Testing models for the compositions of chondrites and their components: III. CM chondrites

Abstract

In continuation of our comprehensive study of the carbonaceous chondrites, we here present data for the CM chondrites. Our study’s aim is to determine the abundances and the average elemental compositions of the major and minor chondritic components. The overarching goal has been to explore the fundamental question of whether chondrules and matrix are complementary, i.e., genetically related, or if these major chondritic components evolved separately before accretion as in the four-component model.Using point-counting and electron microprobe analyses, we investigated the most primitive CMs known to date: Asuka (A) 12169 (CM3.0), A 12236 (CM2.9) and Paris (CM2.7–2.9). Despite their primitiveness, however, none of the samples is completely devoid of signs of terrestrial and/or parent body alteration. For instance, we found evidence for the mobilization and heterogenous redistribution of Ca, as well as for the leaching of Al from chondrules and redeposition into matrix. There is also a trend of the Ca and Al abundances in both, chondrules and matrix, to become increasingly heterogeneous with increasing parent body alteration. We were, therefore, unable to test chondrule-matrix complementarity and the four-component model using Al and Ca. Assuming that Mg and Si are insignificantly affected by alteration and that the Mg/Si ratios of the components in A 12169 are unaltered, our data indicate the pre-accretionary loss from matrix of ∼12 wt% forsterite (matching our results for the primitive COs and CRs) and the addition of a roughly similar amount of forsterite to the chondrules (contrary to our results for the COs and CRs). These results are inconsistent with the four-component model but, based on the matrix/chondrule abundance ratio, complementarity predicts that the addition of forsterite to chondrules should have been significantly higher. This and the possibility that A 12169 component compositions have been modified to some degree means that our results do not unambiguously favor either model.