DFVB: A density-functional-based valence bond method

Fuming Ying, Peifeng Su, Zhenhua Chen, Sason Shaik, Wei Wu*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

40 Scopus citations

Abstract

A new ab initio valence bond method with density-functional-based correlation correction, so-called DFVB, is presented. In the DFVB method, the dynamic correlation energy is taken into account by use of density correlation functional(s), while the static correlation energy is covered by the VBSCF wave function. Owing to incorporation of DFT methods, DFVB provides an economic route to improving the accuracy of ab initio VB theory. Various tests of the method are presented, including the spectroscopic parameters of a series of diatomic molecules, the dipole moments of the NF molecule for different electronic states, and the singlet-triplet gaps of the diradical species, chemical reactions barriers, and total charge-shift resonance energies. These tests show that DFVB is capable of providing high accuracy with relatively low computational cost by comparison to the currently existing post-VBSCF methods.

Original languageEnglish
Pages (from-to)1608-1615
Number of pages8
JournalJournal of Chemical Theory and Computation
Volume8
Issue number5
DOIs
StatePublished - 8 May 2012

Fingerprint

Dive into the research topics of 'DFVB: A density-functional-based valence bond method'. Together they form a unique fingerprint.

Cite this