Quantifying Multireference Character in Multicomponent Systems with Heat-Bath Configuration Interaction

Area of Science:

• Quantum Chemistry
• Computational Chemistry
• Theoretical Chemistry

Background:

• Multicomponent quantum chemical methods aim to incorporate nuclear quantum effects by relaxing the Born-Oppenheimer approximation for specific nuclei.
• Accurate protonic densities in multicomponent methods necessitate the inclusion of electron-proton correlation.
• Existing multicomponent methods often rely on single-reference mean-field wave functions, neglecting crucial electron-proton correlation.

Purpose of the Study:

• To investigate the impact of using single-reference mean-field (Hartree-Fock) orbitals in multicomponent quantum chemical calculations.
• To develop and apply a novel multicomponent heat-bath configuration interaction (HCI) method for accurate electronic structure calculations.
• To analyze the multireference character of systems within a multicomponent framework.

Main Methods:

• Development of the multicomponent heat-bath configuration interaction (HCI) method, a selected configuration interaction technique.
• Application of the multicomponent HCI method to HeHHe+, HCN, and FHF- systems.
• Comparison of results with reference grid-based calculations to validate protonic density accuracy.

Main Results:

• The multicomponent HCI method successfully reproduced the ground-state protonic densities for HeHHe+, HCN, and FHF-.
• All studied systems exhibited significant multireference character (leading configuration coefficient < 0.95) in the multicomponent framework.
• This multireference character persisted even when using natural orbitals instead of Hartree-Fock orbitals for FHF- and HCN.

Conclusions:

• Using Hartree-Fock orbitals in multicomponent calculations can be problematic due to the inherent multireference character of these systems.
• Multireference character appears to be more common or inherent in multicomponent quantum chemical frameworks compared to single-component ones.
• The developed multicomponent HCI method provides an accurate approximation for systems with large active spaces and highlights the importance of electron-proton correlation.