An improved genetic algorithm for structural optimization of Au–Ag bimetallic nanoparticles

Abstract

The structures are crucial for bimetallic nanoparticles (NPs) because they can determine their unique physical and chemical properties. Therefore, structures optimization of bimetallic NPs from theoretical calculation is of increasing importance for understanding their stabilities and catalytic performance. In this article, an improved genetic algorithm (IGA) is proposed to systematically investigate the structural stabilities of Au–Ag NPs. In the IGA, a layered coordinate ranking method is adopted to enhance the structural stability during initialization. Meanwhile, a difference transition fitness function is introduced to keep the population diversity and preserve the best individual of IGA. Furthermore, for improving the global searching ability and local optimization speed, a sphere-cut-splice crossover is employed to replace the classical plane-cut-splice crossover in general genetic algorithm. The performance of IGA has been compared with Monte Carlo simulation method and particle swarm optimization algorithm, the results reveal our algorithm possesses superior convergence and stability.