Ab initio-assisted computational thermodynamics: a modern approach to phase diagram calculation at planetary conditions

<p>Despite the outstanding progress in computer technology and experimental facilities, understanding melting processes and solid-melt phase equilibria at planetary conditions is still an open challenge. In this work a modern computational approach to predict melting phase relations at HP-HT by a combination of first principles DFT and MD calculations, polymer chemistry and equilibrium thermodynamics is presented and discussed. The adopted theoretical framework is physically-consistent and allows to compute multi-component phase diagrams relevant to planetary interiors in a broad range of P-T conditions by a convex-hull algorithm based on the simplex method for Gibbs free energy minimisation. The calculated phase diagrams are in turn used as a source of information to gain new insights on both present-day and early Earth melting processes. Some examples of application of the above method to the CaO-MgO-Al₂O₃-SiO₂ system (CMAS) and relevant ternary and binary subsystems highlights as pressure effects are not only able to change the nature of melting of some minerals (like olivine and pyroxene) from eutectic to peritectic (and vice versa), but also simplify melting relations by drastically reducing the number of phases with a primary phase field at HP-HT conditions. Since the volume-pressure integral contribution to Gibbs free energy become relevant at planetary interior conditions, special attention must be paid to the choice of the P-V-T EoS formalism in order to avoid physical unsoundness or spurious effects in thermodynamic properties (e.g. negative thermal expansion). Ab initio-assisted computational thermodynamics is thus outlined as the main route towards the future development of physically-consistent (besides internally-consistent) thermodynamic databases for global-scale planetary investigations.</p> <p>Financial support by the Italian Ministry of University and Research (MIUR PRIN 2017, Project 2017KY5ZX8 and MIUR PRIN 2020, Project 202037YPCZ) is warmly acknowledged. &#160;&#160;</p>