Main-group pallasites: chemical composition, relationship to IIIAB irons, and origin

We used neutron activation to characterize the metal of 33 main-group pallasites (PMG), widely held to be samples of a core-mantle interface. Most PMG cluster in a narrow range of metal and silicate compositions, but 6 are assigned to an anomalous subset (PMG-am) because of their deviant metal compositions, and 4 others to another anomalous subset (PMG-as) because of their appreciably higher olivine Fa contents. Metal compositions in all PMG are closely related to those in evolved IIIAB irons, and are generally consistent with their formation in the IIIAB parent asteroid. On element-Au diagrams for incompatible elements the normal PMG plot near an extrapolation of IIIAB trends to higher Au concentrations. On element-Au plots of compatible elements such as Ir or Pt the loci of PMG spread out over a broader region explainable by mixing of evolved IIIAB magma with early-crystallized core or mantle-residue solids.Two features of PMG require special models: (1) Ga and Ge contents are generally high (≈1.5×) compared to the IIIAB-based mixing model: and (2) the FeO/(FeO + MgO) ratios span a surprisingly wide range, from 0.11-0.14 in normal PMG to 0.16 to 0.18 in PMG-as This range is larger than expected in a cumulate layer at the base of a mantle. We suggest that both features may be related to the interaction of PMG precursors with a highly evolved magmatic gas phase, and that some or all of these anomalies may have resulted from vapor deposits in voids near the core-mantle interface.An important boundary condition for understanding the detailed PMG origin at the core-mantle interface is the large difference between the solidus temperature of Fa11 olivine (≈2000 K) and the liquidus temperature of an evolved IIIAB melt containing >100 mg/g S and some P (≈1600 K). Following the mixing event that formed the PMG it is therefore reasonable that there would have been olivine rubble floating on top of the IIIAB-like magma, but with appreciable void space present just above the upper level reached by the magma. These voids would have contained gases released from the magma during its flow into the PMG region. We suggest that Ga and Ge, the two most volatile siderophiles in our element set, were added to PMG metal from the magmatic gas. We also suggest that the magmatic gas was oxidizing and that the PMG having high olivine fayalite contents formed in regions where the ratio of void to olivine was high, and that some metallic Fe was oxidized and entered the olivine (or the phosphoran olivine). In support of the latter idea is the observation that both Ni and Co are elevated in the PMG-as (Fa≥16) compared to values predicted by IIIAB trends.We analyzed two Eagle-Station pallasites (PES); after correction for weathering effects in Cold Bay, its composition is found to closely resemble that of Eagle Station but to represent a more evolved composition (i.e., lower Ir, higher Au). Vermillion and Yamato 8451 have been called pyroxene pallasites but have metal compositions (unrelated to those of the PMG or PES) that are too different from each other to even allow assignment to the same grouplet.