Search and design of nonmagnetic centrosymmetric layered crystals with large local spin polarization

Abstract

Until recently, spin polarization in nonmagnetic materials was the exclusive territory of noncentrosymmetric structures. It was recently shown that a form of "hidden spin polarization" (named the "Rashba-2" or "R-2" effect) could exist in globally centrosymmetric crystals provided the individual layers belong to polar point group symmetries. This realization could considerably broaden the range of materials that might be considered for spin-polarization spintronic applications to include the hitherto "forbidden spintronic compound" that belongs to centrosymmetric symmetries. Here we take the necessary steps to transition from such general, material-agnostic condensed matter theory arguments to material-specific "design principles" that could aid future laboratory search of R-2 materials. Specifically, we (i) classify different prototype layered structures that have been broadly studied in the literature in terms of their expected R-2 behavior, including the Bi2Se3-structure type (a prototype topological insulator), MoS2-structure type (a prototype valleytronic compound), and LaBiOS2-structure type (a host of superconductivity upon doping); (ii) formulate the properties that ideal R-2 compounds should have in terms of combination of their global unit cell symmetries with specific point group symmetries of their constituent "sectors"; and (iii) use first-principles band theory to search for compounds from the prototype family of LaOBiS2-type structures that satisfy these R-2 design metrics. We initially consider both stable and hypothetical M′OMX2 (M': Sc, Y, La, Ce, Pr, Nd, Al, Ga, In, Tl; M: P, As, Sb, Bi; X: S, Se, Te) compounds to establish an understanding of trends of R-2 with composition, and then indicate the predictions that are expected to be stable and synthesizable. We predict large spin splittings (up to ∼200meV for holes in LaOBiTe2) as well as surface Rashba states. Experimental testing of such predictions is called for. © 2015 American Physical Society.