Schorl breakdown at upper mantle conditions: Insights from an experimental study at 3.5 GPa

Hydrogen and B input throughout the Earth's mantle is continuously fed through a sequence of dehydration and breakdown reactions of hydrous and B-bearing mineral phases stable at different conditions along the subducting slabs. Therefore, the stability of minerals hosting these elements plays a fundamental role. Tourmaline hosts very large amounts of B (up to 14 wt% of B2O3) along with hydroxyl groups (up to 4 wt% of H2O), thus representing a crucial mineral to investigate the fate of B and H in diverse geological settings. The recent finding of tourmaline minerals in ultra-high pressure metamorphic rocks has raised important questions about the actual tourmaline stability field, paying special attention to the high pressure and temperature stability limits of the various tourmaline species. A single-phase system made of natural schorl with the highest Fe2+ concentration known so far (about 18 wt% of FeO) was studied at a fixed pressure (3.5 GPa) and several temperatures (500, 700, 750, 800, 850 and 950 degrees C) to preliminarily constrain its stability conditions, breakdown mechanisms and breakdown products. Experiments at high pressure-high temperature conditions were performed using a multi anvil apparatus under buffered oxygen fugacity through a Re/ReO2 solid mixture. The experimental products were characterized through a multi-analytical approach consisting in Scanning Electron Microscopy imaging and Energy Dispersive System spectra acquisition, Electron MicroProbe analysis, powder X-Ray Diffraction, 57Fe Mo spacing diaeresis ssbauer spectroscopy and reflectance Fourier Transform infrared spectroscopy. At 3.5 GPa and T ranging from 500 up to 700 degrees C, the schorl experienced a partial Fe oxidation coupled with dehydrogenation: Fe2+ + (OH) -> Fe3+ +O2-+0.5H2(g) The observed Fe oxidation was limited to 30% (significantly lower than the full oxidation observed in ex-periments performed in air at room pressure), suggesting that oxidation-dehydrogenation is indeed a thermally activated process, but both environmental pressure and oxygen fugacity are important governing factors. In the pure schorl system at 3.5 GPa, the structural breakdown started at T = 700 degrees C and ended at 850 degrees C, resulting in the formation of almandine garnet as the first breakdown product together with topaz and a B-rich liquid phase: -> Fe2+3Al2(SiO4)3 almandine Na(Fe2+2Al)(Al5Fe2+)(Si6O18)(BO3)3(OH)3(OH , F) schorl + ( Fe3+ ,Al)2SiO4(OH , F)2 topaz -> + 2SiO2 + Al2O3 + 0.5Na2O + 1.5B2O3 + H2O melt At 3.5 GPa and T >= 850 degrees C, tourmaline, garnet and topaz were not observed anymore and kyanite, prismatine-and boromullite-like phases and corundum became stable. Both prismatine-like and boromullite-like phases identified by stoichiometry can incorporate B from the B-rich hydrous melt formed after schorl breakdown and may carry it to lower depths. From our work it follows that the schorl-bearing granitoid rocks (or sediments) have the potential to form hydrous B-bearing metasomatic melts at 3.5 GPa and T >= 700 degrees C. In cold subduction environments, between the 700-800 degrees C isotherms, the schorl is expected to be stable up to-100 km depth along the subducting slab, although an excess SiO2 might be responsible for a reduction in tourmaline stability. The role of tourmaline companion minerals on its breakdown conditions and products is left as future issue when a multi-phase system will be considered.