Assessment of correlation energies based on the random-phase approximation
Joachim Paier, Xinguo Ren, Patrick Rinke, Gustavo E. Scuseria, Andreas, Grueneis, Georg Kresse, and Matthias Scheffler

TL;DR
This paper systematically evaluates the performance of RPA-based methods and their extensions for calculating correlation energies, dissociation, and activation energies, demonstrating improved accuracy and balanced performance across various chemical systems.
Contribution
It provides a comprehensive benchmark of RPA and beyond-RPA methods, including SOSEX and SE corrections, highlighting their impact on energy calculations and reaction energetics.
Findings
RPA+SOSEX+rSE yields balanced accuracy for diverse energies.
Single excitation corrections improve binding energy predictions.
Activation energies are accurately predicted using RPA-based methods.
Abstract
The random-phase approximation to the ground state correlation energy (RPA) in combination with exact exchange (EX) has brought Kohn-Sham (KS) density functional theory one step closer towards a universal, "general purpose first principles method". In an effort to systematically assess the influence of several correlation energy contributions beyond RPA, this work presents dissociation energies of small molecules and solids, activation energies for hydrogen transfer and non-hydrogen transfer reactions, as well as reaction energies for a number of common test sets. We benchmark EX+RPA and several flavors of energy functionals going beyond it: second-order screened exchange (SOSEX), single excitation (SE) corrections, renormalized single excitation (rSE) corrections, as well as their combinations. Both the single excitation correction as well as the SOSEX contribution to the correlation…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
