Single- and multireference electronic structure calculations for constructing potential energy surfaces. Issue 3 (2nd July 2016)

• Record Type: Journal Article
• Authors: Dawes, Richard; Ndengué, Steve A.
• Is Part Of: International reviews in physical chemistry. Volume 35:Issue 3(2016)
• Page Start: 441
• Page End: 478
• Publication Date: 2016-07-02
• Subjects: potential energy surfaces -- electronic structure theory -- multireference -- coupled cluster -- dynamic weighting; Chemistry, Physical and theoretical -- Periodicals; 541.3
• DOI: 10.1080/0144235X.2016.1195102
• Languages: English
• ISSNs: 0144-235X

Abstract

Abstract : Recent developments in single and multireference electronic structure methods and the approaches suitable to generate ab initio data that may be employed in the construction of global molecular potential energy surfaces are reviewed. The most appropriate, robust, accurate and cost effective strategies are discussed in the context of various applications ranging from cold collisions and weakly interacting clusters, to large amplitude motion in covalently bound molecules, as well as reaction and photodissociation dynamics. The relationships between the types and necessary quantity of ab initio data, and representations through fitting are important, and issues related to symmetry and electronic state degeneracy are mentioned. The impacts of limitations or error in the electronic structure data are discussed in terms of how they are reflected in calculations of spectroscopy, dynamics and kinetics. This discussion includes examples such as the submerged reef feature found along the path to formation of ozone on several published potentials. For that example, a relatively small absolute error in the form of a spurious barrier has profound effects on the dynamics and rates of exchange reactions. The origin of the spurious barrier in ozone and other systems is discussed from an electronic structure standpoint. The effective use of dynamically-weighted state-averaged multireference calculations to obtain robustly convergent global surfaces is detailed.