The random phase approximation applied to ice.

Area of Science:

• Computational physics
• Materials science
• Quantum chemistry

Background:

• Standard density functionals struggle to accurately model ice phases, particularly high-pressure forms, often predicting them as too unstable.
• While van der Waals interactions improve descriptions, significant errors in relative volumes persist.

Purpose of the Study:

• To evaluate the random phase approximation (RPA) for calculating correlation energy in ice phases.
• To compare RPA results with experimental data and diffusion Monte Carlo (DMC) simulations for ice.

Main Methods:

• Utilizing the random phase approximation (RPA) to compute correlation energy for various ice phases.
• Comparing RPA predictions for relative energies and volumes against experimental measurements and high-accuracy DMC data.

Main Results:

• The RPA provides a highly balanced and accurate description across different ice phases.
• RPA results for relative energies and volumes closely approach the accuracy of diffusion Monte Carlo.
• This method significantly improves upon standard density functionals and those including van der Waals interactions.

Conclusions:

• The random phase approximation (RPA) offers a promising and accurate approach for describing molecular water phases.
• This method shows potential for accurate modeling of water ice on surfaces and within cavities.
• RPA presents a viable alternative for precise simulations of condensed matter systems.