SPINTRA - Short Description

Recent progress in information storage and processing technology was able owing to rapid progress in nanotechnology. However, conventional electronics faces already natural borders associated with finite lithography resolution, atomic structure of the matter, quantum phenomena, etc. Thus, further development should be based on new ideas. One of such new ideas emerged after the discovery of giant magnetoresistance effect (GMR) in magnetic metallic multilayers more than a decade ago. It turned out immediately that the electron spin – playing a minor role in conventional electronics – may play a role comparable to that of electron charge. This discovery led to new kind of electronics, known as magnetronics or spintronics. The GMR effect found already application in magnetic field sensors, used in read/write heads of present-day computers. This allowed a further progress in storage density on hard discs. The related phenomenon found in tunnel junctions with ferromagnetic electrodes, and known as the tunnel magnetoresistance (TMR) effect, was used to construct magnetic random access memories (MRAMs). These and related applications proved importance and perspectives for spintronics.

There are two problems in spintronics, which determine functionality of any device. One of them is a source of spin polarized electrons at room temperature, and the other one is spin coherence length. As a source one may use ferromagnetic metals, which have sufficiently high Curie temperature. However, incompatibility of electronic structure of metals and semiconductors makes some difficulties with incorporating ferromagnetic metallic components into semiconductor electronics. This is a driving force for search of ferromagnetic semiconductors with sufficiently high Curie temperature. If this will be achieved, then both the source of spin polarized electrons and operation part of a device can be incorporated into a single element.

Another possibility is offered by hybrid structures, including nonmagnetic semiconducting as well as magnetic metallic or molecular components. In such a case ferromagnetic metallic components can be used as a source of spin polarized electrons, which are injected into an operating semiconducting part. Since the spin coherence length is relatively long in semiconductor structures, such hybrid systems may be of some interest for applications. The main problem here is the relatively low efficiency of the spin injection from ferromagnetic metallic part to a semiconductor, although some progress has already been achieved in recent years.