Non-Hermitian symmetry breaking and Lee-Yang theory for quantum XYZ and clock models

TL;DR

This paper extends Lee-Yang theory to analyze quantum phase transitions in a variety of one-dimensional quantum models with complex magnetic fields, revealing symmetry breaking and fidelity zeros related to phase transitions.

Contribution

It generalizes the Lee-Yang framework to include XY, XXZ, XYZ, and clock models, demonstrating the universality of non-Hermitian symmetry breaking and fidelity zeros in diverse quantum systems.

Findings

Complex fields induce parity symmetry breaking in XY, XXZ, XYZ models.

Fidelity zeros appear within ordered phases due to ground state oscillations.

Finite-size scaling confirms critical exponents match analytical predictions.

Abstract

Lee-Yang theory offers a unifying framework for understanding classical phase transitions and dynamical quantum phase transitions through the analysis of partition functions and Loschmidt echoes. Recently, this framework is extended to characterize quantum phase transitions of quantum Ising models by introducing the concepts of non-Hermitian parity-symmetry breaking and fidelity zeros. Here, we generalize the theory by studying a broad class of quantum models, including the XY, the XXZ, the XYZ, and the $\mathbb{Z}_p$ clock models in one dimension, subject to a complex magnetic field. For the XY, XXZ and XYZ models, we find that the complex field breaks parity symmetry and induces oscillations of the ground state between the two parity sectors, giving rise to fidelity zeros within the ordered phases. For the $\mathbb{Z}_3$ clock model, the complex field splits the real part of the ground-state energy between the neutral sector ($q=0$) and the charged sectors ($q=1,2$), while preserving the degeneracy within the charged sector. Fidelity zeros arise only after projecting out one of the charged sectors. In contrast, for the $\mathbb{Z}_4$ clock model, the ground state oscillates between the neutral sector ($q=0$) and the charged sector ($q=2$), which directly gives rise to fidelity zeros. Finite-size scaling of these zeros yields critical exponents in full agreement with analytical predictions, demonstrating that this approach is applicable not only to the Ising model with $\mathbb{Z}_2$ symmetry, but also to more general Heisenberg-type models and systems with higher discrete symmetries.