Structures and evolution of metal oxide nanoclusters: Bottom-up genetic algorithm and fragment-based energy decomposition model

Abstract

Abstract Theoretical prediction for the structures of large nanoclusters remains a challenge but also presents opportunities for the discovery of new materials. Several computational tools specifically designed to understand large nanoparticles are introduced in the chapter. A novel evolutionary global geometry optimization method, the bottom-up genetic algorithm (GA) has been developed to predict thermodynamically favorable nanocluster structures with sizes up to a few nanometers. The bottom-up GA has minimal requirement for initial geometry guesses and can efficiently search global minima over multiple potential energy surfaces (PESs). A spectrum of low energy nanoclusters with a wide range of particle sizes can be predicted using bottom-up GA, which enables understanding of structural evolution of inorganic materials as well as the structure-property relationships. A fragment-based energy decomposition (FBED) framework has been developed to extract the structure-stability and structure-reactivity relationships with a machine learning methodology for inorganic materials at all scales based only on the computational data for the ultrasmall nanoparticles (USNPs). The derived relationships have implications in nanoparticle growth and controlled synthesis for inorganic materials.