Parameterizing DFT+U+V from Hybrid Functionals: A Wannier-Function-Based Approach for Strongly Correlated Materials

TL;DR

This paper introduces a Wannier-function-based method to derive DFT+U+V parameters from hybrid functional calculations, enabling accurate electronic structure modeling of strongly correlated materials at reduced computational cost.

Contribution

The authors develop a novel approach to parameterize DFT+U+V using hybrid functional data via Wannier functions, improving efficiency and accuracy for strongly correlated systems.

Findings

Accurately reproduces hybrid-functional electronic structures with DFT+U+V.

Validates method on diverse oxide materials with different electronic properties.

Achieves computational efficiency suitable for structural relaxations and many-body calculations.

Abstract

We present an approach to parameterize DFT+$U$+$V$ from hybrid-functional calculations using Wannier-function projections. The method constructs a common localized Wannier basis for both semilocal DFT and hybrid-functional calculations, then determines effective on-site ($U$) and intersite ($V$) Hubbard parameters by minimizing the Hamiltonian mismatch within the correlated subspace. This procedure yields interaction parameters that reproduce the hybrid-functional electronic structure at a fraction of the computational cost and allow efficient structural relaxations and further many-body calculations. We validate the workflow on three oxide systems with different electronic characters: MgO (wide-gap insulator), NiO (antiferromagnetic charge-transfer insulator), and V$_2$O$_5$ (d$^0$ transition-metal oxide). In all cases, the mapped DFT+$U$+$V$ parameters reproduce hybrid-functional band gaps, densities of states, and magnetic moments and improve upon semilocal DFT while maintaining computational efficiency.