SPIDER (Spin Dependent Nanoelectronics)

Considering the great interest in spin-dependent transport and spin-electronics the reader will not be surprised to learn that there is yet another European project in this field. SPIDER is an ESPRIT Long Term Research project. Participants are from IMEC Leuven, Belgium, Thomson CSF, France, University of Groningen and MESA Research Institute, The Netherlands, Weizmann Institute, Israel, University of Nottingham and UCAM-Cavendish, United Kingdom, University of Regensburg and RWTH Aachen, Germany.

The aim of SPIDER is to study and develop quantum low-power nanoscale devices based on spin-dependent transport amd magnetic interactions in semiconductor quantum structures.

The main objectives are:

1) Vertical Spin Transistor

Here a prototype is already available. Monsma et al. (Phys. Rev. Lett. 74, 5260 (1995)), investigated vertical transport through a magnetic multilayer which forms the base electrode of a spin transistor. Electrons are injected via a Schottky contact formed at the interface with a silicon layer underneath. The electrons which make it through the base electrode are collected by a Schottky contact formed with a silicon layer on top.

The operation of the device as a magnetic field sensor is related to the spin-dependent scattering in the base electrode. A change in alignment of the magnetization of adjacent layers from anti-parallel to parallel results in a change in the mean free path of the electrons in the base electode. As a result the fraction of the injeced electrons which reaches the collector is changed, and thus the collector current is modulated.

The realization of this device was made possible by a special vacuum bonding technique which is required to connect the top silicon layer to the base electrode. Currently the effort is directed towards using AlSb or related III/V compound materials, which allow a better control of the energy of the injected and collected electrons.

2) Lateral Spin Transistor

Here the idea is to inject a spin-polarized current in a semiconductor channel in a lateral device. This is usually a two-dimensional electron gas (2DEG), which is located below the surface. This structure is very similar to a field effect transistor, with ferromagnetic contacts. A gate electrodes might be used to provide additional control of the spin-polarized current. The spin-orbit interaction in InAs based quantum wells may even make it possible to control the spin precession by changing the perpendicular electric field with a gate electrode. Several problems are encountered however. First of all, due to the lateral nature of the device, the spacing between the ferromagnetic contacts has to be relatively large (of the order of a micrometer), which requires that the spin-flip length in the semiconductor channel should have a similar, or preferably larger value. However, our current knowledge of spin scattering processes in semiconductors and semiconductor compounds is there is limited. Also the role of the interface between the ferromagnetic contacts and the semiconductor has to be considered. One of the questions is here whether one should inject electrons through a tunnel (or Schottky) contact, or whether one should make the contacts as transparent as possible.

3) Magnetically modulated semiconductor devices

In the third implementation magnetic materials are used to generate a modulated, position dependent magnetic field in the material underneath, where a two-dimensional electron gas is present. It differs from the other two implementations in that the charge carriers themselves are not spin-polarized. The idea is to employ the large magnetoresistance observed in these systems. This magneto-resistance arises from the typical way (along snake-orbits) in which the electrons move in a magnetic field which changes sign as a function of position. A large mean free path is favourable here. Therefore, in order to make the device operate at room temperature, a semiconductor system is required which has a relatively large mean free path at room temperature. InAs/AlSb quantum well systems are suitable candidates, and these systems are currently under investigation. The above is only a brief summary of the activities of the SPIDER project. Related topics studied by the SPIDER partners are: the study of magnetoresistance of magnetic grains embedded within a semiconductor host, the generation of spin-polarized carriers in III/V semiconductors with circularly polarized light, and the study of transport in hybrid systems with ferromagnets coupled to superconductors.

Compared with the OXSEN consortium one sees that SPIDER employs semi conductors as the host for the spin dependent transport. The spin polarization is produced by conventional (Fe, permalloy, Ni or Co) ferromagnets or by epitaxially grown ferromagnetic layers (such as MnAl), which are compatible with the III/V materials used. Although SPIDER works towards the realization of practical devices, nevertheless the exploration of the fundamental physical processes is a major effort. Given the common goal of understanding spin-dependent transport in optimized geometries and materials I think it is wise to maintain regular contacts between both SPIDER and OXSEN projects. I hope that this short contribution is a first step in this direction.

We currently have two positions available in the field of spin-dependent transport in a two-dimensional electron gas (a 4 year PhD position as well as a 1 1/2 year post-doc position)

For further information please write to the address below or contact me by

email:

Bart van Wees, Department of Applied Physics and Materials Science Centre, University of Groningen, Nijenborgh 4.13

9733 GL Groningen, The Netherlands

email: wees@phys.rug.nl

Bart van Wees - University of Groningen