A REASSESSMENT OF BASALTS AS RECORDERS OF UPPER MANTLE OXIDATION STATE

The oxidation state of the oceanic upper mantle controls the speciation of C-O-H volatiles and has been inferred to be near the fayalite-magnetite-quartz (FMQ) buffer on the basis of Fe³⁺/∑Fe ratios of mid-ocean ridge basalt (MORB) glasses determined by wet chemistry [1, 2] and X-ray absorption near edge structure (XANES) spectroscopy [3]. However, MORB are aggregates of near-fractional partial melts formed over a range of pressure-temperature conditions from variably depleted peridotite [4]. Therefore, determining the extent to which their Fe³⁺/∑Fe ratios are representative of oceanic mantle oxygen fugacity requires an understanding of the influence of decompression melting on the Fe³⁺/∑Fe of residual peridotite. I have developed a model that describes the geochemical behavior of Fe³⁺during spinel lherzolite melting to provide a quantitative understanding of the relationship between the oxidation state of MORB and that of the oceanic upper mantle.

The model involves calculating the Fe³⁺/∑Fe of olivine using the point defect model of [5], and determining Fe³⁺/∑Fe of the bulk peridotite from mineral-mineral partitioning derived from Mössbauer data on natural spinel peridotites. The Fe³⁺/∑Fe of the melt is determined by combining mass-balance with an equation relating the Fe³⁺/∑Fe of the melt to the fugacity of oxygen [6].

Modeling results indicate that Fe³⁺ is more compatible in peridotite than previously thought, so that a redox couple with S is not required to explain its behavior during partial melting. The positive correlation between fractionation corrected Na₂O and Fe₂O₃ results from variations in the potential temperature of the oceanic mantle. The oxidation state of the residual peridotite increases relative to FMQ with increasing extent of partial melting. The upper mantle oxygen fugacity calculated from the Fe³⁺/∑Fe ratio of MORB does not directly reflect their source region, but can be constrained by global Fe³⁺/∑Fe systematics.

References. [1] Christie et al. (1986) Earth Planet Sci Lett 79:397-411;. [2] Bézos and Humler (2005) Geochim Cosmochim Acta 69:711-725; [3] Cottrell and Kelley (2011) Earth Planet Sci Lett 305:270-282; [4] Johnson et al. (1990) J Geophys Res 95:2661-2678. [5] Dohmen and Chakraborty (2007) Phys Chem Min 34:409-430. [6] Kress and Carmichael (1991) Contrib Min Pet 108:82-92