Anomalous localization at the boundary of an interacting topological insulator

TL;DR

This paper explores how strong interactions and disorder at the boundaries of topological insulators can lead to a novel anomalous many-body localized phase that preserves boundary anomalies, with potential experimental signatures.

Contribution

It introduces the concept of Anomalous Many Body Localized (AMBL) phases at TI boundaries, analyzing their emergence and properties under strong interactions and disorder.

Findings

AMBL phase preserves boundary anomaly

Anomalous Kramers parity switching observed

Fails for 3D strong TIs, indicating different anomaly restrictions

Abstract

The boundary of a topological insulator (TI) hosts an anomaly restricting its possible phases: e.g. 3D strong and weak TIs maintain surface conductivity at any disorder if symmetry is preserved on-average, at least when electron interactions on the surface are weak. However the interplay of strong interactions and disorder with the boundary anomaly has not yet been theoretically addressed. Here we study this combination for the edge of a 2D TI and the surface of a 3D weak TI, showing how it can lead to an "Anomalous Many Body Localized" (AMBL) phase that preserves the anomaly. We discuss how the anomalous Kramers parity switching with pi flux arises in the bosonized theory of the localized helical state. The anomaly can be probed in localized boundaries by electrostatically sensing nonlinear hopping transport with e/2 shot noise. Our AMBL construction in 3D weak TIs fails for 3D strong TIs, suggesting that their anomaly restrictions are distinguished by strong interactions.