SULFUR PARTITIONING BETWEEN CLINOPYROXENE AND MELT: TACKLING GAS LOADS FROM LIP MAGMAS (Invited Presentation)

Earth history is punctuated by mass extinctions that in most cases were demonstrated synchronous with, and causally related to, magmatism from Large Igneous Provinces (LIPs). Prominent examples in the Phanerozoic are the end-Permian, end-Triassic and end-Cretaceous extinctions, associated with, respectively, the Siberian Traps, the CAMP and the Deccan Traps. Despite the growing body of evidence for causal and temporal links between these events, it is not yet entirely clear how a LIP can wreak havoc on the global environment. Degassing of volatile species such as S, C and halogen compounds directly from LIP magmas, and from contact metamorphism of volatile-rich sediments heated by the intrusions appears as the most realistic mechanism. Modeling the atmospheric response to LIP gas loads requires quantitative estimations of the degassed volatiles and emission rates – both hard to obtain for ancient magmatic systems.

We propose a new approach to calculate the sulfur load of basaltic melts, by measuring sulfur content in natural minerals (clinopyroxene and plagioclase) and combining it with an experimentally determined partition coefficients (KD). We measured partitioning of sulfur between crystals and melt by ion microprobe (Nordsim, Stockholm) on experimentally produced crystals and glasses. Piston cylinder experiments were performed with conditions typical of basaltic, andesitic and dacitic melts (800 or 1000 MPa; 1000°-1350°C), to constrain KD variations as a function of melt composition, oxidation state and water content. We obtained a clinopyroxene/melt sulfur KD of 0.001 for basaltic melts, which can be applied to natural continental flood basalts. Preliminary results from thoroughly-dated lava piles from the Deccan Traps and from the Siberian Traps sills confirm that most of the basalts were at or close to sulfide saturation (ca. 2000 ppm for low fO₂ melts). These results can be compared with the scenario modeled by Schmidt et al. (2016) for Deccan Traps magmatism, for which sulfur from flood basalts can only cause a biotic crisis if released in repeated and closely-spaced pulses. Combined with other studies of H, C, Cl and F partitioning, our study on sulfur allows complete assessment of the volatile cocktail characterizing a basaltic melt.

Schmidt A., et al. (2016). Nature Geosciences 9, 77–82.