A bespoke single-band Hubbard model material

Abstract

The Hubbard model, which augments independent-electron band theory with a single parameter to describe electron-electron correlations, is widely regarded to be the "standard model" of condensed-matter physics. The model has been remarkably successful at addressing a range of correlation phenomena in solids, but it neglects many behaviors that occur in real materials, such as phonons, long-range interactions, and, in its simplest form, multiorbital effects. Here, we use ab initio electronic structure methods to design a material whose Hamiltonian matches as closely as possible that of the single-band Hubbard model. Our motivation is to compare the measured properties of our new material to those predicted by reliable theoretical solutions of the Hubbard model to determine the relevance of the model in the description of real materials. After identifying an appropriate crystal class and several appropriate chemistries, we use density-functional theory and dynamical mean-field theory to screen for the desired electronic band structure and metal-insulator transition. We then explore the most promising candidates for structural stability and suitability for doping, and we propose specific materials for subsequent synthesis. Finally, we identify a regime - that should manifest in our bespoke material - in which the single-band Hubbard model on a triangular lattice exhibits exotic d-wave superconductivity.