期刊
JOURNAL OF COMPUTATIONAL CHEMISTRY
卷 24, 期 8, 页码 973-981出版社
WILEY
DOI: 10.1002/jcc.10273
关键词
principal component analysis; factorial design; sensitivity analysis; Monte Carlo simulation; interaction effects
The multivariate chemometric techniques two level factorial design (TLFD) and principal component analysis (PCA) were used to investigate the TIP4P model potential behavior with respect to perturbations on all intermolecular interaction parameters. The effects of these perturbations were calculated for the enthalpy of vaporization, the density, the first maximum of the radial distribution functions of the O-H and O-O pairs, and the second maximum of the radial distribution function of the O-H pair obtained from Monte Carlo simulations of liquid water at 25degreesC. The principal effects were quantified and rationalized in terms of the pair-wise interaction potential of the TIP4P model. They also corroborate previously published sensitivity analysis results using molecular dynamics and other model potentials. In addition, significant interaction effects between some parameters of the TIP4P model potential were observed and quantified, which hardly could be obtained without such a statistic approach. These interaction effects are very regular and systematic, and their behavior has not been encountered in other chemometric studies and cannot be rationalized in terms of the functional form of the pair-wise potential. (C) 2003 Wiley Periodicals, Inc.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据