Imaging currents in HgTe quantum wells in the quantum spin Hall regime

TL;DR

This study visualizes and confirms the existence and robustness of edge channels in HgTe quantum wells exhibiting the quantum spin Hall effect, using magnetic field imaging with a scanning SQUID.

Contribution

It introduces a new method to image current flow in large quantum spin Hall devices, demonstrating the robustness of edge channels despite disorder and bulk conduction.

Findings

Edge current flows along the device boundary in the QSH regime

Edge channels persist despite disorder and increased temperature

Method enables characterization of new QSH materials

Abstract

The quantum spin Hall (QSH) state is a genuinely new state of matter characterized by a non-trivial topology of its band structure. Its key feature is conducting edge channels whose spin polarization has potential for spintronic and quantum information applications. The QSH state was predicted and experimentally demonstrated to exist in HgTe quantum wells. The existence of the edge channels has been inferred from the fact that local and non-local conductance values in sufficiently small devices are close to the quantized values expected for ideal edge channels and from signatures of the spin polarization. The robustness of the edge channels in larger devices and the interplay between the edge channels and a conducting bulk are relatively unexplored experimentally, and are difficult to assess via transport measurements. Here we image the current in large Hallbars made from HgTe quantum wells by probing the magnetic field generated by the current using a scanning superconducting quantum interference device (SQUID). We observe that the current flows along the edge of the device in the QSH regime, and furthermore that an identifiable edge channel exists even in the presence of disorder and considerable bulk conduction as the device is gated or its temperature is raised. Our results represent a versatile method for the characterization of new quantum spin Hall materials systems, and confirm both the existence and the robustness of the predicted edge channels.