Samarium Hexaboride: The First True 3D Topological Insulator?

Abstract

The recent theoretical prediction of a topologically protected surface state in the mixed-valent insulator SmB$_{6}$ has motivated a series of charge transport studies, which are presented here. It is first studied using a specialized configuration designed to distinguish bulk-dominated conduction from surface-dominated conduction. As the material is cooled below 4 K, it exhibits a crossover from thermally activated bulk transport to metallic surface conduction with a fully insulating bulk. The robustness and magnitude of the surface conductivity, as is manifest in the literature of SmB$_{6}$, is strong evidence for the topological insulator (TI) metallic surface states predicted for this material. This resolves a decades-old puzzle surrounding the low-temperature behavior of SmB$_{6}$. Next, the magnetotransport properties of the surface are investigated using a Corbino disk geometry, which can directly measure the conductivity of individual surfaces. Both (011) and (001) crystal surfaces show a strong negative magnetoresistance at all magnetic field angles, due primarily to changes in the carrier density. The low mobility value accounts for the failure so far to observe Shubnikov--de Haas oscillations below 95 T. Small variations in the mobility and temperature dependence suggest a suppression of Kondo scattering from native oxide-layer magnetic moments. At low fields, a dynamical field-sweep-rate-dependent hysteretic behavior is observed. It persists at the slowest sweep rates, and cannot be explained by quantum interference corrections; it is likely due to extrinsic effects such as the magnetocaloric effect or glassy ordering of the native oxide moments. Pulsed magnetic field measurements up to 60 T at temperatures throughout the crossover regime clearly distinguish the surface magnetoresistance from the bulk magnetoresistance. The bulk magnetoresistance is due to a reduction in the bulk gap with increasing magnetic field. Finally, small subsurface cracks formed in SmB$_{6}$ via systematic scratching or sanding results in a counter-intuitive increase in the electrical conduction due to the unique surface-conducting property of TIs, strengthening the building case for SmB$_{6}$'s topological nature. This material is attractive as a TI because its bulk is fully insulating at a readily achieved 2 K, but it presents a large number of scientific mysteries and experimental challenges for future research.