Force Field Parametrization of Colloidal CdSe Nanocrystals Using an Adaptive Rate Monte Carlo Optimization Algorithm.

In a typical colloidal CdSe nanocrystal more than 50% of the atoms are located at the surface. These atoms can give rise to electronic traps that can deteriorate the performance of optoelectronic devices made of these nanomaterials. A key challenge in this field is thus to understand with atomistic detail the chemical processes occurring at the nanocrystal surface. Molecular dynamics simulations represent an important tool to unveil these processes, but its implementation is strongly limited by the difficulties of finely tuning classical force fields parameters, primarily caused by the unavailability of experimental data of these materials that are suitable in the parametrization procedures. In this work, we present a general scheme to produce force field parameters from first-principles calculations. This approach is based on a newly developed stochastic optimization algorithm called Adaptive Rate Monte Carlo, which is designed to be robust, accurate, easy-to-use, and flexible enough to be straightforwardly extended to other nanomaterials. We demonstrate that our algorithm provides a set of parameters capable of satisfactorily describing nonstoichiometric CdSe nanocrystals passivated with oleate ligands akin to experimental conditions. We also demonstrate that our new parameters are robust enough to be transferable among crystal structures and nanocrystals of increasing sizes up to the bulk.