Investigation of potential parameters effect on stable structures of Pt-Pd alloy nanoparticles by particle swarm algorithm

Because the structure determines the chemical and physical properties, atomic-level understanding of structural features of bimetallic nanoparticles (NPs) is of great importance to their syntheses and applications. In this work, we propose an improved discrete particle swarm optimization (PSO) algorithm to predict the stable structures of bimetallic NPs. In this algorithm, the variable pair of 0 and 1 (which relatively represent the two atoms) is introduced to enhance the speed of searching optimization in the initial stages, and the simulated annealing operator is added to avoid premature convergence and trapping into local optimal solution. Tetrahexahedral (THH) Pt-Pd bimetallic NPs containing 3285 atoms are used to test the effectiveness of the proposed algorithm. Furthermore, the proposed method is employed to investigate and compare the stable structures of the NPs by changing some parameters of the Gupta potentials. The results have demonstrated the superior convergence of the proposed PSO algorithm in structural prediction of bimetallic NPs. The comparison of the stable structures based on three groups of the parameters in Gupta potentials shows that for the parameter I, the structure is in best agreement with the experimental result, while for the parameter III, the Pt-Pd bimetallic NPs is most stable.