The spin Hall effect, Berry curvature, impurities, and application

After the first experimental observation of the spin Hall effect in semiconductors the topic attracted more and more interest from both experimental and theoretical point of view. The potential of the spin Hall effect to overcome the problem of spin current injection from a ferromagnet into a nonmagnetic material is an important reason for the intensive study of the effect in recent years. I will present my work, methods as well as results, on first principle calculations of the spin Hall effect in metals. This talk focuses on the semiclassical approach where intrinsic and extrinsic mechanisms can be separated naturally. The intrinsic mechanism is governed by the Berry curvature of the pure band structure whereas in the extrinsic case electron-impurity scattering has to be described quantum mechanically. We implemented both contributions where we made use of special features of the applied Green-function method and performed broad material scans to identify materials for possible applications. One particular application where the induced spin current is used to switch a ferromagnetic island I will introduce in more detail. Special features of real slabs, more relevant to the experimental situation, will be also discussed.