Fidelity, fidelity susceptibility and von Neumann entropy to characterize the phase diagram of an extended Harper model

TL;DR

This paper uses fidelity, fidelity susceptibility, and von Neumann entropy to numerically characterize the phase diagram of an extended Harper model, revealing phase boundaries and critical phenomena.

Contribution

It introduces a comprehensive numerical analysis of these quantum information measures to fully characterize phase transitions in the extended Harper model.

Findings

Fidelity sharply changes at phase boundaries.

Critical exponents vary with system size for metal-metal transition.

Von Neumann entropy distinguishes metallic and insulating phases.

Abstract

For an extended Harper model, the fidelity for two lowest band edge states corresponding to different model parameters, the fidelity susceptibility and the von Neumann entropy of the lowest band edge states, and the spectrum-averaged von Neumann entropy are studied numerically, respectively. The fidelity is near one when parameters are in the same phase or same phase boundary; otherwise it is close to zero. There are drastic changes in fidelity when one parameter is at phase boundaries. For fidelity susceptibility the finite scaling analysis performed, the critical exponents $\alpha$, $\beta$, and $\nu$ depend on system sizes for the metal-metal phase transition, while not for the metal-insulator phase transition. For both phase transitions $\nu/\alpha\approx2$. The von Neumann entropy is near one for the metallic phase, while small for the insulating phase. There are sharp changes in von Neumann entropy at phase boundaries. According to the variation of the fidelity, fidelity susceptibility, and von Neumann entropy with model parameters, the phase diagram, which including two metallic phases and one insulating phase separated by three critical lines with one bicritical point, can be completely characterized, respectively. These numerical results indicate that the three quantities are suited for revealing all the critical phenomena in the model.