Spintronics or spin electronics has been one of the most rapidly developing research areas in condensed matter physics. Exploiting the electron's spin degree of freedom in addition to its charge is hoped to yield new and novel technological applications. The ability to generate a non-equilibrium spin population, to manipulate the spins, and to be able to detect them have been the most intensely studied subjects in this research area.; In this dissertation, spin injection from an Fe (Iron) ferromagnetic spin polarizing contact into two dimensional and zero dimensional semiconductor heterostructures were investigated using a device called a "Spin-LED" also known as a Spin Light Emitting Diode. Due to the ferromagnetic layer's high Curie temperature (∼1043 K), Fe is a very promising candidate for practical applications at room temperature. This dissertation is comprised of three separate experimental studies outlined as follows. (i) Effects of confined charge carriers in the quantum well of a spin-LED. In this section we have compared the spin-LEDs in which the GaAs quantum well is populated by electrons to those in which the GaAs quantum well does not contain any free electrons. The former's characteristic electroluminescence (EL) spectra and current dependence on the circular polarization differs when compared to the latter. Additional features in the EL spectra are identified and a model which explains the characteristics of these type of spin-LEDs is proposed. (ii) Spin injection from a ferromagnetic Fe layer into GaAs quantum wells grown in the (110) crystallographic direction (as opposed to the normal (100) direction) are presented in this section. The motivation of longer spin lifetimes in (110) oriented quantum wells compared to those grown in the (100) direction, has prompted an investigation into the spin injection efficiency of the (110) based devices. Although the temperature dependence of the circular polarization is a less sensitive function than that of a typical (110) based device, the magnitude of the circular polarization has been consistently much smaller than what has been obtained for its (100) counterparts, in agreement with theoretical calculations. One of the main reasons for the lower than expected magnitude in the spin polarization from (110) LEDs is predicted to be the poor matching of symmetry bands between the ferromagnetic layer and the (110) GaAs quantum well which results in higher than normal spin scattering rates at the interface. (iii) Spin injection from Fe contacts into zero dimensional InAs self assembled quantum dots. With bandgap engineering, zero dimensional systems have become the focus of very intensive research as of late. In addition to potential applications in telecommunications, lasers, photodetectors, and quantum computing, zero dimensional quantum dot systems are often rich in exotic effects. We have achieved spin injection from ferromagnetic Fe contacts into InAs quantum dots at room temperature. Although the circular polarization is small (5%), it is expected to be improved by attaining optimal growth conditions. At low temperatures (5-60K) these spin-LEDs have shown a very sharp reduction in the circular polarization in the vicinity of a specific magnetic field value. Theoretical studies on spin-orbit interaction in quantum dot systems have predicted that a sharp enhancement in spin relaxation rate due to the mixing of spin states at specific magnetic field values. This results theoretically from the Rashba spin-orbit interaction term. Our experimental results along with their characteristics are in agreement with theoretical works.
展开▼
