Spin properties of low-dimensional semiconductor structures

It is obvious for years now that the development of microelectronics eventually will slow down as basic physical limitations, particularly due to shrinking device size, are approached. Therefore, as an alternative for some applications, the utilization of the magnetic moment of an electron (characterized by its spin) is considered instead of its charge which is used at present. Among the possible advantages of "spintronics" there might be less power dissipation and higher speed, since a logical operation could be realized by inverting the spin of an electron without its (dissipative) displacement. In addition, spin based "bits" could have long lifetimes without supply of electric power. Then there is also the idea of realising a "quantum computer" based on the manipulation and interaction of spins. Such a computer is expected of being capable to solve problems of a complexity that cannot be solved with present computers in reasonable time. In order to realize a spintronic device one should be able to prepare a specific spin state, to manipulate it and finally to test its state after some time. In any case, the spin coherence time must be much longer than the time required for manipulation. In this project, we make use of our experience with the spin resonance of electrons in Si quantum wells to investigate the suitability of Si-based structures for spintronic devices. Electrons in a 10 nm wide Si layer keep their spin information up to a few microseconds, limited by potential gradients due to asymmetric doping. Avoiding these we expect a spin lifetime of 50..100 microseconds, whereas the time required for a spin flip by ESR is of the order of 10 ns. For laterally confined electrons the life time will be even longer. This ratio of life time and manipulation time appears to be higher than in any other semiconductor. In this project, we will investigate (i) the effect of external electric fields and lateral confinement of electrons in a quantum well on their spin properties, in particular on the spin relaxation mechanisms and on the g-factor. We expect that both quantities will depend on confinement that can be adjusted by gate voltages. This will allow tuning of the g-factor and thus to "address" a particular electron (or set of electrons). We will also explore (ii) the possibilities to detect the spin state by resistivity measurements through the confined structure..

It is obvious for years now that the development of microelectronics eventually will slow down as basic physical limitations, particularly due to shrinking device size, are approached. Therefore, as an alternative for some applications, the utilization of the magnetic moment of an electron (characterized by its spin) is considered instead of its charge which is used at present. Among the possible advantages of "spintronics" there might be less power dissipation and higher speed, since a logical operation could be realized by inverting the spin of an electron without its (dissipative) displacement. In addition, spin based "bits" could have long lifetimes without supply of electric power. Then there is also the idea of realising a "quantum computer" based on the manipulation and interaction of spins. Such a computer is expected of being capable to solve problems of a complexity that cannot be solved with present computers in reasonable time. In order to realize a spintronic device one should be able to prepare a specific spin state, to manipulate it and finally to test its state after some time. In any case, the spin coherence time must be much longer than the time required for manipulation. In this project, we make use of our experience with the spin resonance of electrons in Si quantum wells to investigate the suitability of Si-based structures for spintronic devices. Electrons in a 10 nm wide Si layer keep their spin information up to a few microseconds, limited by potential gradients due to asymmetric doping. Avoiding these we expect a spin lifetime of 50..100 microseconds, whereas the time required for a spin flip by ESR is of the order of 10 ns. For laterally confined electrons the life time will be even longer. This ratio of life time and manipulation time appears to be higher than in any other semiconductor. In this project, we will investigate (i) the effect of external electric fields and lateral confinement of electrons in a quantum well on their spin properties, in particular on the spin relaxation mechanisms and on the g-factor. We expect that both quantities will depend on confinement that can be adjusted by gate voltages. This will allow tuning of the g-factor and thus to "address" a particular electron (or set of electrons). We will also explore (ii) the possibilities to detect the spin state by resistivity measurements through the confined structure.