4.7 Article

Benchmark Database for Ylidic Bond Dissociation Energies and Its Use for Assessments of Electronic Structure Methods

期刊

JOURNAL OF CHEMICAL THEORY AND COMPUTATION
卷 8, 期 8, 页码 2824-2834

出版社

AMER CHEMICAL SOC
DOI: 10.1021/ct300457c

关键词

-

资金

  1. National Science Foundation [CHE09-56776]
  2. Division Of Chemistry
  3. Direct For Mathematical & Physical Scien [0956776] Funding Source: National Science Foundation

向作者/读者索取更多资源

We report a database of 18 ylidic bond dissociation energies obtained by using highly accurate quantum mechanical methods, and we use it to test approximate electronic structure methods. The new benchmark database is called YBDE18 and is used to test a large number of electronic structure methods, including eight wave function methods and 98 density functional exchange-correlation functionals. Among them, we include some very recent density functionals, including the SOGGA11 GGA functional, the SOGGA11-X hybrid GGA functional, the M11-L local meta-GGA functional, and the M11 range separated hybrid meta-GGA functional. We also consider other functionals of these classes plus a local spin density approximation, global hybrid meta-GGAs, range separated hybrid GGAs, doubly hybrid GGAs, and doubly hybrid meta-GGAs. We found M05-2X-D3, MPWB1K-D3, M05-2X, LC-BLYP, PBE0-D3, and MC3MPWB to be the best DFT methods for this database. Although they do not place in the top four overall, our new-generation functionals show overall competitive performances; each of the new functionals provides the smallest mean signed error within its class, while in terms of mean unsigned errors, SOGGA11 is the best GGA, and SOGGA11-X and M11-L are among the first three best functionals in their categories, global-hybrid GGA and local meta-GGA. The best local functionals are VSXC and M06-L, the best global-hybrids are M05-2X, M08-HX, M06-2X, and MPWB1K, and the best range-separated hybrids are LC-BLYP, omega B97, omega B97X, and M11.

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。

评论

主要评分

4.7
评分不足

次要评分

新颖性
-
重要性
-
科学严谨性
-
评价这篇论文

推荐

暂无数据
暂无数据