Quantum Geochemistry: Theoretical Prediction of Melting Behaviour and Phase Relations of Minerals in the Deep Mantle
IPGP - Îlot Cuvier
Séminaires de Géophysique expérimentale
In this talk, I will discuss a novel computational framework to predict melting phase relations of deep mantle minerals at HP-HT conditions by a combination of first principles DFT calculations, polymer chemistry and equilibrium thermodynamics. The adopted method is thermodynamically-consistent and allows to compute multi-component phase diagrams relevant to Earth’s deep interior in a broad range of P-T conditions by Gibbs free energy minimisation algorithm. The calculated phase diagrams are in turn used as a source of information to gain some insights on the P-T-X evolution of magmas in the deep mantle, providing also a thermodynamic constraint to both early Earth and present-day mantle processes (from the crystallization behaviour of deep magma oceans to the effect of phase transitions on mantle convection and seismic discontinuities). Specific focus will be devoted to those mineral phases (like majorite-pyrope garnet and phase Anhydrous B), which are often disregarded in modelling melting processes of mantle assemblages due to poor knowledge of their thermodynamic and thermo-physical properties. This against the experimental evidence that they are stable phases at solidus or liquidus conditions in a broad range of bulk compositions (e.g. dry and hydrous CMAS-pyrolite, KLB-1 peridotite). Applying the above method to simulation of melting relations in the simplified, but representative, MgO-Al2O3-SiO2 (MAS) system highlights as pressure effects are not only able to change the nature of melting of some minerals (like olivine and pyroxene), but also produce an oversimplification of melting relations in the MAS system by drastically reducing the number of phases with a primary phase field at HP-HT.