Highly Siderophile Element Abundances in Martian Meteorites

Critical evaluation of new and literature data for highly siderophile elements (HSE) in Martian (SNC) meteorites allows several first order conclusions to be drawn. (i) Re concentrations in SNC meteorites are nearly constant (within a factor of two) and do not correlate with rock type. Exceptions to this rule are Chassigny and Dar al Gani (DaG) 476, both of which are inferred to have experienced terrestrial Re contamination. (ii) Fractionations between Rh and Pd are small. Excluding Shergotty, the Rh/Pd ratio of the SNC suite is 0.22\pm0.05. (iii) Os and Ir contents vary by about four orders of magnitude; and positive correlations with MgO, Cr, and Ni suggest that these variations are not controlled by sulfide fractionation. A possible exception is the orthopyroxenite ALH84001, whose HSE's (including Ni, which is compatible in opx) are very low. (iv) Zagami, Shergotty, and Nakhla have nearly identical HSE signatures. Shergotty and Zagami have experienced assimilation-fractional crystallization (AFC) and have "crustal" Sr and Nd isotopic signatures. Conversely, the Nakhla parent was a small degree partial melt of a depleted mantle that interacted little with the Martian crust. These observations suggest that "evolved" HSE signatures can be produced by either fractional crystallization or small degrees of partial melting. (v) Chassigny and other mafic SNC's have HSE signatures that are very distinct from those of Nakhla-Zagami-Shergotty. The HSE elemental ratios of mafic SNC's approach chondritic, implying that the Martian mantle has nearly chondritic relative abundances of the HSE's. (vi) This chondritic HSE signature is observed in SNC's of various ages, suggesting that this is an ancient feature that has not evolved over time. (vii) No correlation is observed between HSE's and signatures of crustal contamination (e.g., Sr isotopes), indicating that the HSE signatures of the SNC suite are not derived from the crust. (vii) The Ru/Pd for the SNC suite ratio is about 0.5 chondritic. We attribute this depletion of Ru to the presence of spinel in the SNC source regions. (viii) Assignment of this depletion of Ru to spinel seems to also imply that spinel is not the cause of Os and Ir fractionations, requiring some other mafic phase to account for the observed variations. -----.