Spin-Selective Quantum Transport through Heterointerfaces

Giant magneto-resistance (GMR) and tunneling magneto-resistance (TMR) arguably are the most spectacular examples to demonstrate the potential of spintronics for realizations of novel fast and low-energy devices. Driven by potential technological applications, spintronics has indeed become a florishing field of research which has developed rapidly into a hot topic of solid-state physics, posing challenges to electron transport physics, ultra-fast spectroscopy, and nano-magnetism alike. While traditional devices have mainly been based on one or two classes of materials, such as semiconductors and oxides in conventional electronic devices, superconductors and insulators in SQUIDS, or ferromagnetic materials, spintronic effects are based on (nano- and hetero) structures which generally are synthesized from materials with vastly different host properties, consisting of a combination of nonmagnetic layers, such as semiconductors, insulators, and metals, and magnetic layers, such as semi-magnetic semiconductors, ferromagnets, half-metallic ferromagnets, and/or antiferromagnets. A sound understanding of the behavior of the spin degree of freedom of charge carriers (transport electrons and holes) at heterointerfaces is therefore one of the key prerequisites for a microscopic understanding of basic processes in spintronics, such as TMR, GMR, spintorque effects, and - particularly- spin injection. Here we propose a theoretical study of spin-selective charge transport through heterointerfaces between semiconductors, semi-magnetic semiconductors, insulators, and (half-metallic) ferromagnets. We set up a theory which combines electronic structure calculations with (steady-state) quantum transport based on Green`s function theory to study spin-selective charge transport through (multiple) heterointerfaces. Self-consistent transport models of increasing complexity are developed. Electronic structure calculations for halfmetallic ferromagnets are performed within the localdensity approximation plus dynamical mean-field theory (LDA+DMFT) providing an input for the transport calculations. A key ingredient for transport calculations, the band-offsets between different materials, will be determined by the former. A second part of work carried out in parallel will investigate microscopically limitations of fast magnetization switching by spin-torque effects and the inverse Faraday effect. Spintorque allows switching of the magnetization by spin-polarized currents. For selected configurations, we will develop a model for this effect and determine optimized switching currents within time-dependent optimization theory developed in our group. Finally, we will investigate theoretically optimization of ultrafast all-optical switching in antiferromagnets. The proposed work constitutes a newly initiated joint effort within Project NAWI Graz between the University of Graz and the Graz University of Technology (TU Graz), linking the expertise in transport theory from the former with the expertise in electronic structure calculation for highly correlated systems from the latter.

Spintronics is a fairly recent field of physics in which the spin degree of freedom of electrons, rather than their charge, is at the center of attention. Advances in material science and demands from technology, such as device miniaturization and low cost paired with energy efficiency, have initiated a closer look at the potential of the spin degree for its use in future electronic devices. There are several focal points in the field, such as the coupling of spin and charge, information storage in spin systems, as well as the generation and subsequent utilization of spin polarized currents. In this project we have concentrated on several of these issues in connection with very specific material systems: heterostructures consisting non-magnetic semiconductors, magnetic semiconductors, half-metals, and/or (metallic) ferromagnets.Heterostructures are layered hybrid nanostructures, ideally lattice matched at the interfaces, which in a particular combination lead to new material properties, different from the properties of the constituents. These frequently arise from quantum effects. In the present project we have concentrated our theoretical studies mainly on heterostructures consisting of layers of GaAs, AlGaAs, CrAs, MnAs and their alloys, and, to a lesser extent, to normal-ferromagnetic-metal heterostructures. As to the former, a self-consistent theoretical model for the current response was developed and used to explore spin-selectivity in the transport through AlGaAs/GaMnAs and GaAs/CrAs heterostructures. In the main part of this project we have studied the dynamics of ferromagnetic order in GaMnAs quantum wells and its importance to the current-voltage characteristics, in particular, regarding its spin polarization. We found that while bias control of ferromagnetic order should be feasible at very high hole concentrations, it is strongly affected by (substitutional) disorder inevitably present in GaMnAs. This model has assisted interpretation of experimental work in explaining the absence of negative differential resistivity and exchange splitting in the system under bias. Strong spin polarization is predicted from (fcc-GaAs)/CrAs heterostructures which may persist even up to room temperature. Our calculations show that such heterostructures may be feasible (in a metastable state), however, experiments have remained controversial. Within this project we have also developed a theoretical tool for the evaluation and design of spin-transfer torque devices based on metallic heterostructures in the diffusive regime, a fundamental spin-selective Boltzmann transport approach, and a variational cluster approach for the nonequilibrium steady state for strongly correlated many-body systems.