Maximally localized Wannier functions in LaMnO3 within PBE+U, hybrid functionals, and partially self-consistent GW: an efficient route to construct ab-initio tight-binding parameters for e_g perovskites
C. Franchini, R. Kovacik, M. Marsman, S. Sathyanarayana Murthy, J. He,, C. Ederer, G. Kresse

TL;DR
This paper develops an efficient method to construct maximally localized Wannier functions for LaMnO3 using various computational approaches, enabling accurate tight-binding models for e_g perovskites.
Contribution
It introduces a new interface and methodology to derive tight-binding parameters from ab initio calculations for complex correlated materials.
Findings
TB parameters vary with the exchange-correlation method used
Explicit and implicit electron-electron interaction models produce consistent band dispersions
The derived TB models accurately reproduce ab initio and MLWF band structures
Abstract
Using the newly developed VASP2WANNIER90 interface we have constructed maximally localized Wannier functions (MLWFs) for the e_g states of the prototypical Jahn-Teller magnetic perovskite LaMnO3 at different levels of approximation for the exchange-correlation kernel. These include conventional density functional theory (DFT) with and without additional on-site Hubbard U term, hybrid-DFT, and partially self-consistent GW. By suitably mapping the MLWFs onto an effective e_g tight-binding (TB) Hamiltonian we have computed a complete set of TB parameters which should serve as guidance for more elaborate treatments of correlation effects in effective Hamiltonian-based approaches. The method-dependent changes of the calculated TB parameters and their interplay with the electron-electron (el-el) interaction term are discussed and interpreted. We discuss two alternative model…
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.
