Constraints on core composition from shock-waves data

Abstract

Seismic data demonstrate that the density of the liquid core is some 8-10 % less than pure iron. Equations of state of Fe-Si, C, FeS 2 , FeS, KFeS 2 and FeO, over the pressure interval 133-364 GPa and a range of possible core temperatures (3500- 5000 K), can be used to place constraints on the cosmochemically plausible light element constituents of the core (Si, C, S, K and O ). The seismically derived density profile allows from 14 to 20 % Si (by mass) in the outer core. The inclusion of Si, or possibly G (up to 11 %), in the core is possible if the Earth accreted inhomogeneously within a region of the solar nebulae in which a C :0 (atomic) ratio of about 1 existed, compared with a G : O ratio of 0.6 for the present solar photosphere. In contrast, homogeneous accretion permits Si, but not C, to enter the core by means of reduction of silicates to metallic Fe-Si core material during the late stages of the accumulation of the Earth. The data from the equation of state for the iron sulphides allow up to 9-13 % S in the core. This composition would provide the entire Earth with a S:Si ratio in the range 0.14-0.3, comparable with meteoritic and cosmic abundances. Shock-wave data for KFeS 2 give little evidence for an electronic phase change from 4s to 3d orbitals, which has been suggested to occur in K, and allow the Earth to store a cosmic abundance of K in the metallic core.