Testing the topological nature of end states in antiferromagnetic atomic chains on superconductors

TL;DR

This study develops an experimental method to distinguish Majorana modes from trivial end states in antiferromagnetic atomic chains on superconductors, revealing that observed end states are likely trivial and not topological Majorana modes.

Contribution

The paper introduces a defect-based perturbation technique to identify the topological nature of end states in magnetic chains, providing a practical approach to verify Majorana modes.

Findings

End states can be split by local defects, indicating trivial origin.

Large spin-orbit coupling is needed to achieve topological phases in antiferromagnetic chains.

Method effectively differentiates topological Majorana modes from trivial end states.

Abstract

Edge states forming at the boundaries of topologically non-trivial phases of matter are promising candidates for future device applications because of their stability against local perturbations. Magnetically ordered spin chains proximitized by an s-wave superconductor are predicted to enter a topologically non-trivial mini-gapped phase with zero-energy Majorana modes (MMs) localized at their ends. However, the presence of non-topological end states mimicking MM properties can spoil their unambiguous observation. Here, we report on a method to experimentally decide on the MM nature of end states observed for the first time in antiferromagnetic spin chains. Using scanning tunneling spectroscopy, we find end states at either finite or near-zero energy in Mn chains on Nb(110) or Ta(110), respectively, within a large minigap. By introducing a locally perturbing defect on one end of the chain, the end state on this side splits off from zero-energy while the one on the other side doesn't - ruling out their MM origin. A minimal model shows that, while wide trivial minigaps hosting such conventional end states are easily achieved in antiferromagnetic spin chains, unrealistically large spin-orbit couplings are required to drive the system into the topologically nontrivial phase with MMs. The methodology of perturbing chains by local defects is a powerful tool to probe the stability of future candidate topological edge modes against local disorder.