Imp Links
Curent Research Areas
Spin injection in semiconductors
Spin, like charge, is an intrinsic property of electrons. However, the spin of the electron is not readily used in devices, as its control can be inefficient or difficult in certain materials of interest, such as in semiconductors. Spintronics research aims to incorporate the use of electron spin in electronic devices.
There are three main challenges to integration of spins into everyday devices. Firstly, the spins must be injected from a polarized source (such as a ferromagnetic metal) into the (usually) semiconductor channels of the e.g. transistor device. Then, the polarized electron must be able to travel in the semiconductor without losing its polarization information. And finally, one must be able to detect (and use) the polarization of the electron at the other end of the semiconductor channel.
Schematic of a spintronics device, with two ferromagnetic electrodes deposited on to a semiconductor. Also depicted are the three main step required for spintronics devices: I: injection, II: travel/manipulation, and III: detection.
The first step – injection – already presents several challenges. Even in an ideal device, the infamous “impedance mismatch” obstacle will prevent efficient injection from a metallic source to a semiconductor. One way of circumventing this is actually to add a tunnelling barrier between the injector and the semiconductor. This can either be a simple Schottky barrier, or an artificial oxide barrier such as MgO, both of which will also aid spin filtering at the interface. Another solution to the mismatch obstacle is to use a ferromagnetic semiconductor as an injector, but these tend only to work at very low temperatures. Transport along a semiconductor channel depends primordially on the material and its quality. Spins are known to have a long diffusion length – the length over which the electrons maintain their spin information unchanged – in several semiconductors. The third step – detection – can be achieved electrically through magneto-resistance measurements in a device. Such measurements detect a change in voltage proportional to the relative alignment of magnetic electrodes, and thus feedback on the spin signal at the point of measurement.
We are studying the efficient achievement of these three fundamental steps in various semiconductors, and plan on using lateral spin valve structures to do so. In such structures, two ferromagnetic electrodes lie across a non-magnetic material (in our case, the semiconductor). Spin accumulation can be generated in the semiconductor, and this spin-polarized signal can propagate along the channel to the second ferromagnetic electrode, even in the absence of a charge current. By varying the injector materials and interfaces, doping concentrations and geometries in our samples, our research will aim to determine an efficient way to inject spins into semiconductors.In addition our research also focuses on hybrid spin filtering devices consisting of MgO tunneling barrier and EuO for spin injection in semiconductor such as Si, GaAs, InGaAs and InAs 2DEG structure.
Contact Person: Dr. Jean-Baptiste Laloe, Marc van Veenhuizen and Dr. J.Y. Chang