Symplectic-Amoeba formulation of the non-Bloch band theory for one-dimensional two-band systems

TL;DR

This paper extends the Amoeba formulation of non-Bloch band theory to one-dimensional two-band systems with Kramers degeneracies, enabling accurate spectrum computation despite boundary effects and symmetry constraints.

Contribution

It introduces a generalized Szeg"o's limit theorem for class AII$^owtie$ systems and demonstrates a method to overcome challenges from multiband degeneracies.

Findings

Successfully computes spectrum and localization length in complex systems

Overcomes limitations of conventional Amoeba formalism for Kramers pairs

Provides a numerical validation of the generalized theorem

Abstract

The non-Hermitian skin effect is a topological phenomenon, resulting in the condensation of bulk modes near the boundaries. Due to the localization of bulk modes at the edges, boundary effects remain significant even in the thermodynamic limit. This makes conventional Bloch band theory inapplicable and hinders the accurate computation of the spectrum. The Amoeba formulation addresses this problem by determining the potential from which the spectrum can be derived using the generalized Szeg\"o's limit theorem, reducing the problem to an optimization of the Ronkin function. While this theory provides novel insights into non-Hermitian physics, challenges arise from the multiband nature and symmetry-protected degeneracies, even in one-dimensional cases. In this work, we investigate one-dimensional two-band class AII$^\dagger$ systems, where Kramers pairs invalidate the conventional Amoeba formalism. We find that these challenges can be overcome by optimizing the band-resolved Ronkin functions, which is achieved by extrapolating the total Ronkin function. Finally, we propose a generalized Szeg\"o's limit theorem for class AII$^\dagger$ and numerically demonstrate that our approach correctly computes the potential and localization length.