High pressure experimental study of natural antigorite dehydration reactions

From the Mid-Oceanic-Ridge to the subduction trench, hydration of peridotite minerals in the upper part of the oceanic lithosphere produces hydrous phases such as serpentine. Because of its high-water content (13 wt% H2O) this mineral family is of particular interest for water fluxes. Depending on the thermal path followed by the lithosphere while sinking into the mantle, antigorite destabilization can either lead to fluid release in the mantle wedge or water transfer to deeper levels. During this thesis we conducted experimental investigations of antigorite dehydration in the framework of these two scenarios.First, we investigated antigorite dehydration under conditions relevant to slab water release, known to trigger partial melting and to generate arc magmatism. Multi-anvil experiments were conducted on a natural serpentinite sample, at 3 GPa and between 600 and 900°C under different redox conditions. We were able to constrain phase assemblages produced by antigorite dehydration as well as the fO2 of such reactions to 5 units above the FMQ (Fayalite-Magnetite-Quartz buffer). These results support the oxidizing character of slab released fluids, that could explain the oxidized character of arc magmas compared to Mid-Oceanic-Ridge basalts or Oceanic-Island basalts.The second experimental work conducted during this thesis allowed to refine phase equilibria involving antigorite and the Dense Hydrous Magnesium Silicates (DHMS) phase A and phase E, in a realistic chemical composition for hydrated ultramafic system. Antigorite destabilization was performed between 6.5 and 10 GPa, for temperatures in the range phase E>phase A for the aluminous and iron-rich hydrated peridotite system. This study allowed the refinement of water budgets that can be stored in relatively cold slabs (<750°C at 8-10 GPa), supporting the hypothesis of water survival down to the transition zone.