On the oxidation state and volatile behavior in multicomponent gas-melt equilibria.

The equilibrium between a 4-component H2O–CO2–SO2–H2S gas phase and a 13-component silicate liquid made of 10 major oxides plus dissolved H2O, CO2, and S, is investigated by means of calculations involving homogeneous reactions in the gas phase and heterogeneous gas–liquid saturation modeling based on classical Gibbs thermodynamics and Toop–Samis polymeric approach. Sulfur is assumed to be present in two different oxidation states in the gas (sulfur dioxide and hydrogen sulfide) and liquid (sulfide and sulfate ions) phase, implying a dependence of the equilibrium conditions on the redox state of the system. Sulfur-bearing solid phases and Fe–O–S immiscible liquid are not accounted for in the modeling. The thermodynamic model is an extension of the one presented in Moretti et al. [Moretti R., Papale P. and Ottonello, G., 2003. A model for the saturation of C–H–O–S fluids in silicate melts. In: Oppenheimer C., Pyle D.M., Barclay J. (eds.) Volcanic Degassing, Geol. Soc. London Spec. Publ., 213, 81–101.] to account for iron speciation at high pressure and dissolved water contents. The consequences on the equilibrium conditions of different assumptions concerning the effective redox buffer in magma are examined through calculations made on two different liquids of shoshonitic and rhyolitic composition, determining the equilibrium conditions on the basis of (i) constant ferric to ferrous mass ratio, (ii) constant hydrogen sulfide to sulfur dioxide fugacity ratio, and (iii) constant oxygen fugacity relative to a solid–gas buffer (DNNOF0.5). Following Giggenbach [Giggenbach, W.F., 1996. Chemical composition of volcanic gases. In: Scarpa R., Tilling R.I. (eds.) Monitoring and Mitigation of Volcano Hazards, Springer-Berlin, 202–226.], the first two buffers are expected to be effective in basaltic and rhyolitic magmas, respectively, according to the most abundant reservoir of redox couples represented by iron in basalts, and sulfur in rhyolite. The model results show strongly nonlinear dependence of the equilibrium compositions in the gas and liquid phases, as well as of the oxidation state of the system, on the assumed redox buffer. Furthermore, for each assumed redox buffer, the pressure dependence of phase composition and oxidation state of the system also shows strongly nonlinear trends. The largest compositional differences are shown by sulfur species; however, the concentrations of water and carbon dioxide in the two phases at equilibrium also show nonnegligible dependence on the redox conditions. For each assumed redox buffer, sulfur dioxide in the gas phase, and sulfate ions in the liquid phase, are found to be present in appreciable quantities or represent the dominating sulfur species even at the largest employed pressures approaching 500 MPa. The more reliable redox buffers represented by constant ferric to ferrous mass ratio for shoshonite, and constant hydrogen sulfide to sulfur dioxide fugacity ratio for rhyolite, show that oxygen fugacity paths during magma depressurization strongly deviate from those parallel to NNO. Therefore, the characterization of the oxidation state in depressurizing magmas on the basis of deviations from solid buffers (usually NNO or QFM) may not be appropriate.