Spin Hall conductivity in Bi$_{1-x}$Sb$_x$ as an experimental test of bulk-boundary correspondence

TL;DR

This study experimentally tests the bulk-boundary correspondence principle for spin currents in topological insulator Bi$_{1-x}$Sb$_x$, confirming that bulk properties determine spin Hall conductivity even with non-conserved spin currents.

Contribution

It provides the first experimental validation that bulk-boundary correspondence applies to non-conserved spin currents in topological materials.

Findings

Measured charge-to-spin conversion efficiency across compositions.

Found spin Hall conductivity matches tight-binding predictions.

Confirmed bulk states' spin-orbit entanglement links to SHC.

Abstract

Bulk-boundary correspondence is a foundational principle underlying the electronic band structure and physical behavior of topological quantum materials. Although it has been rigorously tested in topological systems where the physical properties involve charge currents, it remains unclear whether bulk-boundary correspondence should also hold for non-conserved spin currents. We study charge-to-spin conversion in a canonical topological insulator, Bi$_{1-x}$Sb$_x$, to address this fundamentally unresolved question. We use spin-torque ferromagnetic resonance measurements to accurately probe the charge-to-spin conversion efficiency in epitaxial Bi$_{1-x}$Sb$_x$~thin films of high structural quality spanning the entire range of composition, including both trivial and topological band structures, as verified using {\it in vacuo} angle-resolved photoemission spectroscopy. From these measurements, we deduce the effective spin Hall conductivity (SHC) and find excellent agreement with the values predicted by tight-binding calculations for the intrinsic SHC of the bulk bands. These results provide strong evidence that the strong spin-orbit entanglement of bulk states well below the Fermi energy connects directly to the SHC in epitaxial Bi$_{1-x}$Sb$_x$~films interfaced with a metallic ferromagnet. The excellent agreement between theory and experiment points to the generic value of analyses focused entirely on bulk properties, even for topological systems involving non-conserved spin currents.