WORKSHOP ON THE ISSUE MARTIAN METEORITES: WHERE DO WESTAND AND. WHERE ARE WE GOING?

J . '

November 2-4, 1998 Lunar and Planetary Institute, Houston, Texas WORKSHOP ON THEISSUE MARTIAN METEORITES: WHERE DO WE STAND AND WHERE ARE WE GOING?

Held at Houston, Texas

November 2-4, 1998

Sponsored by National Aeronautics and Space Administration Lunar and Planetary Institute

Lunar and Planetary Institute 3600 Bay Area Boulevard Houston TX 77058-1113

LPI Contribution No. 956 Compiled in 1998 by LUNAR AND PLANETARY INSTITUTE

The Institute is operated by the Universities Space Research Association under Contract No. NASW-4574 with the National Aeronautics and Space Administration.

Material in this volume may be copied without restraintfor library, abstract service, education, or personal research purposes; however, republication of any paper or portion thereof requires the written permission of the authors as well as the appropriate acknowledgment of this publication.

Abstracts in this volume may be cited as

Author A. B. (1998) Title of abstract.In Workhop on the Issue Martian Meteorites: Where Do We Stand and Where Are We Going? LPI Contribution No. 956. Lunar andPlanetary Institute, Houston. 64 pp.

This volume is distributed by

ORDER DEPARTMENT Lunarand Planetary Institute 3600 Bay Area Boulevard Houston TX 77 058-1113, USA Phone: 281-486-2172 Fax: 281-486-2186 E-mail: [email protected].nasa.gov

Mail order requestors will be invoiced for the cost of shipping and handling. LI Contribution No. 956 iii

Preface

This volume contains papers that were accepted for presentation at the Workshop on the Issue MartianMeteorites: Where Do We Stand and Where are We Going? The workshop was held November 2-4, 1998, at the Lunar and Planetary Institute in Houston, Texas. Members of the ScientificOrganizing Committee included D. C. Black, Co-Chair (Lunar and PlanetaryInstitute), C. Pilcher, Co-Chair (NASA Headquarters), D. Des Marais (NASA Ames Research Center), M. Drake (University of Arizona), J. Ferris(Rensselaer Polytechnic Institute), J. Longhi (Lamont-Doherty Earth Observatory),D. McCleese (Jet Propul­ sion Laboratory), D. McKay (NASA Johnson Space Center), H. Mcsween (University of Tennessee), M. Meyer (NASA Headquarters), H. Newsom (University of New Mexico), P. Rogers (Smithsonian Institution), and A. Treiman ( Lunar and PlanetaryInstitute). Logistical, administrative, and publications support were provided by the Publications and Program Services Departmentof the Lunar and Planetary Institute.

LPI Contribution No. 956 v

Contents

Gamma-Ray Sterilization of Mars Analog Rocks and Minerals C. C. Allen ...... 1

Biomarkers in Carbonate Thermal Springs: Implications for Mars C. C. Allen, S. J. Kivett, and D. S. McKay ...... 2

Fluid Inclusions in Allan Hills 84001 and Other Martian Meteorites: Evidence for Volatiles on Mars R. J. Bodnar ...... 3

Rubidium-StrontiumFormation Age of Allan Hills 84001 Carbonates L. E. Borg, L. E. Nyquist, C.-Y. Shih, H. Wiesmann, Y. Reese, and J. N. Connelly ...... 5

Rare Potassium-bearing Mica in Allan Hills 84001: Additional Constraints on Carbonate Formation A. J. Brearley ...... 6

Water on Mars: How Much Was (Is) There, Where Did It Go, and When? M. H. Carr ...... 8

On Independent Proof of a Martian Origin for the Shergotty Meteorite Yu. T. Chuburkov, V. P. Pe rely gin, and V. A. Alekseev ...... 9

The Searchfor Early Life on Mars: A European Space Agency Exobiology Science Team Study P. Clancy, A. Brack, B. Hofmann, G. Horneck, G. Kurat, J. Maxwell, G. G. Ori, C. Pillinger, F. Raulin, N. Thomas, F. Westall, and B. Fitton ...... 9

Studies of Life on Earth are Important for Mars Exploration D. J. Des Marais ...... 10

Optimizing the Search for a Fossil Record of Ancient MartianLife J. D. Farmer ...... 11

Bacterial Mineral Precipitation and the Making of Microfossils F. G. Ferris ...... 12

Organic Carbon in Carbonate and Rim fromAllan Hills 84001 G. J. Flynn, L. P. Keller, C. Jacobsen, and S. Wirick ...... 13

Chains of Magnetite Crystals in Allan Hills 84001: Evidence of Biological Origin E. I. Friedmann, J. Wierzchos, and C. Ascaso ...... 14 vi Workhop on Martian Meteorites

Biomimetic butAbiotic Carbonates: New Geochemical Markers for Primitive Environments J. M. Garcia-Ruiz ...... 16

How Do the Properties of Allan Hills 84001 Compare with Accepted Criteria for Evidence of Ancient Life? E. K. Gibson Jr., D.S. McKay, K. Thomas-Keprta, F. Westall, and C. A. Romanek ...... 17

Xenon and Argon Isotopes in Irradiated, Etched Nakhla: Characterizingthe Host of Martian Atmospheric Xenon J. D. Gilmour, J. A. Whitby, R. Burgess, and G. Turner...... 18

Formation of Carbonates in Allan Hills 84001 by Impact Metasomatism: Cooking with Gas R. P. Harvey...... 20

Recovering More Antarctic Martian Meteorites: Answers to Common Questions R. P. Harvey and W. A. Cassidy ...... 22

Estimates of Oxygen Fugacity in the Basaltic Shergottites fromElectron Microprobe Oxygen Analysis C. D. K. Herd and J. J. Papike ...... 23

Implications of Hydrous Weathering Product on Mars forFuture Exploration L. E. Kirkland and K. C. Herr ...... 24

The Signature of Life: Is It Legible? A. H. Knoll...... 25

Martian Biogenic Activity: Looking for Viruses and DNA Traces Instead of Extant Bacteria Traces L. V. Ksanfomality...... 26

The Structure of Findings in Allan Hills 84001 May Hint at Their Inorganic Origin

L_. V. Ksanfomality ...... _...... 26 Constraints on Martian Volatile History from Studies of Martian Meteorites: Lessons Learnedand Open Questions L. A. Leshin ...... 27

The Future Search for Life on Mars: An Unambiguous Martian Life Detection Experiment G. V. Levin ...... 28 LP/ Contribution No. 956 vii

Curation of U.S. Martian Meteorites Collected in Antarctica M. Lindstrom, C. Satterwhite, J. Allton, and E. Stansbery...... 29

Principal Component Analysis for Biosignature Detection in Extraterrestrial Samples G.D. McDonald and M. C. Storrie-Lombardi ...... 30

Evidence for Ancient Lifein Mars Meteorites: Lessons Learned D. S. McKay ...... 31

Key Research: Martian Meteorites and Their Relationships to Planetary Geology and Perhaps Biology H. Y. McSween Jr ...... 33

What Were the Major Factors That Controlled Mineralogical Similarities and Differencesof Basaltic, Lherzolitic, and Clinopyroxenitic Martian Meteorites Within Each Group? T. Mikouchi, M. Miyamoto, and G. A. McKay ...... 34

Genesis of the Mars Pathfinder "Sulfur-Free"Rock fromSNC Parental Liquids M. E. Minitti and M. J. Rutherford...... 36

Recognizing Lifeand Its Evolution Through Biomarkers J. M. Moldowan ...... 37

Possible Microfossils (WarrawoonaGroup, Towers Formation, Australia, ~3.3-3.5 Ga) P.A. Morris, S. J. Wentworth, C. C. Allen, and D. S. McKay ...... 38

Perspectives on the Future Search for Life on Mars and Beyond K. H. Nealson ...... 39

Oxygen-Isotopic Composition of NakhlaSiderite: Implications for Martian Volatiles J. M. Saxton, I. C. Lyon, and G. Turner ...... 39

Reconnaissance Sampling of AirborneMolecular Organic Contamination in the Meteorite Curation Facility of Johnson Space Center E.A. Schilling and M. N. Schneider ...... 40

Biogenic or Abiogenic Origin of Carbonate-Magnetite-SulfideAssemblages in Martian Meteorite Allan Hills 84001 E. R. D. Scott ...... 42 viii Workhop on Martian Meteorites

Origin of Carbonate in Martian Meteorite Allan Hills 84001 E. R. D. Scott and A. N. Krot ...... 44

A Comparison Between SulfideAssemblages in Martian Meteorites Allan Hills 84001 and Govemador Valadares C. K.Shearer and C. Adcock ...... 45

Evidence for a Late-Stage ThermalOverprint in Allan Hills 84001 and Implications for Biomarkers C. K.Shearer and A. Brearley ...... 47

Overview: The Search forLife on Mars S. W. Squyres ...... 49

The Lateral Distribution of Polycyclic Aromatic Hydrocarbons in Allan Hills 84001: Implications for Their Origin T.Stephan, D. Rost, C. H. Heiss, E. K. Jessberger, and A. Greshake ...... 50

Mineral Biomarkers in Martian Meteorite Allan Hills 84001? K. L. Thomas-Keprta, D. A. Bazylinski,S. J. Wentworth, D.S. McKay, D. C. Golden, E. K. Gibson Jr., and C.S. Romanek ...... 51

Mineralization of Bacteria in TerrestrialBasaltic Rocks: Comparison with Possible Biogenic Features in Martian Meteorite Allan Hills 84001 K. L. Thomas-Keprta, D.S. McKay, S. J. Wentworth, T. 0. Stevens, A. E. Taunton, C. C. Allen, E. K. Gibson Jr., and C.S. Romanek ...... 53

Ancient Martian Lifein Allan Hills 84001? Status of Some Current Controversies A. H. Treiman ...... 54

High Strain-Rate Deformation and Friction Melting as a Possible Origin for "Shock" Features in Allan Hills 84001 C. H. van der Bogert and P. H.Schultz ...... 56

The Common Ion Effect in Deposition of Martian (e.g., Allan Hills 84001) Carbonates P.H. Warren ...... 58

Planetary Core Formation: Evidence fromHighly Siderophile Elements in Martian Meteorites P. H. Warren, G. W. Kallemeyn, and F. T Kyte ...... 58

Alteration Products and Secondary Minerals in Martian Meteorite Allan Hills 84001 S. J. Wentworth, K. L. Thomas-Keprta, and D.S. McKay ...... 59 LP! Contribution No. 956 ix

TerrestrialBiomarkers for Early Lifeon Earth as Analogs for Possible Martian Life Forms: Examples of Minerally Replaced Bacteria and Biofilmsfrom the 3.5-3.3-Ga Barberton Greenstone Belt, South Africa F. Westall, D. S .. McKay, E. K. Gibson, M. J. de Wit, J. Dann, D. Gerneke, and C. E. J. de Ronde ...... 61

Abiotic Synthesis of Hydrocarbons on Mars: Theoretical Modeling of Metastable Equilibria M. Yu. Zolotov and E. L. Shock ...... 62

LP Contribution No. 956 1

Abstracts

GAMMA-RAY STERILIZATION OF MARS ANALOG Rocks and minerals. Fourteen types of materials were tested to ROCKS AND MINERALS. C. C. Allen, Lockheed Martin, assess the efects of y radiation on the physical and chemical prop­ Houston TX 77258, USA ([email protected]). ertes of rocks ad minerals: basalt, haite, chert, Mas soil simulat (weathered volcanic ash), cabonaceous chondrite meteorite, pla­ Samples of rock and soil, collected by robotic spacecraft on gioclase, olivine, pyroxene, CaCO3, clay, quartz, gypsum, water, Mas, will be reted to terestial laboratories ealy in the next ad Sn. century. Plas call fr the saples to be placed immediately in bio­ Triplicate splits of each sample were prepared fr each radia­ logical containment ad tested for signs of present or past lif ad tion dose. The splits (-1 cm3) were placed in snap-top plastic vi­ biological hazard. It is recommended that "contolled distibution als. Sets of samples were iradiated as described below. of unsterilized materials fom Mars should ocur only if rigorous Irradiation. The samples were iradiated in a Gaacell 220 aalyses determine that the materials do not consttute a biological High Dose Research lator at the Centers fr Disease Contol haad. If ay portion of the sample is removed fom containment ad Prevention. Bacteria-loaded cores were expsed to total doses pror to completion of these analyses it should frst be sterilized" of 3 x 1 (, 3 x 106, ad 3 x 107 rad at ambient tempratures, which [I]. While sterilizaton of Mas samples may not be required, a did not exceed 10°C. acceptale method must be availale befre the samples ae reted Gamma Dose Required to Kill Microorganisms in Geologic to Earth. Samples: Both spcies of bacteria were successflly cultured fom Various techniques a routinely used to sterilize biological sam­ the unirradiated contol samples. Bacteria viability decreased with ples. These include d heatng to temperatures of 150°C or higher, increasing radiaton at doses of 3 x 105 ad 3 x 106 rad. No vi­ heatng in the presence of stea, expsure to poisonous gases such ability was detected in the samples irradiated at 3 x 107 rad. Esti­ as formaldehyde and propiolatone, exposure to H2O2 vapor or mates of viability, expressed in terms of light absorbance in liquid plasma, exposure to ultaviolet light, ad exposure to y radiation. media, ae listed in Table 1. The appropriate technique depnds on the physical chaacteristics Effects of Radiation on Rocks and Minerals: Induced of the sample and the desired results. radioactivit. Samples were monitored fr induced radioactivity at Gamma radiation is routinely used to inactivate viruses ad the Johnson Space Center (JSC) Radiation Counting Laboratory, destoy bacteria in medical reseach. The most common commer­ one of the world's best low-background counting fcilities. Entire cial sterilizers utlize 6Co, which emits y photons with energies sample sets (bacteria-loaded cores plus rocks and minerals) were of 1.173 ad 1.332 MeV. Absorbed doses of approximately 10 rad placed in the counter together. (104 y ray= 104 ergs/gm) ae sufcient to destoy most bacteria [2]. The saple set exposed to 3 x 1 O rad produced 36 y-ray peaks, The current study is designed to investgate the efects of lethal all attibutable to natural K, Th, ad U in the samples or to 137Cs doses of 6Co y radiation on geologic materials analogous to the produced by atmospheric nuclear tests. The net count rate attibut­ first samples to be reted fom Mars. The goals are (1) to deter­ able to these peaks across the energy rage 0.05-2.0 MeV was mine the y dose required to kill microorgaisms within geologic 0.521 ± 0.003 counts/s. The sample set exposed to 3 x 107 rad pro­ samples, ad (2) to determine the effects of lethal doses of y radia­ duced the same 36-y peaks ad the identical net count rate. tion on the physical and chemica properties of the samples. Samples of Sn were irradiated to monitor the production of Samples and Methods: Microorganisms in geologic sam­ 117mSn, a y emitter with a half-lif of 13.6 days. After a radiation ples. Cores 8 mm in diameter ad approximately 17 mm long were dose of 3 x 107 rad, a 3.64-g Sn sample showed no detectable prepared fom dense, fne-graned basalt taken fom the Umtanum counts above background, at a sensitivity of 0.004 counts/s (2s). fow, Grade Ronde fraton of the Columbia River basalts. A This count rate sets a upper limit of <7000 atoms of mmsn cre­ slot normal to the long ais was cut halfway through the center of ated, or <5 x 10-18 of the 117Sn atoms present. each core. Half of the cores were sterilized by autoclaving in stea, ad half were not sterilized. Saples of two bacteria species, Clostridium sporogenes ad Bacillus subtilis, were grown fom pure cultures in gel media and TABLE 1. Efects of y rdiaton on bactera. inserted into the slots in the basalt cores. Both ae spre-frming bacteria that are easily cultured. C. sporogenes requires incubation Samle Contol 3 X JOS 3 X J) 3 X 107 in a CO2 environment while B. subtilis is incubated in ar. The bac­ Clostridium 0.63 0.53 ND ND teria-loaded cores were placed in snap-top platic vials and held at (sterle cores) room temperature. Sets of saples were iradiated as described be­ low. Clostridium 0.61 0.50 ND ND After iradiation C. sporogenes was cultured in 90 ml of modi­ (nonsterile cores) fed AC broth (without ascorbic acid) at a pH of 7.2, adjusted with sodium phosphate. The samples were incubated at 37°C, anaero­ Bacillus 0.99 0.50 0.01 ND bically in crimp top vials purged with CO2. B. subtilis was grown (sterile cores) in modifed 90 ml of nutrient broth at a pH of 6. Samples were Bacillus incubated in ai at 30°C. After 116 hr of growth the samples were >0.60 0.22 0.00 ND (nonsterle cores) placed into liquid media and light absorbance through the media was measured at a wavelength of 600 rn. Light absorbance at 600 nm (sterle media = 0). 2 Workhop on Martian Meteorites

Elemental composition. Selected major-, minor-, and trace­ ogy to Earth, specialized microbes may have existed in the heated, element abundances in powdered basalt were measured by instru­ mineralizedwaters, andprecipitates of carbonateand/or silica from mentalneutron activationanalysis (INAA). The samples, standards, these waters may have trapped andpreserved evidence of life.Since and control sampleswere encapsulated in pure SiO2 glass tubes and the geological interactions that produce thermal springs can be rec­ irradiated at the Research Reactor Facility of the University of Mis­ ognized in orbital imagery, directed searches for microfossils in 2 souri for 12 hr at a thermal neutron flux of 5.5 x 1013 n cm- s-1. such deposits are deemed possible [4). The samples were counted three times (roughly 0.5, 1, and 4 weeks We are engaged in a study of the signatures produced by con­ after irradiation) in order to obtain data for nuclides of differing temporary biogenic activity (biomarkers) in carbonate thermal half-lives. Standards used, the data-reduction process, and correc­ springs. We are examining the microbes that live in such environ­ tion procedureswere the same as those previously employed in the ments and the preservation of microbial forms,biofilms, and petro­ Johnson Space Center laboratory [3). graphic fabricsindicative of lifein thermal spring mineral deposits. Twenty-eight major-, minor-, and trace-element concentrations This work is part of a much more extensive study to refine the were determined. For all but one of these elements, the concentra­ appropriate tools, techniques, and approaches to seek evidence of tions (within analytical error) were unchanged following irradia­ life i a range of planetarysamples. A deeper understandingof bio­ tion at any of the three doses. The single exception was Ce, which logical signatures will prepare us forthe detailed search for life on showed a decrease in concentration from 53.9 ± 0.8 to 52.0 ± Mars and eventually on other planets. Overall, the study of 0.8 ppm between the unirradiated control and the sample that re­ biomarkers in rocks and soils will provide insight into the evolu­ ceived a dose of 3 x 107 rad. tion of life because such signatures are a record of how life inter­ Strontium isotopes. The ratios of 87Sr/86Sr were measured for acts with its environment, how it adapts to changing conditions, unirradiatedand irradiated basalt samplesusing standard procedures and how life can influencegeology and climate. for radiometric age dating [4]. The values, 0.705542 ± 12 (con­ Biomarkers in Carbonate Thermal Springs: We are cur­ trol) and 0.705546 ± 12 (3 x 107 rad), were indistinguishable. rently studying samples fromfour active thermalsprings: Le Zitelle Therefore, the effectof this dose of 60Coirradiation on Sr-isotopic in the Viterbo region of Italy [5], Narrow Gauge in the Mammoth compositions of basalt was not detectable. complex of Yellowstone National Park,Wyoming [6], Hot Springs Future Analyses: The effects of y radiation on a variety of National Park,Arkansas [7], and springs at Jemez on the slopes of microorganisms is currently under way. 1n addition, colleagues are the Valles caldera, New Mexico [8). In each case, water reaches measuring effects on the chemical composition, crystal structure, the surface at 60°-72°C and near-neutral pH (6.3-7.5), rapidly pre­ thermoluminescence, and reflectance spectra of the full range of cipitating large amountsof aragoniteand calcite. rock and mineral samples. We are concentrating on samples fromthe hottest areas of each Acknowledgments: Colleagues providing samples, experi­ spring. The Italian, Yellowstone, and Jemez springswere sampled mental support, and analytical expertise for this study include F. at the surface, while the Arkansas springs were sampled below Albert, J. Combie, D. Lindstrom, M. Lindstrom, D. Mittlefehldt, ground. Various samples were fixed in glutaraldehyde, air dried P. Morris, R. Morris, L. Nyquist, Y. Reese, P. Simpson, and H. or critical point dried, etched in 1% HCl, and examined in a high­ Wiesmann. resolution SEM. References: [1] Space Studies Board (1997) Mars Sample Microbes. The near-ventenvironments of thermal springs sup­ Retur Issues and Recommendations, National Research Council, port a variety of microorganisms. Thermothrix thiopara, a pro­ Washington, DC. [2] Battista J. R. (1997) Ann. Rev. Microbial., lificS-oxidizing bacterium [8], is the highest-temperature species 51, 203-224. [3] Mittlefehldt D. W. and Lindstrom M. M. (1993) identified at our sites in Yellowstone and Jemez. Pentecost [9] Proc. NR Symp. Antarctic Meteorites, 6, 268-292. [4] Nyquist identified a species of the photosynthetic, filamentous bacteria L. E. et al. (1994)Me teoritics, 29, 872-885. Chlorojexus as the dominant form immediately downstream from the vent of the Italian spring.Spherical microbes as largeas 15 µm in diameter populate the waters at Hot Springs. Rod-shaped mi­ BIOMARKERS IN CARBONATE THERMAL SPRINGS: crobes several micrometers in length are foundin samples from all IMPLICATIONSFO R MARS. C. C. Allen 1, S. J. Kivett2, and four sites. D. S. McKay3, 1Lockheed Martin, Houston TX 77258, USA The thermal spring depositsalso contain forms,many of which ([email protected]), 2University of Houston-Down­ appear to be biological, significantly smaller than conventional bac­ town, Houston TX 77002, USA, 3NASA Johnson Space Center, teria. Large numbers of 100-200-nm spheres, the "nanobacteria" Houston TX 77058, USA. described by Folk [10], are common in portions of the Italian and Yellowstone carbonates. The spheres arecomposed of C, 0, F, P, Evidence of possible relict biogenic activity has been reported and Ca with detectable Si and S, and areoften found enmeshed in in carbonate inclusions within martian meteorite ALH 84001 [1). organic mucus. Numerous 300-500-nm rods and spheres, some The initial evidence included ovoid and elongated forms 50- apparently dividing, populate the Arkansas samples. 500 nm in length, morphologically similar to but significantly Many of the microbes die and their remains arerapidly destroyed smaller than many terrestrial microbes. More recently, thin struc­ in carbonate thermal spring environments. Organic matter is gen­ turesresembling the remains of organic biofilrnshave been reported erally rare in carbonate sinters deposited at >30°C, suggesting that in the same meteorite [2). decomposition rates in such thermalenvironments are very high Carbonates have also been discussed in the context of Mars [ll). Detailed SEM examination of samples from all four sites sample return missions. Thermal spring deposits have often been consistently reveal mineralized cell remains but only scattered in­ cited as prime locations for exobiological exploration [3). By anal- tact cells. LP/ Contribution No. 956 3

While thermal spring microbes are apparentlypoorly preserved These observations support the interpretation of submicrometer in carbonates, fossilizationby silica canprovide enduring evidence features in ALH 84001 as possible relict life forms and biofilms. of life.At Jemez, amorphous silica spheres 50-300 nm in diam­ They also supportthe sample returnstrategy of searching for a va­ eter preserve the forms of dissolved microbial cells. A wide vari­ riety of biomarkers in thermal spring deposits on Mars. ety of well-preserved thermophilic bacteria has been found in the References: [l] McKay D.S. et al. (1996) Science, 273, 924. silica sinter deposited by many Yellowstone thermal features [12]. [2] McKay D. S. et al. (1997) LPS XXV/ll, 919. [3] An Exo­ Similar fossil assemblages have been described in siliceous ther­ biological Strategy for Mars Exploration (1995) NASA SP-530. mal spring deposits as old as 400 Ma [13]. [4] Walter M. R. and Des Marais D. J. (1993) Icarus, 101, 129. Biofilms. The three-dimensional network of polysaccharide [5] Folk R. L. (1994) Geograph. Phys. et Quat., 48, 233. mucus, living cells, and cell remains that constitutes a biofilmis a [6] ChafetzH. S. andLawrence J.R. (1994) Geog. Phys. et Quat., distinctive macroscopicbiomarker. Carbonate samples fromall four 48, 257. [7] Norsworthy S. F. (1970) Ph.D. dissertation, UCLA. of our sites contain biofilms in various states of preservation. [8] Caldwell D. E. et al. (1984) Geomicrobiol. J., 3, 181. [9] Pen­ ° Biofilms arestable in water ashot as72 C andretain their three­ tecost A. (1995) Microbios, 81, 45. [10] Folk R. W. (1993) J. Sedi­ dimensional nature as the water cools. Upon drying, however, the ment. Petrol., 63, 990. [11] Farmer J. D. and Des Marais D. J. mucus shrinks and deforms but preserves its intercellular binding (1994) in Microbial Mats: Structure, Development and Environ­ structure.Carbonate samples still display extensive biofilmremains mental Significance, p. 61, Springer Verlag. (12] Cady S. L. and after years of desiccation. FarmerJ. D. (1996) in Evolution of Hydrothermal Systemson Earth Morphologic evidence of biofilms can be preserved by miner­ (and Mars?), p. 150, Wiley. [13] Walter M. R. (1996) in Evolu­ alization. Biofilmsin the Jemez samplesare extensively coated with tion of HydrothermalSystems on Earth (and Mars?), p. 112, Wiley. the 50-300-nm-diameter silica spheres. Cady andFarmer [12] have [14] Westall F. (1994) DarmstadterBeitrage zur Naturgeschichte, demonstrated that silica in thermal springs preferentially nucleates 4, 29. (15] ChafetzH. S. and Buczynski C.(1992) Palaios, 7, 159. on and preserves organic mucus. Westall [14] provides evidence of silicified biofilm in deep-sea sediments at least 5.5 m.y. old. Petrographic fabrics. The importance of microbes in promot­ ing the deposition of CaCO3, particularly aragonite, in thermal FLUIDINCLUSIONS INALLAN HILLS84001 AND OTHER springs is a matter of ongoing debate. Pentecost [9] argues against MARTIAN METEORITES: EVIDENCEFOR VOLATILES significant microbial influence in the case of Yellowstone depos­ ON MARS. R. J. Bodnar, Fluids Research Laboratory, Depart­ its. He accepts microbially induced precipitationonly in cases where ment of Geological Sciences, Virginia Polytechnic Institute and aragonite crystals clearly copy the structures of filamentary bacte­ State University, Blacksburg VA 24061-0420, USA (bubbles@ ria. Other investigators, however, provide evidence fora range of vt.edu). bacterially induced lithificationstyles [15]. Thermothrixthiopara, foundnear the vents of our Yellowstone Introduction: The important role that volatiles play in plan­ sampling sites, forms filaments up to 10 cm long. Immediately etary evolution is well known. On Earth, volatiles have and con­ downstream, the filaments become overgrown by aragonite. The tinue to play key roles in the development and properties of the result is a mass of parallel aragonite needles, each millimeters in crust,hydrosphere, and atmosphere. More recently, the recognition diameter andcentimeters long, which closely mimics the Thermo­ that the mantlecontains small but ubiquitous quantities of volatile thrix filaments. In this case bacterially induced precipitation yields components has led to an improved understanding of mantle rhe­ a distinctive petrographic fabric in the carbonate rocks that is spe­ ology and chemistryand the recycling of volatiles in the Earthsys­ cifically attributableto life. This fabrichas been recognized in de­ tem. And, of course, volatiles, especially water, are considered to posits asold as -360 Ma [11]. be a prerequisite forthe origin andevolution of life as we know it. Biological action can also promote the selective dissolution of The recent announcement that evidence for life had been dis­ minerals.Partial dissolution and pitting in calcite crystals is a com­ covered in the ALH 84001 martian meteorite [1] combined with mon feature of the Jemez samples. Pristine crystals show no evi­ the highly successful Mars PathfinderMission has led to renewed dence of associated microbes, while pitted areas are invariably interest in the geochemical evolution of Mars. Much of the scien­ associated with biofilms. tific effort in this area has been devoted to characterizing the Implications for Mars: Extensive previous studies of life in volatiles present in the SNC meteorites and, by inference, the carbonate thermal springs, coupled with our own results, have im­ volatiles present on Mars when the rocks formed.Here, the occur­ plications in the search for evidence of martianlife: rence of "free"volatiles (i.e., volatiles occurring in the fluidstate, l. Specialized microbes are common in carbonate-precipitating and not as minor components of minerals and/or glasses) in the environments as hot as 72°C. ALH 84001and Nakhla meteorites is reported, along with possible 2. Several types of microorganisms in such environments have implications. sizes significantly<1 µm. Background: Fluid inclusions are microsamples of fluid(liq­ 3. Biofilmsare characteristic of thermal spring microbial com­ uid or gas) trapped in rocks and minerals either at the time of their munities. original formation or at some later time asfluids flowthrough rocks 4. Microbes are poorly preserved in thermal spring carbonates. along fractures [2]. The presence of fluid inclusions in terrestrial 5. Silicified microbes and biofilms can be well preserved over rocks is the rulerather than the exception. Indeed, the rock sample geologic time. that contains no inclusions that are visible at high magnification 6. Some petrographic fabrics, recognizable in the geologic re­ (>1000x)is rare, andmany mineralssuch as milky quartz may con­ cord, arespecifically attributable to thermal spring microorganisms. tainup to 109 fluid inclusions per cubic centimeter. 4 Workhop on Marrian Meteorites

The precursor of the martian meteorites is thought to be ultra­ mafic igneous extrusive and/or intrusive rocks that were blasted from the surface of Mars during an impact event. Similar igneous rocks on Earth commonly contain fluid inclusions along with the more abundant silicate, sulfide, and carbonate melt inclusions. As such, it is reasonable to expect that the SNC meteorites should contain fluid inclusions that represent fluid trapped in the rocks while they were on the surface of Mars. A review of the recent literature failed to find any references to fluid inclusions in these well-studied samples, and a systematic search of available SNC meteorite samples was undertaken. Observations: Fluid inclusions have been foundin two thin sections of SNC meteorites: one in Nakhla (NSNM 5891-3) and ·,�···� � . one in ALH 84001,146. .. -:- -�-"-�, Nakhla (NSNM 5891-3). The inclusion in the Nakhla sample is tubular and -8 µm long. The inclusion is one of several dozen - inclusions forming a healed fracture that cuts through an ortho­ . . pyroxene crystal. Most of the inclusions along the fracture are dark with noticeable microfractures extending from the inclusion into ,- . ·:.�:. the surroundingmineral, indicating that the inclusions have decrepi­ . :·'. • 1:1�:.. ., � ... tated (exploded) owing to pressure buildup within the inclusion. This behavior is common for high-density liquid fluid inclusions Va�.1J3liibtiSl£t in terrestrial samples. A few other inclusions along the fracture -- were clear (transparent) but did not contain a visible bubble. This ' is because either the inclusions were empty, contained only liquid . with no vapor bubble, or contained a bubble that was motionless � - and hidden in a comer of the inclusion. The bubble in the one un­ ambiguous fluid inclusion in this fracture plane was in constant, slow motion, proving that the inclusion did indeed contain a liq­ uid phase. Although it was not possible to conduct the types of tests nor­ mally used to determine fluid inclusion compositions because of the manner in which the sample was prepared, the difference in index of refraction between the liquid and the vapor bubble is consistent with the inclusions containing liquid and vapor CO2, although it is possible that the inclusion contained liquid water and water vapor. ALH 84001,146. The second sample that contained liquid in­ clusions with moving bubbles was ALH 84001,146. The inclusions in this sample were spherical to negative-crystal shaped, about 1- 2 µm in diameter, and not along an obvious fracture (Fig. 1). The fact that the inclusions arenot along a fracture is importantbecause it suggests that the inclusions are primary (i.e., trapped during growth of the enclosing pyroxene). The fluidin the inclusions thus represents the magmatic fluid that exsolved from the crystallizing melt as the igneous rock formed on Mars. As such, the inclusions can provide valuable information about degassingearly in Marshis­ tory. As with the inclusion in Nakhla, no direct chemical tests could be conducted on these inclusions, but the optical behavior of the inclusions suggests that they contain liquid and vapor CO2•

Implications: The occurrence of (presumably) CO2 fluid in­ clusions in both the Nakhla and ALH 84001 meteorites suggests Fig. 1. Series of photorcrogaphs captured fom a videotape showing that CO2 was migrating through these igneous rocks at some time a fluid inclusion with a moving vapor bubble in the maruan meteorite ALH duration or after their formation on Mars. If the CO2 was present 8401,146. Te inclusion is about 1.8 µm in diameter. Note that the dark, at or near magmatic temperatures, pyroxene (enstatite) would not rounded vapor bubble, indicated by te aow in the third image fom te be stable in the presence of CO2, and would react to form a car­ top, is in a diferent position in each of the fve images. When examined bonate and a silica-rich phase [3]. This is consistent with a high­ under the microscope at rom temperatre, the bubble is in conunuous rapid temperature origin for the carbonates in the SNC meteorites, as Brownian motion fom the small thennal gradients caused by the micro­ originally proposed by Harvey and McSween [4]. This interpreta- scope light source. LPI Contribution No. 956 5

tion also suggests that the carbonates were present in the rocks be­ mation involving aqueous activity. An intermediate age, combined forethey were ejected fromthe martian surface. However, by anal­ with clearevidence of formationvia aqueousprocesses, would gain ogy with terrestrialsamples, the carbonatesdid not necessarilyhave special significance because it would extend the time period dur­ to formas aresult of the impact event but, rather, may have formed ing which water might have been relatively abundant on the mar­ as a result of natural high-temperature igneous metasomatic pro­ tian surface and available to support life. Carbonate formation via cesses before the impact. impact metasomatism or other high-temperature phenomena could References: [1] McKay D. S. et al. (1996) Science, 273, 924- be consistent with virtually any age, but also might be expected to 930. [2] Roedder E. (1984) Mineralogy, 12, 646. [3] Koziol A. M. be more prevalent at earlytimes. We have approached the difficult and Newton R. C. (1995) Am. Mineral., 80, 1252-1260. [4] Harvey problem of determining the age of ALH 84001 carbonates by iso­ R. P. andMcSween H. Y. Jr. (1996) Nature, 382, 49-51. topic analyses of a programmed sequence of carbonate-rich leachates. Procedure: A 1-g sampleof ALH 84001 was gently crushed RUBIDIUM-STRONTIUM FORMATION AGE OF ALLAN and sieved at 100-200and 200-300mesh. Grains containing car­ HILLS 84001 CARBONATES. L. E. Borg1, L. E. Nyquist1, bonates were hand-picked so that carbonateabundances in the high­ C.-Y. Shih2, H. Wiesmann2, Y. Reese2, and J. N. Connelly3, 1Mail graded sample approached 10%. A series of progressively more Code SN2, NASA Johnson Space Center, Houston TX 77058, reactive reagents were used to leach this high-graded sample;these USA, 2Mail Code C23, Lockheed Martin, 2400 NASA Road 1, reagents were predetermined by experimentation with terrestrial Houston TX 77258, USA, 3Department of Geological Sciences, carbonates. Small aliquots of each leachate were analyzed for Ca, University of Texas, Austin TX 78712, USA. Mg, Fe, K, P, Sr, Nd, and Pb by isotope dilution and atomic ab­ sorption. The remaining aliquots were spiked for Rb-Sr, Sm-Nd, Synopsis: Our preferred age forthe formation of carbonatesin and U-Pb and put through standard cation chromatography sepa­ the martian meteorite ALH 84001 is 3.90 ± 0.04 Ga forA.( 87Rb) = ration procedures. We report results of the Rb-Sr analyses of the 0.01402 Ga-1, or 3.85 ± 0.04 Ga for A.(87Rb) = 0.0142 Ga-1• This Ieachates; the other isotopic analyses are in progress. age is determined by a three-point Rb-Sr isochron defined by Results: The major-element analysesof the leachates are pre­ leachates of high-graded carbonate-richmaterial. Majorcation and sented in Fig. 1. The firstleachates have compositions that areun­ especially phosphorous analyses of the leachates permit contribu­ like any of the carbonates analyzed in ALH 84001 [e.g., 12). Sl tions from igneous whitlockite to be recognized for low-acidity and S2 (the first and second leachates) contain no Fe or P and lie leachates, and the corresponding data are omitted from the isoch­ on the magnesite-calcite join. The leaching agents used for S 1 and ron. Data forthe two highest acidity leachates plot close to the pre­ S2 were extremely weak and are likely to have primarily removed ferred isochron, but are omitted because we believe they contain surface coatings. The S3 leachate has significantly higher abun­ contributions leached from the pyroxene substrate on which most dances of Ca and P. Assuming Sr contents in ALH 84001 whit­ of the carbonates arefound. Nevertheless, the isochron age for all lockite aresimilar to those estimated fromwhole rock Ieachates of five highest-acidity leachates is 3.94 ± 0.04 Ga, and is within er­ the basaltic shergottite QUE 94201 from [13], mass balance cal- ror of the age obtainedfor the more restricteddata set. All leachates used to define the isochron have major cation compositions that are similar to those obtained by microprobe analyses of the car­ bonate rosettes and are consistent with progressive digestion of the carbonatesaccording to their composition. The age thus obtained for the carbonates is about 600 m.y. younger than the crystallization age of ALH 84001 determined by Sm-Nd analyses [1,2), but is within error limits of the age of im­ pact metamorphism inferred from the Rb-Sr and Ar-Ar systemat­ ics of silicates, which yield ages of 3.85 ± 0.05 Ga [3] and 4.05- 3.80 Ga [4] to 4.3-3.8 Ga [5], respectively. Similarities between the carbonate crystallization age and the age of impact metamor­ phism as determined by Ar-Ar and Rb-Sr suggest that the carbon­ u a on ate formation is impact-related. Nevertheless, both high and low­ (Mittlefehldt, 1994 temperature scenarios forthe origin of the carbonates are possible. Introduction: The origin of carbonateminerals in ALH 84001 ALH 84001 has received much attention since [6] reported that the carbonates Carbonates possibly contain evidence formartian microbial life. Several mod­ (Harvey& McSween,19 els have been proposed for the origin of the carbonates in ALH 84001, including (1) impact metasomatism [7], (2) impact melting of preexisting carbonates [8], (3) evaporation of surficial fluids at low temperature [9-10), and (4) low-temperature alteration by hy­ drothermal fluids [11). The time of carbonate formation has im­ Fig. I. Ternary diagram illustrating compositions of leachates from portant implications for martian planetary evolution as well as its ALH 84001. Note compositional similarities between carbonate compo­ possible biological evolution. An old age would be consistent with sitions(shaded area) from[7] and S4-S8 leachates.Also note compositional a wetter, warmer early Marsand thus with models of carbonatefor- differences between the carbonate compositions and Sl-S3 leachates. 6 Workshop on Martian Meteorites

culations based on P abundances suggest that -5 % of the Sr in this (1,2], but is similarto the age of impact metamorphism determined fraction is derived from whitlockite. The S4 leachate has a major­ by Ar-Ar andRb-Sr analysesof silicate phases [3,4]. This suggests element composition that is similar to the bulk carbonate compo­ a relationship between the impact event and carbonate formation. sition estimated by [13], but also has a significant amount of P. Scenarios involving direct formation of the carbonates via impact The maximum Sr contribution from whitlockite to the S4 leachate metasomatism [6] or impact melting of preexisting carbonates [8] can be calculated assuming that 100% of the Sr in the S3 leachate areclearly compatible with our data. However, another possibility is derived from whitlockite. In this case -30% of the Sr in the S4 is that ice buried beneath the martian surface was melted by heat leachate would be of igneous origin. This is obviously anextreme from the impact to produce water that percolated into the newly upper limit; mass balance calculations similar to those made for S3 formed basin and precipitated the carbonates, consistent with the suggest that S4 contains -1% whitlockite-derived Sr. The S5-S8 models of (8,9]. Alternatively, if Marswere warmer and wetter at leachates have no measurable P. That the leachates would become -3.9 Ga, a crater lake may have persisted fora time interval that progressively more Mg-rich with increasing acidity of the leach­ was short compared to the uncertainties of the radiometric ages. ing reagent was expected from experiments with terrestrial carbon­ Thus, the carbonates could have been precipitated at relatively low ates in which dissolution of calcite, then siderite, and then magnesite temperatures and still have had crystallization ages indistinguish­ was observed. able fromthe age of impact metamorphism. A Rb-Sr isochron diagram of our leachate data is presented in References: [1] Nyquist et al. (1995) LPS XXVl, 1065-1066. Fig. 2. The age of 3.90 ± 0.04 Ga is definedby our bestleachates (2] Jagoutz et al. (1994) Meteoritics, 29, 479-479. [3] Wadhwa (S4-S6). Blank contributions to Sr in all of the leachates were and Lugmair (1996) Meteoritics & Planet. Sci., 31, A145. <0.2%, with the exception of S8 (2.5%), which contained <1 ng [4] Turneret al. (1997) GCA, 61, 3835-3850. [5] Bogard and Gar­ of Sr. S4 lies on the isochron,suggesting that the whit-lockite con­ rison (1997) in Papers Presented to the Conference on Early Mars, tribution to its Rb-Sr systematics was indeed small. Those leachates pp. 10-12, LPI, Houston. [6] McKay et al. (1996) Science, 273, having compositions unlike the carbonates (S1- S2), a high P/Sr ra­ 924-930. [7] Harvey and Mcsween (1996) Nature, 382, 49-51. tio (S3), or a very low abundance of Sr (S8) lie off the best isoch­ [8] Scott et al. (1997) Nature, 387, 377-379. [9] Warren (1998) ron. The S7 and S8 leachates appear to lie on a mixing line between JGR, 103, 16759-16773. [10] McSween and Harvey (1998) the S6 leachate and our previous pyroxene analyses [l], suggest­ Meteoritics & Planet. Sci., 33, Al03. (11] Romanek et al. (1994) ing that they contain a contribution from pyroxene leached by the Nature, 382, 655-657. (12] Mittlefehldt (1994) Meteoritics, 29, strong reagents used. Furthermore, visual examination of a control 214-221. [13] Borg et al. (1997) GCA, 61, 4915-4932. suite of carbonates indicated that all were dissolved by the sixth step. The calculated age range fortwo (S4-S5; 3.89 Ga), three (S4- S6; 3.90 ± 0.04 Ga), four (S4-S7; 3.93 ± 0.04 Ga), or five (S4- S8; 3.94 ± 0.04 Ga) data points is -3.9-4.0 Ga. Our preferredage RARE POTASSIUM-BEARING MICA IN ALLAN HILLS of 3.90 ± 0.04 Ga is that determined by S4-S6. 84001: ADDITIONAL CONSTRAINTS ON CARBONATE Discussion: The age of carbonate formation in ALH 84001 FORMATION. A. J. Brearley, Institute of Meteoritics, Depart­ is substantially younger than the crystallization age of 4.50 ± ment of Earth and Planetary Sciences, University of New Mexico, 0.13 Ga derived from Sm-Nd analyses of silicates and phosphates Albuquerque NM 87131, USA ([email protected]).

Introduction: McKay et al. [1] presented several intriguing observations suggesting evidence of fossil life in martian ortho­ pyroxenite ALH 84001. These exciting and controversial observa­ 0.85 tions have stimulated extensive debate over the origin and history ALH84001 Carbonate Leachates of ALH 84001, but many issues still remain umesolved. Among the most important is the question of the temperature at which the carbonates, which host the putative microfossils, formed. Oxygen­ ._ 0.80 isotopic data, while showing that the carbonates are generally out Cl) of isotopic equilibria with the host rock, cannot constrain their tem­ ,.._Cl) perature of formation [2-4]. Both low- and high-temperature sce­ a, narios are plausible depending on whether carbonate growth 0.75 • Leachates S4,S5,S6 o Leachates S1-S3,S7.S8 occurred in anopen or closed system (2-4]. Petrographicarguments I:. Whole r + silicates have generally been used to support a high-temperature origin [5], (Nyquist et al .• 1995) but these appear to be suspect because they assume equilibrium be­ 0. 70 "--L-"'--'---'---'------'---'�,___.___.___.__.____.__._L--J tween carbonate compositions that are not in contact. Some obser­ 0 3 vations appear to be consistent with shock mobilization andgrowth from immiscible silicate-carbonate melts at high temperatures [6]. Proponents o� a low-temperature origin for the carbonates areham­ pered by the fact that there is currently no evidence of hydrous Fig. 2. Rubidium-stontium isochron plot of leachates fom ALH 84001. Our preferred age of 3.90 ± 0.04 Ga is defined by S4-S6; leachates S4- phases that would indicate low temperatures and the presence of S8 defne an age of 3.94 ± 0.04 Ga. Leachates used to defne the isochrons a hydrous fluid during the formation of the carbonates. However, have major cation compositions similar to those of the carbonates (see the absence of hydrous phases does not rule out carbonate forma­ Fig. 1). tion at low temperatures, because the carbonate forming fluids may U'l Contribution No. 956 7

have been extemely COrrich, such that hydrous phases would not have been stabilized. Although phyllosilicates are appaently not present at the scae observable by SEM, they may be present at the submicrometer scale. However, severa detailed TEM studies of ALH 84001 have failed to find evidence of phyllosilicates [l,7]. In this study, I have caied out additional TEM studies of ALH 8401 and have fund evidence of very rae phyllosilicates, which appear to be convinc­ ingly of preterestrial origin. At present these observations ae lim­ ited to one occurence; further studies are in progress to determine if the phyllosilicates ae more widespread. Reults: Trasmission electon mcroscopy studies so far have concentated on fagments of caonate entained within fldspathic glass throughout ALH 8401. The caonate fagments have irregu­ la shaps, but examination at high magnification using BSE and TEM imaging show that may grans are faceted ad these facets ae cleavage surfaces. This obseration suggests that these grains ae fagments of lager, continuous regions of cabonate that have been factured, disrupted ad entained within the fldspathic melt during a post-carbonate frmation impact ad heating event. This is aso consistent with the observauon that adjacent fagments, sepa­ rated by regions of glass, have zoning pattess consistent with their originally being pat of a single grain. BSE images of the fagments show that they have a rage of sizes fom -50 µm down to <10 µm. TEM observations [8] show that many of these carbonate fag­ ments, despite having relatively Mg-rich compositions, contain myriad magnetite particles associated with voids. These magnetite grains have grain sizes ad morphologies consistent with the mag­ netites observed by [1] and other workers [6]. A single carbonate fagment within the fldspathic glass con­ tains the phyllosilicates. They occur as highly elongate, parallel to subparallel ribbonlike crystas, typically <10 r thick and extend­ ing fr several hundreds of nanometers through the cabonate grain. The abundace of phyllosilicates within the fagment is difcult to estmate, but is probably of the order of 10-20 vol%. It is no­ Fig. 1. High-resoluton TM imge of K-bang mica fom a carbonate table that te phyllosilicate grains ae aways tuncated at the edge fagment in AH 841. of caronate grains and never extend into the feldspathic glass. It also appas that close to the interface with the glass the phyllo­ silicates have cured morphologies, ad some of the smallest grains appea to be amorphous. These obserations ae consistent with intergrown with cabonate. Second, the mica appeas to have been thermal decomposition during the postshock heaung that fllowed heated and partially decomposed where it is in contact with the fagmentation ad entrainment of the cabonate fagments within glass, again showing that it must have predated glass formation. the feldspathic glass [8,9]. Electon diffaction and high-resolution Third, terestia weathering products of meteorites ae porly crys­ TEM studies of the phyllosilicates show that they have a baal spac­ talline clay minerals with an extremely fne-grain size, not well­ ing of 1 r and appear to be well-ordered, with no evidence of ordered micas. Fourth, the carbonates appear to be completely stacking disorder or partial interlayers. The composition of the entained within the feldspathic glas ad consequently would have grans ha not yet been obtained with ay great certainty, because been isolated fom altering terestial fuids. The reasonable con­ of their very limited thickness and the fact that they ae closely clusion fom these observations is that the mica formed prior to intergrown with the carbonate. However, EDS spectra show the ejection of ALH 84001 fom Mas. presence of Al, Si, ad K, in addition to Mg, Ca, and Fe fom the Although phyllosilicates are not widespread in ALH 8401, the adjacent cabonate. The AlSi ratio appars to be relatively low. presence of mica clearly indicates that water-beaing fluids were These observations appea to be consistent with this phyllosilicate present at some time during the history of ALH 84001. The ver phase being a K-beaing mica, such as illite. close intergrowth of cabonate with the mica suggests that the two Iplications: A preterestial origin fr the mica is indicated phases grew simultaneously or that the mica may have preexisted by several lines of evidence. First, the mica occurs exclusively before carbonate frmation. Replacement of preexisting carbonate within the carbonate ad does not extend into the surounding fld­ by long, thin ribbons of phyllosilicates does not seem probable. The spathic glass, clealy showing that it pedated frauon of the glass. very elongated morphology of the mica grains does not seem con­ If mica had frmed after glass frmauon, it would most likely have sistent with unimpeded growth fom a solution; rather, it suggests formed by replacement of the feldspathic glass, rather than growth constained by adjacent growing cabonate grains (i.e., the 8 Workshop on Martian Meteorites

two phases grew contemporaneously). The presence of illite and that is, they formed after the end of heavy bombardment, but still its formation from smectiteis widely used as a geothermometer in in the first half of Mars' history. Around the Chryse Basin, roughly terrestrialpelitic rocks that have undergone diagenesis to low-grade 4 x 106 km3 of rock wasremoved, which likely required at least metamorphism. However, this approach has been questioned, and l 07 km3 of water, or 70 m spread over the whole planet.Additional it may be the case that both illite andsmectite are metastable phases outflowchannels aroundElysium, Hellas,and Amazonis bring the with respect to muscovite but form according to the predictions of total volumes of floodwaters to over 100m. This is only the wa­ Ostwald's step rule [9]. As a consequence, illite can occurunder ter brought to the surface in floods.The most likely causes of the diagenetic conditions and up to temperatures in excess of 250°C. outflow channelsare catastrophic eruption of groundwater or cata­ However, at temperatures >~300°C, muscovite would be expected strophic release of groundwater-fedlakes, as, for example, in the to be the stable K-bearing mica and should formbecause tempera­ canyons. This being so, then the 100 m probably represents only a tures are sufficiently high for equilibrium to be attained. This sug­ fraction of the total near-surface inventory. Most of the water re­ gests that mica formation in ALH 84001occurred at temperatures tained by the planet at the end of heavy bombardment probably re­ probably <250° ± 50°C. At low temperatures, such as occur dur­ mained in the ground as ice or water at greater depths. ing diagenesis, clay minerals are extremely fine-grained and are The fate of the floodwaters is controversial. One suggestion is complex submicroscopic intergrowths of different layer lattice min­ that the floods formed under present climatic conditions and that erals such as smectite, illite, and chlorite, which are far from ther­ each flood left a lake that immediately froze over and ultimately modynamic equilibrium. The mica grains in ALH 84001 show no froze solid [1,2]. Successive floods would then have built thick, evidence of this type of interlayering, and hence indicate that they layered, permanent ice deposits in the low-lying northern plains most likely formed at temperatures >150 ° ± 50°C. If the mica where most of the floodsterminated. This water was permanently formedcontemporaneously with the carbonate as is indicated by removed fromthe global aquifer system so that floodingultimately the textural relations, then carbonate formation conditions between ceased.An alternativesuggestion [3] is that ocean-sized(> 107 km3) l00°-300°C are indicated. While this is a broad rangein tempera­ bodies of water periodically accumulated in the low-lying north­ ture, it rules out a high-temperature origin for the carbonates pro­ ern plains, temporarily altering global climates. The oceans ulti­ posedby some workers. mately dissipated, leaving behind shorelines [4,5] as evidence of Acknowledgments: This research was supported by NASA their former presence. Some support for the shoreline hypothesis grant NAG5-4935 to A. J. Brearley, principal investigator. is the similarity in elevations of some sections of the proposed References: [1] McKay D. S. et al. (1996) Science, 273, shorelines, but the lack of detection of carbonates by Mars Global 924-929. [2] Romanek C. S. et al. (1994) Nature, 372, 655- Surveyor [6] is difficultto reconcile with major climate excursions 656. [3] Valley J. W. et al. (1997) Science, 275, 1633-1638. as late as late Hesperian. [4] Leshin L. et al. (1997) GCA, 62, 3-13. [5] Harvey R. P and Martianmeteorites appearto contain water (and other volatiles) McSween H. Y. Jr. (1996) Nature, 382, 49-51 [6] Scott E. R. D. from two sources, a magmatic source and an atmospheric source et al. (1997) Nature, 387, 377-379. [7] Bradley et al. (1996) that has experienced isotopic fractionation, probably as a conse­ GCA, 60, 5149-5155. [8] Brearley A. J. (1998) LPS XXIX, Ab­ quence of losses from the upper atmosphere. This suggests that the stract #1451. [9] Shearer C. K. and Brearley A. J., this volume. atmospheric volatiles were mixed deep into the crust. Under present [10] Essene E. J. and Peacor D. R. (1995) Clays Clay Minerals, climatic conditions, CO2 could diffuse from the atmosphere into 43,540. the ground and be fixedas carbonate, but movement of water from the atmosphere into the ground is more difficult becauseof the thick cryosphere. One possibility is precipitation accompanying climate change, as suggested by the episodic oceanhypothesis, but this pre­ WATER ON MARS: HOW MUCH WAS (IS) THERE, sents other problems and is inconsistent with the low post-heavy­ WHERE DID IT GO, AND WHEN? M. H. Carr, U.S. bombardment erosion rates. Hydrothermal activity has been sug­ Geological Survey, 345 MiddlefieldRoad, Menlo Park CA 94025, gested, but this would be unlikely to be effective in introducing USA ([email protected]). atmospheric water into the ground, given a thick cryosphere. Al­ ternatives arepolar basal melting and reintroduction of water (ice) Optimism about the prospects for development of indigenous into ground at low latitudes during periods of high obliquity. life on Mars has been fueled in large part by perceptions of the The case for a major climate change early in Mars' history is geologic evolution of Mars based mainly on remote-sensing data. much strongerthan the case for changes afterthe end of heavy bom­ The consensus view is that Mars is, or was, a water-rich planetand bardment. Evidence for early climate change includes (1) the al­ that it likely experiencedwarmer climates in the past. Yet this view most universal presence of valley networks in Noachianterrains, has not been proved, and it is difficult to reconcile with most mod­ (2) the dramatic change in erosion rates near the end of heavy bom­ eling studies of climate change and some characteristics of mar­ bardment, and (3) the nonequilibrium distribution of water at the tian meteorites. Moreover, many variants have been proposed for end of heavy bombardment (sources of valleys at high elevations). the climate history of the planet and for the amounts and locations The lack of detection of extensive carbonate deposits is, however, of water present at different times. troublesome. A possible way out of the dilemma is that early in The evidence that water played a major role in the history of the era of heavy bombardment Mars had a large inventory of CO2 Mars, while circumstantial,is strong. A variety of fluidsother than and a thick COz!H20 atmosphere.Loss of CO2 fromthe atmosphere water have been suggested as the erosive agents that cut the large by impact erosion [7] and volcanic burial was partly offset by re­ outflowchannels, but water is by far the most compelling and is cycling of CO2 from carbonates into the atmosphere as a conse­ consistent with the presence of valley networks and numerous pos­ quence of rapid burial and high heat flows [8]. However, as heat sible indicators of ground ice. Most outflowchannels are Hesperian; flow and the total inventory of CO2 declined, recycling could not LPI Contribution No. 956 9

keep pace with losses by impact. As a result the atmosphere was THE SEARCH FOR EARLY LIFE ON MARS: A EURO­ almost entirely lost through impact erosion in late Noachian, and PEAN SPACE AGENCY EXOBIOLOGY SCIENCE TEAM 1 2 no carbonateswere left at the surface. STUDY. P. Clancy , A. Brack , B. Hofmann3, G. Horneck4, G. 5 6 9 We can conclude that Mars emerged fromheavy bombardment Kurat , J. Maxwell , G. G. Ori', C. Pillinger', F. Raulin , N. 1 12 1 with at least a few hundred meters of near-surface water, of which Thomas10, F. Westall1 , andB. Fitton , European Spate Agency, at least 100 m flowed across the surface in the large floods. The Noordwijk, The Netherlands, 2Centrede Biophysique Moleculaire, fate of the flood waters is uncertain. Some was lost from the upper Centre National de la Recherche Scientifique, 3Naturhistorisches atmosphere, some may be in the polar layered terrain, and some Museum, Bern, Switzerland, 4Institute of Aerospace Medicine, may be in permanent ice deposits in the low-lying northern plains Deutsches Zentrumfi1r Luft-undRaumfahrt, Porz-Wahn, Germany, and elsewhere. This entire near-surface inventory may have been 5Naturhistorisches Museum, Wien, Austria, 6Chemistry Department, enriched in D as a consequence of massive losses during hydrody­ Bristol University, Bristol, UK, 'Dipartimento di Scienze, Uni­ namic escape andthe earlier warmclimates. versita d'Annunzio, Pescara, Italy, 'Planetary Sciences Research References: [1] Carr M. H. (1996) Water on Mars, Oxford. Institute, The Open University, Milton Keynes, UK, •centre [2] CarrM. H. (1996) Planet. Space Sci., 44, 1411-1423. [3] Baker National de la Recherche Scientifique, Universites de Paris 7 et 12, 10 V. R. et al. (1991) Nature, 353, 589-594. [4] Parker T. J. et al. Creteil, France, Max-Planck-Institut fur Aeronomie, Lindau, (1989) Icarus, 82, 111-145. [5] ParkerT. J. et al. (1998) JGR, 98, Germany, 11Dipartimento di Protezione Agrolimentare, Universita 11061-11078. [6] Christensen P. R. et al. (1998) Science, 279, degli Studi di Bologna, Italy (currently at Planetary Sciences 1692-1698. [7] Melosh H.J. and Vickery A. M. (1989) Nature, Branch, NASA Johnson Space Center, Houston TX, USA), 338, 487-489. [8] Pollack J.B. et al. (1987) Icarus, 71, 203-224. 12European Science Consultants,Sassenheim, The Netherlands.

Introduction: A recent initiative of the Manned Spaceflight ON INDEPENDENTPROOF OF A MARTIAN ORIGINFOR and Microgravity Directorate of the European Space Agency was THESHERGOTTY METEORITE. Yu. T. Chuburkov1, V. P. concernedwith the identification of specificobjectives in the search Perelygin1 , and V. A. Alekseev2, 1Joint Institute for Nuclear for life on Mars. The Exobiology Science Team, chaired by Andre Research, 141980Dubna, Russia, 2Troitsk Institute of Innovation Brack, consisted of experts in the fields of radiation biology, plan­ and Thermonuclear Studies, 142092 Troitsk, Russia. etary geology, geochemistry,mineralogy, meteorology, and exobi­ ology. The Science Team study emphasized present European The Mendeleev's Periodic Law was used to obtain proof of the capabilities in the areas of drilling, sample acquisition, and prepa­ existense of the hypothetical process of element magnetic separa­ ration, andin situ analysis. tion in protoplanetary nebula. Objectives: Objectives of the initiative were (1) to determine

The share of free neutral atoms, N0, forall elements in proto­ which surface and near-surface environments of Mars might pro­ planetarynebula has been determined with anaccount of their abun­ vide the best evidence for past life and to establish a short list of dance and physical and chemical properties.The lineardependence preferred landing sites fora future exobiologymission; (2) to de­ for the ratio of nonvolatile and volatile element contents in igne­ sign an integrated set of analyticalinstruments with which to search ous rocks of the Earth (R = 1 AU) up to the asteroid belt (R = for evidence of past or present microbial life in subsurface materi­ 2.80 AU) on N0 was obtained. als and rock samples on Mars; and (3) to develop a set of imaging To this end the concentration ratios of elements - analogs and spectroscopic systems to facilitate a search forevidence of ex­ with different N0 in the matters of Venus, Earth, Mars, and chon­ tinct microbial life at all scales down to 0.01 µm. These systems drites - were compared. The data obtained are sufficient forthe should also provide forthe study of the mineralogy and petrogra­ demonstrationof the existence of the process of element magnetic phy, as a function of depth, in the near subsurface region of Mars. separation in protoplanetary nebula. The concentration ratio ZifZ2 Landing Sites: Landing sites in areas that might have sup­ for some pairs of chemicallysimilar elements is presented in Table 1 ported life should be chosen on the basis of (1) the inferred pres­ forthe Earth (IR), Mars(Ms), Shergotty meteorite (Sh), and chon­ ence of sediments, especially in craters; (2) the partial excavation drites (Ch). of sediments by younger impacts to allow access to buried sedi­ ments; (3) stratifiedcanyon walls (possible sediments); (4) chaotic terrains indicating the presence of water and possibly hydrother­ mal activity; (5) high albedo features in craters (possible evapor­ TABLE 1. Data showing thatShergony meteorite element composition ites); (6) ground ice (cratered); (7) polygonal features (sediments/ is much closer to Mars than chondrite meteorites. ice?); and (8) terracing(water flow). In the search for subsurface microfossils, the following rock NIN Z/2-i IR Ms Sh Ch types aresuitable candidatesfor examination: (1) highly porous or Mg/Ca 0.6 1.3 0.9 10 diagenetically cemented sedimentary rocks, (2) voids in impact brec­ 2 Na/K 5 8 cias and impact melts, (3) vesicles in volcanic rocks, and (4).frac­ 3 Ca/Sr 91 530 170 103 tured rocks of any type. Microfossils are most likely to be found 4 Ag/Rb 8 X 10-4 2 X )0-3 J X )0-2 in rocks that have been buried to a considerable depth and have 2 5 Mg/Sr 62 7 xl0 2 X 103 3 x IO' resurfaced due to impact ejection or are exposed on canyon walls. 6 Fe 2 X l04 106 2 X 107 Drill and Sample Distribution System: This is an important 7 lr/K 5 X 10-8 2 X 10- 7 6 X J0-4 feature of the Exobiology Package and will utilize various Euro­ 8 Mg/Ba 55 J X 103 2 X J0-4 pean technological developments in drills, moles, penetrations, and 9 Zn/Ba 0.2 0.4 10 sample distribution systems. 10 Workshop on Martian Meteorites

Programmatic Plans fr an ESA Exobiolog Package: Fol­ Recommended Analytical lowing the recommendation of the Science Team, a nine-month Instrument Package Phase A study is about to start. This phase will cover a technical assessment at system cd subsystem level of the required develop­ ments to produce miniaturized, space-qualifed versions of drilling Microscope: fr examination of samples and sample acquisition systems, as well as miniaturized versions • 3-µm resolution of existing analytical instuments or those in development stages. • mass, 250 g The study will also cover the elaboration of a system-integration Infrared spectroscopy: molecular analysis of minerals cd approach and the mecs to clarify interface issues. Moreover, ESA organics (with Raman) will propose a program to member states of the Europec Com­ • wavelength range, 0. 8-10 µm munity within the framework of the extension of its current • spectral resolution, >100 (1/Dl) Microgravity Programme (to start in 2000), which will include an • spatial resolution, 200 µm element for the development of the Exobiology Package. • expected mass,

following: liquid water availability, temperature, source of meta­ etary exploration by examining the fossil records of key "extreme" bolically usefulenergy, destructiveenergy levels, and temporalsta­ ecosystems that inhabit hydrothermal and/or subsurface environ­ bility of conditions that support growth. How specifically do cell ments and by investigating evaporite and ice deposits. membranes, transport systems, enzymes, and the processes of in­ Exploration efforts will benefit from remote-sensing studies of formation processing, storage, and reproduction cope mechanisti­ those sites and samples that represent meaningful analogs of ex­ cally with extremes in these environmental parameters? How, if at traterrestrial exploration targets (e.g., minerals deposited in hydro­ all, do the dynamics of biological evolution differbetween extreme thermal systems or evaporatinglakes). Indeed, the most significant and"moderate" environments? In the context of the original envi­ rock types might be both relatively rare and hard to recognize. It ronment in which life arose,just exactly which environments are will be importantto cope effectively with mineral mixtures and spa­ extreme- that is, which environmentspresented the greatest adap­ tial heterogeneities in the target. These imperatives create specific tive challenge for life? needs for the development of new technology, instrumentationand Planetaryscience is centrally relevantto the study of life's ori­ methods fordata processing and interpretation. Space exploration gin andsurvival because biospheres must evolve to accommodate will indeed benefit enormously from thoughtfully executed "mis­ environmental changes driven by the normal evolution of silicate sions," both to Earth's biosphere and rock record, as well as to our planets. Because silicate planets followparallel paths of evolution collections of astromaterials. that should produce more or less parallel consequences for their Relevant Literature: surface environments, studies of Earth's long-term habilitability Des Marais D. J. (1994) The Archean atmosphere: Its composi­ help us understand the prospects for life elsewhere. Key are those tion and fate. In Archean Crustal Evolution (K. Condie, ed.), changes in relative arealextent of the particular environmentsthat pp. 505-523. Elsevier, Amsterdam. tend to support relatively high levels of biological activity (e.g., FarmerJ. D. et al. (199S) Site selection forMars Exobiology. Adv. hydrothermal systems, shallow-water coastal environments, etc.). Space Research, 15(3), (3)157-(3)162. Long-term changes in relative rates of processes such as volcan­ Jakosky B. M. (1998) The Search for Lie on Other Planets. Cam­ ism, weathering, and impacts have altered the chemical composi­ bridge Univ. tions of the oceans and atmospheres; thus, these compositional Meyer M. and Kerridge J., eds. An Eobiological Strtegy for Mar trends have been "emergent properties"of planetary evolutionthat Eploration, NASA SP-530. directly affected the biosphere. Excellent examples of such effects Schopf J. W. andKlein C., eds. (1992) Te Proterozoic Biosphere: arethe roles played by planetary and biological changes in the evo­ A Multidisciplinar Study. Cambridge Univ., New York. lution of the redox state of the oceans and atmospheres and how Walter M. R. and Des MaraisD. J. (1993) Preservationof biological this evolution might have affected the development of multicellu­ information in thermal spring deposits: Developing a strategy larlife (plants, animals). Because silicate planets share many key for the search for fossil life on Mars. Icarus, 101, 129-143. long-term changes,parallel evolutionary phenomena certainly could Walter M. R., ed. (1996) Evolution of Hydrotheral Ecosystems occur elsewhere. on Earth (and Mars?). Wiley, Chichester (Ciba Foundation Earth-based research on the molecular and ecological mecha­ Symposium 202). nisms of evolution are relevantto planetary exploration, at least to the extent that they ultimately identify those particular planetary environments that were critical to our biosphere's earliest devel­ OPTIMIZINGTHE SEARCH FOR A FOSSILRECORD OF opment. For example, the confirmation that thermal environments ANCIENT MARTIAN LIFE. J. D. Farmer, Department of indeed were crucibles for our own biosphere's origin and/orearly Geology, Arizona State University, Tempe AZ 85287-1404, USA development could substantially influence our site selection strat­ ([email protected]). egy for Marsexploration. Effortsto understand the extent and diversity of the fossilrecord Introuction: An importantfocus of the decade-long Mars Sur­ of early life carry immediate andimportant benefits for planetary veyor Program is the searchfor evidence of pastor present lifeand/ exploration. Our ability to detect fossils is directly enhanced by or prebiotic chemistry. Based on studies of terrestrial analogs, the improved systematicsfor interpreting the multiple categories of fos­ highest-priority sites fora martianfossil record include such litholo­ sil information, namely textures of cells and their excretion prod­ gies as fine-grained, clay-rich detrital sediments, water-lain pyro­ ucts, rock fabrics created by chemical activity and movements of clastics and evaporite deposits of terminal paleolake basins, the biota, compositions of minerals andorganic matter, and stable iso­ deposits of mineralizing springs(including hydrothermal), and min­ topic compositions of those elements affected by biological pro­ eralized zones (hard-pans)within ancient soils._ The systematic ex­ cesses. Biological vs. nonbiological features mustbe distinguished, ploration for a martian fossil record critically depends on locating even after those features have suffered physical and chemical accessible surface outcrops of these and other aqueously deposited changes subsequent to their depositon (such studies are termed sediments. The following discussion identifies the steps needed to "taphonomy"). These changesare driven by mineral deposition and optimize a plan for implementing the present strategy for Mars recrystallization as well as by alteration by thermal processes and/ exopaleontology. or by fluids that invade rock fracturesand pore space. Should ster­ Mineralogical Mapping from Orbit: Mineralogy is key for ilization be applied to returned samples, it would be yet another accurately assessingpast geological environments andselecting and agent of alteration whose consequences must be understood. We prioritizing landing sites. And fromthe standpointof preservation, need to understand the "taphonomy of planetary protection proce­ mineralogy provides the most logical basis for selecting samples dures!" Earth-basedstudies of life can also be made usefulfor plan- for return to Earth. The simple fact is that fossil preservation is 12 Workhop on Martian Meteorites

strongly biased toward particular geological environments, and de­ copy) have distinct advantages over more precise methods of iden­ posits indicate that site selection base on mineralogy is fundamen­ tification (e.g., X-ray diffraction) that require a powdered sample. tal. In selecting the best sites on Mars for a sample return for Spectral methods that combine mineralogical and organic analyses exopaleontology, we logically begin with orbitaldata to enable us hold the greatest advantages in selecting samples for to narrow the vast number of potential sites. Given the premium exopaleontology. placed on mineralogy in selecting targets, spectral mapping over a Sampling Rock Interiors: Viking and Pathfinder provided wavelength range usefulfor determinative mineralogy is requisite. elemental abundancedata (X-ray fluorescence and APXS respec­ The practical problem is to be able to isolate and identify key min­ tively) for comparatively young sites on the northern plains. Vi­ eralogical signatures within complex mixtures and to specify the king analyses were obtained from wind-transported surface fines actual spatial location of outcrops within targeted terrains.The limi­ and probably had little to do with the local bedrock geology. Path­ tations presently imposed by landing errorsand rover mobility un­ finder advanced our understanding somewhat by providing elemen­ derscore the importance of high-resolution mineral maps in the tal data fora variety of rocks that were transported to the site from selection of landing sites. This is ostensibly the most crucial step a variety of highland sources during largeoutfloods and/or as ejecta leading to a targeted sample returnfor exopaleontology. from local impact craters. But in spite of the seemingly wide vari­ ImprovedLanding Precision and Rover Mobility: Once de­ ety of source materials, the general uniformity of elemental analy­ posits of interest have been identified from orbit, accessibility on ses obtained fromrocks and soils at the Viking and Pathfindersites the ground will require improvements in landingprecision and rover suggests surface weathering on Marscan mask true compositional mobility. At a minimum, landing errors need to match the mini­ differences. This underscores the importance of gaining access to mum traversedistances required for rovers to reach their intended unweathered rock interiors by breakage, abrasion, coring, or other targets within nominal mission times. The landing precision and means. Furthermore, because future sample return payloads will mobility requirements will vary with each mission, depending on likely be very small (a few hundred grams),the ability to subsample the size of the target deposits and terrain trafficability. Based on larger rocks will be required if we are to return the most promis­ terrestrial standards, targets could be quite small, requiring land­ ing samples to Earth. This will necessitate a great deal of flexibil­ ing errorsand equivalent rover mobilities on the order of 5-10 km ity in rover sample acquisition and handling systems. to allow rovers to reach deposits of interest. A practical example Compromises That Make Sense: In selecting potential sites of the connection between the spatial resolution of orbital mapping for a sample return mission, it is important to realize that while data androver mobility is provided by the TES data to be obtained science needs to drive missions, from a practical standpoint, the by the Mars Surveyor orbiter. During subsequent surface missions, constraints imposed by budgets and politics ultimately determine rover traverses will likely all be conducted within a single TES the missions that actually fly. Sample return is one thing. But if pixel. There is also an obvious synergy between improvements in the question of martian life is to remain the focus, then each op­ the spatial resolution of orbital data andthe need for enhanced ro­ portunity must respond to these questions: How canwe best adapt ver mobility. Given that extensive improvements in rover mobility to budgetary challenges while optimizing our science goals? And during MGS seem unlikely, it is all the more important that high what are the acceptable fall-back positions that will still make mis­ spatial-resolution (outcrop-scale) spectralmapping databe obtained sions worth the cost and effort?I submit that compromise positions from orbit as earlyas possible in the MGS program. that settle for soils instead of rocks make no sense in the context Descent Imaging,Compositional Analysis at a Distance, and of exploring forsignatures of past life. And if we must rely on de­ Rover Autonomy: Once on the ground, to optimize planning of livery systems with large landing errorsand rovers of limited range, rover traverses, it will be necessary to specify the highest priority then we should optimize our understanding of surface geology and targets within a site. In this context, descent imaging could be a composition by using each orbital opportunity to progressively en­ great asset in mission planning. Analysis by rovers can beconducted hance the spectral range and spatial resolution of our orbital map­ at two levels. In planning traverses anddefining optimal sampling ping efforts. Once we know where the most interesting targets are strategies, spectral imaging from a distance can be used to survey located, we can design the missions that will optimize our chances rock fields, targetingthe most interesting rocks for close-up min­ of sampling a broad rangeof compositions and ages relevant to the eralogical analysis using rover instruments.Given the distance and searchfor past life. signal delay times between Earth and Mars,real-time telepresence will not be possible. But improvements in rover autonomy may en­ able multiple science objectives to be achieved during single com­ BACTERIAL MINERAL PRECIPITATION AND THE mand cycles. Actual traverse planning is likely to be accomplished MAKING OF MICROFOSSILS. F. G. Ferris, Department of by Earth-based scientists using virtual terrain models constructed Geology, Earth Sciences Centre, University of Toronto, 22 Russell fromrover and lander imaging of the site. Street, Toronto, Ontario, M5S 3B l, Canada (ferris@quartz. Close-Up Imaging and In Situ Spectroscopic Analysis: As geology. utoronto.ca). a first stepin analyzing rocks in the field,geologists normally rely on microscopic (hand lens) imaging of rock surfaces to obtain An important prerequisite forthe precipitation of minerals from microtextural information.This enables identification of rock type aqueous solutions, even where bacteria are involved, is that a mod­ andqualitative mineralogical information.But forrobotic missions, erate degree of oversaturation must be achieved. This requirement microscopic imaging can also provide important contextual infor­ is imposed thermodynamically by an activation energy barrier that mation for more spatially targeted compositional analysis using constrainsthe spontaneousformation of insoluble precipitates from rover instruments.Because of the minimal sample preparation, spec­ solution. Bacteria intervenein mineral precipitation reactions in two tral reflectance methods (e.g., infrared and laser Raman spectros- ways, either directly as catalysts of aqueous geochemical reactions LPI Contribution No. 956 13

or indirectly as chemically reactive solids. In the first case, bacte­ .. rial metabolic activity is often significant and can trigger changes . : ... in solution chemistry that lead to oversaturation (e.g., through the production of reactive ligands like sulfide). This alone can induce ' ·:; 'chromites, and image-stack along a single pixel or over a region of interest by add­ containing only minor amounts of the carbonate, magnetite, and ing the pixels corresponding to that region in each image. At each sulfidethat are major components in the rims described by McKay energy, the referenceintensity (on the substrate)and sample inten­ et al. [2]. sity are measured on the same image, rather than in two succes­ Percent-level organic C was associated with both the carbon­ sive spectra, which are sometimes offset slightly in energy due to ate globule and the opaque material. The C-XANES spectrum of nonreproducibilities in the monochrometer motion (a problem be­ the organic compound in the globule was different from that in the ing addressed in a STXM redesign). opaque material [1]. FfIR spectroscopy indicated the presence of C-XANES Results: Figure1 shows the STXM image of the aliphatic hydrocarbons, having different ratios of C-H2 to C-H3 entire sample.The highest-quality spectra are obtained by averag­ groups, in the two samples [1]. ing over the largest number of pixels on the sample. Figures 2-4 New Sample: To follow up on those measurements, a car­ show C-XANES spectra averaged over the rim material, the glob­ bonate globule with attached rim material was extracted from a ule interior, and the region containing coarse-grained, porous car­ freshlybroken surface of a chipof ALH 84001 (ALH 84001,255). bonate and fine-grainedmagnetite that separates the globule interior This sample was embedded in elemental S, a series of ultrarnicro- fromthe rim. 14 Workshop on Martian Meteorites

Pe-edge (?*) absorptions occur at photon energies corespond­ 0.8 ing to induced electon tansitions fom the K-shell to unoccupied outer orbitals whose energies ae very sensitive to the types ad 0.6 locations of neighboring atoms. The rim material showed fur strong 1* peaks (Fig. 2). Three paks, at 285 eV, 286.2 eV, ad 0 288 eV, are similar to the peaks at 284.8 eV, 286.5 eV, and 0.4 � 288.2 eV detected in the cabonate globule fom ALH 8401 ex­ vi amined previously [1]. Those paks were associated with orgaic C by FI spectoscopic examination of that globule [I]. The fuh 0.2 peak, at 290 eV, is indicative of carbonate. The globule material (Fig. 3) and the porous cabonate (Fig. 4) sepaating the rim fom the globule interior show the same fur 1* peaks. 0,0,.__���-������������ In each C-XANES spectrum we can measure the ratio of the 270 280 290 300 310 320 absorption at 290 eV to that at 288 eV to monitor the ratio of car­ bonate to organic C. Comparing the average C-XANES spectrm Fig. 2. C-XANES spectrum averaged over the rim of the ALH 84001 sample shows three n* peaks, at 285 eV, 286.2 eV, and 288 eV, indicative over the cabonate globule with the average over the rim indicates of organic C and a peak of 290 eV from carbonate. that the rim has a higher rato of organic C to cabonate than does the globule. The porous carbnate appeas to have approximately the same ratio of orgcic C to carbonate as the globule interior. The highest ratio of orgaic C to cabonate was fund in iso­ lated spots within the rim. Thus far, however, we have been un­ 1.0 able to locate regions in the rim that show only the organic or only t the cabonate absorption fature(s), suggesting that organic C is 0.8f- -{ intimately mixed with the caonate on the scale of -100 nm in the rim. Conclusions: The sae three organic absorption paks occur 0 0.6 1 C with roughly the same pak height ratios in both the rim and the O' j vi j globule. This indicates that, in this sample, the rim and the glob­ 0.4 ule contain the sae type(s) of organic compound(s). Analyses of i individual caonates within the globule showed weak, but distinct, 0.2 organic absorptons accompaying the strong carbonate absortion, indicating the presence of the organic component either witn or � associated with the large cabonates. The prous cabonate beneath 0.0 the rim exhibits the same four C-XANES absorptions, and the av­ 270 280 290 300 310 320 erage spectrum of the porous cabonate is indistinguishable fom Fig. 3. C-XANES spectrum averaged over the globule interior. Theratio that of the core carbonate. of the intensity of the carbonate peak to the organic peaks is higher in the These new measurements confir our earlier results indicating globule interior than in the rim (shown in Fig. 2). that relatively high concentrations (percent level) of orgaic C ae spatially associated, at the 100-nm scale, with the carbonates in ALH 84001. They difer fom the earlier results in that this sbple of rim material, consisting of fne-grained cabonate, magnette, and sulfides (i.e., the type of rim described by McKay et al. [2]), con­ 1.0 tains an orgaic component that is identical in its C-XANES spec­ trum to that of the cabonate globule to which it is attached. The 0.8 lI opaque sample we aalyzed previously, which was dominated by ; fldspathic glass ad chromite, has a C-XANES spectum that dif­

..Ji fers in absorption peak energies ad height ratios fom these rim 0 0.6 C � and globule samples. Fourier tansfrm infaed measurements, to O' vi J identif specifc C functiona groups in this sample, are in progress. 0.4 References: [l] Flynn G. J. et al. (1998) LPS XXIX, Abstract #1156. [2] McKay D.S. et al. (1996) Science, 273, 924-927. 0.2 1 CHIS OF MAGNETITE CRYSTALS I ALLAN IILS 0.0 84001: EVIENCE OF BIOLOGICAL ORIGI. E. I. Fried­ 270 280 290 300 2 3 1 310 320 mann 1, J. Wierzchos , and C. Ascaso , Depaent of Biological Fig. 4. C-XANES spectrum averaged over the porous carbonate, located Science, Florida State University, Tallahassee F 3230-110, USA between the rim and interior. The spectrum is very similar to that of the ([email protected]), 2Servicio de Microscopia Electonica, Uni­ carbonate in the interior (shown in Fig. 3). versitat de Lleida, 25198 Lleida, Spain ([email protected]), LP/ Contribution No. 956 15

3Cento de Ciencib Medioabientales, Serano 115, bis, Madrid number is positively corelated with brightness, which is also af­ 2806, Spain ([email protected]). fcted by, e.g., the atomic number of the substances below ad above, so single magnette crystals may appear in digerent shades Introduction: The presence of magnetite crystals in caron­ of grey. Oly chains lying approximately paallel to the image plane ate globules in ALH 8401 was one fact leading to the suggestion will be clearly visible as such. Those oriented obliquely will ap­ that this meteorite may contain relics of martia life [1]. These mag­ pea foreshorened, ad the individual crystals will not be resolved. netite crystals are morphologically similar in shape and size to Similarly, crystals lying below or above the chain will appear su­ magnetosomes fred by magnetotatic bacteria, but less so to typi­ per-imposed, resulting in a stronger, blured signal, as seen in cal inorganically frmed magnetite crystals [2,3]. The size range Fig. lb. We examined fine polished rock surfaces coated with is signifcat, as within this range the magnetic moment of the crys­ 10 nm C. Elemental composition was identified by energy disper­ tals is maimal [4]. Terestial magnetobacteria produce magneto­ sive X-ray spectoscopy (EDS) microaalysis and auger electron somes, magnetite (Fe304), or greigite (Fe3S4) crystals assembled spectroscopy (AES) with a 60-nm beam. into chains, ad the organisms use these chains fr aligning them­ Figures la ad lb show chains fom pacake carbonate glob­ selves along the prevailing magnetic field [4]. No inorganic pro­ ules. Figure 1c, fom a small irregular carbonate globule embed­ cess is known to produce similar stuctures. ded in orthopyroxene, shows on the left a cluster of magnetite Methos and Results: We fund numerous such chains com­ crystals, probably a mixture of single crystas and chains, ad a posed of magnetite crystals in martian meteorite ALH 84001 chain of seven or eight crystals on the right. Figures ld ad le show (Fig. 1) by high-magnifcation scaning electon microscopy op­ chains fom a small, irregular cabonate globule embedded in pla­ eratng in the backscattered electron detection mode (SEM-BSE), gioclbe glass, fom the aea marked by the arrow in Fig. lf. The a method that permits imaging of stuctures embedded in the rock chain in Fig. le appas to comprise about 15, prhaps more, larger substrate [5]. SEM-BSE images depict chemical compositions, not magnetite crystals. The chains in Figs. le ad le illustate the ap­ surface morphologies. Backscattered electrons originate not fom proximate total size range of crystals. Their exact shape and size the sbple surfce but fom below it: Calculations show that in were studied by the method of [3]: cleaning with acetc acid ad our material, fatures can be imaged to a depth of -40-10 nm. examination of the residue by TEM equipped with EDS microaa­ BSE images are fzzy because of scattering of BS electons. Atomic lytical system. We confrmed the results [3] that the magnetite crs-

Fig. I. Backscattered electron (BSE) images of magnetite crystal chains in ALH 84001. (a) and (b): Chains in pancake carbonate. (c)-(e): chains in small irregular carbonate globules. (d) and (e) are from the area indicated by arrow in (f). ca= carbonate, opx = orthopyroxene, pg= plagioclase glass. Further explanation in text. 16 Workshop on Martian Meteorites

tals ae mostly paralellepipedal and, less fequently, bullet- or ir­ Tiera, CSIC-Universidad de Granada, Facultad de Ciencias, 18002- regularly shapd, with a size range of 30-90 nm in length cd 20- Granada, Spain ([email protected]). 50 nm in width. Tiris range of morphologies cd sizes agrees well with those occuring in terestria magnetotactic bacteria. The mag­ The unambiguous detection of ancient lif is a crucial neces­ netite crystals in the chain in Fig. le ae lager (>20 nm). The fe­ sity in assessing the timing of life on Earth.Today, the strategy to quency of magnette crystal chains of digerent lengths (fur crystals reveal features of past lif frms is also of utmost importcce in or more) in ALH 8401 shows a non-Gaussian distibution, cd a seeking out living beings on other planets. negative exponentia relationship between the number of chains cd Among other very few biomakers used today (stromatolite the number of magnetite crstals per chain. This pattes suggests structures, autigenic minerals, biological degradation compunds, that the chains are disrupted fragments of originally longer cd isotopic analysis), mothological recognition of living forms magnetosome chains of magnetotactic bacteria. still plays a critical role in Precambrian micropaleontological stud­ Our fndings also help with the interpretation of the small mi­ ies. The underlying principle supporting lif detection using mor­ crofssil-like stuctures in ALH 84001 [1]. Studies with atomic phological and textural tools derived fom the old idea that frce microscopy [6] have shown that these structures are seg­ inorganic precipitates ae unable to produce neither shaps display­ mented cd morphologically similar to the magnetite crystal chains ing certain symmetry groups nor certain bizae textural aange­ described here. We suggest they are identica. ments. In this fame of mind, there is a substantial morphological Although conditions favorable fr microbial growth have been diference between the incimate and the animate worlds: it was shown to exist in meteorites lying on the Antactic ice sheet [7], thought that certain complex shaps with noncrystallographic sym­ contamination with terestrial magnetobacteria in ALH 8401 can mety were characteristic of life and would be impossible to ob­ be ruled out: (1) The martic origin of the cabonate globules is tain by inorganic precipitation. The most recent and conspicuous not in doubt; (2) several caonate grains contg magnette crys­ applicaton of this "law" is the fossillike microstuctures found in ta chains were enclosed in plagioclase glass (Fig. lf); cd (3) the ALH 84001. insides of fagments of ALH 84001 kept in urcyl acetate solution However, it is known that morphological pbters with symme­ under vacuum remained fee of uranyl contbination. ty pspries characteristc of living frs may also b precipitated Conclusion: We suggest the fllowing scenaio: Decompsed by purely inorganic mechcisms. Many of them ae irelevant to remains of dead martic magnetobacteria, possibly suspended in a lif detection, as they ae unlikely to occur under geochemical con­ cabonate-rich fuid, penetrated fssures of ALH 8401, already ditons. In few cases, on the other hcd, complex inorganic shapes crushed by previous meteorite impact, perhaps after the second im­ frm under laboratory conditions emulating a geological scenario, pat event (12) pstulated by [8]. Single magnette crystas cd chain thus becoming of geological interest as they can be used as fagments were deposited in cabonate globules. Finaly, the rock geochemical makers. This is the cae with cabonate precipitation sugered additiona impacts [8] resulting in melting, geochemical in silica-rich akaline brines forming what I call induced morphol­ and morphological trcsformations, and displacement of mineral ogy crsta aggregates [1]. The morphological emulation is particu­ components. laly dramatic when Ba or Sr carbonate ae precipitated (as shown All known magnetobacteria are microaerophilic or at least fac­ in Fig. 1) while textural emulation is clea in the case of calcite ultatve anaerobes, and it has been proposed [9] that they evolved [2]. on Earth at the time when partial pressure of atmospheric oxygen These biomimetic cabonates (of Ba, Sr, or Ca) patters ae ob­ began to increase. It has also been suggested that, if Earth-like life tained in the laboratory into alkaline silica-rich brines. Extreme as ever appeaed on the surface, then such conditions should have ex­ these conditions cc be (pH > 10; silica concentration > 500 ppm), isted on ealy Mas [10]. Our fndings seem to support this sug­ there ae today fw lakes, particulaly in the Africc Rift Valley, gestion. References: [1] McKay D.S. et a. (1996) Science, 273, 924. [2] Thomas-Keprta K. L. et al. (1997) LPS XXVl, Abstract #1433. Golden D. C. et al. (1997) LPS XXVll, Abstract #1675. [3] Thomas-Keprta K.L. et a.(1998) LPS XXVI, Abstact #1494. [4] Bazylinski D. A. and Moskowitz B. M. (1997) Rev. Mineral., 35, 181. [5] Wierzchos J. cd Ascaso C. (1994) J. Microsc., 175, 54. [6] Steele A. et al. (1998) J. Microsc., 189, 2. [7] Friedmann E." I. et a., personal communication. [8] Treiman A. H. (1998) Me­ teoritics & Planet. Sci., in press. [9] Chang S. R. and Kirschvink J. S. (1989) Annu. Rev. Earth Planet. Sci., 17, 169. [10) Hat­ man H. and McKay C. P. (1995) Planet. Space Sci., 43, 123; McKay C. P. (1995) in Chemical Evolution: Physics of the Origin and Evolution ofLife (1. Chela-Fores cd F.Raulin, eds.), pp. 177- 184, Kluwer.

BIOMIMETIC BUT ABIOTIC CARBONATE S: NEW GEOCHMICAL MARKERS FOR PRITIVE ENVIRON­ MNTS. J. M.Gacia-Ruiz, Instituto Andaluz de Ciencias de la Fig. 1. LPI Contribution No. 956 17

where these precipitation environments cb be fund. Therefre, 5. Is there any evidence of biominerals showing chemical or beyond a chemical curiosity or materials-science innovation, this mineral disequilibria? biomimetic carbonate precipitation is a plausible phenomenon un­ 6. Is there any evidence of stable isotope patters unique to bi- der natural conditions [3]. ology? Obviously, induced morphology crystal aggregation is an inor­ 7. Are there by organic biomakers present? ganic alterative to be considered when using morphology as a 8. Are the features indigenous to the sample? biomaker. I claim in this communication the importbce of con­ For acceptance of past life in a geologic sample, essentially all sidering this bizarre precipitation phenomenon when detecting of these criteria must be met. primiuve lif here on Earh or on other terestal planets. Silica Allan Hills 84001 Data vs. the Established Criteria forPast biomorphs frm into environments that are likely to be widely Life: How do the data fom the balysis bd study of ALH 8401 sprec during ealier stages of the Earth and Mas, plbets which compae to the established criteria? ae though to share a simila geological history during the Archean Geologic context. A matian origin fr ALH 84001 has been (Earth) and Noachian (Mas) periods [4]. In particular, some ter­ shown by both its O-isotopic compositons [3] bd tappd ma­ restial Archean (baritic) cherts ae thought to be the result of in­ tian atospheric gases [4,5]. Althoug the exact mab provenance organic silica-precipitaton under alkaline conditions. Silica leached of this igneous rock is unknown, ALH 84001 does contans cracks fom alkaline volcanic rocks increases the pH and, since there were bd prosity that, baed on textural microstatigahy, clearly frmed no silica absorbing organisms during the Archean period, the pool­ on Mas bd could conceivably have habored water-bore micro­ ing waters become enriched in silica. This draws b extremely bial cells and colonies introduced after the rock cooled (as known fvorable geochemical scenario fr the precipitauon of silica bio­ on Earth, [6]). The presence of secondary carbonate globules or morphs. Similaly, the process leading to precipitate carbonate pa­ pancakes in cracks has been interpreted by most workers as an in­ ticles with properties of induced morphology crstal aggregates is dication of relatively low-temprature secondary mineralization by likely to have been active in the geochemical scenario proposed fr a fuid, possibly water. Thus, the most widely bcepted broad geo­ the cabonate frmauon in ALH 84001 [5,6). logic context of this rock is not incompatible with the presence of References: [1] Garcia-Ruiz I. M. (1985) J. Crstal Growth, past life; if the seconda carbonates frmed at low temperature 85, 251-262. [2) Gacia-Ruiz J. M. (1994) Origin of Lie and Evo­ fom aqueous precipitation, their frmation is completely compat­ lution of the Biosphere, 24, 451-467. [3) Gacia-Ruiz J. M. (1998) ible with pat lif, but would not require it. Geology, in press. [4) Chyba C. F. and McDonald G. D. (1995) Age and histor. The isotopic age of ALH 84001 is 4.5 Ga bd Annu. Rev. Earth Planet. Sci., 23, 215-249. [5] McKay D. et al. it is, therefore, a sample of the original martian crust. The sample (1996) Science, 273, 924-930. [6] Sheaer C. K. et al. (1996) GCA, underwent extensive shocking aound 3.9-4.0 Ga [5,7]. Cabon­ 60, 2921-2926. ate frmation occured aound 3.9 Ga [8], shortly after the period of extensive bombadment and during a period when the plbet had abundant water [9], greater concentations of atmospheric gases, and higher temperatures. This corresponds to the time when life HOW DO THE PROPERTIES OF ALLAN HILLS 84001 appeaed and developed on Earth [10]. Evaporation of the fuids COMPAREWm ACCEPTEDCRITERIA FOR EVIDENCE percolating through the impact-cracked surfce could have resultd OF ANCIENT LIFE? E. K. Gibson Jr.1, D. S. McKay1, K. in the frmation of cabnates [11,12]. The saple was ejected fom Thoma-Keprta2, F. Westa!Jl, bd C. A. Romek3, 1Mail Code SN, the surface of Mars aout 17 m.y. ago and spent 11,00 yr in or Earth Sciences and Solar System Exploration Division, NASA on the Antarctic ice sheets. We suggest that the geologic history Johnson Space Center, Houston TX 77058, USA, 2Mail Code C23, of this rock is understood well enough to relate any possible lif Lockheed Martin, Houston TX 77058, USA, 3Savannah River forms to the geologic history of Mas bd to compae it to the his­ Ecology Laboratory, University of Georgia, Aiken SC 29802, USA. tory of life on Earth. Cellular morphologies. Some bacterialike structures [13-15] Criteria forPast Life: To be confdent that by sample con­ occur in the rims of the carbonate globules that resemble in size tains evidence of past lif or biogenic activity, one must determine bd shape te mineralized casts of moder terrestrial bacteria bd beyond a shadow of a doubt that certain well-established features their appendages (fbrils) or byproducts (extacellular polymeric or biomaker signatures ae present in the saple. In the case of substbces, EPS) [16]. Other bacteriomorphs ae very small but martib saples, the criteria fr past lif have not ben established within the size limit of known nanobacteria (i.e., 50-200 nm because if life existed on the· planet, we have no way of knowing [ 17, 18]). However, frm evidence that these bacterialike structures its detailed characteristics. Lacking independent evidence about the are tuly the fssilized remains of marian bacteria has not been nature of possible past life on Mars, the scientifc community must fund. Although some of the originally identifed features may have use, fr the time being, the criteria established fr ancient saples been coating atifacts or weathered mineral stucture artifcts, some fom the Earth [1,2]: ae defnitely not [16). Some of the fatures in ALH 8401 (e.g., 1. Do we know the geologic context of the sample? Is it com­ filaments) are common biogenic makers on Earth. We conclude patible with past life? that the evidence fr fssilized microbes and their products is not 2. Do we know the age of the sample bd its stratigraphic lo­ conclusive, but cbnot be readily explained by nonbiologic pro­ cation? Are they understood enough to relate possible lif to geo­ cesses bd should not be ignored. logic history? Microbial colonies. We have proposed that some of the features 3. Does the sample contain evidence of cellular morphology? in ALH 8400 l may be the remains of bioflms and their assoiated 4. What structural remains of colonies or communities exist mcrobial communities [13,14]. Bioflms provide major evidence within the samples? fr bacterial colonies in bcient Earth rocks [ 19]. It is possible that 18 Workshop on Martian Meteorites

some of the clusters of microfossil-like featuresmight be colonies, been shown to be violated by any published data on ALH 84001 although that interpretation depends on whether the individual fea­ so as to preclude life, and some evidence exists supporting each tures aretruly fossilized microbes. criterion. Therefore, the jury is still out on early Mars life as re­ Biominerals and disequilibria. The carbonates in ALH 84001 vealed by this meteorite [36]. Evaluation against these criteria is contain a population of magnetites having a highly peaked size dis­ still in progress and more data are needed. tribution and unusual rectangular prism shapes that are indistin­ References: [l] Schopf J. W. and Walker M. (1983) in guishable from some known microbially produced terrestrial Earth'sEarliest Biosphere: Its Origin and Evolution (J. W. Schopf, magnetite but match no known nonbiologic magnetite. Their for­ ed.), pp. 214-239, Princeton Univ. [2] Cloud P. and Morrison K. mation can best be explained by biogenic activity and disequilib­ (1979) Precambrian Res., 9, 81-91. [3] RomanekC. S. et al. (1998) ria of the Fe oxidation potential in the fluid that was the source of Meteoritics & Planet. Sci., 33, in press. [4] Bogard D. and Johnson the Fe [15,20]. Other irregular magnetitegrains could be either bio­ P. (1983) Science, 221, 651-655. [5] Bogard D. and Garrison D. genic or nonbiogenic in origin. Whiskerlike magnetites (<5% total (1998) Meteoritics & Planet. Sci., 33, A19. [6] Stevens T. 0. and magnetites in carbonate) described by [21-23] are quite different McKinley J. P. (1965) Science, 270, 450-455. [7] Ash R. D. in size distribution and shape, and may have had an origin unre­ et al. (1996) Nature, 380, 57-59. [8] Borg L. et al., in prepara­ lated to the rectangular prisms. Nanometer-sized iron sulfides de­ tion. [9] Head J. et al. (1998) Meteoritics & Planet. Sci., 33, A66. scribed in our original paper are also suggestive of disequilibria [10] Mojzsis S. L. et al. (1996) Nature, 384, 55-59. [11] Warren related to microbial activity [24], as is the elemental composition P. (1998) JGR, in press. [12] Mcsween H. Y. and Harvey R. P. of the carbonates. Overall, the mineral assemblages in the carbon­ (1998) Meteoritics & Planet. Sci., 33, Al03. [13] McKay D. S. ates, as well as their extreme chemical variations, are compatible et al. (1997) LPS XXVIII, 919-920. [14] McKay D. S. et al. with known biominerals and known disequilibria related to micro­ (1998) Nature. [15] Thomas-Keprta K. et al. (1998) LPS XXIX, bial activity on Earth, although more work needs to be done to dis­ Abstract #1494. [16] Thomas-Keprta K. et al. (1998) Geology, tinguish true biominerals and biogenically-related chemical in press. [17] Kajander E. 0. et al. (1998) Proc. SPIE 3441, disequilibria fromtotally nonbiologic minerals anddisequilibria. 86-94. [18] Vainshtein M. et al. (1998) Proc. SPIE 3441, Biologic isotopic signatures. Stable isotope patterns haveshown 95-111. [19] Westall F. et al., Precambrian Res., in press. the presence of indigenous C components with isotopic signatures [20] Devouard B. et al. (1998) Am. Mineral., in press. [21] Brad­ of -13 to -18%0 [25-27], which arein the direction of known bio­ ley J. P. et al. (1996) GCA, 60, 5149-5155. [22] Bradley J.P. et al. genic C signatures. Additional detailed study of the C-isotopic sig­ (1997) Meteoritics & Planet. Sci., 32, A20. [23] Bradley J. et al. natures is needed to distinguish between indigenous C components ( 1998) Meteoritics & Planet. Sci., 33, in press. [24] McKay within ALH 84001 and those introduced after its arrival on Earth. D. S. et al. (1996) Science, 273, 924-930. [25] Grady M. et al. Overall, the C-isotopic signatures of the identifiablenonterrestrial, (1994) Meteoritics, 29, 469. [26] Grady M. et al. (1998) GEO­ possibly organic C are compatible with biologic C-isotopic frac­ SCIENCE 98 abstract. [27] Jull T. et al. (1998) Science, 279, 366- tionation, when compared with the signatures of the martian car­ 369. [28] Clemett S. et al. (1998) Faraday Discussions (Royal bonates, but they do not prove that it occurred. Chemical Society), 109, in press. [29] Clemett S. et al., in prepa­ Organic biomarkers. Possible organic biomarkers are present ration. [30] Flynn G. et al. (1997) LPS XXVIl, 367-368. within ALH 84001 in the form of PAHs associated with the car­ [31] Flynn G. et al. (1998) Meteoritics & Planet. Sci., 33, A50- bonate globules [28], some of which may be a unique product of A5 l. [32] Bada G. et al. (1998) Science, 279, 362-365. [33] Gibson bacterial decay [29]. The distribution of the reduced C compounds E. K. et al. (1998) Astrobiology News, in press. [34] Romanek within the globules is irregular [28,30,31]. Clemett's data on PAHs C. S. et al. (1994) Nature, 372, 655-657. [35] Valley J. et al. [28], combined with recent amino acid data [i.e., 32], show that (1997) Science, 275, 1633-1638. [36] Gibson E. K. et al. (1997) the detected PAHs are most likely indigenous to ALH 84001, Sci. Am., 277, 58-65. whereas the detected amino acids aremost likely Antarctic contami­ nation. Exhaustive data must be collected before either component can be used as a biomarker fora specificsample [33]. XENON AND ARGON ISOTOPES IN IRRADIATED, Indigenous features. In our opinion, the recent studies of [28] ETCHEDNA KHLA: CHARACTERIZINGTHE HOST OF have shown conclusively that the PAHs are indigenous to MARTIAN ATMOSPHERIC XENON. J. D. Gilmour, J. A. ALH 84001 and are not contaminants. Based on isotopic compo­ Whitby, R. Burgess, and G. Turner, Department of Earth Sciences, sitions [25,27,34,35] and textures, there is absolutely no question University of Manchester, Manchester, M13 9PL, UK (Jamie. or disagreement that the carbonate globules andtheir included min­ [email protected]). erals formedon Mars and are indigenous to the meteorite. The pos­ sible microfossil structures and some regions of organic C that are Introduction: A noble gascomponent with the elevated 129Xe/ embedded in the carbonates are therefore almo·st certainly indig­ 132Xe ratio characteristic of the martian atmosphere is found in enous, but other possible evidence for life (e.g., amino acids) may Nakhla (and the other nakhlites) [1,2]. It is elementally fraction­ be Antarctic contamination. ated with 84Kr/132Xe values depleted by factor of -5 relative to the Summary: Clearly, we have not completely satisfiedall of the martian atmospheric value (as measured in EET 79001lithology C). criteria needed for general acceptance of evidence for life in a We believethis fractionation reflectsthe processes by which it was sample. We argue that we are close on some (likely biominerals, derived from the martian atmosphere and incorporated in the me­ possible organic biomarkers, and indigenous features) and not so teorite. If, as has been suggested, aqueous alteration is responsible close on others (well-documented geologic context, and evidence for this fractionation, the Xe component should be associated with for cells and colonies). However, not one of the eight criteria has "iddingsite" [2]. Our aim is to locate the host phase(s) of the mar- LP/Contribution No. 956 19

10

0.6

C 0 i Cl) ii C C 0 0.1 (J .E0 0.4 'O Cl) .s E iii ::I E 0.01 iii Ill z0 ' C ' 0 0 \ a. .2 •.. .. C � 0.001 .._.,'' 0.2 ' :c0 (J Olivine\' .001 SI Fe Mg Ca Al Na p s

Fig. 1. The solution above the acid-treated sample (Nak3) is dominated 0 by the dissolution products of olivine. Minor elements suggest additional 0 20 40 60 80 100 dissolution of phases including aqueous alteration products. Olivine and rust compositions from [5]. Percentage K Release Fig. 2. Water treatment (Nak2 = solid black bars) removes -20% of the Cl content of (Naki = open bars). Acid treatment removes nearly all of the chlorine in these samples (Nak3 = gray bars). tian atmospheric Xe component and so determine what led to its incorporationand whether the sameprocess can account forthe el­ emental fractionation. We have previously reported analysesof mineral separatesfrom 4 Nakhla that showed that a componentwith elevated 129Xe/l32Xe is associatedwith olivine andpyroxene but concentratedin mesostasis 3 1.0 [3). Combining our data with literature abundances of the major Ea. minerals [4] suggests that -10% of the martianatmosphere derived .s componentis foundin olivine, with approximately half the remain­ E 2 ::I der being associated with pyroxene and half with mesostasis. We ·;: CD ca 0.5 >< have now supplemented these data with analyses of Nakhla samples m crushed and etched in water and acid. Experimental: Approximately 210 mg of Nakhla was lightly o .:. ground to <400 µm and divided into three subsamples: Nakl 0 0 N (61.42 mg), Nak:2 (97.99 mg), and Nak:3 (57.96 mg). A finergrain 2 3., size was not used because of the anticipated difficultyin handling 1.0 -l and analyzing powders after irradiation. Samples Nak:2 and Nak:3 a. " were derived from the untreated sample by 20 hr of dissolution in .s IC G) ....., water and 0.3 molar HN03 respectively. The solutions above samples Nak:2 and Nak:3 were analyzedfor a selection of cations 0.5 using ICP-OES, while anions from the solution above Nak:2were .2 analyzedby ion chromatography (the ionic strength of the acid so­ lution above Nak:3 prevented use of this technique). Aliquots of each sample were prepared and subjected to neu­ 0 0 tron irradiation. After such irradiation, K, Cl, Ca, I, Ba, andU con­ 0 5 10 15 20 centrationscan be measured as enrichments in the daughter isotopes Release Step of Ar andXe produced after neutroncapture. Xenon-isotopic analy­ sis using RELAX and conventional Ar-isotopic analysis of these Fig. 3. Laser step heating of Nak2 shows that -50% of the excess 129Xe samples are ongoing. Here we report results of laser step heating correlates with I in a low-temperature release. The remaining, high­ analyses of aliquots of Nak:2 and Nak:3. temperature Xe seems associated with a Ba-rich phase. Acid treatment Results: Chemical analysis of the solutions above Nak:2 and removed -50% of the I without removing Xe, and so destroyed the Nak:3 revealed that the major element removed during the water correlation. 20 Workshop on Martian Meteorites

etch was Cl (20% of the expected budget), with traces of Na and S was frst put forth in Havey and Mcsween [1]. This short paper also present. Acid teatment removed -6% of the saple mass, and appeared roughly one month pror to that of McKay et al., in which the elemental concentrations were consistent with olivine dissolu­ specific fatures of ALH 84001 were suggested to be the result of tion and a contibution fom iddingsite and phosphate (Fig. 1). biological activity on acient Mars [2]. In the media fenzy fllow­ Argon. There is no vaation among the 40 Ar/39Ar age data fom ing that anouncement, and subsequent scientifc reseach, the work the three sepaates. The most draatic digerence is a reduction in of Harvey ad Mcsween [l] was ofen cited as a counterpoint to the Cl content of Nak2 (as expected fom analysis of the solution) the McKay groups hypothesis; often this was done as if it were the ad its nea-complete removal fom Na3 (Fig. 2). fag of a "resistace movement," rather tha as a piece of the com­ Xenon. There is no evidence fr any depletion in the martian plex ALH 8401 puzzle. The purpose of this abstract is to explain atmosphere-derived component (as measured by excess 129Xe) be­ the impact metasomatism hypothesis in greater detail and to ofer tween Nak2 and Na3, ad both are consistent with previously re­ an assessment of its validity in view of subsequent reseach. prted bulk values [2]. In Nak2, there is a good corelaton between Impact Metasomatism Hypothesis Restated: Harvey and I and 129Xe at low temperature (Fig. 3). The I concentration is McSween [1] suggested that the carbonates fund in ALH 84001 reduced by a factor close to 2 by acid treatment. Somewhat sur­ frmed when an anhydrous CO2 fluid, heated ad mobilized by prisingly, this dos not remove excess 129Xe and so destroys the impact, was fushed through impact-produced factures of the origi­ correlation. Around 50% of the martia atospheric Xe is released nal orthopyroxene cumulate lithology. This CO2 fuid would ini­ at low temperature in this sample. tally be very hot and col rapidly, with the whole event taking only Discussion: One martia atmosphere component is released a period of minutes or seconds. During the first moments, when at low temprature and is associated with I, ad one is releaed dur­ temperatures and pressures were at a maximum, reactions would ing the high-temprature steps in which Ba-derived gas is evolved. occur between the high-temperature fuid and the mafc components In the light of our work on separates, it is reasonable to assign the of the taget lithology, enching the fuid in Mg, Fe, and Ca. Im­ frst release to mesostasis and the second to pyroxene, although fur­ mediate (a few seconds later) cooling of this fluid would induce ther work will be required to confirm this conclusion. precipitation of the cabonates as obsered. Harvey and McSween The partial removal of I during the acid etch stage of our aaly­ used the ter impact metasomatism to describe this scenario, based ses (along with near-total removal of Cl) and the associaed destuc­ on its similarity to mechanisms suggested fr scrubbing CO2 out tion of the martian atmosphere Xe-I correlation is difficult to of the ancient martia atmosphere (3,4]. The plausibility of impact explain. We do not have accurate temperaure data, but compa­ metasomatism as a mechanism producing ALH 8401 cabonates son of the release pattess suggests I and CJ may not be associ­ was baed on may lines of evidence, as fllows. ated. The low temperature correlation between I and Xe is Mineralogcal and Texural Consideraton: Most reseach­ remakably good ad seems to require a common host phase. Leah­ ers agree that carbonate is found exclusively within te factured ing of I fom this host phase may be capable of producing this ef­ zones of ALH 8401. While some have suggested that several dis­ fct; if so, it seems likely that it is fne grained and mesostasis is tinct types of cabonae exist, evidence fr this is sketchy in that again implicated. The failure of acid etching to remove Xe in spite all are space-flling or coat-facture surfaces and ae composition­ of its dissolution of olivine suggests that, although some excess ally indistinguishable. Indeed, the "type-specimen" carbonate has 129Xe is present [6], iddingsite is not a major carrier. a circular or semicircular habit in thin section and on facture sur­ This work reveals diferences between the Nahlite component faces, with a distinct patter of compsitional changes fom core and the similarly elementally fationated noble-gas component to rim, and all cabonates, regardless of setting, show a recogniz­ present in ALH 8401. In particular, the 129Xe/132Xe ratio of the able part of tis patter. This observation led Havey and Mcsween trappd component in ALH 84001 is 2.0 ± 0.2 [7], while Nakla to the common belief that all carbonate in ALH 8401 originated yields 2.4. Furthermore, the dominat release of excess 129Xe fom fom a single event. This also implies that cabonates grew sequen­ ALH 84001 occurs at high temperature and is associated with tially fom core to rim, nucleating fom point sources, suggesting orthopyroxene[7]. It seems that similar Kr/Xe factionations in the kinetic control of crystallization ad pssible time-sequence infor­ Naites ad ALH 84001 do not indicate a common origin. mation. References: [1] Ott U. and Begemann F. (1985) Nature, 317, While most research on facture-zone phaes has centered on 509-512. [2]Drake M. J. et al. (1994) Meteoritics, 26, 854-859. carbonate, the presence and absence of other phases is imprtant [3] Gilmour J. D. et al. (1998) EPSL, submitted. [4] Treiman to note. Chrome spinels frm ubiquitous trains of crushed grains A. H. (1986) GCA, 50, 1061-1070. [5] Gooding J. L. et al. (1991) that characterize the maximum mechanical ofset occurring during Meteoritics, 26, 135-143. [6] Swindle TD. at al. (1995) LPSXX.VI, impact (on a scale of several millimeters in some cases). Maskelyn­ 1385-1386. [7] Gilmour J. D. et al. (1998) GCA, in press. ite is also fund preferentially associated with facture zones and sometimes intimately intermixed with cabonate. While this ha sug­ gested to some that carbonate frms by replacement of maskelynite FORMATION OF CARBONATES IN ALLAN HILS 84001 (5,6], it cannot be that simple; the Fe/Mg composition of the car­ BY IMACT MTASOMATISM: COOKING WITH GAS. bonates requires a source for these cations, and no alurnous R. P. Harvey, Department of Geological Sciences, Case Wester byproducts have been identifed. Olivine with distinctive amoeboid Reserve University, Clevelad OH 44106-7216, USA (rph@po. gran shapes is another phase fund exclusively within crushed cwru.edu). zones. Finally, the expected products of hydrous alteration of ultramafic rocks, such as chlorite, serpentine, or similar phases, Introduction: The possibility that the carbonates in are for all intents and purposes absent; only a few isolated ALH 84001 were frmed by the process of impact metasomatism submicrometer-sized grains have been identifed, ad it is not clear LPI Contribution No. 956 21

that they ae contemprceous with cabonate frmation [7]. These ate mineralogy at the submictmeter scale. These studies confred fndings ague fr cabonate frmation during a single, anhydrous a complex texture at submicrometer scales and revealed evidence event associated with a fuid percolating through facture zones in for the mechanical cd thermal egects of shok. Several obsera­ ALH 84001. tions best made sense in ters of gas-phase reactions at tempera­ Harvey cd McSween reported that the exteme "core" of ca­ tures above ambient. For example, the magnetite assemblage bonate growths was an Fe,Mg-rich calcite, intermixed with an observed throughout te carbonate included a variety of morpholo­ ankeritic (50% Ca, 50% Mg,Fe) carbonate. This ankerite was in gies, some with specifc crystalline defcts typically associated with tum intermixed with the Fe-rich end member of the relauvely Ca­ high-temperature vapor-phase growth [13). These sbe crystals of­ rich magnesite-siderite cabonates most commonly observed. While ten had a rigorous epitaial relationship to the cabonate substrate, some specifc compsitions had been observed in terestrial rocks, indicating sufcient fee surface energy during magnetite crystal­ the suite as a whole has no known counterparts. Unfortunately the lization to allow cation mobility (i.e., elevated temperatures) [14). initial description of tis in [1] wa termed a "three-carbonate as­ These cd other observations suggest tempratures of at least a few semblage," when it more accurately should have been considered hundred degrees Celsius either during carbonate crystallization or a sequence of two component mixtures. That's how the asemblage during alteration of preexisting cabonate [14,15]. These reactions wa treated when cabonate geothermometry was applied, yielding could have occurred during the last moments of a metasomatic temperatures of 685°C (calcite-ankerite) cd 660° C (ankerite­ impact event, or during subsequent, lower-energy impact events. siderite) [8]. These temperatures were in remarkable agreement Trace-Element and Isotopic Consideration: Trace-element with the previously obsered corespondence between magnesite­ studies strongly suggested metasomatic reactions between siderite composiuons and the Ca-rich border of the 700°C magne­ ALH 8401 and c invasive fuid [16]. The isotopic composiuon site-siderite staility feld [9]. of ALH 84001 cabonates has been more extensively studied, with Carbonate-prpucing Reaction: In the simplest possible interesung if more contoversial results. Ealy studies of AL 84001 scenario, ALH 84001 cabonates would be produced fom compo­ leachates showed enrichments in BC cd 180 that were consistent nents derived solely fom the host ultbafc mineralogy. Unfr­ with either high or low temperatures of frmation; in general, low­ tunately, reactions that produce carbonate and other nonhydrous temperature, aqueous scenarios fr cabonate fration were fa­ products fom ultamafc rocks are not well studied; most of the vored (17]. However, Harey and McSween (1] argued that this literature concess reactions with a hydrous fuid. However, some interpretation was soft fr several reaons. First, BC enrichments reactions are known in both natural and experimental settings cd cannot unambiguously indicate a frmation temperature without provide insight into possible ALH 84001 carbonate formation assuming a fuid composition, given the enriched nature of te ma­ mechanisms. Most notable ae reactions that occur between ultra­ tian atmosphere as a whole. Second, studies of natural impact­ mafc minerals cd fuids with -Xc02 higher than 0.85. Two that produced caonates intermixed with mac impact glases show that ae of direct importance ae isotopic disequilibrium between these phases is the norm, even though peak temperatures in excess of 1200°C must have been 2 frsterite + 2 CO2 = 2 magnesite cd 1 enstatite reached [4). In situ studies of ALH 8401 carbonate using ion microprobe 1 enstatite + 2 CO2 = 2 magnesite + 2 quartz techniques produced equally polarized viewpoints. One study es­ tablished that O isotopes within the observed sequential Mg,Fe Terestal studies predict these reactions will occur in sequence compositional sequence were essentially homogenous, suggesting at about 660°C and 620°C respectively, at pressures between 3.5 a disequilibrum commonly used to indicate frmauon temperatures cd 5 kbar (or ~0.5 GPa) [10-12]. Several features of these reac­ below 300°C (18). However, a much more complete study (includ­ tions coincide well with what is observed in ALH 84001. The pres­ ing calyses of a wider range of carbonate compositions) found sures are reasonable in the context of an impact event (mechanical something distinctly diferent; that there was a distinct corelation twinning of pyroxene cd conversion of plagioclae to makelynite between O-isotopic cd maor-element compositions [19]. While typically require 30 GPa or more), and the temperatures are sug­ this correlation could not unambiguously distinguish between a gestive. Furthermore, these reactions predict a mineralogical se­ high- or low-temperature frmation fr the cabonates, it did show quence that can be readily observed in ALH 84001. Olivine that isotopic comsiuons could be consistent with the impact meta­ exhibiung reacuon textures is surrounded by enstatite cd associ­ somatism scenario fr a small volume system. ated with space-flling carbonate. In addition, spaces between car­ Other Observation and Conclusions: Mcy of the most vo­ bonate growths ae ofen flled with silica. One interesting point cal critics of the impact metasomatism hypothesis have fcused on not noted in [l), however, is that while the interstiual silica requires the notion that it required reaching equilibrium at high tempera­ that the reactions proceed towad production of that phase and ca­ ture, which would reset specifc isotopic or physical systems to val­ bonate, the olivine textures could be the product of decabonation ues other than those observed (e.g., 20). Contrary to this of preexisung carbonate. asumption, Haey cd Mcsween envisioned a rapid, nealy in­ Mcrophase and Microtexture: Havey and Mcsween sug­ stctceous event; reactions at elevated temperatures provide Ca, gested that the typical grain size of ALH 84001 carbonate was Mg, cd Fe to a CO2 fluid, which on fther cooling produces the somewhat smaller thc 1 µm; while some carbonate aeas showed observed cabonates and associated phases. Unfrtunately, without crystal faces cd composiuona homogeneity on the 1-10-µm scale, knowing the exact temperature, pressure, fuid composition, cd plots of cabonate calyses showed linear mixing trends between duration of the event, one cc only explore the reactions supply­ calcite-ckerite and ankerite-siderite compositions. Following that ing cations fr the observed carbonates, rather than set strict limits lead, TEM cd FE-SEM studies were conducted to explore cabon- on foration conditions. To use a common buzz word, the pro- 22 Workhop on Martian Meteorites

posed impact metasomatism is as much a "disequilibrium" process reason is simple; martian meteorites are so physically and mineral­ as any low-temperature precipitation mechanism, with one reaction ogically similar to other meteorite types that they cannot be recog­ supplying cations to a fluidthat later deposits an oversaturated pre­ nized as "martian" during observations possible in the field. cipitate. Such reactions can be initiated either thermally or Thin-section work would be the minimum needed to suggest that allochemically. a specimen is martian, and positive confirmation would require In summary, impact metasomatism still serves as a viable pos­ much more detailed chemistry. Anecdotal evidence suggests that sible mechanism forcar bonate formation in ALH 84001. Perhaps several of the martian meteorites were recognized as "different"in the biggest handicap of the impact metasomatism hypothesis has the field, but such designations have been notoriously inaccurate. been its "alien" nature; i.e., it doesn't seem familiar to those fa­ Systematically collecting all the meteorites froma given stranding voring Earth-like conditions on early Mars, and, to paraphrase surfaceis the best possible way to increase the chances something many, "it just doesn't feel right." But perhapsALH 84001 carbon­ rare will be recovered. ates should serve as a reminder that Mars is not Earth, and that 2. Could we recover more meteorites using technologically ad­ processes such as impact metasomatism, whichare almost certainly vanced techniques? familiar to Mars, may require us to think "like a martian." The answer is no, at least not with present-day technology. References: [l] Harveyand Mcsween (1996) Nature, 382, 49. While technology makes available an amazing array of sensors, [2] McKay D. et al. (1996) Science, 273, 924. [3] James P. et al. none can match the amazing capabilities of the human vision sys­ (1992) in Mars, p. 934, Univ. of Arizona, Tucson. [4] Martinez I. tem. With just a few days' training, a meteorite hunter can visu­ et al. (1994) EPSL, 121, 559. [5] Treiman A. (1995) Meteoritics, ally acquire, recognize, and categorize dozens of rocks within view 30, 294. [6] Kring D. et al. (1998) GCA, 62, 2155. [7] Thomas­ at rates surpassing any mechanicaldata acquisition system by sev­ Keprta K. et al. (1997) LPS XX.VIII, 1433. [8] Anovitz and Essene eral orders of magnitude. Humans store these data, learn as they (1987) J. Petrol., 28, 389. [9] Mittlefehldt D. (1994) Meteoritics, go, and are mobile, easy to care for,abundant, and cheap (particu­ 29, 214. [10] Trommsdorf and Connolly (1990) Contrib. Mineral. larly graduate students). Proposals have been made to try magne­ Petrol., 104. [11] Johannes W. (1969) Am. J. Sci., 267, 1083. tometers, spectral photometers, metal detectors, ground-penetrating [12] Ohnmacht W. (1974) J. Petrol., 15, 303. [13] Bradley J. et al. radar, and brainy robots to help ANSMET find meteorites. Cur­ (1996) GCA, 60, 5149-5155. [14] Bradley J. et al. (1998) Meteor­ rently, humans outperform all of these by a huge margin in terms itics & Planet. Sci., 33, 765. [15] Wentworth and Gooding (1995) of speed, recognition, and cost. LPS XXVI, 1489. [16] Wadhwa and Crozaz (1995) LPS XXVI, One aspect of modem technology that has provedto be of enor­ 1451. [17] Romanek et al. (1994) Nature, 372, 655. [18] Valley J. mous value is highly detailed satellite imagery.While satellites can­ et al. (1997) Science, 275, 1633. [19] Leshin L. et al. (1998) GCA, not "see" meteorites (or tell them from Earth rocks), they can help 62, 3. [20] Treimanand Romanek (1998) Meteoritics& Planet. Sci., us find exposed blue ice, and help us discern the shape and topog­ 33, 737. raphy of an icefield. These images speed up reconnaissance efforts and greatly facilitate systematic searches. They also serve as the only available basemap for precisely locating our finds,given that many Antarctic icefields lie off existing topographicalmaps. RECOVERING MORE ANTARCTIC MARTIAN METEOR­ 3. If we sent an army of professional meteorite hunters, could ITES: ANSWERS TO COMMON QUESTIONS. R. P. we gather martian meteorites more quickly? Harvey1 and W. A. Cassidy2, 1Case Western Reserve University, The answer to this one is "maybe." Success in findingAntarc­ ClevelandOH 44106-7216, USA ([email protected]),2 Department tic meteorites depends on a lot of factors, of which manpower is of Geology and Planetary Science, University of Pittsburgh, only one. Reconnaissance is perhaps the most important factor; you Pittsburgh PA 15260, USA (ansmet@vms.cis.pitt.edu). will recover more meteorites if you know where to go to find the densest concentration. Weather is another vitally important factor. Introduction: Of the 13 martian meteorites reported, almost Too much wind and you can't see because of blowing snow and half were collected in Antarctica. With the increased interest in freezingskin; too much snow and it buries the meteorites; too little martianmeteorite studies accompanying the report of possible bio­ wind and they stay buried. Given good weather and a meteorite genic activity in ALH 84001, many have wondered whether recov­ stranding surface that is a known provider, adding more people to eries of Antarctic martian meteorites could be enhanced or a field team means covering more territory more quickly. Experi­ accelerated. Given that the recovery of Antarctic meteorites is a ence among fieldparty members also helps, and we typicallyaim planned activity, (unlike the recovery of non-Antarctic martian for a 50:50 mix of novices and experts; however, the benefits of a meteorites via random finds and falls), such questions are worth fully"professional" field teamare not asclear. While novices typi­ considering. Listed below are several topical questions, followed cally need a few extra days to learnto recognize meteorites, they by answersbased on experience gathered during the 21 completed are also more likely to be cautious and "over-identify" possible me­ fieldseasons of ANSMET (the AntarcticSearch for Meteoritespro­ teorites, helping ensure that nothing interesting gets left behind. gram). Antarctic field work also demands a significant time commitment 1. Can Antarctic meteorite collection activities be focused on that many professionalswould finddifficult to commit. recovery of martian samples? It should also be recognized that, even if we did recover speci­ The answer is no. Antarctic meteorite hunting is analogous to mens at double or triplethe average pace (-350/yr), there is a bottle­ mining placer deposits - we visit areas where meteorites are con­ neck at the initial characterization and curation end, where martian centrated by ice sheet processes, and only by collecting all the me­ specimens would first be identified. Currently the Meteorite Pro­ teorites can we ensure recovery of a few of special interest. The cessing Facility at Johnson Space Center can handle several hun- LP/ Contribution No. 956 23

dred recoveries per year very well; but recoveries of a thousand or finding more martian samples. However, such searchesare mean­ more dramaticallyslow things down. ingless if unbiased and rapid distribution of samples to interested 4. Do searches miss types of martian meteorites that resemble scientists is not guaranteed. Commercial or private meteorite col­ terrestrial rock? lection is antithetical to this goal. It's possible, but past performancesuggests such potentiallosses Finally, minimize restrictionson the distributionof the martian are minimal. Most of the areas of meteorite stranding surfaces are samples. New martian samples can be most quickly studied if they isolated from nearbysources of terrestrialrock. Areas like the Allan are widely distributed; and the more complex and unique the Hills Western Icefields (where ALH 84001 came from) are many sample, the wider and quicker this distributionshould be. The hast­ kilometers from exposed bedrock and have no terrestrial rock on ily imposed moratorium on distribution of ALH 84001 samples is their ice surfaces; anyrock we find is almost certainly a meteorite a case in point, sincenew researchers were effectively locked out regardless of how "terrestrial"it looks. Regions of some icefields, of this research foralmost a year. however, lie close to exposed bedrock and can have a scattering of terrestrialrock on their surfaces. Only by careful observationof each and every rock can one findthe meteorites included there; looking ESTIMATESOF OXYGEN FUGACITY IN THE BASALTIC for fusion crust, a particular glossy patina, or perhaps just a rock SHERGOTTITES FROM ELECTRON MICROPROBE that looks out of place by virtue of size or mineralogy. Luckily, OXYGEN ANALYSIS. C. D. K. Herd1 and J. J. Papikez, we have just the instrumentto allow such a searchto succeed (see 1Department of Earth and Planetary Sciences, University of New question 2 above). Proofthat martianmeteorites canbe foundmixed Mexico, Albuquerque NM 87131, USA ([email protected]), in with terrestrialrocks is that the lasttwo discovered (LEW 88516 zoepartment of Earth and Planetary Sciences, University of New and QUE 94201) both came from regions with significant earth­ Mexico, Albuquerque NM 87131, USA ([email protected]). rock; the latter from a moraine filled with millions of like-sized pebbles. In summary, while we might miss a fusion-crust-freemar­ The basaltic shergottites consist of four meteorites thought to tiansandstone mixed in with terrestrialsandstone, we wouldn't miss have a martian origin: Shergotty, Zagami, EETA 79001, and it lying on an isolated icefield or mixed in with dolerite; chances QUE 94201. All are basalticin composition, but only QUE 94201 are such losses are minimal. It is also important to note that contains no cumulus pyroxene and may therefore represent a mar­ ANSMET fieldparties always err on the side of caution, routinely tian liquid [I]. bringing back any rock that resembles a meteorite or simply doesn't Previous estimates offO z at the time of crystallization of the fitwithin the observed range of local rock typesand sizes. shergottites range fromthe quartz-fayalite-magnetite(QFM) buffer 5. What can be done to help gather more martian meteorites? [2] to 4 log units below QFM [3], based on coexisting Fe-Ti Scientists and others can take several steps to boost the recov­ oxides, for all but QUE 94201. Oxygen fugacity conditions for ery of martianmeteorites. QUE 94201 are near the iron-wiistite (IW) bufferaccording to [l], First, support annual field parties. Going hunting every year about 4 log units below QFM. minimizes the stochasticinfluence of the weather, maximizing the The current research aims to refine the estimates offO z forthe chancethat ANSMET field parties will be in the right place at the basaltic shergottites, through electronmicroprobe techniques. The right time. If you feel generous, support multiple field parties. This technique of quantitative analysis of O using the electron micro­ likewise minimizes the effectof weather, but also allows simulta­ probe has been developed to the point where O can be analyzed neous reconnaissance and systematic searching. The annual rate of with certainty comparable to the analysis of most cations. Practi­ returnsgoes up because icefieldsfor future searching can be more cal application of the technique to naturally occurring spinels was readily prioritizedwhile recoveries take place, not in their stead. demonstrated by [4]. A set of O standards now exists that can be Second, support survey-level characterization as fully as pos­ readily obtained [5]. The technique also allows an assessment of sible. In order to find the martians in the mix, all the specimens stoichiometry, since the cation total, on the basis of four O atoms, need to be characterized in a timely fashion. Yamato 793605 is a does not rely on the assumption of stoichiometry. Instead, the superb exampleof the importanceof this early step. The 1979 Japa­ amount of Fe3+ is determinedby charge balance. nese Antarctic meteoritecollection effort yielded nearly4000 speci­ The method of analysis of the oxide minerals in the basaltic mens, but because no infrastructurefor rapid characterization was shergottites follows the technique of [4]. Initially, when analyses in place, many of the specimens were never carefully examined. were carried out, 0 was not monitored for shift (constant machine Fifteen yearsafter collection,Y 793605 was recognized as a "new" error) or drift (gradual machine error), as outlined in [4]. Future martian lherzolite. Similarly, support continued study of existing analyseswill include this aspect in the analytical routine. Raw data modem falls. By virtue of their masses, these specimens can in­ were corrected using the off-line CITZAF correction software [6]. clude a more varied lithology within a single sample than is typi­ A comparison of correctionson standard analyses revealed that the cally seen in Antarctic specimens and may contain new clues matrix correctionof [7] corrected O the most accurately. However, concerning the martian crust and its interaction with the martian cation amounts were unacceptable. Therefore, a custom correction environment. Witness the discovery of NaCl in Nakhla last year: was developed using the CITZAF software. The custom correction who knows what lies in the many kilograms of Chassigny not yet uses the absorption correction of [7], the stopping power correc­ studied? tion of [8], and the backscatter correction of [9]. An O area-peak Third, support recovery from dry deserts and searches in factor was also determined and used to adjust the O K-ratios be­ Greenland. Given that 1 of every 2000 meteorites found in Ant­ forethey were input into the customized correction. arcticaturns out to be martian,with even better odds for non-Ant­ The Fe-Ti oxide exchange model software of [10] was used to

arctic recoveries, systematic recovery of more meteorites means initially estimate temperature andf O z for the basaltic shergottites. 24 Workhop on Marian Meteorites

The program assumes stoichiometry, and thus provides an initial brated 1969 Mariner 7 Infrared Spectrometer (IRS) spectra (1.8- estimate with which data based on charge balance can be compared. 14.4 µm) provide evidence forthe presence of a hydrous weather­ The results foran individual shergottite apparently define a trajec­ ing product on the martian surface, interpreted to be goethite. A tory in T-/02 space that parallels the O buffer lines. According to hydrous weathering product such as goethite may mark areaswith this model, EETA 79001 (lithology A) records temperatures of a higher abundance of water vapor or areas where water was more 760°-900°C, and a trajectory -1.75 log units below the QFM abundant in the past. Such areas may contain interesting geologic buffer, or 0.5 log units above the wtistite-magnetite (WM) buffer. structures, and an increased abundanceof water vapor would have Queen Alexandra Range 94201 records temperatures of 950°- implications in the search for areas most conducive to life. These 0 990 C, and an /02 at or slightly below the WM buffer. Shergotty areas may thus warrant more in-depth examination, and may also ° ° records relatively consistent temperatures of 840 -870 C, and/02 make prime targets for futurelanders. only 0.66 log units below QFM. Zagami records temperatures of One point the goethiteinterpretation illustratesthat hasrelevance ° ° 745 -860 C, and/02 from0.75 to 1.3 log units below QFM. for futurespectral instrumentsis the importanceof obtaining a full Estimates based on chargebalance were detennined,in a rough spectrum to search for unexpected constituents. A multi-band ra­ manner, from a graph of coexisting spinel-rhombohedraloxide pair diometer, such as the Viking IRTM or the proposed MarsSurveyor compositions (Fig. 4 of [10]). ElephantMoraine 79001 spine! and 2001 THEMIS, has the advantage of lower data rate requirements, rhombohedral oxide (ilmenite) phases are stoichiometric,and tem­ andit does provide informationon the surface. However, it cannot ° ° peratures range from 590 to 760 C, with/02 at or below the 1W be used to uniquely identify unexpected constituents, such as go­ buffer. Queen AlexandraRange 94201 spinels are nonstoichiomet­ ethite. ric, with cation sums of 2.7-2.8. It should be noted that this Data: The IRS returned high-quality spectra covering 1.8- degree of nonstoichiometryis uncommon andlikely represents un­ 14.4 µm (5550-690 cm-1), and the spectralrange covered remains certainty in the O measurement. Ilmenites are stoichiometric. Ap­ unique. However, much of the dataset and calibration information ° ° proximate temperatures are 500 -700 C, and/02 is at or below the was lost by the late 1970s, so the spectra have not been fully uti­ 1W buffer. Shergotty spinels and ilmenites are stoichiometric, and lized. To reconstruct the dataset, we located the original IRS data ° ° the recorded temperatures are 600 -700 C, with/02 at or belowthe tapes, recovered the missing data, and collected information from 1W buffer.Zagami ilmenites are stoichiometric, but spinels display the original IRS team [l]. Thus, for the first time since the 1970s, cation excess, reflectinguncertainties in the O measurements. Tem­ we have IRS spectra calibrated in wavelength and intensity using ° ° peratures rangefrom 650 to 900 C, again with/02 at or below the the original dataset and calibration information and expertise pro­ 1W buffer. In general, the estimates from the graph of [10] indi­ vided by the IRS team. cate the need forsoftware that does not assume stoichiometry, and Infrared Spectrometer at Thermal Wavelengths: Goethite reflect the uncertainty in measurement of 0. measured in transmission has bandsat approximately 11.25, 12.5, , 1 The rough estimates of /02 based on charge balance instead of and 20 µm (890, 800, and 500 cm- ) [2]. Many Fe-bearing miner­ stoichiometry, are, on average, 3 log units below estimates from als exhibit a 20-µm band; however, an 11.25- and 12.5-µm dou­ the Fe-Ti exchange software of [10]. Ilmenites are essentially sto­ blet seems to be rare. For example, other common ferric oxides, ichiometric, but nonstoichiometry in the spine! phases is common, such as hematite, magnetite (Fe304), maghemite (y-Fe203), in accordance with the study of terrestrial spinels by [4]. Future akaganeite ((3-FeO•OH), and lepidocrocite (y-FeO-OH), do not have analyses using a refined analytical procedure forO will allow the the doublet at 11.25 and 12.5 µm [3]. impact of the stoichiometry assumptionon the estimation off0 2 in Bands centered near 11.25 and 12.5 µm that are a close match the basaltic shergottites to be assessed. to transmission bands in goethite appear in many of the IRS spec­ References: [l] McSween H. Y. Jr. et al. (1996) GCA, 60, tra, although atmospheric CO2 featurespartially obscure the 12.5- 4563-4569. [2] Stolper E. and McSween H. Y. Jr. (1979) GCA, µm band. The intensities of the 11.25- and 12.5-µm bands are 43, 1475-1498. [3] Ghosal S. et al. (1998) Contrib. Mineral. approximately equal andcorrelated, which indicate the same con­ Petrol., 130, 346--357. [4]Kamperman M. et al. (1996) Am. Min­ stituent causes both bands. Neither banddepth correlateswith at­ eral., 81, 1186-1194. [5] McGuire A.V. et al. (1992) Am. Min­ mospheric path length; therefore, both are assigned to the surface eral., 77, 1087-1091. [6] Armstrong J.T. (1995) Microbeam spectrum. Analysis, 4, 177-200. [7] Love G. and Scott V. D. (1978) J. Comparison of the IRS bands to spectra measured in emission Phys. D., 11, 1369-1376. [8] Pouchou J.-L. and Pichoir F. (1991) do not match the sharpness of the features in IRS spectra, indicat­ in Electron Probe Quantitation, pp. 31-75, Plenum Press, New ing the goethite is seen in transmission. This is the expected result York. [9] Duncumb P. and Reed S. J. B. (1968) in Quantitative if the goethite is present as a very thin layer or coating (less than Electron Probe Microanalysis, pp. 133-154, U.S. Government =100 µm thick). Such coatings occur on Earth as a desert varnish Printing Office,Washington, DC. [10] Ghiorso M. and Sack R. 0. or as a coating of very fine (clay-sized) particles on sand-sized (1991) Contrib. Mineral. Petrol., 108, 485-510. grains. Alternatively, a thermal gradient at the surface layer would have the sameeffect. The 11.25-µm band does not correlate with the 9-µm band, in­ IMPLICATION OF HYDROUS WEATHERING PRODUCT dicating it is not from the aerosol dust. It does not correlate with ON MARS FOR FUTURE EXPLORATION. L. E. Kirkland1 topography or 2.2-µmalbedo, and it shows a weak, positive corre­ and K. C. Herr2, 1Lunar andPlanetary Institute andRice University, lation with the 3-µm band depth. We see no correlation with geo­ Houston TX 77058, USA ([email protected]), 2The Aero­ logic units. space Corporation, El Segundo CA, USA. Infrared Spectrometer at ReflectedWavelengths: Goethite has a 2.4-µm band and a very strong 3-µm band. A detailed ex­ Introduction: Analysis of the newly recovered and recali- amination of the 2.4-µm region in IRS spectra must await the full LP/ Contribution No. 956 25

calibration of this region, which is not yet complete. However, pre­ unexpected constituents, and in the search for areas most condu­ liminary analysis indicates the presence of a 2.4-µm band, along cive to life, it may be the unexpected that proves the most infor­ with a stronger 3-µm band. mative. Water Vapor: The stabilityof goethite depends in part on the Determining the presence of a doublet at 11.25 and 12.S µm amount of water vaporpresent, andIRS spectra provide a measure­ caused by goethite requires spectra with sufficient spectral resolu­ ment of the water vapor band at 6.45 µm. However, background tion and spectral sampling rate to measure the shape of the bands variationsin thisregion make water vapor amounts difficultto de­ and thus determine the band centers. Although a multichannel termine fromthis band.Nonetheless, it may be significantthat the instrument may be used to determine the presence of spectral varia­ strongest l 1.25-µm band, recorded along the western side of the tions in the 11.25/12.5-µm region, it cannot return enough infor­ Hellas basin, also contains the strongest6.4S-µm water vaporband. mation to determine the cause. Finally, in the near-IRto thermal-IR Other work [4] also indicates this region exhibits unusual varia­ region, a goethite interpretation is strengthened by coverage from tions in water vapor. -2 to 25 µm, where goethite has spectral features. ThermalEmission Spectrometer: The MarsGlobal Surveyor References: [1] Kirklandand Herr (1998) LPSXX.IX, Abstract Thermal Emission Spectrometer (TES) recently returned spectra #1S18. [2] Hunt et al. (1973) Icarus, 18, 459. [3] Bell et al. (1995) from Mars covering -6-50 µm (1700-200 cm-1), and the spectra JGR, JOO, S297; Estep-Barnes (1977) in Infrared Spectroscopy are in the early stages of calibration and interpretation. Both TES (Farmer, ed.), Ch. 11. [4] McAfee et al. (1998) LPSXX.IX, Abstract and IRS have a spectral resolution of -10 cm-1 at 10 µm. How­ #1315. [SJ Christensen et al. (1998) Science, 279, 1692. [6] Curran ever, TES has a lower spectral sampling rate than IRS, which de­ et al. (1973) Science, 182, 381. [7] Christensen P. (1998) personal creasesits effectivespectral resolution relative to IRS.Both measure communication. this spectralregion withsimilar signal-to-noise ratios (-400for TES and-600 for IRS). We examined four TES spectra taken from the same region as the spectrapublished in (see Fig. 3 in [S]). The main cause of varia­ THESIGNATURE OF LIFE: IS IT LEGIBLE? A. H. Knoll, tions of features in these spectra had been attributed to the pres­ Botanical Museum, HaIVard University, 26 Oxford Street, Cam­ ence of pyroxene [5]. The lower spectral sampling of TES, bridge MA 02138, USA ([email protected]). combined with interference fromthe overlapping CO2 bands, pre­ cludes a detailed examination of the 12.S-µm region in TES spec­ Everyday experience suggests that the gulf between biology and tra. However, the TES spectra exhibit a feature that matches the the physical world is conspicuous. This impression arises,however, 11.25-µm band seen in IRS spectra. because the biology most familiar to us is largely that of organ­ To further examine the 11-µm region, we calculated ratios to isms found on distal branches of the tree of life. The difficulty in enhance the spectral differences between the published TES spec­ distinguishing biogenic from abiogenic forms lies at the other end tra. The close match of the residual spectrum to water ice shows of the tree; life arose as the self-perpetuating product of physical that variations in water ice amounts rather than variations in sur­ processes, andit is likely that the characteristics of Earth's earliest facemineralogy likely cause most of the differencein spectralshape organisms - their size, shape, molecular composition, and cata- from 10 to 13 µm for these TES spectra. The shape of the residual 1 ytic properties - bore a close resemblance to products of the spectra indicates the TES spectrameasured a water ice cloud rather physical processes that gave rise to life. than ice on the surface. The weakness of the band indicates the Organisms have structure, they have a chemical composition, cloud is probably optically thinat visible wavelengths. Both TES and they affect their environment; thus, paleontological evidence and the 1971 Mariner 9 InfraredInterferometer Spectrometer (IRIS) of ancient life can be morphological, geochemical, or sedimento­ reported measurements of water ice clouds on Mars using this spec­ logical. Experience with terrestrialsamples makes it clearthat fea­ tral region [S,6]. tures found in ancient sedimentary rocks can be accepted as Pyroxene or Goethite? An early interpretation of the 9- and evidence of former lifeonly if they meet two distinct criteria. First, 11-µm bands measured by TES attributed them to pyroxene [S], the features must be compatible with formation by known biologi­ but our work indicates the most likely cause of these features is cal processes. Second, they must be incompatible with formation water ice [7]. by physical processes. This is straightforward in principle, but it The shape and depth of the TES 9-µm band in these spectra is requires that we understand the limits of patterngeneration by both consistent with a spectrumof dust, with the 9.4-µm CO2 band caus­ biological and physical processes. There remains a good deal of ing the sharp featureat the base of the band.TES spectraalso show ignorance about the range of features that can be generated by bio­ a weak 20-µm band, and goethite has a 20-µm band. Therefore, logical andphysical processes. We know, for example, that micro­ these TES spectra appear to be consistent with the goethite inter­ bial mats can facilitate episodic carbonate precipitation to form the pretation, with spectral features attributedto the following: 9-µm layered sedimentarystructures known as stromatolites, but we also band = dust as an aerosol and on the surface; 10-13-µm region know that layered carbonate precipitates can form in the absence band depth variations = water ice clouds; 11.25-µm band= goet­ of a templating mat community anddid so in Archean oceans. What hite; and 20-µm band= possibly goethite. we don't yet know is the dimensions of the gray zone within which Advantage of Full Spectrum: One lesson to be learnedfrom the products of biology andphysical processes are difficult or im­ these measurements is the importance of obtaining the full spec­ possible to distinguish. As the ALH 84100 debate illustrates, similar trum in order to search for unexpected constituents. A multiband problems attend the interpretation of microscopic morphologies, radiometer has the advantage of lower data rate requirements, and organic molecules, and isotopic signals found in rock samples. this is animportant consideration fora spacecraft instrument. How­ Taphonomic processes, the processes by which a dead organism ever, a multiband radiometer cannot be used to uniquely identify becomes a fossil, exacerbate the problem. Organisms decompose 26 Workhop on Martia Meteorites

in characteristic ways, but degradation can strip them of their di­ Biochemical Origins and the Search for Life in the Universe (C. agnostic chaacters. Cosmovichi et al., eds.), pp. 391-399, Editrice Corpositori, Bo­ Environment is a key aspect of any discussion of potentially bio­ logna, Italy. [7] Yushkin (1996). logical fatures in ancient terrestrial or matian rocks. The pattess formed by biological and physical processes are environment spe­ cific, so nubced debates about the interpretation of ancient fa­ TH STRUCTURE OF FIIGS I ALLAN HILLS 8001 tures require some estimate of the conditions under which MAY HIT AT THEIR IORGANIC ORIGI. L. V. Ksb­ encompassing rocks were deposited bd diagenetically modifed. fmality,Space Research Institute, Moscow, Russia. To give a single exbple, biologists have recently reprted the pres­ ence of extremely small bateria in the humb blood stream, but The ALH 84001 lessons showed that science at the end of the this places fw limits on the interpretation of objects fred in di­ 20th century is ready for discovery of the simplest organisms on lute aqueous environments. celestial bodies, where minimum conditions fr lif exist. These It is hard to overstate the importbce of understanding the lim­ conditions, as well as the ways of the origin of primitive microor­ its of physics and chemist in pattes formation, because at present ganisms, are already understood and described in the scientific lit­ we have little way of knowing which fatures of contempora ter­ erature. McKay et al. [1] reported that the SNC meteorite ALH restial biology are likely to prove universal and which may be the 84001 possibly contains traces and fossils of bcient primitive lif peculiar products of Earth's specific evolutionary history. The bot­ fom Mars. Neverheless, the authors of this very frst study be­ tom line is that we need to lea a great deal more about morpho­ lieved that a highly comprehensive verifcation bd more caeful logical, geochemica, bd sedimentological pattes frmation by further investigation were needed. In this paper, a few remarks on physical processes. We need to know more about the limits of bio­ the origin of the stsctures in the ALH 84001 meteorite are pro­ logica pattes fration, especially in environments likely to ap­ posed. They may be of interest as a possible proof of their abio­ proximate those on Mars. And at the end of the day, we must admit genic nature. that when we have leaed much more thb we know now, the gray One of the most intguing electron-microscopic photos is the zone will be better defned, but it won't disappea. Certainty about peculiar morphology observed in the globule, shown in Gibbs and martan lif will only become possible if and when we fnd evi­ Powell [2] and in McKay et al. [1]. Te fatures are microscopic dence that lies outside of the gray zone. elongated formations bd resemble the fossils of a colony of an-

MARTIAN BIOGENIC ACTIVITY: LOOKING FOR VIRUSES AND DNA TRACES INSTEAD OF EXTANT BACTERIA TRACES. L. V. Ksanforality, Space Research Institute, Moscow, Russia.

A curent program of investigations of organic components in ALH 84001 meteorite fndings includes a study of the polyaomatic hydrocabns taces [e.g., 1,2], C-isotopic balysis [e.g., 3], a seach fr amino-acid traces [4], O-isotopic analysis [5], etc. All of these studies are based on a hypothesis about a presence of martib pre­ historic primitive lif traces in the SNC meteorite fom Mas. The hypothesis is based on the contemporary existing notions of the ori­ gin of life that originated in natural fashion, through numberless chemical reactions, which were highly probable under the condi­ tions of young Earth [6]. Tere are a number of proofs that these same conditions occured early in martib history, which means lif could have originated on Mars as well. The only known terestial lif form is aino-nucleic-acid life that uses nucleic aids a b infration system. Primitive life frms include both microbes bd viruses. It is known that viruses are able to withstand much more severe conditions than bacteria [7]. Their inactive forms may sutive fr a long time until fvorable condi­ tions occur again. Thus, it could make sense to look fr visses or even DNA traces both in the body of the ALH 84001 meteorite and on Mars in future space missions. References: [1] McKay D.S. et al. (1996) Science, 273, 924- 930. [2] Clemett S. bd Zare R. (1997) Planetar Systems - Te Long View, 9emes Rencontres de Blois, pp. 22-28. [3] Jull A. J. T. et al. (1998) Science, 279, 366-369. [4] Bada J. L. et al. (1998) Science, 279, 362-365. [5] Galimov E. M. (1997) Solar System Res., 31(3),183-190. [6] de Duve C. (1997) in Astronomical and Fig. 1. LPI Contribution No. 956 27

cient terrestrialbacteria from travertinesand limestone. Besides the tian meteorites were the key to establishing the connection between fact that these ovoids areapproximately the same size, which is not the samples and the planet Mars [1,2]. More recent studies have the case for terrestrial bacteria [3 ], there is an another interesting explored the volatile history of Mars as recorded in the samples in detail. All the elongated ovoids are apparently separated from the increasingdetail. solid top layer and the lower layer. If so, it can be presumed that Working Hypothesis #1: Martian magmas are relatively d, all the bacteria constituted a continious layer before the separation. indicating a d martian mantle. It seems more likely that one can see in this photo the process of Observational evidence. Hydrous primary (magmatic) minerals fragmentation of some inorganic film. Such processes are known are rare in all of the martian meteorites. They contain two main to exist and have nothing in common with bacteria [4]. types of primary hydrous minerals: apatite (a Ca-phosphate) and Anotherphoto of a very small structureresembling a worm fossil kaersutite (a Ti-rich amphibole). A single 15-µm biotite has also by its shape was published by Kerr [5] and McKay et al. [1], as been found in Chassigny [3]. However, the dominant phosphate possible evidence of ancient primitive life on Mars. For compari­ mineralogy in the samples is merrillite (also referred to as whit­ son, a set of electron-microscopic photos of four microstructures lockite, although merrillite is probably a more properdesignation), found in kerite frompegmatites in granites (Volin' area in Russia) an anhydrous Ca-phosphate. On Earth, merrillite is essentially ab­ canbe examined (Fig. 1). The images were obtained and published sent, probably attesting to a significant differencebetween the vola­ by Yushkin [6] and reproduced by Galimov [7]. One cansee how tile contents of magmas on the two bodies. Even in the most unusual and strange the shapes of inorganic structures of some min­ evolved rock compositions, such as shergottite QUE 94201, which erals are. Their structures mimic some bacteria and may be mis­ should have the highest concentrationsof volatiles, merrillite is the leading as to their real nature. dominant phosphate, constituting -4% of the sample, with apatite References: [1] McKay D. S. et al. (1996) Science, 273, 924- present at the <1% level [4]. 930. [2] GibbsW.W. and Powell C. S. (1996) Sci. Am., 275, 12- The kaersutites in the martian meteorites arefound only inside 13. [3] Schopf W. (1997) Planetary Systems - The Long View, partially crystallized magmaticinclusions within pyroxene and oli­ 9emes Rencontres de Blois, pp. 22-28. [4] Ksanfomality L. V. vine [e.g., 3,5]. Their presence was initially used to indicate rather (1997) Solar System Res., 31(3), 175-182. [5] Kerr R. A. (1996) substantial water content for primary magmas of -1.5 wt% [3,6]. Science, 273, 864-865. [6] Yushkin N. P. (1996) J. Cryst. Growth, However, the low water content of the amphiboles( <0.2 wt% [7]) 167, 237-247. [7] Galimov E. M. (1997) Solar System Res., 31(3), probably indicates a significantly lower magmatic water content. 183-190. Implications. A low martianmantle water content has been used to support a homogeneous accretion model forMars [8,9] in which volatile-richand metal-rich planetesimals accreted togetheron Mars, CONSTRAINTS ON MARTIAN VOLATILE HISTORY the metal was oxidized and the water reduced to H, which escaped. FROM STUDIES OF MARTIAN METEORITES: LES­ Preservation of this low water content supports the suggestion that SONS LEARNED AND OPEN QUESTIONS. L. A. Leshin, plate-tectonic-style recycling of surface materials (which appear to Department of Geology, P.O. Box 871404, Arizona State Uni­ be more volatile rich - see below) has been unimportant in mar­ versity, Tempe AZ 85287-1404, USA ([email protected]). tianhistory. Open questions. Although the mineralogical observations sup­ Studies of the shergottites, nakhlites, Chassigny, and port low magmatic water content, questions remain about exactly ALH 84001 have provided valuable information on the origin, evo­ what conditions are necessary for the production of the phases lution, and interaction of martian volatile reservoirs. Understand­ present in martianmeteorites. For example, kaersutites such asthose ing these reservoirs provides the broadest possible context within in the meteorites, which contain up to 11 wt% TiO2 [3,5], have not which to consider the possibility of life on Mars, and thus such stud­ been successfully synthesized in the laboratory. By understanding ies are an important component of the quest for addressing issues the parameters under which these types of minerals form, we will related to a martian biosphere. The purpose of this work is to out­ be able to better constrain the volatile contents of the martian mag­ line some of the importantinsights into martianvolatiles provided mas that produced the meteorite samples. by studies of martian meteorites, and to highlight inconsistencies The two-component homogeneous accretion model of [8] may and gaps in our knowledge that could be addressed by further re­ not be capable of producing a Mars with the proper moment of in­ search on the meteorites themselves, or by future missions, includ­ ertia [10]. In addition, the efficiency of the accretion-H reduction­ ing sample returns.This is by no means meantto be an exhaustive loss process for eliminating much of Mars' primordial water is review of all that has been learnedfrom studies of volatiles in mar­ unknown. Thus, many openquestions remain about the mechanism tianmeteorites. Rather, it is intended to highlight major conclusions for accreting Marswith a dry interior. and to stimulate discussion about "what we really know." Working Hypothesis #2: There is a substantial reservoir of In terms of the number of different samples (currently 13, of water (and other volatiles) in the martian crust that interacts with which 12 have been extensively studied), the amount of material crustal rocks. available, and, perhaps most importantly, the diversity in rock types Observational evidence. The evidence for the interaction of ( or relative lack thereof), our current collection of martian meteor­ martian crustal volatiles with crustal rocks comes in two main ites represents an extremely limited sample base. Considering this, forms: mineralogical and isotopic. Secondary minerals, from sul­ the influence of martian meteorites on our current thinking about fates to carbonatesto clays, have been found in all types of mar­ "global" volatile issues on Mars is impressive. This is especially tian meteorites [e.g., 11]. These minerals, which in most cases can true when consideration is given to the overall paucity of volatile­ be convincingly shown to be preterrestrial, attest to the ubiquitous bearing phases in these samples. Initial studies of volatiles in mar- presenc� of fluids in the martian crust. However, it should be re- 28 Workshop on Martian Meteorites

membered that in all the samples, these secondary minerals repre­ H. Y. Jr. et al. (1996) GCA, 60, 4563. [S] Treiman A.H. (1985) sent minor components of the rocks, suggesting rather low fluid­ Meteoritics, 20, 229. [6] Mcsween H. Y. Jr. andHarvey R. P. rock ratios in the alteration environment. Isotopic studies point to (1993) Science, 259, 1890. [7] Watson L. L. et al. (1994) Science, exchange ofH in some magmatic minerals after crystallization [7], 265, 86. [8] Dreibus G. and Wanke H. (1985) Meteoritics, 20, 367. and to the postmagmatic origin of many of the carbonate minerals [9] Carr M. H. and Wanke H. (1992) Icarus, 98, 61. [IO] Bertka in the samples [e.g., 12-16]. C. M. and Fei Y. (1998) LPS X, Abstract#1252. [11) Gooding Implications. The interaction of fluids and rocks appear to be J. L. (1992) Icarus, 99, 28. [12) Carr R.H. et al. (1985) Nature, an important crustal process on Mars, as it is on the Earth. This 314, 248. [13] Leshin L. A. et al. (1996) GCA, 60, 2635. conclusion is especially important because it suggests that "envi­ [14] Leshin L. A. et al. (1998) GCA, 62, 3. [15) Romanek C. S. ronments" on Mars that could besuitable forlife (e.g., "hydrother­ et al. (1994) Nature, 372,655. [16] Valley J. W. et al. (1997) Sci­ mal" environments) are plentiful, since some evidence for such ence, 275, 1633. [17] Pepin R. 0. and CarrM. H. (1992) in Mars, environments is found in all the martian samples we have. p. 120, Univ. of Arizona, Tucson. [18] Jakosky B. M. and Jones Open questions. There aremany open questions about the fluid­ J. H. (1997) Rev. Geophys., 35, I. [19] Farquhar J. et al. (1998) rock interaction in the martian crust. What is the source of the flu­ Science, 280, 1580. [20] KarlssonH. R. et al. (1992) Science, 255, ids in the crust and what is their abundance? Has this reservoir of 1409. [21] Jakosky B. M. (1993) GRL, 20, 1591. volatiles changed with time due either to magmatic influx or at­ mospheric loss? What are the conditions under which the martian samples have experienced fluid-rock interaction?What are the tem­ THE FUTURE SEARCH FOR LIFE ON MARS: AN peratures, durations, and timing of these events relative to the ig­ UNAMBIGUOUS MARTIAN LIFE DETECTION EXPERI­ neous crystallization ages of the samples? More detailed analyses MENT. G. V. Levin, Biospherics Incorporated, 12051 Indian of the alteration products, an active areaof currentresearch, should Creek Court, Beltsville MD 20705, USA. shed light on at least some of these questions. Working Hypothesis #3: Crustal and atmospheric volatiles A simple, robotic method to provide an unambiguous determi­ have interacted over time. nation of extant life in the martian surface or subsurface material Obserational evidence. From groundbased and Viking obser­ is presented. It cannot render a false positive, nor can it mistake a vations, we know that the current martian atmosphere has a unique chemical response fora biological one. isotopic signature in H, C, N, Ar, and Xe, among others [e.g., Not since the landingof Viking has interest in life on Marsbeen 17,18]. In many cases this unique signature can be traced to loss so high. The finding [ 1] of possiblebiochemicals andmicrobial fos­ of atmospheric constituents accompanied by isotopic fractionation. sils in meteorites attributedto Mars has provided incentive for many This "atmospheric" signature can be found in the volatiles intro­ new scientific investigations.Similar inclusions have recently been duced into the martian meteorites by the fluid-rock interactions found [2] in nonmartian meteorites. The latter have rekindled in­ discussed above [e.g., 12,13,15,19,20). In terms of O-isotopic sig­ terest in earlier reports [3,4] of meteorites containing "organized natures, three isotope analyses of water and carbonates from the elements." Exquisite techniques arebeing developed [SJ to exam­ meteorites have shown that the crustal volatile reservoir is not in ine and analyze these minute inclusions, and to distinguish them isotopic equilibrium with the silicate crust [19,20]. frompossible terrestrialcontaminants. Implications. All isotopic data point to interaction of the Intensive efforts are under way in a number of laboratories to atomosphere and crustal reserviors of volatiles. The O-isotopic data develop reliable biomarkers for use in examining meteorites in labo­ that indicate a lack of equilibration between the crustal fluid and ratories, returnMars samples, and samples obtained on robotic mis­ rock O reservoirs would seem to require that fluid-rock exchange sions to Mars and other celestial bodies. Chemical, biochemical, took place for short periods and/or at relatively low temperatures genetic mapping, physical, optical, computer recognition of mor­ to preserve the observed disequilibrium. phological attributes, and laser ablation and sniffing techniques, Open questions. The mechanism by which the interaction be­ along with their appropriate instrumentation, are under develop­ tween crustal and atmospheric volatiles takes place is very uncon­ ment. These techniques are focused on the detection and study of strained. The proposal that atmosphere-crust exchange took place fossilized cellular remains. Other efforts are aimed at drilling to through "hydrothermal systems" [18) fails to explain physically how depths where lenses of liquid water with live microorganismsmay this exchange would take place. Clearly, the fluid-rock exchange exist. However, none of these biomarker recognition methods can can happen in this way, but the atmosphere-crustal fluid exchange distinguish living organisms fromdead ones. remains unconstrained. Recent developments have resuscitated the possibility that live The origin and time history of some of the atmospheric isoto­ microorganisms might be indigenous on the surface of Mars. Liv­ pic anomalies remains unknown. Of particular interest is the 0 ing microorganisms recovered [6] from permanent ice in which anomaly, which could be an indicator of photochemical processes minute quantitiesof liquid water are extractedhave provided a good in the martian atmosphere, of atmospheric loss processes, or of a model for Mars, since evidence of frozenwater deposits was often crustal reservoir that has been out of equilibrium with the crustsince seen on the surface by Viking. An analysis [7] of the dynamics of its formation [19-21). This latter possibility would suggest the the martian atmosphere made possible by Pathfinder data shows crustal and atmospheric volatiles were supplied from some exotic that liquid water is most likely present diurnally on the martian sur­ source such as a late cometary input. face. The calculated amounts aresufficient to sustain soil microor­ References: [1) Bogard D. D. andJohnson P. (1983) Science, ganisms on Earth. 221, 651. [2] Becker R.H. and Pepin R. 0. (1984) EPSL, 69, 225. The new life detection method is based on the legacy of the Mars [3] Johnson M. C. et al. (1991) GCA, 55, 349. [4] Mcsween Viking Labeled Release (LR) experiment [8], which is gaining re- LP Contribution No. 956 29

newed attentionin view of the recent discoverieson Earth and Mars J.C.et al. (1998) Science, 280, 2095-2098. [7] Levin G. V. and cited above. In addition, the salient argumentsagainst the LR find­ Levin R. L. (1998) Proc. SPIE, 3441, 30-41. [8] Levin G. V. ings have been answered: (1) Theories that H202 caused the LR (1997) Proc. SPIE, 3111, 146-161. [9] Krasnopolsky V. et al. results have been negated by recently reported [9] direct observa­ (1997) JGR, 102, 6525-6534. (10] Biemann K. et al. (1977) JGR, tions that established an upper limit of 2 x 10-3 precipitable µm of 82, 4641-4658. [11] Gilichinsky D. (1998) Russian Acad. Sci., H202 forMars (this miniscule amount, were it present, could not personal communication. have produced the LR results); and (2) the failure of the Viking GCMS (10] to detect organic matter on Mars, often cited as proof against the presence of microorganisms, has been contravened by CURATION OF U.S. MARTIAN METEORITES COL­ the finding (11] of fully metabolizing microorganisms in perma­ LECTED INANTARCTICA. M. Lindstrom1, C. Satterwhite2, frost in which no organicC was detected by GCMS. J. Allton2, and E. Stansbery1, 1Mail Code SN2, NASA Johnson The enhanced LR experiment exploits the fact that all known Space Center, Houston TX 77058, USA (marilyn.m.lindstroml@ lifeforms make andutilize L-amino acids and D-carbohydrates pref­ jsc.nasa.gov), 2Mail Code C23, Lockheed Martin, Houston TX erentially over the respectivestereoisomers. On the other hand, no 77058, USA. natural chemical reactions can distinguish between stereoisomers. Therefore, any strong response by an unknown agent to one iso­ Meteorites: To date the ANSMET field team has collected mer of anadministered compound over its stereoisomer constitutes fivemartian meteorites (see below) in Antarctica andreturned them indisputable proof that the agent producing the reaction is biologi­ forcuration at the Johnson Space Center (JSC) Meteorite Process­ cal. A modified,miniaturized LR instrument can administer L and ing Laboratory(MPL). The meteorites were collected with the clean D isomers of amino acids and carbohydrates individually to dis­ procedures used by ANSMET in collecting all meteorites: They crete portions of the same soil sample. A stereospecific response were handled with JSC-cleanedtools, packaged in clean bags, and would be proof of life. Should stereospecificity be found, and shipped frozen to JSC. The five martian meteorites vary signifi­ should it be the same as that of terrestrial life, a possible relation­ cantly in size (12-7942 g) and rock type (basalts, lherzolites, and ship between the two lifeforms would be indicated. Should the ste­ orthopyroxenite). Detailed descriptions are provided in the Mars reospecificity differ from that of Earth life, an independent origin Meteorite Compendium [l], which describes classification,curation, of lifeforms would have been established. A variety of thermal con­ and research results.Table 1 gives the names, classifications, and trols on replicate samples could determine the thermal sensitivity original and curatorial masses of the martianmeteorites. of anylife forms found andwould also serve as aredundant veri­ Meteorite Processing Laboratory: The MPL is a class- ficationof the biological result. 10,000clean lab containing N atmosphere glove-box cabinets and Using new technology, the LR instrumentcan be simplifiedand laminar flow benches for meteorite processing and additional N miniaturized. It canbe deployed on the planetary surface,or, placed cabinets andstainless-steel open-air shelves forsample storage. All in penetrators, it can sampleat depth. The instrument can be lodged tools and sample containers are made of restricted materials and in a cocoon and the entire assembly heat-sterilized before being are carefully cleaned using the same procedures used for lunar mated to the spacecraft. Upon landing, the cocoon door would pop sampletools. All sample handling follows strict published proce­ open, and, in the same motion,the instrument would be ejected be­ dures, andall staff are trained in those procedures. All sample han­ yond the contamination zone of the lander, thereby eliminatingthe dling is alsodocumented by weighing and photographing samples high cost of spacecraft sterilization. Two-way radio communica­ and kept permanentlyin computer and hard-copyrecords. tion between the instrument and the lander would provide for in­ Although the meteorites were handled using procedures stan­ structionsand data exchange. dard for achondrites, those procedures depend on sample size and It is recommended that this simple, small, easily and cheaply have varied over time. Thus, the processing histories of the mar­ developed instrument be flown on the next Mars mission and on tianmeteorites differslightly. Because most of the meteorites were missions to other candidatelife-bearing bodies. identifiedas achondritesby the fieldteam, they were processedearly References: [l] McKay D. S. et al. (1996) Science, 273, 924- in the classification queue (LEW 88516 is an exception), and all 930. [2] Hoover R. B. et al. (1998) Proc. SPIE, 3441, 203-216. samples were stored only in N cabinets. The medium and large [3] Nagy B. et al. (1962) Nature, 193, 1129-1133. [4] Rossignol­ samples (ALHA 77005, EETA 79001, and ALH 84001) were ini­ Strick M. and Barghom E. S. (1971) Space Life Sci., 3, 89-107. tially processed and sawed in N cabinets. Slabs and small splits of [5] Westfall F. et al. (1998) Proc. SPJE, 3441, 225-233. [6] Priscu these meteorites were processed on flow benches. The small

TABLE I. U.S. Antarctc martian meteorites.

Thin Sectons Allocated Chips Name Class Original Mass (g) MWG Mass(g) Curator Mass (g) TS PB No. of Chips Mass (g) EETA 7901 S-basalt 7942 6694 798 Sil 74 20 141 263 QUE 94201 S-basalt 12.0 6.4 0 12 19 3.5 ALHA 77005 S-lherzolite 482.5 185 212 N 32 14 47 45.8 LEW 88516 S-lherzolite 13.2 8.3 0 11 23 3.0 ALH 8401 Opxite 1931 1470 213 S 52 14 160 217 30 Workhop on Martian Meteorites

ContaminationControl: The MPL has alwayshad strict mea­ sures for controlling inorganiccontamination. These measures were patterned after those developed for lunar samples, but were less rig­ orous due to the residence time of the meteorites in Antarctica.With the announcement of possible fossil life in one of our martian me­ teorites, the requirements for contamination control suddenly in­ cluded organic and biological contatninants as well as inorganic ones. The MPL staff did a simple self-assessment of possible sources of organic contamination. The top of the list was human staff working in the lab, followed by plastics in bags and contain­ ers, organic soaps, and solvents. We requested review by an out­ side comtnittee of specialists in organic chetnistry and biology, and NASA Headquarters appointed Dr.Jeff Bada to head the Organic Contamination Review Comtnittee. The group met at JSC in De­ cember 1996 and subtnitted a report to NASA Headquarters in March 1997 [2].Their basic conclusion was that the MPL did not have obvious organic contamination (except as noted above), but Fig. 1. Martian meteorite EETA 79001. that we had not monitored or documented the levels of tninor con­ tamination.JSC set out in 1997 to elitninateas many of the noted contaminants as possible andto docwnent those that remained. samples (LEW 88516 and QUE 94201) were always processed on The first steps were taken to process ALH 84001 samples for a flowbench, but sample containers were later sealed in a N atmo­ both organic and inorganic analysesin 1997. First the meteorites sphere. All samples had some exposure to air in the clean room, were isolated from other achondrites in processing to avoid any as they did formuch longer times in Antarctica. cross-contamination.Next, all plastics were removed fromthe Mars The most intense allocation of a given sample to researchers is Meteorite Processing Cabinet. The major change was the develop­ just after its recovery and classification. The exception is ment of new procedures forcleaning the cabinet. Instead of using ALH 84001, which was originally classifiedas a diogenite and re­ soap and organic solvents, we now use hot (140°C) ultrapure wa­ classified as martian nine years later. The reclassification and the ter (UPW) and monitor the three rinses. The UPW does an excel­ announcement of possible fossillife triggered major allocations in lent job of removing particulates and leaves low levels of total 1994 and 1997. For the small samples, little allocation is done at a organic C (TOC). later date. For the largest sample, EETA 79001, there has been a The second phase of organic contaminationcontrol is in the pro­ steady demand forsamples. For the three largestsamples, some of cess of being implemented with FY98 funds from NASA Head­ the material has been transferred permanently to various meteorite quarters andthe JSC Center Director's Discretionary Fund (CDDF). curatorial partner institutions. Allan Hills 77005 was collected The UPW system is being upgraded and more monitoring and docu­ jointly with Japan and the sample split with NIPR in Tokyo.Both mentation of contamination is being done [3]. The plans for mar­ the Stnithsonian and JSC have display samples of EETA 79001, tian meteorite curation can be used as a starting point for curation while only the Stnithsonianhas a ALH 84001display sample.The of returned martian samples. remaining meteorites remain the propertyof NSF and areallocated References: [1] Meyer C. (1998) Mars Meteorite Compen­ by the Meteorite Working Group. dium, JSC curator. [2] Bada J. et al. (1997) Rept of Org. Contam. Rev. Comm. [3] Schilling E. and Schneider M., this volume.

PRINCIPAL COMPONENT ANALYSIS FOR BIOSIGNA­ TURE DETECTION INEXTRA TERRESTRIAL SAMPLES. G.D. McDonald and M. C. Storrie-Lombardi, Mail Stop 183-301, Jet Propulsion Laboratory, 4800 Oak Grove Drive, Pasadena CA 91109, USA ([email protected]).

Analysis of extraterrestrial samplesfor organic signatures of past or present life presents several problems. Chief among these is dis­ tinguishing bona fide extraterrestrial organic material from terres­ trial contamination, either carried on a spacecraft or present in the terrestrial environment to which the sample is exposed. A related problem is separating biologically derived molecules from those produced by abiotic syntheses in the interstellar medium, on mete­ orite parent bodies, or in planetary atmospheres andoceans. We have begun to address these problems using a technique, principal component analysis (PCA), borrowedfrom classical mul­ Fig. 2. The JSC Meteorite Processing Laboratory. tivariate analysis.Also known as the Karhunen-Loeveor Hotelling LPI Contribution No. 956 31

transfrm, PCA identifes linear combinations of raw paameters tors were multiplied by the amino-acid relative fequencies for 108 accounting fr maximum variance in the dataset. Denoting input terestrial protein families [1,2], seven Murchison meteorite amino­ spectra as S;,1, where i represents n input parameters fr each spec­ acid determinations [3-6], fur Allan Hills meteorite assays [7], tra and I indexes the m taining specta, PCA frst fnds the mean and one Allan Hills ice sample [7]. Ten factors extracted by PCA spectum, 6S;,1, such that 6S;,1 = 6S;,1 - 1, calculates the fom the resulting matix accounted fr >99% of the varicce in digerence between this mean cd each individual spectra, then es­ the test samples. Since PCA trcsposes the data to zero mean cd timates the paameter covaricce matrix, Ci,k• where unit variance, two-dimensional plots of orthogonal factors repre­ sent the dispersion of the amino-acid distributions in units of c. This mecs that in a graphic representation such as that shown in Fig. 1, if the weighted centers of two clusters appear, fr exbple, and decompses this into consttuent eigenvalues cd normalized >2 units apar, then the clusters difer by more thc two stcdard eigenvectors. If we restict the decompsition to the frst p terms, deviatons. Analysis of the fatures comprising the factors provides the method yields the optimum (in a least-squares sense) linea re­ insight into the portions of the amino-acid distbution contibut­ constuction of the input signal using the fewest p pabeters. Since ing to the clustering. In Fig. 1 we have chosen to plot factors 3 the covariance matrix is an average over may spectra, cd since cd 10. Serine, and to a lesser extent proline, isoleucine, and glu­ noise is uncorrelated between specta, the method is quite robust tamic acid, contibute to factor 3, while glycine, and once again to modest bounts of noise. The key advatage of this algorithm proline and serine, contibute to factor 10. is its ability to shift the measurement output of dispaate sensors The presumably abiotc Murchson amino acids frm a cluster to a common statistical metic. Specifcaly, datasets shift to zero distinct fom the terestial protein families, although partial ter­ mean and unit varicce. This transfrmation is a central prerequi­ restrial biological contamination in some of the Murchison analy­ site fr scientifcally testing a null hypothesis and addressing que­ ses is indicated. The profles fom various extacts of ALH 84001 ries about the existence of biomarkers in a specifc unknown are also distinct fom proteins, but less tghtly clustered. The Allan sbple. Hills ice sample also does not cluster tightly with the protein fami­ Ten terestial amino acids (alanine, aspartc acid, glutamic acid, lies, suggesting that geochemical fractionation of the biogenic glycine, isoleucine, leucine, proline, serine, threonine, and valine) amino acids deposited in the ice has taken place at some point. fund in the Murchison cabonaceous chondrite, Mas meteorite The PCA method is particularly suited fr sample analyses in­ ALH 84001, and/or a Allan Hills ice sbple were chaacterized volving datasets fom a variety of disparate instruments designed using 13 physiochemical fatures including hydrophobicity, pKa, to detect and quantif digerent classes of organic compunds. Such tendency to form alpha helices, residual mass, and preferred data can be reduced to common compositional or physiochemical Ramachcdrc angles. PCA feature extaction provided fve fac­ factors ad combined to generate a profle of the overall organic tors accounting fr more than 85% of the variace. These fve fac- composition of a sample. This profle can then be compared with profles of known biological cd nonbiological systems to assess the probability that the extaterestial sample contains evidence of 5 past or present life. 0 References: [l] Dayhog M. 0. (1972) Atlas of Protein Se­ 4 quence and Structure, vol. 5, Natl. Biomed. Res. Found. [2] Day­ hog M. 0. (1978) Atlas of Protein Sequence and Structure, vol. 5, 3 supl. 3, Natl. Biomed. Res. Found. [3] Cronin J. R. cd Pizzaello

2 0 S. (1983) Adv. Space Res., 3, 5-18. [4] Engel M. H. and Nagy B.

0 (1982) Nature, 296, 837-840. [5] Engel M. H. et al. (1990) Na­ ture, 348, 47-49. [6] Engel M. H. and Macko S. A. (1997) Na­ � ture, 389, 265-268. [7] Bada J. L. et al. (1998) Science, 279, s 362-365. M -1 A(ice)

-2 M EVIDENCE FOR ANCIENT LIFE IN MARS METEOR­ ITES: LESSONS LEARNED. D. S. McKay, Mail Coe SN, -3 M :wvi NASA Johnson Space Center, Houston TX 77058, USA (david. [email protected]). -4 A

-5 The lines of evidence we frst proposed [l] as supporting a hy­ -5 -3 -1 pothesis of ealy life on Mas ae discussed by Treimc [2], who Factor3 presents pros cd cons of our hypothesis in the light of subsequent Fig. 1. The interaction of principal component factors 3 and l 0 dis­ research by many groups. Our assessment of the current status of tinguish both ALH 84001 (A) and Murchison (M) amino-acid dis­ the mcy controversies over our hypothesis is given in Gibson et tributions from that of terrestrial protein families ( o) by as much as 1- al. [3]. Rather that repeat or elaborate on that infrmation, I prefer 4 cr along both dimensions. Six of the seven Murchison samples form a to take an overview and present what I think are some of the "les­ cluster in the lower left quadrant distinct from both the protein and the sons leaed" by our team in particula, and by the science com­ Allan Hills meteorite and ice samples. munity in general, as a result of the martian meteorite studies. 32 Workshop on Martian Meteorites

1. Mars meteorites are more complicated than we thought. Re­ mation age of the magnetites, Fe sulfdes, gypsum, bd other sec­ turned Mars samples are likely to be more complicated than we ondar minerals is simply not known. If we could date their fr­ can imagine. mation ages, we would have a better understanding of the rock First, we have leaed that ALH 8401 is b exceedingly com­ histor, the pssibility that some are biomakers, bd even where plex rock that has withstood multiple shock and heating events they were frmed. Reted Mars rocks (as well as samples fom (Treimb - meteoritics). It is an igneous rock with an overprint Europa, Titan, Io, etc.) ae likely to also have both primary fa­ of impact fatures and secondary alteration, some of which is high tures and secondary features, and telling them apart and determin­ temprature bd some of which is low temprature (Goping, Went­ ing their frmation bd alteration timetables becomes very imprtant worth). The time bd place of the low-temperature alterations are to understanding their history. Primary crystallization age dating, not well understood. Some ma have occured on Mars, bd some while important, is not the only gae in town. may have occured in Antarctica. One of the frst lessons is that, at 4. A robotic mission analyzing AUi84001 would have totally leat fr Antactic meteorites, it is very diffcult (but not imps­ missed the data, the hypothesis, and the whole controversy. sible) to sepaate indigenous features fom fatures acquired in Ant­ If we had sent a mission to Mas bd that mission encountered actica or the laboratory. However, it adds enormously to the task ALH 84001 on the surface and analyzed it with the Pathfnder/So­ of reconstucting the histories of these mteorites. For reted Mas jouser instruments or the instruments proposed for subsequent rocks, we will be dealing with saples fom a plbet whose geo­ lbder missions, it is likely that the data and observations would logic history is only poorly understood, whose impact, volcanic, not have included any of the major lines of evidence we and oth­ sedimentary, metamorphic, and weathering processes ae major ers have described in this rock: P AHs and their location and pro­ mysteries. The reted samples are likely to show the egects of fle, submicrometer mnerals of several types, carbonate pbcakes any one of these processes bd may show the efects of all of them. of widely ranging mineral, elemental, bd isotopic compositions, And that's befre we consider the possible efects of living sys­ and small microfossil-like features. Furthermore, without the tems. Unless we ae extemely caefl, we may also have alteration ALH 84001 exprience, it is unlikely that a rock with a basaltic bd contamination acquired during sampling, ret, unloading, bd composition and obvious igneous txture would even be consid­ initial processing. Then we are fced with a situation analogous to ered as a possible site fr microbial lif frs. None of the data the Antarctic contamination problem. that has caused the contoversy about ALH 8401 comes fom the 2. We need better and more reliable biomarkers. bulk major-element composition or the major mineralogy. If we The problem with ALH 84001 is not the lack of potential only had that dba, however accurate, we would likely have missed biomakers. It clealy has potential biomarkers, including orgbic what we consider to be the most interesting features of this rock. compounds (PAHs, amino acids, and kerogenlike frs), carbon­ This is not to criticize the fight instument packages, but to ates, magnetites, sulfides, bd fssillike frms. The problem is that pint out how subtle and small the fatures ae, how important their we do not understbd by of these biomarkers well enough to be spatial locations ae within the rock, bd how diffcult it is to ana­ confdent that the properties of these features in ALH 84001 truly lyze fr them. Clealy, the entire ALH 84001 controversy could indicate life-related processes as opposed to strictly nonbiologic only come fom analysis of reted rocks with complex state-of­ processes. Are the minerals precipitated by stictly inorganic chem­ the-a laboratory equipment. isty, or are they precipitated with the ad of living organisms or 5. lt is sometimes difficult to either prove or disprove a hy­ organic compounds derived fom living organisms? Are the PAHs pothesis. and amino acids all derived fom living systems, or were they gen­ Our orginal hypothesis was that several observed features in erated by inorganic processes? (This is a sepaate question fom ALH 84001 could all be reasonably explained by the presence of possible Antactic contamnation.) Are the fssillike morphologies early life on Mas. This was a hypthesis, and the original paper tuly frmed fom microbes and their products, or are they frmed was filled with qualifers. The hypothesis, was based on circum­ by nonbiologic precipitation? ls the extreme C-isotopic faction­ stantial evidence or scientifc obserations of several diferent typs, ation a result of biologic processing? All these fatures ae present and the hypothesis seemed to explain all of the evidence with a in ALH 84001. But do we understand their proprties and occur­ single kind of event, namely the presence in the rock of microbial rences well enough to be confdent which ae true biomarkers and activit. Many other investigators have been able to reproduce our which ae impostors? I think the answer is no. Before we bring original data, although others have added a number of new kinds samples back fom Mars, we must have in a database, a set of tested, of data to our original observations. Only in one or two cases have reliable, and certifed biomakers with well-doumented distinguish­ completely contadictory data been repred. ing properties so that we can search for and describe these Refting one line of circumstantial evidence supporting a hy­ biomakers in the reted samples and use them to answer un­ pothesis (by either showing that the data were wrong or that the equivocally the question of whether or not there is evidence fr lif proposed explbaton can be replaced by a clealy better one) does in these Mas samples. We need to be confdent that the biomarkers not mean that the hypothesis is wrong; it simply means that that we use cannot be frmed by by conceivable nonbiologic process. line of circumstantial evidence is incorect. Whle it may weaken We also need to decide whether a single biomaker is sufcient or the hypothesis, it does not invalidate it. Only when all proposed whether we need several. Conversely, we need to decide whether lines of evidence supporting a hypothesis ae refuted can that hy­ the absence of known biomakers is sufcient to eliminate the pos­ potheses be considered wrong (bd wrong hyptheses may beneft sibility of Mas lif. scientifc progress as much as right ones). A hypthesis cb be also 3. We need ways to date secondary processes. be considered wrong if none if its predictions ae correct. It has been extraordinaily difcult to date the age of the ca­ Conversely, the hypothesis can only be proved to be correct bonate pancakes in ALH 84001. No one has attempted to date the when either multiple kinds of circumstbtial evidence become over­ multiple shock and heating events recorded in this sample. The fr- whelming, or a single line of evidence is so strong that it cannot LI Contribution No. 956 33

be explained by ay alterate hypothesis - in other words, a smok­ tion in the Antactic, although this remains disputed [8], and a ing gun. A hypothesis can be greatly strengthened if its predictions expriment is needed to determine if a small preterestrial C com­ ae subsequently shown to be te. In our view, the ALH 84001 ponent is organic or merely undigested magnesite. Contnued stud­ hypothesis of early lif on Mas is still valid ad has neither been ies of other SNC meteortes or marta roks may reveal indigenous adequately refted nor adequately proved. The people who propse organics, and futue work should fous on biologically relevant or­ a hypthesis ae not necessarily responsible fr proving it because ganics rather tha PAHs. Microaalytcal techniques fr sepaat­ they may not have the correct tools, technical specialties, or ing and identifying compounds and measuring their isotopic samples, although they may have a moral obligation to either ty composition and optical btivity ae needed. The nature of a pro­ to prove it or to refte it themselves. For ALH 84001 and the other psed martia surface oxidant ad its ability to destoy organic ma­ martan meteorites, the question is still open and may not be an­ terials must be understood fom in situ measurements of martian swered until martian samples are reted by a space mission. soil. Some progress in recognizing biogenic minerals simila to When we do get samples back, we should expct that a number putative biominerals in ALH 8401 has been made, ad the inter­ of digerent hypotheses will be advaced by vaous investigators, relatonships between morphologies, interal stuctures, and mecha­ because, as discussed aove, the rocks (ad soils) may be very com­ nisms of growt in biominerals ad inorganic minerals that resemble plex and may record may overlapping processes. The hyptheses them must be pursued [9]. The biology of nanoscale bacteria re­ advanced to explain their features may even be mutually exclusive mains contentious [10], as does the possibility of their fssiliza­ and contadictory. But we should concentate on the data, give all tion. Are there stable isotop biomakers, such as S [ 11 ], or perhaps the hypotheses a fair hearing, and attack the problems, not the combinatons of isotope biomakers that might be aalyzed in ma­ people. tia rocks? What is the mineralogy of the purorted nanofossils, References: (1) McKay et al. (1996) Science, 273, 924. ad what fatures are artifacts of sample prepaation (12]? The (2) Treima, this volume. (3) Gibson et al., this volume. search for biomakers on the surface of Mars will require the loca­ tion of promising sites (elastic sediments, evaporites, and hot springs deposits) by orbiters, surface mobility, the identfication of KEY RESEARCH: MRTIAN MTEORITES AND THff roks most likely to preserve biomaterials (phosphates, silica?), ad RELATIONSHS TO PLANETARY GEOLOGY AN PER­ signifcat technical advaces for in situ analysis. Most of these HAPS BIOLOGY. H. Y. McSween Jr., Department of Geo­ studies will probably necessitat reted, uncontaminated samples. logical Sciences, University of Tennessee, Knoxville TN 37996- How Much Water? The existence of liquid water on the an­ 1410, USA ([email protected]). cient surface of Mars is unquestioned, but its amount and timing ae disputed. Estimates of the amount of water outgassed fom the Introuction: Alla Hills 8401 ad other SNC meteortes ae martia interior vary over may orders of magnitude, with geo­ bewilderingly complex rocks that have generated as many questions chemical estimates [13] being systematically smaller than geologi­ a aswers. Here ae some remaining questions with profund im­ cal estimates [14). Geochemica estimates might be improved by plications fr the geological ad perhaps biological evolution of more quatitative understading of inital atmospheric gas abun­ Mas. dances and their behavior during hydrodynamic escape ad impact Whence Carbonates? The paeoenvironment of carbonate erosion, estimation of the water contibution made by comet ac­ frmation in ALH 84001 remains disputed. Various authors have cretion, ad exprimental determinaton of magmatic conditions re­ advocated reactions between the rock and aqueous hydrothermal quired to stabilize OH-defcient kaersutite. The global inventory of fuids (1) or hot metasomatic COz-rich fluids [2], precipitation by water on Mars is based on a accretion model using SNC meteor­ matian orgaisms (3), recrystallization during shock [4], or pre­ ites [15], which does not appar to be consistent with new estmates cipitation fom evaprating brnes percolating through subsurface of the plaet's moment of inertia [16). Although considerable wa­ factures [5]. Definition of the fll alteration mineral assemblage ter can potentially be stored within pre spaces in the martian crust might provide a crucial constraint on formation mechanisms. [17), the aount actually present ad its geographic ad vertical Formation temprature and the temperature rage remain critical distibutions are unown. Because of a lack of infration on soil paameters in assessing the plausibility of biologic arguments. De­ mineralogy, the amount of water stored in the regolith is also un­ fnitive rciometic ages of the cabonates have not yet been de­ known. A refined geochemical model for Mars, as well as high­ termined, and there remains some controversy about how many resolution orbital images fr studies of terrain softening, direct carbonate depositional events occurred [6]. The existence of sig­ orbital measurements of H distbutions, ad determination of soil

nifcat carbonate deposits frmed by collapse of a COzrich at­ mineralogy ae needed. An accepted crater stratigraphy, perhaps mosphere has implicatons fr the plaet's climate history. This may anchored to absolute time by the identfcation of SNC launch sites, be addressed by TES orbital data ad soil mineralogy determined might allow determination of when water wa abundant. Also, the by laders. The discovery of evaporites (e.g., carbonates, sulfates, relation of surface water to past climate needs to be resolved. and halides) on Mars, a well as mineralogic, chemical, isotopic, What Is the History of Biogenic Elements and Compounds and fluid inclusion studies of salts in other SNC meteorites [7], on Mars? The histories of H, C, 0, N, and S ae probably best might also bear on this question. Identification ad characterza­ revealed by their isotopic compositions. Measurements of SNC tion of terrestrial analogs fr ALH 8401 carbonates are sorely 0 isotopes reveal the existence of two reservoirs (the atosphere needed. and lithosphere), but there is disagreement about the cause of at­ How Well Can We Recognize Biochemicals, Biominerals, mospherc factionation [18]. However, concentration of heavy H, and Microfossils? Is it pssible to unambiguously distnguish ex­ C, ad N isotopes [19) not affcted by photoecomposition of ozone taterestial biomakers fom terestal contaminants ad artifcts? may favor an escap mechbism. Incorraton of isotopically heavy Most of the ALH 84001 orgaic matter appears to be contamina- H and O into SNC minerals suggests the existence of cycles that 34 Workshop on Martian Meteorites

allow communication between the atmosphere, hydrosphere, and M. H. (1986) Icarus, 68, 187-216; Baker V. R. et al. (1991) Na­ lithosphere, but the way such cycles work is unclear. Fractionation ture, 352, 589-594. [15] Wiinke H. and Dreibus G. (1988) Philos. of S isotopes has also been documented in ALH 84001 and Trans. R. Soc. London, A325, 545-557. [16) Folkner W. M. et al. nakhlites [11), andrecycling of surface S into the interior, perhaps (1997) Science, 278, 1749-1752. [17] Clifford S. M. (1993) JGR, by hydrothermal fluids, has been suggested. Mars Pathfinder data 98, 10973-11016. [18) Karlsson H. R. et al. (1992) Science, 255, confirm earlier results suggesting sulfate is enriched in soils, and 1409-1411; Farquhar J. et al. (1998) Science, 280, 1580-1582. redox state may play a critical role in S fractionation. Isotopic frac­ [19) Wright I. P. et al. (1990) JGR, 95, 14789-14794; Watson L. L. tionations due to inorganic processes must be understood before et al. (1994) Science, 265, 86-90. [20) Nyquist L. E. et al. (1995) these systems can be used as biomarkers. The nature of any mar­ LPS XX.VI, 106-107. [21) Treiman A. H. (1998) Meteoritics & tian organic compounds, as well as the mineralogy of volatile ele­ Planet. Sci., in press. [22] Ash R. D. et al. (1996) Nature, 380, ments, is virtually unknown. Understanding prebiotic chemistry 57-59. [23) Carr M. H. and Wiinke H. (1992) Icarus, 98, 61-71. must await detailed analyses by landers or sample return. Deter­ [24) Longhi J. (1991) Proc. LPSC 19th, 19, 451-464; Longhi J. mining the redox states of Fe and Mn in surface materials is im­ et al. (1992) Mars, 184-208; McSween H. Y. (1994) Meteoritics, portantbecause reduced formsmay serve as a source of energy for 29, 757-779; Lee D. and Halliday A. N. (1997) Nature, 388, 854- organisms. 857. [25] Rieder R. et al. (1997) Science, 278, 1771-1774. What Is the Geologic History of the Ancient Crust? The ancient terrain of Mars, sampled only by one probably nonrepre­ sentative meteorite, provides the geologic context for any poten­ WHAT WERE THE MAJOR FACTORS THAT CON­ tial life. The petrologyof the ancientcrust is likely to be complex, TROLLED MINERALOGICAL SIMILARITIES AND and it is not clear to what degree Mars Pathfinder rocks of andes­ DIFFERENCES OF BASALTIC, LHERZOLITIC, AND itic composition might represent the composition of this crust.The CLINOPYROXENITIC MARTIAN METEORITES WITHIN ages of this crust are unknown because of uncertainty in cratering EACH GROUP? T. Mikouchi1, M. Miyamoto1, and G. A. stratigraphy, but the 4.5-Ga age of ALH 84001 [20) was not ex­ McKay2, 1Mineralogical Institute, University of Tokyo, 7-3-1 pected. Mapping magnetic anomalies by Mars Global Surveyor Hongo, Bunkyo-ku, Tokyo 113-0033, Japan ([email protected]­ might eventually provide a correlationtool. The impact history of tokyo.ac.jp), 2Mail Code SN2, NASA Johnson Space Center, ALH 84001 is complex and the number of impacts it has experi­ Houston TX 77058, USA. enced is disputed [21]. However, a 4-Ga shockage forALH 84001 suggests that Mars also experienced the late heavy bombardment Introduction: Twelve martian meteorites that have been re­ [22). The study of sedimentary and igneous strata is particularly covered so far are classified into five groups (basalt, lherzolite, important forreconst ructing the geologic record. Crustal recycling clinopyroxenite, dunite, and orthopyroxenite) mainly from petrol­ has been discounted [23], but more geologic informationis required ogy and chemistry [e.g., 1). Among them, the dunite and ortho­ to understand lithospheric origin and evolution, including inputs pyroxenite groups consist of only one meteorite each (dunite: from materials cycled through the atmosphere. Some progress has Chassigny, orthopyroxenite: ALH 84001). Thebasalt group is the been made in deciphering the composition and evolution of the largest group and consists of four meteorites (Shergotty, Zagami, crust, mantle, and core through studying SNCs [24] and by Mars EETA 79001, and QUE 94201). The lherzolitic and clinopyrox­ Pathfinder[16,25), but much more needs to be done. enitic groups include three meteorites each (lherzolite: ALH 77005, References: [l) Romanek C. S. et al. (1994) Nature, 372, LEW 88516, and Y 793605; clinopyroxenite: Nakhla, Govemador 655-657; Valley J. W. et al. (1997) Science, 275, 1633-1638; Valadares, and Lafayette). These meteorites within each group are Leshin L. A. et al. (1998) GCA, 62, 3-13. [2] Mittlefehldt D. W. generally similar to the others, but none of them is paired with the (1994) Meteoritics, 29, 214-221; Harvey R. P. and McSween others. In this abstract, we discuss the major factors that controlled H. Y. (1996) Nature, 382, 49-51. [3) McKay D. S. et al. (1996) mineralogical similarities and differences of basaltic, lherzolitic, Science, 273, 924-930. [4] Scott E. R. D. et al. (1997) Nature, and clinopyroxenitic meteorites within each group. Th.is may help 387, 377-379. [SJ Mcsween H. Y. and Harvey R. P. (1998) In­ in understanding their petrogenesis and original locations on Mars tern. Geol. Rev., in press; Warren P. H. (1998) JGR, 103, 16759- in general. 16774. [6) Treiman A. H. (1995) Meteoritics & Planet. Sci., 30, Basalt: Pyroxene and maskelynite (shocked plagioclase glass) 294-302. [7) Gooding J. L. (1992) Icarus, 99, 28-41; Treiman are major minerals in the basaltic martian meteorites. Pigeonite and A. H. et al. (1993) Meteoritics, 28, 86-97. [8) Bada J. L. et al. augite show two distinct textural occurrences of chemical zoning (1998) Science, 279, 362-365; Juli A. J. T. et al. (1998) Science, patterns. In Shergotty and Zagami, pigeonite and augite are usu­ 279, 366-369; Clemett S. J. et al. (1998) Faraday Disc., in press. ally present as separate grains that are zoned from a Mg-rich core [9) Bradley J. T. et al. (1996) GCA, 60, 5149-5155; Steele A. et to an Fe-rich rim, respectively. Both pigeonite and augite usually al. (1997) J. Micros., 189, 2-7; Posfai M. et al. (1998) Science, have homogeneous cores, considered to be cumulus phases.Zagami 280, 880-883; Bradley J. T. et al. (1998) Meteoritics & Planet. pyroxenes are not zoned as extensively as those in Shergotty, but Sci., in press. [10] Nealson K. H. (1997) Science, 276, 1776. their mineralogy is quite similar. On the other hand, pigeonite and [11) Shearer C. K. et al. (1996) GCA, 60, 2921-2926; Greenwood augite in BETA 79001 and QUE 94201 are both present in indi­ J. P. et al. (1997) GCA, 61, 4449-4453. [12) Bradley J. T. et al. vidual composite grains. These pyroxenes are complexly zoned; (1998) Nature, 390, 454; Barrat J. A. et al. (1998) Science, 280, typically the cores are Mg-rich pigeonites, mantledby Mg-rich aug­ 412-414. [13) Yung Y. L. et al. (1988) Icarus, 68, 462-480; ite, and the rims are Fe-rich pigeonite. Maskelynite compositions Dreibus G. and Wiinke H. (1987) Icarus, 71, 225-240; Scambos also correspond to differences in pyroxene zoning. Maskelynites T. A. andJakosky B. M. (1990) JGR, 95, 14779-14787; Mcsween in Shergotty and Zagami are more alkali-rich than those in H. Y. and Harvey R. P. (1993) Science, 259, 1890-1892. [14] Carr EETA 79001 and QUE 94201, suggesting later crystallization. LP/ Contribution No. 956 35

Shergotty and Zagami maskelynites indicate growth fom the sur­ simila, suggesting similar history in general. Augite is firly rounding pyroxene walls in interstitial melts, whereas EET A 7901 homogeneous in composition, dd the core compositions of augite and QUE 94201 normally grew from core to rim. FeO in the ae identical (En39Fs23Wo38) among the three meteorites. They are maskelynite cores is diferent between Shergotty and Zagci (0.5- chemically zoned fom the Mg-rich cores to the Fe-rich rims and 0.6 wt%) dd EETA 79001 and QUE 94201 (0.3-0.4 wt%). The overgrowth of low-Ca pyroxene is observed at some edges. Chemi­ lower Fe content of EETA 79001 and QUE 94201 maskelynite cal zoning of augite in Nakhla and Govemador Valadares shows cores reflects earlier crystallization of plagioclase. Aluminum zon­ more Fe-rich compositions than that of Lafayette. The bulk ing in pyroxenes also marks the bginning of plagioclase crystalli­ composition of overgrowth in Lafayette pyroxene is also more zation. Decrease of the Al/Ti of pyroxenes further suggests Mg-rich and Ca-por than the other two meteorites. Olivines in plagioclase crstallization, dd the Al-Ti distibution shows a clea Nakhla and Govemador Valadaes exhibit chemical zoning in both difference between Shergotty and Zagci on the one hand dd major and minor elements, whereas that in Lafyette is almost EETA 7901 and QUE 94201 on the other. Such mneralogical dif­ homogeneous. Lafayette olivine also difers fom those of Nakhla ferences of EETA 79001 and QUE 94201 fom Shergotty and dd Govemador V aladares because symplectic inclusion composed Zagci can be understood by undercooling of the magmas fom of complex intergrowth of augite and magnetite is absent in which they have crystallized [e.g., 2]. We believe that Shergotty Lafayette olivine. Thus, overall mineralogy of the three meteorites dd Zagami experienced only slight undercoling, resultng initially is simila, but microstctures in pyroxenes and olivines ae slightly in cotectic growth of pigeonite dd augite, later joined by plagio­ distnct. These differences cd be explaned by diferent degees clase. On the other hdd, EET A 7901 dd QUE 94201 exprienced of late-magmatic dd subsolidus difusion [e.g., 4]. That is, Nakhla signifcdt undercooling of their melts, dd their crystallization se­ dd Govemador Valadaes experienced fairly fast subsolidus and quence was pigeonite, augite, plagioclase, and then Fe-rich pigeo­ late-magmatic cooling. As a result, they preserved both chemical nite, with each phase crystallizing metastably alone, because the zoning and symplectte in olivine, and pyroxene became more Fe­ melt compositions did not fllow the equilibrium phase boundaies. rich. On the other hdd, Lafayette exprienced slightly slower late Lherzolite: Pigeonite, olivine, augite, maskelynite, and magmatic cooling, which erased symplectites and rehomogenized chromite ae major phases in the three lherzolitic matid meteor­ olivine. ites. Texturally, the three lherzolites ae very similar dd ae cha­ Discussion and Summar: It is clea thc each group hb gen­ acterized by the presence of two distinctive textural patters: erally simila mineralogy within each group, but is distinct fom poikilitic dd nonpoikilitic. In the pikilitic area, a lage pigeonite the other groups. Such a similarity within the same group is espe­ oiocryst encloses smaller rounded olivine dd euhedral chromite cially remarkable fr lherzolitic dd clinopyroxenitic groups. This grains. The pigeonite oiocryst hb augite bdds, usually along its is also supported by both crystallization and expsure age data [e.g., rim. The pigeonite oikocrst is wealy zoned fom core to rim 5,6]. Unlike them, the basaltic group shows fairly diverse mineral­ (En77Fs19Wo4 - E15Fs20Wo15). Pyroxene in the non-poikilitic area ogy within the group, although they generally shae simila crys­ is mainly pigeonite that is more Ca-Fe-rich (En65Fs28Wo7 - tblization and expsure ages [e.g., 6,7]. Such a diversity can be En58Fs26Wo16) thd that in the poikilitic aea. Olivine is brown to understood by different degrees of undercooling of the magma, as yellowish in both textures. Yamato 793605 olivines in the non­ explaned above. Although it is unlikely that all fur basaltic rocks poikilitic aea ae more Fe-rich (average: Fa34) than those in the shae a common paent magma, their compsitons were probably poikilitic area (average: Fa31), and the Fa content shows bimodal simila. We believe that these basaltic rocks formed nea the sur­ distibution, which is also observed in the LEW 88516 olivines face of Mas, and this might have caused lage diferences in tem­ (Fa25-36). Alld Hills 77005 olivines (Fa23_30) are clealy more Mg­ perature gradient during cooling even though the burial depths were rich than those of LEW 88516 and Y 793605, and mineral com­ not so diferent. That is why the basaltc group shows farly diverse positions are almost identical between the two diferent textures. mineralogy. The lherzolitic and clinopyroxenitic group would be Olivine usually contans magmc inclusions (up to a few microme­ deeper in origin than basaltic rocks, at least in the early stages of ters in size) consisting of Al-Ti-rich pyroxene, Si-rich glbs, and their crystallizaton, dd thus undercooling efects of magmas were chromite. Maskelynites ae rae in the poikilitic area but are abun­ probably minor. The slight mineralogical difference within each ddt in the nonpoikilitic aea. Maskelynites in both textures have group is most evident in olivine compositons. This is probably be­ nearly identical compositions (An55Ab440r 1 - An45Ab520r3). Mi­ cause atomic difusion rates in olivine ae much faster than those nor element chemical zoning of maskelynite is unique. Iron oxide in pyroxene dd plagioclbe. Because most olivines in these groups in the core is about 0.5 wt%. It frst drops down to 0.2 wt% dd still preserve chemical zoning, they originated fom nea the sur­ again slightly increases to 0.4 wt% at the edge, with some decrease face, at least in the late stages of their crystallization history. Py­ and increase. Chromites show chemical zoning toward the roxene preseres the initial compsition at the core. Although the ulvospinel-rich rim. Thus, the mineralogy of all three lherzolitic paent-magma compsitions of these groups were distinct fom each martid meteorites is very similar in contrbt to the basaltic ma­ other, the sequence of crystallization location - that is, fom deep tid meteorites, which show a variety of minerbogical differences. to shallow - would be simila. The basaltic group stated crystal­ This suggests that bl three lherzolitic meteorites generally fllowed lization in shallow locations fom the beginning, dd undercool­ similar crystallization dd thermal histories, although the degree of ing of magma was a major factor that controlled their overall late-stage dnealing or cooling rates was different and chdged oli­ crystallization history. In contast, lherzolitic dd clinopyroxenitic vine compositions in different degrees [e.g., 3]. meteorites originally started crystallization at deep positions (and Clinopyroxenite: Nakhla, Governador Valadares, and thus nealy fee fom the undercooling effects of early-stage mag­ Lafayette manly consist of 80-90% cumulus augite dd 5-10% mas). However, after they moved to nea-surface locations at their olivine in plagioclase-rich mesostases, and their texture is last stages, late-stage cooling rates or thermal dnealing affected remarkably simila to one another. Mineral compositions ae also them a lot. Although such efects would also occur in the basaltic 36 Workhop on Martian Meteorites

meteorites, the infuence wb not so major in the present phases the S-fee rock may be classifed by its SiO2 and alkali content as because they do not contain olivine. an andesite. The S-fee rock has the characteristic low Alp3 bd Acknowledgments: This study was partly supported by high FeO content expected fom martib igneous rocks [3]. Moritbi Science Fundation. According to [3], the compositions of the Pathfnder rocks bd References: [1] Mcsween H. Y. Jr. (1994) Meteoritics, 29, the S-fee rock fall along a tholeiitic tend of bdesite frmation,

757-779. [2] Mikouchi T. et al. (1998) Meteoritics & Planet. Sci., which is chaacterized by strong FeO enrichment without SiO2 en­ 33, 181-189. [3] Harvey R. P. et al. (1993) GCA, 57, 4769- rchment until late in the crystallization sequence. Eventual Fe-Ti GCA, 56, 4783. [4] Havey R. P. bd McSween H. Y. Jr. (1992) oxide crystallization pushes tholeiitic liquids to hgher SiO2 and 1655-1663. [5] Nakamura N. et al. (1982) GCA, 46, 1555-1573. lower FeO contents, thus creating andesitic bd rhyolitic liquids [4]. [6] Eugster 0. et al. (1997) GCA, 61, 2749-2757. [7] Nyquist L. E. Terrestial andesites fred via the tholeiitic diferentiation trend et al. (1995) LPS XXVI, 1065-1066. are associated with the low-pressure crystallization of basaltic pa­ ent liquids, b occurs in Iceland [5] or the Galapagos Spreading Center [6]. Thus, it could be agued that the Pathfinder rocks de­ GENESIS OF TH MARS PATIIDER "SULFUR-FREE" veloped though low-pressure factional crystallization of a tholei­ ROCK FROM SNC PARNTAL LIQUIS. M. E. Minitti bd itic basaltic parent. M. J. Rutherfrd, Department of Geological Sciences, Brown To determine the nature of the basaltic parent liquid bd the evo­ University, Providence RI 02912, USA. lutionary path it fllowed to produce the S-fee rock, McSween et al. [3] attempted to reproduce the composition of the S-fee rock Introduction: In July 1997, the Mars Pathfnder mission via two sepaate pathways. First, they [3] calculated a liquid line touched down in the Ares Valles region of Mas to begin its scien­ of descent fr a basaltic SNC intercumulus melt in Shergotty [7] tific study of the Red Planet. The Sojouer rover used its onboard (Table 1), using the MELTS progra [8]. The liquid line of de­ caeras and a-proton X-ray spctometer (APXS) to study the soils scent produced by the MEL TS calculation illustrates that the and rocks of the landing site. Curently, preliminary balyses of fve Shergotty stating compsiton dos not lead to the S-fee rock com­ rock samples bd six soil samples have been completed. They are position. Therefre, it was concluded that SNC liquids like the based only on the spectra obtained fom the X-ray portion of the Shergotty intercumulus melt could not produce the balyzed rocks APXS [1]. When plotted on oxide vs. S diagrams, the Pathfnder of the Pathfinder landing site under dry, low-pressure conditions. soil and rock data fall on a line extending between the soil data, Second, McSween et al. [3] applied the liquid line of descent rel­ which fall at high S contents, and the rock data, which fall at lower evbt to development of the Galapagos Spreading Center andesitic S contents. Because the solubility of S in bbaltic melts is :0.2 wt% liquids [6] to the S-fee rock composition. According to [3], the [2], the range of S contents in the rocks implies that each rock experimentally determined Galapagos liquid line of descent, which analysis has a difering amount of soil component. The rock com­ taces the evolution of a dr, basaltic paent liquid during factional positions, therefre, fall along a mixing line between the soils and crystallization, successfully reproduces many of the compositional a hypothetical "S-fee" rock obtained by extrapolating the regres­ chaacteristics of S-fee rock. Therefre, it was concluded that the sion line through the soil and rock data back to O wt% S [1]. The conditions associated with the development of the Galapagos liq­ composition of the S-fee rock is contained in Table 1. Accepting uid line of descent are relevant to the frmation of the S-fee rock, the argument that the Pathfnder rocks are igneous, as done by (3), most notably the dry nature of the basaltic starting material.

TABLE l. Relevant melt and rockcompositions.

Oxide Shergotty1 A*2 P0O.82N23 MS-14 S-freeRock 5

SiO2 50.8 50.33 51.35 60.78 62.0 ± 2.7 Alp3 8.0 8.16 13.21 12.27 10.6 ± 0.7 FeO* 19.8 19.87 13.43 10.82 12.0 ± 1.3 MgO 7.7 7.39 6.12 2.74 2.0 ± 0.7 CaO 9.7 8.95 10.74 6.82 7.3 ± I.I Na,O 1.5 1.71 2.30 3.15 2.6 ± 1.5 �o 0.2 0.43 0.10 0.91 0.7 ± 0.2 TiO2 1.0 1.75 2.14 1.31 0.7 ± 0.1 P20s 0.9 0.50 0.18 0.73 na MnO 0.5 0.52 na 0.43 na na = analysis not available. 1 Shergotty intercumulus melt composition [7]. 2 Experimental starting material based on calculated Chassigny parent magma [10]; com­ positionof fused glass was determined via electron microprobe analyses (see text). 3 Composition of experimental starting material used by [ 6]. 4 Preliminary experimental glass composition from run at P = 200 bar, P, ,. = P o• T = 0 1 11, 1050°C; composition normalized to 100%. 5 Sulfur-free rock composition of [1], normalized to 98%. LPI Contribution No. 956 37

This inital work raises two interesting questions. The paent The compositions of the residual liquids fom the suite of water­ basalt fr the Galapagos liquid line of descent is a high-Al2O3, low­ saturated, 20-ba expriments defne a liquid line of descent that FeO basalt unlike that expected fom a low-Al2O3, high-FeO ma­ approaches the S-fee rock composition. A sample composition of tic mantle proposed by [9]. Thus, while the Galapagos liquid line one of the experimental residual liquids is contained in Table 1. of descent leads to liquids resembling the S-fee rock compsition, Thus, the hydrous experiments demonstate that the addition of the composition of the starting basalt is not in keeping with the ba­ small amounts of water (1.0-1.5 wt%) to a crystallizing SNC saltic liquids calculated fr kown martic basalts [e.g., 10]. Sec­ magma yields residual liquids that resemble the S-fee rock com­ ond, the liquid line of descent of the tested SNC melt was calculated position. While the exact composition of the S-fee rock has not with MELTS cd, by nature of the progrb, was thereby a dry liq­ been experimentally obtained, we expect that variations in P , 820 uid line of descent. Water, though, has played a prominent role in P , f , and bulk composition will lead to a residual liquid with 101a1 02 the development of surface fatures on Mars [11], including the the S-fee rock composition. Thus, reproduction of the S-fee rock Mas Pathfnder lcding site. Te presence of water in maran mag­ fom a SNC parent magma does appea possible via near-surface mas is indicated by kaersutitic hosblende fund within melt in­ crystal-factionation processes. clusions in fur SNC meteorites cd a biotite fund in a Chasigny References: [l] Rieder R. et al. (1997) Science, 278, 1771. melt inclusion [10]. Thus, a dry martic paent magma may not be [2) Caroll M. R. and Webster J. D. (1994) Rev. Mineral., 30, 232. approprate to the development of the S-fee rock composition. [3] McSween H. Y. Jr. et al. (1998) LPS XXIX, Abstrbt #1054. Water has a kown and signifcct egect on phae equilibria of crys­ [4] Hess P. C. (1989) Origin of Igneous Rocks, Harvad Univ., tallizing melts, as illustated by the diferences in the liquid lines Cambridge, 336 pp. [5] Carmichael I. S. E. (1964) J. Petrol., 5, of descent of hydrous and anhydrous experiments on a Galapagos 435. [6] Juster T. C. (1989) JGR, 94, 9251. [7] Hale V. P. S. (1997) Spreading Center basalt [12]. Thus, its presence in a SNC melt may LPS XXVI, 405. [8] Ghiorso M. S. and Kelemen P. B. (1987) make it possible to derive the S-fee rock fom a martian basaltic Geochem. Soc. Spec. Pub., 1, 319. [9] Wanke H. and Dreibus G. parent melt. (1988) Philos. Trans. R. Soc. London, A545. [10] Johnson M. C. Experimental: To investigate if the addition of water to a ba­ et al. (1991) GCA, 55, 349. [11] Car M. H. (1981) The Surace of saltic SNC melt could produce a liquid line of descent leading to Mars, Yale Univ., New Haven, 226 pp. [12] Spulber S. D. cd Ru­ the S-fee rock composition, we conducted expriments on a SNC therford M. J. (1983) J. Petrol., 24, 1. [13] Nielsen C. H. and basaltic paent magma, A*, whose composiuon was determined fr Sigurdsson H. (1974) Am. Mineral., 66, 547. Chassigny [10]. This starting composition has the characteristic high-FeO and low-Al2O3 concentations associated with maan ig­ neous rocks (Table 1). A* closely resembles other exprimentally cd theoretically determined SNC paent-magma compositions, in­ RCOGNIZING LIE AN ITS EVOLUTION fROUGH cluding the stang SNC magma composition used by [3] in their BIOMARKER. J.M. Moldowan, Department of Geological MELTS calculations (Table 1). For each experimental run, an ox­ and Environmental Sciences, Stanfrd University, Stanfrd CA ide mix with the A* composition was placed, with sufcient water 94305-2115, USA ([email protected]). to saturate the melt, in a sealed AgPd tube. This sbple tube was then sealed in a lager AgPd tube with water and a QFM bufer Biomakers ae molecula fssils fund in sedimenta roks cd assembly. All expriments were buffred at QFM, as tis is thef02 petoleum. Like fssils, they ae recognized a remncts of enzyme­ believed to be most representative of the conditions on Mas un­ mediated biosynthesis dictated by geneuc code. Tey cannot be con­ der which the SNC meteorite magma crystallized [10]. Experi­ fsed with abiogenic molecules, because they are produced in ments were rn at 20 bars in 1M pressure vessels using a mixture abundances that fa exceed their relative chemical stailities. of Ar cd CH4 gases to provide the pressure. Run temprbures were On a broad scale, evolution can be fllowed in the biomaker between 1089°C and 975°C.Experiments were water quenched at record as chemically more complex molecules ae biosynthesized pressure, cd befre thin sections were made of the experimental by more advanced taa. Numerous corelations have been estab­ products, a grain mount of the bufer wa made to ensure that the lished between various taxa and biomakers. Kingdoms of organ­ experiment maintained QFM. In additon to the hydrous, 200-bar isms cc be diferentiated by hopanes fr prokayotes, extended experiments, 1-atm experiments bugered at QFM were also con­ head-to-head isoprenoids fr archea, sterces fr protista, and vari­ ducted on A* in order to experimentally investigate the evolution ous terpces fr plctae. At lower taonomic levels, relationships of the * magma under dry conditions. The 1-at expriments were include dinofagellates with dinosteroids, porifera (sponges) with run in Fe3O4-saturated AugPd10 tubes in evacuated SiO2-glass 24-isopropylcholestane, and cgiosperms (fowering plants) with tubes. oleanane. At the species level, examples include Botrococcus Major-oxide calyses of the experimental products were con­ braunii and botrococcane and Gloeocapsomorha prisca and a

ducted on the Brown University Caeca Cameba electon micro­ specifc assemblage on odd C-numbered n-alkanes below C 19. Such probe. Analysis conditions included a 15-kV accelerating voltage, taon spcifc biomarkers can be used to tace the origins of lif a 10-nA defocused beam fr glasses, and a 15-nA fcused beam cd its evolutionary lineages. fr oxide cd silicate anayses. During glass calyses, the electron This molecula infration supports fssil records cd provides beam was blanked between analyses and a Na loss progrb [13] important evidence fr lif frms where catomical remains are

was utilized to obtain Na2O concentrations. ambiguous or absent. Recent examples include (1) the prediction Preliminary Results: In the hydrous, 200-bar expriments, of c extended record of Vendic porifra befre Vendian sponge crystallized phases include pigeonite cd c intermediate-Ca py­ spicules were discovered, and (2) identifcation of ancestal dino­ roxene. Magnetite is also fund in experiments rn below 1075°C. flagellates among Lower Cambric acritachs having partial dino- 38 Workhop on Martian Meteorites

fagellate morphology and dinoflagellate biomakers, extending the widely accepted fssil record (Middle Triassic) by -280 m.y. Biomaker asemblages, like microfssils, provide a window into the depositional environment of sedimentary rocks. Factors such as redox potentia, acidity, salinity, and theral maturity infuence biomaker fngerprints; therefore, biomakers can be used to deter­ mine these specific chaacteristics of the depositional environment.

POSSIBLE MICROFOSSILS (WARRA WOONA GROUP, TOWERS FORATION, AUSTRALIA, -3.33.5 Ga). P. A. Morris1, S. J. Wentworth2, C. C. Allen2, and D. S. McKay3, 1University of Houston-Downtown, 1 Main Steet, Houston TX 77002, USA ([email protected]), 2Lockheed Martin, Houston TX 77258, USA, 3NASA Johnson Space Center, Houston TX 77058, USA. Fig. 1. Scanning-electon microscope imge of X-ray element maps for Introduction: Ealy in the twentieth century there were reprts Fe and silica. Arrows indicate some of te biogenic spheres. of Archec stomatolite-like structures tat were similar to orgaic rich stomatolites fom the base of the Cbbrian (60 m.y.) [1,2). It was not until the latter half of this century that fssilized Archec­ age (3.9-2.5 Ga) life frms were found in the Fig Tree Formation of South Afica and the Towers Formauon of Austia [3-7). Some of the acient stromatolites contained streaks ad clots of kerogen, pyrite grains, remnants of microbial cells, and filbents that rep­ resented vaous stages of preservation, while others appeaed to lack fossils [8,9]. A set of physical criteria was established for evaluating the biogenicity of these Archean discoveries [10]: (1) rocks of unquestionable Archec age; (2) microfssils indig­ enous to Archean sediments; ad (3) microfssils occurring in clats that ae syngenetic with depsition of the sedimentary unit. In the case of bedded cherts, the fssils should predate the cherts; (4) the microfssils ae biogenic; and (5) replicate sampling of the fssil­ ifrous outcrop frmly demonstates the provencce of these micro­ fssils. Sbple 002 fom the Pecbbrian Paleobiology Reseach Group (PPRG) was examined. This stromatolitic carbonaceous chert con­ tans microbial remains that meet the established criteria [10]. Us­ Fig. 2. X-ray element map fr Fe. Iron is fund in assoiation with biogenic spheres indicated in Fig. 1. ing a sccning electon microscope (SEM), we have calyzed the morphologies and chemisty of these possible microbial remains [7,10). Methods: Freshly factured chips and petrographic thin sec­ tions were analyzed with an Oxfrd ISIS energy-dispersive spec­ trometer (EDS) cd a Phillips SEM XL 40 FEG (SEM). Discussion: Sccning-electron microscope and EDS analysis of sbple 02 indicates the presence of hematite, quartz, barite, ca­ bonate minerals, including Fe cabonate (siderite), and kerogen [11, 12]. Te kerogens are organic derivatives. Sample 002 includes several morphologic fatures clealy in­ digenous to the sample that may have been produced by biologic activity. These fatures include several sizes of spheres, rods the size of bacteria, ellipsoids, flaments, and smooth blanketing ma­ teria resembling present-day bioflms. Each of these features ae rich in both Fe and C compaed to the surounding quartz matx. X-ray element maps of two of the lagest spheres (16-27 µm, 5- 12 µm) indicate that the Fe cd C ae localized within the spheres, but ae nearly absent fom the surounding quartz matrix (Figs. 1- 3). Iron is known to be important in microbial preservation [13]. The spheres are intimately intergrown with the quartz matx ad Fig. 3. X-ray element map of Si. Toe areas occupied by the biogenic ae clearly indigenous to the sample (Fig. 1). They display a rough, spheres lack this element. LP/ Contribution No. 956 39

Cabridge Univ., 1348 pp. [11) Schopf J. W. (1983) Earth's Ear­ liest Biosphere, Princeton Univ., 543 pp. [12) Moris et. al. (1998) LPS XXIX. [13) Ferris G. et. al. (1988) Geology, 16, 149-152. [14) Liebig K. et. al. (1996) N. Jb. Geo[. Palaeont. Abh, 199, 269- 293.

PERSPECTIVES ON TH FUTURE SEARCH FOR LIFE ON MARS AND BEYOND. K. H. Nealson, Mail Stop 183-301, Jet Propulsion Laboratory, 4800 Oak Grove Drive, Pasadena CA 91109, USA ([email protected]).

One cb view the search fr lif on Mas in two ways: frst, a the initial step in the search fr life elsewhere, bd second, as the one place where in situ methods fr life detection can be tested and proved via saple ret. Afer Mas, most of the lif detection will 0 2 4 6 be done via in situ studies with data retu. Mars ogers us the op­ pority to fine tune our methods - perhaps fr a long time to Fig. 4. Energy-dispersive spectometer analysis of biogenic spheres at low kV. Nouce the relatve abundance of C, 0, and Fe (Fel) in relation come. Our group is involved in the development of methods fr to Si. Te specimen was coated wit Au and Pd for 30 s. life detection that are independent of specifc signals used fr de­ tection of lif on Earth. These approaches include general indica­ tors of metabolic actvity bd orgaismal structure bd composition. Using such approaches, we hope to detect the signals of life uneven texture. The chemical compsition of these spheres is com­ (biosignatures) that ae indepndent of preconceived notions and patible with that of siderite. Siderite can be an indicator of yet ae convincing and unabiguous. Te approaches we are f­ microbially induced precipitation [14). These spheres ae within the cusing on include stable isotopic balyses of metals, mineral fr­ size rbge and appa to be identical to those accepted as biogenic mation and disolution, bd elemental balysis. Tese methods allow by Schopf [6,7,10). us to examine samples at a variety of scales, looking fr nonequi­ Energy-disprsive spectometer balysis of smaller spheres that librium distibution of elements that serve as biosignatures. For fu­ are within the size range of bacterial cocci, as well a flaments and tures studies of Mas and beyond, they, or some variation of them, the smooth, blanketing morphology reveal a simila composition should allow infrence or proof of lif in non-Earth locations. - that is, the presence of signifcbt levels of C and Fe bd the absence of significant levels of Si. Compaison with the suround­ ing matx indicates that although C may occasionally be present OXYGEN-ISOTOPIC COMPOSITION OF NAKBLA in the frm of relict Ca carbonbe minerals, and Fe may occur as SIDERITE: IMPLICATIONS FOR MARTIAN VOLA­ irregula hematite grains, the matx itself is low in both Fe bd C. TIES. J. M. Saxton, I. C. Lyon, bd G. Tuer, Department of We interret all of the C bd Fe-rich features shown in Figs. 1-3 Earth Sciences, University of Manchester, Manchester M13 9PL, as biogenic remains of bacterial colonies, individual cells, and UK [email protected]). bioflms that have been mineralized by replacement with Fe ca­ bonate bd possibly other Fe-bearing minerals. An understbding We have previously [1] reported the O-isotopic measurements of the complex processes by which microbes and their products ae of two grains (Cl bd C3) of Mn-rich siderite in Nakhla. Four O­ mineralized bd presered will be important in analyzing additonal isotopic meaures were identcal within error, yielding meb «5180 = Mars meteorites and reted saples fom Mars. Tis provides 34 ± l%o SMOW. This is the heaviest cabonate yet reported fom one example showing that likely fssilized microbial features may a mab meteorite. We now explore the implicbions of these data retain their distinctve morphologic and chemical identity over sev­ fr martib volatiles. More recently, we have meaured more car­ eral billion years of geologic history. bonate grains (C8, C9, ClO) that difer chemically fom Cl and Acknowledgments: We wish to thank L. Hulse for mby C3, being richer in Mg bd lower in Mn. These new grains still hours of help on the SEM and E. K. Gibson for obtaining the have «51so near 34%0 (the accuracy of each measurement is 1-2%0). sample fom the PPRG. Tis investgation was supported by NASA The carbonate grains ae orbge-brown in transmitted light, sug­ Grbts NAG9-980 bd NAG9-867. gesting some staining by fric oxides; C8 was paticulaly badly References: [1] Rothplez A. (1916) Abh. Bayer. Akad. stained, and its lower «5180 probably refects contamination by ox­ Wiss., 28(4), 1-91. [2] Walcott D. (1912) Geol. Surv. Canada ides bd/or silicates. Mem, 28, 16-23. [3] Baghoom E. S. et al. (1966) Science, 152, Various authors have considered the isotopic composition of 758-763. [4] Barghoom E. S. (1971) Sci. Amer., 224, 30-42. martian volatiles. Clayton bd Mayeda [2] considered production [5] Prashnowsky A. A. et al. (1972) in Advances in Organic of a hydrosphere bd atmosphere by high-temperature outgassing Geochemistry, pp. 683-698, Pergaon, NewYork. [6] Awra of CO2 bd H20, fllowed by cooling to low temperature. Tis pro­ S. M. et al. (1983) Precambrian Res., 20, 357-374. [7] Schopf et cess cb also be considered on a local scale [e.g., 3,4). An altera­ al. (1987) Science, 237, 70-72. [8J Schopf J. W. (1977) Precam­ tive view of martian volatiles was taken by Hutchins and Jakosky brian Res., 5, 143-173. [9J Gutstadt A. A. et al. (1969) Nature, [5], who suggested the hydrosphere/atmosphere might have been ·233, 165-167. [10) Schopf et al. (1992) Proterozoic Biosphere, enriched in heavy isotopes; this would render models with high- 40 Workhop on Martian Meteorites

TABLE 1. outgassing and subsequently unmodifed. Foration at higher tem­ peratures would require heavy isotop enrichment of the hydro­ 6IBQ sphere/atmosphere such as that implicit in the model of Hutchins Sd Rh Mg Ca %oSMOW and Jakosky [S]. We note also that isotopic exchange with the sili­ Cl 3 70 26 +32.4 cate would imply higher values ofH cd C. Cl 3 70 26 +34.8 Bridges and Grady [6] have suggested that the Nakhla siderite Cl 3 70 26 +33.9 frmed fom a melt, derived by melting of an evaprite. A prob­ C3 10 81 8 +34.9 lem with this model is that the present siderite is not in isotopic equilibrium with the igneous assemblage. In principle, the cabon­ cs 27 5 66 2 +28.0 ate textures could be igneous, cd the high 6180 value reflects sub­ C9 38 55 +31.6 5 sequent (low-T) equilibration with c aqueous fuid. This would CIO 33 62 2 +35.6 3 raise the pssibility of isotopic exchcge btween the fuid cd other phbes, such as the fldspa. Since, in fact, the feldspar does not appa to have undergone isotopic exchange (L. A. Leshin, per­ sona communication) it may be more appropriate (and also sim­ temperature fuid-silicate equilibrium inappropriate. It is also pos­ pler) to consider the carbonate as secondary cd fuid-rock isotopic sible that, by the time the Nakhla cabonate fred, the 6180 of exchange as unimportct in Nakhla. We note also that Bridges et the hydropshere could have been reduced through fxation of CO2 al. [8] also fund no evidence for isotopic alteration of Lafayette ° 18 as cabonate. Assuming fxation as CaCO3 at 0 C, we estimate 6 0 fldspar. may be reduced -4-6%0 by this process. The siderite 6180 also places a lower limit on 6180 of present­ In Fig. 1 we relate the siderite 6180 cd frmation temperature day CO2 of abou·t 40%0 ( or about 36%0 if the siderite equilibrated ° to potential matic H2O/CO2 reservoirs. In the left-hand panel, we with water (brine) at -20 C). After maing a simple alowance fr 18 show 6 0 of H2O/CO2 produced by inital outgassing [after 2]; X low-temperature H2O vap,-CO2 factionation, this is consistent with is the mole faction of O b water in the degassed reseroir. Clayton terrestrial infaed spectoscopy [8]. and Mayeda [2] suggested X - 0.8 since it could account for their References: [1] Saxton J. M. et al. (1997) Meteoritics & measurements of the cabonate in EET 79001. In the right- hcd Planet. Sci., 32, Al 13. [2] Clayton R. N. cd Mayeda T. K. (1988) pcel, we calculate the isotopic composition of martic surfce res­ GCA, 52, 925. [3] Saxton J. M. et al. (1998) EPSL, in press. ervoirs capable of generating the Nala cabonate. We assume that [4] Leshin L. et a. (1998) GCA, 62, 3. [S] Hutchins K. S. cd a hydrosphere/atmosphere exists at 0°C, fom which is taken a Jakosky B. M. (1997) GRL, 24, 819. [6] Bridges J. C. cd Grady sample of fuid having mole faction Xiocal O as water. This sample M. M. (1998) LPS X. [7] Bridges et a. (1997) Meteoritics & of fuid is then heated to temperature T, at which temperature sid­ Planet. Sci., 32, A21. [8] Bjoraker G.L. et a. (1998) Bull. AAS, erite frms in equilibrium with it. No isotope exchcge takes place 21, 990. between the fuid cd the silicate minerals. The fgure shows the values ofH cd C required to obtain siderite having 6180 = +34%0. Compaing the two pcels of the fgure, we see that at low tem­ RECONNAISSANCE SAMPLING OF AIRBORNE MO­ peratures, H and C lie within the range of vaues that may be pro­ LECULAR ORGANIC CONTAMINATION IN THE duced in initial outgassing, provided X0 is sufficiently high. METEORITE CURATION FACILITY OF JOHNSON Water-siderite equilibration at O < T < 60°C requires Xo > 0.74. SPACE CENTER. E. A. Schilling1 and M. N. Schneider2, Therefre, our data do not require a hydrosphere/atosphere highly 1Lockheed Matin, Houston TX 77058, USA, curent address: St. enriched in 180, which might have been expected by analogy with Olaf College, Northfeld MN 55057, USA ([email protected]), atmospheric H, C, cd N. If the siderite frmed fom a water rich 2Lockheed Martin, Houston TX 77058, USA, current address: ° fuid (Xioal ➔ l) at tempratures >60 C, the infrred values of Ii Stanford University, Palo Alto CA 94305, USA (schneimn@ and C ae too high to be consistent with fuids produced by initial stanfrd.edu).

Introduction: Deterg the extent of contamination of me­ 6110 of fluid required lo form Nakhla siderile teorites by terestral orgcic compounds has become a question of critical imprtance in the last several yeas. Contamination is­ sues have been considered in the past cd recognized as important, resulting in a good deal of study [1]. However, more thoroughly understcding organic contamination issues is espcially pressing when considering possible evidence of previous lif on Mas and in future Mars sbple-retu missions. "Orgcic contamination," often vague terminology, can be de­ fned in the fllowing way, adapted fom the NASA-appointed Or­ -40 ulgassing -so��� ��,�-�-��-� gcic Contamination Review Comittee [l]. Type 1 orgcics ae 0.0 0.5 1.0 0 100 200 300 400 500 800 the building-block materials of lif. These ae compounds that have X T 'C c essential role in terrestal biochemist, such as amino acids cd metalic byprpucts. Typ 2 organics ae cosmohemical cd bio­ Fig.1. geochemical environmental compounds, those compounds that do LPI Contribution No. 956 41

detecting levels of molecular organic contamination (types 1 ad 2) in the Meteorite Proessing Laborator (MPL) at Johnson Space Center. Sampling Technique: The general room atmosphere of the MPL ad the psitive pressure, high-purity N atmosphere of the Mas meteorite cabinet were sbpled fr trace amounts of arbose molecula contaminats (types 1 ad 2). A stanless-steel tube flled with activated chacoa ad other ptprieta sorbents (Balas Ana­ lytical Laboratories [2]) was attached to a ar pump, which drew air through the tube at a rate of 10 ml/min. Te device was placed in the center of the MPL and air was sampled fr 6 hr while typi­ cal activities took place in the room. Another sampling device was btached to a outlet port of the Mas meteorite cabinet, home of ALH 8401, and allowed to run for 6 hr at a rate of 130 ml/min (Fig. I). Because meteorites are stored and processed in high­ purity N cabinets, ad ofen hadled outside cabinets on a lamina fow bench, these two meaurements are representative of the typs and amounts of organic contaminants to which meteorites ae of­ ten exposed. The two sample tubes, along with a unopened ship­ ping blank, were reted to Balas Analytical Laboratories, where they were aalyzed by Theral Desorpton-Ga Chromatography­ Mass Spectomety. The estimated sensitivity of this method was Fig. I. Airborne Molecular Contamination sampling device attached to 0.1 ng/L, but most compounds were not quantifed below I ng/L. outlet port of Mars meteorite cabinet in JSC' s Meteorite Processing Lab. Reults: Measurale levels of arbose molecula organic con­ taination were detected both in the MPL room center ad in the Mas meteorite cabinet. The compounds detected in highest con­ not have an essential role in biochemistry, such as hydrocabons, centations were silicones [cyclic siloxaes: cyclo(Me2SiO)n1 ad polycyclic aomatic hydrocabons (PAHs) ad nonbiogenic amino low- to medium-boiling hydroaons (Table 1). MPL arbose con­ acids. Typ 3 organics, macromolecula biomolecules, ae polymers tainants included butaol, benzene, toluene, ethylbenzene, xy­ of type 1 compounds such as DNA, proteins, ad lipids. Finaly, lenes, siloxanes, texanol isobutyrate, fuorocarbons, isopropaol, Type 4 orgacs ae viable microorganisms such as bacteria ad propoxypropaol, 2-butoxyethanol, butoxypropaol, dipropylene gngi. Type 1 ad type 2 organic contbinats ae often consid­ glycol, methoxypropoxypropaol, 2-ethylhexanol, nonanal, diethyl ered together ad called molecular contaminaton. phthalate, dibutyl phthalate, and dioctyl phthalate. Mars meteorite While may particulate ad inorganic contanation issues have cabinet containants included toluene, siloxaes, isopropanol, 2- been addressed in the past, curators at Johnson Space Center have ethylhexanol, ad caprolbt. The shipping blak contaned mea­ initiated eforts to determine the baseline levels of organic conta­ surable levels of certan orgaic compunds but at levels below nation in curaton facilities. Described here is a portion of the pre­ those meaured in the MPL ad Mas meteorite cabinet. Although liminary work done on developing and utilizing techniques fr cabinet levels were generally lower tha those of the MPL room

TABLE l. Selected organic compounds measured in an unopened shipping blank, MPL roomcenter, and Mars meteorite cabinet.

Meteorite Lab Mars Cabinet Shipping Blank Room Center Outlet Contaminant (ng/L) (ng/L) (ng/L) Hydrocarbons (C6-Cl0) 8 32 10 Hydrocarbons (Cl 1-Cl 8) l 43 3 Toluene 2 4 3 Isopropanol

cyclo(Me2SiO\ <0.1 2.9 5.8 cyclo(Me2SiO)4 <0.1 2.6 3.0 cyclo(Me2SiO)5 <0.1 26 1.0 cyclo(Me2SiO).

center, several siloxbes, isopropbol, and caprolactam were de­ Acknowledgents: We would like to thank E. Stansbery (As­ tected at higher levels in the Mas meteorite cainet thb in the ML sistant Chief, Planetary Science), J. Allton (Lockheed Martin), and rom center. D. Blanchard (Chief, Earth Science bd Solar System Exploration Discuson: The detection of measurale aounts of molecula Division) fr their guidance and support. orgbic contamination in the MPL and in the Mas meteorite cabi­ References: [1] Bada J. L. et al. (1997) Letter fom Organic net raises a number of concess, such as the potential sources of Contamination Review Committee to Dr. Joseph M. Boyce, NASA the contaminants bd the effect of the contaminbts on the integ­ HQ, 2/5/97. [2] Camenzind M. J. et al. (1995) Micro. [3] Texas rity of meteorite samples. The potential sources fr contaminaton, Natural Resource Conservation Commission (1997) AS-134. as with most indoor air, are numerous. Hydroaons, benzene, tolu­ [4] Ahmed S. A. et al. (1982) Agric. Biol. Chem., 47, 1149-1150. ene, and xylenes are commonly fund as industrial air pollutants bd usually enter through air-hbdling systems. However, meaured levels of these compounds are several times lower than extesal BIOGENIC OR ABIOGENIC ORIGIN OF CARBONATE­ Houston air, which contains benzene, toluene, and xylenes at lev­ MAGNETITE-SULFIDE ASSEMBLAGES IN MARTIAN els between 5 and 10 ng/L (3). Compounds such as butanol may MTEORITE ALLAN HLS 8401. E. R. D. Scott, Hawai'i also be produced as human metabolic byproducts. Fuorocarbons Institute of Geophysics and Plbetology, School of Ocean and Earth within the room air probably result fom residual Freon 113 once Science bd Technology, University of Hawai'i at Mboa, Honolulu used to cleb processing tools or fom refigerbts used in labora­ HI 96822, USA ([email protected]). tor feezers. Compounds such as isopropbol, butanol, propxy­ propbol, 2-butoxyethbol, butoxypropanol, dipropylene glycol, and Introduction: McKay et al. [1,2] suggested that the cabon­ methoxypropxypropbol ae commonly fund in solvents bd may ates and submicrometer grains of manetite, pyrhotite, bd an Fe­ be the result of outgassing of room materials such as fooring, S phae identifed as "probaly griegite" were all biogenic in origin. paints, markers, and wet wipes. Plasticizers such as texanol Their arguments were based on similarities in the compositions, isobutyrate, diethyl phthalate, dibutyl phthalate, bd diotyl phtha­ structures, shapes, and sizes of these minerals with terestrial bio­ late probably come fom the outgassing of sample bags, vinyl minerals bd the apparent absence of plausible abiogenic origins. gloves, gasket materias, or other room plasucs. The variety of de­ Here we compare the cabonate asemblages to pssible martian, tected siloxbes have probable sources in silicone rubber sealant terrestial, and meteoritic analogs bd discuss new bd published materials, silicone adhesives, bd elastomers in ceiling-light gas­ arguments fr bd against abiogenic and biogenic origins fr these kets. minerals. Contamination in the Mars meteorite cabinet may be fom Magnetite: McKay et a. [1,2] argued on the bais of struc­ airlock transfr of tools and bags with adsorbed or electrostatically ture, size, morphology, and chemical composition that the attached contaminbts, fom previous sbples, fom contaminants submicrometer-sized magneutes (Fe304) were lagely frmed in­ due to the attachment technique, or fom impurities in the N sup­ tracellularly by magnetotactic microorganisms. Magnetites in ply. Measurements of the molecula organics in the incoming N ALH 84001 are mostly cuboid, cubo-octahedral, teardrop bd ir­ supply line need to be made to verify which contaminants come regula in shape (1-3]. A small faction are elongated rods or rib­ fom the N and which come fom processing materials and the bons [4] and some are parallelepipedal or hexagonal prisms (5). sbple itself. The source of isopropanol is likely residual material Cubo-octahedra and hexagonal prisms elongated aong [111] and fom the cleaning of heat sealers. Caprolactam is a common break­ bullet-shaped crystals fred by asymmetric growth on cubo-oc­ down product of Nylon-6, often used as sample storage bags, and tahedra ae most characteristic of magnetotactic bactera [6], bd is likely released during heat sealing of these bags. The siloxanes some of these shapes are observed in ALH 84001. However, with present in the cabinet may be due to a heat-sealer malfunction the the exception of the bullet shapes, they can also be fred inor­ day prior to measurement in which the silicone rubber upper plate ganically [7]. Thus although the shapes and sizes of many martian wa overheated. grains ae highly suggestive of magnetosomes, except possibly fr The organic contamination measured in the MPL and in the the teardrop shapes, they cbnot definitely be identifed as crystals Mas meteorite cabinet may or may not have signifcant egects on fom magnetotactic microorganisms. Teardrop-shaped magnetite processed sbples. Though present in the air, it is possible that cer­ crystals were originally reprted in magnetotactc bacteria; tis term tain compounds may become selectively concentated on meteor­ was used fr oblate spheroids and "teardrop-shaped cones" [8,9]. ite surfaces, potentially resulting in misleading geological and However, it is not clea whether the martian teardrops resemble geohemical balyses. Exprimentauon has not yet taken place ex­ these shapes. Contrary to [4], rod-shaped bd twinned magnetites ploring such selective concentration. It is notable that caprolactam ae formed biogenically. However, elongated, ribbon-shaped mag­ is used in a variety of biotechnological processes producing the netites and rods with central screw dislocations ae very unlikely amino acid lysine [4]. Tough the process appeas to b highly cata­ to be biogenic [ 4]. lyzed, the similarity in structure of caprolactam and lysine argues Most martian magnetites are single magnetic domains, as ob­ fr frther consideration of this contaminbt in terms of assessing served fr magnetotactic bactera; however, the proportion that is the possibility of lysine frmation on meteorite surfaces. superaamagnetic [5] is larger than obsered fr magnetotactic Through the acknowledgment of these, and perhaps other, con­ bacteria [10]. Parallelepipedal particles <30 nm in length and tamination sources bd the further development of continuous de­ cuboids <50 nm in size, which are present in ALH 84001 (1-5], tection protocols, the uncertainties associated with organic would not have been classifed as biogenic if they had been found contamination on martib sbples can be reduced. Sample cura­ in a terestial sediment. McKay et al. [1] suggested that the smaller, tors at JSC continue to explore these issues. suprparamagnetic magnetites resembled those frmed extracellu- LP Contribution No. 956 43

lcy by the bacterium Geobacter metallireducens. However, this martib cabonates to concretions but noted that disequilibrium in organism is not motile bd frms magnetites that are iregularly the terestial carbonates may be due to biogenic or abiogenic pro­ shaped and porly ordered [10]. There ae no reprts suggesting cesses. that superpaamagnetic martib magnetites ae more irregular in Barrat et al. [19] compaed the martian carbonates to terestrial shape bd less well ordered than single domain magnetites. calcite aggregates that may have formed biogenically in the Some clusters of elongated magnetites are crystallographically Tatahouine meteorite when it lay in the Saharan desert fr 63 yr. oriented with respct to one another and to the cabonate substrate However, these cabonates ae not chemically zoned and lack mag­ [11-13]. Epitaxial growth is inconsistent with intacelula growth netite rims. Direct evidence fvoring a biogenic origin of the ma­ [12], bd extracellula biogenic depsition is also very unlikely as tib cabonates is lacking. such magnetites ae poorly crystalline and iregula in shape. Al­ Abiogenic Origns: Anders [20] suggested that the ALH car­ though biogenic magnetites cb be crystallographically oriented on bonate asemblages were simila to those in cabonaceous chon­ organic substates with great precision [14], no organism is kown drites and had fred aiogenically at low tempratures. However, to deposit carbonate bd magnetite on the sae organic substate. McKay et al. [2] noted numerous differences in gain size, com­ Thus there is no reason to expct epitaial, biogenic deposition of position, and texture between maib bd chondritic occurences. magnetite on carbonate on by planet. Although Cl bd CM chondrites contain calcite, dolomite, bd mag­ It is very difcult to understbd how magnettes fom magneto­ nesite [21,22], there are no chemically zoned grains like those in tactic organisms could have been intoduced into a plutonic rock ALH 8401. Magnetites in CI chondrites ae larger thb those in like ALH 8401. It is unlikely that magnetotactic organisms would ALH 8401, bd those in CM chondrites are rare and not associ­ navigate bd swim through a maze of cracks in b igneous rock. It ated with peripheral magnesite. Cabonates in cabonaceous chon­ is also very unlikely that by biogenic magnettes fom sediments drites are also not found embedded in pyroxene bd plagioclase or aqueous environments could be washed into factures in glass, as in 8401. ALH 84001 by prcolating fuids bd prefrentially tapped dur­ McKay et al. [1] rejected a high-temperature origin fr the ca­ ing the fnal stages of cabonate crystallization, frming unifrmly bonate assemblages in ALH 84001 on the basis of C- and O­ thin, double shells around cabonates. These factors bd the abun­ isotopic data. However, other authors have fund that calculated dbce of abiogenic magnetites make it very unlikely that by bio­ frmation tempratures could b 40°-250°C [23] or any tempra­ genic magnetites ae present in the ALH 84001 cabonates. ture in the range of 0° to >50°C [24]. The presence of shock­ Iron Sulfdes, Pyrrhotite, and Griegite: Griegite (Fe3S4), fred glass that solidifed in situ in facture zones in ALH 84001 pyrhotite (Fe0_9S), bd other Fe sulfdes cb be fred extacel­ is not consistent with the conclusions of Kirschvink et al. [25], who lularly by sulfte-reducing bacteria but these sulfdes cannot be dis­ argued against high temperatures after fration of the facture tinguished morphologically or chemically fom abiogenic sulfdes zones. [6]. The absence of isotopically light S indicates that the martib Brealey [3] fund evidence for high tempratures in the micro­ sulfdes did not frm by this process [15]. structures at carbonate-glass boundaries bd suggested that shock­ Pyrhotite is not known to frm intacellulaly in magnetotactic melted plagioclase had heated cabonate. The presence of pores in bactera [16]. In adition, some martian pyrhottes ae rounded bd cabonate around magnetites [1] favors decompsition of Fe-rich plycrystalline and thus could not be magnetc single domains like cabonate during shock heating, melting, or vaprization [3,26]. magnetosomes. Griegite magnetosomes ae kown, but it is very Such a proess also appas consistent with Bradley et al. [4], who unlikely that griegite fom magnetotactic mcroorganisms is present concluded that magnette ha condensed above 500°C fom a fuid. in ALH 84001. The particles identifed by McKay et al. [1] as Rapid heating and cooling in seconds of shock-formed fluids griegite ae mor iregular thb known griegite magnetosomes, bd fred above 35 0Pa [26] would be consistent with the constraints the identfcation of griegite is questionable as it was based on the of Valley et al. [18], who argued against high tempratures over similarity in shape (irregula bd approximately rectangula) of a priods of minutes to hours. polycrystalline griegite particle frmed in soil by nonmagnetotactc Conclusions: Arguments fr a biogenic origin fr the carbon­ bacteria. The martian "griegite" particles had pyrhotite composi­ ate, magnetite, bd Fe sulfdes in ALH 8401 ae absent or very tions and were unstable in the electron bea, wherea griegite can weak: An origin fom shock-frmed fuids appears much more be studied in the TEM without decomposing [16]. Thus, no evi­ plausible. Nevertheless, martian sediments bd sedimentary rocks dence suggests that the martib sulfides frmed biogenically, ei­ should be searched fr single-domain magnetites with the unique ther intra- or extacellulaly. chabterstcs of crstals in magnetotbtc bacteria Further seaches Carbonate: McKay et al. [1,2] concluded that the carbonate of cabonates in heavily shocked martian meteorites fr grains of disks in ALH 84001 were simila in size bd shap to some ma­ biogenic minerals cannot be recommended. rine, bacterially induced calcite aggregates of submicrometer crys­ References: [1] McKay D.S. et al. (1996) Science, 273, 924. tallites [17].However, martib carbonate crystals appar to be much [2] McKay D.S.et al. (1996) Science, 274, 2123. [3] Brealey A. J. lager [1]. The unique dumbbell-shaped and brush-shaped aggre­ (1998) LPS XXIX, Abstracts #1451, #1452. [4] Bradley J.P. et al. gates fred by bacteria have not been reported in the martib ca­ (1996) GCA, 60, 5149. [5] Thomas-Keprta K. L. et al. (1998) LPS bonates. XXIX, Abstract #1494. [6] Baylinski D. A. and Moskowitz B. M. McKay et al. [2] also compaed the martib carbonates to chemi­ (1997) in Geomicrobiology (J. F. Banfeld bd K. H. Nealson, eds.), cally zoned concretons of siderite or ankerite in sedimentary rocks. p. 181. [7] Maher B. A. (1990) in Iron Biominerls, p. 179, Ple­ These compsitions ae closer to those of the martian carbonates, num. [8] McNeill D. F. et al. (1988) Geology, 16, 8. [9] Petersen but terrestal exaples lack the extensive Ca zoning fund in small N.et al. (1986) Nature, 320, 611. [10] Spaks N.H. C.et al. (1990) portions of the martan grains. Valley et al. [18] also compaed the EPSL, 98, 14. [11] Bradley J. P. et al. (1998) LPS XXIX, Abstact 44 Workhop on Martian Meteorites

#1757. [12) Bradley J. P. et al. (1998) Meteoritics & Planet. Sci., replacement reactions are totally absent [14). In addition, there 33, in press. [13] Blake D. et al. (1998) LPS XXIX, Abstact #1347. should be compsitional digerences between cabonates in pyrox­ [14) Man S. et al. (1993) Science, 261, 1286. [15] Greenwood J.P. ene and those in plagioclase, but none are observed. Carbonate et al. (1997) GCA, 61, 449. [16) Postfai M. et al. (1998) Science, grains in pyroxene could not have fred by replacement of pla­ 280, 880. [17) Buczynski C. (1991) J. Sed. Petrol., 61, 226. gioclase glass because carbonates are more abundant than glass in [18) Valley J. W. et al. (1997) Science, 275, 1633. [19] Barat J. A. pyroxene, composite grans ae rae, and glass grains ae surounded et al. (1998) Science, 280, 412. [20] Anders A.(1996) Science, 274, by radiating cracks whereas carbonates are not. 2119. [21) Johnson C. A. and Prinz M. (1993) GCA, 57, 2843. Carbonates in Fractures: Studies of disks and veins in thn [22] Endress M. and Bischoff A. (1996) GCA, 60, 489. sections show that cabonate-bearing fctures were closed after car­ [23) Hutchins K. S. and Jakosky B. M. (1997) GRL, 24, 819. bonates had started to frm so that unflled portions of factures [24) Leshin L. A. et al. (1998) GCA, 62, 3. [25) Krschvink J. L. were resealed around the cabonates. The concentration of plaar et al. (1997) Science, 275, 1629. [26] Scott E. R. D. et al. (1996) defcts around cabonate and plagioclase glass inclusions on fac­ Nature, 387, 377. tures suggests that these factures were closed by impact. If ca­ bonates formed fom exterally derived fluids, chemical variations across cabonate grains would refect exteral changes in fluid ORIGIN OF CARBONA TE IN MARTIAN METEORITE composition over time ad factures would have been closed after ALLAN HILLS 84001. E.R. D. Scott and A. N. Krot, Hawai'i cabonate growth.However, if carbonates formed rapidly fom im­ Institute of Geophysics and Plaetology, School of Ocea and Earth pact-derived fuids, the chemical variations must refect interal Science and Technology, University of Hawai'i at Maoa, Honolulu factonation in tappd fuids, and factures would have ben closed HI 96822, USA ([email protected]). during carbonate growth. Several features favor crack closure during growth and interal Introduction: A significant agument advanced by McKay control of chemical factionation. Te tapered edges of some car­ et al. [1,2) in favor of a biogenic origin of the carbonates in bonate disks ad veins, the absence of defration in the tapered ALH 8401 was that abiogenic origins - both low ad high tem­ rims of carbonate grains, ad the presence of well-fceted magne­ perature - appeared to be less plausible. However, subsequent site rims fvor crack closure during carbonate growth. Occurrences shock studies have suggested that impact heating may have modi­ of magnesite at the enclosed end of cabonate-flled cracks are dif­ fied carbonates [3-7). We inferred that plagioclase glass and rae fcult to reconcile with late deposition fom a exterally derived silica grains formed fom impact melts, and Raman spectoscopy fluid, as the cracks would have been efectively sealed with early­ [8] confrmed that plagioclase had been melted at >120°C, sug­ frmed Ca-rich cabonate. In addition, magnesite microdisks are gestng shock pressures exceeded 35-45 GPa. Morphological simi­ fund around some disks ad veins, but not elsewhere, implying larities between grains of plagioclase glass ad iregulaly shaped that chemical factionation of the fuid was caused by carbonate cabonates in fatures and the occurence of intermixed grains in crystallization. We infer that the magnesite microdisks fred fom factures suggested that cabonates may also have been shok heated residual trapped fuid that was squeezed a short distance along the so that a COz-rich fuid was injected into cooler-factured pyrox­ factures as they closed. Carbonates in pyroxene factures are het­ ene during shock decompression causing crystallization in seconds. erogeneously distributed: Heavily defrmed pyroxene crystals Types of Carbonate: Cabonates in ALH 84001 occur in aound interstitial plagioclae glass can contain up to 5-10 vol% three locations: (1) disks, veins, ad irregularly shaped grains in carbonate, consistent with limited transport of cabonate-rich flu­ healed factures in pyroxene; (2) massive grains and globules ad ids along factures. smaller grains associated with plagioclase glass on pyroxene grain Studies of plagiolase glass ad metal-toilite in heavily shocked boundaries; and (3) generally equant grains that appear to L and E chondrites [15) show that monomineralic melts of readily poikilitically enclose pyroxene fagments in facture zones (also defrmed minerals such as toilite and plagioclase are injected into called graular bands ad crushed zones [9,10)). McKay et al [1,2) factures that were frmed by shear stresses and then resealed dur­ studied disk-shaped carbonates, which are abundant on feshly bro­ ing shock decompression. Grains of plagioclase glass ad meta­ ken surfaces; we studied all kinds of carbonates in polished thin troilite are heterogeneously distributed in healed factures in sectons. The close asociation of the three cabonate types ad the pyroxenes as spheroids, sheets, ad iregulaly shaped grains, like uniformity of their chemical zoning trends on MgCOrFeCOr the cabonates in ALH 84001.Metal grains are chemically zoned

CaCO3 plots show that all cabnates in ALH 8401 frmed by the and form crystals > 10 µm because of rapid diffusion in metallic same process [ 4, 11]. melts. Origns of Carbonate: Three different origins have been pro­ Carbonates in Fracture Zones: Fracture zones contain bro­ posed fr cabonates in ALH 8401: (1) biogenic or abiogenic pre­ ken pyroxene crystals 5-10 µm in size that were frmed fom mil­ cipitation fom COz-rich fuids that percolated through factures at limeter-to-centimeter-sized crystals by localized shear stesses, low tempratures [l, 12); (2) reaction between COz-rich fuids and rotated >20° and squeezed together to remove voids. Fracture, ro­ pyroxene, plagioclase, or plagioclase glass at low or high tempera­ tation, and compression occurred during impact(s). We and others tures [13, 14); and (3) shock heating and mobilizaton of preexist­ [e.g., 11,14) fund no evidence fr metamorphic equilibration of ing cabonates [3-5). Because voids ae absent next to cabonates mnerals after facture zones were frmed [9, 16]. Localized, brief (except fr factures that postdate carbonate crystallization), grains fctional heating probably aided void removal. Fracture zones con­ in pyroxene crystals ad plagioclase glass appar to have replaced tain grains of cabonate ad plagioclase glass that are sparsely dis­ the adjacent phase. However, replacement mechaisms ae not plau­ tributed among the pyroxene fagments and are not factured ad sible because phases rich in Si, Al, Na, etc. that form during such deformed like the pyroxene, showing that they formed after the py- LP[ Contribution No. 956 45

roxene crystals were fagmented. Carbonates fll only a tiny fac­ man A. H. (1995) Meteoritics, 30, 294. [10] Mittlefhldt D. W. tion of the spaces that once existed between pyroxene fagments (1994) Meteoritics, 29, 214. [11) McKay G. et al.(1998) Meteor­ and could not have formed after the pyroxene fagments were itics & Planet. Sci., 33, A102. [12) Valley J. W. et al. (1997) Sci­ squeezed together to remove voids.Carbonates must therefre have ence, 275, 1633. [13) Havey R. P. ad McSween H. Y. (1996) fred fom fuids that were present between the pyroxene fag­ Nature, 382, 49. [14) Gleason J. D. et al. (1997) GCA, 61, 3503. ments when the facture zones frmed. Occurences of aligned tails [15) Faga T. J. et al. (1998) Meteoritics & Planet. Sci., 33, A46. of plagioclase glass and chromite grains within facture zones that [16] Treiman A. H. (1998) Meteoritics & Planet. Sci., in press. lead to lager peripheral grains show that plagioclase was molten [17) Schwandt C. S. et al. (1998) Meteoritics & Planet. Sci., 33, ad chromite solid when the facture zones fred by shea stesses. Al39. One large cabonate-plagioclase gain was also fund with ema­ nating paallel tails of cabonate ad plagioclase glas grains [5]. Cabonate-pyroxene boundaries within lage cabonae grains may A COMARISON BETWEEN SULFIDE ASSEMLAGES be subhedral [3,6), inconsistent with low-temprature precipitation I MARTIAN MTEORITES ALLAN IDLLS 84001 AND of carbonate in factures. Cabonates and plagioclae glas in fac­ GOVERNADOR V ALADARES. C.K. Sheaer ad C.Adcock, ture zones both appea to have frmed at high temperatures fom Institute of Meteoritics, Department of Earth and Planetary impact-mobilized fuids that solidifed as voids closed. Sciences, University of New Mexico, Albuquerque NM 87131, Carbonate Mxe with Plagoclase Glass on Pyroxene Grain USA. Boundarie: Textures in these occurences ae highly varied and complex. However, the concentation of carbonate grains in fac­ Introduction: Understading the development, timing, ad tured pyroxene around grain-boundary cabonate-plagioclae glass setting of sulfde ad sulfte depsition is crtical to interpretng S­ intergrowths implies that prior to impact heating, carbonates in isotopic systematics in martian meteorites and future mission ALH 84001 were concentated at interstitial sites together with pla­ sbples. For example, to properly interpret S isotops as a marker gioclase. Small bounts may also have been distributed in fac­ fr biogenic activity in low-temperature hydrothermal mineral as­ tures. The frst cabonates could not have frmed at high tempera­ semblages, it is important to understand the relative sequence of tures, as burial depths of > 120 km at 900°C would have been sulfde and/or sulfate precipitation, the temprature of precipitation needed to prevent carbonate ad pyroxene fom reacting to fr [1], the openness of the system [2], ad the pstdepsitional ther­ CO2.The frst caronates probably crystallized instead fom fuids mal histor of the sulfde-sulfate assemblages [3]. To interpret S at low temperatures in voids formed by gas bubbles during the f­ isotopes in magmatic assemblages, it is critical to decipher the ef­ na stages of igneous crystallization. Carbonates in other marian fct of subsolidus processes. meteorites probably frmed this way, possibly as evaporite depos­ Approach: In this study, the sulfde assemblages in ALH its. The presence of patly flled voids and adjacent plagioclase 8401 (orthopyroxenite) ad Govemador Valadares (nakhlite) were shock melts would have enhanced the shock heating of cabonates. documented using optical microscopy, EMP, and SEM. The Partial decomposition of carbonates by shock heating [6] without orthopyroxenite assemblage of ALH 84001 represents crystalliza­ mobilizaton of shock-frmed fuids appars less plausible, as there tion ad crystal accumulation in a lager intrusive body, whereas is no correlation of chemical zoning with proximity to plagioclase the clinopyroxenite assemblage of Govemador Valadares refects melt. the crystalization and crystal accumulation in either a ultamafc Arguments Against Shock Formation of Carbonates: In baalt fow or a shallow intusion. Identifcation of individual sul­ some shock experiments [17], calcite appas to have been unaf­ fde grains were made using reflectace characteristics and EMP fcted by shock pressures of 50 GPa. However, shock heating of aalyses. Individual sulfide grains were analyzed fr 634S using the minerals is heterogeneous ad probably contolled by shea stesses. Cameca ims 4f at the University of New Mexico using the approach Thus, shea stesses ad matrix efects will be critical. Shock pres­ of Shearer et al. [ 1]. sures in ALH 8401 may also have exceeded 50-60 GPa. Chemi­ cal zoning in cabonates is not explained by this model. However, given our pr understanding of phase relations and kinetcs at high T and P in the terary cabonate system, it is not implausible that zoning resulted fom factionation between solid, melt, and vapor. Impact melt pckets in heavily shoked roks cool at diverse rates according to local conditions so that a unique crystallization se­ quence cannot be specifed. Numerous generations of crosscutting melt veins ad brecciation ae observed in the foors of terestial craters so that many impacts [16] ae not required to account fr defrmation features in ALH 8401. References: [1] McKay D.S. et al. (1996) Science, 273, 924. [2] ] McKay D. S. et al. (1996) Science, 274, 2123. [3] Scott E.R. D. et al. (1997) Nature, 387, 377. [4] Scott E. R. D. et al. (1998) Meteoritics & Planet. Sci., 33, 709. [5] Scott E.R. D.et al. (1998) LPS XXIX, Abstract #1786. [6] Brearley A. J. (1998) LPS XXIX, Abstract #1451. [7] Bradley J.P. (1996) GCA, 60, 5149. [8] Cooney T. F. et al. (1998) LS XXIX, Abstract #1332. [9] Trei- Fig. I. Chromite with minute sulfide inclusions. Scale bar = 200 µm. .ffi Workhop on Martian Meteorites

Results: ALH 8401 and Govesador Valadaes contain both Govesador V aladares and a sulfate associated with pyrite in ALH pyrite and pyrrhoute. Pyrite is the dominat sulfde in ALH 84001, 84001. The sulfdes in ALH 8401 occur in several digerent tex­ whereas pyrhotite is dominant in Govemador Valadaes. Other tural settings that imply digerent origins and thermal histories. S-bearing phases that were identified include chalcopyrite in Sulfidetype J. Small pyrite and pyrrhotite inclusions (1-10 µm) ae associated with large, relatively unbrecciated chromite grains (Fig. 1). Small pyrite grains also occur nea the grain boundary be­ tween the chromite and orthopyroxene. Sulfidetype 2. Lage pyrite aggregates (10-80 µm) ocur within brecciated zones (Fig. 2). Ubiquitous chomite and rare pyrhotite are phases that may be spatially associated with the pyrite aggre­ gates. The interface between the pyrite ad silicate phases ae sharp. There is a trasition between textural type 1 and 2 that is associ­ ated with highly defred chromate. Sulfide type 3. Small, irregular grains of pyrite are bsociated spatially with the base of the cabonate globules. Sulfide type4. Ver irregula-shapd sulfdes occur in the shock glass. The sulfde is probably pyrite. EDS compositional maps in­ dicate that the sulfde is not homogeneous ad may contain abun- dat sulfte (Fig. 3). . Sulfidetype 5. Submicrometer-sized sulfdes have been identi­ Fig. 2. Large pyrite aggegat within brecciated zone. Chomite gins fed in the cabonate. These sulfdes have not been well character­ are comonly associated wit aggegates. Scale bar = 50 µ. ized. In EDS maps of the carbonates, the sulfde abundance as refected by S concentation appeas to be highest in the two Fe­ rich zones in the carbonate globules ad lowest in the Mg-rich zones (Fig. 4). The sulfde mineralogy in Govemador Valadares is not b complex. Pyrhotite and minor pyrite predominantly occur as minute grains (2-25 µm) that ae immersed in the mesostasis that is interstitial to the cumulate augite (Fig. 5). Pyrhotite also occurs infequently b inclusions in the subhedral augite. Discussion: The wide variability in the o34S of sulfdes in ALH 8401 compared to that in other martia samples [l,4,5] re­ fects the multigenerational nature of sulfde frmation ad the ef­ fect of postmagmatic heating-impact events. Te sequence of events that afcted the o34S in the sulfdes is as follows: (1) magmatic crystallization ad accumulation of sulfdes; (2) precipitation of pyrte fllowing the frmation of the impact-induced cataclastic tex­ ture; the S source was both intesal and extesal to the ortho­ pyroxenite; (3) fne-grained, submicrometer sulfides coprecipitated with the cabonate; the Fe-rich carbonates contained a high pro­ Fig. 3. Sulfde gain entained witin shock glass. Iegular surfaces on portion of sulfdes; ad (4) a postcarbonate heating event undoubt­ grain appear to be sulftes. Scale bar = 10 µ. edly affcted the sulfdes and may have perturbed the /34S. Small

Fig. 4. (a) Back-scattered electon image of outer zones in carbonate globule. The two high-Fe zones are bright. The two high-Mg zones are black. (b) X-ray mp of the distibution of S in the cabonate globule. High-S concentatons are assoiated with high-Fe zones. LPI Contribution No. 956 47

Evidence fr a Postcarbonate, High-Temperature Event: There ae at least three lines of evidence tat suggest that these low­ temperature carbonates [1-3] were exposed to a rapid, high-tem­ perature event: dissption of the carbonate globules during the fration and mobilization of the shock glass, the frmaton of oli­ vine inclusions in the orthopyroxene, cd the borphous chaac­ ter of sheet silicates nea the cabonate-glas boundary. The degree of dissption of the cabonate globules by the mo­ bilizauon of shock glass is illustated in Figs. 1-3. Figure 1 shows relatively undisturbed, zoned cabonate. The zoning is paticularly complex as the cabonates precipitated within small cracks in the orthopyroxene, dissoluton voids cd adjacent to csshed fagments of orthopyroxene. Figure 2 shows a similaly zoned cabonate glob­ ule, its base crushed and inected with shock glass. Within the crushed zone, cabonate cleavage fagments provide evidence that Fig. 5. Sulfides in Govemador V aladares. Pyrrhotite inclusion occurs in this texture does not represent caonate growth within shock-glass the augite. Pyrrhotite and minor amounts of pyrite are immersed in the mesostasis and closely associated with the Fe-Ti oxides. Scale bar = voids. Also, the complex zoning observed in the individual fag­ 200 µm. ments was not a product of cabonate melt interaction or growth in glass voids, but rather refects complex carbonate growth a il- sulfdes were entrained in and reacted with the shock glass during this heating event (Fig. 3). In addition, it is possible that the submicrometer sulfdes in the carbonates were frmed during a high-temprature heating event that resulted in prefrential vola­ tilization of the Fe-,S-enriched cabonate zones. Conclusions: Although ALH 84001 and Govesador Vaa­ daes have simila sulfide assemblages (pyrhotite and pyrite), they diger in their complexity. The sulfide assemlage in Govesador V aladares is dominated by pyrhotite cd appears to be magmatic in origin with some subsolidus reequilibrauon. On the other hcd, the sulfde asemblages in ALH 84001 refect magmatic cd mul­ tiple low- and high-temperature subsolidus processes. The more complex genetic history of the sulfdes in ALH 84001 is refected in the relatively lage variaility of o34S (1-9%0) compared to other maan meteorites. This range reflects both high-temperature and low-temperature resetoirs and in situ mixing of these compnents. References: [1] Shearer et al. (1996) GCA, 60, 2921. [2] Sheaer (1997) LPS XXVlll, 1289. [3] Shearer cd Adcock Fig. 1. Complex zoning in carbonate globule. Orthopyroxene wall is highly crushed and partially included in carbonate. (1998) LPS XXIX. [4] Greenwood et al. (1997) GCA, 61, 4449. [5] Greenwood et al. (1998) LPS XXIX.

EVIENCE FOR A LATE-STAGE THRL OVERRT I ALLAN HLS 81 AND ILICATIONS FOR BIO. MRRS. C. K. Sheaer1 cd A. Brearley2, 1Institute of Mete­ oritics, Depament of Eath and Plcetary Sciences, University of New Mexico, Albuquerque NM 87131, USA, 2Depaent of Earh cd Planetary Sciences, University of New Mexico, Albuquerque NM 87131, USA.

Introuction: Decipherng the theral history of ALH 8401 is essenua in evauaung the evidence fr pssible fossil lif in that martc meteorte. The fcus of this debate has primarily involved the temprature of caronate frmation cd ha virtually ignored the pstdepsitional history of the carbonates [l-7]. Here, we pro­ vide evidence fr a tcsient, high-temperature teral pulse fl­ Fig. 2. Carbonates with complexzoning as shown above. The base of lowing carbnate depsition cd describe pssible egects on the these carbonates is highly fracturedand injected with shock glass.Carbon­ cabonates that previously have been attibuted to low-temperature ate fragments are broken along cleavage planes. Olivine (irregular,small processes cd biogenic activity. white masses) occurs in the opx adjacent to the carbonate base. 48 Workhop on Martian Meteorites

temperatures cc be calculated fom its chemical composition [10]. Melts with viscosities of 108 p (log 10 viscosity = 8) have the con­ sistency of cool aphalt or very thick paste. To reproduce the tex­ tural relationships between the carbonate and shock glass, the viscosity of the shock glass must have been <106 p, indicating that it was injected at temperatures exceeding l000°C. Tis agrees with previous qualitative estimates fr the fow of plagioclase glass [11,12]. In such a scenario, would a hgh-temperature thermal event de­ stroy the delicate, low-temprature chemical zoning obseted in the caonate? Te destuction of the chemical zoning by chemical dif­ fusion processes is not necessarily a fnal outcome to this thermal event. The chemical efects the thermal event ha on the carbon­ ates is dictated by the length of the event cd the thermal diffsion Fig. 3. Carbonate globule displaced fom opx wall. Carbonate cleavage fagments and crushed opx are entapped in shock glass. profle across the cabonates. In a simple thermal model, assum­ ing a single 20-µm vein of shock melt at a temperature of 1 l00°C intuded into c infnite cabonate layer at a temperature of 0°C, the caonate adjacent to the shock glass will be heated to 400°- lustated in Fig. 1. Figure 3 also illustes the disruption cd tcs­ 5000C cd will fall to <50°C witin seconds of the event. Under port of cabonate globules by the shock melt. Clealy, in Fig. 3, more complex scenaios, it is conceivable that the caonate was the cabonbe ha been pulled fom the orthopyroxene wall on which exposed to slightly higher temperatures (50°-600°C). The ther­ it had initially crystallized. The csshed zone between the cabon­ mal event, however, is very short lived. If the difsion coefcients ate and the orthopyroxene consists of cabonate cleavage and of Mg and Fe in carbonate are similar to other rock-frming min­ orthopyroxene fagments immersed in shock glass. Displacement erals at these tempratures (10-18 to 10-21 cm2 s-1 ), this theral is on the order of 10-30 µm. Even greater displacement of ca­ event would have heated the cabonates to temperatures of 400°- bonate globules (200-30 µm) has been observed. Evidence for a 6000C without destroying the intrcate, low-temperature chemical late-stage shock event has been recognized by Treimc [8]. zoning. A second line of evidence fr the cabonbe exprencing a high­ Implications fr Biomarkers in ALH 8401: Tis late-stage temperature event is the olivine "inclusions" in the orhopyroxene heating event has a number of important implications fr the ori­ adjacent to the cabonate. The olivine occurs only near factures gin of "biomakers" in ALH 8401. First, this model explains the that contain disrupted carbonate. Note the absence of olivine in occurence of magnetite in cabonate [13]. The concentration of Fig. 1 cd the abundance of olivine in Fig. 2. The inclusions are magnetite in Fe-rich cabonate bcds is produced because this ca­ iregula cd range in size fom -40 µm to submicrometer. The in­ bonate composition is the least thermally stable and will produce clusions are in sharp contact with the enclosing orthopyroxene cd the most magnetite on decomposition. At temperatures as low as often contain small inclusions of chromite. The 6180 values of the 385°C, siderite decomposition can be detected [14, 15], whereas olivine cd orthopyroxene that were analyzed range fom +4.3 to magnesite and calcite ae stable to temperatures of at leat 650°C +5.3%o cd are indistinguishable fom each other within analytical and 985°C respctively [16]. The voids in the cabonates comonly 18 uncertanty. The 6 0 values of the olivine provide stong evidence asociated with magnetite ae produced by the loss of CO2 result­ that the olivine was not in isotopic equilibrium with the carbonate. ing fom the decomposition of the siderite compnent in solid so­ In addition, the specifc location of the olivine adjacent to factures lution. containing disrupted carbonates is highly suggestive of a non­ Second, the wide variability in 6180 observed in the carbon­ magmatic origin. The olivine may have been frmed by a reduc­ ates may be attributed to tis late thermal pulse. The Mg-rch car­ tion of magmatic orthopyroxene and spinel during this late, high­ bonates have 6180 values of 18-22%0, whereas the 6180 of the temperature event. Fe-rich cabonates is much more vaiable cd lower (6-18%0) A third line of evidence fr postdepositional heaung of the ca­ [3,17]. The efect of decarbonation is to decrease the 6180 value bonates has been described by Brearley [9]. Minute sheet silicates of the cabonate [18]. This could explain the lower and variable within the cabonate ae not observed in the shock glas cd ap­ 6180 fr the Fe carbonates, assuming the O-isotopic composition pa to be borphous near the carbonate-glass interface. This im­ of the cabonates was similar prior to heaung. plies that the glass and the cabonate interface were involved in a References: [1] McKay et al. (1996) Science, 273, 924. late-heating event. [2] Shearer et al. (1996) GCA, 60, 2921. [3] Valley et al. Conditions of High-Temperature Event: Mineral geo­ (1997) Science, 275, 1633. [4] Bradley et al. (1996) GCA, 60; thermometers using olivine cd associated phases (orthopyroxene, 5149. [5] Harvey and Mcsween (1994) Meteoritics, 29, 472. spinel) record a high-temperature, subsolidus environment. Iron­ [6] Mittlefehdt (1994) Meteoritics, 29, 214. [7] Scott et al. (1997) magnesium exchange between olivine and chromite inclusions in Nature, 387, 377. [8] Treimc (1998) LPS XXIX. [9] Brealey, this the olivine gives equilibrum temperatures of 850°-90°C. The oli­ volume. [10] Shaw (1972) Am. J. Sci., 272, 870. [11] Stofer et vine-orthopyroxene geotermometer yields temperatures of 800°- al. (1986) GCA, 50, 889. [12] Stofer et a. (1988) in Meteoritics 9000C. and the Early Solar System, p. 165. [13] Brearley (1998) LPS The textures implying mobilization of the cabonates indicate XXIX. [14] Gallagher et al. (1981) Thermochim. Acta, 50, fow of the shock melt. The viscosity of the shock glass at vaous 41. [15] Galagher and Wae (1981) Thermochim. Acta, 50, 253. LPI Conrriburion No. 956 49

[16] Harker and Tuttle (1955) Am. J. Sci., 253, 209. [17] Leshin Overall, better knowledge is needed of the nature of the oxi­ et al. (1998) GCA, 62, 3. [18] Valley (1986) RIMS, 16, 445. dant, the diffusive permeability of the martian regolith, andany­ thing that canbe learned about the vertical profile of the regolith oxidationstate. OVERVIEW: TBESEARCHFORLIFEONMARS. S. W. 2. Where/when has liquid water been available? This ques­ Squyres, 428 Space Sciences Building, Cornell University, Ithaca tion can be partially answered morphologically, since a variety of NY 14853, USA. landforms on Mars provide clear evidence for the flow of liquid water in the past. However, flowingsurface water environments are The search for evidence of life has become the driving force not the most promising target for the search for evidence of life, behind NASA's Mars exploration program. At present, however, since high-energy streambed environments generally do not favor the evidence that Marsonce harbored life is controversial, andour formationof the fine-grainedsedimentary deposits or aqueous min­ understandingof how to search forbetter evidence is rudimentary. erals that tend to best preserveorganics. More attractivetargets are A searchmight be based on the followingassumptions: the remains of standingwater bodies and subsurface hydrothermal At a minimum, life requires biogenic elements, liquid water, and systems. These are more difficult to identify, however, and are a biologically usefulsource of energy. Other requirements are pos­ poorly recognized on the basis of present data. Sites of former sible as well, but just these simple ones can help to narrow the standing water bodies may be identified by subtle morphologic clues search significantly. like ancient shorelines, by topographic data (e.g., closed depres­ The goals of the search should include both ancient and extant sions into which water-carved valleys debouch), and, potentially, life. Because liquid water is now absent near Mars' surface, it may by the mineralogical signatures of evaporites. Sites of past hydro­ be a reasonable strategy to search for evidence of ancient life first thermal systems might be found based on morphologic clues, but andevidence of extantlife subsequently. probably will be best identified on the basis of the mineralogical Ancient life forms, ifthey existed at all, are most likely to have signature of hydrothermal minerals. Exhumation by impacts may been simple, microbial forms. The same is probably true forex­ help in exposing formerlysubsurface hydrothermal environments. tantlife. Subsurface liquid water might be present today, but we have no The best approach to searching for evidence of ancient life is real knowledge of its distribution. Techniques like long-wavelength probably to look for organic materials. Organics of biological ori­ radar sounding could be useful in this regard. gin may be less open to ambiguous interpretations than putative 3. Where/when have biologically useful energy sources been microfossils, andthey have the added virtue of potentiallycontain­ available? Two plausible energy sources have been available in ing important information about prebiotic processes if life never martian history: sunlight and hydrothermal interactions. Sunlight arose. would be available primarily in shallow regions of surface water Working fromthese initial assumptions, several important prob­ bodies. Although cold early climates could cause such bodies to lems must be addressed: be ice covered, light levels beneath meters of ice could still be suf­ 1. In what materials are organics most likely to be pre­ ficientto supportlife. Because of the absence of surface water,sun­ served? The Viking biology experiments failed to detect organic light is unlikely to be a biologically significant energy source on materials in the near-surface martian regolith at the parts-per­ Mars today. Hydrothermal activity was probably common in the billion level. The reasons forthis arenot certain, but it is gener­ ancient crust of Mars. The ancient highlands contain widespread ally attributedto the action of some powerful oxidizing agent pro­ probable volcanic plains and some discrete volcanic features.Abun­ duced photochemically in the martian atmosphere. This putative dant crustal water is indicated by the many fluvial features in the substance diffusesinto martiansurface materials, destroying organic old terrains,and interactions between water and hot volcanic rocks materials that may exist there. Whether or not organics will be found were inevitable. Hydrothermal activity also would have been trig­ in a given martian material depends both on whether they would gered by impacts, since an impact leaves a major thermal anomaly plausibly be there in the first place and whether or not they can be beneath it. Both volcanism and impacts have decreased with time preservedsubsequently against oxidation. Organicsare not expected on Mars,so hydrothermalinteractions probably have decreased and in primary igneous rocks. However, they might be found in sedi­ moved to deeperlevels in the crust. mentary rocks, in the regolith, or in aqueous mineral inclusions 4. What is the appropriate strategy for the search for evidence within rocks of all types. Preservation depends on their distance of martian life? The answerto this question depends on both the from the martian atmosphere and the diffusivepermeability of the scientific considerations discussed above and the practical con­ intervening medium. These considerations suggest two possible straintsimposed by availability of technology andfunding. As noted approaches to the search for organics. One is to look within the above, the depth to which oxidants can diffuseis a function of the martian regolith, although the distance beneath the surface that one diffusive permeability of the medium it is diffusing through. The must look is poorly known and may be considerable. Another is to permeability of a porous regolith may be high enough that plau­ look within sedimentary rocks or within rocks containing aqueous sible estimates of oxidant diffusion depths are many meters. The mineral deposits. This strategy could involve penetratingonly very permeability of solid rock, however, is typically many orders of short distances, but such rocks may be difficult to find. One po­ magnitude lower, so that plausible diffusion lengthscales are cen­ tentially attractive strategy might be to look in portions of the re­ timeters or millimeters. The technology to drill remotely through golith cemented by ice, which can dramatically reduce diffusive many meters of regolith is not available at present, but several tech­ permeability. Our knowledge of the distribution of ground ice on niques exist for breaking or drilling short distances into surface Mars is poor, however, and ice-rich sites may tend to lie at high rocks. This fact, then, suggests a strategy that focuses in the near latitudes where the past presence of liquid water is questionable. term on use of instruments that locate rocks containing aqueous 50 Workhop on Martian Meteorites

mineral precipitates, and on a sampling technique that can obtain voring a terestrial origin fr the majority of organic matter in materials fom within such rocks and retu them to Earth. At the ALH 84001 [7]. same time, it will be importat to explore the vertica oxidation Analytical Technique: To investigate the actua distbution profle in the regolith via penetrators or drills, and the subsurface of PAHs in ALH 84001 on a submicrometer scale, we used time­ distribution of liquid water via radar sounding. The information of-fight secondary ion mass spectrometry (TOF-SIMS) with a lat­ gained would then be used to choose locations and techniques fr eral resolution of about 0.2 µm to analyze areas on three digerent deep sampling activities on subsequent missions. thin sections of ALH 84001 [8,9]. This technique uses a time-of­ fight mass spctometer fr the analysis of psitve or negative sec­ ondary ions sputtered fom the uppermost monolayers during TH LATERAL DISTRIUTION OF POLYCYCLIC ARO­ primary ion bombadment. Besides atomic ions, lage molecules MATIC HYDROCARBONS I ALLAN HILS 8401: J­ also survive the sputterng process, at least as chaacteristic fag­ PLICATIONS FOR THIR ORIGI. T. Stephan1, D. Rost1, ments. Using a fne-fcused Ga liquid metal prima ion source, a C. H. Heiss t, E. K. Jessbergert, and A. Greshake2, tJnstitut fir simultaeous measurement of the latera distrbution fr al second­ Plaetologie/lnterdisciplinary Center fr Electron Microscopy and ary ion species with one polarity is possible. High mass resolution Microanalysis, Westfilische Wilhelms-Universitat Miinster, alows for te sepaation of atomic ions fom hydrocabons at the Wilhelm-Klemm-Strasse 10, D-48149 Miinster, Germany sae nominal masses. ([email protected]), 2Museum fr Naturkunde, Institut fir Results: Our TOF-SIMS studies revealed PAHs on all three Mineralogie, Humboldt-Universitat zu Berlin, Invalidenstrasse 43, sections of ALH 84001, but in no case was a corelaton of PAHs D-10115 Berlin, Geray. with cabonates obsered. On the contrary, a general trend of a lower PAH concentaion in cabonates compaed to orthopyroxene Introduction: Polycyclic aromatic hydrocabons (PAHs) in or feldspatic glass seems to exist [8,9]. ALH 84001 were considered to be evidence fr relic biogenic ac­ Since the infrmation depth in a typical measurement is on the tivity on Mars [l]. Crcia fr their propsed connection to ealy order of a few atomic monolayers, surface contamination is a ma­ lif frms is a suggested spatia association with cabonates that jor concer in TOF-SIS. The three investigated thin sections were contain interal stuctures resembling terestia microfssils. Al­ prepaed fom three diferent pieces of ALH 84001 in two difer­ though each observation ca be explained individually [2,3), a lat­ ent laboraories, minimizing the probability of a common contami­ eral corelation would suggest a genetic link between P AHs ad nation with exteral PAHs during saple preparation. A section the microstuctures. of Chassigny [10) that was treated exactly like one of our A preliminary study of ALH 8401 by Thomas et al. revealed ALH 84001 sections and measured directly before and after no clear corelaton of PAHs and the so-caled carbonate globules ALH 8401 contained no detectable PAHs. Therefre, contamina­ [4]. Three regions on a feshly factured surface with a total size tion with PAHs not indigenous to the sample is unlikely. of 1 x 0.5 mm2 were fund to be enriched in PAHs. Although a On the other had, redistibution of indigenous PAHs may have relatonship seems to exist between the region with the highest P AH occured during saple prepaation, e.g., polishing. This indeed was concentation ad a carbonate spheroid, the two other regions with observed in one of our samples. For this section we observed first enhaced concentrations of P AHs did not corelate with cabon­ a rather unifrm distribution of PAHs on the very surface. After ates. On te other had, not all cabonate globules in the aayzed extensive sputtering with an Ar ion beam that removed several sample contain PAH excesses or even PAH concentrations above atomic monolayers, the indigenous P AH distibution was revealed, background. showing the previously obsered "aticorelation" of P AHs ad car­ Nevertheless, in a later work by the same authors, a spatial as­ bonates [9]. sociation of P AHs ad cabonates was proposed and interpreted as To understad the relative intensities of chaacteristic PAH mass a major clue to relic biogenic activity without presenting convinc­ peaks in the TOF-SIMS spectra [8], which vary signifcantly fom ing new results [l]. Even in a recent paper, some of the authors those observed by µVMS [l], we investigated with TOF-SIMS used the fst study to claim that such a corelation exists [5]. pure PAH substaces, pentacene and coronene (Fig. 1), on Ag sub­ · Crucial fr the investigation of spatial associations is the lat­ strates. Both substances showed factionation during primary ion eral resolution of the applied aalytical methods. In the mentioned bombardment. Pentacene, being more susceptible to factionation, studies, microprobe two-step laser mas spectometry (µL2MS) was revealed simila relative maxima, athough in diferent proprtions, used fr the aalysis of PAHs. With this technique, a unambigu­ ous determinaion of a spatia association is limited by its lateral resolution, about 50 µm, a size that is comparable with the typica dimension of the cabonate globules. Scaning trasmission X-ray microscope (STXM) C mapping and X-ray absostion near-edge structure (XANES) spectoscopic measurements on cabonates fom ALH 8401 indicated the pres­ ence of organic C (7-bounded C) within or associbed with cabon­ ate globules on the scale of -100 nm [6]. A defnite identification of the nature of this organic C, which seems to be locally present at prcent levels, is not pssible with these techniques. Only a smal faction,

in the mas spctm as observed for ALH 8401. Differences fom MINERAL BIOMARKERS IN MARTIAN METEORITE µL2MS specta, especially the high abundance of low-mass PAHs ALLAN HILLS 84001? K. L. Thomas-Kepra1, D. A. Bazy­ observed by TOF-SlMS, ca be explained by differences in ion­ linski2, S. J. Wentworth1, D. S. McKay3, D. C. Golden1, E. K. ization processes, e.g., stronger factionaton during ion bombard­ Gibson Jr.3, and C. S. Romanek4, 1Mail Code C23, Lockheed ment. It is therefore plausible that spectra obtained by µL2MS ad Martin, NASA Johnson Space Center, Houston TX 77058, USA TOF-SlS result fom the same mixture of PAHs. ([email protected]), 2Department of Microbiology, Imu­ Discussion: A major question is where these PAHs, unam­ nology, and Preventive Medicine, Iowa State University, Ames IA biguously present in ALH 84001, come fom. Are they indigenous 50011, USA, 3Mail Code SN, NASA Johnson Space Center, Hous­ [1,5] to this meteorite or were they acquired during residence in ton TX 77058, USA, 4Drawer E, Savannah River Ecology Labo­ the Antactic ice [3]? An importat agument against their teres­ ratory, University of Georgia, Aiken SC 29802, USA. tial origin is believed to be the reprted decrease by a fctor of 5-10 of the PAH signal toward the fusion crust of the meteor­ Introuction: The occuence of fne-grained magnetite in the ite [1,5]. This decrease can at least in part be explained by Fe-rich rims surrounding cabonate globules in the martia mete­ digerent morhological properties of the fusion crst, which has orite ALH 84001, originally described in [I], have been propsed a inherently smaller surface area. If the PAHs observed in ALH as fossil remains of primitive martia organisms. Here we repor 8401 result fom terestrial contamination and now reside along obsesbions on size ad shape distibutions of magnetites fom submicrometer-sized fations or grain boundaries, a corelation of ALH 84001 ad compae them to biogenic and inorganic magne­ PAH concentation with interal surface area would be expected. tite crystals of terestrial origin. While some magnetite morphol­ From the TOF-SIS measurements, such a concentration along ogy is not unequivocally diagnostic for its biogenicity, such as grain boundaes or submicrometer factions can neither be con­ cubodial forms of magnetite [3], which are common in inorgani­ frmed nor ruled out. cally frmed magnetites, other morpholgies of magnetite (parallel­ Conclusions: Irespective of their orgin, a genetic link be­ epiped or elongated prismatic [4] and arowhead frms [5]) are tween PAHs and carbonate globules, the carier of the so-called more likely signatures of biogenic activity [2]. Some ALH 84001 nanofssils, can be excluded based on the lack of spatial associa­ magnetite particles described below have unique morphology and tion. Consequently, these observations have to be taken individu­ length-to-width ratios that ae indistinguishable fom a vaiety of ally and should not be considered collectively. terrestial biogenic magnette and distnct fom all known inorgaic Acknowledgments: This work was supported by a special frms of magnetite. grant fom the Ministerium fiir Wissenschaft und Forschung des Methos: Two ALH 84001 chips containing carbonate glob­ Lades Nordrhein-Westfalen. ules were immersed in -20 ml of 20% acetic acid at -50°C for 72 References: [1] McKay D. S. et al. (1996) Science, 273, 491- and 96 hr. Contols show that magnetite is not affcted by this acid 504. [2] Bradley J. P. et al. (1997) Science, 390, 454. [3] Becker treatent. The residual chips were removed fom the acid; the acid L. et al. (1997) GCA, 61, 475-481. [4] Thomas K. L. et al. (1995) containing the magnetite was placed on C-coated TEM (tasmis­ LPS XXVl, 1409-1410. [5] Clemett S. J. et al (1998) Faraday Dis­ sion electon microscope) grids ad allowed to evaporate. These cussions (Royal Soc. Chem.), 109, 417-436. [6] Flynn. G. J. (1998) grids were examined at 160 ad 200 kV using a JEOL 200 FX Meteoritics & Planet. Sci., 33, A50-A51. [7] Juli. A. J. T. (1998) TEM equipped with a Link System IV energy-dispersive X-ray Science, 179, 366-369. [8] Stepha T. et a. (1998) LPSXIX, Ab­ spectrometer. stact #1263. [9] Stephan T. et al. (1998) Meteoritics & Planet. Sci., Results and Discussion: Thousands of magnetite grains in 33, A149-Al50. [10] Grshake A. et al. (1998) LPS XXIX, Abstact clumps or groups were fund on the TEM grids. The magnetite #1069. ranges fom -10 to 200 nm in size (in the longest dimension) ad

i 0.30 r.,, .,. :.,.

Fig. 1. Two views of a magnetite crystal fom ALH 8401. (a) displays a hexagonal cross section at +17°; (b) shows a hexagonal outline at -22°. Compare with Fig. 4a,b. 52 Workshop on Martian Meteorites

are compsed only of Fe and 0. Of the 526 magnetites documented, ing hexagonal cross sections when viewed along the (111) ais (see we have identified three distinct populations: elongated prisms, Figs. la,b on previous page). More tha 99% of these magnetites whisker, ad iregula. Elongated prismatc magnetites constitute have length/width or axial ratios between 1 and 2 (Fig. 2). Whis­ 27% of the population and are defned as moderately elongated ker magnetites, defined as having length/width ratios >3, consti­ (length/width <2.6) magnetites with rectangular projections hav- tute 6% of the total (Figs. 2 ad 3). We fund two typs of whisker magnetites (rounded ends with rounded cross sections and whis­ kers that tum platy when viewed frm aother angle of rotation (see Fig. 3). Iregular magnetites (including cuboidal, teadrop, and gen­ I01 Irregular M (no whiskers); n=52 •M-1Bloenl M; n=15 erally iregula frs) made up 67% of the population. For com­ CALH801 Prismatic M; n=142 parison, we performed similar analyses of magnetite crystals isolated fom bacteria strain MV-1 ad specific abiotic magnetite; A1 Whiks; n = we also compared these results to those described in the literature 0 2 fr abiotic magnetite [6,7]. The elongated prismatic magnetite par­ 5 ticles in ALH 8401 have high chemical purity, a restricted range of axial ratios, ad distinct size distributions. Additionally, they hae a unique morphology chaacterized by elongated growth along o..._�------' the (111) ais, a hexagonal cross section when viewed along that 1 3 5 7 9 11 13 15 17 axis, and a rectangula outline when viewed normal to the (11 I) Axial Raio (Lngtidh) ais (Figs. la,b). These specifc properties make them identical to Fig. 2. magnetite crystas produced by specifc extant terrestial magneto­ tactic bactera (Figs. 4a,b; [8]). These unique properties ae not ob­ served in any known natural, inorganically produced magnetite crystals. Inorganic magnetite particles we observed ad those fom the literature ae nearly equidimensional (axial ratios -1), and none obsered or described fom the literature have the elongated pris­ matic shaps and aial ratios of the elongated prismatic magnetites in ALH 8401 ad specifc biogenic magnetite crystals (e.g., MV- 1). Unless a inorganic analog fr these crystals is fund, the pres­ ence of elongated prismatic magnetite crystals in marta meteorite ALH 8401 must be considered as likely evidence fr primitive life on early Mas. References: [1] McKay et al. (1996) Science, 273, 924. ,,_.;,,-,.. ,,' [2] Bazylinski ad Moskowitz (1997) in Geomicrobiology: Inter­ ' ' �.t actions between Microbes and Minerals (Banfeld and Nealson, eds.), p. 181. [3] Balkwill et al. (1980) J. Bacteriol.,141, 1399. Fig. 3. [4] Baylinski et al. (1988) Nature, 334, 518. [SJ Thorll et al.

Fig. 4. Two views of an elongated prismatc biogenic magnetite [fom 8]; scale bar = 20 r. LP/ Contribution No. 956 53

(1994) FEMS Microbial. Lett., 115, 169. [6] Schwertmann and Space Center, Houston TX 77058, USA, 3Battelle, Pacific Corell (1991) in Iron Oxides in the Laboratory: Preparation and Northwest Laboratory, P.O. Box 999, Richlad WA 99352, USA, Characterization, p. 111. [7] Nanophase magnetite produced by 4Department of Chemistry and Biochemistry, University of Nanophase Technology at www.nanophase.com/TEXT/PROD­ Arkansas, Fayetteville AR 72701, USA, 5Mail Code SN4, NASA UCTS/1ron0xide.html. [8] Sparks et al. (1990) EPSL, 98, 14. Johnson Space Center, Houston TX 77058, USA, 6Drawer E, Savana River Ecology Laboratory, University of Georgia, Aiken SC 29802, USA. MINERALIZATION OF BACTERIA IN TERRESTRIAL BASALTIC ROCKS: COMPARISON WITH PO SSIBLE The identifcation of biogenic fatures altered by diagenesis or BIOGENIC FEATURES IN MARTIAN METEORITE mineralization is important in deterg whether spcifc fatures ALLAN HLS 801. K. L. Thomas-Keprta1, D. S. McKay2, in terestial rocks [e.g., 1,2] and in meteorites may have a biogenic S. J. Wentworth1, T. 0. Stevens3, A. E. Taunton4, C. C. Allen1, origin [3]. Unfortunately, fw studies have addressed the forma­ E. K. Gibson Jr.5, ad C. S. Romanek6, 1Mail Code C23, Lockeed tion of biogenic fatures in igneous rocks, which may be impor­ Martin, NASA Johnson Space Center, Houston TX 77058, USA tat to these phenomena, including the contoversy over possible ([email protected]), 2Mail Code SN, NASA Johnson biogenic fatures in basaltic martian meteorite ALH 84001 [3]. To

ALH84001 Columbia River Basalt

Fig. 1. 54 Workhop on Martian Meteorites

explore the presence of biogenic features in igneous rocks, we ex­ ized weblike biofilms may also be preserved and suggest previous amined microcosms growing in basaltic small-scale experimental biogenic activity. growth chambers or microcosms. Microbial communities were har­ Here we present new size measurements and photomicrographs vested from aquifers of the Columbia River Basalt (CRB) group comparing the putative martian fossilsto biogenic material in the and grown in a microcosmcontaining unweathered basaltchips and CRB microcosms (Fig. 1 ). The range of size and shapes of the bio­ groundwater (technique described in [4]). These microcosms simu­ genic featureson the CRB microcosm chips overlaps with and is lated natural growth conditions in the deep subsurface of the CRB, similarto those on ALH 84001chips. Although this present work which should be a good terrestrial analogfor any putative martian does not provide evidence for the biogenicity of ALH 84001 fea­ subsurfaceecosystem that may have once included ALH 84001 [9]. tures, we believe that, based on criteria of size, shape, and general Inoculated, fresh (control), and sterile-solution rock chips were morphology, a biogenic interpretation for the ALH 84001 features examined with a high-resolution Philips XL 40 field emissiongun remains plausible. scanning electron microscope (FEGSEM) and a JEOL 2000 FX References: [1] Hoffman H. J. and Schopf J. W. (1983) in transmissionelectron microscope (TEM). FEGSEM samples were Earth'sEarliest Biosphere: Its Origin and Evolution (J. W. Schopf, coated with -2-5 nm of Au-Pd conductive coating. Uncoated ed.), PrincetonUniv. [2] FerrisF. G. et al. (1986) Nature, 320, 609. microorganismswere also examined in the TEM at 160 kV. [3] McKay D. S. et al. (1996) Science, 273, 924. [4] Stevens T. 0. We examinedmore than30 surfaces of rock chips fromthe in­ andMcKinley J. P. (1995) Science, 270, 450. oculated basaltic microcosms. A detailed study of five chips of uninoculated basalt chips revealed no cell-like forms or append­ ages. The inoculated chips displayed three dominant morphologi­ ANCIENT MAR TIAN LIFE IN ALLAN HILLS 84001? cal types of forms that we interpret as microorganisms: Type 1 is STATUS OF SOME CURRENT CONTROVERSIES. A. H. an oval-shaped (coccobacillus) form with a smooth surface texture Treiman,Lunar and Planetary Institute,3600 Bay Area Boulevard, that ranged from -1 to 2.5 µm in length and slightly Jess in width. Houston TX 77058, USA ([email protected]). These organisms arecomposed chieflyof Fe, Mn, and O with mi­ nor elements including P. Type 2 is a similarlyshaped coccobacil­ Four lines of evidence were taken to suggest that ALH 84001 lus form in the same size range but with a textured surface contained traces of ancient martian life preserved in its carbonate composed of thousandsof thin (-2-5 nm diameter) interwovenfila­ mineral masses [1]: abundance of organic compounds(PAHs), dis­ ments composed mainly of Fe and O as ferrihydrite, an Fe+3 hy­ equilibrium mineral assemblages, morphology of submicrometer droxide with minor elements including P. Type 1 and 2 organisms magnetite crystals, and presence of objects comparable in size and are significantlymineralized; both typesdisplay organisms that ap­ shapeto bacteria This evidence is predicated on the carbonateglob­ pear hollow. It is not clear if the formation of a mineralized shell ules having formed at temperatures conducive to life. Here, I re­ was a replacement for the original cell wall or a coating that de­ view evidence on carbonate formation temperature,martian origin veloped external to the cell. Nevertheless, there was no evidence of organic compounds,and bacteria-shaped objects. for the preservation of any ultrastructure, suchas the cell wall. Type Formation Temperature: A precondition for signs of lifeas­ 3 is a tubular (bacillus or rod) form to which a single appendage sociated with the carbonate masses [1] is that the masses formed may be attached at one end. Type 3 formsrange from -0.30 to 2.4 at a temperature amenable to life as we know it [2], <-150°C. The µm in the longest direction without appendage. Microcolonies of formation temperature of the carbonatesis poorlyconstrained [3,4); bacillus forms are frequently observed embedded in biofilm. Type most current hypotheses suggest either formation at T < -200°C 3 organisms are composedmainly of C, 0, and Na with minor P, from water-rich fluid(possibly saline) or formation at T > -250°C S, and Cl; they are unmineralized cells. All three types of organ­ from CO2-rich fluid. Relevantevidence cannot be interpreted un­ isms contain minor P whether they are mineralized or not. In addi­ ambiguously, in part because the geologic context of ALH 84001 tion to the organisms, we observed attached and unattached is unknown, and in part because of its complex history [5]. filaments. Filaments are generally longer andthinner than organ­ The rangeof elemental compositions of the carbonates has been isms; they have distinctivetubular morphologies.Attached filaments taken as indicating quasi-equilibrium at T -700°C [4,6]. The fine­ are similar in composition to the type 3 unmineralized cells. Unat­ scale chemical zoning of the carbonatesrequires that this high tem­ tached filamentsare composed of ferrrihydrite and other minor el­ perature must have lasted only a short time, as might be expected ements inlcuding P; their composition is similar to the type 2 in an impact event [4,7). On the other hand, the elemental compo­ organisms. Some unattached filaments are hollow (Fig. 1). We be­ sitions could have formed at chemical disequilibriaat low tempera­ lieve that these unattached filaments are mineralized cellular ap­ ture [8,9]. The absence of hydrous minerals in ALH 84001 can be pendages because they are not present in the uninoculated control explained in either model. At higher temperature, the absence of microcosms; we conclude that they were likely produced by bio­ hydrous minerals may reflect the presence of CO2-rich fluid logical mechanisms. In addition, the presence of P indicates that [4,7,10,11). At low temperature, one can invoke slow reactions or these filaments are most likely biogenic. If so, it is clear that the very saline aqueous fluids [12-14]. mineralization of bacteria is not necessarily restricted to the main The O-isotopic compositions of the carbonates may suggest body of the organisms; much smaller appendages can also be pre­ equilibrium at low temperature from hydrous fluids [3], although served. The present observations suggest that microfossils in ba­ the inferred temperatures depend on what O-isotopic composition saltic rocks may contain little C, lack evidence of cellular is taken for the depositing fluids (15]. On the other hand, the 0 ultrastructure,and include mineralized subcellular appendages. The isotopes also appear consistent with carbonate growth either from presence of abundant biofilm may also serve as a substrate upon aqueous solutions at variablelow temperature or from carbonicflu­ which minerals may be deposited. If so, it is likely that mineral- ids at T > -500°C [11]. LP/ Contribution No. 956 55

The cabonates in ALH 84001 occur as facture fllings, lacy Summary. It seems likely that the PAHs in ALH 8401 ae (at patches, cd rounded volumes bong cd within pyroxene cd pla­ least in part) martian, but that other organic matter is terestrial. It gioclase glass. These textures have been interpreted as void fill­ is not known whether the PAHs ae biogenic or not, nor whether ings (low temperature) [1,16], products of solution/precipitation this issue cc be resolved. (indeterminate T) [10,14], or solidifed carbonate melt (high tem­ Posible Biogenic Objets: The most publicly complling line perature) [7]. of evidence is the presence of bacteria-shaped objects (BSO) in car­ Considerable work has gone to the temperature signifccce of bonate masses [1,34]. However, their identifcation as biogenic is submicrometer magnetites in the· carbonates. Equant magnetites ambiguous in the absence of intesal structures (contary to early have been interpreted as biogenic magnetosomes fom magneto-tac­ hints [35]), undisputed martic biochemicals (see abve), cd bio­ tic bacteria [1,17], although these shapes might equally well fr logic community structures [e.g., 36]. Morphology alone is prob­ without lif. Elongate magnetite grains, some epitaxial on cabon­ lematc; BSOs cc arise in many ways [e.g., 37]. ates cd some with axial screw dislocations, have been suggested Too Small? At the 1996 press conference on ALH 8401, the as products of vapor deposition at high temperature [18,19], al­ BSOs were claimed to be too small fr life: 20-100 nm [1] vs. though simila magnette shapes ocur in bacteria [20]. Also, it is >20 nm diameter fr fee-living bacteria [38]. The BSOs enclose pssible that the elongate magnetites frmed at high temprature, too little volume to contain the requisite molecules fr Earth lif but after the cabonates frmed [5]. Magnetites ae also present in [2,38], a point contested by some [39]. The BSO size rcge has void spaes in the cabonate, either a the product of decabonation been extended to 7 50 nm, cd there is some evidence of Ea bac­ at high temperature [21] or a entrapped fom the aqueous fuid teria as small as 80 nm [40]. The smaller objects in ALH 8401 that deposited the cabonates [22]. could also be bacterial appndages, not whole organisms [41]. Summary. There is no consensus on the formation tempra­ Inorganic? The 100-nm elongate BSOs may be magnetite ture or mechanism of carbonate masses. The bulk of evidence, in crstals that frmed at high temprature [14,19]. Aligned "swas" my opinion, is not consistent with T > 400°C [4,6,7]. At present, . of -100-nm BSOs [1] may be magnetite crystals epitaxially ori­ there seems no way to tell whether the cabonates frmed at mod­ ented on cabonate mineral substates [19 ,42,43] or irregularities erate or low temperature (hotter or colder than -150°C). on mineral surfaces [14]; McKay et al. may no longer consider the Organic Matter: McKay et al. [l] presented evidence that the swarms biogenic [40]. cabonate pancakes in ALH 8401 ae enriched in polycyclic ao­ Terrestrial? In the absence of (bio)chemical characterizatons, matic hydrocarbons, PAHs, which are organic compounds com­ it remains possible that the BSOs ae terrestal. Endolithic organ­ monly fred fom (or by) biological orgcisms. Absence of PAHs isms ae well known [44]; fungi and bacteria can live in Antarctic nea the meteorite's fsion crust implied that the PAHs ae ma­ meteorites [45]. Spores of these orgcisms, cd lager biologic tic. Confirming evidence included the high abundcce of organ­ objects like diatom tests [46], are distibuted worldwide by wind. ics in ALH 84001 [23]. Work here ha fllowed mcy intercon­ BSOs have been found in lunar meteorites fom Antarctica, which nected threads. presumably never experienced clement conditions before land­ Are PAHs associated with the carbonate globules? The aso­ ing in Antarctica [34]. This last result suggests that BSOs in ciation was amplifed by [24] and confrmed by XANES calyses ALH 8401 ae terestial, although McKay [47] reprts never hav­ [25]. An ealy abstact fom the McKay group states that some ca­ ing found BSOs in lunar rocks or other meteorites. bonate globules have low PAR content [26]. Recent TOFSIMS Artificial? It has been suggested that some BSOs fgured in analyses on polished surfaces found P AHs evenly distributed [1] ae artifats of sample prepaation [18]. This claim is disputed through the silicate cd oxide minerals, but less abundant in the [40], cd scanning frce microscopy measurements appa to sug­ cabonates [27]; this result seems improbable. gest the BSOs ae not artifcial [48]. Yet, the morphology of some Are the PAHs martian? The absence of PAHs nea ALH images, especially the ppula "wor," is suggestve of coating a­ 84001 's fusion crst suggests that they are preterrestial [1,24]. It tifat. has been claimed that the PAHs ae similar to those of Antactic Summary. The bacteria-shaped objects in ALH 84001 ae vi­ ice, cd hence contaminants, cd this is supported by experiments sually appaling, but their actual identity remains uncertain. on solubility, transport, cd adsorption behavior of PAHs [28]. Conclusions: The hypthesis that ALH 84001 contains taces These results have been stongly criticized [24]. of ccient martan lif [1] has not been proved nor disproved, and Are the PAHs biogenic? The PAHs in ALH 8401 ae claimed al parts of the hypthesis remain contoversial. So far, it has not to be like those produced by degradation of bacteria [1,24] (pssi­ been shown that the carbonate masses frmed at a high tempra­ bly not martic [28]). But the variety and abundcces also show ture inimical to lif. This failure is not proof of the hypothesis; it similarities to PAHs in CM chondrite meteorites and common in­ is merely not a disproof. It seems likely that the PAHs in terplcetary dust paticles [29]. The PAHs could also have frmed ALH 8401 formed on Mars, but this is not proof of martic biol­ inorganically (Fischer-Tropsch reaction), possibly catalyzed by ogy. The missing link is the correlation between observed PAHs magnetite grains [30]. But it may be impssible to determine the cd those of degraded bacteria, cd that link may have been per­ origin of the P AHs, because Antarctic weathering cc degrade most mcently severed [31]. The "Mars Bugs" ae under siege, attacked distnguishing chaacters of biogenic PAHs [31]. fom all sides ·cd so far with few reinfrcements. Are other organics martian? Probably not. Nealy all the or­ Whether or not it is proved correct, the hypthesis of [l] has ganic C in ALH 84001 has live 14C, meaning that it frmed fairly stimulated interest cd research on ALH 84001, other martan me­ recently on Earth [32]. Similaly, amino acids extacted in bul fom teorites, and Mars itself. Renewed interest in science is good fr ALH 84001 have abundcces and chirality typical of Earth biologi­ the nation, and renewed funding is benefcial to the reseach com­ cal matter [3 3]. munity. But this debate has stirred some diffcult questions: (1) Are 56 Workhop on Martian Meteorites

the standards fr publication less stringent fr "hot-topic" papers eration and presesation of evidence fr lif [4]. To fully answer than for "routine" papers? (2) Why has it been so difcult to reach these questions, we must determine the processes that frmed the a consensus about ALH 8401? Are the carbonates so intractable textures in the meteorite and what implications these have fr the that our best scientists cannot agree on their frmation tempera­ environment of cabonate depositon. Discussions of the histor ad ture within 700°C? Or are we not nearly as clever as we thought? origins of the textures in ALH 84001 have been synthesized by (3) How will NASA fae if the "life in ALH 8401" hypothesis is [1,3,5,6], among others. All such histories propose the occurence conclusively refuted? of shock melting and metamorphism for ALH 84001 despite References: [1] McKay et al. (1996) Science, 273, 924. nonunique evidence fr shock. Here we present an alternative [2] Nealson K. H. (1997) JGR, 102, 23675. [3] Romanek et al. mechanism fr the frmation of crsh zones in ALH 8401, which (1994) Nature, 372, 655. [4] Havey ad McSween (1996) Nature, may also explain the occurence of feldspa, silica, and cabonate 382, 49. [5] Treima A. (1998) Meteoritics & Planet. Sci., 33, in melts. press. [6] Mittlefhldt D. (1994) Meteoritics, 29, 214. [7] Scott E. Int roduction: One of the puzzling aspects of ALH84001 is its et al. (1997) Nature, 387, 377. [8] Valley J. et al. (1997) Science, shock state. Treiman [5], in his interpretation of the shock hstory 275, 1633. [9] Treima ad Romaek (1998) Meteoritics & Planet. of ALH 84001, describes the orthopyroxene clasts withn the granu­ Sci., 33, in press. [10] Kring D. et al. (1998) GCA, 62, 2155. la bands ("crush zones" of [1)) as having irregular or undulose [11) Leshin L. et al. (1998) GCA, 62, 3. [12] Waen P. (1998) JGR, extinction, pervaive rcrofacturing, and polygonization. These 103, 16759. [13] McSween and Harvey (1998) Meteoritics & observations, coupled with the presence of makelynite ad a glassy Planet. Sci., 33, A103. [14) Treima A. (1995) Meteoritics, 30, 294. matx, lead Treiman [5] to propse that ALH 84001 undetent [15] Hutchins K. ad Jakosky B. (1997) GRL, 24, 819. [16) McKa shock levels equivalent to S5 or S6 of Stofer et al. [7]. Their clas­ G. ad Lofgren G. (1977) LPS XXVIII, 921. [17] Thomas-Keprta sifcation scheme, however, requires the analysis of olivine gains K. et al. (1998) LPS XXIX, Abstact #1495. [18] Bradley J. et al. rather than orthopyroxene grains, due to the wide-ranging nature (1997) Nature, 390, 454. [19] Bradley J. et al. (1996) GCA, 60, of shock defrmation in orthopyroxene [7]. The presence of melt 5149-5155. [20] Thomas-Keprta K. et al. (1997) LPS XXVlll, veins, pckets, and dikes must occur, with high-shock indicators 1433. [21] Brearley A. J. (1998) LPS XXIX, Abstract #1757. in olivine ad maskelynite to warat a classifcation of S5 and S6 [22] Blake D. et al. (1998) LPS XXIX,Abstract #1347. [23) Grady materials [7]. As fther illustated by fction melting experiments M. M. et al. (1994) Meteoritics, 29, 469. [24] Clemett S. et al. in meteortes, the presence of veins of melted materials alone can­ (1998) Faraday Disc., 109, in press. [25] Flynn G. et al. (1998) not be used as a unique indicator of a high-shock process [8]. Un­ Meteoritics & Planet. Sci., 33, A50. [26] Thoma K. et al. (1995) frtunately, the presence of olivine in ALH 84001 is limited ad LPS XXVI, 149. [27] Stepha T. et al. (1988) Meteoritics & Planet. its orgin unconstained [e.g., 9,10]. Consequently, the clasifca­ Sci., 33, A149. [28) Becker L. et al. (1997) GCA, 61, 454. [29] Bell tion scheme of [7] is not directly applicable. J. (1996) Science, 274, 2121. [30] Anders E. (1996) Science, 274, Minerals show even less evidence fr shock outside the crsh 2119. [31] Sephton M. and Gilmour I. (1988) Meteoritics & Planet. zones. Scott et al. [3] observe only minimal mosaicism of the py­ Sci., 33, A142. [32] Juli A. et al. (1998) Science, 279, 366. roxenes outside the zones. As a result, they interpret shock pres­ [33] Bada J. et al. (1998) Science, 279, 362. [34] Sears D. ad Kral sures outside the crush zones to b small, whereas shock pressures T. (1998) Meteoritics & Planet. Sci., 33, in press. [35] Swartz M. of greater than -50 GPa are thought to be required fr the melting (1996) Texas Monthly, November. [36] Schopf J. (1993) Science, and mobilization of silica and plagioclase melts [11] within the 260, 640. [37] Peirng W. and Odler I. (1990) Proc. 12th. Intl. crush zones [3]. It is suggested that these shock pressures would Conf Cement Micros., pp. 382-402. [38] Mailog J. et al. (1997) be lower if the taget rock were warmer [e.g., 3,11,12]. However, Science, 276, 1775-1776. [39] Folk R. L. (1997) Science, 274, this scenario still requires extreme pressure gradients across only 1287; Science, 276, 1776. [40] McKay D. S. et al. (1997) Nature, millimeters of section from the pyroxenes outside of the crsh zones 390, 455. [41] Thomas-Keprta K. et al. (1998) LPS XXIX, Abstract to those within it. Besides the presence of plagioclase and silica #1495. [42] Bradley J. et al. (1998) Meteoritics & Planet. Sci., 33, melts, there is no other evidence fr high shock levels within the in press. [43] Bradley J. et al. (1997) Meteoritics & Planet. Sci., crsh zones [3]. 32, A20. [4] Friedmann E. ad Weed R. (1987) Science, 236, 703. In addition, no high-pressure polymorphs have been identifed [45] Steele I. et a. (1998) Meteoritics & Planet. Sci., 33, A149. in ALH 8401; all evidence for high shock pressures is textural. [46] Burckle L. ad Delaney J. (1998) Meteoritics & Planet. Sci., In Shergotty, the presence of glasses with radiating factures may 33, A26. [47) McKay D. et al. (1996) Science, 274, 2123. indicate rapid quenching fom high-pressure melts and expasion [48] Steele A. et al. (1998) J. Microsc., 189, 2. upn decompression [13]. Maskelynite in ALH 84001 and is also normal rather tha diaplectic glas [e.g., 14,15]. Shock levels ex­ ceeding 80 GPa ae infrred fr these typs of glasses [13]. Rama HIGH STRAIN-RA TE DEFORATION AND FRICTION spcta of martian "makelynites" in ALH 841 are similar to those MELTING AS A POSSIBLE ORIGIN FOR "SHOCK" that have been experimentally subjected to only 31 GPa [15]. These FEATURES I ALLAN HILS 8001. C. H. van der Bogert obsesatons require either that the glass was totally melted at a and P. H. Schultz, Department of Geological Sciences, Brown higher shock level than that fr the frmation of diaplectic glass, University, Box 1846, Providence RI 02912, USA. or the melting event included temperatures high enough to melt the plagioclase grains in situ. Summary: May of the debates surounding maan meteorite An Alternative: The presence of glass and defrmation tex­ ALH 8401 address the questions of the temperature ad age of tures, yet ambiguous evidence fr shock, may indicate that an al­ carbonate depsition [e.g., 1-3] and its ultimate efect on the gen- terate process controlled the petrologic evolution of ALH 84001. U'l Contribution No. 956 57

Treiman [5] suggested that the crsh zones could be generated dur­ which could explain the random orientation of crush zones in ing high stain-rate defration or shock associated with either a ALH 8401 while allowing fr c origin other thc shock. Further­ tectonic or impact event on Mas. Here we suggest that the crsh more, high stain-rate experiments on meteorites suggest that non­ zones and even melts cc develop fom high strain-rate defra­ shock-related melting cc occur during impacts [8]. High-stain-rate tion without high shock levels. Our scenario is based on the evi­ processes with potentially low shock levels that occur during im­ dence fr pervasive shear in ALH 84001, while evidence for shock pact include superfault fration during the modifcation stage of is minimal. For exbple, Treimc [5] describes "augen or ribbon" crater fration [22] cd impact-directed material flow during ob­ structures of orthopyroxene within the crush zone, with chromite lique impacts [23,24]. stringers, which ae "wavy or swirled." Small (< 10 µm) Fe sulfdes Concusions: The psaive high stain-rate metborphic tex­ are aso present in elongated bands within the crush zone [16]. Oli­ tures similar to those in a terestial ultramafc pseudotachylite, vine inclusions ae elongate cd "boudinage-like" [10]. Such tex­ lack of conclusive shock classification, cd the absence of high­ tures ae indicators of shea defration [e.g., 17,18]. In additon, pressure polymorphs lec us to propose thb the "shock" features the crsh zones themselves have been descrbed as cataclastic, with in ALH 8401 may instead reflect high stain-rate defrmation cd larger clats having rounded edges cd smaler clats having sharper fiction melting with minimal shock. High temperatures created edges [3]. These observations, along with evidence for recrystali­ during deformation cc account fr recrystallizaton textures, melt­ zaton suggest that these crush zones ae pseudotachylites or fc­ ing plagioclase grains in situ, cd generating plagioclase, silica, cd tion melts, a type of high strain-rate rock typically associated with carbonate melts. The friction meltng proess is also c egectve ultracataclaites [e.g., 19]. In fact, Treimc [5] notes that the crush mechanism fr disprsing melts cd blebs of metals cd metal com­ zones could texty "be called recrystallized mylonite or pseudo­ pounds along shea zones [8]. This process could take place in a tachylite." Further, "the stuctures of the grcula bcds ae typi­ tectonic settng simila to the Balmuccia peridotite, northes Italy, cal of dynamic recrystallizaton (high strain rates), which can occur or during the formation of pseudotachylite bodies like the Sudbury during rapid tectonic motions or during shock" [5]. Scott et al. [3] cd Vredefort pseudotachylites. The defration and melting aso note that the "crsh zones were probaly heated cd welded by fc­ may have occured during the event that ejected the meteorite fom tion during crushing, as in pseudotachylites." Such comments, Mas. This scenario does not exclude the presence of shock fea­ coupled with the absence of shock features, support a high stain tures within the pseudotachylites; rather, it does not necessitate their rate rather thc shock origin. presence. The best terestrial calog fr ALH 84001 is a terestial ultra­ The asence of high-pressure shock fatures requires that ejec­ mafc pseudotachylite, which occurs in the Balmuccia pridotite, tion occured at low shock levels. Also, if the plagioclase glass Ivrea-Verbano zone, northes Italy.Textues in the pseudotachylites frmed prior to ejection, temperatures during ejection could not of the Ivrea-Verbano zone closely resemble the textures observed have exceeded those that alow the recrystallizaton of plagioclase in ALH 84001. These pseudotachylites contain several distinctive glass (e.g., 900°C fr 1 hr [25]). features: dynamic recrystalization; chaotic cataclastic to mylonitic In agreement with previous studies (3,5], the crush zones ae textures; olivine cd pyroxene with undulose extncton cd densely likely impact related. Here, we consider the possibility that the tex­ spaced laminar kink bands; cd hosblende locally replacing py­ tures presese a high stain-rate process that allows both localized roxene, which shows undulatory extinction [20]. The pseudo­ high temperatures and low shock levels. For example, the crsh tachylite zones occur in two types: (1) fault vein typ, which is zones cd melt veins may be related to the launch process or pre­ parallel to the defrmation plane, cd (2) injection vein typ, which existing pseudotachylite bodies sampled by the event. is anastomosing and cc form networks of veins [20]. The wall References: [1] Mittlefehldt D. W. (1994) Meteoritics, 29, rocks show a high degree of recrystallization, which drops off 214-221. [2] Havey R. P. cd McSween H. Y. Jr. (1996) Nature, sharply away fom the pseudotachylite vein [20]. These observa­ 382, 49-51. [3] Scott E. R. D. et al. (1997) Nature, 387, 377-379. tions, aong with small amounts of interstitial glass [20], closely [4] McKay D. S. et al. (1996) Nature, 273, 924-930. [5] Treiman resemble the ptrography of ALH 8401. Additonally, the tempra­ A. H. (1995) Meteoritics, 30, 294-302. [6] Treiman A. H. (1998) ture during pseudotbhylite foration exceeded 130 K due to fic­ LPS XXIX, Abstract #1195. [7] Stofer D. et al. (1991) GCA, 55, tional heatng. This was adequate to almost completely melt the 3845-3867. [8] van der Bogert C. H. et a. (1998) LPS XXIX, Ab­ minerals within the shea zone [20]. Such ptographic fatures are stact #1693.[9] Havey R. P. and Mcsween H. Y. Jr. (1994) Me­ quite simila to those selected criteria used to defne a shock state teoritics, 29, 472. [10) Sheaer C. K. and Adcock C. T. (1998) LPS fr ALH 84001, and further suggest that ALH 8401 underwent XIX, Abstact #1281. [11) Stofer D. et al. (1988) in Meteorites high stain-rate defrmation cd fricton melting, rather than severe and the Early Solar System (J. F. Kerridge cd M. S. Matthews, shock. eds.), pp. 165-202, Univ. of Arizona, Tucson. [12) Schmitt R. T. Treimc [5] proposes that the textures in ALH 84001 are likely et al. (1994) Meteoritics, 29, 529-530. [13) El Goresy A. et al. the result of a shock event rather than a tectonic event on Mars, (1997) Meteoritics & Planet Sci., 32, A38-A39. [14] Treimc A. H. because the random orientations of the crsh zones ae not typical cd Treado P. (1998) LPS XI, Abstract #1196. (15) Cooney T. F. of tectonic fabrics and there ae more impact thc tectonic features et al. (1998) LPS XIX, Abstract #1332. [16) Foley C. N. et al. on Mas. Lage-scale friction melt bodies, however, are associated (1998) LPS XXIX, Abstact #1928. [17) Spry A. (1969) Metamor­ with lage craters on Eath such as the Sudbury structure, Ontao, phic Textures, 350 pp. [18) Boradaile G. J. et al. (1982) Atlas of and the Vredefort structure, South Afica (e.g., 21]. These Deforational and Metamorhic Rock Fabrics, 551 pp. [19) Sibson pseudotachylite bodies, athough impact related, do not necessa­ R.H. (1977) J. Geol. Soc. London, 133, 191-213. [20) Obata M. ily show evidence fr shock. They have anastomosing zones cd cd Kaato S.-1. (1995) Tctonophysics, 242, 313-328. [21) Spray veinlets simila to those described in the Balmuccia peridotite, J. G. and Thompson L. M. (1995) Nature, 373, 130-132. 58 Workhop on Martian Meteorites

[22] Spray J. G. (1997) Geology, 25, 579-582. [23] Schultz P. H. that the main ions in the solution are Mg2+, CO?-, and SO/-, then (1996) GSA Abstr. with Progr., A384. [24] Schultz P. H. (1996) by the time the solution could be evaporatively concentrated to the

JGR, 101, 21117-21136. [25] Mikouchi T. et al. (1998) Antarct. point of MgSO4 x 7H2O saturation, the solubulity of MgCO3 would Meteorites XXlll, 77-79. be diminished by a fctor of 400, compared to its solubulity in the absence of SOl- (Fig. 1). The actual solution chemist would not be so simple, but given TH COMON ION EFFECT IN DEPOSITION OF MAR­ the magnitude of the common ion efect, potentially operating in TIAN (e.g., AL LAN HILS 84001) CARONATES. P. H. conjunction with the water-level drawdown ad pore occlusion Waen, Institute of Geophysics ad Planeta Physics, University efects, it hardly seems surprising that the evaporitic caonates of Califra, Los Angeles CA 90095-1567, USA (pwaen@ucla. in ALH 8401 are unaccompanied by signifcat sulfates. edu). In general, evaporitic martian sulfates probably formed only fom extemely concentated waters, whereas evaporitic cabonates The most plausible model fr origin of the cabonates in the probably frmed as more diguse depsits, filling pores in rocks ALH 8401 meteorite involves depositon fom a playa lake or zone (e.g., ALH 84001) and the megaegolith. Extremely concentrated of groundwater that underwent evaporative concentation at or nea waters, and thus sulfates, probably developed mainly in places the surface of Mas [1,2]. A key constraint fr such a model is the where foods (surface ad groundwater) terminated in playas. In virtual absence of sulfate in ALH 84001. In a simple closed­ contrast, carbonates developed as diffuse deposits in the mega­ system evaporation sequence, abundant sulfate would be expected regolith along the routes fllowed by the waters on their way to to frm fom any plausible martian lake or shallow groundwater. the playas. Evaporitic sulftes are notoriously fagile ad tend to However, several factors would potentially alter the course of the easily disintegrae upon exposure to eolian erosion. This model evaporation-deposition sequence enough to engender extemes of helps explain the extemely high abundance of SO3 (-6 wt%) in sulftecarbonae factionation, including the extemely low sulfte/ the surface global regolith [3] ad the virtual absence of cabonate cabonate ratio of ALH 84001. Assuming the evaporating water signals among spectal measurements of the maa surface. body was tansient in nature (i.e., the product of a flood), the wa­ References: [l] Warren P. H. (1998) JGR, 103, 16759- ter level could easily have been receding during the evaporation­ 16773. [2] Mcsween H. Y. Jr. and Havey R. P. (1998) Meteor­ deposition process, and we need only assume that carbonate itics & Planet. Sci., 33, A103. [3] Rieder R. et al. (1997) Science, deposition, but not sulfte deposition, occurred befre the water 278, 1771-1774. level receded below the level where ALH 8401 was perched [l]. Occlusion of pores by incipient sulfate precipitation, in the outer fnges of the rok that ultately became ALH 801 ad/or in over­ PLANETARY CORE FORMATION: EVIENCE FROM lying solid materials, might also have reduced the yield of sulfate HIGHLY SIDEROPHILE ELEMENTS IN MARTIAN depsition in ALH 8401 [1,2]. Another factor that probably played METEORITES. P. H. Warren, G. W. Kallemeyn, and F. T. a key role is the common ion efect. Kyte, Institute of Geophysics and Planetary Physics, University When two salts sharing a common ion ae present in a aque­ of Califrnia, Los Angeles CA 90095-1567, USA (pwarren@ ous solution, the dissolved ions of the more soluble salt act to ucla.edu). substantially decrease the solubility of the less soluble salt. For ap­ plication to Mars and ALH 84001, relevant sulftes have much We present new bulk compositional data fr six martian mete­ greater solubilites than relevat cabonates. Assuming, fr example, orites, including the highly siderophile elements Ni, Re, Os, Ir, and Au. These and literature data are compared with the sidero­ phile systematics of igneous rocks fom Earth, the Moon, ad the HED asteroid. The composition of ALH 84001 is anomalously 0.01 siderophile-poor. Whether this refects a more reducing environ­ ment on primordial Mas when this ancient rock frst crystalized or secondary alteration is unclea. Queen Alexadra Range 94201 i 0.001, shows remakable similarity with EET 79001-B fr siderophile a well as lithophile elements; both ae extraordinarily depleted in the (J "noblest" siderophiles (Os ad Ir), to roughly 0.0000lx CI chon­ a, ::!; 0.000100 drites. As in terrestrial igneous rocks, martia rocks display stong 0 corelations of Ni, Os, and Ir vs. MgO. In the case of MgO vs. Ni, l the martian tend is displaced toward lower Ni by a large factor

� 0.00001 (5), but the Os and Ir trends ae not signifcatly displaced fom rn their terestral counterparts. For Mars, Re shows a rough corela­

t common-ion e tion with MgO, indicating compatible behavior, in contrast to its 0.000001 -t-���...-��...... --� ...... --�...... t mildly incompatible behavior on Earth. Among martan MgO-rich 0.001 0.010 0.100 1.000 10.000 roks, Au shows a weak anticorrelaton vs. MgO, resembling the MgS04 · 7H,O in solution (mol/L) terrestal distrbution except fr a displacement toward 2-3x lower Au. The sbe elements (Ni, Re, Os, Ir, and Au) show similar cor­

Fig. 1. Common ion effect on solubility of MgCO3 in presence of relations with Cr substituted fr MgO. Data fr lunar ad HED dissolved sulfate. rocks generally show less clear-cut tends (relatively fw MgO-rich LP! Contribution No. 956 59

terrestrial mantleis probably aneffect traceable to Earth's size. For

♦ the more highly siderophile elements like Os and Ir, the simplest ♦ • model consistent with available constraints is the veneer hypoth­ esis. Core-mantledifferentiation was inefficienton the largestter­ restrial planets, because during the later stages of accretion these ♦ + bodiesacquired sufficientH 2O to oxidize most of the later-accreting ♦ ♦ i ++ * Fe metal, thus eliminating the carrier phase forsegregation of sid­ D.1 • ♦ .s ♦ erophile elements into the core. 1/) • ♦ ♦ References: Warren and Kallemeyn (1998) GCA, submitted. + • • ... . 0.01 + ♦ At• ... . ALTERATION PRODUCTS AND SECONDARYMINERALS IN MARTIAN METEORITE ALLAN HILLS 84001. S. J. O.II01 " Wentworth!, K. L. Thomas-Keprta1, and D.S. McKay2, 1Mail Code 10 C23, Lockheed Martin, 2400 NASA Road I, Houston TX 77058, • () ■ • "' ♦ USA ([email protected]), 2Mail Code SN2, NASA ♦ ♦t.. i.t ♦ A ♦ ♦• ♦ • ♦ Johnson Space Center, Houston TX 77058, USA. ♦ �•4J••• :. ♦ ■ • + ♦ .♦ • ♦♦ � � ♦ ++ . < ♦ • . . • ♦ ♦ The martianmeteorites contain alteration products and second­ ♦ ♦ ♦ .. I + + ary minerals that are a critical part of understanding their near­ + ,.... + + ♦ • surface histories on both Mars and Earth. In some martianmeteor­ 0.1 ...-: t( iC • ites, suspected martian (preterrestrial)alteration products can bedis­ - ·, • A1 ♦ f:. ,. ...• -= + •• tinguished from terrestrial weathering effects [e.g., 1-5]. Using "" . . scanning electron microscopy (SEM), field emission SEM (FE­ + .... OJlt ■• • SEM), transmission electron microscopy (TEM), and energy­ + . ,.. . dispersive X-ray analysis (EDS), we are studying natural fracture . .. surfacesof ALH 84001chips, including samplesfrom both the in­ .+ ♦ ·* Moon • [-♦♦ J terior and the exterior of the meteorite. Exterior samples include O.IIOt fusion crust surfaces, which are important in determining the ex­ 10 21) 30 «> = 50 MgO (w) tent of terrestrialweathering of meteorites. The focus of this study is weathering features and secondary minerals other than the dis­ Fig. 1. MgO vs. Os and Ir for igneous rocksfrom Mars, Earth, and the tinctive carbonateglobules that continueto be studied by many re­ Moon (mare basalts). "A" is main lithology of EET 79001, possibly searchers. meteorite contaminated. Lunar star point is 15426 VLT green glass Both the interior and the fusion crust of ALH 84001 generally (literatlll'e data). show very little evidence of terrestrial weathering. In [1,6], we re­ ported the presence of tracesof typical Antarctic secondary miner­ als (Ca sulfate, Mg sulfate, NaCl, and SiO2) on the fusion crust surface. The absence of these terrestrial minerals in thechips be­ samples areavailable). These trends are exploited to inferthe com­ low the fusion crust surfaceis a good indicator of the lack of sig­ positions of the primitive Earth, Mars, Moon, and HED mantles, nificant Antarctic alteration. We have now found one small by assuming that the trend intercepts the bulk MgO or Cr content occurrenceof Ca sulfate slightly (-200µm) below the fusion crust of the primitive mantle at the approximate primitive mantle con­ surface (Fig. la). It is most likely of terrestrialorigin, since no Ca centration of the siderophile element. Results for Earth show good sulfate has been identified in interior samples. Another weathering agreement with earlier estimates. For Mars, the implied primitive feature not previously described is shown in Fig. lb. It consists of mantlecomposition is remarkablysimilar to the Earth's, except for amorphous-looking C-rich material (not carbonate)partly filling a 5x lower Ni. The best constrained of the extremely siderophile el­ vug in the fusion crust. Traces of this C-rich material are found ements, Os and Ir, arepresent in the martian mantle at 0.5x CI, in both on and near the fusion crust (e.g., Fig. la). More work is comparison to 0.7x CI in Earth's mantle. This similarity constitutes needed to determine if it is identifiablein interior samples with the a key constraint on the style of core-mantle differentiation in both SEM. Similar C-rich material was found on the fusion crust of Mars and Earth. Successful models should predict similarly high Chassigny [2]. concentrations of noble siderophile elements in both the martian Evidence of alteration in the interior of ALH 84001 consists and terrestrial mantles ("high" compared to the lunar and HED mostly of pitted carbonate globule surfaces, etched silicates (mostly mantles and to models of simple partitioning at typical low­ pyroxene), and tracesof smectite-type clay associated with pyrox­ pressure magmatic temperatures), but only predict high Ni forthe ene [1,6]. In addition to the carbonate globules, the interior con­ Earth's mantle.Models that engender the noble siderophile excess tains traces of another type of Mg carbonate, as reported in [6]. in Earth's mantle through a uniquely terrestrial process, such as a This Mg carbonate (Fig. le) has a distinct blady morphology that Moon-forminggiant impact, have difficulty explaining the similarity is differentfrom that of typical ALH 84001 carbonate globules, and of outcome (except for Ni) on Mars. The high Ni content of the it also has a differentcomposition (higher 0, no minor elements). 60 Workshop on Martian Meteorites

Fig. 1.

It is probably secondary to the globule carbonate. This O-rich ca­ It is clear that the interior of the ALH 84001 meteorite has in­ bonate is fund in the interior of the meteorite cd within a few teracted with aqueous fuids, but the fluids ae not obviously the hundred micrometers of the fsion crst surface, but it has not been sbe as those that deposited salts on the fsion crst. Further stud­ fund on the fsion crust itself. In [1] we reprted that Fe sulftes ies of these typs of fatures may alow us to determine where they were present in the interior of ALH 84001. It is not clear, how­ originated cd how they formed. It may also be possible to use ever, that they ae actually sulftes because the amount of O pre­ mineralogical cd textural evidence of the degree of terrestrial sent is not well defined; TEM and digraction work still need to weathering to help determine the amount of terestral biological be done. Exbples of these Fe-,S-rich occurences are shown in contamination. Distinctions between terestrial and preterrestrial Figs. ld-f. Figure Id is a backscattered electron image of a typical weathering efects should prove to be usefl in determining the ori­ occurence (the brightest patches with fne-grained grcula to f­ gins of possible biogenic fatures [7] in the martian meteorites. brous textures ae the Fe-,S-rich phase, the smoother medium gray References: [1] Wentworth and Gooding (1995) LPS XXVI, grains ae chromites, cd the dak gray masses ae silicates). The 1489. [2] Wentworth cd Gooding (1994) Meteoritics, 29, 860. Fe-,S-rich material is commonly found in clusters closely bsoci­ [3] Treiman et al. (1993) Meteoritics, 28, 86. [4] Gooding et al. ated with chromite grains. The equct occurrence in Fig. le appeas (1991) Meteoritics, 26, 135. [5] Gooding et al. (1988) GCA, 52, to be a partly weathered equant pyrte grain. EDS indicates that the 909. [6] Wentworth et al. (1998) LPS XXIX, Abstract #1793. proportion of S to Fe decreases with increbed alteration (Fig. If). [7] McKay et al. (1996) Science, 27, 924. LI Contribution No. 956 61

TERRESTRIAL BIOMARKERS FOR EARLY LIFE ON tial energy source, since biofilms trap nutrients and carbonaceous EARTH AS ANALOGS FOR POSSIBLE MARTIAN LIFE material. Biofilmsare invariably present in modem-day, naturally FORMS: EXAMPLES OF MINERALLY REPLACED occurring bacterial colonies. They are as much representative of BACTERIA AND BIOFILMS FROM THE 3.5-3.3-Ga bacterial presence as the bacterial bodies themselves, since they are BARBERTONGRE ENSTONE BELT, SOUTH AFRICA. F. produced by bacteria. Both bacteria and biofilms are mineralized Westall1, D. S. McKay 1, E. K. Gibson1, M. J. de Wit2, J. Dann2, by the chelation of mineral ions to the active groups in the poly­ D. Gemeke3, and C. E. J. de Ronde4, IMail Code SN, Earth saccharide-rich polymers that make up the EPS [3]. In this case, Sciences and Solar System Exploration Division, NASA Johnson the bacteria andbiofilms have been silicified. Stromatolitesare good Space Center, Houston TX 77058, USA (frances. westall l@ examples of mineralized biofilms: In fact,these examples fromthe jsc.nasa.gov), 2Department of Geological Sciences, University of BGB can be termed tabular stromatolites. Biofilrns are also sinks Cape Town, Rondebosch, South Africa, 3EM Unit, University of for heavy metals andmay be implicated in the concentrationof pre­ CapeTown, Rondebosch, South Africa,4 Institute of Geological and cious metals in economically significant deposits.The association Nuclear Sciences, P.O. Box 31-312, Gracefield, Lower Hutt, New of heavy minerals with filmlike structures in an unknown sample Zealand. may therefore be a biomarker for the presence of lifeforms. Bacteria. Numerous spherical and rod-shaped bacteriomorph Introduction: Thesearch for extraterrestrial life and especially structures and bacteriomorph molds, ranging in size from 0.6 to martian life hinges on a variety of methods used to identify ves­ 4 µm, are embedded in the quartz matrix and in smooth­ tiges of what we could recognize as life, including chemical signa­ surfaced biofilm. They have been interpreted as a fossil bacteria tures, morphological fossils, andbiogenic precipitates.Although the followingwell-defined criteria [4]: possibility of extantlife on Mars (subsurface) is being considered, 1. Their sizes andshapes are similar to those of modem bacte­ most explorationefforts may be directed towardthe searchfor fossil ria, including rod-shaped (bacillus or coccobacillus) and spherical life. Geomorphological evidence points to a warmer and wetter morphologies (coccoid), and range from <0.2 µm (nanobacteria) Mars early on in its history [1], a scenario that encourages com­ to tens of micrometers (cyanobacteria) in size. parison with the early Earth. For this reason,study of the earlyter­ 2. Cellular complexity such as cell division is a fundamental restriallife forms and environment in which they lived may provide bacterial characteristic and is manifested by our microfossils as clues as to how to searchfor extinct martian life. simple pairings to multiple linear or pluridimensional cell divisions. As a contribution to the early Archean database of terrestrial 3. The cell wall texture of bacteria varies.dependingon the os­ microfossils, we present new data on morphological fossils from motic gradient between an organismand its external environment: the 3.5-3.3-Ga Barberton greenstone belt (BGB), South Africa[2]. In perfect osmostic conditions, the cell is turgid and the cell wall This study underlines the variety of fossil types already present in smooth, whereas where the gradient is steep, the outer wall may some of the oldest, best-preserved terrestrial sediments, ranging be puckered and wrinkled. Both situations are preserved in our fromminerally replaced bacteriaand bacteria molds of various mor­ bacteriomorphs. phologies (coccoid, coccobacillus, bacillus) to minerally replaced 4. The association of large numbers of cells is typical of bacte­ biofilm. Biofilmor extracellular polymeric substance (EPS) is pro­ rial colonies. We observe large numbers of microfossils as clus­ duced by bacteria and appears to be more readily fossilisable than ters as well as extending in matlike sheets over bedding planes. bacteria themselves. The BGB fossilsoccur in shallow water to sub­ S. Bacterial colonies usually consist of a coexisting variety of aerial sediments interbedded with volcanic lavas, the whole being species (consoria). Some of the colonies of bacteriomorphs from deposited on oceanic crust. Penecontemporaneous silicification of the BGB cherts contain a number of different morphotypes, e.g., sediments and volcanics resulted in the chertification of the rocks, coccoid, rod-shaped, andlinear associationsof coccoid forms.Each which were later subjected to low-grade metamorphism (lower species, however, presents only a limited size and shape range. greenschist). The association of the microfossils with fossil biofilm, as well Results and Discussion: Petrographic and SEM observation as the -27%0 13C value from these sediments, supports the inter­ of rock chips and thin sections of finely laminated cherts revealed pretation of the bacteriomorphs as bacteria. (1) smooth to granular films (fossil biofilm) and strands coat­ The above criteria are sufficient to distinguish minerally re­ ing bedding-plane surfaces and embedded within the quartz, and placed bacteria from small crystals. Although many of the (2) bacterimorph structures (fossil bacteria). A fuller understand­ bacteriomorphs presented rounded cross sections, some of them had ing of both of these biogenic features in Archean terrestrial rocks slightly crystalline terminations. However, observations of micro­ will help us document similarfeatures, if present, in martianrocks fossils with one end coated by a wrinkled surface and the other by and meteorites. a smooth surface coated with crystalline termination demonstrate Biojilm. Thin sections of the finelylaminated cherts arechar­ that mineral replacement of bacteria can, in some cases, lead to acterized by fine, wavy, discontinuous, brown-colored texturing of subeuhedral crystalline structures.This has also been demonstrated the order of 50-100 µm, parallel to the bedding. These laminae in the laboratory [5]. In the case of the BGB fossils, the morpho­ consist of filmsthat irregularly coat bedding planes, sometimes ex­ types with slightly mineral terminations occur together with rounded tending over thousands of square micrometers. The films are char­ morphotypes and satisfyall the other biogenic criteria. acterized by either a smooth or a ropy, twisted, interwoven Conclusions: Our data show that minerally replaced bacteria ultrastructure,although they may also be granular. and biofilm can be useful morphofossils. The fossil types in the We interpret these filmsas representing the EPS slime or biofilm BGB cherts, together with fossilbacteria and biofilm fromthe (con­ in which bacteria live and grow. Biofilms provide protectionagainst tiguous) early Archean Warrawoona sequence in northwest Aus­ a hostile outer environment and also provide a nutrient and poten- tralia [6] show that bacteria were well established on Earth by that 62 Workhop on Martian Meteorites

Mars, and the study of early Archean samples provides important insights into those processes. Acknowledgments: F. Westallacknowledges the NRCfor fi­ nancialsupport. L. Hulse is warmly thanked forhis SEM help. References: [1] Head J. W. et al. (1998) Meteoritics & Planet. Sci., 33, A66. [2] Westall F. et al. (1998) Precambrian Res., in press. [3] Westall F. et al. (1995) Palaeontology, 38, 495-528. [4] Schopf J. W. andWalter M. R. (1983) in Earth's Earliest Biosphere (J. W. Schopf, ed.), pp. 214-239, PrincetonUniv. [5] Castanier S. et al. (1989) Bull. Soc. Geol. Frances. 8V3, 589-595. [6] Morris P. et al. (1998) LPS XXIX.

ABIOTIC SYNTHESIS OF HYDROCARBONS ON MARS: Fig. 1. Colony of a consortum of rod-shaped and coccoid fossil bacteria THEORETICAL MODELING OF METASTABLE EQUI­ 1 2 1 fom the BGB. LIBRIA. M. Yu. Zolotov , and E. L. Shockl, Department of Earthand Planetary Sciences, WashingtonUniversity, St. Louis MO 63130, USA ([email protected]; [email protected]. edu), 2VernadskyInstitute Russian Academy of Sciences, Kosygin Street 19, Moscow 117975, Russia.

Introduction: The martian meteorite ALH 84001 contains polycyclic aromatichydrocarbons (PAHs) and other hydrocarbons of endogenous origin [1-4]. McKay et al. [1] proposed that the presence of PAHs in carbonate globules of ALH 84001 indicates biological activity on ancient Mars. On the other hand, the pres­ ence of hydrocarbons in other meteorites, IDPs, interstellarclouds, and Earth's pristine and altered igneous rocks (e.g., serpentinites) is not attributed to biological sources. In many cases these hydro­ carbons areformed abiotically due to reduction of C oxides by H. We used thermodynamic calculations of metastable equilibria to evaluate the possibility of abiotic synthesis of PAHs and n-alkanes Fig. 2. Two silicified, rod-shaped, dividing bacteria (arrows) with in possible hydrothermal fluidsand thermal gases on Mars. slightly cryste terminaton on the lower individual and wrinkled sur­ Model: Hydrocarbons are thermodynamically unstable com­ face on the upper one. pounds and may form if the formation of stable methaneor graph­ ite is inhibited [e.g., 5]. On Mars and Earth, this inhibition could have happened in dynamic cooling systems: volcanic/impact gases, fumaroles, and hydrothermal systems around magmatic intrusions or hot impact craters. We considered metastable equilibria among condensed P AHs and n-alkaneswith gaseous and aqueous CO2 and H2, which could have existed in these systems on ancient Mars. In addition, we considered some aromatization andPAH methylation equilibria. We assumedthat the fugacity (f) of Hz(gas) and activ­ ity (a) of H2(aq) are controlled by the quartz-fayalite-magnetite (QFM) or hematite-magnetite (HM) buffers, which represent mar­ tianigneous rocks and oxidized soil/sediments. We used 5.4 mbar and1 bar forf correspondingto the present and a possible an­ co2, cient atmosphere. In the terrestrial analogy, for carbonate-bearing hydrothermal systems we used f values that are governed by co2 calcite-silicate equilibria (4-7]. We assumed the activity of con­ densed hydrocarbonsand water wasin unison. Pressure corresponds to liquid-water vapor saturation. The thermodynamic data for un­ Fig. 3. Smoth, silicifed biofil covering large colony coccoid bacteria. substituted and methylated PAHs and n-alkanes(C numbers 5-20) are from[7] and [8] respectively. period, and,not only do they look very similarto modern bacteria, Results: The calculated activities/fugacitiesof H2 and CO2 for but their intimate association with biofilm is also a fundamental the following metastable equilibria characteristicof modernbacteria. Similar associations in returned martian samples or meteorites aH2 + bCO2 = cHC (PAH or n-alkane) + dH 2O may be the key to documenting early life forms. Deciphering the effects of fossilization, mineralization, and aqueous or metamor­ are plotted in Figs. 1-3 as saturation lines. PAHs and n-alkanes phic alteration will be a vital part of the search forancient life on can be formed if a and f exceed the metastable equilibria values. LP/ Contribution No. 956 63

The presentend an encienl 11 ,00 Teeel ge 101 2 Ss atmosph«e .e -6 - \2 mren ' 0' •L.:1 3 I log10 fCOi '" 0.01681(.°C). 3.781 j 100°c. 1.014 barsj ., -1 2

·2 -2

l -3 f-3 0• I ... : i -4 -4

·5 .5

-6 L..-Ja.llCl..__ ...... {...__,1._...,C....1... __ ..__ _ __, 0 50 100 150 200 250 300 T. !CJ

Fig. 1. Toe metastable equilibrium saturation lines fr PAHs and n­ Fig. 3. The metastable equilibrum hydroarbn saturation lines at the ° alkanes at 10 C. Te shaded box represents te proposed condiuons in CaC03-silicate equilibria [6,7]. martian thermal system.

I QFM ber I 0 -1

·1 -2

-2

.3 iii .II -4 -4

.5 -5

100 150 200 50 100 150 200 20 300 Tapa !C) T-.,.ature(°C)

Fig. 2. The metastble equilibrium saturation lines fr hydroarbons at Fig. 4. The metastble equilibrium conditons btween pairs of n-alknes the QFM bufer. Dashed lines show CO2 in terestrial [6] and proposed ad PAHs with the same C number as a fncton of log aH2 and tempera­ man geotenl systems. Dotted lines show aC02 equilibrated with the ture. The hexadecane to pyrene curve is indicated, other curves are labled present and possible ancient atosphere of Mars. by the PAHs.

Saturation lines for P AHs and n-alkanes lie close to each other; High temperature and/or oxidizing conditions favor aroma­ therefore, the formation of a metastablemixture is more likely. tization of n-alkanes (Fig. 4). At QFM, aromatizationcan proceed ° At the fH 2 controlledby the QFM buffer, the saturation condi­ above -I 70 C. tions for hydrocarbons match with the f conditions proposed for Calculations of equilibria for methylation reactions for PAHs co2 Mars(Fig. 1). However, the HM buffer is too oxidized to stabilize show that low temperatureand reduced conditions favor alkylation hydrocarbons. of parent compounds. At QFM and at temperatures <-350°C, me­ A decrease in temperaturefacilitates the condensation of hydro­ thylated PAHs are more stable than the unsubstituted species. carbons (Figs. 2 and 3). At the QFM buffer and CO2 governed ei­ Discussion: The predominanceof mafic and ultramaficrocks ther by the CaCO3-bearing systems or atmospheric CO2, PAHs and on Mars would favor high/ H in magmatic gases and high aH2 in n-alkanes can be metastably formedbelow 90°-l80°C. Light PAHs aqueous solutions formed during2 the alteration of these rocks (e.g., can condense together with heavy alkanes. More reduced condi­ serpentinization). High CO2 contents proposed in martian thermal tions and more CO2 raise the condensation temperatures. In addi­ systems favor the synthesis of hydrocarbons and provide high tion, possible high-temperature quenching of H2 (see dotted line (a,f)CO due to high-temperature CO-CO2 equilibrium. In dynamic in Fig. 3) increases the condensation temperatures and facilitates systems, C-0-H gasesmay quench below 300°-700°C and produce

hydrocarbonsynthesis. higher disequilibrium amounts of CO and H2, which favors the 64 Workshop on Martian Meteorites

Fischer-Tropsch (F) type synthesis of hydrocarbons. The F syn­ magmatic or impact heating on Mas could increase the PAH con­ thesis ad the propsed C02-H2 interaction could be catalyzed by tent and drive their dealkylation. magnetite [9,10] in igneous rocks and the martian regolith. Acknowledgments: This work was sponsored by the NSF Conclusion: Thermodynamic analysis of metastable equi­ grant CHE-9714060. libria show that PAHs could be frmed abioticaly together with References: [1] McKay D.S. et a. (1996) Science, 273, 924- n-akanes in dynamic cooling gaseous (volcaic/fmaolic/impact) 930. [2] Stephan T. R. D. et a. (1998) LPS XXIX, Abstract #1263. and aqueous geothermal systems on Mas below -180°C. Reduced [3] Flunn G. J. et al. (1998) LPS XXIX, Abstract #1156. [4] Clem­ compositions, high initial temperature of gases/fuids, and quick ent S. J. et al. (1998) Farady Discussions (Royal Soc. Chem.), 109, quenching fcilitate the synthesis of hydrocarbons. However, the in press. [5] Eck R. V. et al. (1966) Science, 153, 628-633. oxidized martia regolith and Oz-bearing subsurface fuids do not [6] Giggenbach W. F. (1997) GCA, 61, 3763-3785. [7] Richad fvor such synthesis. L. and Helgeson H. C. (1988) GCA, 68, in press. [8] Helgeson The possibility of joint frmation of PAHs ad n-alkaes indi­ H. C. et al. (1998) GCA, 68, 985-1081. [9] Anderson R. B. (1984) cates that these hydrcarbons in ALH 84001 could be frmed in a TheFischer-Tropsch Synthesis. [10] Berdt M. E. et al. (1996) Ge­ single process. Our calculations fr the aromatization and methy­ ology, 24, 351-354. [11) Anders E. (1996) Science, 274, 2119- lation equilibria also supprt the suggestions of Anders [11] that 2120.