This study evaluates the role of mantle-to-core segregation of the “Hadean matte” (sulfide liquids) occurring in the solidifying magma ocean in determining the chalcophile and siderophile element (CSE) abundances of the silicate Earth. The partition coefficients of CSEs between sulfide liquid and basaltic to peridotitic melt (DCSESul/Sil) were determined at 1–14 GPa and 1300–2100 °C. The variations in the obtained DCSESul/Sil (30–160 for Co, 50–1200 for Ni, 40–940 for Cu, 20–210 for Mo, 50–210 for Ag, 20–90 for Cd, 4–60 for In, 30–150 for Sb, 3900–30,000 for Re, 15–210 for Pb, 140–1700 for Bi, 0.3–7 for Zn, 0.7–7 for Ge, and 0.1–0.9 for Ga) can be explained and parameterized as a function of the experimental pressure, temperature, and composition of the silicate melt and sulfide liquid. Application of the DCSESul/Sil parameterization to the mantle-to-core segregation of sulfide liquids in a deep magma ocean at 75 GPa shows that less than 10 of each of Co, Ni, Cu, Zn, Ga, Ge, Mo, Ag, Cd, In, Sb, Pb, and Bi in the silicate Earth can be sequestered in the core; whereas, in a shallow magma ocean at 10 GPa, 50–80 of each of Cu, Ag, and Bi and less than 50 of the other CSEs can be sequestered in the core. In contrast, mantle-to-core segregation of sulfide liquids could have extracted more than 90 of the Re in the silicate Earth in both cases, requiring the addition of a late veneer to explain the present-day mantle value. Our results demonstrate that if Earth's volatile CSEs (Cu, Zn, Ga, Ge, Ag, Cd, In, Sb, Pb, and Bi) were delivered when metal segregation to the core was largely inactive, the depletion pattern of volatile CSEs relative to the lithophile elements of similar volatility in the silicate Earth cannot be explained by mantle-to-core segregation of sulfide liquids. Also, previous models that used mantle-to-core segregation of sulfide liquids as an important approach to explain the depletion of volatile CSEs in the silicate Earth need to be re-examined.
Zhang et al. (Wed,) studied this question.