Multireference Model Chemistries for Thermochemical Kinetics

Area of Science:

• Computational Chemistry
• Theoretical Chemistry
• Quantum Chemistry

Background:

• Accurate prediction of chemical reaction energies and barrier heights is crucial for understanding chemical reactivity.
• Traditional single-reference methods often struggle with static correlation inherent in many chemical reactions.
• Existing multireference methods can be computationally expensive, limiting their application to large systems.

Purpose of the Study:

• To develop new multireference (MR) theoretical model chemistries for accurate electronic structure calculations.
• To introduce a computationally viable approach for treating static correlation in chemical reactions.
• To assess the accuracy of the new methods for reaction barrier heights and transition structure geometries.

Main Methods:

• Combined generalized valence bond ansatz of correlated participating orbitals (CPO) with complete-active-space configuration selection.
• Employed multireference second-order perturbation theory (MRMP2) to include dynamical correlation.
• Defined three levels of MR theoretical model chemistries: nominal (nom-CPO), moderate (mod-CPO), and extended (ext-CPO).

Main Results:

• The MRMP2/nom-CPO method achieved an average error of only 1.4 kcal/mol in barrier heights, significantly outperforming single-reference MP2 and several density functionals.
• The accuracy for transition structure bond lengths and donor-acceptor distances was excellent, with a mean unsigned error of 0.007 Å.
• Introduced a new multireference diagnostic, the M diagnostic, to quantify the importance of static correlation.

Conclusions:

• The developed CPO-based MR theoretical model chemistries provide a computationally efficient and accurate approach for studying chemical reactions.
• These methods effectively capture static correlation, leading to improved predictions of reaction barrier heights and geometries.
• The CPO approach offers a viable alternative for complex reacting systems where traditional methods fail.