一种改进的动态格子算法在Au团簇基态结构预测中的应用

A modified dynamic lattice searching method to predict the lowest energy structure of clusters for the Au clusters

借助计算机强大的计算能力通过数学的方法预测团簇的最低能量结构,是本文的主要思想和工作.针对Au团簇的结构优化问题,本文提出了一种改进的动态格子搜索算法.将该算法用于Gupta势能建模的Au团簇对其进行优化求解,在若干国际已知算例上,找到了它们其中许多新的最低能量结构.为了用Gupta势能描述Au团簇原子之间的相互作用,本文采用了两组不同的参数.利用参数一(A=0.11844,B=1,p=10.15,q=4.13),优化了原子数N=38?100的Au团簇.其中,对于原子数N=38,55这两个算例,本文算法的结果优于此前文献中的最好结果.另外,利用参数二(A=0.2061,B=1.79,p=10.229,q=4.036),优化了原子数N=100?200的Au团簇,其中对于原子数N=100,110,120,130,140,150,160,170,180,190,200的Au团簇,本文算法所达到的势能均优于此前文献中的最好结果.结果表明了本文算法对于团簇结构优化问题求解的高效性.

It is the main idea and work of this paper that mathematical methods are used to predict the lowest energy structure of clusters with powerful computing capability of the computer. A modified dynamic lattice searching method is proposed in this paper for Au clusters. It is applied to the Au clusters with Gupta potential, for a series of international examples of known, they are optimized by the algorithm, and many new lowest energies and structures are found. To describe the interactions between Au atoms in a cluster with Gupta potential, two sets of parameters are taken in this paper. With the first set of parameters（A=0.11844, B=1, p=10.15, q=4.13）, the Au clusters with the number of atoms from 38 to 100 are optimized by us. The obtained results indicate that our results of N=38, 55 are better than the reference. On the other hand, with the second set of parameters（A=0.2061, B=1.79, p=10.229, q=4.036）, the Au clusters with the number of atoms from 100 to 200 are optimized by us, and the results of N=100, 110, 120, 130, 140, 150, 160, 170, 180, 190, 200 are better than the reference. The experimental results show that the proposed method for predicting the lowest energy structure of clusters is efficient.