Efficient Structural Relaxation Based on the Random Phase Approximation: Applications to Water Clusters

Area of Science:

• Computational Chemistry
• Quantum Chemistry
• Materials Science

Background:

• Accurate prediction of molecular structures and energies is crucial in chemistry.
• Random Phase Approximation (RPA) is a powerful method for electron correlation, but computationally intensive.
• Water clusters are fundamental systems for understanding hydrogen bonding and solvation.

Purpose of the Study:

• To present an efficient implementation for analytical gradients of RPA electron-correlation energy.
• To investigate the structures and energy orderings of small to medium-sized water clusters.
• To assess the accuracy of RPA and its corrections for binding energies.

Main Methods:

• Improved analytical gradients for RPA using atomic orbitals and localized resolution of the identity.
• Structure relaxation of water clusters (H2O)n, n=21, 22, 25.
• Basis set convergence studies and comparison with double hybrid functionals.

Main Results:

• Efficient RPA force calculations enabled structure relaxation of water clusters.
• RPA energy ordering of low-energy isomers is sensitive to basis set choice.
• RPA underbinds water clusters, exacerbated at the complete basis set (CBS) limit.
• Renormalized single excitation (rSE) correction mitigates underbinding, showing overbinding at finite basis sets.
• RPA+rSE achieves accuracy comparable to double hybrid functionals near the CBS limit for binding energies.

Conclusions:

• The improved RPA implementation allows for efficient structural analysis of water clusters.
• Basis set choice significantly impacts the predicted energy landscape of water clusters.
• RPA+rSE offers a computationally viable and accurate method for predicting binding energies, especially relevant for systems like water clusters.