CARBONATES IN THE MARTIAN ORTHOPYROXENITE ALH 84001:  EVIDENCE OF FORMATION DURING IMPACT-DRIVEN METASOMATISM.  R. P. Harvey and H. Y. McSween, Jr., Department of Geological Sciences, University of Tennessee, Knoxville TN 37996-1410, USA.

Published in Lunar and Planetary Science XXVI, pp. 555-556, LPI, Houston, TX.

An enigmatic feature of ALH 84001 (A84) is the widespread occurrence of carbonate minerals in this orthopyroxenite member of the “martian” or “SNC” group of meteorites [1]. The presence of these minerals, whether attributed to high- or low-temperature processes, has wide-ranging ramifications for our understanding of Martian crustal fluids [2,3].

Petrography:  Carbonates in A84 occur predominantly in crushed zones, often associated with maskelynite, olivine, and ubiquitous orthopyroxene. Carbonate has apparently grown outward from a few nucleation sites to fill available space and is fine-grained, possibly indicative of rapid crystallization. Concentric growth patterns are recognizable in X-ray maps of larger carbonate regions, filling voids within the rock (Fig. 1). Silica can occasionally be found at the interstices where late magnesite rosette rims have grown together. Both orthopyroxene and maskelynite appear to have served as nucleation sites, suggesting growth controlled by kinetic rather than compositional considerations. Carbonates replace maskelynite in some areas, filling or replacing along crystallographic planes. The preferential siting of carbonate in crushed regions, and the preservation of relatively delicate features in the carbonates suggest formation after the major brecciation, although minor fracture offsets show that the carbonates predate later shock events.

Mineralogy:  Carbonate rosettes show a consistent sequence of compositions, with Ca-rich cores zoning outward to intermediate and Mg-rich compositions, with two distinct reversals during crystallization; one early, with a minor Ca recycling, and a late narrow enrichment of Fe (Fig. 2). This sequence is ubiquitious in the regions examined so far, although individual areas rarely show the complete sequence due to irregular exposure. Rapid Mg, Fe, and Ca diffusion rates in carbonate suggest that the survival of this pattern requires very rapid cooling and little or no subsequent heating or presence of fluids [4].

Figure 2 is a carbonate ternary plot showing over 6000 semi-quantitative EDS analyses from two representative carbonate areas mapped at resolutions between 1 and 2 µm; analytical sensitivity is around 2.5% for Mg and Ca and 1.5% for Fe. Representative quantitative (ZAF corrected) analyses, shown as circles, fall within the observed trend, which contains a full suite of calcite, dolomite, magnesite, siderite, and intermediate compositions. The linear trends between calcite-dolomite and dolomite-ankerite, with a sharp inflection at dolomite, strongly suggest tie-lines between coexisting mineral pairs. These tie-lines correspond will with the three-phase equilibria at 700°C of [5], as initially suggested by [1]. In addition, the major trend from intermediate magnesite-siderite to nearly pure magnesite appears to follow the phase boundary at this temperature, suggesting compositional changes in the depositional fluid rather than a steady drop in temperature. The major deviation from this trend occurs for compositions with more than 20 mol% Fe. However, phase boundaries for this region are quite speculative, as little experimental or natural data are available, and it is not clear whether the deviation from the 700°C trend is a result of compositional or thermal changes [5]. Calculated equilibrium temperatures (using the methods of [5]) for the quantitative compositions shown on Fig. 2 suggest that calcite-dolomite pairs approached equilibrium at temperatures around 680°C.

This temperature is in relatively good agreement with the observed mineralogy of the rock and known phase relationships in the CaO-MgO-FeO-SiO2-H,O-CO: system. In hydrous systems, reactions that produce mafic carbonates also produce other characteristic minerals such as talc, antigorite, and tremolite, phases absent from A84. Only when XCO2 values exceed 0.85 can reactions occur that produce carbonates from silicates without also producing hydrous minerals [6]. Two reactions in particular are of interest:  forsterite + CO2 –> magnesite + enstatite, which occurs with dropping temperature at 540°C (at 2 kbar fluid pressure), and enstatite + CO2 –> magnesite + SiO2, which occurs at 480°C. The presence of embayed olivine and silica suggest both reactions occurred, and in the correct order. Recent studies on metasomatic carbonates show that kinetics favor the production of a wide variety of carbonate compositions with very minor changes in the composition of a Mg- and Fe-rich fluid [7].

Trace elements and isotopic chemistry:  Further evidence for reactions involving CO2-rich fluids may come from the REE pattern of A84 carbonates. SIMS analyses (by B. Paterson and L. Riciputi at ORNL) of 10-µm sized regions within carbonate rosettes reveal REE contents somewhat enriched over whole rock values. Rosette cores show higher overall REE abundances than rims, and the pattern is HREE-enriched, like that seen in SNC pyroxenes. A large positive Eu anomaly is present that may be due to dissolution of maskelynite. Previous studies suggest that CO2-rich fluids are efficient carriers of REE’s, and particularly the HREE’s, through preferential complexing with CO3-2 [8].

Although C and O isotopes have been used by some authors to support lowtemperature (<100°C) production of the carbonates, fractionation between CO2 and carbonates can produce the observed 13C-enrichments in precipitates formed at temperatures above 200°C, and do not require O-isotopic equilibrium between reacting silicates and CO2-rich fluids [3]. These findings are supported by studied of carbonates formed during terrestrial impacts, which release significant CO2 and show significant 13C-enrichments and minor O fractionation [9].

Origins:  The lack of hydrous minerals and the 3-phase assemblage shown by A84 carbonates are the strongest evidence for a high-temperature origin in the near absence of water, while fine-scale major and trace element zoning patterns, poor nucleation, and small grain size suggest rapid formation. Impact-driven metasomatism provides an appropriate timescale for formation as well as conduits for fluid flow [9]. Surface frosts on Mars are known to be composed of nearly pure CO2 [10]; volatilization of these or pre-existing crustal carbonates could have produced a metasomatic fluid of appropriate composition in the hours following the impact. Formation of A84’s carbonates during a short period when hot, CO2-rich fluids percolated through a plutonic orthopyroxenite (or lherzolite) brecciated by impact is a plausible origin for these phases.

References:  [1] Mittlefehldt D. W. (1994) Meteoritics, 29, 214-221. [2] Harvey R. P. and McSween H. Y. Jr. (1994) Meteoritics, 29, 472. [3] Romanek C. S. et al. (1994) submitted to Nature. [4] Pingitore N. E. Jr. (1982) J. Sed. Petrol., 52, 27-39. [5] Anovitz L. M. and Essene E. J. (1987) J. Petrol., 28, 389-414. [6] Trommsdorff V. and Connolly J. A. D. (1990) Contrib. Mineral. Petrol., 104, 1-7. [7] Woods T. L. and Garrels R. M. (1992) Geochem. Cosmochem. Acta, 56, 3031-3043. [8] Lee J. H. and Byrne R. H. (1993) Geochem. Cosmochem. Acta, 57, 295-302. [9] Martinez I. et al. (1994) EPSL, 121, 559-574. [10] James P. B. et al. (1992) in Mars, U. Arizona Press, 934-968.