| The spin Hall effect (SHE) is a name given to a collection of diverse phenomena which share two principal features: (i) longitudinal electric current flowing through a paramagnetic semiconductor or metallic sample leads to transverse spin current and spin accumulation of opposite sign at opposing lateral edges; (ii) SHE does not require externally applied magnetic field or magnetic ordering in the equilibrium state of the sample, instead it relies on the presence of spin-orbit (SO) couplings within the sample. This thesis elaborates on a new type of phenomenon within the SHE family, predicted in our recent studies [Phys. Rev. B 72, 075361 (2005); Phys. Rev. Lett. 95, 046601 (2005); Phys. Rev. B 72, 075335 (2005); Phys. Rev. B 73 , 075303 (2006); and Europhys. Lett. 77, 47004 (2007)], where pure spin current flows through the transverse electrodes attached to a clean finitesize two-dimensional electron gas (2DEG) due to unpolarized charge current injected through its longitudinal leads. If transverse leads are removed, the effect manifests as nonequilibrium spin Hall accumulation at the lateral edges of 2DEG wires. The SO coupling driving this SHE effect is of the Rashba type, which arises due to structural inversion asymmetry of semiconductor heterostructure hosting the 2DEG. We term the effect "mesoscopic" because the spin Hall currents and accumulations reach optimal value in samples of the size of the spin precession length---the distance over which the spin of an electron precesses by an angle pi. In strongly SO-coupled structures this scale is of the order of ∼100 nm, and, therefore, mesoscopic in the sense of being much larger than the characteristic microscopic scales (such as the Fermi wavelength, screening length, or the mean free path in disordered systems), but still much smaller than the macroscopic ones.;Although the first theoretical proposal for SHE, driven by asymmetry in SO-dependent scattering of spin-up and spin-down electrons off impurities, appeared in 1970s, it is only in the past few years that advances in optical detection of nonequilibrium magnetization in semiconductors have made possible the detection of such extrinsic SHE in groundbreaking experiments. The experimental pursuits of SHE have, in fact, been largely motivated by very recent theoretical speculations for several order of magnitude greater spin Hall currents driven by intrinsic SO mechanisms due to SO couplings existing not only around the impurity but also throughout the sample. The homogeneous intrinsic SO couplings are capable of spin-splitting the band structure and appear as momentum-dependent magnetic field within the sample which causes spin non-conservation due to precession of injected spins which are not in the eigenstates of the corresponding Zeeman term.;Besides deepening our understanding of subtle relativistic effects in solids, SHE has attracted a lot of attention since it offers an all-electrical way of generating pure spin currents in semiconductors. (Abstract shortened by UMI.). |
