Any Axion Insulator Must be a Bulk Three-Dimensional Topological Insulator

TL;DR

This paper demonstrates that axion insulators must be bulk three-dimensional topological insulators and shows that zero Hall plateau observations are not definitive evidence of axion insulator states, highlighting the importance of bulk properties.

Contribution

The study provides a detailed magneto-transport analysis showing that axion insulators are inherently bulk 3D topological insulators and challenges previous assumptions linking zero Hall plateau to axion insulators.

Findings

Evidence for a quantized topological magneto-electric effect

Zero Hall plateau is not exclusive evidence of axion insulators

Axion electrodynamics signatures observed in magneto-transport data

Abstract

In recent attempts to observe axion electrodynamics, much effort has focused on trilayer heterostructures of magnetic topological insulators, and in particular on the examination of a so-called zero Hall plateau, which has misguidedly been overstated as direct evidence of an axion insulator state. We investigate the general notion of axion insulators, which by definition must contain a nontrivial volume to host the axion term. We conduct a detailed magneto-transport analysis of Chern insulators comprised of a single magnetic topological insulator layer of varying thickness as well as trilayer structures, for samples optimized to yield a perfectly quantized anomalous Hall effect. Our analysis gives evidence for a topological magneto-electric effect quantized in units of e$^2$/2h, allowing us to identify signatures of axion electrodynamics. Our observations may provide direct experimental access to electrodynamic properties of the universe beyond the traditional Maxwell equations, and challenge the hitherto proclaimed exclusive link between the observation of a zero Hall plateau and an axion insulator.