Interpreting current-induced spin polarization in topological insulator surface states

TL;DR

This paper critically examines claims of large current-induced spin polarization in topological insulator surface states, showing that such polarization is minimal and that previous experimental interpretations are flawed.

Contribution

It provides a theoretical and experimental critique demonstrating that observed signals are not indicative of significant spin polarization in topological insulators.

Findings

Current-induced spin polarization in TIs is minimal.

Magnetic field effects can be mimicked in trivial metal films.

Previous claims of spin-momentum locking are likely misinterpretations.

Abstract

Several recent experiments on three-dimensional topological insulators claim to observe a large charge current-induced non-equilibrium ensemble spin polarization of electrons in the helical surface state. We present a comprehensive criticism of such claims, using both theory and experiment: First, we clarify the interpretation of quantities extracted from these measurements by deriving standard expressions from a Boltzmann transport equation approach in the relaxation-time approximation at zero and finite temperature to emphasize our assertion that, despite high in-plane spin projection, obtainable current-induced ensemble spin polarization is minuscule. Second, we use a simple experiment to demonstrate that magnetic field-dependent open-circuit voltage hysteresis (identical to those attributed to current-induced spin polarization in topological insulator surface states) can be generated in analogous devices where current is driven through thin films of a topologically-trivial metal. This result *ipso facto* discredits the naive interpretation of previous experiments with TIs, which were used to claim observation of helicity, i.e. spin-momentum locking in the topologically-protected surface state.