Thermodynamics of an Exactly Solvable Model for Superconductivity in a Doped Mott Insulator

TL;DR

This paper analyzes the thermodynamics and electronic properties of an exactly solvable doped Mott insulator model, revealing unique first-order superconducting transitions and other phenomena distinct from BCS theory, with implications for cuprate superconductors.

Contribution

It provides a detailed thermodynamic and electronic analysis of the HK model, highlighting novel first-order transition features and the role of Mottness in superconductivity.

Findings

First-order transition at a critical interaction strength.

Emergence of a tri-critical point separating transition types.

Enhanced condensation energy and altered heat capacity behavior.

Abstract

Computing superconducting properties starting from an exactly solvable model for a doped Mott insulator stands as a grand challenge. We have recently shown that this can be done starting from the Hatsugai-Kohmoto (HK) model which can be understood generally as the minimal model that breaks the non-local $\mathbb Z_2$ symmetry of a Fermi liquid, thereby constituting a new quartic fixed point for Mott physics [Phillips et al., Nature Physics 16, 1175 (2020); Huang et al., Nature Physics (2022)]. In the current work, we compute the thermodynamics, condensation energy, and electronic properties such as the NMR relaxation rate $1/T_1$ and ultrasonic attenuation rate. Key differences arise with the standard BCS analysis from a Fermi liquid: 1) the free energy exhibits a local minimum at $T_p$ where the pairing gap turns on discontinuously above a critical value of the repulsive HK interaction, thereby indicating a first-order transition, 2) a tri-critical point emerges, thereby demarcating the boundary between the standard second-order superconducting transition and the novel first-order regime, 3) Mottness changes the sign of the quartic coefficient in the Landau-Ginzburg free-energy fuctional relative to that in BCS, 4) as this obtains in the strongly interacting regime, it is Mott physics that underlies the generic first-order transition, 5) the condensation energy exceeds that in BCS theory suggesting that multiple Mott bands might be a way of enhancing superconducting, 6) the heat-capacity jump is non-universal and increases with the Mott scale, 7) Mottness destroys the Hebel-Slichter peak in NMR, and 8) Mottness enhances the fall-off of the ultrasonic attenuation at the pairing temperature $T_p$. As several of these properties are observed in the cuprates, our analysis here points a way forward in computing superconducting properties of strongly correlated electron matter.