THE STABILITY OF IGNEOUS ANHYDRITE IN EXPERIMENTAL ANDESITIC MELTS

Experiments were conducted in which anhydrite was added to andesite rock powder in order to investigate the stability of igneous anhydrite, which can serve as an indicator of the sulfur content and oxygen fugacity of a melt. Capsules containing andesite with 2 wt% or 6 wt% anhydrite and either water or hydrogen peroxide were run in cold-seal reaction vessels at 900° C and approximately 240 MPa pressure for one to seven days. The melt from andesite-anhydrite-hydrogen peroxide runs had about 40% less CaO than andesite-only runs with either water or hydrogen peroxide, while the melt from andesite-anhydrite-water runs had about 20% more CaO than andesite-only runs with either water or hydrogen peroxide. No anhydrite or other sulfide mineral grew in any of the runs, including those containing 2 wt% or 6 wt% anhydrite, and glass analyses of the runs showed only minimal amounts of sulfur in the melt; thus, the added sulfur must have gone into the gas phase. Only runs with added anhydrite grew plagioclase or clinopyroxene (much as overgrowths on amphibole). Runs with anhydrite and hydrogen peroxide grew more amphibole (often euhedral) than runs with anhydrite and water. The hydrogen peroxide run with 6 wt% anhydrite grew abundant amphibole and little plagioclase, while the hydrogen peroxide run with 2 wt% anhydrite grew abundant plagioclase and somewhat less abundant amphibole. Both andesite-only runs (no anhydrite) with either water or hydrogen peroxide grew the iron oxide magnetite; andesite-anhydrite-water runs grew ilmenite, and andesite-anhydrite-hydrogen peroxide runs grew magnetite. The oxygen-richer hydrogen peroxide (relative to plain water) may have increased the oxygen fugacity, at least temporarily, to allow greater growth of amphibole and plagioclase and a shift of the plagioclase to a more sodic composition. Additional shorter-duration experiments under HM-buffered conditions were also run and are under investigation.