Quantification of electron correlation effects - Quantum Information Theory versus Method of Increments

TL;DR

This paper compares Quantum Information Theory and the Method of Increments for quantifying electron correlation, demonstrating their similar patterns in a model system and proposing ways to leverage each approach for more efficient electronic structure calculations.

Contribution

It shows the correlation measures from QIT and MoI are similar, enabling cross-utilization to improve computational efficiency in strongly correlated electron systems.

Findings

QIT and MoI correlation measures exhibit similar patterns.

MoI can be used to screen DMRG active spaces.

QIT results can inform MoI correlation contributions.

Abstract

Understanding electron correlation is crucial for developing new concepts in electronic structure theory, especially for strongly correlated electrons. We compare and apply two different approaches to quantify correlation contributions of orbitals: Quantum Information Theory (QIT) based on a Density Matrix Renormalization Group (DMRG) calculation and the Method of Increments (MoI). Although both approaches define very different correlation measures, we show that they exhibit very similar patterns when being applied to a polyacetelene model system. These results suggest one may deduce from one to the other, allowing the MoI to leverage from QIT results by screening correlation contributions with a cheap ("sloppy") DMRG with a reduced number of block states. Or the other way around, one may select the active space in DMRG from cheap one-body MoI calculations.