Highly accurate electronic structure of metallic solids from coupled-cluster theory with nonperturbative triple excitations

TL;DR

This paper introduces a new coupled-cluster method, ring-CCSDT, capable of accurately computing electronic structures of metals by including nonperturbative triple excitations, overcoming previous divergence issues.

Contribution

The paper develops ring-CCSDT, an iterative method with $N^7$ scaling, enabling accurate electronic structure calculations for metallic solids.

Findings

Ring-CCSDT achieves high accuracy in metallic systems.

Including connected triple excitations is crucial for precise results.

Semiempirical CC methods improve accuracy of ab initio approaches.

Abstract

Coupled-cluster theory with single, double, and perturbative triple excitations (CCSD(T)) -- often considered the "gold standard" of main-group quantum chemistry -- is inapplicable to three-dimensional metals due to an infrared divergence, preventing its application to many important problems in materials science. We study the full, nonperturbative inclusion of triple excitations (CCSDT) and propose a new, iterative method, which we call ring-CCSDT, that resums the essential triple excitations with the same $N^7$ run-time scaling as CCSD(T). CCSDT and ring-CCSDT are used to calculate the correlation energy of the uniform electron gas at metallic densities and the structural properties of solid lithium. Inclusion of connected triple excitations is shown to be essential to achieving high accuracy. We also investigate semiempirical CC methods based on spin-component scaling and the distinguishable cluster approximation and find that they enhance the accuracy of their parent ab initio methods.