Program > Browse abstracts by speaker > Richard Thomas

Sulfur in silicate melts primarily exists as sulfide (S²⁻) and sulfate (S⁶⁺). While extensive 1-atm experiments have clarified how temperature and composition influence sulfur speciation [1,2], pressure effects remain less constrained. Prior estimates rely on assumptions based on FeS solubility, but direct experimental constraints are needed.

To investigate pressure-dependent sulfide capacity (logCₛ²⁻ = logS²⁻(melt)(wt%) + 0.5log(fO₂/fS₂)), we conducted experiments at 1400°C across 1 bar to 2.5 GPa, with f(S₂) controlled via Ag-Ag₂S and oxygen fugacity fixed at the CCO buffer. Our results indicate that logCₛ²⁻ increases linearly as pressure decreases. By integrating our data with [1,2], we derive the following relationship:

logCₛ²⁻ = 0.405 + (-0.05732P + 24661XFeO + 5804XCaO - 1312XSi₀.₅O + 25366XK₂O - 9092)/T

To assess sulfate capacity (logCₛ⁶⁺ = logS⁶⁺(melt)(wt%) - (1.5log fO₂) - (0.5log fS₂)), we performed high-pressure experiments using CaSO₄ as the sulfur source. Results indicate increasing S⁶⁺ content with pressure, leading to the derived equation:

logCₛ⁶⁺ = -13.174 + (-0.128P + 9453XMgO + 4242XCaO + 27255XNa₂O + 4255XAl₀.₆₇O + 19672XMnO + 29414)/T

These findings suggest that ascending melts become enriched in Fe³⁺ and S²⁻, implying greater mantle stability for S⁶⁺ than previously assumed. This also raises concerns that Fe³⁺/ΣFe ratios from XANES may misrepresent mantle f(O₂).

Additionally, sulfur diffusion experiments in Icelandic basalt reveal that S⁶⁺ diffuses much more slowly than S²⁻, contrasting with previous findings [3]. This slow diffusion likely limits S⁶⁺ fractionation into bubbles during decompression of silica-rich melts.

References:
[1] O'Neill (2022), [2] Boulliung & Wood (2023), [3] Baker & Rutherford (1996)