Theoretical study of Cu(38-n)Au(n) clusters using a combined empirical potential-density functional approach

Thirty-eight atom Cu-Au clusters have been studied because this is a magic size for a complete truncated octahedral cluster. The clusters are investigated using two approaches, at different levels of theory, which are complementary. The first is an empirical potential (EP) approach which is used together with a genetic algorithm (GA) to tackle the problems of global optimization-i.e., searching for lowest-lying energy structures. The second is an ab initio approach based on density functional theory (DFT) which is used to reoptimize the initial EP structures (both global minima and other low energy isomers). Structural distributions and energy landscapes, including calculations of electronic energy gaps for all compositions of Cu(38-n)Au(n), are investigated. The energy competition between different structural motifs and different configurations are studied at the DFT level. The analysis of mixing and segregation effects results in confirmation of the preference for Cu(core)-Au(shell) configurations at the DFT level. Charge transfer is calculated for different structural motifs of Cu(19)Au(19) to study the role of this phenomenon in driving cluster configuration.