Experimental petrology of upper mantle materials, processes and products

Experiments on rock materials and volatile components provide an array of phase equilibrium boundaries, giving the depth-temperature framework for the phase transitions experienced by rock masses as they move up or down within the Earth in response to the dynamic processes of mantle convection and plate tectonics. Sub-solidus phase transitions in mantle materials correlate well with upper mantle structure determined from seismic studies, but debate continues about whether there is a change in composition at some seismic boundaries. The interface at 650 km correlates with the transformation of most minerals into the perovskite structure, which may have significant effects on mantle dynamics. Recent research on mantle xenoliths has been concerned with extension of standard thermometers and barometers to higher pressures and their practical assessment, along with the development and refinement of new geothermobarometers. The experimental partial melting of mantle peridotite has been elucidated by detailed studies of model systems and new experimental techniques. Parameterization of the data makes prediction possible. The origin of MORBs involves a complex process of fractional fusion. Evidence from static olivine flotation experiments and shock wave compression experiments on molten komatiite indicates that the densities of mantle melts may exceed that of the residual rock at depths greater than 400 km, which would prohibit ascent of the magma. Recent investigations have identified many dense hydrous magnesian silicates (DHMS), stable through the upper mantle between ~ 300 and 650 km, and reaching the peridotite-volatile solidus curve through part of this interval. The near-solidus liquid composition in peridotite-CO_2-H_2O above ~ 2 GPa is carbonatitic (calcic dolomite), potentially a powerful agent for metasomatism. Experimental studies of trace element distributions between mantle minerals and carbonatitic liquids are beginning to permit quantification. Determination of the effect of reduced oxygen fugacity has introduced mantle scenarios with melting induced by redox changes There is indirect experimental evidence that the solidus may terminate at a critical end-point at depths of a few hundred kilometers. Recent experiments related to subduction of oceanic crust include measurements of liquid composition from the melting of H_2O-undersaturated peridotite, the vapour-absent melting of amphibolite, and partial melting of pelagic clays. Much H_2O is expelled during subduction, but the current estimates of low temperatures support the deep subduction and longterm storage in the mantles of both H_2O and CO_2