Accurate Correlation Potentials from the Self-Consistent Random Phase Approximation.

Area of Science:

• Computational Physics
• Quantum Chemistry
• Materials Science

Background:

• Density-functional theory (DFT) using Kohn-Sham (KS) formalism is a cornerstone for electronic structure calculations.
• Developing accurate exchange-correlation functionals is a long-standing challenge in DFT.
• Existing approximate density functionals often yield poor-quality correlation potentials, hindering DFT development.

Purpose of the Study:

• To investigate the generation of highly accurate correlation potentials within DFT.
• To establish a basis for the further development of Kohn-Sham DFT by providing accurate potentials.

Main Methods:

• Utilized the random phase approximation (RPA) to obtain accurate correlation potentials.
• Combined RPA correlation potentials with exact exchange potentials via the optimized effective potential method.
• Employed a numerically stable approach for self-consistent RPA calculations.
• Used a KS inversion method with coupled cluster electron densities to derive reference correlation potentials.

Main Results:

• Demonstrated that highly accurate correlation potentials can be obtained using the RPA.
• Generated accurate exchange-correlation potentials that yield Kohn-Sham highest occupied molecular orbital eigenvalues close to negative ionization potentials.
• RPA-derived highest occupied molecular orbital eigenvalues show competitive accuracy with GW approximation for ionization potentials.

Conclusions:

• Accurate correlation potentials are achievable within the random phase approximation.
• The developed method provides a foundation for advancing Kohn-Sham DFT accuracy.
• This approach offers a promising alternative for calculating electronic structures and ionization potentials.