Sulfur isotope fractionation during metal–silicate equilibration and implications for Earth’s mantle

Sulfur (S), a moderately siderophile and volatile element, may have been delivered to Earth during the planet’s main stage of accretion. Alternatively, it may have been supplied by the late veneer, after the core was fully segregated from the mantle. Mid-ocean ridge basalts present a systematic depletion in the heavier S isotope, 34 S, relative to chondrites. This has been suggested to reflect isotopic fractionation during core–mantle differentiation or during evaporative loss from Earth’s magma ocean. Two studies have estimated the metal–silicate S isotope fractionation but yielded conflicting results. Laboratory experiments suggest that equilibrium fractionation favors the heavy 34 S in metal during metal–silicate equilibria. However, the magnitude of the 34 S/ 32 S fractionation between metal and silicate appears too small to explain the composition of mid-ocean ridge basalts. In contrast, ab-initio calculations predict fractionation in the opposite direction, with a much smaller magnitude. Here, we present a series of experiments conducted with the triple-isotope technique. This approach allows for a robust determination of sulfur isotope fractionation between metal and silicate. We also report a set of conventional time-series experiments with variable starting isotopic compositions, and compare the results from both approaches. The reported results confirm that metal–silicate equilibration produces significant 34 S-depletion in silicates, with a magnitude of 0.65 ± 0.28‰ at 1673 K, in agreement with previous experimental work. This fractionation is parametrized as a function of temperature to account for Earth’s mantle composition following 1000 ∗ ln α M e t . - S i l . ≈ a / T 2 = 1.819 ± 0.784 × 10 6 K - 2 . This fractionation can explain most depleted mantle estimates; however, additional processes such as the effect of high-pressure equilibrium, sulfide–melt equilibria, or long-term mantle recycling may contribute to the observed Earth’s mantle regional variations.