Abstract:
One of the most widely used methods to estimate magmatic oxygen fugacity involves the use of empirical equations relating fO2 to the iron redox state in quenched silicate liquids; however none of the equations have been calibrated experimentally under subliquidus conditions at temperatures and oxygen fugacities that are relevant to natural magmas. To address this problem, we tested two empirical relationships [Eq.(1) in Kress and Carmichael 1991; Eq.(6) in Nikolaev et al. 1996] on synthetic glasses synthesized from a ferrobasaltic and a transitional alkali-basaltic composition at sub- to superliquidus temperatures (1,132–1,222C) and controlled oxygen fugacities (FMQ=–2 to +1.4). Fe3+/Fe was determined using conventional and milliprobe Mssbauer spectroscopy, and verified using wet chemical analysis on selected samples. For the ferrobasaltic bulk composition SC1-P, both empirical models reproduce the Fe3+/Fe ratio of the quenched liquids generally within 0.03 for sub- as well as superliquidus temperatures, although agreement is worse at higher oxygen fugacities (FMQ>+1) at subliquidus temperatures. For the transitional alkali-basaltic composition 7159V-P, both models reproduce the Fe3+/Fe ratio of the quenched liquids generally within 0.04, although agreement is worse for both models at high oxygen fugacities (FMQ>+1). Such behaviour may be related to a change in melt structure, where a progressive change in Fe3+ coordination is inferred to occur as a function of Fe3+/Fe based on Mssbauer center shifts. Recasting the data in terms of oxygen fugacity shows that calculated oxygen fugacities deviate from those actually maintained during the equilibration of the sample material by generally no more than 0.5log-bar unit, with maximum deviations that only rarely exceed one log-bar unit.
