Further Development of iCIPT2 for Strongly Correlated Electrons

TL;DR

This paper enhances the iCIPT2 method for strongly correlated electrons by introducing new algorithms and data structures, significantly improving efficiency and memory usage, enabling it to handle larger configuration spaces in complex molecules.

Contribution

The paper introduces a new ranking criterion, a particle-hole algorithm, and a data structure for iCIPT2, significantly boosting its efficiency and capacity for strongly correlated electron systems.

Findings

Achieved up to 20x speedup in iCIPT2 performance.

Handled an order of magnitude more CSFs than previous version.

Demonstrated effectiveness on chromium dimer and [2Fe-2S] cluster.

Abstract

The efficiency of the recently proposed iCIPT2 [iterative configuration interaction (iCI) with selection and second-order perturbation theory (PT2); J. Chem. Theory Comput. 16, 2296 (2020)] for strongly correlated electrons is further enhanced (by up to 20x) by using (1) a new ranking criterion for configuration selection, (2) a new particle-hole algorithm for Hamiltonian construction over randomly selected configuration state functions (CSF), and (3) a new data structure for the quick sorting of the joint variational and first-order interaction spaces. Meanwhile, the memory requirement is also reduced greatly. As a result, this improved implementation of iCIPT2 can handle one order of magnitude more CSFs than the previous version, as revealed by taking the chromium dimer and an iron-sulfur cluster, [2Fe-2S], as examples.