The joint programme of planned research involves four main objectives. They are closely related, with many possibilities for cross-fertilization.

Spin injection into semiconductors is a vital ingredient for any emerging spintronics technology. However, despite considerable research efforts over the past decade, initial progress has been remarkably slow. Very recently, massive spin injection from a II-VI dilute magnetic semiconductor into a non-magnetic III-V LED has been demonstrated. While these so-called Spin-LEDs offer many possible applications for electro-optic technology, what spintronics still needs is a lateral all-electrical device technology. We aim to lay the ground stones for such a technology by exploring, experimentally and theoretically, lateral spin injection from the two types of spin injectors that seem most promising:

(1a) dilute magnetic semiconductors (either ferromagnetic ones, or paramagnetic ones that are magnetized by nearby bias magnets);

(1b) conventional metallic ferromagnets, using high-impedance junctions to the semiconductor to overcome the so-called 'impedance-mismatch' problem that has hitherto hampered this injection scheme.

From the emerging spintronic nanodevices, we expect progress in our abilities to manipulate the electron's spin in semiconductors and our understanding of the underlying phenomena. Issues to be studied (some of which will be applied in those parts of Objective 3 that involve semiconductors) include:

- decoherence and diffusion lengths of spin currents in semiconductors;

- transport properties of spin transistors consisting of these new materials (influence of Coulomb blockade, spin excitations and nonequilibrium effects);

- entanglement of spin states (spin-qubits) in quantum dots and wires.

We shall study spin-dependent transport in hybrid structures involving a combination of ferromagnetic (F) and normal (N) or superconducting (S) materials. The interplay between the different types of interactions and correlations present in each can produce a host of interesting spin-dependent effects, many of which have direct potential for applications (e.g. the GMR effect in layered ferromagnet-metal hybrid structures). In particular, the following studies are planned:

(2a) General hybrid structures: Development of a general semiclassical theory for spin-dependent transport in hybrid structures, as well as the equivalent circuit theory, and its application to some of the points listed below.

(2b) FN structures: Experimental studies of the spin flip length in the metal, and how it can be increased by changing material and/or junction parameters. Development of devices exploiting the non-local aspects of spin accumulation. Theoretical studies of the interplay of spin and heat transport in thermoelectric effects.

(2c) FS structures: Exploration of the interplay between spin coherence and electron-hole coherence. This includes proximity effect in the ferromagnet, spin injection into the superconductor, spin life-time of quasiparticles in the superconductor, nonequilibrium effects due to an applied bias voltage.

This objective has a major overlap with the previous two, but the emphasis here is on confined geometries. They will be used both as tools to study fundamental aspects of spin-dependent transport (e.g. spin coherence times, or many-body effects such as the Kondo effect or spin-charge separation), and to develop and optimise device applications (e.g. spin transistors, spin qubits).

(3a) Quantum Dots: We shall explore and exploit the unique possibilities offered by quantum dots to manipulate and utilize the spin of electrons in individual quantum states. One goal is to build a 100% spin filter by Zeeman-splitting the energy levels of a quantum dot in a large (in-plane) magnetic field, and to test it by studying the interference of two such dots in an Aharonov-Bohm ring. Another goal, crucial for recent proposals to use quantum dots as spin qubits, is to study, by experiment (electron spin resonance) and theory, the single-spin and two-spin decoherence times and to identify and minimize the sources of decoherence. Theoretical proposals will be worked out for determining decoherence times (including that of a single electron spin the holy grail in spintronics) via transport and noise measurements. In the context of quantum computation, theories will be developed for coherent spin dynamics generated by electron-spin resonance sources, and for implementing an active error feedback control loop via an external macroscopic circuit. Coherent spin dynamics will also be studied in the context of the recently-observed Kondo effect in quantum dots, which will be explored in novel situations (multi-level dots, multi-dot systems, dots in an Aharonov-Bohm ring, strong magnetic fields, strong nonequilibrium). We shall also explore dots with non-spin-conserving barriers, and dots near spin-related instability points (e.g. the Stoner instability).

(3b) Quantum Wires: The prediction of spin-charge separation in 1D systems will be explored, by measuring the dispersion relation of the spin-dependent excitations in 1D semiconductor wires, and the magnetoresistance of a carbon nanotube between two leads (ferromagnetic and ferromagnetic or superconducting). Detailed theories for the corresponding geometries will be developed.

(3c) Quantum Point Contacts: Conductance quantization and the spin-valve magnetoresistance in magnetic quantum point contacts (and also quantum wires) will be analyzed theoretically. Realistic descriptions, based on band structure calculations, will be developed, and the role of defects at the sample/lead interface will be investigated. In particular, the influence of magnetic impurities on energy relaxation in strongly nonequilibrium transport, in the presence of a large applied bias and strong microwave irradiation, will be analysed and compared to experiment.

(3d) Magnetic Single-Electron Transistors: The interplay between Coulomb blockade and spin-dependent effects (e.g. tunnelling magnetoresistance, spin accumulation) in single-electron transistors with ferromagnetic and superconducting components will be explored. Particular emphasis will be put on magnetoresistance phenomena with potential for applications, e.g. in field sensors, magnetically controlled diodes, etc.

(3e) Hybrid multilayers: Hybrid layered systems will be studied, that include combinations of metals (N or F), semiconductors (N or F) and magnetic tunnel junctions. Mmuch less is known for such hybrids than for the much-studied metallic magnetic multilayers. Poorly-understood aspects to be studied include spin accumulation, spin relaxation, nonequilibrium interlayer exchange coupling, magnetic switching, crossover from coherent to incoherent transport, and interference corrections to the low-temperature conductivity.

(3f) Thin Ferromagnetic Films: Collective dissipationless spin transport has recently been predicted to be possible in itinerant thin-film ferromagnets, based on the observation that spiral magnetic ordered states support spin currents for equilibrium quasiparticle populations. We plan to consider spin transport across weak links of ferromagnetic films and discuss relations to the Josephson effect in superconductors.

Itinerant magnetism in metallic nanoparticles with discrete electronic states has recently been studied experimentally using tunnelling spectroscopy. This necessitates the development of a theory of nanoscale itenerant ferromagnetism. Our aim is to find a microscopic model that describes the complete crossover from the Stoner model for bulk magnetism to Hund's rule in molecular systems. Using this model, nonequilibrium effects (e.g. spin accumulation), higher order cotunneling processes, and the dynamics of the magnetization reversal of the grain in an external field (quantum tunnelling or thermal activation?) will be studied. We shall also explore in detail the interplay between superconducting and exchange correlations in the low-temperature thermodynamic properties of (nonferromagnetic) metallic nanograins.