Atomic and molecular clusters: discovering low-energy structures with evolutionary algorithms
By: Jorge Marques
From: Universidade de Coimbra
At: Faculdade de Ciências, Ed. C3, 3.1.10
[2015-09-24]
Clusters are formed by a relatively small number of atoms or molecules, which leads these systems to have rather different properties in comparison to the condensed phase. The knowledge of low-energy structures of such systems is particularly relevant and, hence, some effort has been developed to build efficient algorithms for discovering the most stable spatial arrangements of particles that are subject to an interaction potential. In general, this is a very difficult problem, due to the enormous number of local minima and the roughness of the potential energy surface. In the last decade, we have developed state-of-the-art evolutionary algorithms (EAs) that have shown to be successful in discovering the global minimum structure of several atomic and molecular clusters. In the seminar, I will present the main features of the EAs and showcase their application to discover global minimum structures of binary atomic clusters, to study the microsolvation of alkali-metal ions with different solvent molecules, and to get insight about charge-repulsion effects in colloidal aggregates.Â
