GiovanniMaria Piccini Università di Modena e Reggio Emilia
Abstract:
Molecular dynamics (MD) serves as a crucial tool for studying the evolution of complex chemical systems. However, the exploration of significant metastable states, such as reactants and products, is often hindered by large free energy barriers separating them. To surmount these challenges and facilitate transitions between metastable states, the integration of enhanced sampling methods with machine learning-derived potentials proves to be a powerful combination.
These methods become of paramount importance when delving into the sampling of chemically activated events within catalytic systems, aiming to emulate realistic operando conditions. The objective is to describe, with atomistic precision, the thermodynamics, mechanisms, and ultimately, the kinetics of the process. This presentation will showcase practical applications addressing nanoconfined catalytic reactions. Examples include sustainable organic synthesis within biomimetic calixarene capsules and the heterogeneous conversion of biomass over solid/liquid interfaces in zeolites.
