The Stability of Hydrous Silicates in Earth’s Lower Mantle: Experimental constraints from the System MgO-Al2O3-SiO2-H2O

Michael J Walter1, Andrew R Thomson2, Weiwei Wang3, Oliver T Lord4, Annette K Kleppe5, Jennifer Ross2 and Simon C Kohn6, (1)University of Bristol, Bristol, BS8, United Kingdom, (2)University of Bristol, Bristol, United Kingdom, (3)University College London, London, United Kingdom, (4)University of Bristol, School of Earth Sciences, Bristol, BS8, United Kingdom, (5)Diamond Light Source, Didcot, United Kingdom, (6)University of Bristol, School of Earth Sciences, Bristol, United Kingdom
Abstract:
Laser-heated diamond anvil cell experiments were performed at pressures from ~ 30 to 125 GPa on bulk compositions in the system MgO-Al2O3-SiO2-H2O (MASH) to constrain the stability of hydrous phases in Earth’s lower mantle. Phase identification in run products by synchrotron powder diffraction reveals a consistent set of stability relations for the high-pressure, dense hydrous silicate phases D and H. Experiments show that aluminous phase D is stable to ~ 55 GPa. Aluminous phase H becomes stable at ~ 40 GPa and remains stable to higher pressures throughout the lower mantle depth range in both model peridotitic and basaltic lithologies. Preliminary FEG-probe analyses indicate that Phase H is alumina-rich at ~ 50 GPa, with only 5 to 10 wt% each of MgO and SiO2. Variations in ambient unit cell volumes show that Mg-perovskite becomes more aluminous with pressure throughout the pressure range studied, and that Phase H may become more Mg- and Si-rich with pressure. We also find that at pressures above ~ 90 GPa stishovite is replaced in Si-rich compositions by seifertite, at which point there is a corresponding increase in the Al-content of phase H. The melting curves of MASH compositions have been determined using thermal perturbations in power versus temperature curves, and are observed to be shallow with dT/dP slopes of ~ 4K/GPa. Our results show that hydrated peridotitic or basaltic compositions in the lower mantle should be partially molten at all depths along an adiabatic mantle geotherm. Aluminous Phase H will be stable in colder, hydrated subducting slabs, potentially to the core-mantle boundary. Thus, aluminous phase H is the primary vessel for transport of hydrogen to the deepest mantle, but hydrous silicate melt will be the host of hydrogen at ambient mantle temperatures.