Electronic structure of LaNiO$_{2}$ and CaCuO$_{2}$ from self consistent vertex corrected GW approach

TL;DR

This study applies self-consistent GW methods to analyze the electronic structures of LaNiO$_{2}$ and CaCuO$_{2}$, revealing their correlation effects and differences from previous DFT+DMFT results, with implications for ab-initio approaches.

Contribution

It demonstrates the application of self-consistent GW and vertex-corrected GW methods to nickelate and cuprate materials, highlighting differences from prior DFT+DMFT studies and emphasizing the potential of ab-initio diagrammatic approaches.

Findings

LaNiO$_{2}$ is more correlated than CaCuO$_{2}$.

Correlation effects are weak enough for ab-initio GW methods.

Fermi surface differences suggest need for experimental validation.

Abstract

Electronic structure of one of the nickelates (LaNiO$_{2}$) and one of the cuprates (CaCuO$_{2}$) is studied with three self consistent GW-based methods: scGW, sc(GW+Vertex), and quasiparticle self-consistent GW. Low energy features obtained in our study are in many respects similar to the features reported in previous DFT+DMFT studies. Consistent with the DFT+DMFT conclusion, we find LaNiO$_{2}$ as more correlated than CaCuO$_{2}$. However, correlation effects included in our study change the DFT Fermi surface near the $\Gamma$ point differently than it was reported in the DMFT studies. Features which are a few electron-volts away from the Fermi level are broader in our calculations than in the DFT+DMFT which reflects the differences between the DFT and the GW methods. Our results are in qualitative agreement with previous G0W0 results, but the self-consistency brings in the quantitative differences. Generally, correlation effects are found to be sufficiently weak in both materials which allows one to use totally ab-initio diagrammatic approaches like sc(GW+Vertex) and to avoid the methods with adjustable parameters (DFT+U or DFT+DMFT). However, the possibility of some strong correlations at low energy which cannot be captured by perturbative methods cannot be completely excluded. For instance, differences in the Fermi surface should be resolved: experimental studies are necessary.