Present trends in magnetism of solids and nanostructures call for novel techniques of manipulation of spin polarization which avoid the traditional application of external magnetic fields. This proposal is focused on theoretical investigation of phenomena and systems yielding alternative routes in this direction. We will study mechanisms leading to spin accumulation in nonmagnetic solids and to spin currents and spin torques in ferro- and ferrimagnets. Our effort will be focused on phenomena related to spin-orbit interaction (spin Hall and Nernst effects, spin-orbit torques) and to spin dynamics on a femtosecond scale (response of magnetization to short laser pulses). We will develop the relevant ab initio theory, which will be applied to complex bulk solids, layered structures and magnetic domain walls. Special attention will be paid to realistic systems with substitutional randomness. We will also provide detailed analysis of individual contributions to the resulting properties aimed at explanation of recent experiments and prediction of prospective materials and systems.
Aims of the project: Development of ab initio theory for efficient generation of spin accumulation, spin currents and torques based on spin-orbit interaction and ultrafast dynamics. Application of this theory to realistic systems with layered structures and substitutional disorder and to prediction of new materials.