Experimental investigations on the role of sulfur in magmas

Sulfur (S) is the third volatile element in importance in Earth's magmas, after water and carbon dioxide. Recent work has shown that sulfur abundance (melt+gas) in typical arc magmas commonly exceeds 0.1 wt%, with maxima in the range 0.5-1 wt% [Scaillet et al., 2003]. In this presentation we will review recent advances with respect to the solubility and partitioning of S in silicate melts and give a general outline of on going, as well as next, research priorities. The solubility of S in metaluminous rhyolitic and phonolitic melts has been extensively investigated by Clemente et al. [2003] and Moncrieff et al. [in prep], respectively. Experiments have been performed at 800- 1000°C, 1-4 kb, and fO2 ranging from NNO-2 up to NNO+4. Both studies have shown that S solubility is sensitive to T, fO2 and fS2 variations, pressure control being less important. At any given fO2, the higher the fS2, the higher the S solubility. Rhyolite and phonolite melts follow the same pattern. The modeling of S solubilities has been attempted using either empirical or thermodynamical approaches. Empirical equations are available for either a given melt composition (rhyolite) or are valide over a large compositional spectrum (rhyolite-basalt). The thermodynamic modelling of S solubilities can be done considering that the total dissolved S results from the addition of H2S and SO2 species dissolution reactions, whose relative abundances depend on the prevailing fO2. This model has been calibrated only on rhyolite compositions. The S2-/S6+ proportions predicted by this model are in good agreement with experimental observations. The partition coefficient of S between melt and gas has been determined mostly in silicic magmas [Scaillet et al., 1998; Keppler, 1999], for which there appears to be a strong control of fO2. Thermodynamic calculations predict that the partition coefficient between gas and melt should decrease with melt silica content, from ca 1000 in rhyolite, to 100 in andesite, to 10 in basaltic melts [Scaillet and Pichavant, 2003]. Recent experimental work has explored the S behaviour in peralkaline rhyolites, which appears to dissolve up to 20 times more sulfur than their metaluminous counterpart [Scaillet and Macdonald, in prep]. Despite its relative low concentrations, S affects the stability of phases such as pyroxenes, amphiboles and biotite. In moderately oxidized dacite magmas, the addition of a few thousands ppm of S enhances the thermal stability of biotite by as much as 60°C [Costa et al., submitted]. In strongly oxidized dacites, the incorporation of S leads to the breakdown of horblende at the expense of orthopyroxene [Scaillet and Evans, in prep]. These effects on phase relations suggest the existence of various sulfur complexes in the silicate melt. We are currently concentrating our efforts on (1) calibrating existing solubility models on mafic compositions, (2) determining the S partitioning between melt and gas in mafic melts. We suggest that additional investigations should be devoted to (1) investigating the effects of S on silicate melt transport properties (density, diffusivity and viscosity), (2) determining the nature and proportion of S species dissolved in silicate melts so as to build more realistic thermodynamic models. References Clemente, B., Scaillet, B. and Pichavant, B. (2003). The solubility of sulphur in rhyolitic melts. Journal of Petrology, in press. Keppler, H. (1999). Experimental evidence for the source of excess sulfur in explosive volcanic eruptions. Science, 284, 1652-1654. Scaillet, B., Clemente, B., Evans, B. and Pichavant, M. (1998). Redox control of sulfur degassing in silicic magmas. Journal of Geophysical Research, 103, 23937-23949. Scaillet, B. and Pichavant, M. (2003). Experimental constraints on volatile abundances in arc magmas and their implications for degassing processes. Geol. Soc. Spec. Pub., in press. Scaillet, B., Luhr, J. and Carroll, M. (2003). Petrological and volcanological constraints on volcanic sulfur emissions to the atmosphere. In Volcanoes and the Earth atmosphere, A. Robock and C. Oppenheimer (eds.), AGU, in press