Executive Summary : | The purpose of this proposal is to develop and apply advanced simulations methods to understand one of the most complex phenomena, phase transitions – both in single-component and multi-component molecular systems. Molecular crystals are an important class of materials that have diverse applications from chemical industries to pharmaceuticals production. Understanding the early stage of crystallization is of paramount importance to optimize crystallization conditions and guide experiments. Moreover, since most of the pharmaceuticals are produced by solution crystallization, the study of concentration- and solvent-dependent crystal shape and morphology prediction could aid the crystals design and engineering. Our aim of this project would be many-fold: first - the development of Machine-Learning (ML)-based accurate force fields for molecular systems, second – finding suitable order parameters (physics and ML-based) to study ordered-disordered phase transitions using Enhanced Sampling (ES) simulations methods, and finally investigating solution crystallization at realistic experimental conditions. Special emphasis will be given to the study of cocrystals' growth/dissolution and prediction of solubility. |