Spin Mesoscopics
EU RTN "Computational Magnetoelectronics"
NEDO "NanoMagnetoelectronics"
ESF Working Group "Magneto electronics"
Joint Workshop
on "Spin Mesoscopics"
University of Twente, Enschede, March 15 - March 18, 2003
supported by:
EU RT-Network on "Computational Magnetoelectronics"
NEDO Network "NanoMagnetoelectronics"
ESF Working Group "Magnetoelectronics"
MESA+ Research Institute, University of Twente
A workshop on "Spin Mesoscopics" was organized by Paul Kelly (Twente) and Gerrit Bauer (Delft) on the campus of the University of Twente over the weekend of March 15-183, 2003 and was attended by 63 participants. It was sponsored by the RTN "Computational Magnetoelectronics", the NEDO Network "NanoMagnetoelectronics", the ESF Working Group "Magnetoelectronics" and by the MESA+ Research Institute, University of Twente.
One of the principal aims of the workshop was to identify important basic physics issues which need to be resolved if spin-transport devices are to be reduced in size, and especially those where materials-specific computational techniques can play an important role. A topic discussed in considerable detail was current-induced magnetization reversal (CIMR), an alternative to the Lorentz switching of memory elements in magnetic random-access memory (MRAM) devices. We were fortunate to have some of the principal experimentalists working on CIMR attend the workshop: Buhrmann (Cornell), Sun (IBM, Yorktown) and Fert (Orsay) presented results of recent experiments on CIMR in very small structures. These talks were complemented with theoretical presentations by Zhang (Columbia, Missouri), Edward (Imperial College, London), Mathon (City U. London) and Stiles (NIST). The talk by Pratt (Michigan State) on measuring the angular magnetoresistance was closely related because the two effects involve transport between non-collinearly aligned magnets and a "spin-mixing" conductance is central to describing both. The same spin-mixing also formed an important element in the talks on spin-pumping by Brataas (Trondheim), on magnetization dynamics by Tserkovnyak (Harvard) and on enhanced Gilbert damping by Zwierzycki (Twente). The liveliness of the discussions during and after these talks was a good indicator of the interest and controversy raised by this subject.
Researchers very much want to realize spin-injection into semiconductors in order to add a spin degree of freedom. Molenkamp (Wiirzburg) reviewed his work using II-VI dilute magnetic semiconductors (DMS) as spin injectors while III-V DMSs were discussed by Ohno (Tohoku U. Sendai) and by Turek (Brno). Van Roy (IMEC, Leuven) discussed the difficulty of detecting the spin-polarization achieved on spin-injection into a semiconductor through tunnel junctions. Filip (Eindhoven) talked about the possibility of using a magnetic semiconductor or insulator as a spin filter. van Wees (Groningen) described his experiments of spin-injection and -detection through a tunnel barrier into one- and zero-dimensional non-magnetic metals and subsequent observation of the spin-precession. Jansen (Twente) and Parkin (IBM Almaden) introduced spin-dependent hot electron transport as the basic principle underlying the operation of hybrid devices such as the spin valve transistor (SVT) and the magnetic tunnel transistor (MTT). The difficulty of studying spin-transport through tunnel junctions in a realistic way was underlined by Wortmann (Jiilich) who discussed the difficulties presented by the "simple" model system of Fe/Mg0/Fe. A recent experimental observation of oscillations of the TMR when the thickness of a Cu "dusting" layer is varied was addressed theoretically using a tight-binding model and including disorder by Itoh (Nagoya). The state-of-the art in modelling disorder with first-principles methods was reviewed for a variety of systems by Xia (Beijing/Twente) while Ebert (Miinchen) illustrated for ferromagnet/semiconductor heterostructures the wide range of information which can be gained on such systems using a variety of band structure methods.
Ways of circumventing the problems bedevilling efficient spin-injection using half-metallic ferromagnets were discussed theoretically by Chioncel (Nijmegen) while Mavropoulos (Jiilich) discussed theoretical calculations for a new class of zinc-blende compound which would be compatible with III-V semiconductors and are predicted to be half-metallic ferromagnets. Temmermann (Daresbury) reviewed the most important issues relating to half-metallic ferromagnetism of double perovskite structured materials.
The ultimate in miniaturization is achieved when devices have atomic dimensions. The state of the art of measuring electronic transport though a single molecule was presented by von Lohneysen (Karlsruhe) and calculations of spin-transport through atomic wires were discussed by Bagrets (Halle).
Paul Kelly
More details may be found in newsletter 57 from page 7