Electronic Structure of Strongly Correlated Systems Emerging from Combining Path-Integral Renormalization Group with Density Functional Approach

TL;DR

This paper introduces a first-principles computational scheme combining density functional theory, GW, and path-integral renormalization-group methods to accurately study strongly correlated electron systems, demonstrated on Sr2VO4.

Contribution

The novel approach integrates multiple advanced methods to better capture low-energy physics in strongly correlated materials, improving predictive power over existing techniques.

Findings

The scheme accurately reproduces experimental results for Sr2VO4.

It predicts a nontrivial orbital-stripe order in Sr2VO4.

The method effectively considers spatial and dynamical fluctuations.

Abstract

A new scheme of first-principles computation for strongly correlated electron systems is proposed. This scheme starts from the local-density approximation (LDA) at high-energy band structure, while the low-energy effective Hamiltonian is constructed by a downfolding procedure using combinations of the constrained LDA and the GW method. Thus obtained low-energy Hamiltonian is solved by the path-integral renormalization-group method, where spatial and dynamical fluctuations are fully considered. An application to Sr$_2$VO$_4$ shows that the scheme is powerful in agreement with experimental results. It further predicts a nontrivial orbital-stripe order.