Show simple item record

dc.contributor.authorXin Xu
dc.contributor.author徐昕
dc.contributor.authorWilliam A. Goddard
dc.date.accessioned2011-12-16T01:57:47Z
dc.date.available2011-12-16T01:57:47Z
dc.date.issued2004-09
dc.identifier.citationJ. Phys. Chem. A, 2004, 108 (40):8495–8504zh_CN
dc.identifier.issn1089-5639
dc.identifier.urihttp://dx.doi.org/doi:10.1021/jp047428v
dc.identifier.urihttps://dspace.xmu.edu.cn/handle/2288/11405
dc.description.abstractIn this paper we present a systematic assessment of the Handy-Cohen optimized exchange density functional (OPTX), comparing results from OLYP and O3LYP with those from BLYP and B3LYP. We find that OPTX significantly outperforms Becke88 in the calculations of the atomic exchange energies, and O3LYP leads to the best total atomic energies (H to Ar) among these four functions. We find OLYP and O3LYP are competitive or even better than BLYP and B3LYP in the predictions of ionization potentials, electron affinities, and proton affinities against the extended G2 set of 75 atoms and molecules. For thermochemistry of the extended G2 set of 148 molecules, we find that the mean absolute deviation (in kcal/mol) follows the order BLYP (7.10) > OLYP (4.66) > O3LYP (4.13) > B3LYP (3.14). Thus OLYP is the best pure DFT, but B3LYP is the best overall. The histogram of error distribution of the G2 set indicates that O3LYP has more predictive power than B3LYP, although O3LYP has a tendency for overbinding. OLYP and O3LYP significantly outperform BLYP and B3LYP in describing van der Waals interactions, but OLYP and O3LYP underestimate hydrogen bond strengths even more than BLYP and B3LYP and, hence, cannot be recommended for studying hydrogen bond systems.zh_CN
dc.language.isoenzh_CN
dc.publisherAMER CHEMICAL SOCzh_CN
dc.titleAssessment of Handy-Cohen optimized exchange density functional (OPTX)zh_CN
dc.typeArticlezh_CN


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record