Effects of interaction strength, doping, and frustration on the antiferromagnetic phase of the two-dimensional Hubbard model

TL;DR

This study uses cellular dynamical mean-field theory to explore how interaction strength, doping, and frustration influence the antiferromagnetic phase in the 2D Hubbard model, revealing complex phase boundaries and in-gap states.

Contribution

It provides a detailed phase diagram of the 2D Hubbard model considering multiple parameters and identifies the nature of the AF transition and in-gap states, advancing understanding of doped Mott insulators.

Findings

Néel phase boundary is non-monotonic with U and doping

Frustration reduces Néel order more at small U

In-gap states are found at large U, with slight spin polarization

Abstract

Recent quantum-gas microscopy of ultracold atoms and scanning tunneling microscopy of the cuprates reveal new detailed information about doped Mott antiferromagnets, which can be compared with calculations. Using cellular dynamical mean-field theory, we map out the antiferromagnetic (AF) phase of the two-dimensional Hubbard model as a function of interaction strength $U$, hole doping $\delta$ and temperature $T$. The N\'eel phase boundary is non-monotonic as a function of $U$ and $\delta$. Frustration induced by second-neighbor hopping reduces N\'eel order more effectively at small $U$. The doped AF is stabilized at large $U$ by kinetic energy and at small $U$ by potential energy. The transition between the AF insulator and the doped metallic AF is continuous. At large $U$, we find in-gap states similar to those observed in scanning tunneling microscopy. We predict that, contrary to the Hubbard bands, these states are only slightly spin polarized.