derivatives at infinity), it can be 
demonstrated that the curves which represent all 
the possible Cyjs or C|r functions for a given line 


of sight have to constrict in two foci or nodes, 
where the local scattering or thermal cross- 
section of the unit volume can be determined with 
less uncertainty than elsewhere. Details on the 
method can be found in Dumont and Levasseur- 
Regourd (1985a) for the optical case or in 
Dumont and Levasseur-Regourd (1988) for the 
infrared case. 

One of the nodes, always located at an 
heliocentric distance of the order 1 .5 au, is called 
the martian node. In the visual case, the 
determination of Cvis (Ro = 1-5 ua, 9) at the 
martian node allows to disregard the heliocentric 
dependence and therefore to retrieve the 
scattering phase function near 1.5 ua. 

The other node, which remains located near 
the middle of M1M2 chord, is called the quasi 
radial node. In the visual case, the determination 
of Cvis(Ro. 6= 90°) at this node allows to 
disregard the scattering angle dependence and to 
retrieve the scattering cross-section at = 90°, at 
least in the 1 to 0.5 au range. 

In the infrared case, the simultaneous 
determinations of C|r (R©, A.i ) and C|r(Rs, X2) at 
the quasi radial node, together with the values of 
the thermal cross-section for the same two 
wavelengths at the martian node, provide (with a 
grey-body assumption) the radial dependence of 
the local colour temperature. 

Decrease of polarization degree with 
decreasing heliocentric distance 

The nodes of lesser uncertainty method allows 
to derive local parallel and perpendicular 
scattering cross-sections from the measured 
zodiacal light polarized components (parallel and 
perpendicular to the scattering plane, i.e. to the 
ecliptic plane). The local polarization degree is 
then computed from the two local scattering 

The inversion at the martian node provides the 
evolution of the local polarization degree at 1 .5 au 
from the Sun as a function of the scattering (or 
phase) angle (Dumont and Levasseur-Regourd, 
1985b). It may be of interest to mention that the 
smallest values on the negative branch and the 
slope at inversion compare quite well with the 
cometary data, as presented in Dollfus et al. 

The inversion at the quasi-radial node 
provides the evolution of the local polarization 


Fig. 3. Solar dependence of local polarization 

degree at 90° scattering angle as a function of the 
heliocentric distance. As can be seen on fig. 3, it 
drops from = 30% at the Earth's level to = 20% 
at 0.5 au. Such a decrease is in excellent 
agreement with Helios 1 and 2 measurements 
(Leinert, 1975). 

This result clearly demonstrates that the 
interplanetary dust particles properties depend 
upon their heliocentric distance. It is most likely 
that their surfaces are quite rough (negative 
branch in polarization) and that their porosity 
decreases with decreasing heliocentric distance. 

Increase of albedo with 
decreasing heliocentric distance 

As previously mentionned, the inversion 
method also allows to derive local temperatures 
from the thermal emissions measured at two 
wavelengths. There is indeed a discrepancy (even 
after correction for the effective bandpasses) in 
local temperatures as deduced from IRAS or 
rocket observations, which reflects the 
discrepancy in the raw data. However, the 
agreement on the gradient of temperature 
with heliocentric distance (-0.35±0.05) is 
excellent. It should be noted that the deviation of 
the gradient from the value that could be expected 
in a grey-body case (-0.5) is likely to originate 
in the distinct heliocentric changes of C|r and 


Once both the thermal energy reemitted by 
unit volume in the infrared and the energy 
scattered by unit volume in the optical domain are 
known, the local albedo can be computed. The 
former is deduced from the local temperature and 
from a monochromatic thermal cross-section 
C|R. The latter is precisely the directional 
scattering cross-section Cvis which, from 
zodiacal light observations, is mostly derived at 


90° scattering angle. The albedo at 90°, as defined 
by Manner et al. (1981) by analogy with the 
geometric albedo, is therefore accessible for 
various heliocentric distances. 

The gradient of albedo is negative from all 
observational data, in agreement with our 
preliminary analyses (Levasseur-Regourd and 
Dumont, 1985) and with similar suggestions 
made, from very different approaches, by Fechtig 
(1984) or Lumme and Bowel! (1985). Fom IRAS 
survey, the local albedo at 90° scattering angle is 
found to be of the order of 0.7 Ro"0-7. This result 
is in good agreement with the evolution of local 
polarization previously mentioned, since 
increases in albedo are usually correlated with 
decreases in polarization on various planetary 
surfaces (Dollfus, 1985). 

Once both the local temperature and albedo are 
obtained as a power law function of heliocentric 
distance, an empirical law of evolution of the bulk 
albedo A versus the local colour temperature T is 
derived in the ecliptic plane. IRAS leads to A = 
10-6 X t2. 

These previous results demonstrate that the 
in-ecliptic dust is most likely to originate in 
rough, porous and dark grains. When the 
particles spiral towards the Sun under Poynting- 
Robertson effect, their roughness, porosity, 
darkness and size may decrease, possibly because 
of some breaking off and evaporation due to 
sublimation and sputtering. 


Nodes of lesser uncertainty towards the 
ecliptic poles 

As well as the ecliptic case, various 
constraints are imposed to the polarized 
components of the local optical brightness and to 
the monochromatic thermal brightnesses. The 
integrals towards the ecliptic pole are known, 
together with the local values at the Earth (found 
by differenciation of the measurements 
performed in the ecliptic plane) and the 
derivatives at the Earth (assumed to be equal to 
zero for symmetry reasons). Also, the local 
values and their derivatives at infinity are equal 
to zero (absence of dust and flattened dust cloud). 



line of signi 

Ecliptic plarw 

Fig. 4. Observations towards the ecliptic pole 

It has been already noticed (Levasseur- 
Regourd and Dumont, 1988) that all the curves 
which represent the local functions have to be 
constricted at a node of lesser uncertainly, the 
ecliptic latitude of which is of the order of 20° 
(fig. 4). A partial inversion can therefore be 
performed at this node for the visual local 
brigthnesses and polarization, for the thermal 
local brightnesses and temperature, and 
ultimately for the local albedo. 

Smaller polarization 
out of the ecliptic plane 

From Dumont and Sanchez (1975) optical 
observations, the smallest uncertainty is obtained 
for the ecliptic latitude 21°, i.e. for a nodal 
scattering point at 1 .07 au from the Sun and at 
0.36 au above the Earth and the ecliptic plane. 
The corresponding local polarization at 111° 
scattering angle is found to be of the order of 9%. 
Comparable results are obtained from Fechtig et 
al.(1981) compilations, with a local polarization 
at 110° scattering angle of the order of 11%. 

Since the polarization curves (as measured on 
different interplanetary samples, or deduced 
from various modellings) are usually flat in the 
80° to 110° scattering angles range, these 
results can be compared to the values obtained in 
the ecliptic plane at 90° scattering angle. At the 
same heliocentric distance, the local polarization 
is equal to 30.5%, much greater than at the nodal 
point towards the ecliptic pole. 

This result demonstrates that the 
interplanetary dust particles found at 0.35 au 
towards the pole are different from the grains 
observed at 1 au in the ecliptic plane. It is likely 
that their porosity is significantly smaller. 


Larger albedo out of the ecliptic 



BETA ANGLE ( deg. ) 

BETA ANGLE ( deg. ) 

Fig. 5. Nodes for thermal emission 
towards the pole 

As can be seen on fig. 5, the same inversion 
method is used to derive local thermal 
brightnesses at various wavelengths, whence a 
local colour temperature. From IRAS data, the 
lesser uncertainty on the temperature is obtained 
for the ecliptic latitude p© = 18°, i.e. for a nodal 
point at 1.05 au from the Sun and at 0.32 au 
above the Earth's orbit. The temperature has to be 
in a 195-240 K range, much smaller than it 
would be in the ecliptic plane at the same 
heliocentric distance. 

From both the visual and thermal 
brightnesses, a local albedo near p© = 20" is 
derived for 110° scattering angle. As could be 
suspected from the steep decrease of temperature, 
this albedo, with a value of the order of 11%, is 
greater than in the ecliptic plane at the same 
solar distance. 

As noticed previously, the decrease in the 
local polarization goes together with an increase 
in bulk albedo. The dust particles found at 0.35 au 
above the Earth's orbit are less porous or dark 
and, maybe, smoother and smaller than in the 
ecliptic at the same heliocentric distance. 


The optical properties of the interplanetary 
dust grains are found to strongly depend upon 
their location in the zodiacal cloud. They are 
consistent with a scenario of evolution of dark 
porous cometary grains evaporating as they 
spiral towards the Sun and, later, being 
isotropically pushed away as small light p 
meteoroids under the solar radiation pressure. 
However, our purpose here is not to propose 
speculative interpretations of observational 
results, but rather to present some constraints 
on polarization and albedo of cometary dust, and to 
emphasize the evolution of porosity and fluffyness 
of. cometary grains as their age and surface 
temperature increases. 


Asteroids, comets, meteors II, ed. C.I. Lagerkvist 
et al., HSC, Uppsala, 583-594,1986 
DOLLFUS, A., Photopolarimetric sensing of 
planetary surfaces. Adv. Space Res., 5, 8, 47-58 

polarimetry of P/Halley : synthesis of the 
measurements in the continuum, Astron. 
Astrophys., 206, 348-356,1988 
Zodiacal light gathered along the line of sight ; 
retrieval of the local scattering coefficient from 
photometric survey of the ecliptic plane. Planet. 
Space Sci., 33, 1-9, 1985 a 
Remote sensing of the zodiacal cloud along secants 
to Earth's orbit, in : Properties and interactions 
of interplanetary dust, ed. R.H. Giese and P. Lamy, 
D. Reidel, Dordrecht, 207-213 ,1985 b 


Properties of interplanetary dust from infrared 
and optical observations I, Astron. Astrophys., 
191. 154-160. 1988 

Zodiacal light photopolarimetry II, Astron. 
Astrophys., 38,405-412.1975 
FECHTIG. H..The interplanetary dust environment 
beyond 1 au. and in the vicinity of the ringed 
planets. Adv. Space Res. 4, 9, 5-11, 1984 
planetary dust and zodiacal light, in Landolt- 
Bornstein, Springer, VI, 2a, 228-243, 1981 
R., On the definition of albedo and application to 
irregular particles, Astron. Astrophys., 104, 

WALKER, R.G., IRAS observations of the diffuse 
infrared background. Astrophys. J., 278, L15- 

HAUSER, M.G., Models for infrared emission from 
zodiacal dust, in : Comets to cosmology, ed. A. 
Lawrence, Springer-Verlag, Berlin, 27- 

LEINERT, C. Zodiacal" light - a measure of the 
interplanetary environment, Space Sci. Rev., 18, 

Evidence for scattering particles in meteor 
streams, in : Interplanetary, dust and zodiacal 
light, ed. H. Elsasser and H. Fechtig. Springer- 
Verlag, Berlin. 58-62,1976 
Absolute photometry of zodiacal light, Astron. 
Astrophys. 84, 277-279,1980 
Determination des temperatures locales et de 
gradient d'alb6do dans le nuage zodiacal ^ partir 
des donn^es radiom6triques d'IRAS, Compt. Rend. 
Acad. Sci. Paris, 300, II. 109-112.1985 
IRAS observations and local observations of 
interplanetary dust. Adv. Space Research, to be 
published, 1988 

LUMME K and BOWELL, E., Photometric 
properties of zodiacal light particles, Icarus, 
62, 54-71,1985 

MURDOCK, T.L. and PRICE, S.D.. Infrared 
measurements of zodiacal light, Astron. J. 90, 

SHIVANANDAN, K., Measurements of near and far 

infrared zodiacal dust emission, Astron. J. 92, 



LOW, F.J., The discovery of dust trails in 

comets, Science, 232, 1115-1117,1986 



Frans J. M. Rietmeijer 

Department of Geology 

University of New Mexico 

Albuquerque, New Mexico 

Ian D. R. Mackinnon 

Electron Microscope Centre 

University of Queensland 

St. Lucia, QLD 4067, Australia 



Frans J. M. Rietmeijer, 

Department of Geology, University of New Mexico, Albuquerque, NM 87131, USA 


Ian D. R. Mackinnon, 

Electron Microscope Centre, University of Queensland, St. Lucia, QLD 4067, Australia. 


The degree of diversity or similarity detected in comets depends primarily on the lifetimes of the 
individual cometary nuclei at the time of analysis. It is inherent in our understanding of cometary orbital 
dynamics [Weissman, 1985] and the seminal model of comet origins by Oort [Oort, 1950] that cometary 
evolution is the natural order of events in our Solar System. Thus, predictions of cometary behaviour in 
terms of bulk physical, mineralogical or chemical parameters should contain an appreciation of temporal 
variation(s). Previously, Rietmeijer and Mackinnon [1987] developed mineralogical bases for the chemical 
evolution of cometary nuclei primarily with regard to the predominantly silicate fraction of comet nuclei. 
We suggested that alteration of solids in cometary nuclei should be expected and that indications of likely 
reactants and products can be derived from judicious comparison with terrestrial diagenetic environments 
which include hydrocryogenic and low- temperature aqueous alterations. In a further development of this 
concept, Rietmeijer [1988] provides indirect evidence for the formation of sulfides and oxides in comet 
nuclei. Furthermore, Rietmeijer [1988] noted that timescales for hydrocryogenic and low-temperature 
reactions involving liquid water are probably adequate for relatively mature comets, e.g. P/comet Halley. 

In this paper, we will address the evolution of comet nuclei physical parameters such as solid particle 
grain size, porosity and density. In natural environments, chemical evolution (e.g. mineral reactions) is 
often accompanied by changes in physical properties. These concurrent changes are well-documented in 
the terrestrial geological literature, especially in studies of sediment diagenesis [Berner, 1980] and we 
suggest that similar basic principles apply within the upper few meters of active comet nuclei. 

The database for prediction of comet nuclei physical parameters is, in principle, the same as used for 
the proposition of chemical evolution [Rietmeijer and Mackinnon, 1987]. We use detailed mineralogical 
studies of chondritic interplanetary dust particles (IDPs) as a guide to the likely constitution of mature 
comets traversing the inner Solar System. While there is, as yet, no direct proof that a specific sub-group 
or type of chondritic IDP is derived from a specific comet [Mackinnon and Rietmeijer, 1987], it is clear 
that these particles are extraterrestrial in origin [Bradley et al., 1988] and that a certain portion of the 
interplanetary flux received by the Earth is cometary in origin [Brownlee, 1985]. Two chondritic porous 
(CP) IDPs, sample numbers W7010A2 and W7029C1, from the Johnson Space Center Cosmic Dust 
Collection have been selected for this study of putative cometary physical parameters. This particular type 
of particle is considered a likely candidate for a cometary origin [Bradley et al., 1988] on the basis of 
mineralogy, bulk composition and morphology. While many IDPs have been subjected to intensive study 
over the past decade, we can develop a physical parameter model on only these two CP IDPs because few 
others have been studied in sufficient detail [Mackinnon and Rietmeijer, 1987]. 



The data used in this analysis have been obtained solely from Analytical Electron Microscope studies 
of individual CP IDPs W7010A2 and W7029C1. The latter ID? was provided in two separate allocations 
W7029A23 and W7029A24 by the JSC Curatorial Facility. In each case the majority of grains within each 
allocation has been examined for mineral identity (i.e. structure and composition determinations) and 
grain size and shape. The dimensions of typically platey grains in both IDPs have been measured from 
transmission electron micrographs (with a precision of -1%) and calculated as the root-mean-square (rms) 
grain size. This size is calculated by the relation {a^+b^}^'^ where a and b are two orthogonal dimensions 
across a grain. 

Further details on individual mineral analyses and abundances, as well as their interpretations, are 
given in Mackinnon and Rietmeijer [1984, 1987], Rietmeijer and Mackinnon [1985a] and Rietmeijer 
[1989]. For simplicity, not all grains within each IDP are utilised in this study. Only non-carbonaceous 
grains and grains which are part of the IDP matrix are included in the grain size histograms shown in 
Figures 1 and 2. Thus, the data for IDP W7010A2 excludes measurements of large (> 1.0 fim) euhedral 
and rod-shaped silicate crystals [Rietmeijer, 1989] while the data for IDP W7029C1 exclude poorly 
graphitised carbon [PGC] grains which constitute -45% of all grains in this IDP [Rietmeijer and 
Mackinnon, 1985a]. 

The ultrafine platey grains in IDP W7010A2 are embedded in amorphous carbon and, as yet, 
unidentified hydrocarbons [Bradley, 1988; Rietmeijer, 1989]. The presence of these carbonaceous species 
suggests a low thermal regime (< -250°C) in the anhydrous IDP parent body(ies) [Rietmeijer, 1986; 
Rietmeijer and Mackinnon, 1985b]. In the case of IDP W7029C1, the degree of ordering inferred from 
the PGC basal spacing is consistent with a thermal regime of ~300°C in this IDP [Rietmeijer and 
Mackinnon, 1985b]. While this IDP is nominally an anhydrous variety, -11% of all grains are layer 
silicates and qualitatively, the mineralogy of IDP W7029C1 is similar to that interpreted from the 
chemical signature of ultrafine-size silicate dust in P/comet Halley [Rietmeijer et al., 1989]. 

The omission of grain size data for carbonaceous phases, as well as data related to the presence of Ti- 
rich minerals which have pseudomorphic textures due to a temperature dependant transformation, limits a 
thorough interpretation of grain size distributions for these two CP IDPs. Nevertheless, the choice of 
grains for this size distribution comparison implies an analysis of processes which have affected the bulk 
of the IDP (and, by implication, the IDP parent body(ies)). The size distribution for 254 grains in IDP 
W7010A2 and for 157 grains in IDP W7029C1 are shown in Figures 1 and 2, respectively. 


Interpretations of grain size distributions are, to a first approximation, model dependant and for the 
sake of discussion, we list below important assumptions for these interpretations: 

(1) chondritic porous IDPs are samples of cometary dust, 

(2) hydrocryogenic and low-temperature aqueous alterations of anhydrous IDPs occurs on comet nuclei, 

(3) the chemical and mineralogical diversity of chondritic IDPs is a good argument for similar diversity 
in comet nuclei. 

While we have argued for mineralogical diversity, and thus evolution, in cometary nuclei [Rietmeijer 
and Mackinnon, 1987], there is as yet little understanding of the spatial variations of comet mineralogy 
with time. Nevertheless, if we compare the behaviour of terrestrial sediments during diagenesis, it is 
apparent that grain size distributions follow well-defined and predictable trends [Berner, 1980]. For 
example, grain sizes during terrestrial diagenesis generally show initially a strongly peaked size 






800 1000 1200 1400 

RMS GRAIN SlZE(Nanometer) 






FIGURE 1: Root- mean- square grain size distribution for 254 mineral grains in chondritic porous 
interplanetary dust particle W70IOA2. 

coo "00 1600 1800 2000 

RMS GRAIN SEE (Nanometer) 

2200 2400 

FIGURE 2: Root-mean-square grain size distribution for 157 mineral grains in chondritic porous 
interplanetary dust particle W7029C1. 


distribution in the original sediment which flattens and shifts to a higher mean grain size with further 
diagenesis. Also a concomittant decrease in porosity accompanies an increase in the median grain size in 
terrestrial sediments [Berner, 1980]. 

The rms grain size distribution for IDP W7010A2 [Figure 1] shows a distinct positive skewness which 
is markedly different from the much flatter distribution for IDP W7029C1 [Figure 2]. Also, the mean rms 
grain size for each distribution differs, viz. 96.85 nm (W7010A2) and 562.1 nm (W7029C1). These data 
suggest comparatively advanced diagenesis for IDP W7029C1 relative to IDP W7010A2 and is consistent 
with the higher thermal regime indicated for the former. Both IDPs are of the chondritic porous subtype. 
Yet, the moderately higher median grain size for IDP W7029C1 (1325 nm) compared to 1125 nm for IDP 
W7010A2 indicates a slightly lower porosity for the former and suggests that mineralogical evolution of 
cometary nuclei will be accompanied by subtle changes in grain size, and consequently also in nucleus 
porosity and density. Measured densities for chondritic IDPs are between 0.7 and 2.2'^ [Flynn and 
Sutton, 1988; van der Stap, 1986]. Unfortunately, the porosity of chondritic IDPs is poorly known but it 
may be as high as 90% for anhydrous chondritic IDPs and -70% for a layer silicate-rich IDP [Mackinnon 
et a!., 1987]. 

Assuming that (1) terrestrial diagenesis can be used to model the chemical, mineralogical and physical 
evolution of chondritic IDPs and (2) chondritic IDPs are samples of cometary dust, it will be a 
prerequisite to assess grain size distributions of chondritic IDPs. Comet nucleus models should consider 
differences in physical properties (grain size, porosity and density) on length-scales of at least -60 nm 
which is the size of the largest chondritic IDP presently collected from the Earth's stratosphere. The 
extent and spatial variations with time of these differences within cometary nuclei will be different for 
individual comets and will depend on inherent comet nucleus properties such as ice-dust ratios, the 
structural state of dust, .the evolution of comet orbits and comet lifetime. 


Petrological analyses of chondritic porous IDPs suggest that grain size, density and porosity of comet 
nuclei may evolve during their lifetime in the Solar System. Effects of physical evolution, as well as 
chemical and mineralogical evolutions, in cometary nuclei may be subtle. The extent and spatial variations 
with time are presently unknown but it seems imperative for models of active short-period comets to 
consider the possibility of dust evolution. 

We believe that Analytical Electron Microscope analyses of chondritic IDPs, in conjunction with 
astronomical observations and theoretical modelling, will yield the data to model comet nucleus evolution. 
It seems obvious that putative evolutions of comet nucleus physical properties can place engineering 
constraints on a Comet Nucleus Sample Return Mission. For example, the mode of penetration (rotation 
or percussion) selected for a "smart nucleus penetrator" as a function of resistance encountered during 
descent into the nucleus may critically depend on pre-programmed density differences. Modelling the 
physical, chemical and mineralogical evolution of cometary dust properties will have a qualitative 
character until a successful Comet Nucleus Sample Return Mission. However, the possibility of comet 
nucleus evolution may have important implications for mission planning and the type of sample that will 
be returned. 


Berner, R. A.: Early Diagenesis. A theoretical Approach. Princeton University Press, 1980. 
Bradley, J. P.: Analysis of chondritic interplanetarv dust thin-sections. Geochim. Cosmochim. Acta, vol. 


52, 1988, pp. 889-900. 
Bradley, J. P.; Sandford, S. A.; and Walker, R. M.: Interplanetary Dust Particles. In: Meteorites and the 

Early Solar System (Kerridge, J. F. and Matthews, M. S., Eds.), University Arizona Press, 1988, pp. 

Brownlee, D. E.: Collection of cosmic dust: Past and future. In: Properties and Interactions of 

Interplanetary Dust (Giese, R. H. and Lamy, P., Eds), D. Reidel Publ. Co, Dordrecht, Holland, 1985, 

pp. 143-147. 
Flynn, G. J.; and Sutton, S. R.: Cosmic dust particle densities inferred from SXRF elemental 

measurements. Meteoritics, vol. 23, 1987, pp. 268-269. 
Mackinnon. I. D. R.; and Rietmeijer, F. J. M.: Bismuth in interplanetary dust. Nature, vol. 311, 1984, 

pp. 135-138. 
Mackinnon. I. D. R.; and Rietmeijer, F. J. M.: Mineralogy of chondritic interplanetary dust particles. 

Reviews Geophys., vol. 25, 1987, pp. 1527-1553. 
Mackinnon. I. D. R.; Lindsay, C; Bradley, J. P.; and Yatchmenoff, B.: Porosity of serially sectioned 

interplanetary dust particles. Meteoritics, vol. 22, 1987, pp.450-451. 
Oort, J. H.: The structure of the cemetery cloud surrounding the solar system and a hypothesis 

concerning its origin. Bull. Astron. Inst. Netherlands, vol. 11, 1950, pp. 91-110. 
Rietmeijer, F. J. M.: Olivines and iron-sulfides in chondritic porous aggregate U2015*B formed at low- 
temperature during annealing of amorphous precursor materials. Meteoritics, vol. 21, 1986, pp. 492- 

Rietmeijer, F. J. M.: Sulfides and oxides in comets. Astrophys. J., vol. 331, 1988, pp. L137-L138. 
Rietmeijer, F. J. M.: Ultrafine-grained mineralogy and matrix chemistry of olivine-rich chondritic 

interplanetary dust particles. Proc. 19th Lunar Planet. Sci. Conf., 1989, pp. 513-521. 
Rietmeijer, F. J. M.; and Mackinnon. I. D. R.: Layer silicates in a chondritic porous interplanetary dust 

particle. Proc. 16th Lunar Planet. Sci. Conf., part 1, J. Geophys. Res., vol. 90, Suppl., 1985a, pp. 

Rietmeijer, F. J. M.; and Mackinnon. I. D. R.: Poorly graphitized carbon as a new cosmothermometer 

for primitive extraterrestrial materials. Nature, vol. 316, 1985b, pp. 733-736. 
Rietmeijer, F. J. M.; and Mackinnon. I. D. R.: Cometary evolution: Clues from chondritic interplanetary 

dust particles. European Space Agency, SP-278 (September 1987), 1987, pp. 363-367. 
Rietmeijer, F. J. M.; Mukhin, L. M.; Fomenkova, M. N.; and Evlanov, E. N.: Layer silicate chemistry 

in P/Comet Halley from PUMA-2 data. Lunar Planet. Sci., vol. 20, 1989, pp. 904-905. 
van der Stap, C. C. A. H.: Experimental studies of meteorites and cosmic dust, Ph.D. thesis Free 

University, Amsterdam, the Netherlands, 1986. 
Weissman, P. R.: The origin of comets: Implications for planetary formation. Protostars & Planets II 

(Black, D. C. and Matthe\ys, M. S., Eds.) University of Arizona Press, Tucson, 1985, pp. 895-919. 

^ ' This work was supported by NASA Grant NAG 9-160. 



R. E. Johnson 

Department of Nuclear Engineering and Engineering Physics 

University of Virginia 

Charlottesville, Virginia 


Laboratory Simulations: The Primordial Comet Mantle 

R.E. Johnson 

Dept. of Nuclear Engineering and 

Engineering Physics 

University of Virginia 

Charlottesville, VA 22901 

Laboratory data are needed to understand the formation of organics in 
cometary and precometary materials and for deciding on the fate of the 
volatiles. Appropriate experiments were described in the talk at Milipitas. 
Because of its importance for the comet sample return mission, I discuss here 
the relevance of this data for predicting the thickness, nature, and ability to 
survive of the cosmic-ray produced primordial comet mantle ('crust') . That 
part of the mantle which becomes predominantly refractory is 30 gm/cm"^ thick. 
The tensil strength of this outer mantle is such that it might survive the 
comet's entrance into the inner solar system. In addition, important 
modifications to the ices occur to depths 300 gm/cm'^ . Based on this it is 
expected that a deep probe is needed to obtain minimally altered material. 



The outer layers of a comet, the comet's mantle, will be sampled 
during the proposed Rossetta comet-sample-return mission. It has been pointed 
out by a number of authors that this region of the comet is significantly 
altered by cosmic -ray particle processing of the ices and organics during the 
comet's 4.5x10^ years residence time in the Oort cloud^-'-'^) . That such an 
alteration occurs would appear to be reenforced by the recent measurements of 
the ortho-para ratio of water molecules effusing from a new comet^-'^. Since a 
prime goal of the proposed mission is, as indicated by its name, to establish 
the connections between the comet constituent materials and the precursor 
materials, any post -formation alterations to the region from which the sample 
is taken is of great importance. There are, of course, other processes, some 
of which were recently described by Stern^^'^^, that can also affect the 
structure and state of the comet mantle before the comet is ejected from the 
Oort cloud into the inner solar system. These will not be dealt with here: see 
paper by Weissman in this volume. Whereas estimates of the effect of the 
Stern processes are based on statistical considerations involving a nxomber of 
likely interactions but involving uncertain physical quantities (e.g. Oort 
cloud comet densities, supernova events near our solar system, etc.), 
determination of the cosmic- ray particle processing of the mantle is based on 
spacecraft measurements of cosmic-ray particle fluxes and on laboratory 
measurements of energetic particle alterations of materials. Therefore, the 
effect of cosmic-ray particles on the comet mantle can, in principle, be 
described with some certainty, especially if this mantle is static during the 
comet's life in the Oort. If other mantle altering processes are also 
important then the effects described here are superimposed on them. 


In my talk in Milipitas I described the general nature of the laboratory 
simulations of interest to comet science, I attempted to outline the various 
stages in the 'history' of comet materials during which charged-particle 
alterations occurred, and I discussed the differences between UV and charged- 
particle irradiations . In this paper I will not discuss those topics in any 
depth as that material has been recently incorporated into a chapter on 
particle irradiation effects for the volume to be published following the 
Bamberg meeting this year^°^ and some of that material also occurs in a 
recently completed text on charged particle irradiation effects to appear 
shortly^^^. One of the most important topics from my talk for the proposed 
Rossetta sample-return mission is the nature of the primordial comet mantle, 
and the survivability and evolution of this mantle when the comet is ejected 
into the inner solar system. Therefore, this paper will be devoted primarily 
to this topic. In the next section I will review the laboratory data relevant 
to mantle formation and mantle stability. I will then combine this with the 
cosmic -ray particle energy deposition profile in order to expand on the picture 
of the mantle described in an earlier paper ^ ■'■ ^ . Finally I will consider the 
fate of this mantle and its relevance for the Rossetta mission. 

Laboratory Results 

There has been a considerable body of literature devoted to the study of 
energetic particle alterations of materials. Whereas the GeV energies of 
interest for the primordial mantle formation are not easily obtainable in the 
laboratory it has been shown that the effects of interest generally scale with 
the electronic energy deposition per unit path length in the material, 
(dE/dx)g' ■'-^' . This energy produces, primarily, ionizations and excitations of 
the electrons on the molecules in the sample. The decay of these excitations 


leads, with high frequency, to bond breaking and atomic displacements. In ices 
or organics this can initiate chemical activity, whereas in more refractory 
materials structural alterations and defects may occur. Because the nature of 
the primary excitation processes produced in the laboratory by keV electrons 
and MeV ions are the same as the dominant processes for GeV particles, 
laboratory data can be applied to describe the comet mantle formation knowing 
only the total dose of electronic energy deposited. The GeV cosmic-ray 
particles produce nuclear reactions more efficiently than the MeV ions, but 
much of the energy lost in such processes is eventually also transferred to the 
electrons by the stopping of the energetic products and it has been shown the 
net effect of the nuclear alteration is small ^ •'■■'■ ^ . 

Alterations produced in molecular ices, organic solids and liquids, and 
molecular gases are often characterized by G values ^ ^^^ . G is the average 
number of a particular type of alteration (e.g., 2 CH4 -► C2Hg + H2) per lOOeV 
of electronic energy deposited in the material. Such results have been 
published for those gas and liquids relevant to biological processes^^^ . 
Typical G-values are in the range 0.1-2 depending on the process. However, 
in most cases studied diffusion of some species (i.e. temperature) plays a 
role. The low temperatures in the Oort cloud inhibit such effects, therefore, 
only studies on ices and organics at low temperature (<30K) are relevant. In 
this region thermal diffusion of any species except H2 and He is highly 
inhibited. Irradiation induced 'diffusion', due to jostling of the atoms 
following displacement events, does occur, however. Further, only experiments 
involving low fluxes in which events happen on a particle-by-particle basis are 
relevant. In such an environment unrecombined radicals can be relatively 
stable in the solid matrix, as has been shown experimentally^^-^^ . Of course, 
even though the mobilities are small, extrapolating them to 4.5xlo" years is 


risky. In this regards, one of the major effects of the Stem-processes 
mentioned earlier is a temporary elevation of temperatures in the mantle region 
which may allow radical recombination. Such chemical activity also adds heat 
and, therefore, a super nova shock or a comet-comet collision might enhance the 
alteration of the comet mantle. 

In a closed system long-term irradiation will lead to a material which is a 
stable mixture of species, as bonds are broken and reform. However, the 
presence of a vacuum interface allows the continuous ejection of volatile 
species from the surface. Therefore, the material is driven to an end point, a 
highly refractory material resistant to alteration by subsequent irradiation. 
The most volatile species formed, H2 , is lost 'immediately' by diffusion 
through the damaged solid to the vacuum. Other volatile species formed in an 
ice mixture or from organics (e.g. CO, O2, N2, etc.) are "lost' more slowly at 
these low temperatures by incident particle assisted diffusion and ejection 
from the surface'^). However, the loss of H is the controlling effect. In a 
mixture with atomic composition of C, H,0,N,S, the removal of H permits the 
enhanced formation of for example C-C, C-N, C-O, C-S and S-S bonds. This 
converts ice mixtures (e.g. H2O, NH3 , CO, CH4, H2S , etc.) to 'organics' with 
residual volatiles (e.g. O2 , CO) and some unrecombined radicals. 

In this irradiation process the solid also becomes inhomogeneous so that the 
material partially segregates due to preferential bonding. Therefore, local 
carbonized regions form and, possibly, regions with enhanced S content. As the 
material becomes altered the efficiency of the radiation eventually 
decreases'-'- ' , and in almost all mixtures studied, containing either C or S 
atoms, refractory solids, referred to as residues, result after long-term 
irradiation^-'-^' •'■^ . Further, the refractory materials themselves experience 
enhanced adhesion due to the irradiation' ■'-'^ . In an organic solid such 


processes happen more efficiently (i.e. at lower doses). This residue 
material, superimposed on damaged refractory grains (e.g. silicates), is the 
material that forms the outer part of the mantle. 

In order to quantify the above it is sufficient to note that in gases of 
small molecules an ionization event is produced, on the average, for every 30 
eV of energy deposited (G ' 3) . One such event generally leads to breaking of 
a typical covalent bond and the formation of a radical. In solids and liquids 
ionizations (electron-hole pairs) are produced more efficiently. In charged 
particle irradiation the average primary excitation energies are of the order 
of 60 eV/molecule resulting in pairs of ionization events occurring in close 
proximity due to the secondary electrons produced. Therefore, in addition to 
radical formation and hot-atom chemistry, prompt reactions can occur between 
the affected neighbors. Based on the results of Foti et al.^-'-°^ the 
polymerized fraction of an organic component at low doses varies, roughly, as 

f = (1 - exp (-a<i>)) (1) 


where 4> is the fluence of MeV ions (ions/cm'^; a dose). They find a a (dE/dx) 
Writing a = (dE/dx) g/nj^W^ , where n^ is the molecular number density and V^, is 
the energy deposited per molecule for loss of H, formation of new C-C bonds, or 
cross-linking^^°\ In this form 

a4> = Dji/Wc (2) 

where Dj^ is the dose in eV/molecule and W^ = 140 eV based on their data for MeV 
ions, or G = 0.7. This energy is about twice the average energy deposited per 


methane molecule found by Lanzerotti et al.^-*-^^ for initiating the loss by 
diffusion of newly formed H2 from the complete depth of penetration of the ion 
in solid methane. The G-value above energy corresponds to about two primary 
ionization events per initial small molecule . In an ice mixture , the average 
energy deposition for forming C-C bonds will increase somewhat due to reduced 

Water is lost from a low temperature solid both by sputtering and 
conversion, H2O -* ^2 + (1/2) O2 . The G for the latter process is " 0.7 for a- 
particles and 0.3 for /9-particles. These are comparable to the values for 
organic alteration above. As GeV ions have very low (dE/dx)g, more comparable 
to the P's, the smaller values may be more relevant at large depths into the 
mantle. The H2 is lost by diffusion through the lattice 'immediately' after a 
dose 60 eV/molecule is received (G " 1.5) and then the accumulated O2 is 
driven off. 

Cosmic-Ray Doses 

The energy deposited by cosmic rays in the comet mantle has been estimated 
by a number of authors'^' ■'-^' ^^^ . As cosmic ray fluxes have been measured in 

the earth's atmosphere for years the measured energy deposition vs thickness, 

given in grams/cm^, can be used as a lower limit and applied to the comet, as 

was done initially by Whipple ^^^. 

Since the primary energy loss mode for fast ions is to the electrons, the 

standard Bethe-Bom expression for energy deposition, with relativistic 

corrections at high speeds, can be used to calculate the energy loss. A 

recommended expression for protons for energies of the order of or greater than 

1 Mev(21) is 


(g)^- (5.1x10 '■^) (eV cm2) (Z^n g) (3) 

I (1-r) 


which is shown for water in Fig (1). Here ^=(v/c) , c is the speed of light, 
Zgng is the number of electrons on the typical target atom or molecule having 
number density ng, and I' is called the mean ionization energy, here given 
relative to the hydrogen atom. Since (dE/dx)g is only weakly dependent on I' , 
approximate expressions are useful (I'= 5.5 for water, I' =1.2 Zg • for 
refractory components) . In addition there is a small component of stopping due 
to nuclear elastic collisions and a large component due to inelastic nuclear 
interactions, since the inelastic nuclear interaction length is 100 g/cm for 
1 GeV protons in most materials ^■^'^ ' . For an average energy loss in such 
processes of 100 MeV the mass stopping would be 1 MeV cm^/gm in water or 
about half the minimum in the ionization effect in Fig(l) . 

The dose of electronic energy deposited vs. depth is calculated from the 
particle flux spectrum, <^(E) , and the total stopping power, (dE/dx)^, and the 
energy loss by the primary and all subsequent particles. If the (dE/dx)'^ 
mostly relaxes to electronic loss then it can be used alone. Giving the dose 
in energy deposited per initial molecule, Dj^, for a depth z into the comet 

0(E') r(^)T~| <iE' (^a) 

Ignoring deflections, the energy of a ion of initial energy E at depth z is 

E(z) , so that 



10 - 













10 - 


- 10 



- 10 






- 10 

- 10 

I 10 10 10' 



Fig.l The electronic stopping power from Eq.(3) and the penetration depth 
using this. Axes as indicated. 


In Eq (4a) E^ is the minimum ion energy, an ion which reaches depth z with 
E(z)-o. Using Eq (4a) and (4b) with (dE/dx)i' equal to (dE/dx)g an approximate 
Dj^ at large penetration depths was obtained by Moore ^^-^^ for an incident flux 

(^(E)-k e"^ (ions/(m2 sec ster MeV/nucleon) ) (5) 

where a-2 . 5 , k-2.5xlO^, and e-E+mQC-^ . At z « 10 cm , Dj^ = 100 eV/mol, for E < 
IGeV. Ryan and Draganic^-'--'-^ approximated the nuclear aspect of the stopping 
for E > IGeV. The results given in ref (1) based on ref (11) were too large as 
the data in ref (11) were plotted incorrectly. Strazzulla^ ' obtained 
alteration depths which were too large due to neglect of the high velocity 
energy loss effects. 

The actual atmospheric measurements can be used as lower bounds to Dj^ in 
Eq (4a) at large z. Since ionization, due to the incident particles or the 
secondaries, is the dominate energy loss mode and is directly connected to 
material alteration, the measurements of ion pairs (ion plus electron) produced 
per cm-^ per sec in air'^^ can be used. Assuming the atmospheric molecules (O2 > 
N2 , CO2 , H2O, etc.) exhibit energy loss curves similar to that of an average 
comet material (H2O, CO, C2O, Si02, etc.), the number of times an 'original' 
molecule would be ionized in 1.4 x lO-*-' sec (4.5 x 10' years) can be 
calculated. Accounting for the differences in stopping between water and air 
and for the differences in W values results about (1.5) times Whipple's 
estimates are obtained and are given in Fig (2) for energy deposition in water. 
Assuming " 30 eV deposited per ionization, the right hand axis gives the number 


DEPTH (METERS AT 0.2 g/ cm ) 
10 15 20 25 



- 5 

- 4 

— 3 








— 1 


DEPTH (lOOg/cm^) 

Fig. 2 Cosmic-ray dose. Solid curves :U Whipple's estimate from the atmospheric 
measurements multiplied by 1 . 5 due to accovmting for differences of ionization 
and stopping in air and water. MDRD sum of the Moore and Donn results and the 
Ryan and Draganic' results as discussed. Dashed cuirve an interpolation. 


of times, on the average, an original water molecule has been ionized. (In the 
solid electron/hole pair formation requires, on the average, less energy.) 

The Ryan and Draganic^ ■'••'■) results for E > iGeV are combined with the 
Moore^^-^) results, E < 1 GeV, in Fig (2). The depth is given in g/cm^ since 
stopping is roughly proportional to the mass density in any solid (i.e. the 
number of electrons is roughly proportional to the mass). Therefore, for water 
ice lOOg/m"^ implies Im depth; for a comet mantle or 'dust' density ( 0.2g/cm-^), 
lOOg/cm^ is "5m. The choice of this density accounts for some of the confusion 
in stating mantle 'thickness'. The latter results provide lower limits to the 
dose at small depths and the results obtained from air provide lower limits at 
large depths. The dashed line interpolates between these (note that these 
doses are lower than those in ref . (1) for the reasons discussed) . 

The Comet Mantle 

It is seen that at a depth of 100 g/cm^ into the comet any original water, 
methane, ammonia, etc. molecule will have been 'ionized' at least once in its 
life in the Gort cloud. Using Eqs . (1) and (2) with the net dose in Fig (2), 
the fraction of the initial organic material in the mantle which is polymerized 
is shown vs. depth into the mantle in Fig (3) based on G 0.7. 

Water molecules are depleted from the mantle due to ion bombardment. This 
depletion is due to sputtering and chemical conversion H2O -► H2 + (1/2) O2 with 
the latter process dominating. The loss of H2O is likely to be given by 
something in between the solid (G 0.7) and dashed curves (G 0.3). As the 
damage to water occurs in a mileau containing other species the back reactions 
are less likely so that the larger G value may be more relevant. Conversion of 
CO2 to CO + (1/2)02 is even more ef f icient^^ ' ■'■*^) , affecting greater depths. 

Regions containing sulfur molecules such as CS2 , H2S and SO2 in a water or 


carbon mix will be converted to sulfur suboxide, S2, or chain-sulfur with G- 
values comparable to those used for obtaining Fig (3). Therefore, the 
observation of S2 near the nucleus of a new comet may be ^understandable . 

The recent observations of the altered ortho-para ratio ^-"^ can be understood 
by noting that at low temperatures H transfer occurs efficiently on ionization 
(" 30 eV/mol., G ~ 3) . If this is the case then the alteration depth (f ' e'-'-) 
is 200-300 gm/cm^. This is a region of little water loss. Writing the 
fraction of the original water which remains and has its ortho-para ratio 
changed ( exp ( - Dj^/W^, ) - expC-Dj^/W)) with W^, = 140eV and W = 30eV, that fraction 
of the initial water which has an altered ortho-para ratio is given in Fig (4) . 
At small depths the water is lost, at large depths the ortho-para ratio is 
altered. Therefore, it is seen that a considerable volume of ice is so 
altered. This region is also a region in which stored radicals are produced 
which may result in additional activity on first pass^^'^^. 


Because the proposed comet -sample return penetration does not go very deep 
into the mantle , it is not clear from the above results that the Rossetta 
mission can obtain a record of early solar system or pre -solar system (ISM) 
material with a shallow probe . A rajunber of observations on new comets are 
indicative of outer mantle alterations. However, a key issue is the tensil 
strength of the refractory material in the altered layers which would indicate 
whether the primordial mantle would survive . The region of depth < 30 gm/cm^ 
has received a total dose of radiation comparable to that experienced by an 
IPDP (interplanetary dust particle) prior to collection at earth^^^) . These 
particles are structurally sound, due in part to the irradiation produced 
adhesion^-"-' ^ . Therefore, a tensil strength in the outer layer equivalent to 




_^ lO 











o — 




T- ^ 



























TNOiiDvad QByBinv 





^ in 







[L io 


O — 



— CO 


— ^ 




ro Q- 

































^NOiiovdd G3a3inv 


such particles would be expected. As pointed out earlier^ ■'■^ , the outer mantle 
('crust') will not be uniform, due to the initial irregularity of the surface 
which allows shadowing. Therefore, there will be unstable -regions of the 
mantle which will be lost quickly and crevices may exist allowing volatiles to 
have access to the vacuum. These form the active regions which will be 
controlled by inner solar system mantle forming processes'^^^ . Such processes 
will also increase the thickness of the mantle in the regions from which the 
primordial mantle it is not lost. The comet then acts like a large reservoir 
for volatiles and loose dust with a few active spots, as was observed to be the 
case for Halley. The warming of the comet fills the reservoirs below the 
active region and increasing temperature near perihelion makes these regions 

Acknowledgement: The author acknowledges the support of NASA grant NAGW 186 
and discussions with M. Moore and R. Streitmatter about cosmic-ray doses. 

(1) Johnson, R.E., J.F. Cooper, L.J. Lanzerotti and G. Strazzulla (1987) 
Radiation formation of a non-volatile comet crust Astron. Astro. 187 889- 


(2) Whipple, F.L. (1977) The constitution of cometary nuclei in Comets . 
Asteroids. Meteorites : ed. A.H. Delsemmone , (Univ. of Taledo Press) pp. 

(3) Donn, B. (1976) The nucleus: panel discussion in The Study of Comets (B. 
Donn et al . , eds . ) (NASA SP-393) pp 611-621. 


(4) Strazzulla, G. (1986) "Primitive" galactic dust in the early solar system? 
Icarus 67 63-70. 

(5) Mumma, M.J., W.E. Blass , H.A. Weaver, and H.P. Laron (1989) Measurements 
of the ortho-para ratio and the nucleus spin temperature of water vapor in 
comets Halley and Wilson and the implication for their origin and 
evolution. (submitted) . 

(6) Stern, S.A. and M.J. Shull (1988) The reflectance of supemovae and 
passing stars on comets in the Oort cloud Nature 332 407-411. 

(7) Stem, S.A. (1988) Collisions in the Oort cloud. Icarus 73 499-507. 

(8) Strazzulla, G. and R.E. Johnson (1989) Irradiation effects on comets and 
cometary debris (submitted) . 

(9) Johnson, R.E. (1989) Energetic Charged- Particle Interactions with 
Atmospheres and Surfaces ( Springer -Verlag, Berlin) in press. 

(10) Magee, J.L. and A. Chatterjee, (1987) Track reacting of radiation 
chemistry in Kinetics of non homogeneous processes (ed. G.R Freeman) (J. 
Wiley and Sons, N.Y.) pp. 171-214. 

(11) Ryan, H.P. and I.G. Draganic (1986) Radiation dosimetry of a cometary 
nucleus Astron. Space Sci. 125 49-69. 

(12) Cooper, D.C. and R.W. Wood (1974) Physical Mechanisms in Radiation Biology 


(CONF- 721001, U.S. Dept. of Commerce, Wash.) 

(13) Hart, E.J. and R.L. Platzman (1961) Radiation Chemistr-y, in Physical 
Mechanisms in Radiation Biolopy 1 (Academic Press, N.Y.) pp. 93-120. 

(14) Lanzerotti, L.J., W.L. Brown and K.J. Marcantonio (1987). Experimental 
study of erosion of methane ice by energetic ions and some consideration 
for astrophysics. Astrophvs . J . . 313 910-919. 

(15) Moore, M.H. and B. Donn (1982). The infrared spectrxim of a 
laboratory-synthesized residue: implications for the 3.4 micron 
interstellar absorption feature. Astrophvs . J . 257 L47-L50. 

(16) Foti, G. , L. Calcagno, K.L. Sheng and G. Strazzulla (1984) 
Micrometer-sized polymer layers synthesized by MeV ions impinging on 
frozen methane. Nature 310 5973-5974. 

(17) Johnson, R.E. (1985) Comment on the evolution of interplanetary grains in 
Ices in the Solar System (eds. J. Klinger) (D. Reidel, Amsterdam) pp. 

(18) Foti, G. , L. Calcagno, F.Z. Zhu, and G. Strazzulla (1987). Chemical 
evolution of solid methane by keV ion bombardment. Nucl . Inst, an d Meth. 
B 24/25, 522-525. 

(19) Lanzerotti, L.J., W.L. Brown and R.E. Johnson (1989) Threshold for H2 
release from methane Astrophvs. J. (submitted) 


(20) Moore, M.H. and B. Donn (1983). Studies of proton- irradiated ice mixtures 
Icarus 54 388-392. 

(21) Inokuti, M. J.L. Dehmer, T. Baer, and J.D. Hauson (1981) 
Oscillator- strength moments, stopping powers, and total 
inelastic- scattering cross sections of all atoms through strontium. 
Phvs. Rev. A 23 95-109. 

(22) Particle data Group (1984) Review of particle properties. Rev. Mod. Phys. 
56 supplement. 

(23) Moore, M.H. (1982) Ph.D. Thesis. University of Maryland: 

(24) A'Heam, M.F. P.D. Feldman, and D.G. Scheicher (1983) The discovery of $2 
in comet IRAS-ARAKI-ALCOCK 1983d. 

(25) Johnson, R.E. and L. J .Lanzerotti (1986) Ion bombardment of interplanetary 
dust. Icarus 66 619-624. 

(26) Prialnik, D. and A. Bar-nun (1988) The formation of a permanent dust 
mantle and its effects on cometary activity Icarus 74 272-283. 



Joseph A. Nuth m. 

Astrochemistry Branch 

Laboratory for Extraterrestrial Physics 

NASA Goddard Space Flight Center 

Greenbelt, Maryland 



Nucleation is a non-equilibrium process: the products of this process are seldom 
the most thermodynamical ly stable condensates but are instead those which form 
fastest. It should therefore not be surprising that grains formed in a circumstellar 
outflow will undergo some degree of metamorphism if they are annealed or are exposed 
to a chemically active reagent. Metamorphism of refractory particles continues in 
the interstellar medium (ISM) where the driving forces are sputtering by cosmic ray 
particles, annealing by high energy photons and grain destruction in supernova 
generated shocks. Studies of the depletion of the elements from the gas phase of the 
interstellar medium tell us that if grain destruction occurs with high efficiency in 
the ISM, then there must be some mechanism by which grains can be formed in the ISM. 
Various workers have shown that refractory mantles could form on refractory cores by 
radiation processing of organic ices. A similar process may operate to produce 
refractory inorganic mantles on grain cores which survived the supernova shocks. 
Most grains in a cloud which collapses to form a star will be destroyed; many of the 
surviving grains will be severely processed. Grains in the outermost regions of the 
nebula may survive relatively unchanged by thermal processing or hydration. It is 
these grains which we hope to find in comets. However, only those grains encased in 
ice at low temperature can be considered pristine since a considerable degree of 
hydrous alteration might occur in a cometary regolith if the comet enters the inner 
solar system. Some discussion of the physical, chemical and isotopic properties of a 
refractory grain at each stage of its life cycle will be attempted based on the 
limited laboratory data available to date. Suggestions will be made concerning the 
types of experimental data which are needed in order to better understand the 
processing history of cosmic dust. 


Silicate dust is observed in a wide variety of astrophysical environments: in 
dense, circumstellar shells and the diffuse interstellar medium, in the cold, dark 
interiors of molecular clouds and in the energetic plasma of planetary nebulae. It 
would seem only logical that each of these environments could "leave its mark" on 
grains which once resided within. If a large enough sample of interstellar grains 
were collected it might be possible to distinguish features which could reveal 
something of the history of an individual grain based on that grain's deviation from 
characteristics common to an "average" grain. Such general techniques are already 
used to distinguish pre-solar components of meteorites (Anders, 1988) and 
interplanetary dust particles (Zinner, 1988) from more common matrix material. In 
these studies specific isotopic anomalies are thought to be correlated with grain 
formation in particular astrophysical environments. 

There is no doubt that individual components of any class of meteorite have 
undergone a significant degree of processing during the formation and evolution of 
the meteorite parent body. In addition the processes used to concentrate "pre-solar" 
meteoritic material in the laboratory completely destroy the petrographic context of 
the individual grains. Similarly, the IDP's have undergone a number of processing 
steps which make it difficult to distinguish primitive features from more recent 
processes including the possibility for metamorphism on the comet, in the cometary 
coma and the interplanetary medium or upon entry into the earth's atmosphere. The 
return of samples from a cometary nucleus promises to eliminate many of these 
impediments to interpretation of the astrophysical context of such grains provided 


that the samples are carefully chosen. In this paper I will discuss the 
characteristics which might be observable in grains from a sample of the cometary 
nucleus based on both theoretical considerations of the ]\fe cycle of refractory 
grains and laboratory experiments using simple analogs to the more complex natural 

In section II the lifecycle of a typical silicate grain is briefly discussed, 
from condensation in a circumstellar outflow to incorporation into a cometary parent 
body. In Section III I discuss the possibility for grain metamorphism within the 
cometary environment together with the results of relevant laboratory experiments 
which have been performed to date. In Section IV, the results of experiments which 
may have implications for the properties of interstellar grains are presented 
together with some speculative comments concerning the likelihood that evidence for 
any of these characteristics might be observable in cometary grains. Section Visa 
summary of what might be learned about pre-solar refractory grains by studying 
samples returned from a cometary nucleus. 

II. Lifecycle of Refractory Silicates 

a. The Circumstellar Environment 

At equilibrium in the photosphere of a cool star of normal (oxygen-rich) 
composition, most of the grain forming elements exist as atoms or diatomic molecules 
(Tsuji, 1973). SiO, atomic Fe and atomic Mg are present in approximately equal 
amounts and together account for more than about 75% of the material which will 
nucleate to form circumstellar grains. Other materials, e.g., Al , Ca or Na, together 
account for less than 5% of the grain mass and are therefore unimportant as a first 
approximation (Duley, Millar and Williams, 1979). Most of the remaining mass of the 
initially nucleated grains will be derived from reaction of the grain cores with OH. 
The Initially nucleated grains will be oxygen poor since nucleation Is an inherently 
non-equilibrium process, and much of the mass of the grain forming material exists in 
the gas phase as metal atoms. 

Once the grain nuclei begin to form in the supersaturated stellar outflow (due to 
the nucleation of SIO), a sufficient surface area becomes available for many of the 
remaining metal atoms to condense rapidly: this Is especially true of metals such as 
Fe and Al which have low vapor pressures. More volatile metals such as Mg, Ca or Na 
may not condense until later depending upon the exact temperature and pressure at 
which the nucleation of the Initial grains occurs. Both SiO and Al are efficient 
reducing agents and will tend to become more oxidized at the expense of other metals 
(such as Fe or Mg). Both of these species will tend to react rapidly with OH or HjO 
molecules which Impinge on the growing surface. If grain formation occurs rapidly, 
then it Is unlikely that either the SiO or Al will have sufficient time to become 
fully oxidized before they are shielded from directly impinging OH by the growing 
grain surface. Rapid grain formation will also result in a large number of grain 
defects and unsaturated chemical valences. 

The grain temperature in typical outflows appears to be on the order of 500-lOOOK 
(Rowan-Robinson and Harris, 1982) and It would therefore seem unlikely that the 
grains would be thermally annealed to a more stable "glassy" structure (Nuth and 
Donn, 1984). However, because the grains are already at a relatively high 


temperature they should be quite readily heated by the absorption of UV-visible 
photons. Depending upon the oxidation and structural state of the grain at the 
moment of heating it is possible that a significant amount of chemical energy could 
be released as the grain anneals, (e.g., Clayton, 1980) leading to large degree of 
internal mobility of the grain constituents. This would result in the formation of 
tetrahedrally coordinated silicon and aluminum oxides, the reduction of some of the 
iron and magnesium oxides to metal, some degree of segregation between metallic and 
oxide phases and the formation of a glassy silicate. 

In the outermost regions of circumstel lar winds, thermally driven annealing 
reactions have been completely quenched for a considerable time and the glassy 
silicate grains may undergo some degree of low temperature surface reaction with 
volatile species such as OH, HS, H„0 or HjS. If the (C + Si)/0 ratio is 
approximately equal to one, sulfides will be more prevalent than oxides as carriers 
of the other metals and the silicate grains will develop sulfide crusts as metallic 
magnesium and iron react with SH or H-S (Nuth et al., 1985). These effects have been 
observed in a number of stars (Goebel and Moseley, 1985). However, as the sulfide 
coated grains mix into the general interstellar medium, one would expect such 
coatings to be destroyed via reaction with the large abundance of water and OH in the 
ISM. Such reactions would produce MgO or FeO from any unoxidized metal atoms near 
the grain surface. Only metal nuggets which are well shielded from possible gas- 
grain surface reactions would be expected to survive as metals. Of course, jn stars 
with (C + Si)/0 considerably less than one, most of the exposed metal atoms will have 
been oxidized before the grains enter the ISM. 

Despite the wide variation in the structures of the silicate grains added to the 
interstellar medium by each class of oxygen-rich star, one might expect that 
conditions in the interstellar medium would tend to drive all silicate grains toward 
some steady-state average. The primary forces responsible for grain metamorphism are 
cosmic-ray induced radiation damage, annealing by the absorption of UV photons and 
the reaction of grain surfaces with oxidizing vapor-phase species. The later would 
tend to fully oxidize SiO and any metals such as iron or magnesium which remain 
exposed at grain surfaces. Constituents buried within grains are not likely to be 
effected by such processes since rapid diffusion below the grain surface would be 
Inhibited due to the low grain temperatures and low pressures in the interstellar 
medium. Cosmic ray bombardment would tend to destroy any crystalline structure in 
silicates which formed by annealing in very hot stellar environments. Conversely, 
chaotic amorphous silicates could become more ordered after cosmic ray bombardment or 
the absorption of a UV photon if the absorbed energy is greater than the activation 
energy for the formation of glassy silicates. Experimental observations have been 
reported which support both effects: Kratschmer and Huffman (1979) used high energy 
ions to convert crystalline olivine to glassy olivine while DeNatale and Howitt 
(1983) have shown that a degree of crystal Unity may be introduced in some glasses 
due to bombardment by 100 keV electrons in an electron microscope. 

The net observational effect of these processes would be to homogenize the 
properties of interstellar grains condensed in different circumstel lar outflows. 
Effects of peculiar stellar chemistry, such as the formation of a MgS crust on 
otherwise normal silicate grains, would be quickly eliminated in the ISM. Similarly, 


the preservation of highly crystalline or extremely chaotic silicates would be 
equally unlikely. Interstellar silicates would tend to be glassy and have a higher 
overall oxidation state than that of an average of silicates in circumstel 1 ar 
outflows. Iron metal inclusions might be preserved if they become buried within the 
glassy silicate grain. Magnesium and iron on grain surfaces would tend to oxidize. 

b. Destruction and Reformation of Silicates in the ISM 

Until recently, it was assumed that grains which formed in circumstel lar 
outflows comprised the bulk of the interstellar grain population: the masses of these 
grains might be increased by the accretion of both volatile and refractory organic 
mantles, but the silicate cores were expected to survive most interstellar processes 
intact. Several years ago, Seab and coworkers presented a series of calculations 
which indicate that the lifetime of a typical silicate grain might be much shorter 
than had previously been expected (Seab and Shull, 1985). In fact, some of these 
calculations Indicate that as many as 10 times the number of grains are destroyed in 
the ISM as are replaced by circumstellar condensates. Unless these calculations are 
grossly in error, a significant fraction of the interstellar grain population may 
have formed in the ISM rather than in circumstellar outflows. What properties might 
be used to distinguish grains formed in the interstellar medium from those formed in 
a circumstellar environment? 

If a large fraction of interstellar grains were atomized in a shock wave, 
recondensation would occur slowly on the surfaces of the surviving grain cores. As 
the metals (e.g., SI, Fe and Mg) are accreted onto such surfaces, reaction with the 
more abundant molecules (e.g., HjO, CO or NH.) or with adsorbed hydrogen atoms could 
be more probable than reactions which yielcf silicates (A.G.G.M. Helens, personal 
communication, 1984). Such reactions would undoubtedly produce a variety of metal 
radicals (e.g., SiOH, FeCO, or MgH) as well as sllanes (SI^Hj^,^-), metal hydrides or 
metal carbonyls. These species would coexist In the grain mantle with ices of water, 
ammonia and various organic compounds. Exposure of these mantles to cosmic rays and 
ultraviolet radiation would increase the concentration of metallic radicals in the 
matrix and create a variety of new organic radicals as well. As this ice/radical 
mixture warms, reaction of the radicals will produce a refractory residue. 
Experiments by Greenberg and coworkers (See Greenberg, this volume) and by Moore and 
Donn (1983) on organic/ice mixtures have demonstrated the formation of refractory 
organic residues, while experiments by Nuth and Moore (1988a, b) on metallic/ice 
mixtures have shown that silicate residues can be formed by such processes. 
Experimental studies of organic/metalllc/ice mixtures are needed before a realistic 
assessment can be made of the Importance of this process compared to a scenario in 
which silicates are made by epitaxial growth of individual atomic species onto 
surviving grain cores. The properties of these inorganic residues will be discussed 
in Section IV. 

c. Processing in the Solar Nebula 

Undoubtedly the bulk of the grains present in the portion of the giant 
molecular cloud which formed the sun were destroyed as they became Incorporated into 
the solar interior. Grains which passed close to the sun before being carried into 
the outer nebular regions could have been annealed to crystalllnlty, partially melted 
or even distilled to a more refractory composition. Implantation of ions from the 


intense solar wind into these crystalline grains is likely to have occurred at many 
different times as the sun underwent outbursts during the accretion process and 
during its T-Tauri phase. Gas-solid reactions must have occurred between grains and 
active molecular species such as 0, OH, H^O, S, HS or H.S. These reactions could 
have been mediated by nebular lightning, intense UV radiation from the sun or the 
nebular corona as well as by the elevated temperatures in the disk. All of these 
processes will be represented in the grains returned from a cometary nucleus provided 
that turbulent mixing in the disk was sufficiently vigorous. 

For comets, the accretion process itself is unlikely to have been energetic 
enough to have effected the grains: the process was sufficiently gentle that a 
significant quantity of volatile material remained within the comet. It is also 
unlikely that the comet underwent a period of intense radioactive heating: again 
because comets today retain a large volatile inventory which would have been lost 
during any period of even moderately high temperature (e.g., T > 250K) . Cometary 
grains sampled by CNSR will therefore be representative of the population of grains 
present in the outer solar nebula at the time the comet accreted provided that no 
grain metamorphism occurs on the comet over its lifetime. The probability of 
metamorphism in such an environment will be assessed below. 

III. Grain Metamorphism in the Cometary Environment 

a. Cosmic-ray bombardment 

Comets have spent roughly 4.5 billion years in the unshielded environment of 
the Oort Cloud without benefit of the protection from cosmic rays provided by the 
Sun's magnetic field. Over this time period it is probable that at least the topmost 
10 meters of cometary ice and dust have received a significant dose of radiation 
(Moore and Donn, 1982). The above estimate assumed a cometary density of ~ 1 g/cm^ 
and so might underestimate the depth to which such radiation penetrates if the 
density of the nucleus is significantly lower. It is usually assumed that this 
topmost layer is lost within the first one or two orbits through the inner solar 
system and so might not be of interest in a mission to an older (periodic) comet. 
Unfortunately, this may not be completely correct: the extent to which radiation 
processed dust is a component of the cometary regolith will depend a great deal on 
the mechanism by which volatiles escape from the cometary interior. It may even be 
possible that a significant fraction of this processed refractory material remains in 
the regolith if most outgassing comes from vents to the comet's interior rather than 
from a more or less continuous mass flux through the regolith (and random opening of 
vents which are nearly at the comet's surface). These possibilities will be 
discussed further in Section V. 

The effects of cosmic-ray interactions on the structure of refractory grains is 
relatively straight forward. Energetic particles may destroy the crystal linity of 
refractory minerals by creating defects in the previously ordered structure: this 
could be a simple matter of a few cosmic ray tracks in an olivine grain or could 
result in the complete loss of all order in a less stable phyllosilicate. On the 
other hand, the energy deposited within an amorphous silicate grain by the passage of 
a cosmic ray particle may sufficiently exceed the activation energy required to 
initiate oxidation/reduction reactions in a chemically inhomogeneous particle that 


these reactions could become self-sustaining. The resultant energy release might be 
sufficient to partially anneal the previously amorphous grain. 

b. Thermal Annealing 

Except in the case of sun grazing comets, it is fairly safe to assume that 
thermal annealing of even nanometer sized amorphous silicates will play little or no 
role in the metamorphism of cometary grains. Prior to the time when such materials 
were incorporated into the comet however it is probable that at least some thermal 
metamorphism will have occurred either in the c1 rcumstel lar environment, the 
Interstellar medium or the solar nebula. The rate and the effects of such processing 
depend upon both the chemical composition and initial structure of the grains. 

The technological importance of fiber optics has led to numerous studies of the 
kinetics of devitrification of various glass compositions. It has been shown by 
several authors that the rate at which polymorphs of SiOj crystallize is dependent 
upon the initial structure (Wagstaff, 1968; 1969), the presence of trace impurities 
(Zaplatynsky, 1974, 1988; Bihuniak, 1983; Boden et al., 1984), the bulk composition 
of the glass (Brown and Kistler, 1959) and the ambient atmosphere in which it is 
annealed (Wagstaff and Richards, 1966). All of these factors are important in 
considering the effect of thermal annealing on an initially amorphous grain of 
"cosmic" composition which passes through several very different astrophysical 
environments between the time it nucleates in a circumstellar outflow and the time at 
which it finally becomes incorporated into a comet. A more detailed understanding of 
the initial composition and structure of refractory condensates is needed before 
accurate predictions of the rate at which such grains crystalline can be attempted. 
Very preliminary x-ray diffraction studies of the rate at which amorphous Mg-SiO 
smokes become crystalline (Nuth and Donn, 1983) show an extreme temperature 
dependence. Because we have few constraints on the temperature history of an 
individual grain in the solar nebula before it was incorporated into the comet it 
will be difficult to directly unravel the temperature-time dependence of such 
material if it is well crystallized. More amorphous samples however, may provide 
upper limits to the time-temperature history of associated materials. 

Nuth and Donn (1984) have carried out a series of experiments designed to place 
limits on the stability of initially amorphous SijOj and Mg-SiO condensates as a 
function of temperature and time. In these experiments changes in the infrared 
spectra of the samples were monitored as a function of processing history. Figure 1 ( 
taken from Nuth and Donn, 1982) shows the changes in the infrared spectrum of 
amorphous Mg-SiO smoke annealed in vacuo at lOOOK for 0, 1, 2, 4, 8, 15.5 and 30 
hours (top to bottom). Three aspects of these spectra should be noted. First, no 
sharp features (indicative of crystalline material) appear in the spectra for at 
least 16.5 hours. Second, one of the earliest changes to occur in this series is the 
loss of a minor absorption feature near 11.5 microns. Third, a series of very large 
shifts in both the positions and relative Intensities of the major silicate features 
at 10 and 20 microns occur as the sample anneals: these latter changes indicate an 
increased level of polymerization and chemical equilibration within the sample. 


10 15 

20 25 






1400 1200 

1000 800 600 


Figure 1. 

Infrared spectrum of an unannealed Mg-SiO smoke (A) and of smoke samples 
annealed in vacuo at lOOOK for 1 hr. (B) , 2 hrs. (C) , 4 hrs. (D) , 8 hrs. 
(E), 16.5 hrs. (F), and 30 hrs. (G) (Nuth and Donn, 1982). 


Nuth and Donn (1984) have measured the rate at which the 11,5 micron feature (due 
to SijOg) disappears as a function of temperature. The rate of this reaction is 
given by k(hr-i) = 10® exp(-40 kcal /mole/RT) and represents a lower limit on the 
stability of an amorphous condensate. As the example in Figure 1 shows, if the 11.5 
micron feature is still present in the amorphous Mg-SiO condensate then one can infer 
that little metamorphism had occurred since the material nucleated. The half-life 
for survival of the 11.5 micron feature calculated using the above rate equation is 
given In Table 1 as a function of temperature. It is obvious from this table that If 
nucleation occurs at temperatures less than lOOOK and transport to cooler 
environments occurs relatively rapidly then the grains will remain amorphous and may 
preserve internal chemical heterogeneities which could lower the activation energy 
for other processes such as hydration. 

Infrared spectroscopy is sensitive to small changes in the average chemical 
bonding in the solid and is therefore very sensitive to the earliest stages of 
metamorphism (e.g., transitions from a chaotic state to a more ordered glassy 
structure). Unfortunately, because Infrared spectroscopy averages over a sample 
containing millions of small grains, conclusions derived from such studies apply only 
to "average" grains and cannot easily constrain the range of the deviations which 
might occur in an individual grain during processing. 

Rietmeijer et al . (1986) have shown using analytical election microscopy that 
although the bulk Mg-SiO sample is Indeed amorphous as shown by its infrared spectra, 
some degree of microcrystal 1 ini ty exists in even the unannealed smokes. The 
crystalline fraction Increases as a function of annealing time. It should be noted 
however that the Initial crystalline components (Olivine and tridymite) are not 
predicted to be the most thermodynamically stable mineral assemblage. As annealing 
continues, olivine and tridymite gradually convert to the more stable clinopyoxene. 
Several alternative explanations for these observations are discussed in the paper by 
Rietmeijer et al . (1986) and will not be further discussed here. 

Table 1 

Half-life of Si203 Disproportionation vs. Temperature 

Temperature (K) SUO ^ half-life 

1200 43 seconds 

1000 20 minutes 

800 50 hours 

600 24 years 

400 0.4 billion years 

c. Hydration 

The first general misconception which must be dealt with in any discussion of 
the hydration of cometary grains is that macroscopic quantities of liquid water are 
necessary for hydration to occur. Rietmeijer (1985) has argued, based on analogy 
with weathering studies of antarctic soils, that a significant degree of alteration 
can be expected in the presence of ice at temperatures below the freezing point of 
water. In addition Clayton and Mayeda (1984) have shown that hydration in at least 
one meteorite parent body occurred at temperatures which may have been below 273K. 
Such conditions may occur at the base of the cometary regolith in contact with the 


"icy snowball" surface in older models of a cometary nucleus. Hydration may also 
occur within the regolith as water vapor percolates outward through ever hotter 
grains to the cometary surface. It is well known to anyone who has ever worked with 
a high vacuum system that water bonds quite strongly to both silica and metal 
surfaces; transport of water vapor from the interior of a comet to its surface 
through a silicate regolith will therefore result in some degree of H^O absorption on 
grain surfaces. Studies of antarctic weathering have shown that significant 
hydration can occur from the presence of a single monolayer of water (see Rietmeijer, 
1985). If the regolith pores are not "well connected" then water vapor percolation 
will be slow, the surface coverage will be higher, and the probability of hydration 
will be increased. 

Studies of the rate at which amorphous Mg-SiO smokes hydrate in the presence of 
liquid water have been performed in order to set limits on the rate at which 
hydration may occur in a cometary regolith (Nelson, Nuth and Donn, 1987) or a 
meteorite parent body (Nuth et al., 1986). These studies used amorphous smokes as a 
starting material since these smokes are likely to be more reactive than cometary 
grains due both to their large surface to volume ratio and to the metastability of 
amorphous materials. In addition, liquid water was used rather than water vapor to 
ensure complete surface coverage of each grain by water. Reaction was followed by 
monitoring changes in the infrared spectra of the exposed samples as a function of 
both temperature and time. Results of these experiments are presented. in Table 2 
using several different infrared bands as indicators of the hydration process. 

As can be seen from Table 2, little hydration occurs below about 200K and 
therefore grains which are encased in ice at the temperatures postulated for the 
nucleus (T ^ 170K) will probably not be affected by hydration unless they are much 
more reactive than our amorphous Mg-SiO analogs. From these results it would appear 
that more crystalline materials which may have been produced in the solar nebula via 
several processes will be preserved quite well if encased in ice. Grains in the 
regolith however may be severely altered if the temperatures reported for Comet 
Halley (T 1 400K) are typical of most comets (Combs et al., 1986). Our amorphous 
materials are altered on timescales of hours at such high temperatures so that it 
would seem likely that more crystalline material would be altered to some extent over 
the longer periods available to grains in such a regolith. Alteration would of 
course proceed faster at higher temperatures or with shocked grains. 

Table 2 


life of Mg-SiO Infrared Features in Liqu- 

id Water 



11.5 micron 

16 micron 

20 micron 





4.1 seconds 

0.03 seconds 

41.4 seconds 


12.2 seconds 

0.16 seconds 

1.6 minutes 


1.3 minutes 

2.2 seconds 

6.6 minutes 


51.2 minutes 

7.6 minutes 

1.8 hours 


13.5 hours 

6.8 hours 

15.2 hours 


56.3 days 

220 days 

21.5 days 


1541 years 

360,000 years 

67.8 years 

Grains Produced 

in the Interstell 



As mentioned in Section II. b, Greenberg has long discussed the possibility that 
refractory organic mantles or "yellow stuff" will be formed on the surfaces of 
silicate grains in the interstellar medium. The properties of such material is 


discussed in the review by Johnson (this volume) and will therefore not be treated 
further here. However, a similar process may act to produce refractory inorganic 
grains in the interstellar medium (Nuth and Moore, 1988) or^ in the solar nebula as 
previously irradiated grains are gradually warmed during the collapse phase. The 
measured and expected properties of these grains are discussed briefly below. 

Infrared spectra of the amorphous iron silicate residues have been reported (Nuth 
and Moore, 1988a, b) and show strong and very broad absorption features near 10 and 20 
microns due to the silicate stretching and bending fundamentals, respectively. Also 
evident in the laboratory spectra are features at 4.6 and 4.9 microns due to the SiH 
stretch and the CO stretch. Both of these features are quite stable and are easily 
observed in the residues even after vacuum annealing to 400K for 30 hours. Features 
at these wavelengths are also observed in the spectrum of W33A, a strong infrared 
emission source thought to be a young protostar. The persistence of reactive 
entities such as SiH or volatile species such as CO after vacuum annealing and 
exposure to air at room temperature implies that silicates produced by slow reaction 
of icy radicals may serve as very efficient traps for more volatile materials. 

Experiments to measure the efficiency with which noble gases are trapped in these 
grains are currently in progress. Preliminary measurements show that both Kr and Xe 
are very efficiently trapped in the growing silicate matrix while Ne and Ar are 
trapped much less efficiently (Hohenberg, private comm.). It may be possible that 
the "planetary" component of the noble gases trapped in meteorites formed via these 
processes in the interstellar medium. Trapping would therefore have occurred both In 
carbonaceous and in inorganic grains. If noble gases are trapped then it might 
easily be possible to trap more reactive species such as CO or HjO. Interstellar 
materials may have been a major source of volatile materials for planetesimals in the 
early solar nebula. 

The morphology of the grains produced in our experiments is also quite 
interesting. Some grains produced from Irradiated mixtures of SiH^-HjO show quite 
distinct rhomboldal structure indicative of crystal Unity even though these grains 
were never heated above 300K (Mackinnon, personal comm.). This blocky texture is 
also evident in grains produced from Fe(C0)g-SiH^-H20 mixtures. In addition we have 
observed a number of Iron poor fibrous grains comprising approximately 10-15% of our 
sample. These grains probably also contain additional light elements such as carbon, 
oxygen, and hydrogen. More detailed analysis of the composition, crystal structure 
and morphology of these materials is in progress (Mackinnon, personal comm.). In 
comets these materials would be identified as an inorganic coating over a refractory 
core much in the same way that "yellow stuff" would be expected to occur on silicate 
cores in the Interstellar medium. 

V. Sample Collection Strategy 

The major uncertainty in developing a strategy for the collection of samples of 
refractory cometary particles is the thickness and history of the regolith. Of 
course this consideration is also of major importance in collecting samples of 
cometary ices since, if the regolith is too thick, then a core sample may not extend 
down to the ice layer. In the old "dirty snowball" model of the cometary nucleus it 
was assumed that any dust which came to the surface was quickly lost. Despite 
evidence for a very high surface temperature (T > 400K at 1 AU) many models for the 


cometary surface still assume a very thin regolith which has a very low thermal 
conductivity. These models also assume that the active jets are "bare spots" in the 
regolith where the ice resides at the surface. If these models are correct then the 
Comet Nucleus Sample Return mission described at this workshop will be successful. 
Unfortunately the current mission does not have a viable backup sample collection 
strategy if the current model of the cometary nucleus is incorrect. 

The very low apparent density of Halley's Comet « 0.5 g/cc) supports the fractal 
model of a cometary nucleus proposed by Donn (1981, 1987). Such a comet would have 
numerous internal voids which could interconnect to allow the escape of volatiles 
from the nucleus and the transport of energy into the interior via gas transport. 
Laboratory experiments in which comet-like mixtures of ice and dust have been allowed 
to sublime into a vacuum (Saunders et al., 1986; Storrs et al . , 1988; reports of the 
KOSI team In this volume) have shown that the resulting residue may occupy nearly the 
same volume after loss of its volatile component as it did initially and that this 
residue can be surprisingly strong. If we combine Donn's model with the laboratory 
experiments one possible result is that the active jets are vents to the interior of 
the comet through which volatiles escape. It is quite conceivable that the outside 
of the comet could be almost all regolith - a large number of completely degassed - 
but still largely intact - dust balls which adhere to one another because the dust 
filaments of adjacent "units" are interconnected. 

In the above scenario the "jets" which are observed could still become active or 
dormant due to the opening or closing of interior connections via the condensation or 
evaporation of volatiles or the shearing caused by shrinkage of individual ice-dust 
balls during loss of the unit's volatiles. Stephens (this volume) has noted that if 
individual ice-mineral mixtures contain organics then the resulting residue is even 
sturdier than a typical ice-dust residue. Given the large number of CHON particles 
detected during the Halley encounter (Kissel and Kruger, this volume) one might 
suppose that the dust regolith could be quite strong and that once the regolith 
formed it would be quite difficult to erode. If a layer of cosmic-ray produced 
organic residue were produced in the Oort cloud (e.g., Johnson, this volume) then 
this organic glue might serve to hold a large number of near surface grains together 
as the comet first approached the sun by increasing the relative proportion of the 
stickier organics in the near surface region relative to the more volatile organics. 
Once this regolith forms it can only grow. 

If the model just described is more realistic than the thin regolith models, then 
several predictions can be made about the nature of the refractory particles returned 
from the nucleus and about the likelihood of recovering any ice using a 3m coring 
device. To take this latter point first, it would be very unlikely that such a short 
core would succeed in recovering an ice sample since it would seem more likely that 
in an older comet the top 10 to 100 meters of the comet would be completely degassed. 
Grains near the surface (e.g., the top 10 meters or so) may contain cosmic ray tracks 
from the time the comet spent in the Oort cloud. Grains in the regolith should also 
be completely hydrated. Cometary grains ejected in jets may be pristine materials 
newly released from the interior ice core or may be hydrated grains swept out of the 
interior from the walls of the regolith: both populations of grains are likely to be 
represented in such material. Cryogenic collection and preservation of grains 
ejected in jets may be one of the only ways to return "pristine" unaltered refractory 
grains to earth for analysis since grains collected near the surface may have 
undergone some degree of metamorphism within the cometary environment. 



Anders, E., 1988 in Meteorites and the Early Solar System , eds. J, Kerridge and M. 
Matthews (Univ. Ariz. Press, Tucson) Chap. 13.1. 

Bihuniak, P. P., 1983, J. Am. Cer. Soc. 66, C-188. 

Boden, G., Richter, E. and Wollschlager, K., 1984, Sillkattechnik 35, 149. 

Brown, S. D. and Kistler, S. S., 1959, J. Am. Cer. Soc. 42, 263. 

Clayton, D. D., 1980, Ap. J. 239 , L37. 

Clayton, R. N. and Mayeda, T. K., 1984, EPSL 67, 151. 

Combs, M. and 19 co-authors, 1986, Nature 321, 266. 

DeNatale, J. F. and Howitt, D. G., 1983, Proc. 41st Annual Meeting of the Microscopy 
Society of America (San Francisco Press, San Francisco) p. 354. 

Donn, B. D., 1981 in Comets and the Origin of Life , ed. C. Ponnemperuma (D. Reidel, 
Dortrecht) p. 21. 

Donn, B. D., 1987 in Exploration of Halley's Comet , eds. B. Battrick, E. J. Rolfe and 
R. Reinhard (ESA SP-250, Vol. Ill) p. 523. 

Duley, W. W., Millar, T. J. and Williams, D. A., 1979, Astrophys. Space Sci . 65, 69. 

Goebel. J. H. and Moseley, S. H., 1985, Ap. J. (Lett.) 290 , L35. 

Kratschmer, W. and Huffman, D. R., 1979, Astrophys. Space Sci. 61_, 195. 

Moore, M. H. and Donn, B. D., 1983, Icarus 54, 388. 

Nelson, R. N., Nuth, J. A. and Donn, B. D., 1987, Proc. 17th Lun. Plan. Sci. Conf. , 
1n J. Geophys. Res. , 92, E657. 

Nuth, J. A. and Donn, B. D., 1982, Ap. J. (Lett.) 257 , L103. 

Nuth, J. A. and Donn, B. D., 1983, Proc. 13th Lun. Plan. Sci. Conf. , in J. Geophys. 
Res. 88, A847. 

Nuth, J. A. and Donn, B. D., 1984, Proc. 14th Lun. Plan. Sci. Conf. , in J. Geophys. 
Res. 89, B657. 

Nuth, J. A., Moseley, S. H., Silverberg, R. F., Goebel, J. H. and Moore, W. J., 1985, 
Ap. J. (Lett.) 290 , L41. 

Nuth, J. A., Donn, B. D., DeSeife, R., Donn, A. and Nelson, R. N., 1986, Proc. 16th 
Lun. Plan. Sci. Conf. , in J. Geophys. Res. 91, D533. 


Nuth, J. A. and Moore, M. H., 1988a, Ap. J. (Lett.) 329 , L113. 

Nuth, J. A. and Moore, M. H., 1988b, Proc. 19th Lun. Plan.^Sci. Conf. , (LPI, Houston) 
in press. 

Rietmeijer, F. J. M., 1985, Nature 313 , 293. 

Rietmeijer, F. J. M., Nuth, J. A. and Mackinnon, I. D. R., 1986, Icarus 66, 211. 

Rowan-Robinson, M. and Harris, S., 1982, Mon. Not. Roy. Astr. Sec. 200 , 197. 

Saunders, R. S., Fanale, F. P., Parker, T. J., Stephens, J. B. and Sutton, S., 1986, 
Icarus 66, 94. 

Seab, C. G. and Shull , J. M. , 1985 in Interrelationships Among Circumstellar, 

Interstellar and Interplanetary Dust , eds. J. Nuth and R. Stencel (NASA CP 2403, 
GPO, Wash. DC) p. 37. 

Storrs, A. D., Fanale, F. P., Saunders, R. S. and Stephens, J. B., 1988, Icarus 76, 

Tsuji, T., 1973, Astr. Astrophys. 23, 411. 

Wagstaff, F. E., 1968, J. Am. Car. Soc. 51, 449. 

Wagstaff, F. E., 1969, J. Am. Cer. Soc. 52. 650. 

Wagstaff, F. E. and Richards, K. J., 1966, J. Am. Cer. Soc. 49. 118. 

Zaplatynsky, I., 1974, NASA TM X-2969. 

Zaplatynsky, I., 1988, NASA TM-101335. 

Zinner, E., 1988 in Meteorites and the Early Solar System , eds. J. Kerridge and M. 
Matthews (Univ. Ariz. Press, Tucson) Chap. 13.2. 



A. Bar-Nun 

D. Prialnik 

I. Kleinfeld 

D. Laufer 

Department of Geophysics and Planetary Sciences 

Tel Aviv University 

Tel Aviv, Israel 



A. Bar-Nun, D. Prialnik, I. Kleinfeld and D. Laufer 

Dept. of Geophysics and Planetary Sciences 
Tel Aviv University, Tel Aviv, Israel 


The realization that water ice at low temperatures is the major constituent of 
comets, the satellites of the outer planets and their rings particles, and of icy 
grain mantles in dense interstellar clouds, prompted in recent years a number of 
groups to study experimentally the properties and behavior of ice at very low 
temperatures and apply their findings to icy bodies. Among them are the groups of 
Delsemme, Donn and Moor, Fanale , Greenberg, Ibadinov, Kajmakov, Klinger, Mayer, 
Miller, Rossler, Sanford and Allamdola, Schmitt and Wallis. Because of the 
limitation on the space of this article, the over 100 works of these groups will 
not be included in the list of references. 

In most of these experiments, several microns to several cm thick ice samples 
were studied under, presumably, isothermal conditions. Thus, almost no experimental 
data exists on the processes by which a heat wave penetrates into the ice, the 
formation of a dust crust by partial water sublimation and other large scale 
phenomena. The works of Kajmakov, Ibadinov and their groups were aimed mainly at 
the formation of a filamentry residue when the ice sublimated completely, as were 
the works of Storrs, Fanale and Stephans. A major step in the direction of 
measuring bulk behavibr of large (-15 cm thick, -30 cm diameter) ice-gas-dust 
samples was made in recent COST (Comet Simulation) experiments. These results 
will be described by Gr\in in this voltome. 

All this experimental and theoretical effort would not have led anywhere 
without some knowledge of the structure, composition and behavior of icy bodies. In 
this field, major advances were made by the two Voyagers' flybys among the icy 
satellites of Jupiter, Saturn and Uranus, with Neptune yet to come. A tremendous 
advance in our knowledge of comets was obtained by the six spacecrafts which studied 
comet Halley, with Giotto and Vega contributing most of the new and surprising data. 
Only the combined efforts of the spacecrafts experimenters, modellists and 
laboratory experimenters could lead to the far better londerstanding of comets which 
we now have. The next major advances will certainly occur during the coming CRAF 
and ROSETTA missions. 

In what follows, because of space limitations, we will describe only the 
experimental results obtained at the Comet Simulation Laboratory of Tel Aviv 
University (1-5), as well as our thermal models (6-10), which incorporate the 
experimental results. These will be applied to the temperature and composition of 
the nebula in the region of comet formation; the timescale of comet formations; the 
small explosions observed on Comet Halley and the formation of its large active 
craters, as well as to Miranda's chaotic terrain; the possible contribution of 
comets to the noble gas inventory on the terrestrial planets and finally, to a 
detailed model of the thermal evolution of comet P/Temple-1. 



The experimental setup and procedures were described in detail in our papers 
(1-6) and will therefore be described here only very briefly. A premixed gas -water 
vapor mixture was flowed onto a cold plate at 20- lOOK and was co- deposited on it, 
at a pressure of -10'^ Torr, for -45 min. , through a capillary tubing with a 
diffuser on its tip. When an ice layer containing 10^^ -10^° water molecules was 
formed, the chamber was pumped for -10 min, until a constant pressure of -10'* Torr 
was reached. The plate was then uniformly warmed, at a constant rate of 0.1 -IK 
min' ^ . 

The evolution of gas and water vapor from the ice was monitored by a 
precalibrated quadrupole mass -filter and the amounts of gas and water vapor emerging 
at each temperature range were obtained by integrating their fluxes over the time of 
their evolution from the ice. The sensitivity of the mass-filter spanned 8-9 orders 
of magnitude of gas flvix. Thus, very subtle changes in the ice, which were 
manifested by gas evolution, could be detected. During the evolution of gas at the 
various temperature ranges , spikes of water and gas were monitored by the quadrupole 
mass -filter. Each spike corresponds to a gas jet or to an ice grain, which enter 
the mass filter's probe. 


A. Structure of the amorphous ice (3-4) 

When a layer of amorphous ice is viewed from the edge or at an oblique angle, a 
hairlike structure is revealed (Fig. 1). The ice looks like a shaggy woolen carpet, 
which explains its very large surface area - 86 m^ g'^- Tl^e amorphous ice needles, 
which grow by ballistic deposition of water vapor, are smooth, with no side 
branches, unlike the dendritic structure of hexagonal ice, which grows on a cold 
surface by diffusion of water vapor in air. The same structure is formed with water 
vapor deposition rates between 10^^ and 10^^ water molecules cm"^ min' ^ , at 
temperatures between 20 and lOOK. Below this deposition rate, the ice layer is too 
thin for anything to be seen. It is estimated that about 80% of the ice is in the 
form of needles. The ice needles are not altered during the evaporation of the ice, 
between -140 and -180K and retain their shape until their complete evaporation. 

B. Gas trapping in the ice and its release upon warming (3-4) 

The trapping of various gases in amorphous ice and their stepwise release when 
the ice is warmed up, was used to probe the ice's structure and dynamics. A typical 
plot of gas release from ice which was co-deposited with argon (HjO water vapor 
Ar=l:l) is shown in Figure 2. Eight regions of gas release, labelled a-g can be 
seen. The first one (a) is due to the evaporation of gas which was not trapped 
internally, and froze on the surface. Range (c) which starts at 44K is due to the 
desorption of the last monolayer of adsorbed gas from the surface . From the amount 
of gas released in range (c) , a surface area of 86 m^ g'^ is found for ice at 44K, 
diminishing to 78, 55 and 38 m^ g'^. for ice which was deposited at 75, 100 and 
120K, respectively. Ranges (b) , (d) and (d') which start at 35K -85K and -120K, 
respectively, are attributed to three distinct annealing processes in the amorphous 
ice. Each one proceeds in a stepwise, temperature dependent, manner. Thus, if the 
temperature is kept constant for tens of hours in the middle of these three ranges , 
gas emission declines slowly, until it reaches the limit of detections - 10' 
molecules cm"^ sec"^. A temperature increase by a mere 2-3 degrees results in the 


Fig. 1 - A view of an amorphous ice layer, seen at an oblique angle. Needles about 
0.1 nun long are seen, with an ice layor containing -10^° water molecules/cm^ . 








11 - 

10 H 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I I 

20 40 60 80 100 120 140 160 180 200 


Fig. 2 - A plot of the fluxes of evolved argon and water vs temperature, 
representing the eight ranges of gas evolution. The fluxes vary by up to 8 orders 
of magnitude as the ice temperature varies. The rise in the water flux at 30K is 
due to ice-grain ejection. An evaporation curve of frozen argon from an ice-free 

plate ( ) is added for comparison. 


resumption of gas emission at the level reached before the temperature was kept 
constant. Range (e) is due to the release of gas during the exothermic 
transformation of the amorphous ice into cubic ice. The onset of the transformation 
occurs at 137K. If the ice is kept at a constant temperature anywhere in range (e) , 
all the gas which was supposed to come out in this range leaks out and, unlike in 
ranges (b) , (d) and (d' ) , a slight increase in temperature does not cause another 
surge of gas. At 160K, the cubic ice transforms into hexagonal ice, in an almost 
thermoneutral process. During this transformation, an additional portion of the 
trapped gas comes out in range (f ) . Gas release in range (g) , which follows, is 
symmetrical with the evaporation of the ice (Fig. 1) . It is attributed to the 
evaporation of the clathrate hydrate, where the water cages evaporate, releasing the 
gas molecules which were trapped in each cage. 

Very large fluxes of gas jets, each containing about 5x10^° gas molecules, and 
ice grains, each containing about 10^° water molecules, are emitted from the ice 
under a variety of conditions (Fig. 3): whenever Ar, CO, CH^ , N2 and Ne , but not 
Hj or D2 , are emitted from gas- rich ice; during the deposition of a 1:1 gas- water- 
vapor mixture below 29K; and during gas flow into pure amorphous ice, below 29K at 
a pressure exceeding 2.6 dyn cm' 2. With an ice layer containing -10^^ water 
molecules cm"^, the ice needles are -1 /sm long and -0.2 /xm wide, and their speed is 
at least 1.67 m sec"^. In addition to the large gas jets, all gas emission in the 
various ranges is not quiescent, but consists of numerous minijets, -100 times 
smaller than the large ones, which result in a "noisy" gas signal at a frequency of 
-10^ sec"*. Whenever a large fltjx of ice grains is emitted, the ice remaining on 
the plate is not hairlike, but has a very rough texture. Therefore, it seems that 
the ice grains are indeed the ice needles, which are broken from their base and are 
propelled by the large gas jets. The frequency of ejection of the needles is 
correlated with the amount of gas trapped in the ice . 

Up to 63% (by number) of Hj , Dj and Ne are trapped in the ice at 16-25K (Figure 
4). Three ranges of gas release are observed for these three gases: 16-35K, 35-85K 
and 85-120K. The second and third ranges of gas release correspond to ranges (b) 
and (d) , whereas the first range corresponds to the penetration of the narrow Hj and 
Dj molecules or the somewhat bigger Ne atom into the hexagonal channels of the ice. 
According to Hobbs (12), in cubic ice, the hexagons which are formed by the oxygen 
atoms have the configuration shown in Figure 5. With a radius of 1.4 A between 
adjacent oxygen atoms, the free channel through each hexagon has a diameter of 2.5 
S. The diameters of Ar, Ne , H and D atoms are 4.7, 3.8 and 2.2 A respectively. 
Therefore, hydrogen and deuterixjm molecules, but not neon or argon, can penetrate 
freely the lattice of cubic ice . Only the antorpbous ice lattice , in which the 
distances between adjacent oxygen atoms are larger, is permeable to neon, and some 
channels are even large enough to allow argon atoms to penetrate the ice lattice. 

This is clearly demonstrated by the reversible trapping in the first range, of 
H2 and Dj in cubic ice, which was cooled to 19K, and the failure of Ne and Ar to be 
trapped under these conditions. Yet, considerable amounts of H2 and D2 are held in 
the ice until ranges (b) and (d) are reached, even though the ice matrix is open to 
them. Apparently, some of the open channels are blocked by domains in the ice which 
are perpendicular to each other. Hence the amounts of Hj and Dj which are emitted 
in (b) and (d) versus the amount in the first range, provide a measure of the 
fraction of blocked domains in the ice. Of the 0.63 Hj molecules per water molecule 
in the ice, 0.42 reside in open channels and 0.21 reside in blocked regions, which 
open when the ice anneals in ranges (b) and (d) . 

As for the larger argon atoms, when gas is flowed into preexisting amorphous 











Fig. 3 - (top) A plot of the fluxes of water vs temperature, from a co-deposition 
of a 1:1 Ar-H,0 mixture at 25K. showing the ranges of ice-grain ejection. The 
timescale in this measurement did not allow the separation of single grains, which 
is shown at the bottom. 



1/ ■ 



// 1 w \ 


U / vv \ 


/ v^ \ 


/ 'v \ 





\ ^''-^,.._^_^ — --v,,^^^ 

^_^ - — — .....^^^^ 


^^ ^^"^^-vj^e ^^"""-^^ 

^^--^^ ^^^ 

^""-.^^^^^ ^""-C^ ^'*"*-*-^_^..-^ 

j^"^ ^v..^^^^ V 





\ \\ 



\ \\, 


\ Ni. 



1 1 1 1 1 ) 1 1 


^ig- 4 - A ploc of rhe fluxes of evolved H, , D, and Me vs cemperacure. from co- 
daposicion, ac I0-I8K, of 1;1 gas-HjO mixcures'. 



rij. 5 - (13?) Tha cotiiiguracion of 1 hexa^or. of oxygen icons in 1 cuoic ice 
la::tce (af:ec Ref. 12) Tao hydrogen acorns (noc shoxm) are aajacenc Co each 

oxygen acorn. The dtaaecer of Che free channel chrough Che hexagon is 2.5 .-». 
(boccon) a scnemacic drawing of Che hexagonal channel in cryscalline and anorphous 


ice, the channels are blocked by the constant annealing of the ice, whereas in co- 
deposition of gas -water-vapor mixtures, the underlying gas -filled channels are 
blocked by the added layers of ice as well as by the annealing of the ice . Thus , 
the argon atoms are locked inside the wide channels in the amorphous ice matrix, 
either completely isolated from each other or in small clusters. Upon warming, when 
the ice anneals, the rearrangement of the water molecules opens some of the blocked 
channels. However, only when a channel opens all the way to the ice surface can the 
gas in it escape out. Thus, a dynamic percolation behavior results, where only a 
connection of many gas- filled domains in the ice, up to the surface, forms a 
minijet. While the larger gases (Ar, CO, CH^ , Nj and Ne) emerge in minijets, the 
smaller H2 and Dj molecules emerge continuously and reversibly, since the channels 
in both amorphous and cubic ice are wide enough for them to get through without 
obstruction. The large jets are formed when many small channels open 

simultaneously, not to the surface, but to a weak region in the ice, where the gases 
build-up a pressure large enough to rupture the overlying ice. This rupture breaks 
the ice needles from their base and the outflowing gas propels them. 

C. Competition among CO, CH^ , N, and Ar (5) 

Up till now we discussed the trapping of siagle gases in the ice, where the gas 
could occupy all the available sites in the ice. With gas mixture, however, the 
gases have to compete on these sites. With co-deposition of a mixture of HjO: CH^ : 
Ar: CO (or Nj) - 1: 0.33: 0.33: 0.33, the results shown in Figure 6 were obtained. 
It can be seen that all three gases come out at exactly the same temperature ranges, 
but differ in their qviantities . The quantities of trapped gases from this mixture 
as a fiinction of deposition temperature between 28 and lOOK are shown in Figure 7 . 
It can be clearly seen that, while at 28K all four gases are trapped in the ice in- 
discriminantly , at 50-100K there are strong enhancements among the gases. These 
enhancements are observed with a 1:1 gas -water vapor mixture, where there is a 
competition among the various gases on the available trapping sites in the ice. 
With a gas-poor mixture (water vapor: gas = 100) , the gases are trapped in the ice 
exactly in their proportion in the gas mixture, because there are sufficient 
trapping sites in the ice. The reasons for these enhancements include factors such 
as gas- water molecules interaction energy, size of the trapped gas atom or molecule 
and type of clathrate -hydrate formed (I for CH^ and CO and II for Nj and Ar) . Note 
that between 28 and lOOK, the total amoiint of trapped gas in the ice drops 
exponentially by many orders of magnitude. This sharp temperature dependance, 
together with the gas content found for comet Halley, might serve as an indicator to 
the temperature in the region of comet Halley' s formation: several percent of 
trapped gas may suggest a formation temperature of about 50K (Figure 7) . 

This formation temperature is only for condensation of water vapor in the 
presence of various gases and their subsequent trapping in the amorphous ice , which 
is very temperature sensitive. The question then arises whether ice which was 
formed at higher temperatures and, consequently, trapped very little gas, was cooled 
in the solar nebula to 50K or even lower and then trapped some additional gases , 
bringing the gas content to the observed value. This situation was studied by us as 
well. In these experiments, pure water ice was condensed from the gas phase at 
various temperatures, was cooled down to 25K and gas was then flowed into it. These 
experiments show very clearly that the amount of gas which is trapped in the ice 
depends only on the highest temperature which the ice reached at some time and not 
on the temperature at which the gas was flowed into it. This is because the 
amorphous ice anneals in a stepwise, temperature dependant process, which closes a 
fixed fraction of the available trapping sites at each temperature. The conclusion 
from these experiments, regarding the formation of comets, is straightforward and 








50 70 


110 130 150 










Fig. 6 - A.ploc of Che fluxes of gas And vacer vapor vs zetaperacure . Tne gas-ricn 

aaorphous ice was condensed ac 50K. from a ci,0; CH, ; CO:Ar - 1 :0 . 23 . 33 :0 . ;3 
aiixcure. The various ranges of gas evoLucion (b>-(gj are laoeled. Soce :he 
changes Ln gas evolucion over 6 orders of aia^niCudc . 


1 i 

-A § 


1 1 

1 i 

1 1 1 


^ o 






1 1 

1 1 

1 1 

1 1 1 







Deposition Temperolure (°K) 

rig. 7 - The rocai araouncs of crapped gases ^n daposicion cemperacure of che vacer 
vapor-gas siixcures. - CH,. i - CO. • S- and - Ar. Open symbols represenc 

The resuics of che deposition of 1:1 single gas-wacer vapor. Solid symbols 

represenc che resxiics of che depoSLCion of gas aixcures H-0:CH, :Ar:CO (or N*- ) - 1:0. 
33:0.33:0.33. £ach poinc is an average of ac lease cwo experiiDencs . (Some poincs 
were moved slighcly on che cearparanure acaXe . co prevenc overlapping;. 


will be discussed below. 

Since the Giotto, Vega, lUE and sounding rockets observations all showed that 
CO is the major gas component of comet Halley, and it is more abundant than CH^ by a 
factor -3.5, (13) we focused our attention on the CO/CH^ ratio in the gas which was 
trapped in the ice, at 50K, as a function of this ratio in the gas mixture. The 
amount of CH^ which is trapped in the ice at 50K drops below that of CO only when 
the CO/CH4 ratio in the gas mixture is increased to 100. The implication to the 
composition of the solar nebula is obvious, and will be discussed below. Hydrogen, 
diluting 35 times a 1:1 gas: water vapor mixture, had no effect on any of the gases. 
Hydrogen itself is trapped in the ice only below 20K. 

D. Competition among the noble gases Ar, Kr and Xe 

The competition among the noble gases was studied, with a mixture containing 
Ar:Kr:Xe - 10,000: 8:1, similar to the solar ratio of 20,000:8:1 (Table 2). This 
mixture was co-deposited with water vapor (1:1) between 30 and 75K. The enhancement 
factors for Kr/Ar and Xe/Ar (the ratio of the amounts of trapped gases divided by 
their ratio in the gas mixture), as a function of deposition temperature, are shown 
in Table 1. A hundredfold dilution with CO did not affect these enhancement 
factors . 

Table 1 
Trapping of an Ar:Kr:Xe = 10,000:8:1 mixttire at 50-75K 
resulted in the following enrichment factors 






















The possible implications of these, temperature dependent, enrichment factors to the 
noble gas distributions on the terrestrial planets will be discussed later. 


Our experimental work was complemented by models, which describe the heating of 
comets and icy satellites by solar radiation or by the radionuclides in their dust. 
The experimental results on gas release were, of course, incorporated into these 

A. Thermal models of pure ice (6) 

Assvime that at a given time t a cometary nucleus is a fast rotating perfect 
sphere of mass M(t) and radivis R(t), consisting of water ice, either amorphous (at 
low temperatures), or crystalline. 

If the mass m is chosen as the independent space variable (0<m<M) , the 
structure of the nucleus will be given by the following time dependent functions : 
the temperature T(m,t), the density p(m,t), and the mass fractions of amorphous (A) 
and crystalline (C) ice X^(m,t) and Xc(m,t) respectively. In a pure ice nucleus, X;^ 
+ Xc = 1, but, in principle, other constituents of the nucleus may be included, such 


as gases, radioactive isotopes or dust. 

The rate of change of the internal energy at each point of the nucleus is 
determined by the heat flux F(m,t) and the local energy sources (or sinks) q(m,t): 

au(m,t)/at = aF(m,t)/am + q(m,t), (1) 


F(m,t) = /c(T)aT(ni,t)/am ergs s'^ (2) 

with F(0,t) = d (3) 

at the center, and at the surface: 

F(M,t)=47rR(t)2[(l - A)S<cos ^>/dH(t)2 - £crT(M,t)«], (4) 
where A is the albedo, S is the solar constant, <cos 4>> is the average value of the 
local solar zenith angle, which will be taken as 0.25 (the value corresponding to a 
uniformly illuminated sphere) , d^ is the heliocentric distance (in AU) , e is the 
emissivity, and a the Stefan- Boltzmann constant. 

The crystallization reaction, which sets in when amorphous ice is heated to 
136. 8K, is practically instantaneous and may be described by 

aXA(m,t)/at - -AXA(m,t), (5) 

where the time constant X'^ should be short (on the order of seconds to minutes) for 
T > 136. 8K and infinite for T < 136. 8K. The definition adopted for X was 
/ 1/10 s-^ T > 136. 8K, 
^ = ^ 0, T < 136. 8K (6) 

Equations (1) - (5) , constitute the evolution equations for the cometary 
nucleus, to be solved numerically for an assumed orbit, i.e. a given function djj(t) . 

The orbit of comet P/Halley was chosen. Other relevant parameters were: R = 
2500 m, p -= 0.7 g cm'^, porosity p= 0.3, A =0.1, £=0.5 and the latent heat of 
sublimation H — 2490 J"g"^. The density and porosity were assumed constant and 
uniform throughout the calculations. The gas content of comet Halley (gas/water > 
0.2), together with our experimental results, suggest that Halley-type comets were 
formed in a region of the solar nebula where the temperature was -50K. This is well 
below the transition point from amorphous to cubic ice (136. 8K ± 1.6K). Hence, it 
seems quite certain that the water ice of cometary nuclei was formed in the 
amorphous state. In the region where the comets spent -4.5 x 10^ yr, they had ample 
time to cool down to the ambient temperature of -lOK. Therefore, the initial model 
for the evolutionary calculations was an amorphous ice sphere at a uniform 
temperature of lOK. 

B. Thermal model with dust (7) 

In ntimerical simulations of dust-covered cometary nuclei, it is generally 
assumed that the water vapor may escape freely through the dust mantle. Indeed, in 
the experiments of Storrs et al. the entire dusty ice ball evaporated, although more 
slowly than from a ball made of pure ice, leaving behind the fluffy network of dust. 
Thus, even if the entire comet is enveloped in a dust mantle, the cometary activity 
is not completely quenched. The dust layer, if sufficiently thin to permit easy 
diffusion of water vapor and gases, provides mainly thermal insulation. The main 
question addressed is, then, to what extent would the secular cometary activity be 
impaired by a growing, permanent dust mantle, permeable to water vapor? The initial 
stages of mantle formation would proceed as follows: Consider an outer layer of 
mass Am of a spherical cometary nucleus . Assiime that a fraction X<j of the mass is 
dust and the rest Xj^ = 1 - X^j is water ice and other volatiles . The ratio X^j/Xi in 
comets is of the order of 1. When the ice sublimates and the gases escape, the 


small dust particles are carried away with the gas flux. Only a fraction rj of the 
dust mass Am(j — Xd.Am will be stored on the surface, forming a permanent dust mantle. 
It can be envisioned that rj will increase with time, as more dust particles will 
stick to the thickening dust mantle. This process might be augmented if the dust 
particles are covered by a layer of organic material which, when initially cold, 
behaves like the minerals but, when warmed up near the surface, becomes sticky. The 
variation of the dust mantle's thickness and the rate of erosion of the nucleus with 
repeated revolutions may be obtained by numerical simulations of the evolution of a 
cometary nucleus in a given orbit for given r?. The initial model for the one- 
dimensional ("fast rotator") numerical simulations is a homogeneous sphere of ice 
and dust, with an initial radius of 2500 m (the assumed radius is of no significant 
consequence) . The crucial parameters of the model are the density of the nucleus 
p-^, the mass fractions of ice and dust Xj^ and X^, respectively, and rj - the fraction 
of dust mass which is assumed not to be blown away. Evolutionary sequences were 
calculated for different combinations of these parameters, to be compared with the 
pure ice model. The method of computation was described above, with some changes, ( 
e.g. heat capacity, thermal conductivity) which take account of the dust. 

C. Solar heatitiR and crystallization (5,7) 

A major feature of the comet model is the penetration of a crystallization 
front inward. When, due to solar heating, the outer layer of amorphous ice reaches 
136. 8K, it crystallizes into cubic ice. The energy released by this transformation, 
90 J gr"^, further heats the ice to 160K, where it transforms into hexagonal ice 
and, also, drives the transformation wave inward, until a layer is reached, where 
the temperature is lower than 136. 8K. The next transformation event occurs when the 
surface has sublimated far enough so that solar heating will raise the temperature 
of an underlying amorphous ice layer to 136. 8K, thus triggering another wave of 
transformation. A typical distance between the surface and the underlying amorphous 
ice botmdary at which transformation can be triggered is 15 m. 

Due to the insulating effect of the dust mantle, the rate of erosion of the 
surface is much slower than in the case of a bare nucleus and is constantly, 
although slowly, decreasing. The inward propagating heat wave, generated by the 
onset of crystallization, is weaker than in the case of pure ice. An appreciable 
fraction of the absorbed solar energy is absorbed by the dust, and only part of the 
energy induces further crystallization. Consequently, the thickness of the 
crystalline ice layer varies between -15 and -40 m. These findings are summarized 
in Figure 8, where the variation with time, over many orbits, of the outer radius R, 
due to sublimation and of r j. , the boundary between the outer-crystalline ice layer 
and the inner -amorphous ice core, are shown for a comet in Halley's orbit. Note the 
effects of ice density (p) , dust to ice ratio (X^j/Xj^) and the fraction of dust which 
remains permanently on the surface (r;) . The variation of temperature with 
heliocentric distance in comet Halley's orbit at the dust surface (T^) and the ice 
surface (Tg) just below the dust mantle are shown in Figure 9. T^^x "° 400K, is the 
temperature reached by the dust surface at perihelion. 

D. Radiogenic heating (8) 

Both comets and icy satellites have considerable amounts of dust mixed with 
their ice. in the dust are incorporated radionuclides such as *°K, ^^^Th, ^ssy and 
238U, all with decay constants of the order lO^i" - lO'^^ yr"i. In addition, the 
dust is very likely to contain 26^]_^ (whose decay constant is 9.37 10"^ yr"^) as 
suggested by the Mg excesses in the Allende meteorite and, independently, by the 7- 
ray detectors on the HEA03 and SMM satellites. The radioactive heat output per gram 



24 50h 







2 500 



24 50 





2 350 

ice ona dust 
P = 055 qm cm"' 
i„/x, =0 2 
7 = 0001 

20 30 40 10 20 30 



Fig. 3 - Variation vich rime of zhe oucer radios R (inicially 2.5 lea) and =he 
boundary r _ becueea amorphous and cryscallinA ice vlchin a comec nucleus aodcL. 
the orbLc assvased is coaeu ciaLlay' s and Che else is measured in unics o£ orbical 
periods. (a; corresponds :o pure ice models o£ different densicies: 0.7 g cm*' 
Csolid lines) and 0.2 g c='' (dashed lines). (b) , (c) and (d) correspond Co aodels 
oc nixed cosiposicion; zhe densicy, dusc :o ice mass racio »iui value of 7 are 
indicaced. The dashed line in (b) represents che oucer radius of a pure ice model 
of sizLilar properties. 



-' 180 






Q. 140- 




-\ — I I I I 1 1 

T r- 


I I I I I I I 

_1 ' ■ 

5 12 5 10 20 30 


Fig. 9 - Variation of ceoperacures ••rith heliocencric dtscance in comcc HaLiey' s 
or&ic cor che dusc surface T^ (solid Line) and zhe ice surface T^ below :he dusc 
aancle (dashed Line) . '^•jmx ^^ ^^ naxiarum cemperacure reached by che dusc surface 
ac perihelion. 


of comet nucleus material as a function of time is shown in Figure 10. The dust to 
ice ratio is 1 and the mass fractions of the radionuclides: X(^^Al) = 5.6x10'®, 
XC^°K) = 5.65 10-'^, X(232xh) = 2.76x10-8, xC^s^U) = 3.13x10-9 and XC^a^U) = 
1.09x10- 8. Contrary to solar heating, which propagates from the surface inward, the 
heating of comets or icy satellites by the decay of radionuclides is most effective 
at the center, because the outer layers get rid of the heat more easily through the 
surface. Therefore, in an icy body made of amorphous ice, once a temperature of 
137K is reached at the center, a wave of transformation from amorphous into 
crystalline ice will propagate outward. Similarly, melting will also begin at the 
center, if enough heat is generated. 


A. The temperature in the region of comet Halley's formation (11) 

The results of the measurements of the Giotto and Vega spacecrafts , together 
with our experimental findings on the amount of gas trapped in ice at various 
temperatures, can set a rather strict limit on the temperature of the solar nebula 
in the region of comet Halley's formation: From the latest results on gas emission 
from comet Halley, as siimmarized by Lammerzahl (13), CO is the major gas released 
directly from the nucleus . Its amount, relative to water, is 5-7%. In order to 
trap in comet Halley's ice this much CO, its formation by condensation of water 
vapor in the presence of CO, had to take place in a region of the solar nebula where 
the temperature was 48±2K, as seen from Figure 7. The very steep slope of the curve 
of the amoiint of trapped gas vs deposition temperature sets a strict limit on this 
temperature. As shown in section IIIC, the condensation of water vapor into 
amorphous ice could not have taken place at a higher temperature, with subsequent 
cooling and flow of gas into the ice, since in this case, as shown experimentally, 
the amount of trapped gas depends only on the highest temperature at which the ice 
was formed or resided. 

The experimental results on the amount of gas trapped in amorphous ice which 
was deposited at various temperatures from a 1:1 gas: water vapor mixture, are 
directly applicable to the gas content of comet Halley, since in the solar nebula 
the ratio of CO or CH^ to water vapor was near unity, as shown in Table 2. 

Table 2 

Abundances of gases in the solar nebula , relative to Hj , 

calculated from the atomic abtindances of Anders, F. and Ebihara, M. 

(Geochim. Cosmochim. Acta. 46, 2363-2380, (1982)) asstiming that the 

molecixLes listed are the major ones in the nebula 

All C 


; CH4 

All C as CO 




























Xe 3.2 10-1° 

*If other oxygen bearing molecules, such as MgO, are taken into account, the HjO 
abundance should be lower by about 20%. 


7 8 9 

LOG [time (years)] 

Fig. 10. - R-adioaccive heat output per gram of comec nucleus material as a function 
of time. The mass fraction of ^«A1 is 5.6xlO-»: those of *°K. ^^jjh. 2350 ^nd 238^ 
are given in the text. Before z = 10' yr. when the decay of ^*aJ. is the main 
energy soxircc, the contribution of the other radioisotopes is shown by the dashed 
line. Lower curve shows the heac ourpuc due to the heavy elements alone. 


A formation temperature for comet Halley of 48K is surprisingly close to the 
temperatures observed by IRAS for the circumstellar dust shells around a PsA (55K) 
and £ Eri (45K) . The temperatures of the dust shells around q Lyr (Vega - 85K) and 
/3 Pic (lOOK) are considerably higher, and comets formed there would have gas/ice 
ratios of 10"^ - lO'*. The agreement between the comet formation temperature which 
was deduced from the present experimental study, and those observed for two out of 
four circumstellar dust shells, lends credence to the suggestion that Halley type 
comets were formed outside the region of planet formation. Moreover, Duncan et al . 
(14) , Weismann (16) and Delsemme (15) propose that the orbits of short and long 
period comets can be explained by two different regions of foirmation: the short 
period comets were formed in the extended dust shell (the "Kuiper belt") at -40K and 
remained there unpreturbed, whereas the long period comets were formed in the Uranus - 
Neptune region, at -80K and were ejected from there into the Oort cloud and, 
simultaneously, into the inner region of the solar system. Consequently, a much 
larger flux of high temperature comets, as compared with the flxix of low temperature 
comets, can be expected on the Terrestrial Planets. A formation temperature of 80K 
for the long period comets should mean that their gas content is in the -10'* range, 
as compared with 10" ^-10"^ for the short period comets. 

B. The gas composition in the region of comet Halley 's formation (11) 

With a ratio of CO or CH4 to H2O in the nebula being close to 1 (Table 2) , the 
regime of competition among the various gases on the available trapping sites in the 
amorphous ice should be applied. As discussed in section III C, at 50K, the 
trapping of CO overcomes the trapping of CH^ only in a mixture with CO: CH^ = 100. 
Thus, in order for CO to be the dominant gas in the icy nucleus of comet Halley, the 
CO/CH4 ratio in the region of comet formation had to be > 100. Simonelli et al . 
(17) also reach the conclusion that CO comprised 75- 90% of the carbon in the outer 
solar nebula, if Pluto and Charon were formed directly from the solar nebula. 
Lunine (18) also reaches the conclusion that the CO/CH^ in the solar nebula was 
100. Knacke et al . (19) set a limit of CO/CH^ > 100 in several molecular 
clouds . 

C. On the timescale of comet formation and thermal evolution 
of icy satellites (8) 

A study of the radiogenic heating of icy bodies by ^^Al, ''°K, ^^^Th, ^asy ^^^^ 
^U (section IV D) , showed that in order to maintain the cometary ice in the 
amorphous and, hence, gas-rich form, the main heat generating radioisotope - ^^Al - 
had to be in an initial abundance of <4xl0"^. This initial abundance is a 100 times 
lower than that in the Allende meteorite. Hence, a lower limit of -5x10^ years for 
the formation time is implied, in order for the ^^kl to decay from its initial 
abundance. In addition, the coexistance of molten cometary cores and extended 
amorphous, gas- rich, ice mantles is ruled out. 

Larger icy spheres (r > 100 km) reach the transformation temperature even in 
the absence of ^^Al, due to the decay of the other radionuclides. As a result, the 
outermost icy satellites in the solar system, which might have been formed from ice 
in the amorphous state, have probably undergone crystallization. Since this process 
proceeds from the center outward, huge amounts of trapped gases are pushed in front 
of the outward propagating transformation wave. This could give rise to eruptive 
activity when the gas is released near the surface , and to chaotic terrain such as 
observed on Miranda. 



D. Coiild have comets provided the terrestrial planets vltji their noble gases? 

The experimental results on the trapping of Ar, Kr and Xe in amorphous ice are 
discussed in section III D. The enrichment factors Kr/Ar and Xe/Ar, as a function 
of deposition temperatures are shown in Table 1. On Earth, these enrichment factors 
are: Kr/^*Ar = 74 and Xe/^^Ar = 48. From Table 1, they can be obtained (from an 
Ar:Kr:Xe -= 10,000:8:1 mixture) by co-deposition this noble gas mixture with water 
vapor at 52-55K, quite close to the 48±2K, which was deduced to be the formation 
temperature of comet Halley, from its CO content. Thus, the Earth could have 
obtained its entire budget of Ar, Kr and Xe from a single hit by a R = 42 km, p = 
0.5 g cm'^ comet, which was formed around 50K. Mars, which has a similar Kr/^®Ar 
ratio, could have been supplied with its noble gases by a smaller comet which, 
still, had to be formed at about 50K. On Venus, the Kr/^^Ar and Xe/^^Ar ratios are 
solar and the abiindance of ^®Ar is 2-3 orders of magnitude larger than on Earth. 
The solar ratio could have resulted from a hit by a comet which was formed at 30K, 
since at this temperature the noble gases are trapped in the ice in their proportion 
in the gas mixture. (Table 1). Moreover at 30K, Ar, Kr and Xe are trapped 2 to 3 
orders of magnitude more efficiently than at 50K (Figure 5). Hence, the larger 
amounts of these gases on Venus . 

In conclusion, the Ar, Kr, Xe patterns on the Terrestrial Planets could have 
been obtained by a single hit of each planet by a comet which was formed at either 
-30K (Venus) or -50K (Earth and Mars). A single hit is required since, if many hits 
are required, mixing between 30K and 50K comets could not be avoided. These low 
temperature comets came from the stable Kuiper Belt, from which a large flux of 
comets should not be expected. The large flxix of comets which hit the Terrestrial 
Planets and the icy satellites originated in the Uranus- Neptune region, from which 
they were scattered both inward and outward, into the Oort cloud. These, however, 
were formed at 80- 90K- and their noble gas content was 3-4 orders of magnitude 
smaller (Figure 6) . 

E. The activity of cometary nuclei (6,7) 

The experimental results on gas release from the ice upon its warming, were 
combined with the modelling results on the penetration of a heat wave into the 
nucleus, to produce a reasonable picture of the activities observed on comet Halley: 

In our dusty comet models (section IV C) an outer layer of crystalline ice, 
about 15-40 m thick, was found to overlay the inner amorphous ice bulk of the 
nucleus. With comets formed at 48K, most of the trapped gas (except the clathrate- 
hydrate) is released from the ice in ranges (e) and (f) - during the transformations 
of the amorphous ice into cubic and then hexagonal ice (Figure 2) . As shown 
experimentally, with only a -100 /xm thick ice layers (section III B) , gas emission 
from the ice cannot proceed freely and is accompanied by the ejection of ice grains, 
which are propelled by gas jets. A 15-40 m thick crystalline ice layer would 
certainly block gas escape more efficiently than a 100 /sm thick ice layer. Hence, 
it can be easily envisioned that gas filled pockets will be formed in the ice, when 
the gases will be released from the amorphous ice upon its transformation into cubic 
and hexagonal ice. These pockets could be permanent or form temporarily by dynamic 
percolation processes. When the pressure in such a pocket exceeds the tensile 
strength of the overlying crystalline ice layer, an explosion will occur. Numerous 
such explosions were indeed obsejrved on comet Halley both from the ground and by 
lUE. In one (20) an amount of gas equivalent to a frozen 30x30x10 m^ chunk was 
observed. These dimensions are in good agreement with our calculation of a 15-40 m 


thick crystalline ice layer. 

Such an explosion will remove the overlying, choking, dust layer. Moreover, 
its vibrations might trigger the explosion of adjacent gas- filled pockets. Thus, 
by enhanced evaporation of a single small crater or by combination of several 
adjacent small craters, a large crate such as observed on comet Halley can be 

F. The possible formation of a hydrogen coma aroxmd comets 
at large heliocentric distances (9) 

An experimental test - the detection of a hydrogen coma around comets at large 
heliocentric distances - can be proposed for determining whether comets were formed 
by the agglomeration of unaltered, ice-coated, interstellar dust grains. The 
laboratory experiments (section III B) showed that amorphous water ice traps H2 , Dj 
and Ne below 20K and does not release them completely vmtil the ice is heated to 
150K (Figure 4). Gas/ice ratios as high as 0.63 are obtainable. Thus, if the ice- 
coated interstellar grains were not heated above -llOK prior to their agglomeration 
into cometary nuclei, the inward propagating heat wave, when the comets approach the 
sun, should release from the comets a continuous flux of molecular hydrogen. This 
flux would exceed that of water molecules (and, hence, Hj production by photolysis) 
at -3 AU preperihelion and -4 AU post-perihelion. 

G. Thermal evolution of comet P/Temple 1 (10) 

Comet P/Temple 1 was chosen for detailed thermal modelling, as a representative 
of the group of targets for the CRAF and ROSETTA missions. It has a period of 5 . 5 
years, an eccentricity of 0.52 and a perihelion distance of 1.5 AU. The model was 
constructed in the same manner as the ones described in sections IV B and C. 
Several evolutionary tracks were computed, spanning more than 100 revolutions, for 
various combinations of the following parameters: p - 0.2 or 0.55 g cm"^; dust/ice 
ratio Xd/Xi — 1 or 0.2; the fraction of dust mass which accumulates permanently on 
the surface ri = 0.001, 0.01 or 1. Six parameters were calculated for each 
combination: Tg - the temperature of the dust layer on the surface; Tj^ - the ice 
temperature of the ice jtost tindemeath the dust layer; T .. . - the ice temperature 10 
m below the surface; Rg - Rs . in m - the difference between initial radius (5 km) 
and surface radius at a given orbit, or the thickness of the ice which has been lost 
by sublimation; Rg - r^^, in m - the difference between the surface radius and the 
radius of the interface between crystalline and amorphous ice or, the thickness of 
the crystalline ice layer overlying the inner core of amorphous ice, Ar, in cm - the 
thickness of the dust mantle. 

The following conclusions may be drawn from these results: 

1 . The temperature at the surface of the dust layer is weakly dependant on the 
assumed parameters, for given albedo (0.04) and emissivity (0.5). Thus, at 
perihelion, Tg varies between 235 and 268K, with an average value of 252117K. 

2. The temperature of the ice surface just beneath the dust mantle is very nearly 
constant. Tj^ at perihelion is between 176 and 187K, except for the extreme case 
of r; = 1. 

3. For all models Tg increases, Tj^ decreases and R decreases slowly with repeated 
revolutions, because of the growth of the dust mantle, which is of the order of 
cm-m, depending on r? . 


4. The temperature at a depth of 10 m is practically constant with time and varies 
very little from one set of parameters to another. Thus, we may expect a 
temperature of 163K with an uncertainty of 3%. Moreover, because this temperature 
is above the transition point from amorphous into crystalline ice, the ice down 
to 10 m below the surface is crystalline. 

5. The crystalline ice shell overlying the inner core of gas- laden amorphous ice 
is at least 40 m thick and may be as thick as 240 m. The crucial parameter in 
this respect is the dust to ice ratio. The higher this ratio, the thinner the 
crystalline ice layer becomes. The layer thickens with niimber of revolutions. 

In conclusion, since even the lower limit of 40 m for the thickness of the 
crystalline ice layer is significantly larger than the depth of penetration 
currently planned for the CRAF and ROSETTA probes, we may conclude that they will, 
most probably, sample crystalline ice. They may tap a gas -filled pocket, with 
unpredictable results - explosions etc. If the target comet resembles comet Halley, 
where several active craters were found, it might be worthwhile to dig into such a 
crater, at the bottom of which the more active gas -laden amorphous ice should be 


1. A. Bar-Nun, G. Herman, M.L. Rappaport and Yu. Mekler. Ejection of HjO, 0^, Hj 

and H from water ice by 0.5-6 KeV IT*" and Ng"*" ion bombardment. Surface Sci . , 
150 , 143-156 (1985). 

2. A. Bar-Nun, G. Herman, D. Laufer and M.L. Rappaport. Trapping and release of 

gases by water ice- and implications for icy bodies. Icarus , 63 , 317-332 

3. A. Bar-Nun, J. Dror, E. Kochavi and D. Laufer. Amorphous water ice and its 

ability to trap gases. Phys. Rev. B . 35, 2427-2435 (1987). 

4. D. Laufer, E. Kochavi and A. Bar-Nun. Structure and dynamics of amorphous water 

ice. Phys. Rev. B. 36, 9219-9227 (1987). 

5. A. Bar -Nun, I. Kleinfeld and E. Kochavi. Trapping of gas mixtures in amorphous 

water ice. Phys. Rev. B . 38, 7749-7754 (1988). 

6. D. Prialnik, and A. Bar-Nun. On the evolution and activity of cometary nuclei. 

Astrophys. J. 313 893-905 (1987). 

7. D. Prialnik and A. Bar-Nxin. The formation of a permanent dust mantle and its 

effect on cometary activity. Icarus , 74, 272-283 (1988). 

8. D. Prialnik, A. Bar-Nun and M. Podolak. Radiogenic heating of comets by ^^Pil 

and implications for their time of formation. Astrophys. J . 319 , 993-1002 

9. A. Bar-Nun and D. Prialnik. The possible formation of a hydrogencoma around 

comets at large heliocentric distances. Astrophys . J . Lett . 32 , L31-L34 


10. A. Bar-Nun, E. Heifetz and D. Prialnik. Thermal evolution of comet P/Temple-1, 

representing the group of targets for the CRAF and ROSETTA missions. Icarus , 
in press (1989). 

11. A. Bar-Nun and I. Kleinfeld. On the temperature and gas composition in the 

region of comet formation. Icarus , in press (1989). 

12. P.V. Hobbs. Ice Physics , Clarendon Press , Oxford (1974). 

13. P. Lammerzahl. Gas emission from Comet Halley. in E. Grun. 3 Kometenwerstatt. 

14-15 Nov. 1988. 

14. M.T. Duncan, T. Quinn and S. Tremain. The origin of short period comets. 

Astrophys. J. 328 , L69-L73 (1988). 

15. A.H. Delsemme. Have comets played a role in the primary organic synthesis. 

Preprint F.7.1.2 COSPAR, Helsinki (1988). 

16. P.R. Weissman. The Oort cloud and the galaxy: dynamical interactions. in The 

Galaxy and the Solar System , eds . R. Smoluchowski , J.N. Bachall and M.S. 
Matthews. University of Arizona Press, pp. 204-237 (1986). 

17. D.P. Simonelli, J.B. Pollack, C.P. McKay, R.T. Reynolds and A.L. Suntmers . The 

carbon budget in the outer solar nebula. Submitted to Icarus (1988) . 

18. J.I. Lunine. Primitive bodies: molecular abundances in comet Halley as probes 

of cometary formation environments. Preprint (1989). 

19. R.F. Knacke, Y.H. Kim, K.S. Noll and T.R. Geballe. Search for interstellar 

methane. Astrophys. J. 298 , L67-L69 (1985). 

20. P.D. Feldman, H.A. Weaver, T.N. Woods, M.F. A'Heam, L.A. McFadden and M.C. 

Festou. The 18-19 March 1986 outburst of Comet Halley as observed by the lUE. 
Abstract No. 20.10, DPS Meeting, Paris. Bull. Amer. Astron. Soc . 18 , 795 



E. Griin, et al. 

Max-Planck-Institut fiir Kemphysik 

Heidelberg, FRG 



E. GrUnK J. Benkhoff^, A. Bischoff^. H. DUren^, H. Hellmann^, P. Hesselbarth^ P. Hsiung*. 
H.U. Keller^, J. Klinger^, J. Knblker^, H. Kochan^, G. Neukum'', A. Oehler^, K. Roessler'^, T. 
Spohn^, D. Stbffler^ and K. Thiel^ 

*Max-Planck-Institut fiir Kernphysik, Heidelberg, FRG; ^Institut fur Planetologie, WWU, 
Miinster, FRG; "^Institut fiir Raumsimulation, DLR, Kbln-Porz, FRG; ^Institut fiir Chemie I, 
KFA, Jiilich, FRG; ^Max-Planck-Institut fiir Aeronomie, Katlenburg-Lindau, FRG; ^Laboratoire 
de Glaciologie et Geophysique de I'Environment, Saint-Martin-d'Heres, France; ''institut fiir 
Optoelektronik, DLR, Oberpfaffenhofen, FRG; ^Abteilung Nukelarchemie, Universitat Koln, 
Koln, FRG. 


Sublimation experiments with ice-mineral mixtures were carried out at the DLR Space 
Simulator- in order to study cometary processes. First experiments were done with 
cylindriced samples of 30 cm diamter and 15 cm thickness which consisted of water-ice or 
water^ and C02-lce mineral mixtures. These experiments have cdready yielded important and 
new insights into the modifications of the sample which are caused by the sublimation of 
the ices due to insolation: (1) Spectral reflectance measurements show the reduction of 
volatile materials in thq surface layers of the sample and the formation of a permeable 
refractory dust mantle, (2) the dust mantle as well as the residueds of emitted dust 
particles have a low density (~ 0.1 g/cm^) aggregate structure, (3) metamorphosis of the 
original non-coherent ice into hard but still porous water ice has been observed under the 
dust mantle and (4) fractionation of ices of different volatility occurs during their 
sublimation. A queilitative model is described which can explain the observed modifications 
of the sample material. 

Key words: Comet simulation, sublimation fractionation, ice-mineral mixtures 


Halley measurements. An active comet like comet Halley loses by sublimation a surface 
layer of the order of 1 m thickness per perihelion passage. In situ measurements 
(Krankowsky and Eberhardt, 1989) showed that water ice is the main constituent which 
contributes to the gas emission although even more volatile species have been identified 
(Table t). 

Dust particles which were embedded in the ices are carried by the sublimating gases. 
Measurements of the chemical composition of cometary grains indicate that they are 
composed of sihcates of approximate chondritic composition (Jessberger et al., 1988) and of 
refractory carbonaceous material (Kissel and Krliger, 1987) at a mass ratio of about 2:1. 


Table 1: Gas composition of Comet Hal ley (Krankowsky and Eberhardt, 1989) and typical 
sublimation temperatures <i.e. temperature at a vapor pressure of 0.1 Pa). 


Abundances (Vol%) 

Temperature (K) 














1. 5 











Previous comet simulations. In the past there have been several attempts to experimentally 
study the sublimation process of mixtures of ices, minerals and carbonaceous compounds. 
Extensive work was carried out by Soviet groups in Dushanbe and Leningrad, who heated up " 
ice samples electrically or irradiated them with light sources inside a cold chamber 
(Kajmakov and Sharkov, 1972; Ibadinov, 1989). Among the interesting findings was the 
formation and ablation of dust mantles during the sublimation process. In another approach 
(Saunders et al., 1986; Storrs et al., 1988), silicate minerals and organic compounds covered 
with water ice were exposed in a vacuum chamber. After sublimation of the water ice 
highly porous filamentary sublimate residues were found for some classes of phyllosilicate 
minerals or in cases when organic compounds (tar) were present. 

KOSI experiments. The approach by the KOSI-team (Kometensimulation) (Griin et al., 1987; 
Kochan et al., 1989a; Klinger et al., 1989a) focuses on the investigation of cometary 
processes at relevant scales. The scale of the simulation experiments is determined by the 
scale for the heat transport into the interior (diurnal thermal skin depth) and the scale of 
the gas interaction (mean free path length) above the surface which both have been 
estimated to be of the order of 10 cm. Comet simulation experiments are performed in the 
big Space Simulator of DLR, Cologne, which allows the study of model comets of up to 
one meter dimensions. 

Experimental Set-Up 

Space Simulator. The Space Simulator is a large vacuum chamber with an inner liquid 
nitrogen cooled shroud and a separate light source (Xenon lamps) which insolates a < 1 m 
sized target with up to 1.3 kW light power (Kochan et al., 1989b). The spectrum of the 
lamps matches roughly the solar spectrum in the visible wavelength regime. The 
background pressure in the chamber is 5l0~* Pa. The shroud keeps the background 
temperature at -77 K and thereby acts as an effective pump for less volatile gases like 
water and CO2. 

Sample container. The cylindrical sample container for the first 3 KOSI experiments has a 
diamter of 30 cm and a hight of IS cm (Fig. 1). For easy charging the container with the 
ice-dust mixture is mounted horizontally, while during the experiment it is tilted by 45- in 


Fig. 1 Sample container used for the comet simulation experiments. Its dimensions 
are 30 cm diameter and 15 cm depth. Thermocouples sticking from both sides into 
the interior. A teflon cylinder insulates the sample from the liquid nitrogen cooled 
outer copper shell. 

order to be effectively insolated. The bottom of the sample container is a liquid nitrogen 
cooled copper plate. The inner wall of the container is made out of teflon in order to 
shield the sample material from the outer copper wall. A motor driven cover protects the 
sample during preparation from ambient temperature and contaminants. Several temperature 
sensors are installed in order to measure the sample temperatures at various depths. 

Diagnostics. At some distance from the sample container a range of instruments are 
mounted to a rectangular support structure in order to analyse the emitted gas and dust as 
well as to observe in situ the modifications of the sample during insolation (Fig. 2). Dust 
collectors, piezo-electric impact detectors as well as television cameras determine the rate, 
the size and the speed of the emitted dust particles. Ionization gauges and mass 
spectrometers measure the flux, the composition and the speed of the released gases. The 
sample itself is monitored by TV cameras. 

Sample materials. The sample material consists of ices (H2O, CO2) and minerals simulating 
cometary dust. The dust analogue materials were selected on the basis of (1) the observed 
mineralogical composition of solar system materials and (2) the availability of the analogue 
materials in large quantities. The mineralogy of carbonaceous chondrites (Kerridge and 
Matthews, 1988), of interplanetary dust particles (Mackinnon and Rietmeijer, 1985; Sandford 
and Walker, 1985), and the data of the comet Halley obtained by the GIOTTO mission 
(Jessberger et al., 1988) justify the selection of Mg-rich silicates of olivine and pyroxene 
composition, of sheet silicates and of carbonaceous material. Carbon (soot) has been chosen 
as a simple substitute for carbonaceous matter, and montmorillonite and kaolinite as 
representatives of the sheet silicates. Olivine and pyroxene are the main constituents of 
powdered dunite. The dust components, compositions and grain size characteristics are 
listed in Table 2. All components are in fact mineral mixtures as they contain accessory 
minerals in the order of 5 to 10%. Carbon which has a grain size of only 23 nm is 
suspended in water activated by a sodium salt of a naphtaline sulfoacid condensation 
product. The grain size of the other mineral powders is predominantly below ~ 4 y.m. 


Fig. 2 View into the opened Space Simulator. In the center the sample contsuner is 
visible. The rectangular structure supports mirrors (top and right), gas diagnostics 
(upper left) and dust detectors and collectors (bottom). Outside the cylindrical cold 
shroud can be seen. 

Sample preparation technique. So far, all samples were prepared from dust mixtures 
suspended in water. Non-coherent, fluffy ice-dust mixtures were prepared by spraying these 
suspensions into liquid nitrogen (cf. Saunders et al., 1986). This method was chosen for 
simplicity and efficiency (abuot 10 kg of sample material are needed for a simulation). 
Because of the high content of minerals in the water (ca. 10% by weight) the individual 
grains were not completely separated from each other in the suspension. Therefore dust 
aggregates found after sublimation of the ice were preformed during the freezing of the 
suspension. Sample preparation techniques which avoid the mutual contact of dust particles 
in the presence of liquid water are under development (e.g. condensation from the gas 

The propellant gas used for the spraying was nitrogen in the first two experiments (KOSI-1 
and 2). For the KOSI-3 experiment we used CO2 as propellant gas. Thereby another volatile 
ice component was produced. The content of C02-ice in the mixture was measured by gas 
chromatography of released C02-gas upon heating of witness samples (Roessler et al., 
1989). The composition of the samples of the first 3 KOSI experiments is given in Table 3. 

The technique of sample preparation has been standardized. The water- mineral suspensions 
were treated for 8-12 hours by shaking and were kept for 15 minutes in an ultrasonic 
bath at a frequency of 20 kHz (Bischoff and Stoffler. 1988). The grain size distribution of 
the mineral dust was controlled by a laser granulometer. In mixtures of all components 
including carbon the median grain size was ~4 [im. The control of the grain sizes in the 
suspension together with the variation of the dust composition was intended to get an idea 
of the effects caused by the contact of liquid water with the different mineral grains which 
could not be avoided in these initial experiments. 

For spraying the suspension we used a special device which is shown in Fig. 3. The 
suspension was constantly stirred before it flowed through an ultrasonic bath and 


Table 2: Properties of mineral components for synthetic cometary samples (weight %) 

Material name 

Mineralogical composition 

Nominal grain size 

Density of the individual 
minerals (g/cm-^) 


90-92% kaolinite 
5% illite 
93% montmorillonite 

min. 90% < 2 [im 



9 3% montmorillonite 

median diameter: 

1 .95-2.06 

ASB 350 

3% feldspar 

4 Jim 


2% mica 

99% < 35 [Ltn 


2% quartz and 





84-9 1% olivine 

median diameter: 


5 - 1 1 % orthopyroxene 

4 y.m 


1.7-2.2% chlorite 

95% < 30 [im 


1.0-2.2% serpentinite 


0.1-0.4% talc 


0. 5% spinel 



2 5% soot in H2O with ion 

ca. 23 nm 

max. 2.23 

VU 25/L 

activated wetting agent 

Table 3: KOSI samples; initial composition (main constituents) and properties 




May 87 

April 88 

Nov. 88 

Composition (weight %) 

H,0 ice 




CO2 ice 




total dust content 




Relative mineral composition 


















~ 0.2 


0. 18 

density (g/cm3) 








penetration strength (MPa) 




formed large aggregates; 

n.d.: not determined 



ultrasonic bath 

spraying pistol 
fuel gas 


Fig. 3 Method for the production of synthetic cometary samples (schematic). 

entered a spraying pistol (nozzle diameter: 1.5 mm) which was directed into a sieve catcher 
positioned at some depth within the liquid nitrogen. The produced "snow" was taken 
directly into the sample container of the space simulator. At the same time a second 
sample container was filled in the same way and analysed in a glove box (Roessler et al., 
1989). There it was cooled by liquid nitrogen and finally kept in a low temperature room at 
about 250 K for physical characterization of the sample. 

Considering the relatively short times the minerals were exposed to liquid water and ice 
before the sublimation experiment is performed we can exclude chemical reactions between 
H2O and olivine, pyroxene and montmorillonite, respectively. Such reactions may take place 
in permafrost over geological time periods (Rietmeijer, 1985). 

Sample characterization. Analyses performed with the samples before and after the 
sublimation experiment include the measurement of reflectance spectra (albedo) both at 
visual and near-infrared wavelengths, determination of the penetration strength (Thiel et al. 
1989) and of the density and porosity (cf. Table 3). Attempts to produce polished sections 
of aliquots of KOSI-2 samples were made (Stoffler et al., 1989) based on techniques used 
in snow research (e.g. Good, 1982, 1987). In this method the porous samples are 
impregnated with liquid diethyl phtalate at ~268 K and then frozen at ~255 K. Polished 
sections are prepared by a sledge microtome. Preliminary studies of these sections with a 
polarized microscope indicate that the dust grains are contained either in spherical 
aggregates of dust and ice or in irregular polycristailine sections of ice where they occupy 
the grain boundaries of the ice crystals preferably (Fig. 4). The spherical aggregates ranging 
from some 100 tim to more than 2 mm (Bischoff and Stoffler, 1988) are obviously formed 
when the dust-water suspensions are sprayed into liquid nitrogen. 


Fig. 4 Micrographs of a polished section of a KOSI-2 sample obtained by the 
polarizing microscope in reflected light, horizontal width of microphotograps: 1.4 
mm. (a) In the center: spherical dust-ice aggregate; upper right corner: 
polycristalline ice with interstitial dust; white matrix with diagonal lines: pore 
space filled with crystallized diethy Iphtalate. (b) Polycristalline ice; dark lines: grain 
boundaries between ice crystals where some dust aggregates are located (worm-like 

Initial Results 

Experimental sequence. Besides with different sample materials the KOSI experiments were 
conducted with different insolation profiles. During the KOSI-1 experiment, the main 
purpose of which was technology verification, the sample was insolated with 1.1 SC (Solar 
Constant, light power onto surface) for 13 hours. The insolation profile during KOSI-2 was 
more complex: 1 SC for 16.5 h, off 4 h, 1 SC 4 h, off 4 h, 1 SC 4 h, off 4 h, ~2 SC 2 h 
and during KOSI-3: 1.3 SC 10.3 h, off 6 h, 1.3 SC 30.8 h. As an example for the in-situ 
measurements we want to discuss in some detail the temperature and the released gas 

Temperature measurements. Figure 5 shows the time-history of the temperatures within the 
sample during the KOSI-3 experiment. The temperatures are taken at 1 cm distance 
intervals from the backplate of the sample container (no measurements at and 3 cm have 
been taken). The higher the temperatures the closer they are measured to the sample 
surface. The thermoelement at 13 cm from the backplate penetrates the surface of the 
sample after about 3 h and becomes exposed to direct irradiation, at which time the trace 
is discontinued. The measurement at 12 cm never became exposed to direct light but 
reached temperatures as high as 300 K in the upper layers of the sample. The sensitivity of 
the temperatures to the incident light in the lower part of the sample is an experimental 
artefact. Two distinct levels of temperatures are recognized within the sample: at the 
beginning of the experiment at about 120 K and in the later stages just below 210 K. They 
are attributed to the discrete levels of sublimation of CO2 and H2O, respectively (cf. Spohn 
and Benkhoff, 1989). 

Emitted gas. The gas release during the KOSI-3 experiment is shown in Fig. 6. Immediately 
after the initial switch-on of the irradiation both CO2 and H2O emission rates jumped to 




-" — I — '- 

20.0 30.0 

Time (h) 



Fig. S Evolution of temperatures in the KOSI-3 sample. The sample had been 
insolated with 1.3 SC O h to 10.3 h and from 16.3 h to 47.1 h. Temperature 
measurements are shown at distances 1, 2. 4, ,S. 6, 7, 8. 9, lO, 11, 12 and 13 cm 
from the back plate. The trace of the measurement at 13 cm is discontinued at ~3 
h when it became exposed to direct irradiation. 










I r 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 I I [ I I I I I I I I I I I 1 I I I I I I I ■ 

o H,0 

+ CO, 


^ 10 

5 10" 

N. .• 

*! Q I ■'■■ I ■'■■ I ■■■■'■■■■ I I ■■■■■■■■ I ■■■■■■■■■ I ■ ■ ■ 

10 20 30 40 50 

Time (h) 

Fig. 6 Measurement of the H2O and CO, gas flux density at <»0 cm distance from 
the sample and at an angle of 38 from the surface normal (KOSl-3). For the 

insolation periods cf. Fig. 5. 



1000 1500 

Wavelength (Nanometer) 



Fig. 7 Reflectance spectra of the KOSI-3 sannple. Trace 1: original sample; traces 2: 
irradiation processed sample (2a upper and 2b lo^ver part of the inclined sannple 
surface); trace 3: pure carbon. Sample temperature «> ISO K, measurement in sannple 
normal, illumination at 32" phase angle. 

maximum values. Thereafter both emission rates decreased with time. The ratio of 
H2O/CO2 emission decreased with time from about 6 to 3 at the end of the experiment, 
although the mole fraction of both constituents was ~14 in the original sample material. 
During the off-period 'of irradiation and shortly thereafter this ratio varied over a large 
range. When the lamps were switched off the water emission ceased rapidly, while the CO2 
emission followed much more slowly. During the sun-off period only CO2 emission was 
observed. After switch-on of the lamps water emission started immediately, reaching its 
maximum value about three hours later. C02-emission continued to decrease for two more 
hours after switch-on before it started to increase again. This is an indication that CO2 
sublimates in deeper layers of the sample. After it had reached its maximum value about 5 
hours after switch-on the C02-emission declined roughly in proportion with the water 
emission. From the evaluation of the gas flux data we estimate that about 60% of the total 
CO2 content left the sample during the experiment. 

Optical properties. The samples are characterized by optical reflectance spectroscopy. The 
radiance coefficient (Hapke, 1981) is measured in the wavelengths range from 0.45 y.m to 2.5 
[im. The spot size on the sample is about 4 cm diameter. The sample is contained in a 
glove box and is liquid nitrogen cooled. The sample is illuminated and viewed through four 
quartz windows which allows the measurement of reflectance spectra at phase angles of 
32 , 50 and 70 . Fig. 7 shows reflectance spectra of the sample before and after insolation 
at 32' phase angle. The reflectance (albedo) averaged over the wavelength range is 18.3% 
for the original KOSI-3 sample, 16.0% In the upper part and 13.7% in the lower part of the 
irradiation processed sample. For comparison pure carbon powder has an average reflectance 
of 3.1%. The albedo of the KOSl-3 sample was much higher than that of the KOSl-2 sample 
(6%) because of the admixture of bright C02-ice. In the KOSl-1 experiment a different type 
of carbon was used which formed large agglomerates and therefore was not as effective. 

The reflectance spectra show characteristic absorption bands for carbon as well as for CO2 
and H2O ice. The abundances of the minerals olivine and montmorillonite are too low in 


Fig. 8 SEM micrographs of dust grains emitted during the KOSI-3 experiment. The 
dust composition is 90 % (by weight) olivine and lO % montmorillonite. (a) Fluffy 
grain, (b) Compact grain. 

order for their absorption bands to be noticeable in the presence of the dark carbon. 
Narrow absorption bands of COg-ice at 2020 nm can barely be seen in the original sample 
on top of the heavy absorption band of water ice at 2000 nm. Because of the high 
transmission of both ice components between 450 nm and 1000 nm wavelength and because 
of the low concentration of the silicates the spectrum in this range is dominated by 
carbon. Below 500 nm the carbon shows a strong increase of its reflectance. This increase 
is caused by Rayleigh scattering of the very small carbon grains. This increase at short 
wavelength is also visible for the KOSI sample material but the effect of carbon is 
somewhat reduced because of an opposite effect of the silicates. The balance between the 
reddening by the minerals and the increase towards shorter wavelengths by the carbon 
depends on the distribution of both minerals in the mixture and is therefore dependent on 
details of the sample preparation technique. 

A comparison of the spectra of the original sample and the processed sample shows a 
decrease of the total reflectance as well as a reduction of the water absorption bands. The 
C02-signature vanishes already after little processing by insolation. The difference in the 
spectra from the upper and lower parts of the surface of the processed sample are 
explained by the formation of a dust mantle of variable thickness on top of the ices. 

Dust mantle. Inspection of all samples after insolation shows that a layer of dry dust of a 
few mm to ~1 cm thickness had formed overlaying the ice mixtures. This layer was 
generally thicker on the lower parts of the sample surface than on the upper parts. 
Observations of the sample during insolation showed that some of the bigger particles 
which were not completely dragged away by the gas stream just rolled down the inclined 
surface of the sample and accumulated on the lower portions of the sample surface. 

SEM analysis of the mantle material (Thiel et al., 1989) and the residues of emitted dust 
particles (Fig. 8) which were collected during the experiment showed that both were of 
very similar structure. The density of the mantle material is of the order of 0.1 g/cm"^. The 
fluffy texture of the mantle indicates an almost complete lack of volatiles in it. This 
structure is mainly controlled by (1) the mineralogical composition of the dust component 
in the ice-mineral mixture, (2) by the consistency of the original material and (3) by the 
preparation method. 


As mentioned above phyllosilicates (montmorillonite, kaolinite) and nesosilicates <olivine) are 
mainly used to simulate the comet nucleus analogue. SEM-investigations reveal that the 
fraction of phyllosilicates relative to olivine is the main parameter that determines the final 
structure of the residuals (cf. Storrs et al., 1988). Higher fractions of phyllosilicates yield 
spongy particles of regular pore shape, the pores being separated by straight partition 
walls made of silicate platelets. A high olivine fraction on the other hand leads to irregular 
fluffy material made up of roundish constituents glued delicately together, forming 
completely irregular pore spaces. 

The structure of the dust mantle as a whole is influenced by the consistency of the 
original ice (snow) dust mixture. Starting with a massive block of dirty ice yields a 
fine-grained coherent dust mantle occasionlly separated in series of fine layers. Compact 
snow with small pores (mud-like appearence) produces a loose dust layer of small grain 
sizes (~I [im to ~100 (im), spongy snow with large pores (snow-like appearance) leads to a 
highly fluffy dust mantle of medium to large grain sizes (100 [im to >1000 ^im). 

Hardness test. Both before and after a KOSI experiment the hardness of the sample is 
measured. The measuring device is a motor driven force-meter which pushes a cylindrical 
piston of 5 mm diameter into the sample (Thiel et al., 1989). The penetration force is 
recorded as function of depth. After the KOSI-3 experiment it was found that right 
beneath the dust mantle a hard crust had formed. The hardness had risen from originally 
~0.2 Mpa to 1.3-5.1 Mpa over a thickness of 28 to 70 mm. The actual values varied over a 
large range with the location on the sample, however, the general trend that the hardness 
of a sub surface layer increased considerably during insolation has been confirmed by a 
number of experiments. In addition, the thickness of this layer increased with time. Below 
this hard layer the hardness showed the original value. This observation indicates that a 
major metamorphosis of the near surface layer of the sample material occured during the 
sublimation experiment. 

Sample structure and composition. After transfer of the irradiated KOSI sample to the 
glove box (Roessler et al., 1989) the sample was inspected and a number of small specimen 
including drill cores were taken from different positions within the sample. These specimen 
were used for chemical, isotopic and petrographic characterization of the different phases 
of the sample. 

Visual inspection of the KOSI-3 sample showed beneath the dust mantle the several cm 
thick, hard and coherent but porous ice crust. This material was somewhat brighter than 
the non-coherent material underneath the crust which resembled most the original sample 
material. The bottom layer (of several mm to cm thickness) was very bright and contained 
some larger ice platelets. 

The COj-content at various depths was determined by gas chromatography (Roessler et al., 
1989). Except for the bottom ~2 cm no significant COg abundance was measured within the 
KOSI-3 sample. Even there it was mostly reduced from its original value of ~14% to about 
2% except for the lowest few mm where it had even increased to 20%. 

The isotopic ratios O/ O and D/H have been measured at various depths of the KOSI-2 
sample (Klinger et al., 1989b). It turned out that in the near surface layers (^ 4 cm) a 
strong enrichment of heavy isotopes occured, whereas in the deeper layers the depletion 
was smaller or even negligible. But these results need further confirmation. 

The mass densities were determined on volumetricaiiy defined aliquots. For the KOSI-3 
samples the density before the sublimation experiment was 0.5 g/cm"^ and afterwards it 
differed distinctly between the ice crust (0.5 g/cm^) and the non-coherent lower layer (0.3 
g/cm^). The results from the first attempts of a petrographic characterization of the 


ice-mineral mixtures have been discussed above (cf. Fig. 4). Several specimen were molten 
for grain size measurements of the dust component by laser granulometry. The dust in the 
suspensions from which the KOSI-3 sample was prepared had a median grain diameter of 
4.9 (im. a rather large standard deviation and a slight excess of coarse grains. The 
suspensions produced by melting the ice-dust samples after the sublimation experiments 
had a median grain diameter of ~6.4 (jm and a slight excess of fine grains. Comparison 
with measurements made with KOSl-2 samples showed that the CO., admixture had a major 
effect on the mean grain size measured after the sublimation experiment. 


The results from the sublimation experiments are discussed in terms of a proposed 
qualitative model which accounts for the observed temperature profile, the gas release 
characteristics and the metamorphosis of the sample material (i.e. the formation of a dust 
mantle and an ice crust) 

Sublimation fractionation model. The KOSI sample material had a density of about 0.5 
g/cm . Such a low density material necessarily contains a great number of pores that 
communicate with each other. As long as such a system is confined in a closed volume and 
kept at a constant temperature, the pressure in the pores is equal to the vapor pressure of 
the ice at the given temperature. When a thermal gradient is maintained within the sample 
material, a pressure gradient builds up as a result of the variation of the vapor pressure of 
the ice as a function of temperature. This pressure gradient acts as a driving force for the 
diffusion of the vapor phase through the pore system. In this way a supersaturation occurs 
locally and leads to a recondensation of the vapor phase. This phenomenon can persist in 
deeper laj'ers of the sample material even when the vapor phase can leave the sample 
through a semi-permeable top surface. In this case a net sublimation occurs in the 
near-surface ice layers. The vapor phase that circulates through the pore system 
contributes to the heat exchange between the surface and the deeper layers of the sample 
(Smoluchowski, 1982, Klinger et al., 1989a, Spohn and Benkhoff, 1989). This heat transfer to 
colder ice layers has three components, heat conduction by solids (1) and by gas (2) and (3) 
deposition of latent heat due to the net mass transport by recondensation. A consequence 
of the redeposition of water vapor is the hardening of the initially non-coherent material. 
This mass and heat transport occurs both for water and CO.,, for the latter only at greater 
depth within the sample and at a lower temperature level. 

Figure 9 demonstrates the result of sublimation fractionation of the KOSI sample after 
some period of insolation and the corresponding temperature and pressure profiles. The two 
top panels show the stucture and the composition of the sample. At the bottom of the 
sample (depth 4) it is still in its original state. At depth 3 effective sublimation of CO2 
occurs, inwards of which its concentration is slightly increased due to inward migration, 
outward this level no CO^-ice is found. Water sublimation occurs at depth 2, outward of 
which a dust mantle has normed. 

The temperature profile reflects the energy input at the top surface of the dust mantle. Its 
surface temperature is determined by the balance between absorbed light energy, emitted 
thermal radiation and heat conducted into the interior by solid state and gas heat 
conduction. The temperature of the sublimation surface of water ice is elevated above the 
temperature of an ice surface which sublimates freely into vacuum, although the energy 
flux into that surface is already reduced from the energy absorbed by the dust mantle. A 
flat temperature profile characterizes the heat transport by the water vapor down to some 
distance from the water sublimation surface (Spohn et al., 1989). A similar behaviour of the 
temperature profile is seen inward the CO., sublimation surface. Only at great depth the 
temperature profile has the steep gradient expected for solely solid state heat conduction. 


Sublimation Fractionation 


I 4 I ; 




Temperature (K) 

Pressure (Pol 

Fig. 9 Model of sublimation fractionation: sample structure, connposition and 
internal temperature and pressure profiles . The conditions of the KOSI-3 
experiment are shown after some time of insolation. 

The pressure profile follows the temperature profile insofar as inside sublimation surfaces 
the partial pressure corresponds to the vapor pressure at this temperature. Outside 
sublimation surfaces the pressure is controlled by the flow through the partially permeable 
overlaying material. 

Future developments. Because of the close coupling of the energy and mass transport 
within the sample a quantitative model has to describe both effects together. Since the 
system is evolving, ultimately, non-stationary solutions have to be investigated. Also the 
experiments have to be improved in order to determine quantitatively the energy and mass 
balance of the KOSI experiments. Methods have to be developed to measure the thermal 
emission from the surface and the heat flux into the back plate of the sample container as 
well as to determine the mass loss from the sample container. In addition the sample 
materials and their phases have to be better characterized: both thermal conductivity, heat 
capacity, porosity and permeability for gases have to be determined. It is estimated that for 
the present sample dimensions the temperatures at the sublimation level of CO2 are 
affected by the back plate temperature already after a few hours of experiment time. 
Therefore, an increase of the dimensions of the sample by at least a factor of two would 
allow us to follow the development of the temperature wave for longer times and at lower 
temperature levels. 

Relevance to comets. There are two lines of evidence that cometary nuclei in general and 
their near-surface layers in particular are of high porosity (s 0.5). Firstly, estimates of the 
mean density of Halley's and other comets" nuclei indicate that their densities are low (< 1 
g/cm"^), lower than any reasonable solid non-porous ice-dust mixture would have. Secondly, 
the detection of gases more volatile than water suggests that these gases are released 
from greater depths than water and that these gases have to percolate through the 
overlaying material. Therefore, the fractionation effects which we have found to take place 
during the sublimation of porous ice-dust mixtures in the laboratory should occur also 
under cometary conditions. 


However, there is a major difference between the simulation experiments and the cometar> 
situation which has to be taken into account. That is the about lO* times higher gravity in 
the simulation experiments. The effect of which is. that only grains smaller than 10 [im to 
a few 100 [im (depending on their densities, cf. Griin et al., 1989) will be draged away by 
the gas flow. Under cometary conditions the limit will be 10cm to 100 cm sized particles. 
In addition, the thickness of a layer which is able to sustain by its weight alone the 
pressure of the evaporating gases has to be much thicker in the cometary case. Therefore, 
the maximum pressure gradient will be only ~10"''' of that in the simulation experiments 
and the scale lengths will be accordingly larger. Because of these larger scale lengths the 
time period to reach quasi stationary equilibrium will be much longer than in the 
simulation experiments and it will only be reached if the insolation period (~ nuclear spin 
period) is long enough. However, this time scale applies only to the upper layers of a few 
diurnal skin depths deep. For the more volatile ices which sublime even further down the 
relevant time scale is the time period of perihelion passage (~ orbital period). 

The thickness which is required for a layer to sustain a given pressure gradient will be 
reduced if the cometary surface material has some coherence. Since we have found in our 
simulation experiments that at least icy materials develop coherence during the sublimation 
process some effects of the low cometary gravity will be reduced. Another conclusion from 
our sublimation experiments is, that clathrates or other trapped mixtures of volatile gases 
in water (cf. Bar-Nun, 1989) will probably be separated during the sublimation fractionation 
process and therefore will not be found in the surface layers of cometary nuclei. These 
inferences have to be confirmed by future comet simulation experiments. 

This research has been supported by DFG within the SPP 'Kleine Korper im Sonnensystem". 


Bar-Nun A., 1989: Experimental studies of gas trapping in amorphous ice and thermal 
modelling of comets- Implications for Rosetta; Workshop on Analysis of Returned 
Comet Nucleus Samples, Milpitas, USA 

Bischoff A. and Stoffler D., 1988: Comet nucleus simulation experiments: Mineralogical 
aspects of sample preparation and analysis; Lunar and Planet. Sci. XIX, Lunar and Planetary 
Insitute, Houston, 90-91 

Good W., 1982: Structural investigations of snow and ice on core III from the drilling on 
Vernagtferner, Austria; Z.f. Gletscherkunde u. Glazialgeologie 18, 53-64 

Good W., 1987: Thin sections, serial cuts and 3-D analysis of snow; Avalanche Formation, 
Movement and Effects, lAHS Publ. /S2, 35-48 

Griin E., Kochan H., Roessler K. and Stoffler D., 1987: Simulation of cometary nuclei; Proc. 
Symposium on Diversity and Similarity of Comets, eds. E.J. Rolfe and B. Battrick, ESA- 
SP 278, 501-508 

Grun E., Benkhoff J., Fechtig H., Hesselbarth P., Klinger J.. Kochan H., Kohl H., Krankowsky D., 
Lammerzahl P., Seboldt W., Spohn T. and Thiel K., 1989: Mechanisms of dust emission 
from the surface of a cometary nucleus, Proc. COSPAR workshop on Comet Nucleus 
Modeling and Cometary Materials, Adv. Space Res., Vol. 9. No. 3. 133-137 

Hapke B., 1981: Bidirectional reflectance spectroscopy, 1. Theory, J. Geophys. Res., 86, 

Ibadinov K.I., 1989: Laboratory investigation of the sublimation of comet nucleus models. 
Adv. Space Res., Vol. 9. No. 3. 97-112 

Jessberger E.K., Christoforidis A. and Kissel J., 1988: Aspects of the major element compo- 
sition of Hal ley's dust; Nature, 332. 691-695 


Kajmakov E.A. and Sharkov V.I., 1972: Laboratory simulation of icy cometary nuclei; in The 
Motion, Evolution of Orbits, and Origin of Comets, eds. Chebotarev et al., Reidel Publ. 
Comp., Dordrecht, 308-314 

Kerridge J.F. and Matthews M.S., (editors), 1988: Meteorites and the Early Solar System, 
Univ. of Arizona Press, Tuscon 

Kissel J., Krueger F.R., 1987: The organic component in dust from comet Halley as measured 
by the PUMA mass spectrometer on board Vega 1, Nature, 326, 755-760 

Klinger J., Benkhoff J.. Espinasse S., Griin E., Ip W., Joo F., Keller H.U., Kochan H., Kohl 
H., Roessler K., Seboldt W., Spohn T. and Thiel K., 1989a: How far do results of 
recent simulation experiments fit current models of cometary nuclei?, Proc 19th 
Lunar Planet. Sci. Conf., Lunar and Planetary Institute, Houston, 493-497 

Klinger J., Eich G., Bischoff A.. Joo F., Kochan' H., Roessler K., Stichler and Stbffler D., 
1989b: "KOSI" comet simulation experiment at DFVLR: Sample preparation and the 
evolution of the '^0/*^0 and the D/H ratio in the icy component, Proc. COSPAR 
workshop on Comet Nucleus Modeling and Cometary Materials, Adv. Space Res., Vol. 
9. No. 3. 123-125 

Kochan H., Benkhoff J., Bischoff A., Fechtig H., Feuerbacher B., Grlin E., Joo F., Klinger J., 
Kohl H., Krankowsky D., Roessler K., Seboldt W., Thiel K., Schwehm G. and 
Weishaupt U, 1989a: Laboratory simulation of a cometary nucleus: Experimental setup 
and first results, Proc. 19th Lunar Planet. Sci. Conf., Lunar and Planetary Institute, 
Houston, 487-492 

Kochan H., Feuerbacher B., Joo F., Klinger J., Seboldt W., Bischoff A., DUren H., Stbffler 
D., Spohn T., Fechtig H., GrUn E., Kohl H., Krankowsky D., Roessler K., Thiel K., 
Schwehm G. and Weishaupt U., 1989b: Comet simulation experiments in the DFVLR 
Space Simulators, Proc. COSPAR workshop on Comet Nucleus Modeling and 
Cometary Materials, Adv. Space Res., Vol. 9. No. 3. 113-122 

Krankowsky D. and Eberhardt P., 1989: Evidence for the composition of ices in the nucleus 
of comet Halley; Comet Halley 1986, World-Wide Ivestigations, Results and Interpretations, 
Ellis-Horwood Limited, Chichester, in press 

Mackinnon I.D.R. and Rietmeijer F.J.M., 1987: Mineralogy of chondritic interplanetary dust 
particles; Review of Geophysics, 25, 1527-1553 

Rietmeijer F.J.M., 1985: A model for diagenesis in proto-planetary bodies; Nature, 313, 

Roessler K., Hsiung P., Heyl M., Neukum G., Oehler A., Kochan H., 1989: Handling and ana- 
lysis of ices in cryostats and glove boxes in view of cometary samples; Workshop 
on Analysis of Returned Comet Nucleus Samples, Milpitas, USA 

Sandford S.A. and Walker R.M., 1985: Laboratory infrared transmission spectra of individual 
interplanetary dust particles from 2.5 to 25 microns; Astrophys. J., 291, 838-851 

Saunders R.S., Fanale F.P., Parker T.J., Stephens J.B. and Sutton S., 1986: Properties of fila- 
mentary sublimation residues from dispersions of clay in ice; Icarus, 66, 94-104 

Smoluchowski R., 1982: Heat transport in porous cometary nuclei; J. Geophys. Res., 87, 
Supp. A., 422-424 

Spohn T., Benkhoff J., Klinger J., Griin E. and Kochan H., 1989: Thermal modeling of two 
KOSI jcomet nucleus simulation experiments; Proc. COSPAR workshop on Comet Nucleus 
Modeling and Cometary Materials, Adv. Space Res., Vol. 9, No. 3. 127-131 

Spohn T. and Benkhoff J., 1989: Thermal history of models for KOSI sublimation experi- 
ments, submitted to Icarus 

Stbffler D., Diiren H. and Knblker J., 1989: Concepts for the curation primary examination, 
and petrographic analysis of comet nucleus samples returned to Earth; Workshop on 
Analysis of Returned Comet Nucleus Samples, Milpitas, USA 

Storrs A.D., Fanale F.P., Saunders R.S. and Stephens J.B., 1988: The formation of filamentary 
sublimate residues <FSR) from mineral grains, Icarus, 76, 493-512 

Thiel K., Kochan H., Roessler K., Griin E., Schwehm G., Hellmann H., Hsiung P. and Kolzer 
G., 1989: Mechanical and SEM analysis of artificial comet nucleus samples; Workshop on 
Analysis of Returned Comet Nucleus Samples, Milpitas, USA 



James Stephens 

Planetology and Oceanography Section 

Jet Propulsion Laboratory 

Pasadena, Cahfomia 


The Nature of Comet Materials and Attachment to Them 

James Stephens 
Planetology and Oceanography Section 
Jet Propulsion Laboratory- 
Pasadena, California 91109-8099 

Because cometary surfaces are likely to be far colder and of a different 
composition and surface topography than other planetary surfaces with which we 
have experience, there are some new considerations that must be examined in 
regards to placing and attaching instnmiented packages or sample return devices 
in or on their surfaces. The qualitative analysis of the problem of embedding 
hardware in a comet icy core is limited to only one of several means for the 
purposes of this discussion. This means can be characterized as a kinetic impact 
piercing device. Such kinetic impact piercing device may be used to attach the feet 
of an instrumented package on the surface or it may be the means for implanting 
the package in the icy core below the mantle. 

The functional requirement is to implant a device in the icy core and by 
mechanical means, prevent the device from being ejected back into space. The 
requirement can be divided into two parts. 

The first requirement is to pierce the mantle and obtain access to the icy core 
because the mantle may not be attached to the core and instrument sensors must 
have access to the icy core. The impact piercing device may ricochet off of the 
mantle if it cannot be directed approximately perpendicular to the impact surface 
(Fig. 1). The surface geometry may be closer to the local vertical than to the 
horizontal because solar heat focusing, fluid dynamic channeling and electrostatic 
filament forming forces will likely prevail over the low comet gravity to form 
very grotesque surface topography. Furthermore, if the mantle that covers the icy 
core has mineral particles that are bonded together by a tar-like substance and if 
the surface tension forces of this "tar" prevail over other bonding forces; then the 
mantle may shrivel under solar heating to form an "asphalt" like brittle, high 
density, high strength material. In addition, old cometary mantles may be formed 



from the thermally stress-fractured remnants of earlier mantles. Radar 
observations of comet mantles and laboratory experimental investigations into 
the formation of comet mantles suggests that such non horizontal surface 
structures are likely to exist. 

The second requirement is to overcome the forces produced by the 
transformation of the impact kinetic energy into forces that try to eject the 
piercing device back into space. The mantle and icy core can absorb some of the 
innpact kinetic energy in the form of fracture formation and friction energy (Fig. 
2). What energy that is not absorbed in these two ways is for the most part stored 
by the icy core as elastic deformation of the icy core. This elastic deformation 
energy is returned to the piercing device and the fragmented mantle and 
fragmented core material that surrounds it after the piercing device comes to zero 
velocity. The elastic deformation rebovmding force is assisted by the pressure force 
of the gas that is formed by the previously mentioned fracture formation and 
friction energy. Much of the fracture formation and friction energy is converted 
into heat that is ultimately converted into gas because the icy core is more than 
likely to be in thermodynamic equilibrium with its under-mantle environment. 
An additional source of gas pressure is supplied by the new equilibrium that the 
core must achieve when the conductance through the mantle is increased by the 
additional venting of the mantle by the piercing device. Only if the icy core is 
subcooled to below the new equilibrium temperature will the production of 
additional gas not occur. Clathrate, free radical, and/or phase change produced gas 
may also be triggered by the impact. 

The impact piercing device must develop hold-down forces that can 
overcome the elastic deformation rebounding force and the pressure force of the 
evolved gas. 

Hold-down forces that depend upon friction between the pierdng device and 
the icy core may be insignificant because a gas bearing may form at the contact 


between the piercing device and the icy core. Experiments in drilling simulated 
comet icy core materials in vacuum suggest that gas bearing lubrication is to be 
expected between highly loaded metal objects and ice containing materials. 

Hold-down forces that depend upon cohesion between the piercing device 
and the icy core are likely to be insignificant because there are no liquids that are 
likely to form at the interface and upon refreezing bond the pierdng device to the 
core. Even if the icy core can rebound in the horizontal direction and clamp the 
piercing device in the crater, the friction forces and the cohesive forces remain 

Hold-down forces that depend upon fixed or deployed barbs may be 
insignificant because they are likely to shatter the core material (Fig. 3) during 
entry as does the piercing device and it is unlikely that the material that they 
engage will still be attached to the core. It is unlikely that the shattered material 
will rebond because no liquid can form at the expected comet pressure and 

Even if anchoring devices could be deployed horizontally below the surface of 
the unfractured icy core, the hold-down force might be as little as the gravity force 
of the pieces of the core that may be fractured by the deployment forces of the 
barbs (Fig. 3). Such fracturing is possible because subsurface wedging forces are 
likely to be very large in a low porosity icy core and the resulting cracks are likely 
to extend to the surface. Observations of comet fracturing and laboratory 
experimental investigations into the fracturing of simulated comet cores suggest 
that brittle fracture behavior of comet icy core material is to be expected. 

It is concluded that: 

1) because the comet mantle topography is likely to be non horizontal and 
thus may deflect the kinetic impact piercing device; some means must be devised 
to assure perpendicular entry into the local mantle. 



2) because the comet mantle and icy core are almost certainly brittle, the icy 
core is likely to be self lubricating and nonbonding; the elastic rebound and gas 
pressure expulsion forces produced by an impact piercing device must be 
counteracted by forces greater than those that may be provided by a piercing 
device with fixed or deployable barbs (Fig. 4). 




Abt. Nuklearchemie 

Universitat Koln 

Koln, FRG 

H. Kochan 

Institut fiir Raumsimulation 

Koln, FRG 

K. Roessler 

Institut fiir Chemie, Nuklearchemie 

JuHch, FRG 

E. Griin 

Max-Planck-Institut fiir Kemphysik 

Heidelberg, FRG 

G. Schwehm 

European Space Agency 

Noordwijk, The Netherlands 

H. Hellmann 

Institut fiir Raumsimulation 

Koln, FRG 

P. Hsiung 

Institut fiir Chemie, Nuklearchemie 

Julich, FRG 

G. Kolzer 

Abt. Nuklearchemie 

Universitat Koln 

Koln, FRG 



K. Thiel'', H. Kochan^, K. Roessler^, E. Grun"^, G. Schwehm^, H. Helimann^, p. Hsiung3, 

andG. Kolzer''. 

Vbt. Nuklearchemie, Universitat Koln, D-5000K6ln 1, FRG 
^Institutfijr Raumsimulation, DLR WB-RS, D-5000 Koln 91, FRG 
^institutfurChemie, Nuklearchemie, KFA, D-5170 JiJIich, FRG 
^Max-Planck-institutfurKernphysik, D-6900 Heidelberg, FRG 
^European Space Agency, NL-2200 Noordwijk, The Netherlands 


In 1987 approx. 20 scientists from different disciplines started a 6 year program (KOSI) to 
simulate physically and chemically relevant processes of cometary nuclei (E. Griin et al., 1989, 
H. Kochan et al., 1988;). The experiments are mainly carried out in two simulation chambers of 
the German Aerospace Research Establishment, DLR at Cologne, FRG. Experiments in the Big 
Space Simulator are dedicated to effects and processes induced by artificial solar irradiation 
(~ 1 solar constant) on a 30 cm diameter model comet of well-defined properties. Supporting in- 
vestigations are performed in a smaller space simulation chamber with ice-dust targets of typically 
10 cm diameter. A detailed description of the chambers is given by Kochan et al. (1988). Several 
groups of theorists who are part of the KOSI-team process the experimental results and provide 
relevant boundary conditions for the design of new experiments. 

The main objective of these studies is a better understanding of 

♦ the temperature behaviour of ice-dust mixtures under given irradiation 

the total mass and energy budget of the target 
the mobilization of material (dust and gas) within the target body 
physical and chemical alterations of the sample as a function of the experi- 
mental parameters, especially: 
crust formation 
gas emission 

dust emission and dust properties 
gas-dust interaction 

The KOSI-project is intended to allow a better interpretation of ground based and space 
mission gained cometary data and to support the planning of future sample return missions. 

*) This work is financially supported by Deutsche Forschungsgemeinschaft Bonn, under 
9 different contracts. 



On-line Monitoring of Crustal Strength 

During experiments in the small simulation chamber the development of a near surface ice 
crust is monitored by hardness measurements using a remote-controlled hardness tester. The 
force meter can be equipped with cylindrical boring tools of various dimensions, depending on 
the penetration forces to be measured (typical diameter of the boring head: 5 mm). The observa- 
tion of an in -situ crust development during artificial sunlight exposure was a first step to confirm 
diffusion and re-freezing of water vapor in the porous sample as expected from theoretical consi- 
derations (Klinger et al., 1988; Grun et al., 1988a). The crustal strength as a function of irradiation 
time (using 1 solar constant intensity) was found to gradually increase during the first 50 minutes. 
After 1 hr of light exposure strength values rapidly raised by a factor of ~6 relative to the initial 
value and again showed a slower increase for longer irradiation times (see also figure 1 ). 

Off-line Measurement of Crustal Strength 

Before and after experiments in both simulation chambers the hardness of the different 
sample materials was measured by means of a motor driven boring device supplied by ESA 
(fig. 2a). A boring tool similar to that used in the on-line investigation in the small chamber is pro- 
pelled vertically into the sample with a penetration velocity of 0.2 mm/s. The penetration force is 
measured and recorded on an x-y-plotter as a function of penetration depth. The resulting depth 
profiles of mechanical stress provide a feasible diagnostic means to check the reproducibility of 
the sample preparation procedure in the case of not irradiated samples. In the case of irradiated 

Mechanical pressure at crust breakthrough 











1 ' r- 

' "~ "T 1 

SC 02.12.87 

y — 




Boring tool: p=O.Z cm 



/ . 













3ss Kaolinite 


^^ / 

3s6 Montmorillonite 



3s6 Olivine 


0.083!S Carbon 



I \ 1 ; 1_ 


40 80 120 

irradiation time Cmin) 


Figure 1: In-situ crust formation as a function of irradiation time 


samples stress depth profiles are the most direct way of measuring crustal evolution of the comet 
nucleus analogues induced by radiation. 

It was found that soil mechanical data of a given sample are closely related to the method 
of sample preparation. The present procedure of sample production is based on a simple 
spraying technique using aqueous suspensions of mineral powder. For a more detailed discus- 
sion of the technique cf. Roessler et al. (1989). The following parameters affect the structure and 
texture of the sample: ( 1 ) ultrasonic treatment of the aqueous suspension, (2) type of the propel- 
lant spray gas [N2 or CO2], (3) pressure of the propellant gas [variations of a few hundred Pa are 
essential], (4) flow rate of the suspension in the spray gun at zero propellant gas pressure (5) 
distance between spray gun nozzle and surface of the liquid nitrogen the suspension is sprayed 
into, (6) content of C02-ice in case of two-component ice samples [H2O and C02-ice]. 

Table 1: Typical parameters of small chamber experiments 8-12 (18.10., 19.10., 20.10., 25.10., 28.10.1988) and 

the big chamber experiment KOSI 3 (29.11.-02.12.1988) 








Mineral composition , 







Content of C02-ice [wt%] 







Density [g/cvir] 







Porosity [%] 














Intensity of irradiation [SC] 



2.0 - 2.4 

2.0 - 2.4 

2.0 - 2.4 


Period of irradiation [h:min] 







Tj, Tf [K] 2 cm below surface *) 

149 - 196 

143 - 209 



164 - 222 

-100- -150 

Max. dust activity [/ig/cm^/min] 






Thickness of crust [mm] 







Strength of crust (MPaj 

0.30 - 1.3 

0.43 - 0.55 


0.75- 1.10 

0.35 - 0.88 


Strength below crust [MPa] 


0.04 - 0.06 

0.03 - 0.05 

0.08 - 0.24 

0.03 - 0.08 

0.2 - 0.5 

No. of strength measurements 







' Tj, Tf initial and final temperature of the sample during the experiment 


Figure 2: a: Sketch of the ESA/ESTEC boring device. During measurements the 
sample is shielded against room temperature and air humidity by floo- 
ding with 77-100 K nitrogen gas. 

b: Stress depth profile of a sample which was irradiated for 2.2 hrs with 
an intensity of 2 solar constants (olivine : montmorillonite ratio 9:1). 


Styropor box 




02 Oi. 


Exp. No.: 9 
Test No.: 4 

-i 1 \- 

2 3 




X 50-- 





H H 

] -:: 

Test No.:2 

Figure 3: Stress depth profile of asample which was irradiated for 41.2 hrs with an 
intensity of 1.3-1.6 solar constants (olivine : montmorillonite ratio 9:1). 


In some cases a complex interrelation between these parameters make it difficult to 
produce ice-dust mixtures that meet the essential requirements of a comet nucleus analogue at 
the same time (low mass density, high porosity, desired mineral dust admixture, content of CO2- 
ice etc.). The present technique does not allow to produce e.g. "fluffy" material of a snow-like 
consistency with a nesosilicate to phyllosilicate ratio of the dust admixture of 7 : 3 and an overall 
C02-content of - 15 wt%. 

In table 1 a summary is given of experiment parameters and some soil mechanical data 
obtained during a series of test runs in the small simulation chamber and during the KOSI 3 expe- 
riment in the Big Simulator at DLR. One of the main factors controlling the gas and dust emission 
of a sample when irradiated with light is the composition of the dust admixture. Most of the 
samples investigated so far contained nesosilicates (olivine) and phyllosilicates (montmorillonite) 
in ratios (by weight) of 7:3 and 9:1. The samples listed in table 1 all contained C02-ice in the 
range of 4 to 14 wt%. Experiment no. 1 1 gives an example of a high C02-content in a fine grained 
sample of "mud"-like appearance when stored in liquid nitrogen. With the spray technique used 
for the sample preparation it is difficult to produce "snow"-like samples with CO2 -con- 
tents > 10 wt% and phyllosilicate contents > 3 wt%. 

The strength data cover a relative wide range between 150 kPa and 5.1 MPa due to lateral 
and vertical inhomogeneities with respect to the sample hardness. Especially the depth variation 
of stress shows some general features which become obvious in typical depth profiles of the dif- 
ferent target materials investigated after the experiment (cf. figures 2 and 3). The presence of a 
loose ice-free dust mantle is indicated by low compressive forces during the first few mm of 
penetration depth. When the solidified ice-dust layer ("crust") is reached 3-5 cm below the surface 
a steep increase of the stress values is observed in the depth profile. A common feature of most 
of the samples is the occurrence of a stronger upper crust and a weaker lower crust where the 
stress values drop by a factor of ~2. A more thorough study of material transport within the 
sample is devoted to the question of a possible alteration of the target structure due to moving 
depth layers of vapor re-freezing. 


Dust Mantle 

All samples investigated so far showed the formation of a surface dust layer (dust "mantle") 
when irradiated with light. There are however differences in the appearance of the mantle depen- 
ding on the original structure of the sample: 

(i) Compact, pore free ice-dust mixtures suffer ice sublimation only at the outermost surface. 
Samples containing initial dust fractions of > 10 wt% form a thin ice-free dust "skin" of low porosity 
made up of tightly cohering mineral grains. This "skin", due to the lack of ice, can now be heated 
to higher surface temperatures. Eventually the gas pressure of the sublimating ice underneath 


exceeds the cohering forces of the mineral grains of the "skin" and an extended dust "sheet" of 
~ 50-200 fjm thickness is slowly peeled off, releasing the gas pressure. M 

This process may recur several times leading to a periodic separation of up to -10 "skin" 
generations. The dust "mantle" of such samples is a sequence of loosely bound thin dust layers 
showing a typical substructure. In several cases single layers were regularly covered with 
"bubbles" of some 10 /L/m in diameter, similar to "blister" phenomena induced by energetic gas 
implantation in solids. Most of the "bubbles" were burst open, indicating the effect of gas 
pressure. A cracked dust "bubble" as seen from the bottom is shown in the SEM-micrograph 
figure 4. 

(ii) Compact but porous ice-dust samples, which have a "mud"-like appearance when stored 
in liquid nitrogen (cf . table 1 ) produce quite different dust mantles. A sample made of "mud"-like 
material, due to its fine porosity, undergoes ice sublimation in a surface layer of several 100 /L/m 
thickness when irradiated with light. This leads to the formation of a porous but coherent crust of 
mineral residuals covering the ice-containing material underneath (fig. 5). Due to its greater thick- 
ness (mechanical rigidity) and higher porosity (gas permeability) compared to massive ice-dust 
samples this mineral "mantle" can not easily be lifted by the gas stream originating from deeper 
ice layers. 

Figure 4: Coherent dust layer showing features of deformation caused by gas 
pressure. The SEM-micrograph shows the high pressure side of the 
layer, which is pan of a multi-layer dust mantle formed on a compact 
dust-ice block under light exposure (-2 solar constants). Upper right: 
broken "pressure bubble". 


Figure 5: Dust mantle of a sample with "mud"-like texture. 

Figure 6: Dust mantle of a sample with "snow"-like texture. 


The dust mantle quenches the sublinnation of ice, the gas emission of the sample is 
reduced, and dust emission completely stops. By increasing the irradiation intensity, the surface 
temperature and consequently the vapor pressure of the deeper ice layers can be raised to high 
values. When the pressure forces due to enhanced ice sublimation exceed the mechanical 
strength of the mantle, big flakes of the dust mantle (diameters > 1 cm) are blown off, and dust 
emission abruptly starts again ("dust activity threshold"). 

(iii) Highly porous ice-dust mixtures of a "snow"-like texture (cf . table 1 ) suffer ice sublimation in 
a much thicker surface layer than compact samples, when exposed to light. The emanating gas 
stream is sufficiently strong to carry away particles which are only weakly attached to the surface 
or which become loose due to the steady ice corrosion by sublimation. Dust emission activity is 
maintained for much longer time periods (days) and the quenching of ice sublimation by a dust 
cover is effectively delayed by heat transport into greater depths via the vapor phase. Structural 
and compositional changes of "fluffy" snow samples are mainly controlled by their porosity and 
heat permeability. The typical surface texture of a "snow"-like sample is demonstrated in figure 6. 

Residuals of Emitted Dust Particles 

Ice-free dust particles deposited on the dust collectors during the experiments look very 
similar to the grains found in the dust mantle of the target. The emission process obviously 
causes no structural alteration of the residuals. Their structure and texture is mainly controlled (1) 
by the mineral composition of the dust in the original ice-dust mixture and (2) by the sample pre- 
paration technique which may lead to very specific target materials (cf. "Dust Mantle"). 

(1) The amount of clay minerals (e.g. montmorillonite and kaolinite) contained in the 
dust component strongly influences the appearance of the residuals. Especially phyllosilicates of 
the montmorillonite mineral group carrying a small negative charge on their cleavage planes may 
strongly adsorb traces of cations present in the water of the sample suspension. It is assumed 
that montmorillonite platelets are "glued" together by electrostatically effective water layers in a 
parallel orientation and on a scale of many grain diameters. The formation of parallel oriented 
grain agglomerates is obviously favored by shock freezing during the sample preparation process 
(cf. (2)). The resulting texture of the dust in the ice matrix (which is essentially preserved in the 
texture of the ice-free residuals) is characterized by highly regular pores and a delicate network of 
straight or bent walls made up of phyllosilicate platelets. 

Reducing the phyllosilicate fraction and increasing the amount of nesosilicates, which are 
represented by olivine in the samples investigated so far, yields residuals of less regular texture 
and commonly smaller pores. This can be achieved e.g. by changing the ratio oli- 
vine : montmorillonite from 7:3 to 9: 1 . 

(2) In the KOSI simulation experiments a comet nucleus analogue material is used, 
which is produced by spraying aqueous suspensions of mineral dust into liquid nitrogen (cf. also 
Roessler et al., 1989). The parameters of the spraying process (cf. "Off-line Measurements of 
Crustal Strength"), especially the pressure of the propellant gas, and the flow rate of the sample 
suspension control the droplet size of the spray and thus the grain size of the ice-dust mixture. A 


Figure 7: SEM-micrograph of a coarse porous dust residual containing 30 wt% 

Figures-. SEM-micrograph of a fine porous dust residual containing 10 wt% 


higher gas pressure generally yields smaller grain sizes and vice versa. In the case of a two- 
component ice mixture of H2O and CO2 the carbon dioxide is used as a propellant gas and the 
fraction of both ices can be controlled by the gas pressure. 

The droplets which are injected into the liquid nitrogen suffer shock freezing. The freezing 
starts at the surface of the droplets and rapidly proceeds into the interior. Suspended mineral 
grains may be predominantly oriented along proceeding ice fronts yielding the textures described 
above. Typical dust residuals with different amounts of phyllosilicates are shown in figures 7 and 
8. One of the objectives of the sample preparation is, to produce material that yields dust resi- 
duals which resemble stratospheric Brownlee particles of the "fluffy" type. With the spraying tech- 
nique used in the KOSI experiments this is achieved only for a phyllosilicate fraction in the dust 
component of < 10-30%. To overcome the disadvantages due to the presence of liquid water 
during sample preparation, a method to produce material from the vapour phase is presently 
being developed. 


GriJn, E. and KOSI team: Modifications of comet materials by the sublimation process: Results 
from simulation experiments. Workshop on Analysis of Retumed Comet Nucleus 
Samples, Jan. 16-18, 1989, Milpitas, Cal. 

Grun, E., J. Benkhoff, H. Fechtig, P. Hesselbarth, J. Klinger, H. Kochan, H. Kohl, D. Krankowsky, 
P. Lammerzahl, W. Seboldt, T. Spohn, and K. Thiel: Mechanisms of dust emission from 
the surface of a cometary nucleus, Proc. of the twenty-seventh plenary meeting of the 
Committee on Space Research (COSPAR), 18-29 July 1988, Espoo, Finland, Advances 
in Space Research, (1988a), Pergamon Press, Oxford (in press) 

Klinger, J., J. Benkhoff, S. Espinasse, E. GriJn, W. Ip, F. Joo, H. U. Keller, H. Kochan, H. Kohl, 
K. Roessler, W. Seboldt, T. Spohn, and K. Thiel: How far do results of recent simulation 
experiments fit current models of cometary nuclei ?, Proc. of the Nineteenth Lunar and 
Planetary Science Conf. (1988) (in press) 

Kochan, H., J. Benkhoff, A. Bischoff, H. Fechtig, B. FeuertDacher, E. GriJn, F. Joo, J. Klinger, 
H. Kohl, D. Krankowsky, K. Roessler, W. Seboldt, K. Thiel, G. Schwehm, and 
U. Weishaupt: Laboratory simulation of a cometary nucleus: Experimental setup and first 
results, Proc. of the Nineteenth Lunar and Planetary Science Conf. (1988) (in press) 

Roessler, K., G. Eich, M. Heyl, H. Kochan, A. Oehler, A. Patnaik, W. Schlosser, R. Schuiz: Hand- 
ling and analysis of ices in cryostats and glove boxes in view of cometary samples. 
Workshop on Analysis of Returned Comet Nucleus Samples, Jan. 16-18, 1989, Milpitas, 



Thomas J. Wdowiak 
Edward L. Robinson 
Gregory C. Flickinger 

David A. Boyd 

Physics Department 

University of Alabama at Birmingham 

Birmingham, Alabama 





Thomas J. Wdowiak, Edward L. Robinson, 
Gregory C. Flickinger, and David A. Boyd 

Physics Department, University of Alabama at Birmingham 
Birmingham, Alabama 35294 


Simple molecules frozen as mantles of interstellar and circumstellar grains and 
incorporated into comets are subjected to ion bombardment in the form of cosmic rays, 
stellar flares, stellar winds, and ions accelerated in stellar wind shocks. The total expected 
dosage for the variety of situations range from 10 eV/molecule for interplanetary dust 
subjected to solar flares to 10^ eV/molecule for material in the T Tauri envirormient. 
Utilizing a Van de Graaff accelerator and a target chamber having cryogenic and mass 
spectrometer capabilities, we have bombarded frozen gases in the temperature range of 10 
K to 30 K with 175 keV protons. After irradiation, removal of the ice by sublimation at an 
elevated temperature in vacuum reveals a fluffy residue. These experiments suggest that 
processes resulting in the formation of organic particles found in the coma of Comet Halley 
,"CHON", may have included ion bombardment. Also, the moderate energy (100 keV to 
500 keV) shock accelerated ion environment of bipolar outflow of stars in the planetary 
nebula stage such as the Red Rectangle, could produce complex molecular species which 
emit the observed unidentified infrared bands at 3.3 fxm, 6.2/zm, 7.7fjm, 8.6/im, and 11.3/im. 


During the Giotto and Vega encounters with Comet Halley, mass spectrometers were 
used to determine the elemental compostion of impacting dust particles (Kissel et al. 
1986a; 1986b). A significant component of the dust popdation was found to be free of 
mineral constituents and was composed of carbon, hydrogen, oxygen, and nitrogen. The 
name "CHON" has been coined to identify this apparent organic material. The formation 
of these organic solids and their incorporation into a comet nucleus are among the more 
interesting questions of cometary science. Greenberg, in many discussions, (see Greenberg 
1989) argues for the mechanism of photoprocessing of precursor interstellar grain mantles. 
However, other processing mechanisms such as charged particle irradiation warrant 
consideration. Inventories of possible charged particles fluxes in various environments 
(Strazzula, 1988; Johnson, 1989) indicate particle fluxes from 10 cm'2 s"i for low energy- 
galactic cosmic rays to 10^ cm"2 s'l for solar flares at 1 AU. The estimated ion irradiation 
dosage for interstellar, circumsteUar, and interplanetary situations are listed in Table 1. 



Interplanetary Dust 
Comet in Oort Cloud 
Interstellar Dust 
Interplanetary Dust 
T Tauri Flare Environment 

^Johnson 1989 





100 keV 

0.1 fim 


> IMeV 

< 0.5 m 


> IMeV 


10^ (loV) 


100 A 


< IMeV 

1 fim 


Still to be explored are the ion environments associated with shock activity in the 
bipolar outflow and rotating molecular disk during the T Tauri stage of the sun. In 
addition, it is important to remember the sun was most likely formed as a member of a 
galactic cluster, and hence, the solar nebula could have been subjected to bipolar outflows 
of other members of cluster. Perhaps at some point the sun may have been in the 
Herbig— Haxo object of a nearby massive protostar. This scenario seems to have neglected 
in discussions regarding the formation of the solar system, and one only has to consider the 
current situation in Orion to recognize its possible impact on models. 

The acceleration of 1 keV ions in the solar wind to >: 100 keV energies in shocks is well 
established from spacecraft measurements of solar system phenomena (see papers in volume 
edited by Arons, McKee, and Max, 1979). The diversity of situations involving shocks as 
particle accelerators in the solar system has been catalogued by Verkatesan (1985) and 
includes: coronal shocks, bow shocks, interplanetary propagating shocks, corotating 
interaction forward and reverse interplanetary shocks, and the solar wind termination 
shock. Voyager 2 measurements in a propagating interplanetary shock have demonstrated 
that shock energy is translated into accelerated ion energy at an efficiency of approximately 
40 percent (Sarris and Krimigis, 1985). During the T Tauri stage of the sun the solar wind 
mass flow could have been 7 to 8 orders of magnitude greater than its present value. This 
suggests an intense moderate energy (>. 100 keV) ion environment as probable. 

Attempts have been made to characterize the ion environment in the early solar system 
through examination of meteorites (Heyman and Dziczkaniec, 1976; Audouze et al., 1976) 
and observation of T Tauri stars ( Worden et al., 1981). Using a scaling argument, 
Strazzulla (1985) has argued that ion— induced solid— state effects are important for the 
chemical evolution of planetary systems in their early stages. The difficulty in specific 
analysis of the ion environment and its impact on chemical processes is that its complexity 
of multiple sources requires more extensive analysis than that which estimates the outcome 
of photoprocessing. In the latter case an estimate of the available ultraviolet radiation is 
accessable from the parameters of photospheric temperature, luminosity, distance, and 
extinction. An attractive aspect of ion irradiation is that it is free of extinction by dust in 
the environment and hence should be an important process in dark clouds. 



If ion— irradiation has been an important process in the chemical evolution of the early 
solar system, then comets are expected to be repositories of the products. Therefore, we 
have undertaken the task of determing the characteristics of materials processed by ion 
bombardment for eventual comparison with cometary samples and contemporary and 
future telescopic observations. Also, experience with laboratory analogs is necessary for 
planning with integrity the sampling techniques and examination procedures to be utilized 
in spacecraft missions to comets. Emphasis to date has been on obtaining infrared spectra 
of non— volatile residues produced by irradiation of ice mixtures with 112"" ions (175 

Gas mixtures composed of Ar, CO, H2O, D2O, N2, CH4, C2H2, and C2D2 are mixed in a 
one hter bulb to desired specifications. Typically, a total pressure of 400 torr is used. The 
gas mixture is bled through a glass capillary into the bombardment chamber where it is 
frozen onto a sapphire disk maintained in the temperature range of 10 K to 15 K. A one 
hour deposition results in approximately 200 torr hters being deposited on the 2.5 cm 
diameter sapphire disk that is cooled by an Air Products Incorporated model 202 closed 
cycle refrigerator. After deposition, the refrigerator and sapphire are rotated 90^ bringing 
the sample into view through two quartz windows. As shown in Figure 1 an ion beam from 
the Van de Graaff accelerator can bombard the frozen matrix at an incident angle of 45°. 
Material eroded from the matrix during ion bombardment can enter the ionizer of the 
quadrupole mass spectrometer. Prior to cooling the sapphire to cryogenic temperatures, 
the chamber is evacuated to a pressure of < 1 x 10"'^ torr using a Varian Star Cell triode 
ion pump. Also prior to gas deposition, the sapphire is raised briefly to a temperature of 60 
K to remove any residual gases other than H2O and CO2 that may have condensed during 
the coohng phase of the experiment. 

Concern about contamination in the form of hydrocarbons from the accelerator beam 
line led to control experiments where Ar and Ar/H20 mixtures were irradiated with H2* 
ions accelerated by a potential of 350 KV (175 keV/proton). Negligible amounts or an 
abscence of residue on the sapphire after sublimation of the .-^r or Ar/H20 mixture leads to 
the conclusion that the level of contamination is < 1% of residues formed from reactive 




Figure 1. Astrophysics bombardment chamber 
for the preparation of species by the bombard- 
ment of cryogenic ice mixtmres with 175 
keV protons. The beam collimator on the left 
serves to reduce contamination from the beam 
line by a reduction in conductance. Gas mix- 
tures may be ionized by electrical discharge 
prior to deposition and may be exposed to UV 
radiation after deposition. 

Quartz window 

0(1 Initt 

lofl collimator 

A nionltor collector 

Quartz window 

Ion gaug« 


Quadrupole mass spectroscopy of ejected species (Figure 2.) during 175 keV proton 
bombardment (as H2* ions accelerated by a potential of 350 KV) of an ice mixture of in 
parts: 170 CO, 170 Ar, and 25 H2O indicates the formation of species at m/q peaks 29, 30, 
32, and 44 suggestive of the CHO radical, CH2O, CH3OH, and CO2. 

Figure 2. Mass spectrum of species 
ejected from a CO, Ar, and H2O ice 
at 20 K during ion bombardment. The 
CO* and Ar* peaks are in excess of 
the partial pressure scale. 



Figure 3 is an SEM image of the residue remaining after sublimation of bombarded ice 
composed in parts of: 170 CO, 25 H2O, 20 N2, and 15 CH4. This mixture was chosen to 
reflect cosmic molecular abundances (Wdowiak et al. 1989). The SEM image was made 
with 1.1 keV electrons and the sample was not coated with a conductive film. This 
indicates it may not be necessary to coat samples while using a spacecraft SEM avoiding an 
obvious complication. Strazzula (1988) warns that conclusions regarding morphology of 
samples prepared at high dose rates may not be valid for astrophysical situations. 

Figure 3. SEM image. 



H2"' ion beam currents of less than 1 /zamp to as great as 10 ^m are used to irradiate a 
target area of approximately 0.25 cin2. This translates into a proton flux between ~ 1 x 
10^3 cm'2 s"i and ~ 5 x 10 i'* cm"2 s"i or an integrated flux between ~ 2 x lOi^ cm"2 and ~ 4 x 
1018 cm"2. Assuming a maximum penetration of 10 //m, dose estimates range from ~ 700 
eV/molecule to ~ 1.4 x 10^ eV/molecule for 175 keV protons incident on a CO ice. The 
experiments cover the higher part (~ 10^ — ~ 10^ eV/molecule) of the dose range listed in 
Table 1. The integrated flux of 4 x IQis cm'2 compares well with the estimate of Andouze 
et al. (1976) of ~ 3 x 10^8 cm"2 for the irradiation of the amorphous grains of carbonaceous 
chondrites with protons having energies > 100 keV. 

The color of the residues depends upon dosage and range from brownish yellow to black 
with a brown tinge on the periphery. Samples are prepared for FTIR absorption 
spectroscopy by dispersing the residue into 100 mg spectroscopic grade KBr in a two ball 
agate vibrating mill. The mixture is pressed into a 0.25 cm2 pellet at a pressure of 20,000 
PSI. Spectra are obtained with a Mattson Polaris FTIR spectrometer operating in the 
4000 cm"i to 400 cm'i (2.5/mi to 25/fln) range. 

The spectrum of the bombarded ice brown residue (displayed in Figure 3) is shown in 
Figure 4 along with material vaporized from the acid insoluble residue of the Orgueil CI 
carbonaceous chondrite onto a KBr crystal (Wdowiak et al., 1988). The ice residue 
exhibits aliphatic C— H stretch at 2940 cm'i (3.4 /m), C=0 stretch at 1710 cm'i (5.85 /im), 
C=C stretch at 1620 cm'i (6.2 /mi), and features at 1380 cm-i (7.24 fjm) and 1100 cm-i (9.1 
/^m). These features are in common with some of those of the volatile fraction of the acid 
insoluble residue of the Orgueil CI carbonaceous chondrite. This particular residue was 
prepared at an estimated dose of ~ 1 x 10^ eV/molecule composed of parts: 170 CO, 25 
H2O, 20 N2, and 15 CH4. 

Wavelength O^m) 


Figure 4. FTIR spectrum of bombarded 
ice residues prepared at a dose of 
^ 1 X 104 eV/molecule using 175 keV 
protons, and that of the acid insoluble 
residue of the Orgueil CI meteorite. 

S 910 





25 HjO - 

20 Nj. AND 15 a^^ 


« ' I ' ' ' I ' ' ' ' ' ' ' ' I ' ' ' t I I I . I r ■ I I . I I I 










Irradiation at a current and duration leading to an estimated dose of ~ 5 x 10'* 
eV/molecule results in black with brown tinged periphery residue prepared from an ice in 
parts of: 170 CO, 170 Ar, 25 H2O, 20 N2, and 15 CH4. The FTIR spectrum (Figure 5) 
besides exhibiting aliphatic C-H stretch at 2940 cm"i (3.4/mi) and the feature at 1400 cm"i 
(7.1/im), is significant in that it has strong features that correlate with the observed 
celestial unidentified infrared bands (UIRs) emitted by numerous cosmic sources ranging 
from bipolar nebulae to galaxies. First discovered by Gillett, Forrest, and Merrill (1973) 
the UIRs are found in emission at 3.3 fjxa (3030 cm-i), 6.2 /zm (1620 cm'i), 7.7 //m (1300 
cm'i), 8.6 fjxn (1163 cm"i), and 11.3 fim (885 cm'i). Because the laboratory material has a 
brown component, the two aliphatic (and non— UIR) features may be due to that 
component and not the material responsible for the features at UIR wavelengths. That 
possibility is currently under investigation. Figure 6 shows the high degree of correlation of 
the features of the laboratory material with the four of the five predominant UIRs. 
Especially significant is the exact coincidence with the 7.7 /an and 8.6 fxm UIRs. The 
principle deviation is with the 11.3 /mi UIR thought to be due to the out-of— plane 
aromatic hydrogen wag (Cohen et al., 1985). However, the 11.6 /im feature of the 
laboratory material is a better match than that of the 11.9 /mi feature of coronene as 
proposed by Leger and Puget (1984), although not as good as the 11.3 /zm match achieved 
by thermal treatment of the Orgueil Cl acid insoluble residue (Wdowiak et al, 1988). 

tlPta MO» rosiTlou or j.j. «.i. 7.7, ,.4, ,„a i,.j kjciok oi»-» 
thl«*r>lt)f of llilno - Blraln^n PILARIS 
' I ' I '''■'■■■■'■■■ ■'■■■■ I ■ 

Figure 5. FTIR spectrum of bombarded 
ice residue prepared at a dose of ~ 5 
X 10'* eV/molecule showing coincidence 
with four of the five UIR bands at 
frequencies marked by vertical lines. 

•l^iol gain i 

SOD 2000 


rMolutlon I 4 GB-1 



The formation of solid material from the ion bombardment of reactive cryogenic ices is 
of interest as a laboratory analog for the CHON particles detected during the Giotto and 
Vega encounters with Comet Haliey (Kissel et al., 1986a; 1986b). During the T Tauri 
stage of the sun, extensive flare activity and a solar wind mass flow of 7 to 8 orders of 
magnitude of its present value would have produced an intense moderate energv- ion 



Figure 6. FTIR spectrum shown in Fig. 5 
plotted against a micron scale showing 
the high degree of coincidence with four 
of the five UIR bands. 

environment. Considerable shock activity would have been present resulting in 1 keV T 
Tauri wind ions accelerated by shock processes to > 100 keV. These ions impacting on 
cryogenic reactive ice mantles of dust grains and centimeter size ice "chunis" would 
produce organic sohds. Aggregation of the bombarded ice mantled grains and "chunks" 
into a cometary nucleus would result in particulates being distributed throughout. The 
chemical nature of the organic solids would depend upon the dosage with aliphatics 
dominant at low dosage and aromatics at higher dosage. Aromatic material has been 
detected by nuclear magnetic resonance in the acid insoluble residue of the Orgueil CI 
carbonaceous chondrite (Cronin et al., 1986), and by Raman spectroscopy in the 
interplanetary dust particle (IDP) Essex (AUamandola et al, 1987). Bipolar outflow 
appears to be ubiquitous occurring in planetary nebulae such as the Red Rectangle 
(Schmidt et al., 1980; Warren-Smith et al., 1981) and would be expected to be 
accompanied by shock activity. Ion acceleration in shocks and the irradiation by these ions 
of cryogenic reactive ice mantles may be a mechanism through which the species 
responsible for the UIR bands are formed. Ion irradiation may also be an important 
mechanism in active galaxies such as M 82 and in dusty regions in general. 

Continuing experiments are focusing on the relationship between dosage and the 
development of infrared spectral features as suggested by comparison of the ice residue 
spectra in Figures 4 and 5. 

TJW and GCF wish to acknowledge the support of NASA Grant NAGW-749. 



Allaxnandola, L. J., Sanford, S. A., and Wopenka, B. (1987) Interstellar Polycyclic 
Aromatic Hydrocarbons and Carbon in Interplanetary Dust Particles and 
Meteorites. Science, 237, 56-59. 

Arons, J., McKee, C, Max, C, eds (1979); Particle Acceleration Mechanisnas in 
Astrophysics, American Institue of Physics. 

Audouze, J., Bibring, J. P., Dran, J. C, Maurette, M., and Walker, R. M. (1976) Heavily 
Irradiated Grains and Neon Isotope Anomalies In Carbonaceous Chondrites. Ap. J. 
(Letters) 206, L185-L189. 

Cohen, M., Tielens, A. G. G. M., and AUamandola, L. J., (1985) A New Emission Feature 
in IRAS Spectra and the Polycyclic Aromatic Hydrocarbon Spectrum. Ap. J. 
(Letters) 299, L93. 

Cronin, J. R., Pizzarello, S., and Frye, J. S. (1987) ^^C NMR Spectroscopy of the Insoluble 
Carbon of Carbonaceous Chondrites. Geochim. Cosmochim. Acta 51, 299—303. 

Gillett, F. C, Forrest, W. J., and Merrill, K. M. (1973) Ap. J. 183, 87. 

Greenburg, J. M. (1989) From Interstellar Dust To Comets. LPI Contribution No.691. 
Analysis of Returned Comet Nucleus Samples, 22—23. 

Heymann, D., and Dziczkaniec, M. fl976) Early Irradiation of Matter in the Solar System: 
Magnesimn (Proton, Neutron) Scheme. Science 191, 79-81. 

Johnson, R. E. (1989) Radiation Modification of Cometary Materials: Laboratory 
Simulations. LPI Contibution No. 691, Analysis of Returned Comet Nucleus 
Samples, 32. 

Kissel, J., Sagdeev, R., Bertaux, J., Angarov, V., Audouze, J., Blamont, J., Buchler, K., 
Evlanov, E., Fechtig, H., Fomenkova, M., von Hoerner, H., Inogamov, N., 
Khromov, V., Knabe, W., Krueger, F., Langevin, Y., Leonas, V., 
Levasseur— Regourd, A., Managadze, G., Podkolzin, S., Shapiro, V., Tabaldyev, S., 
Zubkov, B. (1986a) Composition of Comet Halley Dust Particles from Vega 
Observations. Nature 321, 280-282. 

Kissel, J., Brownlee, D., Buchler, K., Clark, B., Fechtig, H., Grun, E., Hornung, K., 
Igenbergs, E., Jessberger, E., Krueger, F., Kuczera, H., McDoimell, J., Morfill, G., 
Rahe, J., Schwehm, G., Sekanina, Z., Utterback, N., Volk, H., Zook, H. (1986b) 
Nature 321, 336-337. 

Leger, A., and Puget, J. L. (1984) Identification of the Unidentified IR emission features of 
Interstellar Dust? Astr. Ap. 137, L5-L8. 

Sarris, E. T., and Krimigis, S. M. (1985) Quasi-Perpendicular Shock Acceleration of ions 
to ~ 200 MeV and Electrons to ~ 2 MeV Observed by Voager 2. Ap. J. 298, 


Schmidt, G. D., Cohen, M., and Margon, B. (1980) Discovery of Optical Molecular 
Emission From the Bipolar Nebula Surrounding HD 44179. Ap. J. (Letters) 239 . 

Strazzulla, G. (1984) Modifications of Grains by Particle Bombardment in the Early Solar 
System. Icarus 61, 48—56. 

Strazzulla, G. (1988) Ion Bombardment: Techniques, Materials and Applications. In 
Experiments on Cosmic Dust Analogues (eds. E. Bussoletti, C. Fusco, and G. 
Longo). Klower Academic Publishers, Dordrecht, 103—113. 

Venkatesan, D. (1985) Cosmic Ray Picture of the Heliosphere. Johns Hopkins APL 
Technical Digest 6 No. 1 . 4-19. 

Warren-Smith, R. F., Scarrott, S. M., and Murdin (1981) Peculiar Optical Spectrum of 
the Red Rectangle. Nature 292, 317-319. 

Wdowiak, T. J., Flickinger, G. C, and Cronin, J. R. (1988) Insoluble Organic Material of 
the Orgueil Carbonaceous Chondrite and the Unidentified Infrared Bands. Ap. J. 
(Letters) 328, L75-L79. 

Wdowiak, T. J., Donn, B., Nuth, J. A., Chappelle, E., and Moore, M. (1989) Laboratory 
Experiments On Carbonaceous Material As a Source For the Red Rectangle Visual 
Emissions. Ap. J. 336, 838-842. 

Worden, S. P., Schneeberger, T. J., Kuhn, J. R., and Africano, J. L. (1981) Flare Activity 
On T Tauri Stars. Ap. J. 244, 520-527. 



Friedrich Begemann 

Max-Planck-Institut fur Chemie 

Mainz, FRG 



Friedrich Begemann 
Max-Planck-Institut fiir Chemie, Mainz, FRG 

The understanding was that this presentation should consist of two 
parts, one about what one may hope to learn from the isotopic analysis in 
the laboratory of cometary matter, and a second part on whether 
present-day analytical methods are adequate to reach these goals, where 
improvements are required and what needs to be developed in order to 
optimize the scientific return. The understanding was, furthermore, that I 
should report on isotopics and the analysis of heavy elements and noble 
gases by conventional mass spectrometry, but that I should neither concern 
myself with light elements like H, C, and N, nor should I deal with the 
potential of ion probes. 

I. Problems of interest 

Comets are still believed to be a conglomerate of ices and 
meteoritic dust in a ratio of about 5:1 (if we take the ratio ices/dust to 
be equal to the ratio gas/dust). When Whipple proposed his "dirty 
snowball" model in 1950 he envisaged a single, well consolidated body; 
recent refinements are the icy-glue model for the cometary nucleus of 
Houpis and Gombosi (1986) or the "primordial rubble pile" of Weissman 
(1986). These refinements appear to pertain essentially to the macroscopic 
structure of the nucleus on the scale of tens of centimeters to hundreds 
of meters, however, so they need not concern us in the present context, 
since we are interested in the structure on a much smaller scale. What we 
should like to know is the size distribution of non-gaseous matter 
("gaseous" at elevated temperatures like room temperature) which we may 
expect in a kg-sized sample. The most recent and most comprehensive data 
available on this topic are those obtained for comet Halley which may or 
may not be typical of comets - but then we do not know, of course, how 
"typical" the first returned cometary sample will be either. 

According to McDonnell et al. (1987) the grain size distribution, at 
the cometary surface, is such that the nximber of grains decreases steeply 
with increasing grain size, but the exponent in the power law distribution 
appears to be < 3 over the whole size range analysed which is from about 
1 |im to 1 cm, corresponding to a range in mass from ca. 10" •'•-'g to 1 g. 
The fact that the exponent in the power- law is < 3 means, of course, that 
most of the mass occurs in large grains. Still, if one kg of cometary 
matter were to be returned, it should contain a hundred gram or so of 
silicates, and a few milligram of these would be grains with masses of 
10 ixg or less. 


The question what one can do with these grains, and what one may 
hope to learn from their analysis, is related to the nature of the grains. 
We anticipate to find brittle, friable, loosely connected aggregates of 
primordial grains, a collection of "stardust" and of pristine grains from 
the proto-solar cloud. And we hope that the "yellow stuff" of Greenberg 
[cf. e.g. Greenberg and Grim, 1986] plays a minor role, at least in the 
sense that individual fragments of the fragile grains may be covered by 
it, but the grains as a whole are not. Presimiably, this would make the 
disintegration of the grains into their original constituents much easier 
as if everything were embedded in "yellow stuff". 

The reason why one would wish to do this and why one wants to have a 
look at the grains in their original state is, of course, the experience 
we have with carbonaceous chondrites. From several of them, in particular 
the CM 2's Murray and Murchison, it has been possible to isolate rare 
constituents which are more or less anomalous in the isotopic composition 
of all their elements (H, C, N, Ne, Si, Kr, Xe; (Swart et al., 1983; Lewis 
et al., 1983; Yang and Epstein, 198A; Zinner et al., 1987; Ott et al., 
1988; Tang Ming and Anders, 1988). Since the anomalies are as they are 
predicted for different processes of nucleosynthesis, the explanation is 
that these grains are "stardust", i.e., that they are unadulterated 
primary condensates from the stars in which the elements were produced. 
Such "stardust" as it is known from astronomical observations and from 
meteorites is in the sub-micrometer size range, and it is therefore grains 
of this size which should be analysed, one by one, or which should be 
pooled. The only reason why this has been possible in meteorites is that 
nature has been kind to us by providing the isotopically anomalous 
elements in phases with an extreme chemical resistivity - carbonaceous 
matter, diamonds, SiC (Zinner et al., 1987, and references therein) - 
which makes them the insoluble residue when one subjects the carbonaceous 
chondrites to an extended and harsh chemical treatment. But there is no a 
priori-reason why all "stardust" should be chemically refractory; perhaps 
the inertness of a few such components is just meant as a hint for us to 
look for others which might be as interesting and as rewarding as the ones 
which have been found so far. 

Aside from attempts to identify and isotopically analyse individual 
grains in order to better characterize the different nucleosynthetic 
contributions to the proto-solar nebula there are other questions which 
can perhaps be answered by looking at larger ensembles of grains. One is 
the age of the solid matter. There may be two problems, however. The first 
is related to the small grain size of the "stardust". Since all 
radiometric clocks depend on one kind or other of radioactive decay, there 
is always a certain fraction of the total decay energy imparted to the 
daughter product and this may make it recoil out of where it belongs. How 
serious this problem is depends on which decay one wants to utilize - 
fission transfers a much larger kinetic energy to its fragments than 
does a -decay which, in turn, is orders of magnitude more effective than 
3-decay. Typical ranges of the recoil particles are 15 pm for fission 
fragments, 15 nm (i 150 A) for a-decay and distances of a few atomic 
diameters for p-decay. The severity of the problem of losses by recoil 


depends then on the commensurability of these recoil ranges with the grain 
size of the host material, on the average distance between grains, and on 
what the interstitial space is filled with. What should be done in order 
to avoid these problems is a "whole rock" analysis on chunks of cometary 
matter which include all the original constituents - solid, liquid and 
gaseous (under normal lab conditions). And one might look at the ices 
separately and see whether they are enriched in radiogenic nuclides which 
would give an indication whether the effect is something to worry about. 

The second problem is one of interpreting the data. If the meteoritic part 
of cometary matter was never isotopically and elementally equilibrated it 
will not be possible to determine an internal isochron, which would date 
the time since equilibration, and the meaning of an age will not be 
immediately obvious. After all, if cometary matter were a representative 
sample of pre-solar system matter, it would make no difference whether the 
radiometric clocks were running in the dilute pre-solar cloud or in the 
condensed cometary matter, except that condensed matter would be cut off 
from galactic nucleosynthesis. As to this latter point it would be 
interesting to know the time since cometary matter was decoupled from 
galactic nucleosynthesis. One way to measure this is to look at the 
relative abundance of radionuclides like '^^K and °'Rb, or to measure 
the ^^^U/^^°U abtindance ratio. Both U-isotopes are produced in the 
r-process of nucleosynthesis which is, of course, spotty in time and 
space, but on time scales of the half life of ^-^^U and 2^°U it is 
fairly regular. In the simplest case of a production rate constant in time 
the situation is as depicted in Fig. 1. The concentration of both isotopes 
increases with' time until an equilibriiom is reached where the production 
rate equals the decay rate. ^-^^U, because of its shorter half life, will 
reach this equilibriim faster than ^-^^U and consequently the abundance 
ratio '^-^^U/'^-^°U changes with time, it slowly decreases. Once uranium 
is cut off from its source so that nothing is being produced any more the 
ratio decreases much faster, almost with the half life of ^^^U. 
Actually, the ratio changes so fast that its exact value is rather 
sensitive to the time of decoupling: a difference of 100 Ma ensues in a 
change of the ratio of ca. 8%. 

The results presently available for solar system matter are 
notoriously messy. For terrestrial and lunar samples the ratio has 
invariably been found to be constant, but for meteorites different 
laboratories have measured quite contradictory results (cf. Shimamura and 
Lugmair, 1981 and references therein). Still, for bulk samples the total 
range reported so far is less than 2% which would indicate that 
meteoritic, terrestrial and lunar matter all were decoupled from r-process 
nucleosynthesis within 20 Ma or so. (It should be mentioned, though, that 
most of the production of the U isotopes occurs via their radioactive 
precursors and that there are other explanations possible for variations 
in the 235u/238u ratio (cf. e.g. Tatsumoto and Shimamura, 1980)). For 
cometary matter one might expect a less strict contemporaneity, be it 
because the outer reaches of the proto-solar cloud were still in contact 
with interstellar matter at a time when the inner parts had been decoupled 
already, or because (some) comets are errant messengers from other worlds 
altogether . 


Figure 1: Idealized sketch of the development of r-process ^^^U and 
-■^°U and of their abundance ratio, a) The absolute 
concentrations increase with time until an equilibrium is 
reached between production and decay. Note, that the assumption 
of a production rate constant in time is an oversimplification 
and that the exact value for the ratio of the production rates 
(which include production via radioactive precursors) as well as 
the duration of r-process nucleosynthesis are contentious 
[Thielemann et al. , 1983,; Fowler, 1987; Clayton, 1988]. This, 
however, does not affect the essential point discussed here, 
b) The abundance ratio 235u/238u ^^ interstellar matter 
decreases with time because 2-^°U (T2^y'2 ~ A.A7 x lO^a) 
approaches its equilibrium concentration slower than does ^-^^U 
(T]_/2 ~ 7.04 X 10°a). At the time, matter is decoupled from 
galactic production the ratio starts to change very fast, due to 
the much faster decay of ^-^^U as compared to -^-^"U. If the 
decoupling of two parcels of matter is separated in time by AT, 
the present-day •^•^-'U/'^-^ U ratio in the two parcels will be 
different by 6R. 

10 - 

10 - 

P =2xp = constant in time 

H'u -^'' 

"'^ 235,, 






■ ^^^^ ^ 






Decoupling from galactic V" 

production \ \ 



terrestrial = 

7?6xin' \ 


1 1 t I 1 1 1 1 1 1 

, , , ,n 


15 AE 


II. Experimental requirements 

A solution of the problems I have been sketching requires mass 
spectrometric isotope abundance analyses. Since we do not know what to 
expect, it is difficult to say which accuracy will be necessary in order to 
reach meaningful conclusions. What we should strive for, however, are 
uncertainties in isotope abundance ratios of less than 1 percent. If we 
take this number as a basis for our discussion what, then, is the minim\im 
amoiint of material required to reach this goal? 

A mass spectrometric analysis consists essentially of three steps 


Transmission of ions 

Detection of ions. 
Independent of the exact mode of detection, whether it is measuring 
currents or ion counting, the entity which determines the ultimate 
precision that can be reached is the number of collected ions. Since the 
statistical error connected with registering n independent events is Vn the 
relative error is o = Vn/n = 1/Vn so that a statistical uncertainty of 1 
percent requires that 10 ions be registered. Let us assume that the 
relative abundance of the particular isotope we want to measure is 10% and 
that the atomic weight of the element is 60. What is needed then is 10 x 
1/0.1 X 60/6 X 10^-^ = 10"-'-'g of ions of the element in question. If we 
assume further that the abundance of our element in the grains is 1% by 
weight, then the grain must have a mass of 10~-'--'g which corresponds to a 
diameter of ca 0.1 \sm. 

This is the absolute minimiom amount of material required since the 
estimate implies, first, that there are no errors other than the 
statistical ones connected with the number of coimted ions and, second, 
that all atoms are ionized and registered. In reality the total efficiency 
for detection is 

^ total ~ '^ionisation ^ ^transmission ^ '^registration 

Table 1 shows a compilation of data from the literature for the product 

^ionisation ^ ^transmission- ^^^ measurements were performed with 
modem mass spectrometers for which the transmission is better than 50% so 
that, within a factor of two, the entries in column 3 are essentially the 
ionisation efficiencies. Note that in all cases except the noble gases, 
emitters have been used which for some not well -understood reason greatly 
enhance the ionisation efficiency. Actually, they do this in such ein 
irregular way as to obliterate the dependance on ionisation energy of the 
ionisation probability as one finds it for thermal ionisation without 
emitter, and they do it in such an unpredictable way that so far trial and 
error is just about the only way to find efficient emitters. 

From this compilation it is obvious that, using present technologies, the 
amount of matter required is typically at least hundred times the minimiim 
amount calculated above. "At least", since the registration efficiency is 
not vinity which would require, first, that all ions at the receiving end of 
the mass spectrometer are collected and, second, that the sample is run to 
exhaustion. Collecting all ions is not possible in the normal sequential 


scanning mode where the magnetic field is changed in steps and one 
isotopic species is collected while all others go xindetected. It rather 
necessitates using a multiccllector which has the additional advantage 
that, at least in principle, irregularities in the ion beam intensity 
cancel out so that abundance ratios can be measured even if the signal is 
very noisy. Modern technology is sufficiently advanced that one need not 
worry any more about the vagaries of working with a number of independent 
amplifiers or counting systems, at least not at the level of precision we 
are talking about. 

Element Ej[eV] '^ion^ "trans. Emitter Reference 





Ti (as TiO+) 











7. A 



























Si-gel, H3PO4 


















e ~ - bombardment 
e ~ - bombardment 



(1) Jungck, M.H.A. et al. , 1984 

(2) Niederer, F.R. et al., 1980 

(3) Birck, J.L. and Allegre, C., 1984 

(4) Birck, J.L. Lugmair, G.W., 1988 

(5) Poths, H. et al., 1987 

(5) Chen, J.H. and Wasserburg, 

G.J., 1980 
(7) Hohenberg, CM., 1980 

Kirschbaiom, C. 1988 

Running a sample to exhaustion may not appear to be difficult. 
Presently, however, the emphasis is more on stable ion beams, and for many 
elements these have only been attained at rather low intensities. In this 
case running a sample to exhaustion means long times of measurements. 
This, by itself, is no problem, but any ion detection device has its 
background and it clearly would be advantageous to have a signal which 
normally lasts for 20 hours compressed into, say, 20 minutes if one had a 
corresponding increase in the signal/noise ratio. Again, use of a 
multiccllector should alleviate much of the need for stable ion beams, but 
it remains to be shown how much compression in time of the signal can be 
done without deterioration of the ionisation yields. 


Finally, in order to avoid contamination, handling of the cometary 
samples should be kept to an absolute minimum. Preferably, individual 
crystals should be loaded directly onto the filament of the ion source 
whithout any chemical treatment whatsoever. Lee et al. (1977) have shown 
that <0.4ng of Mg from an anorthite crystal with Al/Mg >280 can be 
isotopically analysed with a precision of better than 1 percent. But 
again, it remains to be shown that "direct loading" can 3e done without 
much reduction in ion yields and in particular that a combination of 
"direct loading" and signal compression still allows to distinguish 
between a genuine anomalous isotopic composition of an element and an 
apparent one caused by compromising interferences on one or more of the 

Problems with such compromising interferences must indeed be 
anticipated once the analyses are performed in such a way that control 
measurements are not possible. One way to regain specificity in the sense 
that only one element at a time is ionized and detected is to utilize 
laser resonance ionisation. This together with laser evaporation of the 
samples would certainly be a very clean method of analysis. Samples could 
be measured without doing chemistry, thus avoiding the inevitable problem 
of contamination, and the amount of hot structural material in the ion 
source would be kept to a bare minimum which again would help to reduce 
contamination. Of course, this would make it impossible to take advantage 
of the enhancement by orders of magnitude of the ion yield by using 
emitters so that it remains to be seen whether the advantages outweigh the 
drawbacks. And it must be realized that by this method it will be a 
non-trivial matter to analyse more than one element per grain which makes 
the search for correlated anomalies difficult, if not impossible. 

For noble gases the situation is somewhat different and in a sense 
much more advanced already and closer to the theoretical limits. High- 
sensitivity ion sources require only ca. 15 000 atoms of Ar or Xe to yield 
a count rate of 1 count per second (cps) which should be compared with a 
background count rate of 0.1 cps or less (Kirschbaum, 1988). Transmission 
X ionisation yield are >0.5 (Baur, 1980; Hohenberg, 1980) so that more 
than 50% of all gas atoms are available for registration. So far this 
potential has not been fully utilized, however, because ion collection is 
still done sequentially with single collectors. In the case of Kr and Xe 
with their many stable isotopes this means that a factor of 10 or more is 
given away in the total number of ions. Use of a multicollector would 
again be the way out but in the case of gases it would not only be 
beneficial. It would result in a considerable increase in volume of the 
mass spectrometer and, since noble gases are always measured under static 
conditions, a corresponding decrease of gas pressure in the spectrometer. 
A reduced particle density in the ion source, in turn, results in a lower 
ion beam intensity so that the advantage of the multicollector that a 
larger fraction of the total number of ions is collected is offset in part 
by a lower signal/noise ratio. 

III. Conclusion 

It would appear that no major break-throughs or innovations are required 


in order to exploit the isotopic information we expect to be contained in 
cometary "solid" matter. What is needed, however, is combining a number of 
technologies into a single instrument and a demonstration that the whole 
is not more than but equal to the sum of its parts. Preferably, such an 
instrument should also provide the possibility to characterize by 
appropriate non-destructive means single grains prior .to their destructive 
isotope analysis. 

References : 

Baur, H.: Numerische Simulation und praktische Erprobung einer 

rotationssymmetrischen lonenquelle fiir Gasmassenspektrometer. Ph.D 
dissertation, Swiss Fed. Inst. Technology, 1980. 

Birck, J.L.; and Allegre, C.J.: Chromium Isotopic Anomalies in Allende 
Refractory Inclusions. Geophys. Res. Lett., vol. 11, 1984, pp. 

Birck, J.L.; and Lugmair, G.W.: Nickel and Chromium Isotopes in Allende 
Inclusions. EPSL, vol. 90, 1988, pp. 131-143. 

Chen, J.H. ; and Wasserburg, G.J.: A Search for Isotopic Anomalies in 
Uranium. Geophys. Res. Lett., vol. 7, 1980, pp. 275-278. 

Clayton, D.D*-: Nuclear Cosmochronology within Analytical Models of the 

Chemical Evolution of the Solar Neighbourhood. Mon. Nat. R. astr. Soc. 
vol. 234, 1988, pp 1-36. 

Fowler, W.A. : The Age of the Observable Universe. Quarterl. J. R. astr. 
Soc. vol. 28, 1987, pp. 87-102. 

Greenberg, M.J.; and Grim, R. : The Origin of Comet Nuclei and Comet 

Halley Results. 20th ESLAB SYMPOSIUM on the Exploration of Halley's 
Comet. Proceedings of the International Symposium Heidelberg, Germany 
27-31 October 1986, pp. 255-263. 

Hohenberg, CM. : High Sensitivity Pulse -Counting Mass Spectrometer System 
for Noble Gas Analysis. Rev. Sci. Instrum. , vol. 51, 1980, pp. 

Houpis, H.L.F.; and Gombosi, T.I.: An Icy-Glue Nucleus Model of Comet 
Halley. 20th ESLAB SYMPOSIUM on the Exploration of Halley's Comet. 
Proceedings of the International Symposium Heidelberg, Germany 27-31 
October 1986, pp. 397-401. 

Jungck, M.H.A. ; Shimamura, T. ; and Lugmair, G.W. : Ca Isotope Variations in 
Allende. Geochira. Cosmochim. Acta, vol. 48, 1984, pp. 2651-2658. 

Kirschbaum, C: Carrier Phases for Iodine in the Allende Meteorite and 
their Associated ^^^Y.e.^l^^'1 ratios: A Laser Microprobe Study. 
Geochim. Cosmochim. Acta, vol. 52, 1988, pp. 679-699. 

Lee, Typhoon; Papanastassiou, D.A.; and Wasserburg G.J. : Mg and Ca 

Isotopic Study of Individual Microscopic Crystals from the Allende 
Meteorite by the Direct Loading Technique. Geochim. Cosmochim. Acta, 
vol. 41, 1977, pp. 1473-1485. 


Lewis, R.S.; Anders, E. ; Wright, I. P. et al.: Isotopically Anomalous 

Nitrogen in Primitive Meteorites. Nature, vol. 305, 1983, pp. 767-771. 

McDonnell, J. A.M.; Alexander, W.M.; Burton, W.M. et al.: The Dust 

Distribution within the Inner Coma of Comet P/Halley 1982i: Encounter 
by Giotto's Impact Detectors. Astron. Astrophys. vol. 187, 1987, pp. 

Niederer, F.R. ; Papanastassiou, D.A. ; and Wasserburg, G.J.: Endemic 

Isotopic Anomalies in Titanium. Astrophys. J., vol. 2A0. 1980, pp. 

Ott, U.; Begemann, F. ; Jongmann Yang; and Epstein, S.: S-Process Krypton 
of Variable Isotopic Composition in the Murchison Meteorite. Nature, 
vol. 332, 1988, pp. 700-702. 

Poths, H. ; Schraitt-Strecker, S.; and Begemann, 7.: On the Isotopic 

Composition of Ruthenium in the Allende and Leoville Carbonaceous 
Chondrites. Geochim. Cosmochim. Acta, vol. 51, 1987, pp. 1143-1149. 

Shimamura T. ; and Lugmair G.W. : U-isotopic Abundances. Lunar Planet. Sci. 
XII, 1981, p. 976-978. 

Swart, P.K.; Grady, M.M.; Pillinger, C.T. et al.: Interstellar Carbon in 
Meteorites. Science, vol. 220, 1983, pp. 406-410. 

Tang Ming; and" Anders, E. : Isotopic Anomalies of Ne, Xe, and C in 

Meteorites. III. Local and Exotic Noble Gas Components and their 
Interrelations. Geochim. Cosmochim. Acta, vol. 52, 1988, pp. 

Tatsumoto, M. ; and Shimamura, T. : Evidence for live •^^'Cm in the Early 
Solar System. Nature, vol. 286, 1980, pp. 118-122. 

Thielemann, F.-K.; Metzinger J.; and Klapdor H.V. : Beta-Delayed Fission 

and Neutron Emission: Consequences for the Astrophysical r-Process and 
the Age of the Galaxy. Z. Phys. A - Atoms and Nuclei, vol. 309, 1983, 
pp. 301-317. 

Weissman, P. A.: Are Cometary Nuclei Primordial Rubble Piles? Nature, vol. 
320, 1986, pp. 242-244. 

Whipple, F.L.: A Comet Model. I. The Acceleration of Comet Encke. 
Astrophys. J., vol. Ill, 1950, pp. 375-394. 

Yang, Jongmann; and Epstein, S.: Relic Interstellar Grains in Murchison 
Meteorite. Nature, vol. 311, 1984, pp. 544-547. 

Zinner, E. ; Tang Ming; and Anders, E. : Large Isotopic Anomalies of Si, C, 
N and Noble Gases in Interstellar Silicon Carbide from the Murray 
Meteorite. Nature, vol. 330, 1987, pp. 730-732. 



John R. Cronin 

Arizona State University 

Tempe, Arizona 



John R. Cronin 

Arizona State University 
Tempe, Arizona 


Comets are generally believed to be primitive bodies that preserve solar system 
matter in, or nearly in, its primordial state. This expectation has been at least 
partially borne out by the 1986 flyby missions to Comet Halley which provided data 
indicating that, with the exception of hydrogen, the light elements (C, N, 0, and S) 
occur in approximately their solar abundances (Delsemme, 1988). Although mass 
spectrometers carried aboard the spacecraft provided much additional data from which 
to speculate about the molecular forms of these elements (Kissel and Krueger, 1987), 
a detailed understanding of cometary organic chemistry will ultimately require the 
laboratory examination of returned samples. 'Some of the problems that will be 
encountered in such studies, for example, sensitivity to trace constituents, 
resolution of numerous isomeric forms, and avoidance of terrestrial contaminants, 
have already been faced in analyses of the organic compounds from carbonaceous 
chondrites. (See Cronin et al., (1988) for a recent review.) Furthermore, there is 
reason to believe that the progenitors of the carbonaceous chondrites were volatile- 
rich planetesimals (Bunch and Chang, 1980) similar to those which, at greater radial 
distances, formed comets. Thus, the organic chemistry of carbonaceous chondrites may 
represent the outcome of a process of chemical evolution that parallels, although is 
perhaps further advanced than, that which occurred in comets. These meteorites may 
then represent not only a useful model for the development and refinement of 
analytical methods, but also a guide to the types of organic compounds that may be 
encountered in analyses of cometary matter. 

In what follows, I have (i) briefly reviewed the results of amino acid analyses 
of CM chondrites, (ii) discussed the origin of these compounds and the implications 
for comet organic chemistry, and (iii) described some recent developments in 
analytical instrumentation for amino acids and their implications for analyses of 
extraterrestrial materials. Although the emphasis is on amino acids, their general 
characteristics are common to the other classes of organic compounds in CM chondrites 
and inferences regarding their origins should be generally relevant (Cronin et al., 


Amino acids can be extracted from crushed samples of CM chondrites with hot 
water and isolated, free of coextraoted inorganics, by adsorption onto a cation 
exchange resin and elution with dilute NH^OH (Kvenvolden et al . , 1970). When 
analyzed by ion-exchange chromatography with o-phthal aldehyde detection, the 
Murchison extract gives a complex chromatogram (Cronin and Pizzarello, 1986). The 
majority of the amino acids, certainly all the more abundant ones, have now been 
positively identified by GC-MS using ion exchange and/or reverse phase chromatography 


for preparative fractionation. All the amino acids identified to date fall into one 
of two general classes; they are either monoaminoalkanoic acids or 

monoaminoalkandioic acids. The monoaminoalkanoic acids also include cyclic forms and 
N-alkyl derivatives. Structures of these amino acids are illustrated in figure 1 . 
It is interesting to note that, of the more than 80 amino acids now positively 
identified, only eight protein amino acids and 11 non-protein, but biologically 
occurring, amino acids have been found. Thus, the majority of the amino acids are 
uniquely extraterrestrial. 

As might be expected, given the large number of amino acids and only two 
structural classes, they show little structural selectivity, that is, all the stable 
amino acid structures are apparently represented in the meteorite within the limits 
set by the two general classes. Analyses of the seven-carbon a-amino 
monocarboxylic acids have provided the most stringent test of this point (Cronin and 
Pizzarello, 1986). This group is comprised of 14 chain isomers, four of which have 
two chiral centers and thus have diastereomeric forms. As a result, there are 18 
isomers separable by normal, i.e., achiral, chromatographic systems. All 18 isomers 
have been identified in extracts of the Murchison meteorite. It should also be noted 
that both amino acid enantiomers have been found to be present in many instances, and 
in nearly equal amounts (Kvenvolden et al., 1970). 

Quantification of the Murchison amino acids is summarized in figure 2. Both the 
total amino alkanoic acids and the total a-amino alkanoic acids show declining 
amounts with increasing carbon number but with a pronounced maximum 
at C^ (fig. 2A) ; the abundance of the amino position isomers is a>Y>B. When 
the a-amino acid abundances are presented as a function of carbon number within 
homologous series (fig. 2B), a smooth, exponential decline in content is observed 
within each homologous series. At each carbon number, the content of individual 
branched-chain isomers is greater than that of the straight-chain isomer. This fact 
seems to explain the maximum at C^ in the plots of total amino alkanoic acids and 
total a-amino alkanoic acids, in that C^ is the minimal chain length at which chain 
branching becomes possible. In summary, the Murchison amino acids have the general 
characteristics listed in figure 3. These characteristics are shared by many of the 
other classes of organic compounds found in carbonaceous chondrites, e.g., the mono- 
and dicarboxylic acids (Lawless and Yuen, 1979; Peltzer et al., 1984), thus we 
believe them to be products of a common synthetic process. 


The origin of the organic compounds of carbonaceous chondrites has been a 
controversial question. Two hypotheses have dominated the discussion. The first 
involves reactions of CO and Hj catalyzed by condensed particles in the solar nebula, 
the so-called Fischer-Tropsch type (FTT) process (Studier et al., 1958). The second 
is a planetary (parent body) process in which atmospheric reduced gases react under 
the influence of one or more possible energy sources, the so-called Millei — Urey (MU) 
synthesis (Wolman et al., 1972). A third possibility, that intact interstellar 
materials were incorporated into meteorites, was also suggested some time ago 
(Cameron, 1973), but has only recently received serious consideration with the 
discovery that some meteoritic carbonaceous materials are isotopically unusual. The 
latter possibility now appears to provide the key to understanding the origin of the 
chondritic amino acids, although parent body processes probably also played a role in 
their formation. 





Acyclic monoamino 
alkanoic acid 

N-Ali(yl monoamino 
atitanoic acid 

/ X 
CH2 CH2 
\ / 


Cyclic monoomino 
oikanoic acid 



Monoamino oilcan- 
dioic acid 

Figure 1 . " General structures of amino acids of CM chondrites. 


straight chain 

-I 1 i_ 

234567 234567 


Figure 2. - Amino acid concentration in the Murchison meteorite related to carbon 








CO •— • trt 





• • 












• • 















O 4-> 








• • 


























































































'-' E 

I c 


C7> O 

































f— » 

<U T3 















=3 -O 


































• • 













































Deuterium in Carbonaceous Chondrites 

Inferences about meteorite synthetic processes based on molecular analyses are 
likely to be valid only insofar as the results accurately reflect the product yield 
of the synthetic process and are unaffected by secondary processing, decomposition, 
and contamination. In contrast, the stable isotopic signature of a particular 
synthetic process may survive substantial alteration of the original product 
mixture. (It can, of course, be seriously affected by contamination.) Consequently, 
increasing emphasis has been placed on determining stable isotope ratios of meteorite 
organics, i.e., ^^C, '*N, and ^H(D). 

The results of selected deuterium analyses are given in Table 1. Deuterium 
content is expressed as 6D.' The first analyses of deuterium in carbonaceous 
chondrites were carried out by Boato (195^). Hydrogen is found in meteorites in 
essentially three forms: hydrated salts, hydrous silicates, and organic compounds. 
The water associated with hydrated seilts can undergo exchange with terrestrial water 
and is therefore of little interest, however, this water can be removed by heating 
the samples in a vacuum at 180°. As seen in table 1, the remaining hydrogen was 
found to be enriched in deuterium relative to terrestrial matter, although this is 
not always the case (Kerridge, 1985). Since two reservoirs of deuterium still 
remained after heating, the question became whether the deuterium enrichment 
was associated with the organic matter, the clay minerals, or both. Smith and Rigby 
(1981) attacked this question directly by carrying out isotopic analyses of the 
insoluble, kerogen--like matter obtained by demineralizing a carbonaceous chondrite 
sample with HF-HCl. They found this organic matter to be significantly enriched in 
deuterium in all five meteorites examined. Later, Robert and Epstein (1982) showed 
by stepwise pyrolysis/combustion that this material was not homogeneous, but 
contained a very D-rich component(s) diluted by less enriched material, and Becker 
and Epstein (1982) found the methanol-soluble compounds also to be quite D-rich. 

Although several processes could have contributed, in principle, to deuterium 
enrichment in the early solar system, it is now generally believed that the deuterium 
enrichments observed in carbonaceous chondrites result from fractionation by ion- 
molecule reactions (Watson, 1976) in the gas phase of the precursive interstellar 
cloud, followed by condensation of D-rich compounds and their incorporation into the 
meteorite parent body(ies) (Geiss and Reeves, 1981), a possibility first suggested by 
Kolodny et al. (1980). Deuterium enrichment may be a hallmark of interstellar 
organosynthesis in low-temperature regimes as a consequence of the low zero-point 
energy of the D-C bond. 

If meteorite amino acids were themselves interstellar molecules, or were derived 
from interstellar precursors, they might be expected to show a significant deuterium 
enrichment. Therefore, we extracted and isolated amino acids from the Murchison 
meteorite and, in collaboration with Epstein, obtained stable isotope ratios (Epstein 
et al., 1987). The results, given in table 1, showed the amino acids to be 
substantially enriched in deuterium. 

(D/H) , - (D/H) ^ . 

6D (o/oo) = [ ^^^, ^^] X 10 = 


Standard mean ocean water (SMOW) is taken as the standard for deuterium 
analyses; (D/H)smow = 0.0001557. The terrestrial 6D range is about -^400 to +100 
(Pillinger, 1984). 







Whole rock, >180° 




HF-HCl residue 




HF-HCl residue, 700" 




MeOH extract 




Amino acids 




a) Becker & Epstein, 1982; b) Boato, 1954; c) Epstein et al., 1987; 
d) Robert & Epstein, 1982; e) Smith & Rigby, 1981. 


Formation Hypothesis 

As before, this deuterium enrichment has been interpreted as indicating 
chemistry under interstellar cloud conditions. But how direct is the relationship to 
interstellar molecules? Are the meteorite amino acids themselves interstellar 
molecules, or are they derived from interstellar precursors? Thus far, amino acids 
have not been identified in interstellar clouds although glycine, the simplest amino 
acid, has been sought (Snyder et al . , 1983). 

It has been suggested that the meteorite amino acids are products of a Strecker- 
cyanohydrin synthesis (Peltzer and Bada, 1978), a reaction sequence shown in figure 
4. These reactions generate both hydroxy acids and amino acids and both classes of 
compounds have been identified in the Murchison meteorite (Peltzer and Bada, 1978). 
Moreover, the ratios of four pairs of analogous hydroxy /ami no acids are consistent 
with their equilibration at a common ammonia concentration as would be expected were 
the Strecker synthesis responsible for their formation (Peltzer et al., 1984). The 
facts that the reactants in this process, aldehydes or ketones, HCN, and ammonia, 
occur in interstellar clouds (Irvine and Hjalraarson, 198^1) and have been shown to be 
D-rich (Penzias, 1980) can account, at least qualitatively, for the D-enrichraent of 
the meteorite amino acids. 

Where did the reactions occur? The coexistence of organic matter and clay 
minerals in carbonaceous chondrites was apparent to the chemists who first examined 
carbonaceous chondr-ites over 150 years ago and the analyses of numerous carbonaceous 
chondrites since then have shown this correlation to be without exception. 
Furthermore, there is substantial evidence that the clay mineralogy is secondary, 
that is, that clays were formed in a parent body by hydrothermal alteration of pre- 
existing anhydrous silicates (DuFresne and Anders, 1962; Bunch and Chang, 1980). 
These indications of liquid water in the meteorite parent body raise the question 
whether the synthesis of amino acids from interstellar precursors by the Strecker 
reaction, an aqueous phase process, occurred concomitantly. If so, might there not 
be a correlation between amino acid content and some petrologic indicator of 
hydrothermal activity? For example, it might be predicted that amino acid content 
would correlate with the meteorite matrix content since the matrix components, 
serpentine minerals plus complex Fe-Ni-S-0-bearing phases (so-called "poorly 
characterized phases" (PCP) now known to be composed of cronstedtite and tochilinite 
(MacKinnon and Zolensky, 1984; Tomeoka and Buseck, 1985)), are believed to have been 
produced by aqueous processes. Determination of the apparent matrix content 
(McSween, 1979) and amino acid content (Cronin and Pizzarello, 1983) for a set of CM 
chondrites have provided data with which this prediction can be tested. A 
correlation with matrix content was not found. However, more recently, McSween 
(1987) has determined the relative proportions of serpentine minerals and PCP in the 
matrices of several CM chondrites. Using these data along with the previously 
determined matrix contents it was possible to test for a correlation between amino 
acid content and PCP content. Interestingly, as shown in figure 5, there appears to 
be a positive correlation. Tomeoka and Buseck (1985) have proposed an alteration 
sequence in which PCP appears at an early stage and then is decomposed by further 
alteration. Thus, these results seem to suggest that amino acids were formed early 
in the aqueous alteration process and then were destroyed as alteration progressed 
further. (Their destruction may have involved a process akin to diagenesis of amino 
acids in terrestrial soils, shales, etc., i.e., incorporation into kerogen-like 
material.) In any case, both the evidence for a Strecker synthesis and the 
correlation of amino acid content with the aqueous alteration process are consistent 
with aqueous processing of interstellar precursors in the meteorite parent body. 


?" .> 



•H2O I II +H2O 



■'^ r. i 




+ NH3^ R-CH-CN — ^-* R-CH-C-NH2 


— » 


I ^ 

Str«ck«r-CyanohydrIn Reaction 

Figure 4. - Reactions of the Strecker-cyanohydrin synthesis of ct-hydroxy acid 
and a-amino acids. 














CrO ® / 





/ 1 t 




0.2 0.3 


0.4 0.5 

Figure 5. - Variation of amino acid content with PCP content of CM chondrites. The 
fractional volume of PCP (fp^p) was obtained from McSween (1979, 1987). 


The origin of the chondritic amino acids can be understood in the context of the 
general scenario outlined in figure 6. In this scheme, ion-molecule reactions in 
cold interstellar clouds generate simple precursors, aldehy-des and ketones, ammonia, 
and HCN (IS Molecules) in the case of the amino acids. These precursors may or may 
not be processed to some degree in interstellar grain mantles, e.g., to form 
branched-chain aldehydes and ketones (IS Grain Molecules). Aggregation of grains and 
the condensation of gas-phase components give volatile-rich planetesimals which 
further agglomerate to form the meteorite parent body(ies) and comets. Warming of 
the parent body, perhaps by decay of short-lived radioisotopes, leads to the 
generation of an internal aqueous phase (DuFresne and Anders, 1953), the conversion 
of anhydrous silicates and metals to PCP, and the conversion of interstellar 
precursors to amino acids (Meteorite Organics). The final stage suggests the 
delivery of amino acids to the primitive earth by meteorite infall, a probable 
occurrence, but one of unknown significance for terrestrial chemical evolution. 


The implications of this scheme for the organic chemistry of a pristine comet 
are straight-forward. Unless it has experienced sufficient heating to produce a 
liquid water phase, a comet nucleus would be expected to be rich in the small, 
volatile, highly reactive molecules, ions, and radicals that characterize the gas 
phase of interstellar clouds and to contain, in addition, a significant contribution 
from interstellar grain mantles where the gas-phase species may have been processed 
further (Greenberg, 1984). The irradiation of icy mantles by cosmic rays is expected 
to generate abundant hydroxyl radicals which would react in part with organic species 
to produce alcohols, polyols, hydroxy acids, etc. Indeed, a laboratory model of such 
a process has been found to produce a suite of oxygenated organics of this type 
(Agarwal et al., 1985). 

If a comet were heated above the melting point of water, as may well be possible 
even in the Oort cloud (Weissman, 1989), numerous aqueous phase reactions would come 
into play. In addition to the rapid reaction of radicals and unstable ions, the 
formation of hydroxy nitriles by the Strecker reaction would be expected, and in the 
neutral pH range and above there would be a significant yield of amino nitriles. The 
hydrolysis of these compounds would produce hydroxy acid amides and amino acid 
amides, respectively, as intermediates in the formation of a-hydroxy acids 
and a-amino acids, the stable, hydrolytic end products found in carbonaceous 
chondrites. (See Fig. 4.) 

It is interesting to inquire whether the conditions of a cometary melt would be 
suitable for amino acid formation. The necessary reactants, HjO, HCN, NH3 and at 
least one aldehyde (CH2O) are well documented comet constituents. However, as 
mentioned above, amino acid formation by the Strecker reaction is pH dependent. 
Miller and Van Trump (1981) have calculated the hydroxy acid/amino acid ratio from a 
Strecker synthesis as a function of pH. They find that the ratio becomes favorable 
for amino acid formation only at pH values -5. Lunine (1988) has critically reviewed 
the Halley molecular abundance data from several types of observations. Based on 
these data, the pH of a melt of Halley composition would be set by the NH3/CO2 
ratio. Based on the abundance ranges of these molecules given by Lunine, this buffer 
would have a pH in the range 5.3-5.4, sufficiently high to provide a significant 
yield of amino acids by a Strecker reaction. The conversion of anhydrous silicate 
grains in the melt to clay minerals, e.g., olivine to cronstedtite, is an acid- 
consuming process and would tend to drive the pH even higher if it were occurring 

























05 >, 
>, ,-1 

£- 0) 








(0 JO 
>. >i 

■— 1 r-H 
1. JO 

CO — ( 





x: Q. 

C C 
— 1 CO 


c 0: 


—1 <D 






> - 




•— < 



»— 1 4-5 

CO —1 
C 5. 


^ 5^ 


e c 










>4 3 






OS 03 







£- CO 



















03 C 






03 -O 

S, 03 






C3 *J 

CO c 








^ 4-1 
CO <!-. 
S- (0 







C 4J £ 







03 — c 4J 







^. ^^ 


03 (0 

< CO Ci3 


^^V ^^^^H 


i^M^^^ MBM 

'^ — — ' 


























In addition to the Strecker reaction, other hydrolytic and facile addition 
reactions among interstellar molecules can be envisioned, for example, the conversion 
of (i) alkyl nitriles to monocarboxylic acids, (ii) cyanate- and ammonia or amines to 
various ureas and other amides, (iii) cyanoethylene and HCN to dinitriles and on to 
dicarboxylic acids, and (iv) cyanoacetylene to pyrimidines. The well-known HCN 
polymerization reactions that generate purines and amino acids, in addition to 
complex polymeric species (Sanchez et al., 1967), could also come into play. 
However, given a formaldehyde/HCN ratio of 10-100 (Lunine, 1988) these reactions 
would appear to be less favored than the Strecker reactions. 

Thus, the nature of the organic analytical problem may depend on the extent to 
which an aqueous phase has conditioned the chemical evolution of comets. Assuming 
some parallel with the carbonaceous chondrites, the nature of silicate grains in 
comets can have important implications, since the secondary formation of hydrous 
silicates is indicative of the development of the typical chondritic organic 
chemistry. On the other hand, a comet in which an initial complement of interstellar 
molecules has remained locked in an icy matrix at Oort cloud temperatures may have 
had little opportunity for chemical evolution, and analyses must be primarily 
designed to detect volatile and reactive molecules, radicals, and ions. Intermediate 
possibilities can also be envisioned. A brief episode of melting would allow radical 
reactions and facile addition reactions to occur, generating what might be called 
transitional species such as hydroxy- and aminonitriles, imines (Schiff's bases), 
etc., without their extensive hydrolysis to stable end-products. The possibility of 
regimes within a comet that have differed significantly in their temperature history 
implies the presence of a range of organic compounds spanning the categories 
described above. The cometary surface, which is subject to the competing effects of 
cosmic radiation (accretion) and particle bombardment (erosion) even in the Oort 
cloud, provides unique conditions for organic reactions that have not been dealt with 


Although developments in the technique and instrumentation of amino acid 
analysis have been driven mainly by the needs of biochemists and molecular 
biologists, the technology has been readily adapted for analyses of carbonaceous 
chondrites. Sensitivity is frequently a limiting parameter when considering the 
analysis of scarce and irreplaceable materials like meteorites, however, the 
development of instrumentation for amino acid analysis has been accompanied by such 
remarkable increases in sensitivity that sample size limitations have not generally 
been a hindrance. In figure 7, a time line is sketched illustrating the advances in 
sensitivity accompanying instrumental evolution in this area over the last 35 years. 
The progression begins with automation (Spackman et al., 1958) of the chromatographic 
amino acid analyzer invented by Moore and Stein (1951) and proceeds through the 
introduction of HPLC (high performance liquid chromatography) technology (Hare, 
1977), the use of fluorescence rather than absorbance for detection (Benson and Hare, 
1975), the formation of derivatives prior to chromatographic separation (Chang 
et al., 1977), the introduction of microcolumn technology (Einarsson et al., 1986), 
and finally, laser-induced fluorescence detection (Roach and Harmony, 1987). A 
parallel progression, beginning about 1980, has exploited capillary electrophoresis 
for the separation of fluorescent amino acid derivatives (Jorgensen and Lukacs, 
1981). This technology has outstripped chromatographic techniques in sensitivity, at 
least temporarily, by the use of laser-induced fluorescence and a device "borrowed" 
from cell biology, the sheath-flow cell, as a detection system (Cheng and Dovichi, 
1988). These developments have brought us to a point at which sensitivity can be as 
conveniently expressed in numbers of molecules as in fractional moles. 


The earliest amino acid analyses of meteorite amino acids were made using 
commercial Moore-Stein analyzers. This early work was done__ without a full 
appreciation of the pitfalls of terrestrial contamination and was ultimately of 
little scientific value (Hayes, 1967), except as a lesson that must still be heeded, 
i.e., that each advance in detection sensitivity is accompanied by a correspondingly 
heightened risk that terrestrial contamination will confuse, distort, or obscure the 
analysis of indigenous amino acids. 

The first valid analyses of chondritic amino acids were made by Kvenvolden 
et al. (1970) using a commercial Moore-Stein analyzer and GC/GC-MS instruments of 
comparable sensitivity. This level of sensitivity required meteorite samples of the 
order of grams for an analysis. For about the last ten years, we have been using an 
HPLC system with post-column fluorescent detection (Cronin and Hare, 1977) which has 
reduced the sample requirement for a single analysis to tens of milligrams. However, 
the recent advances in sensitivity now open the way, not only for further decreases 
in sample size requirements, which are always desirable, but also for new experiments 
that will be designed to answer questions at a petrologic level, i.e., the 
correlation of amino acid content with individual grains and particular mineral 
phases. In this regard, it will be very interesting to approach the question of 
correlation of amino acid content with PCP-phases by direct analysis. It will also 
be possible to analyze amino acids in materials that are available in only minute 
amounts, for example, IDP's, although such work will require the development of 
radically different techniques for sampling and contamination control. The fact that 
terrestrial sources of amino acids add almost exclusively the L-enantiomer has 
provided an effective way to assess contamination in extraterrestrial materials. It 
is important to note that enantiomeric separation has not been neglected in the 
development of ultra-sensitive methods (Gozel et al., 1987). 

The implications of these analytical advances for chondrite and IDP sample size 
requirements are illustrated in Table 2. Here are listed the minimum number of 
molecules detected by Cheng and Dovichi (1988) using capillary electrophoresis with 
lasei — induced fluorescence detection of fluorescein isothiocyanate derivatives. The 
weight of Murchison meteorite required to yield this number of molecules was 
calculated as well as the number of IDP particles. The latter calculation assumed an 
IDP of 20 pm diameter, density of 1 .0 gm cra~^ , and amino acid content equivalent to 
the Murchison meteorite. It is clear that if detection at these levels can be 
achieved in practice, analyses of CM chondrites can be made on samples as small as a 
few picograms if the total sample can be transferred to the analytical system. In 
the case of an IDP, even if only a small fraction of the total extract were analyzed, 
e.g., 1/100, single particle analyses would be feasible. 

Given the development of contamination-free techniques for sampling a comet 
nucleus specimen, transferring, and extracting the sample, it should be possible to 
analyze the amino acids in very small solid and liquid phase samples. As a result, 
the relationship between comet structure and organic composition could be ascertained 
with a high degree of spatial resolution. 

Finally, for reasons discussed in preceding sections, it is by no means certain 
that amino acids will occur in a comet nucleus. Although these bodies are rich in 
the biogenic elements and appear to contain in abundance the reactants for amino acid 
and hydroxy acid synthesis via the Strecker pathways, whether the necessary aqueous 
phase has existed remains a crucial question. Even if this has not been the case, it 
will be important to attempt to verify the potential for such syntheses by carrying 
out experiments in which comet samples are melted, concentrated, and subjected to 
hydrolysis before amino acid analysis. 

390 ^ 

= o 































. I 















OS 51 

i SZ5T < 

- 0Z5I 




























































(SSIOk,' 501) iiwn AllAIiISK5S 







Amino Acid 




(pi cog ram) 







Aspartic acid 





Glutamic acid 






















1 . Analyses of amino acids in CM chondrites over the past 20 years have 
required the development of methods- for quantification of these compounds in scarce 
and irreplaceable samples, at trace levels, as components of very complex mixtures, 
and in an environment where there is enormous potential for contamination. Analyses 
of returned comet nucleus samples pose similar problems and appropriate analytical 
methodology can build on, and be tested in, meteorite studies. 

2. A current view of the origin of chondritic amino acids suggests their 
formation by aqueous processing of interstellar precursors, a scheme that may also be 
relevant to comets. Depending upon the prior existence of an aqueous phase, comets 
can be expected to contain organic molecules ranging from simple, volatile, reactive 
interstellar molecules, to the more complex and stable chondritic compounds. Organic 
analysts should be prepared to deal with this entire range of organic species, as 
well as grain mantle organics and possible intermediates formed in a transient 
aqueous phase. It may also be desirable to carry out analyses after laboratory 
simulation of aqueous processing. 

3. Returned samples should represent the full range of comet structural and 
compositional diversity, i.e., samples taken at various depths and representing 
various material phases. 

4. Maintaining sample integrity is essential, especially avoiding temperature 
increase and the introduction of contaminants. 

5. Recent developments in analytical technology, e.g., for amino acid analysis, 
suggest that much information including spatial distribution will be obtainable from 
analyses of the small samples now made feasible by recent advances in analytical 



Agarwal, U. K. ; Schutte, W. ; Greenberg, J. M. ; Ferris, J. P". ; Briggs, R. ; Connor, S. ; 
Van de Bult, C. P. E, M. ; and Baas, F. : Photochemical Reactions in Interstellar 
Grains - Photolysis of CO, NH3 , and H2O. Origins of Life, vol. 16, 1985, pp. 21- 

Becker, R. H. ; amd Epstein, S. : Carbon, Hydrogen and Nitrogen Isotopes in Solvent- 
Extractable Organic Matter from Carbonaceous Chondrites. Geochim. Cosmochim. Acta, 
vol. -46, 1982, pp. 97-103. 

Benson, J. R. ; and Hare, P. E. : o-Phthalaldehyde: Fluorogenic Detection of Primary 
Amines in the Picomole Range. Comparison with Fluorescamine and Ninhydrin. Proc. 
Nat. Acad. Sci. USA, vol. 72, 1975, pp. 619-622. 

Boato, G. : The Isotopic Composition of Hydrogen and Carbon in the Carbonaceous 
Chondrites. Geochim. Cosmochim. Acta, vol. 6, 195^, pp. 209-220. 

Bunch, T. E. ; and Chang, S. : Carbonaceous Chondrites - II. Carbonaceous Chondrite 
Phyllosilicates and Light Element Geochemistry as Indicators of Parent Body 
Processes and Surface Conditions. Geochim. Cosmochim. Acta, vol. 44, 1980, 
pp. 1543-1577. 

Cameron, A. G. W.: -Interstellar Grains in Museums. In Interstellar Dust and Related 
Topics (J. M. Greenberg and H. C. Van de Hulst, eds.) lAU, 1973, pp. 545-547. 

Chang, J.-Y.; Knecht, R. ; and Braun, D. G. : Amino Acid Analysis in the Picomole Range 
by Precolumn Derivatization and High-Performance Liquid Chromatography. Meth. 
Enzymol., vol. XLVII, 1977, pp. 41-48. 

Cheng, Y.-F.; and Dovichi, N. J.: Subattomole Amino Acid Analysis by Capillary Zone 
Electrophoresis and Laser-Induced Fluorescence, Science, vol. 242, 1988, pp. 562- 


Cronin, J. R. ; and Hare, P. E. : Chromatographic Analysis of Amino Acids and Primary 
Amines with o-Phthalaldehyde Detection. Anal. Biochera., vol. 81, 1977, pp. 151- 

Cronin, J. R. ; and Pizzarello, S. : Amino Acids in Meteorites. Adv. Space Res., 
vol. 3, 1983, pp. 5-18. 

Cronin, J. R. ; and Pizzarello, S. : Amino Acids of the Murchison Meteorite. III. 
Seven Carbon Acyclic Primary a-Amino Alkanoic Acids. Geochim. Cosmochim. Acta, 
vol. 50, 1986, pp. 2419-2427. 

Cronin, John R. ; Pizzarello, Sandra; and Cruikshank, Dale P. : Organic Matter in 

Carbonaceous Chondrites, Planetary Satellites, Asteroids and Comets. In Meteorites 
and the Early Solar System (J. F. Kerridge and M. S. Matthews, eds.), Univ. Arizona 
Press, 1988, pp. 819-857. 

Delserame, A. H. : The Chemistry of Comets. Phil. Trans. Roy. Soc. Lond. A, vol. 325, 
1988, pp. 509-523. 

DuFresne, E. R. ; and Anders, E. : On the Chemical Evolution of the Carbonaceous 
Chondrites. Geochim. Cosmochim. Acta, vol. 26, 1962, pp. 1085-1114. 


Einarsson, S. ; Folestad, S. ; Josef sson, B. ; and Lagerkvist, S. : High-Resolution 
Reversed-Phase Liquid Chromatography System for the Analysis of Complex Solutions 
of Primary and Secondary Amino Acids. Anal. Chem., vol. 58, 1986, pp. 1638-15^^3. 

Epstein, S. ; Krishnamurthy, R. V. ; Cronin, J. R. ; Pizzarello, S. ; and Yuen, G. U. : 
Unusual Stable Isotope Ratios in Amino Acid and Carboxylic Acid Extracts from the 
Murchison Meteorite. Nature, vol. 326, 1987, pp. 477-479. 

Geiss, J.; and Reeves, H. : Deuterium in the Solar System. Astron. Astrophys., 
vol. 93, 1981, pp. 189-199. 

Gozel, P.; Gassmann, E. ; Michelsen, H. ; and Zare, R. N. : Electrokinetic Resolution of 
Amino Acid Enantionmers with Copper (II)-Aspartarae Support Electrolyte. Anal. 
Chem., vol. 59, 1987, pp. 44-49. 

Greenberg, J. M. : Chemical Evolution in Space. Origins of Life, vol. 14, 1984, 
pp. 25-36. 

Hare, P. E. : Subnanomole-Range Amino Acid Analysis. Meth. Enzymol., vol. XLVII, 
1977, pp. 3-18. 

Hayes, J. M. : Organic Constituents of Meteorites - A Review. Geochim. Cosmochim. 
Acta, vol. 31, 1967, pp. 1395-1440. 

Irvine, W. M. ; and Hjalraarson, A. : The Chemical Composition of Interstellar Molecular 
Clouds. Origins of Life, vol. 14, 1984, pp. 15-23. 

Jorgenson, J. W. ; and Lukacs, K. D. : Zone Electrophoresis in Open-Tubular Glass 
Capillaries. Anal. Chem., vol. 53, 1981, pp. 1298-1302. 

Kerridge, J. F. : Carbon, Hydrogen and Nitrogen in Carbonaceous Chondrites: Abundances 
and Isotopic Compositions in Bulk Samples. Geochim. Cosmochim. Acta, vol. 49, 
1985, pp. 1707-1714. 

Kissel, J.; and Krueger, F. R. : The Organic Component in Dust from Comet Halley as 
Measured by the PUMA Mass Spectrometer on Board Vega 1. Nature, vol. 326, 1987, 
pp. 755-760. 

Kolodny, Y.; Kerridge, J. F. ; and Kaplan, I. R. : Deuterium in Carbonaceous 
Chondrites. Earth Planet. Sci. Lett., vol. 46, 1980, pp. 149-158. 

Kvenvolden, K. ; Lawless, J.; Pering, K. ; Peterson, E. ; Flores, J.; Ponnamperuma, C. ; 
Kaplan, I. R. ; and Moore, C. : Evidence for Extraterrestrial Amino Acids and 
Hydrocarbons in the Murchison Meteorite. Nature, vol. 288, 1970, pp. 923-926. 

Lawless, J. G. ; and Yuen, G. U. : Quantification of Monocarboxylic Acids in Murchison 
Carbonaceous Meteorite. Nature, vol. 282, 1979, pp. 396-398. 

Lunine, J. I.: Abundances of Molecular Species in Halley's Comet— Their Role in 
Understanding the Chemistry of Cometary Formation Environments. In The Formation 
and Evolution of Planetary Systems (H. A. Weaver, F. Paresce, and L. Danly, eds.) 
Cambridge Univ. Press, 1988, pp. xxx. 


MacKinnon, I. D. R. ; and Zolensky, M, E. : Proposed Structures for Poorly- 
Characterized Phases in C2M Carbonaceous Chondrite Meteorites. Nature, vol. 309, 
198^, pp. 2^40-2^2. 

McSween, H. Y . : Alteration in CM Carbonaceous Chondrites Inferred from Modal and 
Chemical Variations in Matrix. Geochim. Cosmochim. Acta, vol. -43, 1979, pp. 1761- 

McSween, H. Y.: Aqueous Alteration in Carbonaceous Chondrites: Mass Balance 
Constraints on Matrix Mineralogy. Geochim. Cosmochim. Acta, vol. 51, 1987, 
pp. 2469-2iJ77. 

Miller, S. L. ; and Van Trump, J. E. : The Strecker Synthesis in the Primitive Ocean. 
In Origin of Life (Y. Wolman, ed.) D. Reidel Publ., 1981, pp. 135-141. 

Moore, S. ; and Stein, W. H. : Chromatography of Amino Acids on Sulfonated Polystyrene 
Resins. J. Biol. Chem., vol. 192, 1951, pp. 663-68I . 

Peltzer, E. T. ; and Bada, J. L. : a-Hydroxycarboxylic Acids in the Murchison 
Meteorite. Nature, vol. 272, 1978, pp. 443-^444. 

Peltzer, E T. ; Bada, J. L. ; Schlesinger, G. ; and Miller, S. L. : The Chemical 

Conditions on the Parent Body of the Murchison Meteorite: Some Conclusions Based on 
Amino, Hydroxy and Dicarboxylic Acids. Adv. Space Res., vol. 4, 1984, pp. 69-7^. 

Penzias, A. A.: Nuclear Processing and Isotopes in the Galaxy. Science, vol. 208, 
1980, pp. 663-669. 

Pillinger, C. T. , Light Element Stable Isotopes in Meteor ites~from Grams to 
Picograms. Geochim. Cosmochim. Acta, vol. 48, 1984, pp. 2739-2766. 

Roach, M. C; and Harmony, M. D. : Determination of Amino Acids at Subfemtomole Levels 
by High-Perf ormance Liquid Chromatography with Lasei — Induced Fluorescence 
Detection. Anal. Chem., vol. 59, 1987, pp. 411-415. 

Robert, F. ; and Epstein, S. : The Concentration and Isotopic Composition of Hydrogen, 
Carbon, and Nitrogen in Carbonaceous Meteorites. Geochim. Cosmochim. Acta, vol. 
46, 1982, pp. 81-95. 

Sanchez, R. A.; Ferris, J. P.; and Orgel, L. E. : Studies in Prebiotic Synthesis. II. 
Synthesis of Purine Precursors and Amino Acids from Aqueous Hydrogen Cyanide. 
J. Mol. Biol., vol. 30, 1967, pp. 223-253. 

Smith, J. W. ; and Rigby, D. : Comments on D/H Ratios in Chondritic Organic Matter. 
Earth Planet. Sci. Lett., vol. 5>i, 1981, pp. 64-66. 

Snyder, L. E. ; Hollis, J. M. ; Suenram, R. D. ; Lovas, F. J.; Brown, L. W. ; and Buhl, 
D. : An Extensive Galactic Search for Conformer II Glycine. Astrophys. J., 
vol. 268, 1983, pp. 123-128. 

Spackman, D. H. ; Stein, W. H. ; and Moore, S. : Automatic Recording Apparatus for Use 
in the Chromatography of Amino Acids. Anal. Chem., vol. 30, 1958, pp. 1190-1206. 

Studier, M. H. ; Hayatsu, R. ; and Anders, E. : Origin of Organic Matter in Early Solar 
System - I. Hydrocarbons. Geochim. Cosmochim. Acta, vol. 32, 1968, pp. 151-173- 


Toraeoka, K. ; and Buseck, P. R. : Indicators of Aqueous Alteration in CM Carbonaceous 
Chondrites: Microtextures of a Layered Mineral Containing" Fe, S, 0, and Ni. 
Geochira, Cosmochira. Acta, vol. •49, 1985, pp. 2149-2163. 

Watson, W. D. : Interstellar Molecule Reactions. Rev. Mod. Phys., vol. 48, 1976, 
pp. 513-552. 

Weissman, P. R. ; and Stern, S. A.: Physical Processing of Cometary Nuclei. This 
volume, LPI, 1989, pp. xxx-xxx. 

Wolman, Y.; Haverland, W. J.; and Miller, S. L. : Nonprotein Amino Acids from Spark 
Discharges and Their Comparison with the Murchison Meteorite Amino Acids. Proc. 
Nat. Acad. Sci. U.S.A., vol. 69, 1972, pp. 809-811. 




D. Stoffler 

H. Diiren 

J. Knolker 

Institut fiir Planetologie 

Westfaiische Wilhelms-Universitat Miinster 

Miinster, FRG 


Concepts for the curation, primary examination, and petrographic analysis of comet 
nucleus samples returned to Earth* 

D. StOffler, H. DUren, and J. KnOlker 

Institut fUr Planetologie, Westfalische Wilhelms-Universitat MUnster, MUnster, F.R. 

of Germany 


One of the fundamental requirements of the planned ESA/NASA Comet Nucleus Sample 
Return Mission (ROSETTA) is to develop concepts and instrumentations for the 
curation and analysis of the returned samples in an adequately equipped Receiving 
Laboratory. This laboratory must meet the condition that the handling, inspection, 
and analysis of the samples be performed at the temperature and pressure of the 
parental comet nucleus environment. This restriction must hold for the complete 
period of what may be called 'Primary Sample Examination and Analysis" (PSEA). Only 
after this period subsamples may be released to less restrictive environments. 

As foreseen in the "Sampling and sample storage requirements" of the ROSETTA 
mission definition report (ESA Publ. Div. , 1987), different types of samples will be 
returned in sealed containers at a temperature of less than 160 K, preferably at 
130 K. The total sample mass is expected to be near 15 kg. This sampling plan which 
includes a 1-3 m long core sample provides a first set of boundary conditions of a 
Receiving Laboratory. A second set of limiting conditions has to be derived from a 
'best estimate model' of the mineralogical composition and texture of the returned 
samples. It is the aim of this article to help defining (a) the objectives of a 
Primary Sample Examination and Analysis, (b) the properties of best estimate 
cometary model samples, and (c) the requirements and instrumentation for PSEA with 
special emphasis on microscopic bulk analysis of the samples. 


According to the ROSETTA mission goals (ESA Publ. Div., 1987) three types of 
cometary samples will be returned to Earth: 

(1) A core sample of 1 to 3 m length and about 10 cm diameter; it will be stored 
most probably in segments of 0.5 m, 

(2) a small sample of volatile material with a mass of up to 100 g, and 

(3) a surface sample enriched in refactory materials with a mass of up to 5 kg. 

Although the actual design of the sample containers is not yet known, it is clear 
that they will be under high vacuum and at temperatures as low as 130 K. In 
principle, the handling of the containers arriving at the Receiving Laboratory will 
require a facility which can transport and open the containers by remote operation 
without change of the P-T-conditions. Furthermore it must be capable of handling and 
subdividing samples and to transfer whole samples or aliquots to special inspection 
devices and analytical instruments by which all types of investigations considered 
to be necessary for the PSEA can be adequately performed with a minimum degradation 
of the samples. 

The basic idea for the remotely operated primary examination and analysis of 
the samples (PSEA) is to proceed in two subsequent modes in a system of evacuated 
cryogenic cabinets and transfer locks: 

(1) Non-destructive inspection and analysis by special tomographic methods 

(Albee, this volume) and non-destructive bulk chemical and physical analyses 

* Contribution No. 12 of the Institut fUr Planetologie, MUnster, work in part 
supported by the Deutsche Forschungsgemeinschaf t (Sto 101/23-1) 


by photometric and spectrocopic methods 
(2) Inspection and analysis of the bulk samples by destructive methods such as 
cutting of samples and thin sectioning of specially prepared aliquots or 
mechanical separation of particular constituents 

Basically PSEA should provide information on the bulk mineralogical and 
chemical composition, and on textural and fabric characteristics in the sense of an 
overall petrographic and chemical sample characterization. The aim of the PSEA is to 
provide a thorough assessment of the sample properties in as much detail as will be 
necessary for the definition of a post-PSEA sample analysis and allocation plan. In 
particular, PSEA should be capable of identifying heterogeneities of the bulk 
samples, textural subunits, and certain physical properties such as porosity and 
material strength in order to optimize the decision on type, mode, and sequence of 
further analyses. 

This concept for PSEA undoubtedly requires the development of new laboratory 
techniques and robotic operations adapted to these techniques. Such instrumentation 
must be tested on artificially produced comet nucleus model samples which have to be 
defined on the basis of our present knowledge about comets, primitive solar system 
material, and interstellar matter. 


Information which is relevant for a model of the composition and texture of 
comet nucleus material can be obtained from very different sources: (1) Direct 
observations of comets, meteors, and of interstellar and circimstellar dust 
(Wilkening, 1982; Grewing et al., 1988; ESA Public. Division, 1986; JeBberger et 
al., 1988; Mathis, this volime), (2) laboratory analysis of primitive solar system 
material such as interplanetary dust particles (IDP's) and carbonaceous chondrites 
(Brownlee, 1985; Sandford and Walker, 1985; Mackinnon and Rietmeijer, 1987; Bradley, 
1988; Kerridge and Matthews, 1988; Wasson, 1985; McSween, 1987), (3) theory and 
laboratory simulation of grain formation in solar and stellar nebulae and 
interstellar clouds (Black and Matthews, 1985; Nuth and Stencel, 1986; Morfill. 
1985; Wood and Morfill, 1988; Cassen and Boss, 1988), and (4) accretion and 
evolution models of comets (Whipple, 1950; Wilkening, 1982; Fanale and Salvail, 
1984; Klinger et al., 1985; Houpis et al., 1985; Weissmann, 1986, ESA Publ. Div. , 
1986; Stem, 1988). 

How can we make use of this extreme diversity of information in defining model 
samples representative of- the upper 3 to 5 m layer ("regolith") of a comet nucleus? 
Clearly, it is necessary to derive a 'regolith model" which takes care of the 
sample properties on the macro- and microscale. The interpretation of the presently 
available data leads to the basic conclusion that the regolith is heterogeneous in 
vertical and horizontal sections on a macroscopic as well as microscopic scale. This 
model actually reflects the generally accepted idea of a stochastic nature of the 
processes that formed and reprocessed cometary nuclei. Consequently, the 
instrumentation for sample analysis has to be capable of dealing with the following 
potential macro- and microproperties of the regolith (Fig. 1): 

(1) Fractal structure at all scale lengths 

(2) Layered structure subparallel to the surface 

(3) Existence of textural and compositional subunits at all scale lengths 

This may be considered as a 'worst case heterogeneity scenario" which we take as 
a basis for defining the range of variation of comet regolith model samples. If we 
consider such samples as petrographic objects it is obvious that we have to develop 
best estimates for three classes of bulk properties given the size and geometry of 
samples to be returned to Earth (see section above) : 

(1) The modal composition, i.e. the nature and volumetric abundance of the 

(2) The textural properties, i.e. the size, morphology, and intergrowth 
characteristics of the constituents and cavities (pore space) 

(3) The fabric characteristics, i.e. the spatial distribution and orientation of 





































































































































End iMtoer constituents 

wonoonase gra ins 

(1) single crystals 
C2) disordered crystals 
(3) anorphous grains 

CfitSCfiQl aggregates 

(1) mantled nonociOase grains 

(2) oolycrystalllne grains 
(with crystals and 
aRDrohous grains) 

Porous aooreoates 

(1) aggregates of renoohase 

or Bultlohase grains 

(2) aggregates of corcoslte 
■ono- and aultlphase grains 

Ices and 
(H, C. H. 
0, S) 


NHj etc. 

CO2- CH,, 

and clatn- 

rates containing 
these and other 

(Si, Bg, Fe. 
Ca, Al, 0. 
S. C) 

Anhydrous silicates 
(Olivine, pyroxene, 
feldspar and other 
Ca-ng-Fe-sil Rates) 
Hydrated silicates 
(serpentine, smectite, 
kaollnite. talc, 
trennllte etc.) 
Oxides (Spinel, 
oerovsklte. magnetite 

Carbides (SiC, 
Sulfides (FeS, 
pentlandlte. etc.) 
Sulfates, carbonates, 
phosphates, and hydro- 
xides of ng. Ca. Fe 
Betals (Fe-m. noble 
netals, Tl etc.) 
Cartxxi (graphite, 
anorphous cart>on 

(C, H. 0. K) 

simple hydrocarbons 
oolycycllc hetero- 
aronatics etc. 

Ices and 





more than 

one type of 


minerals (or 


or amorphous 


coatings on 
other mono- 
phase grains 
matrix 'cement' 
bodies (?) 


(a) grains 
Ices and 

(b) grains 
and hydro- 





grains with 
all types of 
volatile and 




of all 
types of 




inorganic minerals I > 

hydrocarbons (K) 



the constituents and the pore space. 

These bulk porperties determine the chemical composition and the physical 
porperties which are therefore no variables to be defined independently. However, 
all properties depend now on the definition of the term "constituent" . In a 
petrographic sense, constituents of a cometary sample can be homogen"eous grains of 
variable composition, heterogeneous grains (e.g. mantled crystals or coherent 
multiphase particles), and aggregates of homogeneous and/or heterogeneous grains all 
of which can be considered as textural subunits formed as such by one particular 

Erocess. At a smaller scale, we expect three kinds of solid phases as building 
locks of the "constituents" : ices (including clathrates), inorganic minerals, and 
hydrocarbon compounds. Therefore, constituents of comet samples containing these 
phases are classified as follows (Table 1): 

(1) Single monophase grains (crystalline to amorphous) 

(2) Coherent aggregates of one type or different types of monophase grains 

(3) Porous aggregates of either type (1) or type (2) particles or composites of 

The grain size of these constituents in the comet regolith may range from the 
nanometer to submillimeter level for type (1) constituents and from the submicron to 
the meter scale for type (2) and (3) constituents (compare Fig. 1). With these 
formalities in mind, a comet model samples can be seen as a comet nucleus subunit or 
regolith subunit with a particular volume proportion of "end member" constituents or 
components as they have been defined above and in Table 1. Therefore comet nucleus 
samples may differ in the following important parameters: 

(1) The relative volume proportion of the end member constituents which includes 
the ratio of refactory to volatile material and of organic to inorganic 

(2) the grain size distribution of the constituents, and 

(3) the structure and porosity of the bulk sample. 



Parameter Dimension Range 

Fraction of refractory dust 2 - 100 

Mass ratio of refractory 

inorganic minerals to 

refractory carbonaceous 

compounds (includ. C) - 1-4 

Porosity of refractory 

surface material 2 0-95 

Porosity below dust mantle 2 10 - 80 

Density of dust mantle g/cm^ 0.005 - 2.2 

Density below dust mantle g/cm^ 0.1-1.5 

Grain size of monophase 

grains ym 0.001 - 100 

Grain size of multiphase 

grains and composite 

aggregates ^m-m lym-lm 

Compressive strength of 

bulk samples MPa 10 - 10 

-■♦ . - 2 


Estimated ranges for these parameters of cometary model samples as established 
by the ROSETTA Science Definition Team are given in Table 2. The data and 
informations discussed so far (Fig. 1, Tables 1 and 2) translate into a comet 
regolith model which is visualized in Figs.l and 2. The regolith can be seen 
petrographically as a polymict geological body defined as a mixture of 4 ma^or 
textural components (Fig. 2). These components have a certain matrix-inclusion- 
relationship typical of polymict rocks vith finer grained components forming the 
matrix and coarser grained components representing inclusions (Fig. 2). Any actual 
sample returned by the ROSETTA mission may represent only part of this multi- 
component system, e.g. a surface sample consisting only of refactory dust or a core 
sample segment composed solely of a rock-like aggregate of refactory material or 
alternatively of pure, non-porous polycrystalline ice. This must be kept in mind for 
the following discussion on required instrumentation of a comet sample receiving 


The basic technical requirements for an adequate handling and primary 
examination and analysis of returned comet nucleus samples (PSEA) depend on a 
variety of boundary conditions. They are dictated by the general scientific goals of 
the sample analysis (ESA Publ. Div., 1987) and by the mass, geometry, and nature of 
the samples themselves. Since most of these conditions cannot be exactly predicted, 
the instrumentation for an effective PSEA will be highly complex and demanding. The 
analytical capabilities of a Receiving Laboratory must meet any of the extreme 
conditions given by the wide range of possible sample properties discussed in the 
last section. Compared to the case of the preliminary examination of lunar samples 
(McKay, this volume) a much higher degree of quality and quantity of investigations 
must be achieved during the PSEA of cometary material because of the complexity and 
unusual P-T-state of the study objects. 

A tentative and simplified concept of the mode and sequence of investigations 
during the PSEA of returned nucleus samples is sketched in the flow diagram of Fig. 
3. It is based on a system of cryogenic cabinets and transfer locks providing the 
necessary environmental conditions with respect to temperature and pressure. Until 
the sample characterization has reached a stage where subsamples may be released 
into less restrictive environments, it is assumed that all studies are performed 
under cometary conditions, i.e. at 130-160 K and ultrahigh vacuum. The PSEA plan of 
Fig. 3 is subdivided into several steps or sections: 

(1) Detailed assessment of all remote sensing data related to the sample 
acquisition on the comet nucleus; this includes the reduction and 
interpretation of data obtained from the photo documentation of the sampling 
site and of the sampling procedures, from the temperature monitoring during 
sampling, and from the borehole logging if available 

(2) Inspection and analysis of the samples inside the closed containers by non- 
destructive methods, in particular by modem three-dimensional X-ray 
computer assisted tomography (CAT scans; Albee, this volvmie) 

(3) Opening of the sample containers and further inspection of the samples by 
imaging and non-destructive microscopic, bulk chemical and physical analyses 
including reflection spectrophotometry and spectrometric analyses (e.g. X- 
ray fluorescence, gamma-ray and alpha-backscatter spectroscopy; e.g. Englert 
et al. . 1987) 

(4) Dissection of samples and recovery of suitable aliquots for further detailed 
petrographic and microchemical analysis based on destructive methods of 
sample preparation 

(5) Preparation and investigation of sample aliquots in more specialized 
sections of the cryostat system; these comprise a preparation unit for 
polished section and thin section production with attached facilities for 
optical and electron microscopy and micro-chemical analysis; a mineral (or 
phase) separation unit; and a long-term storage vinit 




Figure 2.- Schematic representation of the mixing of 'end member constituents" 
in a cometary model regolith 

Cryoscat Bencfi 

for ntnerai 

Cryostat for 



Hlneral scoarduon 
Beaote insoecnon 

Cryosiai Bencn for 
Preliminary Examination 
ana Petrograonic Analysis 

lOBte tnsoection 

Ov nonoestructlve 


Cutting Preoaracton 

Preoaratlon o' relisnea 

for SKtlonirej '"'" s«:'lon 

(F nation or 


SaBle SiMKlslcn 

/licroscooic and siereOBttric 
tecture analysis; onoto OKumtauon 

Electron ■icroscoov 

RlcrocJVBical analysis 

Cryostat Bencti 
for 'Bean- Analysis 

Figure 3.- Schematic plan for sample examination and analysis in a Comet Nucleus 
Samnle Receiving Laboratory 


(6) Transfer of samples to highly specialized instruments for further bulk or 
grain by grain analysis. Such instrimients could be in part attached to the 
Receiving Laboratory or located in outside laboratories of principle 
investigators. Part of the sample suite will probably be released to non- 
cometary environmental conditions at this point . 

It is obvious that most of the instrumentation needed for the PSEA. is not yet 
available and must be developed in the next decade or existing analytical tools must 
be adapted to the extreme P-T-conditions of the cometary environment. For the 
petrographic (mineralogy, texture) and chemical bulk characterization of the 
samples the X-ray tomography and the non-destructive spectroscopy using sensor 
heads are the most important methods of non-destructive PSEA. They will not be 
discussed here in any detail and the reader is referred to other sections of this 
volume (Albee, this volume; Lindstrom and Lindstrom, this volume; Sutton, this 
volume) and to special literature (Englert et al., 1987). These non-destructive 
investigations however, are insufficient for a quantitative assessment of the 
texture and mineralogy of the bulk samples. A stereometric analysis in two- 
dimensional sections is required and will be discussed in the remaining part of this 

Any combined stereometric and modal analysis of a sample must be performed on 
well-defined plane sections which are either polished sections of "thick samples" or 
polished thin sections. The analytical instrvunents will be either the polarizing 
microscope or any type of electron microscope with attached image analysis. 
Currently used techniques for snow and ice microscopy are the most appropriate 
starting point for further developments (Bader et al., 1939; De Quervain, 1954; 
Kinosita and Wakahama, 1960; Narita, 1969, 1971; Kry, 1975; Gubler, 1978; Good, 
1980, 1982, 1987; Perla and Ommanney, 1985; Perla, 1985; Perla and Dozier, 1985; 
Dozier et al., 1987). 

In snow and avalanche research the preparation of polished sections and thin 
sections is made in the following steps (Fig. 4; e.g. Good, 1987): 

(1) Impregnation of the snow sample by an organic liquid (e.g. diethylphthalate, 
aniline) at moderately low termperature, e.g. at about -4°C in the case of 

(2) Freezing of the impregnated sample at -20 to -30 t in order to crystallize 
the pore-filling liquid 

(3) Cutting of the sample and preparation of polished sections by means of a 
sledge microtome at about -13 t 

(4) Staining of the impregnating medium for better discrimation of ice crystals 
and pore space in. microphotographs 

(5) Digital image analysis of microphotographs 

The stereological analysis of the two-dimensional sections can be done by 
conventional methods of stereology and pattern recognition (Good, 1980, 1987; Dozier 
et al., 1987). It yields the volumetric fractions of the constituents (ice and pore 
space in the case of snow) as well as the textural parameters characterizing grain 
size and intergrowth properties. In recent years it was recognized that a more 
sophisticated three-dimensional analysis of the snow microstructure is required for 
a better understanding of the correlation of stereological parameters with the 
mechanical, thermal, and electromagnetic properties of snow. Therefore, the method 
of "serial cuts" was introduced (DeHoff, 1983; Perla et al., 1986; Good, 1987) ^rtiich 
allowed the reconstruction of the three-dimensional snow microstructure (Fig. 5). 
The serial polished sections are typically taken with spacings of 50 to 100 pm (Fig. 

The application of the described techniques for the textural analysis of snow to 
the much more complex cometary samples requires several important changes of the 
method. The impregnation technique must be adapted to much lower temperatures that 
means liquids with freezing points near 160 K must be used or a completely different 
method has to be developed. A different sectioning techniques is necessary in order 
to cope with the expected wide range of hardness of the constituent phases in comet 
samples (Mohs hardness ranging probably from ~1.5 to ~7 or 8). Diamond tools will 
be needed for cutting and polishing. The whole preparation procedure must be 
performed at very low temperature and by remote robotic manipulation. 


norous ,■ -.K solidifying top surface 

porous impregnating with ' ^ j 

"^^''^"^' organic liquid ^^ polished thin serial polished 

freezing section 





(fa, 4 

"yd-* Y- 




' ■' ^g' ca •IS'C I J i„^ 


' ' 

^ \ " 

J ■» -■ 

" '* «■' 

/ 7 

electron microscopy transmitted light microscopy 
and imaging and imaging 
/ L 



computerized stereological 
image analysis 

taken m 

Figure 4. 

PreparaciOQ of porovu ice-dust svspies for polished thin sectioning 
asd polished serial sectioning, subsequent aicroscop;. and quantitative 
textural analysis: teaperatures are those used for terresttlal snov (see 
text for references) 

Figure S.- Micrograph of snov taken in reflected light (ice crystals are black, por* 
space is vhlte) and reconstruction of the thzee-dlaeaslonal shape of the 
center particle by stereological analysis of serial sections of the 
area as shovn in the top image: reproduced from Good (1987) 


We have made first successful tests to prepare polished sections of synthetic 
cometary samples (Grlln et al . , this volume) which consisted of 90Z HjO-snow, 9Z 
olivine, IZ montmorillonite, and 0.082 carbon and had a bulk density of about 0.5 
g/cm^. The sample was prepared according to steps (1) and (2) described above. For 
the polishing procedure a sledge microtome (Polycut E of Relchert and Jung, 
Cambridge Instruments Germany) equipped with a rotating double-diamond milling head 
was used while the sample temperature was varied between 100 K and 260 K. Additional 
samples which consisted of solid H2 0-ice and mineral powder (olivine, 
montmorillonite, carbon) were sectioned and polished by the same method. In some 
cases a sledge micotome with a tungsten carbide knife was used. The preparation work 
and the study of thin sections by a polarizing microscope was carried out in a low- 
temperature laboratory at about -15 "C. Selected microphotographs of the samples are 
shown in Figs. 6 to 10. 

Our preliminary findings can be summarized as follows. A method for producing 
polished thin sections of porous ice- silicate mixtures containing olivine can be 
developed by modifying commercially available microtomes which use rotating diamond 
tools (Grasenick and Warbichler, 1979). Sledge microtomes with metal knifes as used 
in snow research can only be used for samples with dust rich in sheet silicates (low 
Mohs hardness). Future developments must concentrate on new impregnation techniques 
for porous samples which can be applied at temperatures below 200 K, and on new 
techniques for better discrimination of ice, cemented pore space, and various types 
of mineral phases in optical microscopy. Also the problem of transferring polished 
thin sections at temperatures below 200 K from the preparation unit (e.g. microtome) 
to the optical or electron microscope needs attention. The microscopic 
investigations can be carried out with the aid of commercially available cryogenic 
tables where appropriate P-T-conditions can be achieved. Techniques applicable to 
optical microscopes have been reviewed by Roedder (1984). 


Following the science goals and mission concepts for a Comet Nucleus Sample 
Return Mission as expressed by the joint ESA/NASA Mission Definition Team (ESA 
Publ.Div., 1987) we have reviewed the potential requirements for the curation and 
investigation of returned samples in a Receiving Laboratory. We arrive at the 
following main conclusions: 

(1) The instrumentation of a Receiving Laboratory must be capable of handling, 
characterizing, subdividing, and transferring returned samples to special 
inspection devices an'd analytical instruments. All operations must be performed 
at cometary P-T-conditions by remotely operated robotic manipulators in a 
sustained system of cryogenic cabinets and transfer locks. 

(2) A thorough and quantitative "Primary Sample Examination and Analysis" (PSEA) is 
required before any subsample will be released to non-cometary environments and 
subsequent specialized analyses. This PSEA will be more complex and demanding 
that the corresponding Preliminary Examination of lunar samples or of any other 
extraterrestrial samples. 

(3) The PSEA must provide sufficient information on bulk sample properties 
(mineralogy, chemistry, texture) in a way that further destructive analyses of 
the samples can be carefully planned without loss of any global information. 

(4) It is proposed to perform the PSEA in two subsequent modes with a first phase of 
non-destructive investigations by "tomographic" methods and remote bulk chemical 
and physical analyses and a second phase of optical and electron microscopy 
requiring destructive modes such as preparation polished sections and thin 

(5) The methodological and instrumental developments required for achieving the PSEA 
call for a test phase during which synthetic comet samples must be analyzed. 
Cometary model samples have to be defined on the basis of the current knowledge 
of comets and other relevant solar system, stellar, and interstellar matter. 


♦ # 

I I 

Figure 6.- Polycrystalline ice produced from a water-carbon suspension (soot of ca. 
25 nm grain size) with carbon aggregates occupying grain boundaries of 
ice crystals; polished section in reflected light; horizontal width of 
image : 0.7 ^w"' 




Figure 7.- Polycrystalline ice with interstitial olivine-powder (average grain-size: 
4 um) : polished section in reflected light; horizontal width of image: 
2.3 am ^^^ 

Figure 8.- Same as Figure 7 but with monunorillonite (average grain siie:-8uni) 
instead of olivine; horizontal, width of image: 0.7 am 

Figure 9.- Micrograph of a polished section of a synthetic cometary sample 
with 90 Z H20-snow. 92 olivine Cav. grain size: 4 u"") . 1 I 
montmorillonite (av. grain size: 8 um) and 0.08 I carbon (soot) by 
weight: reflected light: center: spherical dust-ice aggregate: upper 
right comer: polycrystalline ice with interstitial dust; white matrix 
vith diagonal lines; pore space filled with crystallized diethylphthalate; 
horizontal width of image: 1.4 imn 












i S 

jr -w — 


(6) Comet nucleus model samples have been defined on the basis of a four component 
cometary reeolith model in -which "snow", "polycrystalline ice", "dust grains" 
and "rock-like aggregates" are mixed in variable proportions. The constituent 
phases of these "end member" components are (1) ices and clathrates, (2) 
inorganic minerals, and (3) hydrocarbons. These phases may occur in 3 major 
textural subunits: single monophase grains, coherent aggregates of phases, and 
porous aggregates of mono- and multiphase grains. 

(7) For the petrographic and microchemical characterization of cometary samples 
during the PSEA we propose to improve and modify techniques which are currently 
used in snow research for the fabrication of thin sections and serial sections 
and for the subsequent stereological analysis of microscopic images. Ongoing 
studies on olivine-smectite-carbon-snow mixtures in our laboratory show that 
such techniques can be developed. 


Albee, A.L.: Analytical study of comet nucleus samples, this volume, pp. 1-2, 


Bader, H.; Haefele, R. ; Bucher, E. ; Neher, J.; Eckel, 0.; and Thams , C: Der 

Schnee und seine Metamorphose. Beitrfige zur Geologie der Schweiz, Geotechnische 

Serie, Hydrologie, 1939 

Black, D.C.; and Matthews, M.S. (eds.): Protostars & Planets II. The University 

of Arizona Press, Tucson, 1985 

Bradley, J. P.: Analysis of chondritic interplanetary dust thin-sections. 

Geochim. Cosmochim. Acta 52, pp. 889-900, 1988 

Brownlee, D.E.: Cosmic dust: Collection and research. Ann. Rev. Earth Planet. 

Sci. 13, pp. 1A7-173, 1985 

Cassen, P.; and Boss, A. P.: Protostellar collapse, dust grains and solar system 

formation, in J.F. Kerridge; and M.S. Mathews (eds.): Meteorites and the early 

solar system. The University of Arizona Press, Tucson, pp. 30A-328, 1988 

DeHoff, R.T.: Quantitative serial sectioning analysis: preview. J. Microsc. 131, 

pp. 259-563, 1983 

De Quervain, M.R. : Snow as a crystalline aggregate. CRREL Trans. 21, pp. 1-7, 


Dozier, J.; Davis, R.E.; Perla, R.: On the objective analysis of snow 

microstructure. Avalanche Formation, Movement and Effects lAHS Publ. 162, pp. 

A9-59, 1987 

Englert, P.; BrUckner, J.; and Wfinke, H.: Planetary gammaray spectroscopy, a 

special form of prompt charged particle and prompt neutron activation analysis. 

J. Radioanal. and Nucl. Chem. 112, pp. 11-22, 1987 

ESA Publication Division: ROSETTA, The Comet Nucleus Sample Return Mission, 

Report of the Science Definition Team, Space Science Department of ESA, ESTEC, 

Noordwijk, The Netherland, 1987 

ESA Publication Division: The Comet Nucleus Sample Return Mission Proc. 

Workshop, Canterbury, UK, 15-17 July 1986, ESA SP-249, 1986 

ESA Publication Division: Symposium on the Diversity and Similarity of Comets, 

Brussels, Belgium, ESA SP-278, 1987 

Fanale, P.P.; and Salvail, J.R.: An idealized short-period comet model: surface 

insolation, H20 flux, dust flux, and mantle evolution. Icarus 60, pp. 476-511, 


Good, W. : Structural investigation of snow, a comparison of different parameter 

sets, in E.S. Gelsema; L.N. Kanal: Pattern recognition in practice. North 

Holland Publ. Comp. , 1980 


Good, W. : Structural investigations of snow and ice on core III from the 

drilling on Vernagtfemer , Austria, in 1979. Z. Gletscherk. und Glacialgeol. 18, 

pp. 53-64, 1982 

Good, W. : Thin sections, serial cuts and 3-D analysis of snow. Avalanche 

Formation, Movement and Effects lAHS Publ. 162, pp. 35-48, 1987 

Gra^enick, F.; and Warbichler, P.: The effects of different methods of 

preparation on reproducing the surface of porous materials. Z. Prakt. 

Metallographie 16, pp. 537-546, 1979 

Grewing, M. ; Praderie, F.; Reinhard, R. (eds.): Exploration of Halley's Comet. 

Springer-Verlag Berlin, 1988 

GrUn, E.; and KOSI-team: Modifications of comet materials by the sublimation 

process: results from simulation experiments, this volume, 1989 

Gubler, H.: Determination of the mean number of bonds per snow grain and of the 

dependence of the tensile strength of snow on stereological parameters. J. 

Glaciol. 20, pp. 329-341, 1978 

Houpis , H.F.L.; Ip, W.-H.; and Mendis, D.A.: The chemical differentiation of the 

cometary nucleus: the process and its consequence. Astrophy. J. 295, pp. 654- 

667, 1985 

Jefiberger, E.K.; Christoforidis, A.; and Kissel, J.: Aspects of the major 

element composition of Halley's dust particles. Nature 332, pp. 691-695, 1988 

Kerridge, J.F.; and Matthews, M.S. (eds.)i Meteorites and the Early Solar 

System. The University of Arizona Press. Tucson, 1988 

Kinosita, S.; and Wakahama, G.: Thin section of deposited snow made by the use 

of aniline. Contr. Inst. Low Temp. Sci. 15, pp. 35-45, 1960 

Klinger, J.; Benest, D.; Dollfus, A.; and Smoluchowski , R.: Ices in the Solar 

System. D. Reidel Publ. Company, Dordrecht, 1985 

Kry, P.R.: Quantitative stereological analysis of grain bonds in snow. J. 

Glaciol. 14, pp. 467-477, 1975 

Lindstrom, D.J.; and Lindstrom, R.M, : Prompt gamma activation analysis (PGAA): 

Technique of choice for nondestructive bulk analysis of returned comet samples? 

this volume, 1989 

Mathis, J.S.: Interstellar and cometary dust, this volume, 1989 

McKay, D.S.: Description and analyses of core samples: the lunar experience. 

this volume, 1989 

Mackinnon, I.D.R.; and Rietmeijer, F.J.M.: Mineralogy of chondritic 

interplanetary dust particles. Rev. Geophy. 25, pp. 1527-1553, 1987 

McSween, H.Y., Jr.: Aqueous alteration in carbonaceous chondrites: Mass balance 

contraints on matrix mineralogy. Geochim. Cosmochim. Acta 51, pp. 2469-2477, 


Morfill, G.E.: Physics and chemistry in the primitive solar nebula, in R. Lucas; 

A. Omont; and L.R. Stora: The Birth and Infancy of Stars. North Holland Publ., 

Amsterdam, pp. 693-792, 1985 

Narita, H. : Measurement of the specific surface of deposited snow I. Contr. 

Inst. Low Temp. Sci. A 27, pp. 77-86, 1969 

Narita, H. : Measurement of the specific surface of deposited snow II. Contr. 

Inst. Low Temp. Sci. A 29, pp. 69-79, 1971 

Nuth, J. A.; and Stencel, R.E. (eds.): Interrelationships among circumstellar, 

interstellar, and interplanetary dust. NASA Sci. Technol. Info. Branch, NASA 

Conf. Publ. 2403. 1986 

Perla, R. : Snow in strong or weak temperature gradients, Part II: section-plane 

analysis. Cold Regions Sci. Technol. 11, pp. 181-186. 1985 

Perla, R.; and Dozier, J.: Observations of snow structure. Proc . Int. Snow Sci. 

Workshop, pp. 182-187, 1985 


Perla, R.; Dozier, J.; and Davis, R.E.: Preparation of serial sections in dry 

snow specimens. J. Microsc. 141, pp. 111-114, 1986 

Perla, R. ; and Ommanney, C.S.C.: Snow in strong and weak temperature gradients. 

Part I: experiments and qualitative observations. Cold Regions Sci. Technol . 11, 

pp. 23-35, 1985 

Roedder, E. : Inclusion measurements - heating, cooling decrepitation and 

crushing, in Reviews in mineralogy 12, Fluid inclusions, pp. 182-219, 1984 

Sandford, S.A. ; and Walker, R.M.: Laboratory infrared transmission spectra of 

individual interplanetary dust particles from 2.5 to 25 microns. Astrophy. J. 

291, pp. 838-851, 1985 

Stem, S.A.: Collision in the Oort Cloud. Icarus 73, pp. 499-507, 1988 

Sutton, S.R.: Non-destructive trace element microanalysis of as-received 

cometary nucleus samples using synchrotron x-ray fluorescence, this volume, 1989 

Wasson, J.T.: Meteorites. W.H. Freemann and Company, New York, 1985 

Weissman, P.R.: Are cometary nuclei primordial rubble piles? Nature '320, pp. 

242-244, 1986 

Whipple, F.L.: A comet model I: the acceleration of Comet Encke. Astsrophys. J. 

Ill, pp. 375-394, 1950 

Wilkening, L.L.: Comets. The University of Arizona Press, Tucson, 1982 

Wood, J. A.; and Morfill, G.E.: A review of solar nebula models, in J.F. Kerridge 

and M.S. Matthews (eds.): Meteorites and the early solar system. The University 

of Arizona Press, Tucson, pp. 329-347, 1988 



P. G. Magnani 

C. Gerli 

G. Colombina 


P. Vielmo 



P.G.Mag?%ani, C. Gerli and G. Colombina, Tecnospazio • P.Vielmo, Tecnomare 


The Comet Nucleus Sample Return (CNSR) Mission, is a cornerstone of ESA scientific 

While Giotto, Vega I and n provided the first picture of a comet nucleus, a much im- 
proved understanding of the nucleus and processes on it will result from in situ measurements 
and Earth based analysis of the material samples collected on the nucleus siuiace. 

The CNSR baseline mission foresees the landing and anchoring of a spacecraft on the 
comet nucleus surface (see Fig. l.a), and the collection of the following three types of samples 
by means of a dedicated "Sample Acquisition System" (SAS) : 

. a core sample gathered from siuface down to a maximum depth of 3 meters to be cut in 0,5 
m. long sections for storage; 

• a volatile material sample, to be gathered at the bottom of the core sample hole; 

• a surface material sample, gathered from one or more locations on the surface. 

These samples will have to be placed in a storage canister in the capsule (to be returned 
on Earth) and preserved therein at a temperature not higher than 160° k. 

If on board sensing instrumentation identified comet nucleus features not allowing a safe 
landing, a back-up system, -based on a "harpoon" sampler, would be laimched from the 
spacecraft hovering the comet, and recovered via a tether line; degraded sample quality would 
be accepted in this case (no surface and volatile samples and limited core sampling depth) 
(Fig. l.b). 


The most likely composition of cometary nucleus, as nowadays generally accepted, is 
given by a finely grained structure of amorphous and/or crystalline ices (water ice or hydrates 
including CO2, 03, and other gases) including micron sized dust particles (carbonauceous 
and/or sUiceous minerals). 

The expected range of physical properties of comet nucleus layers is shown in Table 1. 



































^ LLI 

^ O 
5 LU < 

^ < 





















■^ 5 


























O O 

a 5 
i2 ^ 










h* CO 








































mpling depth °K 


irbonaceous compounds 





:= to 







• • 




























































■ ^■" 

















The structure is expected to be very porous, and because of irradiation phenomena (from 
solar or cosmic origin) with a layering of modification of structural properties, being mainly 
attributed to migration and recondensation of volatile components mobilized by thermal 
radiation during orbital phases closer to Sun. 

At nucleus mantle, the effect of sublimation of most volatile components is expeaed to 
cause an enrichment of dust components and, because of radiation, formation of (complex) 
organic molecular species. 

The most likely stratigraphy of nucleus surface is, therefore, assumed to consist of an 
upper layer made of the less (or not) volatile components (dust grains) bounded together in a 
(highly) porous matrix overlying layers with prevaHing ice content (mostly water ice). 

Because of nucleus activity coimected with sun irradiation, migration of nucleus parts in 
burst phenomena has been observed, hence a not uniform and regular surface pattern is likely 
to occur: some loose material such as dust, pebbles or even boulder size parts could be found, 
overlaying a rough, irregular surface with possible escarpements or areas in which inner layers 
are exposed, being the upper stratum (or "crust") blown away; a pictorial representation of 
comet nucleus morphology is depicted in Fig. 2 where some of the possible operational 
scenario features, affecting sampling process, are shown. 


The main requirements for the design of the SAS derive from the need of achieving mis- 
sion scientific goals for what regards : 

• sampling depth and sample quantity; 
. preservation of stratigraphy pattern; 

. preservation of sample microstructure; 

. avoidance of contamination of sample with hetereogeneous chemical species; 

• representativity of gathered samples. 

Said requirements bring to the following main criteria for design : 

. the system must be capable of trimming the operational parameters to encountered 
nucleus material features; 

. all samples must be stored in the same housing in which have been gathered "in situ"; 

• heat generation during sample cutting process must be monitored and controlled to not 
disturb sample (micro) strucmre; 

• the sampling system design has to be such as to minimize the interaction forces developed 
by nucleus material reactions to cutting process; 

. the maximum integration of sampling system functions will have to be achieved in order to 
minimize their overall number (increasing functional reliabihty and operational 




On the base of the analysis of mission requirements and of trade-off between the pre- 
viously described system concepts, the most suitable solution that has been identified is based 
on a special telescopic device supported at the end of a robotic arm (see Fig. 3). 

This handling device has the possibihty of being connected, by means of a standard inter- 
face at its end, with all the tools which are provided for sample gathering (core, surface, 
volatile samples and back-up harpoon units); this device has the possibility of extending teles- 
copically up to a maximum of 3.6 meters length applying controlled torque to the tools it is 
connected with. All tools are stored in a suitable rack, at the reach of handling device. 

The robotic arm can extend telescopically in order to reach possible sampling sites in a 
proper area below spacecraft; a rotation function allows tilting of telescopic handling device 
in order to perform the main SAS mission phases : 

• unlatching robotic arm from storage configuration; 
. grasping a sampling tool in its storage rack; 

• positioning the same on sampling site; 
. operating the tool together sample; 

. recovering the tool from sampling site (with sample inside); 

• positioning the samples (in their housing) in the return capsule. 

The basic function provided by robotic arm and telescopic device is, the capability to 
operate and activate all the sampling tools. 

In the following, a brief description of the selected concepts for the sampling tools is 

Core Sampling Tools : Among the many different approaches, the core sample cutting process 
by rotary drill is considered the most suitable while other solutions (e.g. impact coring, vibra- 
tion coring, thermal coring, etc..) are not meeting the mission scientific requirements. 

In Fig. 4 the basic arrangement and operating sequence of a coring tool system is 
presented. The particular features of the system are : 

• the hole sides are contained by an outer tube in order to avoid bore hole wall instability 
and missing of stratigraphy information; 

• the bottom of core sample(s) is cut by a proper device, in order to avoid loss of sample 
material and minimize disturbance of its layering. 

In the proposed concept each tool section is made of a inner sample housing tube, carry- 
ing the bottom cutting device, of an intermediate torque tube carrying the drill bit and of an 
outer guide tube. 


Figure 3 - Robotic arm system concept 

























The tool sets have diameters progressively decreasing in order to allow the drilling of 
each tool in the guide tube of previous one left in the hole to provide guidance and bore wall 

Cuttings are conveyed in the gap between torque tube and guide tube and are swept to 
surface by proper springs when each sample is recovered. 

Volatile Sampler : The tool (see Fig. 5) collects material in the bottom of the deep core sam- 
pling hole; it allows preservations of volatile chemical species even if the tool is subject to 
temperature increase during handling to storage bay. 

Surface Sampler : This tool concept (see Fig. 6) is based on a configuration similar to core 
sample tools : the surface material sampler is made of a tube having a special drill bit to dis- 
gregate nucleus "crust" and convey material to an imier sample housing pipe. 

After penetration of tool into nucleus upper layer (about 0,05 m. deep) a sample bottom 
cutter confines the gathered material into the tool and allows its collection into the sample 
housing pipe which is lowered to catch it. 

The main advantage of this concept is the capability to deal with every type of comet soil, 
due to the rotary drill action. 

Harpoon : The harpoon (see Fig. 7) has to be used as a back-up sampling tool should the 
nominal scenario be not feasible; therefore maximum reliability must be guaranteed. For 
these reasons, a spinned harpoon concept has been selected for its capability to withstand dif- 
ferent kinds of soil by varying spiiming speed and for its stability during flight and impact 

The possibility of penetration by using the rotational speed allows a soft impact with the 
soiJ comet, with the consequent minor risks of damaging electrical and mechanical parts and 
to better preserve the collected material properties. 

Anchoring System : The main characteristic of this device (see Fig. 8a. and b.) shall be the 
capability to assure a proper S/C anchoring on the comet to counteract drilling forces and tor- 

Because of imcertainties on comet nucleus mechanical characteristics, a single system 
sizing is not likely to cover the full range of possible real operating conditions : at least two 
sizes of the anchoring tool will have to be foreseen, to be selected accordingly to information 
previously acquired on nucleus features (by the onboard sensing instrumentation). 




















•^r-v.: ^-^.(-^--=^- 















Figure 7 - Spinned harpoon operative sequence 




" . '.^^ ol 


TTj. |I 

, -i 

\\ °l 

\ ^ 3 


\ '•* K 

^ ./•- 


T ^ : ;^ ^ ■■ ^ J r r iV ^"'' .-.-■; '..' :---■■-■ ■- ;^,S-::.. ^w .ij ■ 

















David S. McKay 
NASA Johnson Space Center 

Judith H. Allton 
Lockheed Engineering and Sciences Co. 


Description and Analysis of Core Samples: The Lunar Experience 

David S. McKayl and Judith H. Allton2 
iNASA, Johnson Space Center and 2Lockheed Engineering and Sciences Co. 


Although no samples yet have been returned from a comet, extensive experience from sampling another 
solar system body, the Moon, does exist While, in overall structure, composition, and physical properties the 
Moon bears little resemblance to what is expected for a comet, sampling the Moon has provided some basic lessons 
in how to do things which may be equally applicable to cometary samples. In particular, an extensive series of core 
samples has been taken on the Moon, and coring is the best way to sample a comet in three dimensions. 

Data from cores taken at 24 Apollo collection stations (Duke and Nagle, 1976) and 3 Luna sites have been 
used to provide insight into the evolution of the lunar regolith. It is now well understood that this regolith is very 
complex and reflects gardening (stirring of grains by micrometeorites), erosion (from impacts and solar wind 
sputtering), maturation (exposure on the bare lunar surface to solar winds ions and micrometeorite impacts) and 
comminution of coarse grains into finer grains, blanket deposition of coarse-grained layers, and other processes (fig. 
1). All of these processes have been documented in cores. While a cometary regolith should not be expected to 
parallel in detail the limar regolith, it is possible that the upper part of a cometary regolith may include textural, 
mineralogical, and chemical features which reflect the original accretion of the comet, including a form of gardening. 
Differences in relative velocities and gravitational attraction no doubt made this accretionary gardening qualitatively 
much different than the lunar version. Furthermore, at least some comets, depending on their orbits, have been 
subjected to impacts of the uppermost surface by small projectiles at some time in their history. Consequently, a 
more recent post-accretional gardening may have occurred. Finally, for comets which approach the sun, large scale 
erosion may have occurred driven by gas loss. The uppermost material of these comets may reflect some of the 
features of this erosional process, such as crust formation, and variations with depth might be expected. Overall, the 
upper few meters of a comet may be as complex in their own way as the upper few meters of the lunar regolith have 
proven to be, and by analogy, detailed studies of core samples containing this depth information will be needed to 
understand these processes and the details of the accretional history and the subsequent alteration history of comets. 


Two types of coring devices were used on the Moon during the American Apollo program: drive tubes, 
which were manually pushed or driven with a hammer into the regolith (fig. 2), and a battery-powered, rotary- 
percussion driU corer, which allowed the collection of deeper cores, down to 3 meters in depth (fig. 3). 

Unlike the planned comet sampling mission, the six voyages of Apollo allowed drive tube design to evolve 
in response to factual information about soil characteristics. Initially, the drive tubes were designed to collect a soil 
column 2 cm in diameter and up to 32 cm in length (two tubes screwed together doubled the column length). They 
were designed to be easily opened in the laboratory without disturbing the sample, a feature also desirable in comet 
coring devices. However, these narrow diameter tubes did not easily penetrate the denser-than-expected lunar soil (1.6 
to 2.0 g/cm'^), resulting in only 50 to 60 percent recovery of soil with significant disturbance of the original 
stratigraphy (Carrier et al., 1971). The first design change, after actually trying to core the lunar regolith, was to 
modify the drive tube bit from an inverted funnel shape (originally designed to collect loose, fluffy soil) to a straight 
wall bit (fig. 4). However, these core tube walls were relatively thick (0.5 cm) which contributed to sample 
distortion. Consequently, new drive tubes were designed and flown on die last three lunar missions. 


^^ O 
CD 45 

"CO ^ 






o) E .E i5 























•5 S 






























































Figure 2. An Apollo 1 1 astronaut uses a hammer to drive a 2-cm diameter core tube into the lunar regolith. An 
extension handle is attached to the top of the core tube (NASA Photo ASll-40-5964). 

Figure 3. The Apollo Lunar Surface drill is tested prior to flight. The batteries are contained in the gray box and the 
motor is contained in the wire cage. Fiberglass-epoxy bore stems, used for drilling holes for heat flow probes, are 
shown in place of titanium drill stems in this photo; however, except for the length of individual sections, they are 
similar (NASA photo S-70-29673). 








1.97 cm 


1 1.95 cm, 


2.92 cm 

1.82 cm 

3.32 <m 


2.92 cm 

.09 cin \ 

3.32 cm 

4.13 . 



2.54 cm 



0.14 ( 

Figure 4. - The evolution of drive tube bits is illustrated by the ApoUo 1 1, the Apollo 12 and 14, and the Apollo 
15, 16, and 17 bits. The drill bit is shown for comparison (adapted from Carrier et al., 1971 and Carrier et ai, 

Figure 5. - 4-cm diameter drive tube. This tube has threads near the bottom which allow it to be screwed into 
another tube section. The mbe on the bottom (not shown) has a steel bit formed into the aluminum tube in place of 
the threads (NASA photo S-71-16526). 


Those new tubes (fig. 5) were able to collect a soil column 4 cm in diameter and up to 35 cm in length 
(two tubes screwed together doubled the column length). The wall thickness was 1 mm, the soil column collected 
was disturbed very little and soil recovery was nearly 100% (Carrier et al., 1972). However, these tubes were 
designed, built and flown on very short notice; consequently, no provision was made for easily opening the tubes in 
the laboratory, so the ability to carefully open these tubes was not achieved until 2 years after the last Apollo 
mission. An additional advantage of larger diameter cores is that they provide sufficient sample at any one depth 
interval so that multiple analyses can be performed using a variety of techniques. For example, in the 4 cm diameter 
tubes, a 0.5 cm interval contained enough material (about 10 g) to allow for full grain size distribution for at least 
six size intervals, chemical and rare gas analyses of each size fraction, thin section grain mounts of each size fraction 
with enough sample left over for several other types of analyses, reserves for future analyses, and for making 
continuous thin sections of the core (Fruland et al., 1982). The smaller diameter core tubes (2 cm) did not contain 
enough sample in a 0.5 cm depth interval for many of these analyses. Based on this experience, a comet sampler 
core should be at least 4 cm in diameter. 

The battery-powered Apollo Lunar Surface Drill obtained soil colunwis 2 cm in diameter and up to 3 meters 
in length via rotary-percussive action. The 16 silver oxide-zinc batteries delivered 2270 blows per minute at 280 
rpm. The length of the drill string resulted from screwing together individual tubes 42.5 cm in length (fig. 6). The 
bit contained 5 tungsten carbide cutting tips. Threaded flutes on the exterior of the tubes conducted the cuttings to the 
surface. Detailed dimensions, weights and materials of construction for both the drive tubes and the drill core can be 
found in Allton, 1989. Overall the drill worked well in the lunar regolith, although there was some difficulty in 
extracting the first stem drilled due to the very high bulk density of the soil (1.9 g/cm^). The difficulty was resolved 
on later missions by running the drill in place (without further penetration) to clear the flutes of cuttings. The 3 
meter deep soil columns taken by the Apollo drill were probably the most useful scientifically because the depth 
encompassed radiogenic profiles resulting from galactic cosmic ray bombardment. One disadvantage of the 2-cm 
diameter drill core, however, was lack of enough material in a narrow size interval (< 0.5 cm) for complete size 
fractionation and analysis using multiple techniques on each size fraction. 

Three unmanned Soviet Luna missions obtained lunar regolith cores with an impact and rotary drilling 
device. The drill, which operates 30° from the vertical (fig. 7), automatically changed woiidng regime in response 
to the density of the material being drilled. The Soviet drill differed from Apollo in that the rigid drilling tube 
contained an iimer tube of flexible material, 260 cm long for Luna 24. The drill head interior diameter was 8 mm, 
while the flexible tube diameter was 12 mm. This diameter difference, the vibration during drilling or incomplete 
recovery of the uppermost, loose fluffy regolith may have contributed to the soil column being shorter than the 
drilled depth on the Luna 24 mission (Florensky el al., 1977). The flexible tube, full of soil, was coiled around a 
drum, in spiral fashion, for return to Earth. Again, the restricted diameter of this tube, along with the disturbances 
of the sampling and handling technique, precluded some of the multiple analyses and detailed textural studies. 


The first three Apollo mission samples were handled under a biological quarantine protocol. Sample 
containers were initially brought into vacuum cabinets for opening and for preliminary volatile composition 
sampling. Working with samples in a vacuum cabinet with space suit gloves was chosen over electro-mechanical 
manipulators to allow flexibility to execute or recover from unexpected events. This approach turned out to be 
expensive and not practical; sudden system leaks were exciting events. Core tubes were opened, described and 
sampled in nitrogen atmospheres at slightiy less than ambient pressure. The use of vacuum cabinetry was dropped 
during Apollo 14, afterwhich the biological quarantine was lifted. Samples were processed in pure nitrogen at 
slightly positive pressure thereafter, without significant loss of scientific information. 

The canisters carrying the coiled Soviet regolith cores were opened and initially examined in a helium 
atmosphere (Florensky et al, 1977). However, Surkov et al, 1974 and 1979, reported on a chamber for receiving 
samples that was capable of achieving vacuum conditions. However, the chamber was filled with an inert gas before 
the sample canister was opened. Apparendy, the Soviets too decided that sample handing in a vacuum chamber was 
not worth the expense and the problems. A transparent cylindrical glove box for handling lunar soil samples was 
described by Stakheyev et al, 1974. 









10 CM 

Figure 6. - Sections and bit from Apollo drill. Up to 8 sections were screwed together. All sections, except the one 
holding the bit, were the same length (NASA S-89-25295). 


Figure 7. - Luna vehicle shown with drilling device (cylindrical shape) on arm in two positions. Drilling occurred in 
the lower position. The core in a flexible tube was extracted from the ground, and wound around a cylinder 
(enlargement) which was placed in the Earth return sphere atop the Luna vehicle (arm in upper position) (adapted 
fiomSoviet Cosmonautics, 1981 and Tarasov, et al., 1980). 


A low temperature and low pressure environment will be much more important for comet sample integrity 
than it was for lunar samples. However, the lunar experience with vacuum cabinets illustrates the need to have a 
reliable, practical system for sample handling. The sample handling system should be defined early in the mission 
planning. For example, totally robotic sample handling and human handling via gloves each place different 
requirements on the design of the sample containers. 


The extraction of maximum scientific information from Apollo cores was a complex process consisting of 
three phases, depending on whether the information from the core was obtained before, during or after the core was 
opened and separated into individual samples - a process called dissection. The first phases, pre-dissection and 
dissection, were carried out as a curatorial function. The curatorial information was quickly and impartially 
distributed to all interested scientific investigators. The post-dissection phase involved work on samples requested by 
individual investigators or small consortia through a peer review panel, the Lunar Sample Analysis Planning Team 
which evolved into Lunar and Planetary Sample Team. An extensive database for all curatorial activity evolved from 
first-hand notes and chronological computer entries to an online accounting and tracking system for 70,000 lunar 


Pre-dissection included field Hata collected during actual coxing operations on the lunar surface. This data 
included field relationships to other features, physical properties of the regolith (color, texture, density), and 
penetration depth. These data provided the geologic context for interpretation of the laboratory core aiialyses. 
Investigation of the returned core began with weighing to determine the density of the cored regolith. Next, most 
Apollo cores were photographed with x-rays in two orientations to provide stereo images of the contents of the tubes 
(fig. 8). These x-rays were used mainly to determine the degree of fill, find voids, identify major units, and locate 
smaU fragments, and they did not provide much quantifiable scientific data. Measurement of a nearly in situ thermal 
conductivity was attemped on unopened cores under vacuum (Horai et al., 1980). In general the pre-dissection phase 
of ApoUo cores provided the geologic setting, soil geotechnical properties and some basic textural information. 
However, recent advances in technololgy including three-dimensional x-ray computer assisted tomography (CAT 
scans), nuclear magnetic resonance imaging, and sophisticated digital image analysis might be used to extract more 
useful scientific data fit)m future cometary cores during this pre-dissection phase. Measurements of gas pressure and 
composition can also be made on closed core containers. 


The dissection phase began by opening, inside of nitrogen-filled cabinets, the tubes by one of three 
methods, depending on the type of tube. The most convenient to open was the 2-cm diameter drive tube, in which a 
thin inner tube containing soil was extracted after the cap was removed. This thin iimer tube was constructed of two 
aluminum semi-circular halves held together by teflon tubing. The teflon tube was easily cut open with a scalpel 
and the top aluminum tube half lifted off to reveal the soil. (This double tube construction contributed to the 
increased wall thickness which impeded the penetration of the tube into the dense soil and resulted in redesigning the 
drive tubes.) The second type core tube to be opened was the sections from the driU. These titanium alloy tubes 
were placed horizontally and milled open without lubricant in the nitrogen atmosphere. Without lubricants this 
process inflicted significant vibrations on the soil in the tube. The last method of opening cores was applied to the 
4-cm diameter drive tubes. A large plunger device pushed the soil out of the drive tube and into a fixture with 
exacfly the same sized bore. However, this fixture was constructed of layers, running the length of the core, which 
could be removed one by one revealing the soil. 

Once opened, the exposed core soil was photographed and carefully subdivided into 0.5 cm increments (fig. 
9). The 4-cm diameter drive tubes were of large enough diameter to allow for several sets of 0.5 cm increments to be 





to restrain 
soil after 



Figure 8. - X-ray photograph of 4-cm diameter Apollo drive tube showing individual fragments and regions of 
compacted soil. The steel bit is visible at the bottom, and a device for restraining the soil is visible at the top 
(NASA photo S-89-25298). 


taken at the same depth. Small tools were used to remove soil, which was then sieved. Fragments larger than 1 
mm were measured, photographed and color, texture, coherence, luster and cleavage were described. Preliminary 
gross mineral identification was sometimes made based on binocular microscope observations. The fragments and 
the fine soil were saved in separate vials for long term storage and allocation to investigators for further research. A 
portion of each increment was left untouched so than a continuous length of soil could be preserved. This 
undissected soil was removed from the nitrogen atmosphere in the clean room laboratory and taken to another 
laboratory to be preserved using both an acrylic plastic peel and impregnation by epoxy (fig. 10). Thin sections 
along the length of the core were produced from the epoxy impregnated portion. 

Every step of the dissection was documented by written notes, extensive photography and numerous 
sketches. Dissection for a single core often took up to one year to complete. After the core tube was opened and 
before dissection began, many of the early cores had samples removed for basic petrography and chemical analysis by 
x-ray fluorescence in nearby laboratories. This was done as part of the preliminary examination conducted on all 
Apollo samples at that time. A very useful way of measuring the extent of soil exposure to the solar wind and 
micrometeorites, a maturity index based on ferromagnetic properties, was established in the mid-1970s (Morris, 
1976). Small samples from all depth increments in cores were removed as the dissection progressed and passed on to 
an investigator's laboratory for measurement of this maturity index. Once established as a reliable index, 
ferromagnetic resonance of the sample was routinely used to survey COTes during dissection, for this method requires 
very little sample (25 mg) and is a rapid and non-destructive measurement Information generated during the 
dissection phase, including the preliminary data fix)m the early cores and the maturity measurements, was quickly 
disseminated to scientific investigators via catalogs and newsletters. A similar protocol may be desirable for returned 
cometary core samples. 

This information usually identified the major textural uiuts within the core, described the contacts between 
the units, and located fragments larger than about 0.5 cm. Textural data collected during dissection were usually 
more detailed then the pre-dissection textural information and also included color variations and classification of rock 
and mineral types for larger particles. Descriptions from this dissection stage also provide the basis for accurate 
determinaton of true depth beneath the lunar surface of any sample. This basic sample location and textural 
information provides the framework for interpretation of the data from the more detailed and sophisticated post- 
dissection arudyses which were usually performed by investigators in their own laboratories. While the amount of 
detailed description necessary during the dissection phase has been the subject of some debate, no one has questioned 
the value of basic core description and dissection data in providing a framework for subsequent analyses and 
interpretation. It is particularly important that the dissection phase be carefully planned because some information 
may be permanentiy lost if not acquired during this phase. Once dissected, the core cannot be reassembled to take a 
closer look at features such as uitit boundaries. 


Post-dissection analyses consists of a wide variety of analyses performed by individual investigators on 
subsamples provided in small vials or on thin sections. While in theory, the post-dissection phase does not begin 
untiil the dissection is complete, it was practical to begin allocating and shipping samples after the first dissection 
unit was completed and preliminary information on the sample set was distributed. TTun sections of the nearly in 
situ soil showed the soil fragments in their lunar orientation, such as the agglutinate particle with it's glass 
splattered side up shown in figure 11. However, fewer scientific conclusions were drawn from particle orientation 
and micro-stratigraphy in these thin sections than originally thought Petrographic grain population studies from 
thin sections, did however, sometimes reveal unexpected exotic layers, such as the green gla^s sphere strata in an 
Apollo 15 core from the Apennine Front or the coarse-grained olivine layer from the Apollo 12 site. 

Grains extracted from cores, sorted by size and mounted in thin sections revealed a wide variety of grain 
types which not only varied with depth in the core, but with grain size. Figure 12 illustrates grain size variation 
with depth in a core from the Apollo 15 site. Each core had its own grain size distribution pattern, and the pattern 
was seldom systematic or predictable. However, vertical variations in many properties were common; homogeneous 
cores, even short ones, were the exception. Chemical and isotopic analyses showed variation with core depth and 
also with grain size. Figure 13, for example, shows that the finer size firaction of this core section is enriched in rare 
earths compared to an intermediate size, and this pattern is somewhat different from the pattern in the lower section 
of this core (not shown). 




Figure 9. - Du"!scciion of 4-cm diamclcr Apollo core in 0.5 cm incrcmcnis using small lools inside of a nitrogen- 
nilcd cabinet. 


remaining soil 


♦ .* Acrylic peel 

Figure 10- - Distinct color boundaries arc easily observed in this 4-cm diameter core, mkcn from the Apollo 16 site, 
after a tbin bycr of soil has been removed wiiji a su-ip of acrylic glue. The soil-ridden acrylic strip is shown on the 
nghL The remaining soil on the left will be impregnated with epoxy which can be made into thin sections (NASA 
photo S-77-22213). 


J**" ?^»»: 


♦ •-< 



V- «•» 

Figure 1 1. - Thin section of a 4-cm diameter Apollo 15 core with soil in position as collected in tube. An 
agglutinate is shown with its glass-coated surface facing up. The field of view is 1 mm. 


uj u 
a ~ 








10 - 





20 - 






30 - 



1 '"" 





50 - 

_ _ 55 - 




50 100 150 


Figure 12. - Grain size variation with depth in 4-cm diameter core from ApoUo 15 site (McKay et al., 1980). 


One of the major achievements of lunar core science was the formulation of a model for the variation of 
maturity properties with depth and exposure time. Trends in some vertical regolith sections could be shown to fit 
this model. Communition of soil fragments into smaller particles and formation of agglutinates by micrometeorite 
bombardment, both indicators of soil maturity, are correlated (fig. 14). Soil maturity is also evidenced by the 
abundance of solar flare tracks (fig. 15). The ferromagnetic resonance based index proved to correlate well with 
agglutinate abundance (fig. 16), mean grain size and solar flare tracks. Not only did this index vary with core depth, 
but also with grain size as shown in figure 17. 

More and more complex scenarios had to be developed to explain the data from cores. Samples from 
various depths in this Apollo 16 drive tube core (fig. 18) could be related to each other only by using a combination 
of mixing three end members followed by a superimposed maturation evolution for only one of the mixed sets. 
Studies of individual grains or groups of grains from die cores has also been useful. Figure 19 shows a cluster of 
micrometer-size grains from an ApoUo 15 core which exhibit rounding interpreted as resulting from sputter erosion. 
At about the same scale, figure 20 shows a cluster of mostly submicrometer grains from a porous interplanetary dust 
particle, the type which has been interpreted as a likely sample from a comet If these are really comet samples, 
understanding of a comet would be as complete as an understanding of the Moon from a few random grains. Much 
understanding of the history and evolution of the lunar regolith has come from the detailed, integrated core studies set 
in the proper geologic framework 

Soviet Luna cores 

Although each Soviet Luna core was less supported by additional samples and first-hand field data to form a 
framework for study, the extraction of information from these cores also occurred in pre-dissection, dissection and 
post-dissection phases. During the pre-dissection phase an in situ survey of magnetic properties was made. The 
flexible tube was then coiled in a flat spiral and an x-ray photograph taken. Dissection proceded by uncoiling the 
tube and cutting it into 25-30 cm long sections, with the position of the cut based on the magnetic and x-ray data. 
Each of the tube sections was placed on a special tray and then slit open. The samples were handled under a heUum 
atmosphere. The regolith material was then visually examined, described, and photognqjhed both in visible light and 
by x-ray. Small samples were removed for subsequent processing in nitrogen-filled cabinets. However, the coarsest 
sieve fractions were taken into an air atmosphere for hand-picking of individual fragments of rocks and minerals . 
Small samples of 150-200 mg were taken along the entire core at limited intervals (unspecified). In addition, larger 
samples of about 2 grams were taken from selected 3-4 cm intervals to carry out comprehensive studies on the 
separate soil zones. Half of the material at each depth was reserved for fiiture studies (Rorensky et ai, 1977). Post- 
dissection information came from more sophisticated studies conducted in research laboratories in the Soviet Union 
and abroad 


In summary, Apollo core analysis and handling experience provides a good starting point for the detailed 
planning of a cometary core program. Core collection and analysis is a complex task which must be carefully 
planned from sample collection at the comet to sample allocation in the receiving laboratory. 

1. Core samples from comets are likely to have complex vertical variations and even horizontal variations, 
if preserved by the sampling process. These variations may reflect accretion history and accretion gardening, later 
impact and reworking history, and changes resulting from near-sun passage, with gas and dust disturbances. Possible 
endogenetic processes such as aqueous alteration or gas transport may also be reflected in cometary regoliths. 

2. Collection and analysis of cores from the lunar regolith was a complex process which evolved over time 
so that subsequent core hardware and procedures were improved over earUer ones. This luxury will not be available 
for a comet mission which must be done right the first time. Very careful planning is obviously required. 

3. Not only must the science requirements for the sample be integrated with the mission operation and 
engineering aspects, but designers of flight hardware should also be aware of the conditions under which the samples 
will be opened and examined upon return to Earth. 


CORE 15008 

100X - 

SOX - 




FeO = 20.4 

J L 



Sm Eu 


yb lu 

Figure 13. - The <20 ^m size fraction of this Apollo 15 core is enriched in rare earths compared to the 90-150 nm 
fraction, illustrating differences in chemical composition with grain size. Chemical composition varies with depth 
in the core also; the rare earth pattern at a lower depth in this core is different (Blanchard, unpublished data). 





Figure 14. - Good conelaiion is shown for two indicators of soil mamrity, the % agglutinates in the 90-150 \im size 
fraction and the mean grain size of the soil, in 42 Apollo 17 soils (McKay et al., 1974). 





























O 60009 
• 60010 

J— . 

















40 60 

(90 - 150 /.m) 



Figure 15. - Soils from a 4-cm diameter Apollo 16 core show correlation between the % agglutinates in the 150-250 
]im size fraction and the solar flare particle track density in individual grains from the same soil, another indicator of 
soil maturity (Blanford ec al., 1977). 


15 20 

25 SO 7S 



Figure 16. - Two maturity indices, Ij/FeO (a ferromagnetic based index) and percent agglutinates vary widi depth in 
an Apollo 16 core. The soil at the surface shows the greatest maturity (McKay et al., 1977). 


10 15 20 25 30 35 40 45 50 55 60 65 70 75 
Ij/FeO (ARB) 

Figure 17. - The ferromagnetic index varies with depth in this ApoUo 16 core and also with grain size. Increasing 
values of Ig/FeO indicate greater maturity (McKay et al., 1976). 



• 60009 
O 60010 

250-500 ^m 








Figure 18. - Soil samples from a single Apollo 16 core can be related by superimposing a maturation evolution, 
evidenced by increasing agglutinates, for only one of three end-members in a soil mixture (McKay et al., 1977). 


Figure 19. - Cluster of micrometer-sized grains from an Apollo 15 core which exhibit rounding interpreted as 
resulting from sputter erosion (NASA photo S-78-29566). 

Figure 20. - Submicrometer grains from a porous inteiplanetaiy dust particle, the type which has been interpreted 
be from comets. 



4. Different types of information will be collected during each phase of cometary core analysis: pre- 
dissection, dissection, and post-dissection. Some information may be permanendy lost if not collected during the 
appropriate phase. 

5. Much of the rich experience with lunar core samples can be directly applied to cometary core handling 
and analysis. Examples include the three phases of data collection and analysis, the preliminary examination 
concept, the early and subsequent allocation of samples to outside investigators, the mechanics and bookkeeping of 
the curation and allocation system, and the preservation of a significant portion of the cores for future studies. 

6. Cores returned from a comet can be expected to be just as complex as the lunar regolith cores, and 
vertical variations should be expected. Judging from the lunar experience, these vertical variations may be neither 
systematic nor predictable and may be highly complex. Only a wide variety of analyses interpreted in the framework 
provided by a geologic setting and careful dissection can begin to allow us to understand the history and evolution of 
a comet 

Soviet Cosmonautics, Moscow Machine Building, 1981. 

Allton, Judith R: Catalog of Apollo Lunar Surface Geological Sampling Tools and Containers. JSC 23454, 
LESC-26676, Johnson Space Center, Houston, TX. 1989. 

Blanford, George E.; McKay, David S.; and Wood, G. C: Particle track densities in double drive tube 60009/10. In 
Proc. Lunar Sci. Cortf. 8th, pp. 3017-3025. 1977. 

Carrier, W. David HI; Johnson, Stewart W.; Werner, Richard A.; and Schmidt, Ralf: Disturbance in samples 
recovered with the Apollo core tubes. In Proc. Lunar Sci. Conf. 2nd, vol. 3, pp. 1959-1972. M. I. T. Press, 1971. 

Carrier, W. David HI; Johnson, Stewart W.; Cairasco, Lisimaco R; and Schmidt, Ralf: Core sample depth 
relationships: Apollo 14 and 15. In Proc. Lunar Sci. Conf. 3rd. vol. 3, pp. 3213-3221. M. I. T. Press, 1972. 

Duke, M. B.; and Nagle, J. S.: Lunar Core Catalog. JSC 09252. NASA, Johnson space Center, Houston, TX. 

Florensky, C. P.; BasUevsky, A. V.; Ivanov, A. V.; Pronin, A. A.; and Rode, O. D: Luna 24: Geologic setting of 
landing site and characteristics of sample core (preliminary data). In Proc. Lunar Sci. Conf. 8th, pp. 3257-3279. 

Fruland, R. M.; Nagle, J. S.; and Allton J. H.: Catalog of the ApoUo 16 Lunar Core 60009/60010. JSC 17172, 
Lunar curatorial Branch Publication 61. NASA, Johnson Space Center, Houston, TX. 1982. 

Horai, K.; Winkler, J. L. Jr.; Keihm, S. J.; Langseth, M. G.; Fountain, J. A.; and West, E. A.: Thermal 
conduction in a composite circular cylinder A new technique for thermal conductivity measurements of lunar core 
samples. In Philosophical Transactions of the Royal Society of London, vol. 292, pp. 571-598. 1980. 

McKay, D. S.; Fruland, R. M.; and Heiken, G. H.: Grain size and evolution of soils. In Proc. Lunar Sci. Conf. 
5th, pp. 887-906. 1974. 

McKay, D. S.; Morris, R. V.; Dungan, M. A.; Fruland, R. M.; and Fuhrman, R.: Comparative studies of grain 
size separates of 60009. Jn Proc. Lunar Sci. Conf 7th, pp.295-313. 1976. 

McKay, D. S.; Dungan, M. A.; Morris, R. V.; and Fruland, R. M.: Grain size, petrographic, and FMR studies of 
the double core 60009/10: A study of soil evoluton. In Proc. Lunar Sci. Conf 8th, pp.2929-2952 1977. 

McKay, D. S.; Basu, A.; and Nace, G.: Lunar core 15010/11: Grain size, petrology, and implications for regolith 
dynamics. JnProc. Lunar Planet. Sci. Conf 11th, pp.1531-1550. 1980. 


Morris, R. v.: Surface exposure indices of lunar soils: a comparative FMR study. In Proc. Lunar Sci. Conf. 7th, 
pp.315-335. 1976. 

Stakheyev, Yu. I.; Tarasov, L. S.; Krestinina, K. K.; and Ivanov, A. V.: Box for preliminary investigation of the 
lunar soil in nitrogen atmosphere. In Lunar Soil from the Sea of Fertility, ed. Vinogradov, A. P., Publishing 
House "Nauka", Moscow, pp. 34-37. 1974 (Russian). 

Surkov, Yu. A.; Rudnitsky, E. M.; and Glotov, V. A.: Reception and smdies of lunar matter in a medium of inert 
gas. In Lunar Soil from the Sea of Fertility, ed. Vinogradov, A. P., Publishing House "Nauka", Moscow, pp. 29- 
33. 1974 (Russian). 

Surkov, Yu. A.; Heifez, E. M.; Rudnizky, E. M.; Danilov K. D.; Glotov, V. A.; Visochkin, V. V.; and Sherstjuk, 
A. I.: Acceptance and investigation of lunar soil in noble gases atmosphere or/and at ultrahigh vacuum. In Regolith 
from the Highland Region of the Moon, ed. Barsukov, V. L. and Surkov, Yu. A.J^blishing House "Nauka", 
Moscow, pp. 31-40. 1979 (Russian). 

Tarasov, L. S.; Ivanov, A. V.; Vysochkin, V. V.; Rode, O. D.; Nazarov, M. A.; and Sherstyuk, A. J.: Reception 
and preliminary study of Luna 24 regolith core. In Lunar Soil from Mare Crisium, ed. Barsukov, V. L., Publishing 
House "Nauka", Moscow. 1980 (Russian). 



L. J. AUamandola 

NASA Ames Research Center 

Moffett Field, California 




NASA-Ames Research Center MS 245-6 
Moffett Field, CA 94035 


A wealth of information essential to understanding the composition and physical 
structure of cometary ice and hence gain deep insight into the comet's origin and 
history, can be gleaned by carrying out a full range of spectroscopic studies on the 
returned sample. These studies ought to be among the first performed as they are 
generally non-destructive and will provide a broad data bank, which will be crucial in 
planning subsequent analysis. Examples of the spectroscopic techniques discussed 
below, along with relative sensitivities and transitions probed, are listed in table 
1. This table is by no means exhaustive and other techniques will certainly be added 
as more spectroscopists become aware of the unique challenges and opportunities 
afforded by the comet nucleus sample return mission. 

Each "spectroscopy" will be summarized, with emphasis placed on the kind of 
information each provides. Infrared spectroscopy should be the premier method of 
analysis as the mid-IR absorption spectrum of a substance contains more global 
information about the identity and structure of that material than any other 
property. In fact, the greatest strides in our understanding of the composition of 
interstellar ices (thought by many to be the primordial material from vrfiich comets 
have formed) have been taken during the past ten years or so because this was when 
high quality infrared spectra of the interstellar medium (ISM) first became 
available. The interpretation of the infrared spectra of mixtures, such as expected 
in comets, is often (not always) ambiguous. Consequently, a full range of other non- 
destructive, complementary spectroscopic measurements are required to fully 
characterize the material, to probe for substances for which the IR is not well suited 
and to lay the groundwork for future analysis. 

Given the likelihood that the icy component (including some of the organic and 
mineral phases) of the returned sample will be exceedingly complex, these techniques 
must be intensely developed over the next decade and then made ready to apply 
flawlessly to what will certainly be one of the most precious, and most challenging, 
samples ever analyzed. 



(10,000-10 cm~^) 

Infrared spectroscopy gives great insight into a material -s compostion because 
each chemical bond in a molecule produces a characteristic set of well defined, 
fundamental, vibrational frequencies. The spectrum of these frequencies, nearly all 
of which fall in the 4000 - AOO cm" range for molecules made up of the most abundant 
elements H, C, N and 0, is generally very characteristic of the molecular subgroups 
comprising the molecule (Fig. 1). For example, every molecule which has a CH bond has 
at least one fundamental CH stretching mode which falls in the 3300 - 2800 cm" 
region, independent of how the C is further bonded (Fig la). The precise frequency 
within this region however, is determined by how that C is further bonded (Fig lb). 
If the carbon is singly bonded to a C, or N atom, the C-H stretching modes are 
between about 2820 and 2990 cm" , the CH bending (or deformation) modes fall in the 
lAOO - 1500 cm" range and the corresponding C-C, C-0, or C-N stretch lie between 1300 
and 800 cm~^ . If it is doubly bonded the C-H stretch lies between 3000 to 3100 cm" , 
the CH bend in the 900-700 cm" region and the heavy atom C=C stretch (probably very 
weak) falls close to 1800 cm" . Finally, if it is triply bonded, the CH stretch lies 
close to 3300 cm~^ , the C=C stretch near 2100 cm~^ and the C-H bend near 700 cm"-^. 
Thus obtaining the entire mid-IR spectrum allows one to place very important 
constraints on the types of chemical structures present. 

Absorption measurements on ices in the 10,000-4000 and 400-10 cm region are 
scarce, however these should prove important regions as well. The 10,000-4,000 cm 
region will be rich in overtone and combination bands, and thus compounds which have 
modes that saturate in the mid-IR should show characteristic, unsaturated features in 
the near IR. (Reflection spectra of ices have been made in this region to interpret 
the spectra of some planets and satellites. See Nash and Nelson, 1979 and Nash and 
Howell, 1988 and references therein). The 400-10 cm" region is also largely 
unexplored for bona-fide mixed molecular ices. Here one probes very low frequency 
modes such as large molecule bending vibrations and lattice vibrations. The latter 
may well prove important in characterizing the ice structure (i.e.. Are the samples 
crystalline or amorphous?). 

A good example of how infrared spectroscopy has made an important contribution to 
astrophysics is provided by the analysis of the icy component of interstellar dust in 
molecular clouds. As recently as fifteen years ago the composition of this dust was 
largely speculated about, if not ignored. This situation has changed dramatically 
during the last decade primarily because it is now possible to directly measure the 
absorption spectra of interstellar ices and interpret these spectra with the aid of 
laboratory analog studies. Figure 2 illustrates how the direct comparison of 
laboratory spectra with interstellar spectra can be used to determine interstellar ice 
composition. Besides probing the composition each absorption band can also provide a 
quantitative measure of the amount of each material present, give insight into the 
nature of the ice, (i.e., is it amorphous, crystalline, highly polar, nonpolar etc.), 
and, in some cases, it can give an indication of the thermal history (Tielens and 
Allamandola, 1987, Allamandola, Sandford and Valero, 1988; Sandford et al, 1988; 
Sandford and Allamandola, 1988, Sandford and Allamandola, 1990). 

The ices in comets have certainly been processed by photons and energetic 
particles, producing large, complex species. (Moore et. al, 1983; Lanzerotti, Brown 
and Johnson, 1984; Johnson, Cooper and Lanzerotti, 1986; Strazulla, Pirronello, and 
Foti, 1983; Khare et al., 1988; Allamandola, Sandford and Valero, 1988; Schutte, 


Mum I 2.5 




I lart 
(AO 300.0 




Str«tchin9 mod«9 

1000 »0 33 

Lattict modts 




Binding modw 














= C— H 

= C 





BB ■■ BROAD t 





O. .0— H 

Figure 1. a) The fundamental vibrational frequency ranges for various molecular 
subgroups made up of the most cosmically abundant elements H, 0, C, and 
N. b) Expansion of the 2600-3600 cm"^ X-H stretching range showing the 
precise regions in which specific types of molecules absorb. This 
illustrates the power of IR spectroscopy to place strong constraints on the 
major types of compounds present in any given sample. 


4 5 6 8 

10 12 14 


LACY n al. Ap. J. 276. 533 

fir /^ CH3OH 
f AT 10 K 




6 7 

Figure 2. Top: The entire mid-infrared spectrum of an IR source (W-33A) imbedded in 
a dust cloud. The solid line is the presumed black body emission from the 
source and the dots are the spectral data measured at earth (Soifer et al . , 
1979). The absorption bands are due to the intervening dust and reveal its 
composition. The strong band at 10 microns is due to the SiO stretch in 
the silicates, and the strong band at 3 microns is due to the OU stretch in 
U2O ice. Bottom: More recent, higher resolution observations of the 
interstellar ice features compared with laboratory simulations ^ich show 
that CO and a cyano-(CN) containing compound are frozen in the ice (Lacy et 
al., 1984) and that ices made up principally of HjO and CH^OH account for 
the strong absorptions at about 6 and 7 microns (Tielens and Allamandola, 

1988.) The nature of the complex species produced by these very different processes 
will probably differ. It is quite possible that the outer few meters is rich in 
material produced by energetic particle bombardment whereas deeper in, the comet may 
be rich in material produced in the pre-solar system interstellar cloud. Studies of 
these various processes which are under way in only a few laboratories must be 
increased to ensure that when the sample is returned, the correct interpretation can 


be made. Figure 3 shows how the mid-infrared spectrum of a laboratory analog of an 
interstellar ice changes as the ice is UV processed, reflecting its photochemical 

Figure A shows the spectral evolution in the CH stretching region as the 
photolyzed sample is warmed from 200 to 300 K. This serves to --illustrate another 
important point concerning analysis of the returned sample. Various components 
(organic in the example shown in Figure A) evaporate nearly continuously as the 
temperature is raised. Figure 5 shows the behavior of many IR absorption bands as a 
different irradiated mixture is warmed from 20 to 300°K. Iliese results show that the 
comet sample must be returned at the lowest temperature possible. It is erroneous to 
think that above any given temperature all of the volatiles have evaporated leaving 
only a non-volatile residue. 



This also measures the vibrational spectra of material and, as such, is a direct 
complement to the IR. It is not redundant and provides essential additional 
information. An example of how the infrared vibrational spectrum of a material may be 
different from the Raman vibrational spectum and thus yield important, different 
information is provided by interplanetary dust particles (IDPs). IR radiation passes 

(3) AT 10 K 


1800 1600 1400 1200 1000 


(b) AFTER 45 ^flINUTES UV 

3500 3000 2500 2000 1500 1000 

Figure 3. The infrared spectra of an HjO: CH3OH (2:1) ice taken at 10 K a) before 
and b) after A5 minutes of ultraviolet photolysis. H2O and CH^OH are the 
major constituents of many interstellar ices and presumably important 
precursors to cometary ices. These spectra show that HCO, H^CO, CO, and 
CO2 are important photoproducts in H^O and CHoOH ices (Allamandola, 
Sandford, and Valero, 1988). 


Figure 4, 

Figure 5. 



Of H,0 ; CK,0« : CO : NH, : C,K, 

100 : SO : 10 : 10 : 10 

The infrared spectra (taken at 10 K) 
in the CH stretching region of the 
low vapor pressure coaterials pro- 
duced by 15 hours of photolysis 
of an H20:CH3OH:NH3:CO:C3Hg 
(100:50:10:10:10) ice after tem- 
porary warm up to a) 200 K, b) 250 K, 
and c) 300 K. The different subli- 
mation behavior of the various 
aliphatic molecular components shows 
that a mixture of low vapor pressure, 
complex, organic materials are pro- 
duced upon photolysis (Allamandola, 
Sandford, and Valero, 1988). 

Infrared peak absorb- 
ance strengths as a 
function of temperature 
during warm-up for- a 
number of infrared 
features present in an 
H20:CO:NH3 (100:40:20) 
ice which was prepared 
by slow deposition with 
simultaneous photolysis 
at 10 K for about 20 
hours. Note that the 
sample composition 
varies at all tempera- 
tures from 10 to 300 K, 
emphasizing how crucial 
it will be to return 
the comet nucleus 
sample at the lowest 
temperature possible 
(Schutte, 1988). 




-I 1 r 

I / 






-I h- 




-i h 











• 1 1 < 











1 \ 




m v\ 
















1 1 


, , 1 , 

100 200 

T (K) 


through the small particle and probes the bulk. Sandford and Walker (1985) have shown 
that the IR spectrum of an IDP can be used to characterize the dominant mineral phase. 
With this information they demonstrated that IDPs fall into three mineral classes, 
olivenes, pyroxenes and layer-lattice silicates. However, the Raman spectra of the 
same IDPs are very different, revealing the nature of the minor carbonaceous material 
present and showing no evidence of the dominant mineral phase (Allamandola, Sandford 
and Wopenka, 1987; Wopenka, 1988). As Raman spectra are obtained by scattering 
ultraviolet and visible photons, a minor amount of strongly absorbing carbonaceous 
material is sufficient to dominate the spectrum. The Raman spectra imply that IDPs 
contain an amorphous carbon phase vrfiich probably coats the minerals. In addition to 
the vibrational structure, some of these IDPs luminesce in the red, a common 
characteristic of heavily cross-conjugated carbonaceous materials. Luminescence is 
discussed further below. 


While the infrared is well suited to detect many of the major constituents and 
probe the sample's radiation and thermal history, important minor constituents and 
abundant species with low IR activity will remain undetected. Ultraviolet-Visible 
(UV-Vis) spectroscopy provides an important addition to IR spectroscopy. It may be 
difficult to measure resolved structure in the UV-Vis spectra of a comet nucleus 
sample, especially in absorption. Several strong absorptions may overlap and produce 
a broad, structureless band, or one strongly absorbing or luminescing material may 
dominate in ice poor samples. However, if the returned sample is largely icy, then 
the chances are high that the spectra will provide critical additional information. 

Several examples will now be given which illustrate how UV-Vis spectra complement 
the IR and Raman spectra. 


A potentially important cometary species is HCO, a radical which is readily 
produced by the energetic processing of ices containing CO and photolabile H atoms. 
Upon warm-up HCO can diffuse through the ice and react with other HCO molecules and 
produce H2C202(van IJzendoorn et al, 1983, 1986, 1990; d'Hendecourt et al, 1986). H 
addition to HCO may also produce some H2CO and eventually perhaps CH^OH (Tielens and 
Hagen, 1982). The HCO radical can be stored in water-rich ices, up to surprisingly 
high temperatures (120 -130 K). In an ice, HCO absorbs strongly across the visible 
from about A70 to 620 A and is easily detected in absorption (van IJzendoorn et al . , 
1983), whereas, as shown in Figure 3, the IR absorption of HCO in a mixed ice is weak 
(d'Hendecourt et al, 1986; Allamandola, Sandford and Valero, 1988). 

Sulfur, although minor in abundance compared to carbon, nitrogen and oxygen, is a 
very important interstellar element. Sulfur bearing compounds have been suggested in 
interstellar ices (Geballe et al , 1985) and Sj has been detected in a cometary coma 
(A'Hearn, Feldman, and Schleicher, 1983). Sulfur containing compounds often absorb 
strongly enough in the ultraviolet and visual spectral regions to be detected, even if 
present in minor amounts, whereas they might go unnoticed if one only had infrared 

Another potentially important reactive cometary species which may escape IR 
detection, but which absorbs strongly in the visible, is ozone (O^). The visible 
absorption spectrum of an ice containing ozone is shown in figure 6. 


—I 1 1 — 

CH30H/H20/02 (1/1/1) 

deposited through di&chcrge 





500 500 



Figure 6. Single beam absorption spectrum of an H20:CH20H:02 (1:1:1) ice deposited 

through a discharge at 12 K. The strong, broad, absorption band under the 
dashed line is due to O-, (ozone) (van der Zwet, 1986). In atomic H poor 
regions of dense molecular clouds O2 is expected to be an abundant ice 
constituent. Energetic processing of these ices will readily produce 
ozone, making it a likely comet constituent as well. 


Photoinduced Luminescence 

Detection of emission induced by photon absorption is a much more sensitive 
technique than either Raman or IR and UV-Vis absorption spectroscopy. Furthermore it 
is molecule specific and does-not suffer from screening to the same extent that 
ultraviolet and visible absorption measureooents do. By using various sources of 
monochromatic excitation (as from a tunable laser or dispersed light source) one can 
scan through the spectrum searching for frequencies which induce luminescence to the 
red of the excitation frequency. While the frequency which stimulates the emission 
gives some insight into one transition of the carrier, the emission spectra are often 
very characteristic of the emitter. For example, van IJzendoom et al., 1986 gave a 
complete discussion of the laser induced luminescence spectroscopy of glyoxal (H2C202\ 

in various low temperature solids and used luminescence techniques to track formalde- 
hyde (HjCO) as well (Figure 7). Together the excitation/emission spectra can provide 
an unquestionable assignment. 

As with the IR, this method can be used as a probe of thermal history as well. 
Figure 8 gives the luminescence from S^ and figure 9 shows how the laser induced 
fluorescence (LIF) of S2 in an ice varies as a function of temperature. The intensity 
increase of S2 from about 100 K up to 160 K is presumably due to the diffusion of S 
through the ice and its preference for reacting with another sulfur atom rather than 
other constituents in the ice. Above 160 K, the sulfur reservoir diminishes and the 
S2 apparently reacts with other ice constituents. Note that the S2 peaks at 160 K and 








1 1 1 1 


\ r 

— 1 1 1 1 

\ N,(Hf)>500) 

11 Fluortsctnc* 


1 Ptiosphor«tc«nct 



— \ — ^ 

\ Ptiotphorttctnct 

460 460 500 520 530 550 570 590 610 

1 1 3 1 

Figure 7. Laser induced fluorescence (A * A ) and phosphorescence ( A^ * A ) 

spectra from glyoxal (H2C2O2) in N2 and CO ices. The excitation wavelength 

was 440 nm in both ices. The concentration of this species grows when 

irradiated interstellar ice analogs are warmed (van IJzendoorn et al., 

1986; van IJzendoorn et al., 1990). 

380 )400 u;o ^^o geo <*so soo 
wove'e'^gth (nm) 

Figure 8. Laser induced fluorescence of the B-X transition of $2 in a) H20:H2S (10:1) 
and b) H20:CO:CH^:H2S(5:2:2:1 ) ices which had been photolyzed at 10 K 
(Grim, 1988; Grim and Greenberg, 1987). 


I ' I ' I ■ I ' I ■ I ■ 1 ■ I 

b : 

' ■ 

10 eo 120 ISO 200 

tenperoture (K) 

Figure 9. Intensity of the laser induced fluorescence from $2 as a function of 

temperature during the warmup of the photolyzed ices: a) H20:H2S{ 10:1); b) 
H20:CO:H2S(10:1:1); and c) H20:CO:H2S(10:2:1) showing that S atoms are 
trapped m these ices up to about 100 K (Grim, 1988; Grim and Greenberg, 
1987). Above this temperature S can diffuse and apparently preferentially 
reacts with other S atoms up to about 160 K. Thus S atoms may well be 
trapped in the returned comet nucleus sample and luminescence from $2 might 
be expected. 

is not exhausted until 200 K. Thus, LIP of S2 can potentially be used as a measure of 
the thermal history of the sample in the high temperature region while HCO may play 
this role in the low temperature regime (recall that HCO is probably depleted by 130 
K, see adjacent subsections). 

Chemically Induced Luminescence 

All samples irradiated at 10 K which contain CO and an H atom source glow upon 
warm-up (Hagen, Allamandola and Greenberg, 1979; Van IJzendoorn, 1985; Van IJzendoorn, 
et al., 1990). The spectrum of the thermally promoted luminescence is shown in 
figure 10. This has been ascribed to emission form the transition state in the 
reaction: HCO + HCO — > H2C2O2 (glyoxal). 

The temperature range over which this particular luminescence occurs is 
determined by the major constituent of the ice. If it is rich in non-polar molecules 
such as CO and CU,, HCO is not trapped in deep, "high temperature" sices and is 
depleted by the time the sample has reached 40-50K. If, on the other hand, the ice is 
rich in polar molecules such as HjO, the HCO can remain trapped until the ice tempera- 
ture exceeds about 120-150K (van IJzendoorn, 1985; van IJzendoorn et al., 1990). 

This and other trapped radicals which can diffuse and produce luminescence during 
wamr-up may be present in the returned sample. Thus, an attempt to monitor this sort 
of luminescence during the core drilling phase should be made, perhaps by placing 
light pipes or even filtered detecters directly in the drill bit itself. Thermally 
promoted luminescence should certainly be searched for during subsequent sample 
preparation and wars— up of parts of the returned sample. 


— 1 1 






1 — n43c»- I — I7i7cm- — 1 

l-ttiSdl-^l— T7i7en.- 1 

1 1 1 

i.SO 500 



1 1 


— , , 


28 K J 



1— iMScm- 1 TTMcm- 1 



32 K J 

— 1694 c«- — 1 i7Ue«"— t 


[Otc«- 19*.C(n-' 

1 1 


iSO 500 



Figure 10. Chemi luminescent spectra emitted during warm up from CO:CH/ (100:1); 

COrHjOdOO:!); and COiNH^dOO-.l) ices which had been deposited at 10 K with 
simultaneous deposition for two hours. In CO rich ices such as these this 
emission peaks in the 40-50 K range, in H2O rich ices this emission is 
intermittent up to about 125 K (van IJzendoorn, 1985; van IJzendoorn 
et al., 1990). 


A much more sensitive technique, and one which is particularly well suited to the 
detection of all trapped radicals because it is sensitive to all unpaired electrons, 
is ESR. This is especially important since, if radicals are present, some will most 
likely be in low concentration and probably escape detection by most of the other 
techniques, except perhaps luminescence. Even if luminescence from some radicals is 
detected, complete radical characterization of the sample will almost certainly be 
impossible without ESR. Tsay and coworkers have successfully applied ESR techniques 
to many difficult samples for several years. At this meeting they have shown the ESR 
spectrum of many interesting intermediates trapped in ices, many of which are 
reasonable to expect in cometary ices. Many of the radicals they have studied have 
similar chemical structures and consequently very similar UV-Vis and IR spectra. Tsay 
and colleagues have shown, however, that even in ices the ESR spectra of chemically 
similar radicals differ substantially and one can identify the individual species. 


Furthermore they have shown that the ESR technique is sensitive enough to probe the 
radical concentration decay during warm-up to rather high temperatures. For example 
they have probed the HCO decay in an ice up to 150K and the CH^CUO radical in a 
different ice up to 250K. Thus this technique can also be used to probe the thermal 
history of the sample from the low to the high temperature regimes and extends the 
upper temperature limit well above that discussed in the previous sections. 


To the best of my knowledge, this technique has not yet been applied to mixed 
molecular ices but certainly should be. It can be used to probe specific nuclei such 
as H or ^ C in the sample. All of the nuclei are detectable. The "chemical shift" 
displayed is a measure of the local environment of the nuclei. For example, an H 
attached to an aliphatic carbon has a different chemical shift (i.e., resonates at a 
different frequency) from an H attached to an aromatic carbon or to a carbonyl 
carbon. In most cases, even within these chemical classes, smaller shifts are 
detectable which specify how much of the sample is in one subset of molecule 
comprising a particular class and how many are in another subset. 

As all nuclei are sampled, the NMR spectrum gives a rather accurate measure of 
the relative amounts of all of the different kinds of molecules present, and relative 
amounts of each subset therein. Thus classes of molecules present to the level of a 
few percent which may go unnoticed with the other techniques will not escape 

Modern solid state NMR techniques have made a large impact on our understanding 
of coal, an extremely complex material (Miknis, 1988; Stock, Muntean and Botto, 
1988). The analytic challenges presented by the comet sample will be similar to those 
encountered in coal research. A recent, elegant NMR experiment has been carried out 
by Cronin, Pizzarello and Frye, 1987 on the carbonaceous component of several 
meteorites, another very difficult material to analyze. These spectra showed, for the 
first time, the relative amounts of aromatic and aliphatic carbon in these 
materials. This information had not been possible to obtain by any of the other 
techniques described above, yet all have been applied extensively to meteorites. 


The spectroscopic analysis of pristine cometary material promises to provide a 
very important, often non-invasive, probe of the chemical identity of the material 
present as well as of the physical and chemical conditions which prevailed during the 
comet's history. Concerning classical spectroscopy, the spectral regions which will 
most likely prove most useful are the infrared, the visible and ultraviolet. "Newer" 
spectroscopic techniques which have the potential to provide equally important 
information include nuclear magnetic resonance (NMR) and electron spin resonance 

The infrared should be the premier method of analysis as the mid-infrared 
absorption spectrum of a substance contains more global information about that 
substance's identity and structure than any other property. However, the 
interpretation of the infrared spectrum of the mixtures expected in a comet can be 
(not always) ambiguous. Other nondestructive, complementary spectroscopic 
measurements are required to characterize the material and probe for substances for 
which the infrared is not particularly well suited. While the mid- and far-IR span 
frequencies which correspond to skeletal vibrations in molecules and thus provide 
insight into the identity of chemical groups present, the ultraviolet, visible and 


near-infrared span frequencies which correspond to electronic transitions and give 
insight into the molecular bonding structures present. In these regions absorption 
and emission studies are desirable. Absorption measurements have the potential to 
give an indication of the importance of conjugated bond systems, although sample 
porosity will almost certainly make the measurements difficult. More tractable will 
be luminescence studies. Measuring the luminescence spectrum excited by ultraviolet 
and visible photons should be straightforward. The emission spectrum, as well as the 
wavelength dependence of the exciting light (the excitation spectrum), give important 
insight into the nature of emitting materials. In addition to UV-Vis induced lumines- 
cence, thermally promoted chemiluminescence should also be searched for. Irradiation 
of solid materials often produces trapped ions, electrons, and radicals which can 
diffuse through the medium if it is warmed. Reactions involving these diffusing 
species often emit a spectrum which is characteristic of the reacting species. The 
temperature domain over which light is emitted depends on the nature of the solid. 
Volatile rich ices luminesce in the 10-40 K range, HjO rich ices in the 10-150 K range 
and higher melting point materials luminesce at much higher temperatures. Thus the 
monitoring of potential luminescence during core drilling and during subsequent sample 
warm-up is important to consider seriously as it can provide unique information on the 
thermal and radiation history of the sample which cannot be obtained in any other way. 

Three additional spectroscopic probes have also been briefly sunmarized: Raman, 
NMR and ESR. Raman spectroscopy is complementary to IR spectroscopy in that it 
measures the vibrational frequencies of the material. It is not redundant. ESR 
studies can directly determine the total radical content of the ice. ESR is one of 
the most sensitive techniques available, and the concentrations of many radicals can 
be probed. NMR spectra should also be measured as these reveal the fractions of 
various classes of all of the organic compounds present. The full potential of these 
last three techniques can be realized on the returned comet sample only if laboratory 
programs are started to learn how to apply them to realistic mixed ices. Of the 
three, only the ESR technique has been applied to molecular ices. 

Given the likelihood that the icy component (including some of the organic and 
mineral phases) of the returned sample will be exceedingly complex, these techniques 
must be intensely developed over the next decade and then made ready to apply 
flawlessly to what will certainly be one of the most precious, and most challenging, 
samples ever analyzed. 


Table 1. Summary of Spectroscopic Techniques to Apply to the 
Returned Comet Nucleus Sample 





Thin Section (< 0.3 tun) 




Molecular Vibrations 
Molecular Vibrations 



Laser Induced 

Thin Section (0.1-O.Oliun) 

Emission, Sample warm-up 

Chemically Induced 

Very high 

Probably high 





Low to Moderate Molecular Vibrations 


Thin Cylinder Submicron 

Extremly high 

Unpaired electron 


Thin Cylinder 
millimeter diameter 

Very low 

Nucleus spin-flip 


A'Hearn, M.F. Feldman, P.D., and Schleicher, D.G., 1983, Ap. J. Letters, 27A, L99 

Allamandola, L.J., Sandford, S.A., and Valero, G.J., 1988, Icarus 76, 225 

Allamandola, L.J., Sandford, S.A. , and Wopenka, B., 1987, Science 237 , 56 

Cronin, J.R., Pizzarello, S., and Frye, J.S., 1987, Geochim. Cosmochim. Acta, 51 , 

d 'Hendecourt , L.B., Allamandola, L. J. , Grim, R.J. A., and Greenberg, J.M., 1986, 
Astron. Astrophys. 158 , 119 

Geballe, T.R., Baas, F. , Greenberg, J.M., and Schutte, W. , 1985, Astron. Astrophys. 
U6, L6 

Grim, R.J. A., and Greenberg, J.M. , 1987, Astron. Astrophys. 181 , 153 

Grim, R.J. A., 1988, Ph.D. Dissertation, Leiden University, The Netherlands 

Hagen, W. , Allamandola, L.J., and Greenberg, J.M. , 1979, Astrophys. Sp. Sci . 65, 

Johnson, R.E., Cooper, J.F., Lanzerotti, L.J., 1986, in Proc. 20th ESLAB Symp. on 
Exploration of Halley's Comet , ESA SP-250, Vol. II, 269 

Khare, B.N., Thompson, W.R. , Murray, B.C., Chyba, C, and Sagan, C, 1988, Icarus, 
in press 

Lacy, J.H., Baas, F., Allamandola, L.J., Persson, S.E., McGregor, P.J., Lonsdale, 
C.J., Geballe, T.R., and van de Bult, C.E.P., 198A, Ap. J. 276, 533 

Lanzerotti, L.J., Brown, W.L. , and Johnson, R.E., 198A, in Ices in the Solar 
System , eds Klinger, J. et al . (Reidel, Dordrecht) 317 

Miknis, F.P., 1988, in Mew Trends in Coal Science , ed. Yurum, Y., (Kluwer, 
Dordrecht), 117 

Moore, M.H., Donn, B. , Khanna, R. , and A'Hearn, M.F., 1983, Icarus 54, 388 

Nash, D.B., and Howell, R.R., 1988, Science 244, 454 

Nash, D.B., and Nelson, R.M., 1979, Nature 280, 763 

Sandford, S.A., and Walker, R.M., 1985, Ap. J. 29J., 838 

Sandford, S.A., and Allamandola, L.J., 1988, Icarus 76, 201 

Sandford, S.A., Allamandola, L.J., Tielens, A.G.G.M., and Valero, G.J., 1988, Ap. 
J. 329, 498 

Sandford, S.A., and Allamandola, L.J., 1990, Ap. J., in press 

Strazulla, G., Pironello, V., and Foti, G., 1983, Astron. Astrophys. 123, 93 

Schutte, W., 1988, Ph.D. Dissertation, Leiden University, The Netherlands 


Soifer, B.T., Puecter, R.C., Russell, R.W., Willner, S.P., Harvey, P.M., and 
Gilletc, F.C., 1979, Ap. J. (Letters), 232, L53 

Stock, L.M., Muntean, J.V., and Botto, R.E., 1988 in N ew Trends in Coal Science , 
ed. Yurum, Y. (Kluwer, Dordrecht), 159 

Tielens, A.G.G.M., and Allamandola, L.J., 1987, in Physical Processes in 
Interstellar Clouds , eds. Morfill, G.E. and Scholer, M. (NATO ASI Series C210, 
Reidel, Dordrecht), 333 

Tielens, A.G.G.M. and Hagen, W. , 1982, Astron. Astrophys, 114, 245 

van IJzendoorn, L.J., Allamandola, L.J., Baas, F., Komig, S. , and Greenberg, J.M., 
1986, J. Chem. Phys . 85, 1812 

van IJzendoorn, L.J., Allamandola, L.J., Baas, P., and Greenberg, J.M., 1983, J. 
Chem. Phys. 78, 7019 

van IJzendoorn, L.J., 1985, Ph.D. Dissertation, Leiden University, The Netherlands 

van IJzendoorn, L.J., Allamandola, L. J. , de Groot, M.S., Baas, F. , van de Bult, 
C.E.P.M., and Greenberg, J.M. , 1990, J. Phys. Chem., submitted 

van der Zwet, G.P., 1986, Ph.D. Dissertation, Leiden University, The Netherlands 

Wopenka, B. , 1988, Earth and Planet. Sci. Lett. 88, 221 



L. Colangeli 

University of Cassino 


E. Bussoletti 

Istituto Universitario Navale 

Naples, Italy 

Osservatorio Astronomico Capodimonte 
Naples, Italy 



L. Colangeli (^) and E. Bussoletti C^'^) 

(^) University of Cassino, Italy; 

(^) Istituto Universitario Navale, Naples (Italy); 

(^) Osservatorio Astronomico Capodimonte , Naples (Italy) 


The investigation of comets has preceded for long time on remote observations 
from ground. In 1986 several space missions towards comet Halley have allowed, for 
the first time, to have a close look to a comet (Encounters with comet Halley 1986). 
In particular, the GIOTTO mission by the European Space Agency (ESA) has provided "in 
situ" observations and measurements up to a distance of about 600 Km from the 
nucleus. Surface morphology and physical properties have been observed; plasma, gas 
and dust components in the coma have been analyzed (Grewing et al. 1987). It is 
clear, however, that definite answers about the primordial nature of comets and their 
relation with interstellar material can be obtained only from direct analysis of 
cometary samples. Future space missions such as CRAF (NASA) and ROSETTA (ESA) have 
exactly this aim. In particiilar, the ambitious goal of Rosetta mission is to return 
to earth comet samples which can be analyzed carefully in laboratory. 

In preparation to this event a large effort must be placed both in the 
improvement of existing analytical techniques and in the development of new methods 
which will provide as much information as possible on "returned comet samples" 
(hereinafter RCSs) . Handling of extra-terrestrial samples will require to operate in 
carefully controlled and extremely "inert" ambient conditions. In addition, working 
on a limited amount of "unique" cometary material will also impose to use analytical 
techniques which should not produce alteration, contamination or destruction of the 
sample. Many suggestions can come from people working in laboratory on "cosmic dust"; 
in fact, experimental methods which are applied to analyze a) interplanetary dust 
particles (IDPs) collected in stratosphere, b) meteorites , and c) laboratory produced 
cosmic dust analog samples, can be mutuated or properly improved in the future for 
specific application to RCSs. 

Since modem techniques used to analyze IDPs and meteorites are reviewed 
elsewhere in this workshop (see for instance the contributions from J. Bradley and 
A. Albee), we will discuss some of the most powerful techniques which are presently 
applied to characterize physical and chemical properties of micron and/or submicron 
solid grains, synthetized in laboratory with the aim of simulating cosmic dust. 


We will focus here attention mainly on carbonaceous materials, because it is 
commonly accepted that: a) about half of interstellar dust is in various forms of 
carbonaceous compounds; b) about 70 % of "small" grains detected around P/Halley's 

(*) Work performed under contracts MPI 120111 (40 %) and 120257 (60 %) , 
CNR 8800361-02, PSN 88-020, and ASI 88-060. 


nucleus were carbonaceous (Jessberger et al. 1988); c) IDPs contain significant 
fractions of carbon. According to the results of the "working group on carbon" 
(Huffman 1988a) presented at the international workshop on "Experiments on cosmic 
dust analogues" held in Capri (Bussoletti et al. 1988), we have to bear in mind that 
"carbon is an extremely complicated and variable material" and in nature - as well as 
in space and in laboratory experiments - it can occur in various forms as, for 
example: disordered graphitic carbon, diamond-like carbon, ""cyclic and acyclic 
molecules, polycyclic aromatic hydrocarbons (PAHs). 

Actually, a simplified way to classify different kinds of carbon is to consider 
two main parameters: a) the internal ordered-domain length, b) the size. The first 
parameter determines the largest domain length of the lattice over which order is 
mantained; in this scale, graphite (sp^ hybridization) and diamond (sp^ 
hybridization) represent one of the limits since they are perfectly crystalline 
materials and the length over which the lattice structure repeats itself is virtiially 
infinite. On the other side of the scale, amorphous carbon represents a material with 
a complete internal disorder so that no regular lattice structure can be identified 
and the ordered domain size is virtually zero. Of course, we expect that any material 
formed in nature, as well as in space and in laboratory conditions, never reaches one 
of these two extreme configurations, except under very pectiliar formation conditions 
(Marchand 1987). Therefore, in general, we deal with polycrystalline materials, soots 
and/or more or less "disordered" lattice structures. This means that the term 
"amorphous carbon", commonly used to identify laboratory produced carbon grains, must 
be rather interpreted as "internally very disordered carbon". The other parameter we 
have considered is just the dimension of the grains or molecules. For very small 
sizes (< few A) we fall in the field of molecules as, for example, PAHs. As soon as 
the size increases, we talk of submicron or micron grains but, in some cases, 
agglomeration of grains may produce larger "clumps" . The progression from molecirles 
to grains has assumed today a particular meaning since PAH molecules in space are 
probably formed by the destruction of carbonaceous grains and/or are the leftover 
nuclei of condensation processes to form carbon grains, for example arovind carbon 
rich stars (Allamandola et al. 1985). 

The techniques most frequently used to produce "amorphous carbon" submicron 
and/ or micron grains in laboratory are summarized in table I. In the following we 
will critically discuss various experimental methods used to perform physico-chemical 
investigation of grain samples and we will present recent results obtained on 
various kinds of candidate materials for cosmic dust. 


Dealing with powdered samples, the only way to get clear information about 
morphology of single micron/ submicron grains is to use electron microscopy technique. 
Transmission electron microscopy (T.E.M.) allows to determine the size and shape of 
single grains, when used at high magnification (about 400,000 X), and size 
distribution of the samples by measuring the size of a large nximber (> 1000) of 
single particles on photographs recorded at mediiim magnification (about 50,000 X). 
Scanning electron microscopy (S.E.M.), on the contrary, is a powerful tool to 
determine surface properties of samples. However, we have to recall that SEM utilizes 
the electron beam reflected by the sample so that the intensity of the recorded 
signal is proportional to the atomic niunber, Z, of the examined material. This means 
that it is rather difficiilt to get surface details of submicron carbon grains (Z=6), 
while SEM is efficient for silicate particles (Z=12), and - in any case - for large 
particles rather than very small ones. Furthermore, TEM often requires the deposition 
















Focusing of high 

Striking an arc 

Benzene /xylene 

Quenching of an 

power pulsed-laser 

between two 

burning in 


beams on a target 

amorph. carbon 


plasmic gas in 

of bulk pure 

or graphite 


material in Ar 

electrodes in 


Ar atmosphere 

Stephens 1980 

Koike et al. 

Day and Huffman 

Sakata et al. 

1980; Bussoletti 

1973; Bussoletti 


et al. 1987 

et al. 1987 

of the sample onto special micro-grids covered with holey-carbon films. This method 
is clearly non destructive, but it appears rather complicate to retrive the sample 
after TEM analysis since the substrate-film must be removed by using fluids which may 
contaminate the sample. On the other hand, TEM can be performed also on thin sections 
(500 - 1000 A) of grains. In this case a single grain is embedded in epoxy material 
and cutted by a diamond knife into submicron slices. Although destructive, this 
method allows to perform a 3-dimensional analysis of the physical grain structure by 
studying various slices of the same grain (Bradley and Brownlee 1986). In the case of 
SEM, the sample can be simply mounted onto a metallic substrate allowing for an 
easier retrival of the non-destructed sample. Both TEM and SEM investigations are 
performed in high vacuiun conditions and the sample holder can be adapted to a 
cold-finger. This ensures "clean" working conditions and sample thermal control, 
which appear very appropriate for RCSs. On the other hand, the electron beam working 
conditions must be carefully selected in order to avoid the release of high energy 
onto the sample: structure alterations are possible for organic and polymeric 

In figures 1 to 3 some examples of TEM images obtained on various samples 
produced in laboratory are reported. In figure 1 chain-like and fluffy structures of 
AC amorphous carbon grains formed during condensation are evident. At high 
magnification the properties of single AC grains are evidenced and typical spheroidal 
shape appears (fig. 2). In the case of industrially produced silicon carbide (a-SiC) 
the grains appear mainly irregular (fig. 3). Examples of size distributions for AC 
and SiC grains are shown respectively in figures 4 and 5. They are obtained by 
measuring the maximum elongation of more than 1000 grains on various TEM photographs 
for each sample. 

Electron microscopy allows also to investigate the internal structure of the 
material, when used in diffraction mode. Typical diffraction patterns of our a-SiC 
show the presence of isolated spots, suggesting a regular lattice structure, though 
their disordered arrangement in the pattern indicates that grains are mainly 
polycrystalline. The diffraction pattern for AC samples does not present any spot but 
only slightly contrasted concentric dark and light rings, according to a "nearly 


Figure 1.- TEM photograph of AC amorphous carbon sample. Chain-like and fluffy 
structures are evident. 

"*. ■ Jr. 

7M^.^. •^■iffi. 



Figure 2.- High magnification TEM image of a typical AC grain (maximum elongation 
s: 200 A). 





--. $ 

















ij • 

u -1-1 

= a) 

S -H 

Cd > 

H <D 





1 1 > 1 1 1 y 









1 I 1 





Figure 4. 

Grain size distribution of AC sample obtained b>- measuring the inaxiniuin 
eloneation of more than 1000 particles on TEM photographs . 

2.0 4.0 6.0 2.0 

d {\im) 

Figure 5.- Grain size distribution of SiC grains; 


4.0 6.0 

a) raw material; b) 6 h ground 

ajnorphous" structure. We remind here that X-ray diffraction technique is an 
alternative method to get information similar to that obtained by electron 


This spectroscopic technique is generally considered complementary to IR 
transmission spectroscopy (see section 5), since different vibration modes may be 
active in the two cases. This is generally true for highly simmetric molecules 
characterized by an inversion centre of simmetry, meanwhile Raman and IR spectra can 
be very similar in the case of "disordered" non-symmetric materials. However, Raman 
spectroscopy is important not only to characterize optical properties but also in the 
clarification of internal structure of the materials. In particular, carbonaceous 
materials usually show two main first-order Raman scattering features at about 1600 
and 1350 cm~^ (Tuinstra and Koenig 1970; Rosen and Novakov 1978). The first band is 
assigned to Ea^ mode in single graphite crystallites (Tuinstra and Koenig 1970), 
while the attribution of the second feature is more controversial. In fact, this 
signature can be interpreted as a resonance in diamond-arranged C atoms (Tuinstra and 
Koenig 1970) or in rhombohedral graphitic polytype (Wieting and Verble 1979), but the 
possibility exists that it is associated with Ai^ mode in disordered graphite lattice 
(Tuinstra and Koenig 1970; Robertson 1986). Acttially, the exact position of the 
1600 cm~^ peak seems to depend on the size of crystallites: the smaller the crystal 
size the larger the wavenvunber. At the same time, the intensity ratio, 
1(1355 cm~^)/I(1600 cm~^) increases as : i) the amount of "unorganized" carbon in the 
sample increases, ii) the graphite crystal size in the material decreases (Tuinstra 
and Koenig 1970) . 

Recently, Allamandola et al. (1987) have obtained Raman spectra of various IDPs, 
all showing the two mentioned features. Their position and relative intensity suggest 
that the carbonaceous component of IDPs is arranged in aromatic sub-units with 
ordered domains no larger that 25 A, bridged by short aliphatic bonds. Furthermore, 
it has been evidenced that the IR emission spectrum of the Orion Bar nicely resembles 
the Raman spectrum of the examined IDPs (Allamandola et al. 1987) and that of 
"auto-exhaust" solid grains (Allamandola et al. 1985). 

Blanco et al. (1988a, 1989) have recently performed a systematic Raman analysis 
of various "carbonaceous materials" produced in laboratory. In the experimental 
set-up an Ar* laser tuned at 5145 A has been used as a source and the power incident 
on the sample is about 200 mW. In figure 6 examples of Raman spectra on amorphous 
carbon grains show the two main features at 1355 and 1600 cm~^, suggesting a 
structural similarity with IDPs and interstellar dust grains. In the experiments, 
various sample configurations have been tested. In particular, both deposition of 
dust on glass substrate and embedding in KBr matrix allow to obtain good signal to 
noise ratios in the spectra. Both these techniques seem suitable for RCS analysis. 
However, the deposition on clean glass substrates guaranties uncontamined 
preservation of the sample and prevents from risks of mechanical stresses, which are 
possibly produced during embedding processes in matrixes. 

One relevant problem with Raman spectroscopy is the high thermal stress produced 
by the laser beam impinging on the sample. This effect can be partially reduced by 
mounting the sample on special holders which rotate during laser illumination. 
However, this solution presents the disadvantage that the monitored spectrum gives an 
information averaged over the area spanned during the laser scanning; this drawback 
can be critical for a non homogeneous material, as actually it is expected for RCSs. 


glass suppo 

wavelength ( ^m) 

7 8 9 10 







wavenumber(cm ) 

Figure 6.- Raman spectra of AC (a) and BE (b) samples embedded in KBr pellets and 
deposed on glass substrates. 



AbsoiTJtion, scattering and emission spectroscopy extended over the widest 
wavelength range from vacuum ultraviolet (VUV) to far infrared (FIR) is extensively 
used in laboratory to identify the optical properties of materials candidate to be 
present in space. Especially VUV and near IR portions of the spectrum appear 
particularly useful to diagnostics, because they contain "fingerprints" of specific 
chemical compounds. Their detection in laboratory and in astronomical spectra can be 
helpful in the classification of cosmic materials. However, when one deals with 
"dusty" samples in laboratory a certain number of practical problems arise; here we 
will mention some of the most crytical and still difficult to solve in present days. 

5.1 Clustering effects 

Most of the dust samples produced in laboratory occur in clumped agglomerates 
(see section 3). This effect may strongly affect optical measurements, especially at 
wavelengths where "surface modes" are active (Bohren and Huffman 1983). Therefore, 
the deduction of optical properties for submicron/micron grains from extinction data 
is not always straightforward. From a theoretical point of view, Huffman (1988b) has 
shown that in many cases clustering can be properly treated by considering a "shape 
distribution" of randomly oriented ellipsoids in the Rayleigh limit. However, 
clustering is expected to become significant only at wavelengths where the optical 
constants (n,k) of the material are sufficiently high. According to Hanner (1988), 
for glassy carbon this is true at /i > 100 yun. The importance of studying isolated 
solid grains is furtherly confirmed by the agreement existing between optical 
constants measured from glassy carbon films (Edoh 1983) and from single levitating 
grains (Pluchino et al. 1980). 

5.2 Size effects 

The optical properties of submicron-micron particles are also sensitive to the 
actual dimension of the grains. Again, this effect is more significant at wavelengths 
where "surface modes" are active (Bohren and Huffman 1983). A typical example is 
given by the behaviour of the 11.5 um band, characteristic of SiC grains. The 
absorption spectra recorded for a-SiC grains after differentiated grinding, washing 
in ultrasonic chamber and sedimantation in acetone are compared in figure 7. The 
profile of the band changes according to the average size of the grains as it is also 
evidenced from the data collected in table II. We note that, as the size reduces, the 
main peak becomes sharper and more intense, while the actual peak position shifts 
towards shorter wavelengths. These results fit rather well with the theoretical 
predictions on surface resonances, confirming that optical properties of dust samples 
must be correlated with their morphology (Borghesi et al. 1985). 

5.3 Homogeneity of data 

A mandatory requirement in laboratory analysis of cosmic dust analogues is to 
produce homogeneous spectroscopic data of the same sample, under controlled ambient 
conditions and all over the widest spectroscopic wavelength range from VUV to FIR. In 
the case of RCSs, this requirement must be also coupled with a particular care in 
minimizing handling and treatment processes which may lead to a possible 
deterioration of the material. 


Figure 7.- Normalized extinction for a-SiC grains. The curves refer respectively 
to raw material (a), 6 h ground sample (b) , and 6 h ground + 30 min US 
washed + 2 h sedimented grains (c). 



[From Borghesi et al. 1985] 












10.7 12.2 13.2 
10.7 12.0 13.0 
10.7 11.9 12.8 




G: 6h ground; GWS: 6h ground + 30 min US washed + 2h sedimented in acetone; 
LM: longitudinal lattice vibration mode; TM: transverse lattice vibration 
mode; MP : main peak; Tx = 1(MP) / I(continuum) ; Tz = I(MP) / I(LM); 
HW: half -height band width. 

In the laboratory work performed at Lecce and Naples, the full range from 
1000 A to 1 mm is covered, by using various experimental techniques and 
instrumentations as it is summarized in table III. In figure 8 the complete spectra 
of various forms of amorphous carbon grains are reported. A well pronounced peak is 
evident at 235 - 250 nm in all spectra. The peak for smaller AC grains (size = 80 A) 
falls at shorter wavelength and appears sharper than that observed for larger BE/XY 
particles (size = 300 A). In both cases the hump is attributed to a plasmon mode of 
surface electrons in the ground state. This is also confirmed by the absence of any 
absorption feature beyond 300 nm, where the extinction curves fall as /i "", with 
a 5= 1. Only in the 2 - 13 pm range some weak features are detected. Most of them are 
interpreted as due to both C=C skeletal vibrations and CH„ (n=l,2,3) radicals bound 
to active sites onto the grain surface. For more details about these features we 
refer to Borghesi et al. (1987) and Blanco et al. (1988b). 

At this point we have to stress that our spectroscopic results have been 
obtained by using "classical" experimental techniques, which are based either on the 
embedding in trasparent matrix or on the deposition onto window-substrates of the 
material (see table III). These methods have the limitation of leaving vinsolved the 
mentioned problem of clumping and sometimes may imply risks of contamination or 
alteration, especially in the view of application to RCSs. 


In the recent years new experimental techniques have been tested which could be 
succesfully applied also to RCSs, partially solving the problems found by using 
"classical" methods (see section 5.3). A substancial improvement could come from the 
suspension of "individual" isolated grains for optical analyses. Among other 
methods, we recall: a) laser levitation (Ashkin 1970, Ashkin and Dziedzic 1980) 
b) quadrupole trapping (Philip et al. 1983), c) electrostatic levitation (Marx and 
Mulholland 1983, Weiss-Wrana 1983, Giese et al. 1986). 

An example of optical measurements performed on levitating particles is reported 
by Pluchino et al. (1980). In this case a carbon particle with diameter of about 1 \m 
is electrostatically suspended and its position monitored by detecting the scattered 
light from an Ar* laser. The signal drives a servo-system that keeps the particle 
levitating always in the same position. This allows scattering measurements from 10* 
to 170® by moving the detector around the suspension chamber. Recently, Stephens 
(1988) has performed extinction and scattering measurements on carbon grains 



10" 10 




Figure 8.- The mean extinction efficiency curve for AC, BE, and XY amorphous carbon 
samples . 





1000 A 3000 A 2.5 urn 50 urn 

2000 A 2.6 pm 30 urn 

1 mm 


Synchrotron light 
facilities of 
"Adone" (Frascati) 
"BESSY" (Berlin) 



P. E. 580/880 




Deposition on 

Embedding in 



KBr or Csl 


vaporized by laser focusing (see table I) onto bulk graphite target in a floating 
chamber. The submicron (100 - 500 A) particles levitate in the chamber without any 
suspension system for a time lag sufficient to record the spectrum in the 200 - 1200 
nm range by means of an optical multichannel detector. 

Another technique which could present advantages, when applied to RCSs, is the 
so-called "Infrared Photothermal Beam Deflection Spectroscopy" (Low and Morterra 
1983). As it is shown in their figure 2, the sample is deposited on a metallic 
substrate in an isolated chamber and it is heated by IR radiation coming from an 
interferometer. If the IR radiation flux is modulated in time, also the thermal 
gradient and the consequent refractive index gradient in the medium over the surface 
are modulated in time. When a He-Ne laser beam is passed over the surface, it is 
deflected and modulated, so that the detector reveals a "photothermal interf erogram" . 
This can be processed as in a conventional Fourier transform spectrometer to get the 
final spectrum. The main advantage of this technique is that spectroscopy can be 
performed avoiding any interaction of the sample with the external ambient and no 
particular sample preparation is required. On the other hand, the sample must be 
heated by the IR beam, and this effect coxild produce xindesired chemical and strucutre 


Allamandola, L.J. ; Tielens, A.G.G.M. ; and Barker, J.R. : Polycyclic Aromatic 
Hydrocarbons and the Unidentified Infrared Emission Bands: Auto Exhaust along 
the Milky Way !, Astrophys. J. (Letters), vol. 290, 1985, pp. L25-L28. 

Allamandola, L.J. ; Sandford, S.A. ; and Wopenka, B.: Interstellar Polycyclic Aromatic 
Hydrocarbons and Carbon in Interplanetary Dust Particles and Meteorites. 
Science, vol. 237, 1987, pp. 56-59. 

Ashkin, A.: Phys. Rev. Lett., vol. 19, 1970, p. 283. 

Ashkin, A.; and Dziedzic, J.M. : in "Light Scattering by Irregularly Shaped 
Particles", ed. D.W. Schuerman (New York: Pleniom Press), 1980, p. 233. 

Blanco, A.; Borghesi, A.; Bussoletti, E. , Colangeli, L. ; De Blasi, C. ; Fonti, 5.; 


Fusco, C. ; Orofino, V.; and Schwehm, G. : Amorphous Carbon and Carbonaceous 
Materials in Space; Part I: Laboratory Measurements. Nuovo Cimento, in press, 

Blanco, A.; Bussoletti, E. ; Colangeli. L. ; Fonti, S.; and Orofino,. V. : Raman Spectra 
of Submicron Amorphous Carbon Grains and Mixtures of Polycyclic Aromatic 
Hydrocarbons. Infrared Fhys . , vol. 28, 1988a, pp. 383-388. 

Blanco, A.; Bussoletti, E. ; and Colangeli, L. : A Mixture of Hydrogenated Amorphous 
Carbon Grains and PAH Molecules: a Candidate for the Unidentified Infrared 
Bands?. Astrophys. J., vol. 33 A, 1988b, pp. 875-882. 

Bohren, C.F. ; and Huffman, D.R. : Absorption and Scattering of Light by Small 
Particles. John Wiley & Sons, 1983. 

Borghesi, A.; Biissoletti, E. ; Colangeli, L. ; and De Blasi, C. : Laboratory Study of 
SiC Submicron Particles at IR Wavelengths: a Comparative Analysis. Astron. 
Astrophys., vol. 153, 1985, pp. 1-8. 

Borghesi, A.; Bussoletti, E.; and Colangeli, L. : Amorphous Carbon and the 
Unidentified Infrared Bands. Astrophys. J., vol. 314, 1987, pp. 422-428. 

Bradley, J. P. ; and Brownlee, D.E. : Cometary Particles: Thin Sectioning and Electron 
Beam Analysis. Science, vol. 231, 1986, pp. 1542-1544. 

Bvissoletti, E. ; Colangeli, L. ; Borghesi, A.; and Orofino, V.: Tabulated Extinction 
Efficiencies for Various Types of Submicron Amorphous Carbon Grains in the 
Wavelength Range 1000 A - 300 pm. Astron. Astrophys. Suppl. Ser. , vol. 70, 1987, 
pp. 257-268. 

Bussoletti, E. ; Fusco, C. ; and Longo, G. : Experiments on Cosmic Dust Analogues. 
Astrophys. Space Sci. Library, vol. 149 (Dordrecht: Kluwer Academic Publishers), 

Day, K.L. ; and Huffman, D.R. : Measured Extinction Efficiency of Graphite Smokes in 
the Region 1200 - 6000 A. Nature Phys. Science, vol. 243, 1973, pp. 50-51. 

Edoh, 0.; Optical Properties of Carbon from the Far Infrared to the Far Ultraviolet. 
Ph.D. dissertation, Dept. of Physics, Univ. of .Arizona, 1983. 

Encounters with Comet Halley, the First Results. Nature, vol. 321, 1986, pp. 259-366. 

Giese, R.H. ; Killinger, R.T. ; Kneissel, B.; and Zerull, R.H. : Albedo and Colour of 
Dust Grains: Laboratory versus Cometary Results. 20th ESLAB Symposium on the 
Exploration of Halley 's Comet, vol. II, ESA SP-250, 1986, pp. 53-57. 

Grewing, M. ; Praderie, F. ; and Reinhard, R. : Exploration of Halley's Comet. Astron. 
Astrophys., vol. 187, 1987. 

Banner, M. : Grain Optical Properties. "Infrared Observations of Comets Halley and 
Wilson and Properties of the Grains", ed. M. Banner, NASA CP-3004, 1988, 
pp. 22-49. 

Huffman, D.R. : Report of the Working Group on Carbon. "Experiments on- Cosmic Dust 
Analogues", ed. E. Bussoletti, C. Fusco, and G. Longo (Dordrecht: Kluwer 


Academic Publishers), 1988a, pp. 3A5-347. 

Huffman, D.R.: Methods and Difficulties in Laboratory Studies of Cosmic Dust 
Analogues. "Experiments on Cosmic Dust Analogues", ed. E. Bussoletti, C. Fusco, 
and G. Longo (Dordrecht: Kluwer Academic Publishers), 1988b, pp. 25-42. 

Jessberger, E.K. ; Christoforidis, A. ; and Kissel, J. : Aspects of the Major Element 
Composition of Halley's Dust. Nature, vol. 332, 1988, pp. 691-695. 

Koike, C. ; Hasegawa, H. ; and Manabe, A.: Astrophys. Space Sci., vol.67, 1980, p. 495. 

Low, M.J.D.; and Morterra, C. : IR Studies of Carbons - I. Carbon, vol. 21, 1983, 
pp. 275-281. 

Marchand, A. : Various Kinds of Solid Carbon: Structure and Optical Properties. 

"Polycyclic Aromatic Hydrocarbons and Astrophysics", ed. A. Leger, L. 

d ' Hendecourt , and N. Boccara (Dordrecht: D. Reidel Publishing Company), 1987, 
pp. 31-54. 

Marx, E. ; and Mulholland, G.W. : J. Res. Nat'l. Bur. St., vol. 88, 1983, p. 321. 

Philip, M.A. ; Gelbard, F. ; and Arnold, S.: J. Colloid and Interface Sci., vol. 91, 
1983, p. 507. 

Pluchino, A.B, ; Goldberg, S.S.; Dowling, J.M. ; and Randall, CM.: Refractive-index 
Measurements of Single Micron-sized Carbon Particles. Applied Optics, vol. 19, 
1980, pp. 3370-3372. 

Robertson, J.: Amorphous Carbon. Advances in Phys., vol. 35, 1986, pp. 317-374, 

Rosen, H. ; and Novakov, T. : Identification of Primary Particulate Carbon and Siilfate 
Species by Raman Spectroscopy. Atmos. Environ., vol. 12, 1978, pp. 923-927. 

Sakata, A.; Wada, S.; Okutsu, Y. ; Shintani, H. ; and Nakada, Y. : Does a 2,200 A Hxmp 
Observed in an Artificial Carbonaceous Composite Accoiint for UV Interstellar 
Extinction?. Natxire, vol. 301, 1983, pp. 493-494. 

Stephens, J.R. : Visible and Ultraviolet (800-130 nm) Extinction of Vapor- condensed 
Silicate, Carbon, and Silicon Carbide Smokes and the Interstellar Extinction 
Curve. Astrophys. J., vol. 237, 1980, pp. 450-461. 

Stephens, J.R.: Light Scattering from Simulated Interstellar Dust. "Experiments on 
Cosmic Dust Analogues", ed. E. Bussoletti, C. Fusco, and G. Longo (Dordrecht: 
Kluwer Academic Publishers), 1988, pp. 245-252. 

Tuinstra, F. ; and Koenig, J.L.: Raman Spectrum of Graphite. J. Chem. Phys., vol. 53, 
1970, pp. 1126-1130. 

Weiss-Wrana, K. : Optical Properties of Interplanetary Dust: Comparison with Light 
Scattering by Larger Meteoritic and Terrestrial Grains. Astron. Astrophys., vol. 
126, 1983, p. 240. 

Wieting, T.J.; and Verble, J.L. : "Electrons and Phonon in Layered Crystal 
Strucutres", ed. T.J. Wieting and M. Schluter (Dordrecht: D. Reidel Publishing 
Company), 1979, pp. 321-407. 




K. Roessler 

P. Hsiung 

M. Heyl 

Institut fiir Chemie 1 

Kemforschungsanlage Jiilich 

JiiUch, FRG 

G. Neukum 

A. Oehler 

Institut fiir Optoelektronik 


WeBling, FRG 

H. Kochan 

Institut fur Raumsimulation 


Koln, FRG 



K. Roessler^, P. Hsiung^, M. Heyl^, G. Neukum^, A. Oehler^, and H. Kochan^ 

^Institut fur Chemie 1, Kernforschungsanlage Julich, D-5170 JQIIch, FRG 
^Institut fur Optoelektronik, DLR-Oberpfaffenhofen, D-8031 WeBling, FRG 
^Institut fur Raums inflation, DLR-Koln, D-5000 K61n-90, FRG 

The preparation and analysis of frozen volatiles at temperatures < 77 K via condensation 
and optical spectroscopy, resp., is traditionally performed in cryostats in situ such as 
described in (Pross L. et al., 1976; Biel E. et al., 1988). The study of larger ice 
samples, in particular to simulate processes in icy material of the solar system such as 
comets necessitated a somewhat different approach, cf. e.g. (Dobrovolsky O.V. et al., 
1977; Ibadinov K.I. et al., 1987; Saunders R.S. et al., 1986; Storrs A.D. et al., 1988). 
Preparation, the proper study in a cooled sample holder, and analysis of the states 
before and after the experiment have often to be performed at different sites, thus, 
necessitating cryotransport and handling of the samples under protective conditions, i.e. 
inert and cold atmosphere or vacuum. This became even more stringent for the comet 
simulation experiments performed since 1987 by a team of scientists from different 
disciplines in the Big Space Simulator in the German Aerospace Research Establishment 
DLR-Koln (Bischoff A. et al., 1988; Grun E. et al., 1987, 1988, 1989; Klinger J. et al., 
1989; Kochan H. et al., 1989; Roessler K. et al., 1988, 1989; Spohn T. et al., 1989; 
Thiel K. et al., 1989). Three experiments have been performed with water ice and water- 
COp-ice mineral dust mixtures in April 1987 (KOSI-I), April 1988 (KOSI-II), and November 
1988 (KOSI-III). A fourth experiment is prepared for May 1989 (KOSI-IV). These 
experiments deal with icy material of approx. 10 1 volume and 4 to 5 kg weight, each for 
a standard and the proper sample to be irradiated several ten hours with an artificial 
sun of approx. 1-3 SC. Techniques applied for cryohandling, -transport, -storage and 
-analysis will be reported here. They may be considered as first tentative steps in view 
of the development of methods for treatment of icy samples from space brought to earth 
via return missions such as ROSETTA. 


The temperature of comets may range from 80 to 250 K and more at the surface. Thus, 
liquid nitrogen can be considered as a convenient cooling agent for sample preparation 
and handling. Only for special examinations, such as high resolution spectroscopy or 
condensation of CO, CH^ and N2, temperatures lower than 77 K are needed. Table 1 gives a 
list of temperatures for some critical processes. From Table 1 it can be seen that dry 
ice-methanol cooling baths will not be sufficient for studies with volatile componer.ts 
such as CO2. Besides cooling liquids also electrical methods (Peltier-effect) can be 
applied. However, their heat capacity is in general low, and samples cannot be easily 


Conventional bath cryostats with cold fingers (5 or 77 K) such as used for low 
temperature spectroscopy in general do not allow manipulations at the chilled sample nor 
is it possible to introduce an already preexisting ice sample. This can be achieved 

This work is financially supported by Deutsche Forschungsgemeinschaft DFG 


Table 1: Temperatures in K (rounded values) for processes in ices 

4 liquid helium evaporation 

10-14 migration of H starts in H2O ice ■ 

< 72 evaporation of CO at 1 bar - ! 

77 evaporation of nitrogen at 1 bar 

80-120 amorphous to cubic transition in NH3 ice 

100-110 migration of OH starts in HoO ice 

110-130 migration of O2H starts in R2O ice 

135±10 amorphous to cubic transition in H2O ice 

130-150 sublimation of CO2 in vacuum 

165±10 cubic to hexagonal transition in H2O ice 

195 sublimation of CO2 at 1 bar 

195 dry ice/methanol cooling 

196 melting of NH3 

240 evaporation of NH3 at 1 bar 

< 273 melting of H2O with impurities 

273 melting of pure H2O at 1 bar 

by a removable plug system positioned around the cold finger and connected tightly to the 
cryostat body via two rubber-0-rings (Pross L. et al., 1975). In the version shown in 
Fig. la it is provided with optical windows for spectroscopy. It can be removed from the 
cryostat after mounting onto an evacuated cross tube by means of a push-pull fead 
through. The plug can even be separated from the rest by a valve system, and can be 
exchanged for another plug (e.g. change from quartz to KBr windows when going from VIS to 
IR spectroscopy). The sample hangs free on the cold finger and is accessible to 
manipulation. After processing the plug is remounted and the cryostat can be transported 
to another site. In order to load samples from the outside, a glove box system with No 
purging gas allowed to open the cold cryostat without giving atmospheric moisture or CO2 
a chance to condense onto sample and walls. Fig. lb. Special window plugs containing gas 
nozzles were developed for simultaneous condensation of ices on KBr cold plates and 
optical spectroscopy in transmission (Biel E. et al., 1988; Roessler K., Eich 6. et al., 


For transport and storage of closed samples liquid nitrogen containing dewars are best 
suited. The material of the sample containers should be relatively thick. It should 
possess high heat capacity in order to allow short removal of the samples from the bath. 
For the same reason all tools by which samples are touched should be made out of metal 
and chilled before use by immersion into liquid nitrogen. They should be provided with 
insulating handles in order to prevent heat flow from the hands of the operator. 
Furthermore, thermally insulating gloves should be worn. 

For transport and handling of open samples cold N2 gas evaporating from liquid N2 baths 
can be used as a protective medium. Fig. 2a shows the tank used in KOSI experiments. It 
is made out of stainless steel (2 mm) and it is deliberately not a dewar in order to 
induce vigorous evaporation of N2 gas. The KOSI samples (30 cm diameter and 15 cm high) 
are stored in the cylindrical recipient which is surrounded by liquid nitrogen. The cold 
N2 gas (130-150 K) protects the surface of the sample from thermal effects (e.g. sub- 
limation of COp) and condensation of atmospheric HoO or CO2. This effect is improved by a 
steel cover, which is kept cold by heat conductivity. The irradiated KOSI samples and the 
standards are stored and transported in this tank into the central glove box for further 


UOMO H» r u iP wj g 



= Nj out 



lock for htTo- 
dudnq samples 

\ openings for 

p N, in 

removable plug (e.g. with ophcoi wmdows) 

PW«A put 1wm« t*wau9i 

Fig. 1: Bath cryostat with removable (window) plug mounted on an evacuated cross tube (a) 
and glove box system to load icy samples onto the cold finger of the cryostat (b) 

top cover 

N2 gcs outlet 
and Iq. Nj refill 

2mm steel 

J ^ liquitf K] 




Fig. 2: Stainless steel tank with liquid nitrogen and the KOSI sample recipient (a) 
and central glove box (80x80x80 cm) with Iq. N2 bath and windows for optical 
reflexion spectroscopy at three angles (b) 



Fig. 2b shows the 80x80x80 cm glove box used for KOSI experiments. It is made out of 1 cm 
thick plexiglass and contains several holes for gloves (made out of rubber with 
insulating layers), quartz windows for optical spectroscopy from the outside (remote 
sensing), a lock system for samples and tools (not shown in Fig. 2b), and many electric 
connections. More effective than purging the box by dry No gas (5.0) is to use the N2 
evaporating from the bath itself. An overpressure of 1.1 bar has to be maintained to 
hinder penetration of atmospheric gases. The relatively steep temperature gradient to the 
walls prevents freezing from the outside, except for the lower regions of the glove box 
where temperatures of 200 to 220 K were measured at the outer walls. This, however, has 
the advantage that sample containers, tools, etc. which were stored here, were already 


The analysis of the non- irradiated standard and the irradiated KOSI samples consists of 
five steps: 

1) General visual inspection of surface and morphology during sample taking and deepening 

of pits (Fig. 3, for KOSI-III). 

2) Mesurement of temperature. Fig. 4 shows the temperatures of KOSI-III sample after 

insertion into the liquid N2 tank measured by three thermocouples (1 cm, 5 cm and 
11 cm below the actual surface) during the operations in the glove box. Temperatures 
are reasonably lew to prevent sublimation of COo- 

3) Optical reflexion spectroscopy in VIS and IR (4^ to 2500 nm) in remote sensing, in 

particular to determine the Albedo, with Barnes field and similar spectrometers. Fig. 
5 shows the general arrangement around the glove box. Fig. 6a shows the sample before 
irradiation, Fig. 6b the irradiated sample, both with the blank standard. Fig. 7 
exhibits the spectra obtained for KOSI-III (average over three angles). It can be see 
that the sample under irradiation became slightly darker and that H2O and CO2 peaks 
disappeared at least partly. 

4) Test of material strength by drilling boreholes with a penetrator provided with a 

Newtonmeter (ESA/ESTEC, Dr. G. Schwehm). The most important result is the detection 
of crusts formed under the loose dust layer. The steep rise and fall of the force 
indicates the thickness of the layers. Fig. 8a, b are from KOSI-II. The results for 
KOSI-III are reported in (Thiel K. et al. 1989). 

5) Measurement of electrical conductivity of surface and bulk (ESA/ESTEC, 

Dr. G. Schwehm), planned for KOSI-IV. 


Samples were taken from non-irradiated standard and KOSI material for following analyses. 
The first 6 methods require very cold sample taking in order to maintain their specific 

1) CO2 content via collecting gases evolving upon heating of samples to ambient 

temperature and gas chromatography. Fig. 9 gives the CO2 profiles for KOSI-III; 
further details in (Roessler K., Hsiung P. et al. 1989). The starting material 
contained 13.8 wt. % CO2, the non-irradiated standard did not show severe losses upon 
filling, except for the surface layer. After 41 h irradiation at 1.3-2.7 SC artifi- 
cial sunlight, CO2 was almost totally lost from the upper and middle layers. Only nea 
the cooling back plate some CO2 remained. A very CO2 rich white layer (1-2 mm) was 
observed on the back plate itself indicating inward diffusion of COo- 

2) Investigation of crystal structure via X-ray diffraction (in cooperation with 
Prof. E. Mayer, Univ. Innsbruck, Austria). 





former surface _ver^Jhin_dust_la^r(cQ_2jnm2 ac 

~ " "* special 

evaporation effect by 
illuminated walls 

light gray ice crust ^s - 10 


homogeneous gray matter 
white ice byer on backplote ( CO2 rich) 

- 5 



Fig. 3: Results of visual inspection of KOSI-III sample after irradiation for 41 h 
with 1.3 to 2.7 SC artificial sunlight. 

15431600 1700 1800 1900 2001 21« 2200 ZB" 2400 100 


il cm below surface 

Fig. 4: Temperatures of KOSI-III sample in liquid No bath inside the glove box during 
operations for in situ analysis and sample taking 


Fig. 5: Central glove box in operation at KOSI-III experiment, with support for optical 


b) with blank 

Fig. 6: Non- irradiated reference 

(a) and irradiated KOSI-III sample 



1000 1500 

Wavelength (Nanometer) 



Fig. 7: Reflexion spectra and Albedo values for KOSI-III before and after artificial 
sun irradiation 






lower ( 


/ Y /~\ Ihemo couples 



/ / — \ \ 

y / ofrer s«ji trcolmeni V^^N^ 


// viicm- 


// ^ — "^ bef<jre si»i Ireolmen* 


5 penetrolKKi. cm 


Fig. 8: Material strength test and attempts to draw borekernels with ESA/ESTEC 

Penetrator at KOSI-II. a) the penetrator, b) material strength before and 
after irradiation with approx. 26 h 1-2 SC artificial sunlight. 


CO; content (weight %) in comet simulation (KOSI III) 
; original moteriol : 77.9% H^O , 13.6% CO?. 8.3% minerol dust) 


11.6 ± 0.5% 

12.4 ± 0.5% 

13.6 1 0.5% 

after filling ftie material 
into \he box . before 

artificial sun {1.3-2.7SC) 

dust ^ 

^17 tO.05% 

ice crust 

0.08 10.05% 

0.07 10.05% 

0.08 10.05% 

1.8 10.05% 

ice layer. 

/ CO. 20 % 

offer approx. Clti 
ortificial sun irradiation 


77 K plate 

Fig. 9: CO2 content in KOSI-III sample before and after artificial sun irradiation 

N, out 

for gloves 

dffuse reflexion wth 
■^Vayng Monfis" inf 
and k) ^2 coated 
sample holder Uock 

lock systea (or 
nlniduung Iq N, 
devor with ioape 

Fig. 10: Satellite glove box with Fig. 11: New concept of central ylove 

IR-spectrometer and liquid box (for KOSI-IV and following) 
N2 cooled sample tiolder for 
diffuse reflexion 


3) Measurement of microporosity (in preparation by Prof. J. Klinger and Dr. B. Schmitt, 

Glaciologie, CNRS, Grenoble). 

4) Determination of density and overall porosity (Dr. H. Hellmann, DLR-Koln). 

5) IR-diffuse reflexion spectroscopy in satellite glove box (Fig. 10, 
Dr. E. Langenscheidt, KFA-JQlich). 

6) Differential thermal analysis DTA and differential scanning calorimetry DSC (in 

preparation, KFA-JQlich) 
The following methods do not require very low temperatures and samples can be taken or 
stored with less precaution: 

7) Mineralogy and petrography (texture and structure) via drilling boreholes and removing 

borekernels of material (2 and 4 cm in diameter), fixation with gels and analysis 
after thin sectioning in a cold laboratory (233 to 269 K) (in preparation by H. Duren 
Prof. D. Stoffler, Univ. Munster). ' 

8) High performance liquid chromatography HPLC for chemical compounds such as e.g. CH3OH, 

CH2O, HCOOH, etc. newly formed by the interaction of heat, UV light and catalytic 
processes on the^ mineral dust (clay) grains (KFA-JQlich). 

9) Isotope ratios ^"0/-^°0 and H/D via majs^ spectrometry (Dr. H. Stichler, GSF-Neuherberg 
and Dr. H. Forstel, KFA-JQlich). If '■Kr^C ratio of CO2 has to be checked, the 
samples must be taken and stored at low temperatures (< 195 K). 

10) SEM of dry residue (Prof. K. Thiel, Universitat Koln). 

In future experiments, KOSI-IV etc., a new somewhat larger glove box (1x1x1 m) is 
conceived containing a turnable platform for the sample container which can be moved up 
and down in order to yield best access to the sample surface in view of in situ analysis 
(Fig. 11). An important improvement is the addition of a large lock chamber for 
introduction of the samples into the glove box under protective gas and for storage and 
bringing out small samples in larger dewars. 

If further developed and refined some of the methods described above may be considered in 
the planning of a return sample receiving station. 


Biel E., Patnaik A., and Roessler K. (1988) Herstellung binarer Eisschichten durch 
Kondensation aus der Gasphase, Rpt. JQl-2248, Kernforschungsanlage Julich, FRG, pp. 1-50. 

Bischoff A. and Stoffler D. (1988) Comet nucleus simulation experiments: mineralogical 
aspects of sample preparation and analysis (abstract), Lunar and Planetary Science XIX, 
pp. 90-91, Lunar and Planetary Institute, Houston. 

Dobrovolsky O.V. and Kajmakov E. (1977) Surface phenomena in simulated cometary nuclei, 
In Comets, Asteroids, Meteorites (Delserame A.H. ed.) pp. 37-46, University of Toledo, 

Grun E., Kochan H., Roessler K., and Stoffler D. (1987) Simulation of cometary nuclei. In 
Diversity and Similarity of Comets (Rolfe E.J. and Battrick B. eds.) ESA-SP-278, pp. 501- 

Grun E., Kochan H., Roessler K., and Stoffler D. (1988) Initial comet simulation 
experiments at DFVLR, In Experiments on Cosmic Dust Analogues (Bussoletti E. et al. eds.) 
pp. 17-23, Kluwer Acad. Pub!., Dordrecht. 

Grun E. and KOSI-team (1989) Modifications of comet material by the sublimation process, 
this issue (abstract in Workshop on Analysis of Returned Comet Nucleus Samples, 1989, 
Milpitas, Cal., pp. 24-25, Lunar and Planetary Institute, Houston). 


Ibadinov K.I. and Aliev S. (1987) Sublimation characteristics of H2O comet nucleus with 
COp impurities, In Diversity and Similarity of Comets (Rolfe E.J. and Battrick B. eds.) 
ESA-SP-278, pp. 710-719. 

Klinger J., Benckhoff J., Espinasse S., Grun E., Ip W., J06 J., Keller H.U., Kochan H., 
Kohl H., Roessler K., Seboldt W., Spohn T., and Thiel K. (1989) How far do results of 
recent simulation experiments fit current models of cometary nuclei?, Proc. Lunar Planet. 
Sci. Conf. 19th, pp. 493-497. 

Kochan H., Benckhoff J., Bischoff A., Fechtig H., Feuerbacher B., Grun E., J06 F., 
Klinger J., Kohl H., Krankowsky D., Roessler K., Seboldt W., Thiel K., Schwehm G., and 
Weishaupt U. (1989) Laboratory simulation of a cometary nucleus: experimental setup and 
first results, Proc. Lunar Planet. Sci. Conf. 19th, pp. 487-492. 

Kochan H., Feuerbacher B., J06 F., Klinger J., Seboldt W. , Bischoff A., Duren H., 
Stoffler D., Spohn T., Fechtig H., Grun E., Kohl H., Krankowsky D., Roessler K., Thiel 
K., Schwehm G., and Weishaupt U. (1989) Comet simulation experiments in the DFVLR Space 
Simulators, Adv. Space Res., in press. 

Pross L., Hemmerich J., and Roessler K. (1976) Transportable cryostat for charged 
particle irradiation and optical spectroscopy. Rev. Sci. Instrum. 47, 353-355. 

Roessler K., Bischoff A., Eich 6., Grun E., Fechtig H., J06 F., Klinger J., Kochan H., 
Stoffler D., and Thiel K. (1988) Cometary matter in observation and simulation 
experiments (abstract) Lunar and Planetary Science XIX, pp. 996-997, Lunar and Planetary 
Institute, Houston. 

Roessler K., Eich G., Heyl M., Kochan H., Oehler A., Patnaik A., Schlosser W., and Schulz 
R. (1989) Handling and analysis of ices in cryostats and glove boxes in view of cometary 
samples (abstract) In Workshop on Analysis of Returned Comet Nucleus Samples, 1989, 
Milpitas, Cal., pp. 62-63. 

Roessler K., Hsiung P., Kochan H., Hellmann H., Duren H., Thiel K., and Kolzer G. (1989) 
A model comet made from mineral dust and H2O-CO2 ice: sample preparation development 
(abstract) In Lunar and Planetary Science XX, pp. 920-921, Lunar and Planetary Institut, 

Saunders R.S., Fanale F.P., Parker T.J., Stephens J.B., and Sutton S. (1986) Properties 
of filamentary sublimation residues from dispersion of clay in ice, Icarus 66, 94-104. 

Spohn T., Benckhoff J., Klinger J., Grun E., and Kochan H. (1989) Thermal modeling of two 
KOSI comet nucleus simulation experiments, Adv. Space Res., in press. 

Storrs A.D., Fanale F.P., Saunders R.S., and Stephens J.B. (1988) The formation of 
filamentary sublimate residues (FSR) from mineral grains, Icarus 76, 493-512. 

Thiel K., Kochan H., Roessler K., Grun E., Schwehm G., Hellmann H., Hsiung P., and Kolzer 
G. (1989) Mechanical and SEM analysis of artificial comet nucleus samples, this issue 
(abstract in Workshop on Analysis of Returned Comet Nucleus Samples, 1989, Milpitas, 
Cal., pp. 75-76, Lunar and Planetary Institute, Houston). 

Thiel K., Kolzer G., Kochan H., Grun E., and Kohl H. (1989) Crustal evolution and dust 
emission of artificial cometary nuclei (abstract) In Lunar Planet Sci. Conf. XX, pp. 
1113-1114, Lunar and Planetary Institute, Houston. 



Fun-Dow Tsay 

Soon Sam Kim 

Ranty H. Liang 

Jet Propulsion Laboratory 

California Institute of Technology 

Pasadena, California 



Fun-Dow Tsay, Soon Sam Kim and Ranty H. Liang 



Electron Spin Resonance (ESR) studies have been carried out on organic and 
inorganic free radicals generated by gamma- ray and/or UV- irradiation and trapped 
in ice matrices. It is suggested that the concentration of these free radicals 
together with their thermal stability can be used as an accurate built-in 
geo thermometer and radiation probe for returned comet nucleus sample studies. 
ESR studies have also been carried out on paramagnetic (Mn^"*", Ti^* and Fe^"*") and 
ferromagnetic (ferric oxide and metallic iron) centers known to be present in 
terrestrial and extraterrestrial samples. The presence or absence of these 
magnetic centers coupled with their characteristic ESR lineshape can be used to 
investigate the shock effects, quenching/cooli-g rate and oxidation- reduction 
conditions in the formation and subsequent evolution of returned comet nucleus 
samples . 


The most important scientific objective of the planned Comet Nucleus Sample 
Return Mission is to return to Earth the least altered, pristine samples which 
could reflect formation conditions and evolutionary processes in the early solar 
nebula. The returned cometary samples are expected to consist of fine-grained 
silicate materials mixed to some extent with ices. The cometary ices may imbed 
in them such simple molecules as CO, CO2, H2O, NH3 and CH4 as well as organics 
(HjCO, CH3OH, HCOOH and HCONH2) and/or more complex compounds (Kissel et. al . . 
1987: Allamandola et . al . . 1988). Because of the exposure to ionizing radiation 
from cosmic -ray, gamma- ray and solar wind protons at low temperature, free 
radicals in the form of "H, "OH, •CR^ and -CH2 are expected to be formed and 
trapped in the solid ice matrices. The kind of trapped radical species together 
with their concentration and thermal stability can be used as a dosimeter as well 
as a geothermometer to determine thermal and radiation histories as well as 
outgassing and other possible alteration effects since the nucleus material was 
formed. The free radicals that are known to contain unpaired electrons are all 
paramagnetic in nature. Thus, they can be readily detected and characterized in 
their native form by the Electron Spin Resonance (ESR) method. In fact, ESR has 
been shown to be a non-destructive, highly sensitive tool for the detection and 
characterization of paramagnetic (Fe^*, Ti^"*" and Mn^*) , ferromagnetic (magnetite, 
hematite, maghemite and metallic iron), and radiation damage centers in 
terrestrial and extraterrestrial geological samples (Tsay et. al . . 1971, 1972a, 
1973; Ahrens et. al. . 1976; Vizgirda et. al. . 1980; Beckett et. al . . 1988). The 
purpose of this paper is to point out the potential use of ESR as an effective 
method in the study of returned comet nucleus samples, in particular, in the 
analysis of fine-grained, solid state icy samples. 

Work Supported by NASA. 



ESR Spectroscopy : ESR measures the absorption of microwaves by a 
paramagnetic or ferromagnetic center in the presence of an applied external 
magnetic field. Conventional ESR experiments are carried out with varying 
magnetic fields on samples in a tuned resonant cavity operated at a certain 
frequency. The magnetic field position (g-value) at which resonance occurs, 
together with the nvunber of resonance lines (fine and hyperfine structure) and 
the separation between them (zero-field splitting and hjrperfine coupling 
constant) , provides a direct identification of magnetic species , their oxidation 
state and chemical environment. Detailed analyses of ESR signals and their 
response to physical and chemical treatments can furnish further information on 
the nature, origin, thermal stability and formation condition of the magnetic 
species examined. The sensitivity of ESR is such that it can detect 3 x 10^^ 
standard spin (or molecules assuming one spin per molecule). Thus, milligram- 
size cometary samples can be examined non- destructively as a function of 
temperature by ESR. In many instances, ESR has the advantage of detailed 
submicroscopic identification of transient species and/or reaction intermediates 
generated in UV and/or gamma- ray radiation. We intend to use the unique 
capability of ESR to investigate the thermal cycling, radiation exposure, shock 
effects and oxidation/reduction conditions in the formation and evolution of comet 
nucleus samples . 

ESR Experiments : The ESR measurements were carried out on samples sealed 
in 2-3 mm o.d. quartz tubes. An ESR spectrometer (Varian E-Line Century Series 
with 15" magnet) operating at X-band (9.2 GHz) was used with 100 KHz field 
modulation. The spectrometer was equipped with variable temperature accessories 
for low temperature measurements. A ^°Co gamma- ray source and an UV-enhanced 
Xenon lamp (ILC Technology, LX 300 UV) with a water filter (4 cm path length) were 
used for irradiation experiments . 


Thermal and Radiation Histories : Free radicals in the form of •OH and HOj 
have been detected in 7-irradiated ice at TT'K (see Fig. 1). These free radicals 




Fig. 1. ESR spectra observed 
for the 7-irradiated ice at 
various temperatures . Note 
the decaying of ESR signal 
amplitude and the relative 
intensity ratio at high 
temperature . 


are found to be unstable and to decay rapidly above 110°K as reported (Siegel et. 
al. . 1961). No ESR signals of the free radicals can be detected above 160°K. 
The characteristic ESR signals of "H having a hyperfine coupling constant of 500 
gauss and present in 7- irradiated ice can only be detected below 60°K. 

ESR studies have also been carried out on free radicals generated in gamma- 
ray and/or UV- irradiated simple organic compounds and trapped in ice matrices. 
Various organic free radical species are formed in 7- irradiated formaldehyde 
(H2CO) as a function of temperature (See Fig. 2). The characteristic ESR signals 
of •CHO doublet are found to be unstable above 160°K, and no ESR signals can be 
detected at 223°K. 

ESR signals attributable to -CH2 organic free radicals are detected in the 
UV-irradiated polycrystalline samples of simple nitrogen containing organic 
compounds such as glycine (NH2CH2COOH) at room temperature (See Fig. 3) . The 
triplet ESR signals of -CHj are also detected in UV-irradiated frozen aqueous 
solutions containing glycine at 77°K (Fig. 3). The ESR signals detected in ice 
matrices at low temperature are foxind to be unstable at elevated temperature and 
to disappear at 273 °K. 

, 20 G 


Fig. 2 

Temperature -dependent ESR 
spectra observed for 
7- irradiated formaldehyde 
(H2CO) . The broadened 
center peaks detected at 
77"'K and 160°K are the 
composite spectrum of the 
ESR signals arising from 
multiple radical species 
(•CH3, -o5hO- and •OCHO) . 
The dominant spectral 
features are the quartet of 
• CH3 at 197''K. 

Fig. 3 

ESR spectra of -JHj 




A. Before irradiation. 

B. UV-irradiated 
polycrystalline sample of 
glycine at 296°K. 

C. UV-irradiated ice 
containing 0.2M glycine 
at 77''K. 

By ESR definition, the least altered, pristine comet nucleus sample should 
contain all the organic and inorganic free radicals produced and stabilized in 
its environment since the formation of the sample. The total free radical 
concentration a pristine sample should have can be recreated by re- irradiation 
of the returned sample at its prevailing temperature since the sam