An example of atomic scale approach to solid solutions and disordered crystalline systems - Valentina Lacivita

Valentina Lacivita, ISTeP - UPMC

Lundi 5 septembre 2016 à 10 h 30

IMPMC - Université P. et M. Curie - 4, Place Jussieu - 75005 Paris

Salle de conférence - Barre 22-23 - 4e étage - Salle 401

Abstract

Very often materials of scientific and technological interest involve substitutional and/or positional disorder. These materials challenge atomistic simulations because the disorder implies dealing with large lowsymmetry supercells where multitudes of atomic distributions exist. An important methodological progress has come from the recent development of a symmetry-adapted Monte Carlo (SA-MC) method which
overcomes the main flaws of standard MC and MC Metropolis algorithms. It exploits space symmetry (i) to partition the configurations in classes of equivalence - thus avoiding repetitions, and (ii) to ensure a uniform draw probability over all classes regardless of their size. This is important because the most stable configurations usually correspond to “rare events”, namely classes with few configurations. Results obtained from a full ab initio quantum-mechanical investigation of the hydrogrossular solid solution are presented: average geometric features of the actual solid solution are closely approximated and two energy minima are found that can be associated to structures or compositions observed in nature. Further examples of case studies include the calcite carbonate Ca0.75Mg0.25CO3, and soda-melilite, (Na,Ca)AlSi2O7, where Na and Ca cations are randomly located in a layered tetrahedral network. In these cases it is shown how the use of the SA-MC method allows to deal with the issues related to the convergence of the thermodynamic properties and to the choice of the supercell.

Contact : Christian Brouder (christian.brouder @ upmc.fr)