SPC cluster modeling of metal oxides: ways of determining the values of point charges in the embedded cluster model
- 化学化工－已发表论文 
Several criteria for determining self-consistently the magnitude of point charges employed in the embedded cluster modeling of metal oxides have been proposed and investigated. Merits and demerits of these criteria have been compared. Ab initio study has been performed to show the influence of the values of point charges chosen on the calculated electronic properties of the embedded MgO cluster. The calculation results demonstrate that the electronic properties of the embedded cluster are of great dependence on the magnitude of the embedding point charges; that the employment of the nominal charges, +/-2.0, would cause overestimation of the crystal potential even in the case of the so-called purely ionic oxide, MgO; and that certain requirements for the consistence between the embedded cluster and the embedding point charges should be reached. It is further found that errors for the calculated properties of the embedded duster still exist with respect to those of bulk solid even in the case that self-consistence in terms of charge, dipole moment, or electrostatic potential was met between the cut-out cluster and the embedding point charges. As far as spherical expansion is performed upon the embedding point charges, which furnishes the embedding point charges with a continuous distribution of charge density, a global agreement is reached between the calculated properties of the embedded cluster model and those of the bulk solid.
CitationSCIENCE IN CHINA SERIES B-CHEMISTRY，1998,41（2）：113-121
Showing items related by title, author, creator and subject.
冯少荣; 潘炜炜; 林子雨 (2015-9-15)XMl文档由于其自身的可扩展性、半结构化和自描述性等特点,已成为数据表示和交换的数据格式标准。一个高效、快速的XMl文档聚类机制能够大幅缩短信息检索时间,提高数据查询的效率,挖掘出潜在的信息价值。为此,提出一种改进的k-MEdOIdS算法对XMl文档进行聚类。运用模糊聚类方法确定聚类个数,利用遗传算法的全局最优的搜索能力求解最佳聚类中心点或质心,从而提高大规模XMl文档集的聚类质量。实验结果表明,与基于传统k-MEdOIdS算法的聚类方 ...
陈黎飞; 姜青山; 王声瑞 (2008-01-15)确定数据集的聚类数目是聚类分析中一项基础性的难题.常用的trail-and-error方法通常依赖于特定的聚类算法,且在大型数据集上计算效率欠佳.提出一种基于层次思想的计算方法,不需要对数据集进行反复聚类,它首先扫描数据集获得CF(clusteringfeature,聚类特征)统计值,然后自底向上地生成不同层次的数据集划分,增量地构建一条关于不同层次划分的聚类质量曲线;曲线极值点所对应的划分用于估计最佳的聚类数目.另外,还提出一种新的聚 ...
HU Hui-rong; WANG Zhou-jing (Department of Automation; Xiamen University; Xiam; en; China) (《厦门大学学报(自然科学版)》编辑部, 2002-12-30)Hierarchical clustering analysis based on statistic s is one of the most important mining algorithms, but the traditionary hierarchica l clustering method is based on global comparing, which only takes in Q clusteri ng ...