09.12.2011

Perfect layer for the right spin

PETRA III illuminates chemical structure of promising spintronics material

Ever faster and ever more efficient: the next generation of electronics is spintronics. It uses not only the electrical charge of electrons but also their spin. Now, at the world’s most brilliant X-ray source PETRA III at DESY, a group of scientists headed by Martina Müller from Forschungszentrum Jülich explored a promising material for spintronics: europium oxide, which is excellent to produce and readout electron currents with a defined spin in a silicon semiconductor. The scientists’ report appeared in the cover story of the journal "PSS Rapid Research Letters".

With spintronics, it is not only possible to produce faster components but also to create completely novel functions. A prominent example of spintronics on the basis of magnetic metals is the discovery of giant magnetoresistance (GMR), awarded with the Nobel Prize in 2007. Since 2005, this effect is used for read heads of computer hard disks, providing users with abundant memory. A next step would be the development of semiconductor based spintronics, e.g. the so-called spin transistors. “The goal is to find a combination of materials which can easily be integrated into current silicon technology,” Wolfgang Drube, DESY physicist and member of the scientists’ group explains. A suitable material would for example be europium oxide, which is structurally, chemically and electronically compatible with silicon.

The research made evident that an important step on this path was achieved by the Müller team. The materials scientists grew a 4.5 nanometre (millionth of a millimetre) thin europium oxide layer directly on silicon. “Europium oxide is one of the highly traded candidates for the so-called spin filters which allow injecting an electron current with high spin polarisation into a semiconductor” said Drube. “But this only works well when europium, in the form of europium oxide, is available as bivalent europium. If there is an admixture of trivalent europium, this will not work anymore.” Moreover, the silicon in layer production must not oxidise because this too would disturb the desired spin polarisation.

The chemical composition of the produced layers is not recognizable from outside, particularly as the freshly prepared europium oxide nanolayer is protected with a 4 nanometre thick aluminum cover. Photoelectron spectroscopy is generally well suited for a systematic chemical analysis. “However, with this method, I can generally only see the surface, i.e. just a few of the upper atom layers,” Drube explains. “Such a relatively thick structure, a whole eight and a half nanometres on top of the silicone, cannot be penetrated with the standard energies of the electrons.”

However, with a new facility at PETRA III at the Deutsches Elektronen-Synchtrotron DESY in Hamburg, it is possible to use the unique PETRA III brilliance to produce very fast photoelectrons to gaze deep into the samples, right down to the silicon layer.

The analysis showed that the Jülich scientists managed to grow a nearly pure layer of europium oxide directly on silicon, without unwanted admixtures. And the silicon too was nearly undisturbed. “This is decisive” Drube emphasises. „There is practically no silicon oxide.” At another inferior sample, the unwanted oxides are clearly discernible.

This sensitive analysis technology has great potential for the future – multilayers are a growing research field with regard to electronics applications. Frequently, it is a question of complex arrangements of many nanometre-thick layers. This goes along with the fact that the analysis depth of photoelectron spectroscopy was increased by the new facilities, of which there are only a few worldwide. “Currently, it is already possible to gaze some ten nanometres into such multilayers; this is about the scale on which these multilayers are produced,” Drube reports. “This fits together perfectly, and we expect a lot to happen in this field.”