The subject of this proposal is the theory of the decay of electron spin in semiconductors. It aims to understand the time that it takes for non-equilibrium spin distributions to disappear once they are created. All semiconductor materials will be considered, but there will be a particular focus on silicon and gallium arsenide, which have the most technological importance. Furthermore, the devices under consideration confine the motion of the electrons in a way that affects the spin decay. This will also be taken into account. There are multiple causes for this spin decay or relaxation. This will necessitate the development of computer programs to handle the complex interaction of different mechanisms.
Broad Impact Most of the important technological advances in the last 50 years have been closely associated with the progress of electronics, which, in physical terms, is the manipulation of electronic charge. In computer technology in particular, this manipulation is done in semiconductors. This steady advance has shown signs of slowing in the last decade or so, and the future will depend on the manipulation of electronic spin - the field of spintronics. This proposal aims at increasing our theoretical understanding of the behavior of electron spins in semiconductors. This area is the most promising for technology in the short term, since spintronics in semiconductors would allow rapid integration with current manufacturing techniques.
Intellectual merit Many basic mechanisms of spin relaxation in semiconductors are reasonably well understood. However, crucial difficulties remain. The effect of magnetic fields, doping ranges and particularly geometry of devices introduce complications that can change the relaxation in very significant ways. The work proposed here focuses on some novel physical mechanisms that we believe are central to understanding these unresolved issues. The most important of these novel insights is spin transfer between extended and localized states. This is not a spin relaxation in itself, since the transfer conserves spin, but the transfer accelerates spin relaxation. A second point is the unexpectedly large field dependence of the spin-phonon relaxation due to symmetry considerations and the phonon density of states. These insights will be incorporated into computer calculations of spin relaxation from all mechanisms to achieve quantitative understanding of relaxation times. This will be a major advance the theory of spin relaxation in semiconductors.
