The effect of composition on Cr2+/Cr3+ in silicate melts

Abstract

Chromium K-edge X-ray absorption near-edge structure (XANES) spectra were recorded at room temperature for 27 CaO-MgO-Al₂O₃-SiO₂ (CMAS) glass compositions quenched from melts equilibrated at various oxygen fugacities (f_{O2) at 1400 °C. Values of Cr ²⁺/∑Cr were determined from the intensity of a shoulder on the main absorption edge, attributed to the 1s → 4s transition, which is characteristic of Cr²⁺ in these glasses. For each composition, Cr²⁺/∑Cr could be quantified as a function of fO2, using a theoretical expression, from as few as three samples (Cr²⁺/∑Cr 0, 0.5, and 1). This allowed logK′, or the reduction potential of the Cr^3+/2+ half-reaction, and hence the relative change in the ratio γ^meltCr³⁺O1.5/γ^meltCr²⁺O, to be determined for each composition. At constant fO2, log[Cr²⁺/Cr³⁺] was found to decrease linearly with increasing optical basicity. The variation in logK′ with composition is controlled by γ^meltCr³⁺O1.5, corresponding to the capacity of the melt to stabilize both the charge and the preferred solvation site of Cr³⁺. The method was then applied to spectra recorded in situ at 1400 °C for a synthetic mid-ocean ridge basalt (MORB) composition, allowing Cr²⁺/∑Cr to be quantified in a Fe-bearing melt for the first time. Cr²⁺/∑Cr was found to vary from ~0.45 at the nickel-nickel oxide (NNO) fO2 buffer to ~0.90 at iron-wüstite (IW). This indicates that Cr²⁺ is likely to be the dominant oxidation state in terrestrial basaltic melts.}