Super ccCA (s-ccCA): an approach for accurate transition metal thermochemistry

Abstract

Computational chemistry provides a powerful route to determine thermochemical properties. For transition metal thermochemistry, typically, high-level quantum methodologies are required. Ab initio composite methods – which aim to replicate the predictions possible from a high-level method/advanced basis set with predictions from a series of lower-level methods/basis sets to reduce computational cost – have proven to be effective for transition metal species, with better than chemical accuracy for transition metal energetics (<12  kJ mol⁻¹), on average. While useful, to provide a more robust computational approach, Super ccCA (s-ccCA) is introduced herein and varies from its predecessor, ccCA, by utilising higher-level coupled-cluster corrections along with a spin–orbit contribution from a Breit-Pauli Hamiltonian. In this work, s-ccCA has been utilised for the prediction of dissociation energies of 3d and 4d molecules. A set of borides, sulphides and carbides in conjunction with three early first-row transition metals (Sc, Ti, V) and three second-row transition metals (Y, Zr, Nb) were studied with this new composite method. The energies calculated herein were compared with the experiment and shown to be in excellent agreement. The energetic predictions show that for cases where a balance of static and dynamic correlation is of paramount importance, s-ccCA offers an effective approach.

GRAPHICAL ABSTRACT

KEYWORDS:

• Electronic structure
• transition metal thermochemistry
• multireference
• ab initio composite
• ccCA

Acknowledgements

This work utilised computational facilities at the Institute for Cyber-Enabled Research (ICER) at Michigan State University. The authors acknowledge the tremendous influence that John Stanton has had upon the field, particularly upon spectroscopy and thermochemistry.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This material is based upon work supported by the National Science Foundation under [grant number CHE-1900086].