Metal-silicate partitioning and the light element in the core (Invited)

Most attempts to constrain the concentrations of “light” elements in the Earth’s core rely either on cosmochemical arguments or on arguments based on the densities and equations of state of Fe-alloys containing the element of concern. Despite its utility, the latter approach yields a wide range of permissible compositions and hence weak constraints. The major problem with the cosmochemical approach is that the abundances in the bulk Earth of all the candidate “light” elements- H, C, O, Si and S are highly uncertain because of their volatile behavior during planetary accretion. In contrast, refractory elements appear to be in approximately CI chondritic relative abundances in the Earth. This leads to the potential for using the partitioning of refractory siderophile elements between the mantle and core to constrain the concentrations of light elements in the core. Recent experimental metal-silicate partitioning data, coupled with mantle abundances of refractory siderophile elements (e.g. Wade and Wood, EPSL v.236, 78—95,2005; Kegler et. al. EPSL v.268, 28-40,2008) have shown that the core segregated from the mantle under high pressure conditions (~40 GPa). If a wide range of elements, from very siderophile, (e.g. Mo) through moderately (Ni, Co, W) to weakly siderophile (V, Cr, Nb, Si) are considered, the Earth also appears to have become more oxidized during accretion. Metal-silicate partitioning of some elements is also sensitive to the light element content of the metal. For example, Nb and W partitioning depend strongly on carbon, Mo on silicon and Cr on sulfur. Given the measured mantle abundances of the refractory elements, these observations enable the Si and C contents of the core to be constrained at ~5% and <2% respectively while partitioning is consistent with a cosmochemically-estimated S content of ~2%.