HYDROGEN ISOTOPE EVIDENCE FOR THE ORIGIN AND EVOLUTION OF THE CARBONACEOUS CHONDRITES

Abstract

We present new hydrogen isotope data for separated matrix, hydrated chondrules, and other hydrated coarse silicate fragments from nine carbonaceous chondrites. These data were generated using a micro-analytical method involving stepped combustion of tens to hundreds of micrograms of hydrous solids. We also re-evaluate hydrogen isotope data from previous conventional stepped combustion experiments on these and other carbonaceous chondrites. Hydrogen isotope compositions of matrix and whole-rock samples of CM chondrites are correlated with oxygen isotope indices, major and minor-element abundances, and abundance and isotope ratios of other highly volatile elements. These correlations include a monotonic decrease in δD with increasing extent of aqueous alteration and decreasing abundances of highly volatile elements (including C, N and Ar), between extremes of ∼0‰ (least altered, most volatile rich) and −200‰ (most altered, least volatile rich). In plots involving only abundances and/or isotope ratios of highly volatile elements, CI chondrites fall on the high-δD, volatile rich end of the trends defined by CM chondrites; i.e., CI chondrites resemble the least altered CM chondrites in these respects. These trends suggest the protoliths of the CM chondrites (i.e., before aqueous alteration) contained an assemblage of volatiles having many things in common with those in the CI chondrites. If so, then the volatile-element inventory of the CI chondrites was a more widespread component of early solar system objects than suggested by the scarcity of recognized CI meteorites. Differences in volatile-element chemistry between the CI and average CM chondrites can be attributed to aqueous alteration of the latter. Previous models of carbonaceous chondrite aqueous alteration have suggested: (1) the protoliths of the CM chondrites are volatile poor objects like the CO or CV chondrites; and (2) the CI chondrites are more altered products of the same process producing the CM chondrites. Both suggestions appear to be inconsistent with hydrogen isotope data and other aspects of the volatile-element geochemistry of these rocks. We present a model for aqueous alteration of the CM chondrites that reconciles these inconsistencies and suggests revised relationships among the major subtypes of carbonaceous chondrites. Our model requires, among other things, that the water infiltrating CM chondrites had a δD value of ∼−158‰, consistent with initial accretion of CM parent bodies at ∼4 AU.