Homotopy, Symmetry, and Non-Hermitian Band Topology

TL;DR

This paper develops a unified classification of non-Hermitian band structures with symmetries using homotopy theory, revealing new stable topological phases and invariants in physical systems.

Contribution

It introduces a comprehensive framework for classifying non-Hermitian topologies with symmetries, encompassing band gaps and separations, and uncovers novel stable and fragile topological phases.

Findings

Reveals Abelian and non-Abelian topological phases in $\\mathcal{PT}$-symmetric systems.

Identifies invariants robust to symmetry-preserving perturbations.

Connects spontaneous $\\mathcal{PT}$ breaking to Chern-Euler and Chern-Stiefel-Whitney invariants.

Abstract

Non-Hermitian matrices are ubiquitous in the description of nature ranging from classical dissipative systems, including optical, electrical, and mechanical metamaterials, to scattering of waves and open quantum many-body systems. Seminal line-gap and point-gap classifications of non-Hermitian systems using K-theory have deepened the understanding of many physical phenomena. However, ample systems remain beyond this description; reference points and lines do not in general distinguish whether multiple non-Hermitian bands exhibit intriguing exceptional points, spectral braids and crossings. To address this we consider two different notions: non-Hermitian band gaps and separation gaps that crucially encompass a broad class of multi-band scenarios, enabling the description of generic band structures with symmetries. With these concepts, we provide a unified and comprehensive classification of both gapped and nodal systems in the presence of physically relevant parity-time ($\mathcal{PT}$) and pseudo-Hermitian symmetries using homotopy theory. This uncovers new stable topology stemming from both eigenvalues and wave functions, and remarkably also implies distinct fragile topological phases. In particular, we reveal different Abelian and non-Abelian phases in $\mathcal{PT}$-symmetric systems, described by frame and braid topology. The corresponding invariants are robust to symmetry-preserving perturbations that do not induce (exceptional) degeneracy, and they also predict the deformation rules of nodal phases. We further demonstrate that spontaneous $\mathcal{PT}$ symmetry breaking is captured by Chern-Euler and Chern-Stiefel-Whitney descriptions, a fingerprint of unprecedented non-Hermitian topology previously overlooked. These results open the door for theoretical and experimental exploration of a rich variety of novel topological phenomena in a wide range of physical platforms.