遗传算法结合密度泛函理论研究Cu13团簇的几何结构

基于Gupta 多体势采用遗传算法详细计算了金属团簇Cu13的几何结构，对所得到的大量低能（基态及低激发态）结构，利用密度泛函理论方法作进一步优化计算。结果表明：尽管遗传算法所得基于Gupta原子间多体势的基态结构（Ih高对称性的紧致结构）并不对应第一性原理计算结果（非紧致低对称性基态），但遗传算法给予的大量候选结构经密度泛函理论再次计算仍然可以高效地得到真实基态结构，体现出遗传算法在计算具有非紧致低对称性基态体系时的有效性。

Study of Geometries of the Cu13 Clusters using the Genetic Algorithm and Density-functional Calculations

Geometrical structures of Cu13 cluster are calculated by a genetic algorithm combined with the Gupta inter-atomic potential systematically, and the obtained structures are re-optimized in further by density-functional theory (DFT) calculations. Our results show that: though the high-symmetric compact ground-state structure obtained by Gupta potential is not the same one as that obtained by DFT which favors a non-compact low symmetric ground-state, from re-optimization by DFT of the large number of initial structures provided by Gupta potential, one still can find the true ground-state geometrical structure, which means that the genetic algorithm can still offer the efficiency of structure optimization in dealing with a system with non-compact low symmetric ground-state geometry.