Intrinsic axion insulating behavior in antiferromagnetic MnBi$_6$Te$_{10}$

TL;DR

This paper demonstrates that MnBi6Te10 exhibits intrinsic axion insulating behavior due to antiferromagnetic order, leading to a gapped surface state protected by lattice symmetries, combining theoretical, computational, and experimental evidence.

Contribution

It reveals the intrinsic axion insulating state in MnBi6Te10, a magnetic topological insulator, protected by inversion and fractional translation symmetries, with experimental confirmation of surface gapping.

Findings

Identification of a quantized axion term in MnBi6Te10

Experimental observation of gapped Dirac surface states

Theoretical and computational validation of symmetry protection

Abstract

A striking feature of time reversal symmetry (TRS) protected topological insulators (TIs) is that they are characterized by a half integer quantum Hall effect on the boundary when the surface states are gapped by time reversal breaking perturbations. While time reversal symmetry (TRS) protected TIs have become increasingly under control, magnetic analogs are still largely unexplored territories with novel rich structures. In particular, topological magnetic insulators can also host a quantized axion term in the presence of lattice symmetries. Since these symmetries are naturally broken on the boundary, the surface states can develop a gap without external manipulation. In this work, we combine theoretical analysis, density functional calculations and experimental evidence to reveal intrinsic axion insulating behavior in MnBi6Te10. The quantized axion term arises from the simplest possible mechanism in the antiferromagnetic regime where it is protected by inversion symmetry and a fractional translation symmetry. The anticipated gapping of the Dirac surface state at the edge is subsequently experimentally established using Angle Resolved Spectroscopy. As a result, this system provides the magnetic analogue of the simplest TRS protected TI with a single, gapped Dirac cone at the surface.