Hubbard vs. Emery model: spectra, transport and relevance for cuprates

TL;DR

This study compares the spectral and transport properties of the Hubbard and Emery models for cuprates, identifying key differences and aligning theoretical results with experimental data to inform model selection.

Contribution

It provides a detailed comparison of two prevalent models for cuprates, highlighting quantitative differences and suggesting experimental ways to distinguish them.

Findings

Both models show similar but not identical physical pictures.

Quantitative differences can help discriminate between models.

Results suggest the effective Hubbard coupling is larger than previously thought.

Abstract

Understanding the transport properties of cuprate superconductors is one of the central challenges in the physics of strongly correlated electrons. The most common approach is to define and solve a low-energy lattice model, but it is still unclear what the minimal model is to capture all relevant mechanisms and provide quantitative predictions. The main uncertainty concerns the choice of the orbital degrees of freedom to be included in the model, as well as the definition of the effective coupling. In this paper, we study the two most commonly considered models, namely the single-orbital Hubbard model and the three-orbital Emery model. We investigate and compare their spectral and transport properties, and find that the two models present a similar, but not the same, physical picture. We identify several strong quantitative differences which might allow one to discriminate between the two models by comparing theory with experiments. We compare our results for several physical quantities with 7 different experiments on 3 different La$_2$CuO$_4$-based cuprates, and in general find excellent agreement. The dc resistivity and the effective mass results suggest that the coupling constant in the effective Hubbard model is larger than expected. We find several more properties that are sensitive to the precise value of the coupling constant, including the critical doping for the Lifshitz transition, and the local spectral weight in the vicinity of the Fermi level; the latter provides a promising way to estimate the effective coupling constant in future photoemission experiments.