Consistent LDA'+DMFT approach to electronic structure of transition metal oxides: charge transfer insulators and correlated metals
I. A. Nekrasov, N. S. Pavlov, M. V. Sadovskii

TL;DR
This paper introduces a consistent, parameter-free LDA'+DMFT approach that improves the accuracy of electronic structure calculations for transition metal oxides, addressing double counting issues and aligning well with experimental data.
Contribution
The paper presents a novel LDA'+DMFT method that effectively resolves double counting problems and enhances the accuracy of electronic structure predictions for strongly correlated materials.
Findings
LDA'+DMFT qualitatively improves results for charge transfer insulators.
Better quantitative agreement with experimental band gaps and spectral features.
Inclusion of transition metal 4s-states enhances modeling accuracy.
Abstract
We discuss the recently proposed LDA'+DMFT approach providing consistent parameter free treatment of the so called double counting problem arising within the LDA+DMFT hybrid computational method for realistic strongly correlated materials. In this approach the local exchange-correlation portion of electron-electron interaction is excluded from self consistent LDA calculations for strongly correlated electronic shells, e.g. d-states of transition metal compounds. Then the corresponding double counting term in LDA+DMFT Hamiltonian is consistently set in the local Hartree (fully localized limit - FLL) form of the Hubbard model interaction term. We present the results of extensive LDA'+DMFT calculations of densities of states, spectral densities and optical conductivity for most typical representatives of two wide classes of strongly correlated systems in paramagnetic phase: charge transfer…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
