Part B – The Joint Programme of Work

1. Project Objectives

Objective 1. Spin Injection into Semiconductors

The goal is to achieve electrical injection of spin-polarized currents into semiconductors, which is an essential ingredient for any emerging spintronics technology. Two main types of spin injectors will be explored: (a) dilute magnetic semiconductors; (b) metallic ferromagnets with high-impedance junctions.

Objective 2. Spin-Dependent Transport through Hybrid Structures

Spin-dependent transport in hybrid structures involving a combination of ferromagnetic (F) and normal (N) or superconducting (S) materials will be studied systematically. A general quasiclassical theory for such hybrid systems will be developed. For FN structures, the technologically vital question of which factors effect the spin-flip length in the metal will be studied systematically. For FS structures, novel effects arising from the interplay between spin coherence and electron-hole coherence will be explored.

Objective 3: Spin-Dependent Transport in Confined Geometries

Spin-dependent transport in a variety of confined geometries will be studied in detail, both to explore their fundamental properties and with a view to possible applications. Particularly, we shall study (a) quantum dots [Kondo effect, spin coherence times, dots as qubits for quantum computation]; (b) quantum wires [ballistic and diffusive spin transport, spin-charge separation]; (c) quantum point contacts [conductance quantization, spin-valve magnetoresistance]; (d) magnetic single-electron transistors [tunnelling magnetoresistance, spin accumulation]; (e) magnetic multilayers [spin-valve effect]; (f) spin supercurrents in thin ferromagnetic films.

Objective 4: Mesoscopic Magnetism

We shall develop microscopic models for small magnets on the nanometer scale, and study spin-dependent transport through magnetic nanograins so small that their energy spectrum is discrete.

2. Research Method

Objective 1: Spin Injection into Semiconductors:

Groningen and Würzburg will continue their experimental efforts to achieve spin injection into semiconductors, by pursuing both of the following strategies:

(1a) Dilute magnetic semiconductors can be lattice matched via molecular beam epitaxy, as already demonstrated by the Würzburg group. Aachen and Karlsruhe will expand on the theoretical model for diluted F-semiconductors worked out by König et al. [3]. Transport properties will be studied by using real-time diagrammatic techniques developed in the Aachen and Karlsruhe groups.

(1b) Conventional metallic ferromagnets will be prepared in situ by shadow evaporation techniques, to have a good control over the quality of the junctions.

Objective 2: Spin-Dependent Transport through Hybrid (F,N,S) Structures:

(2a) General hybrid structures: The Delft theory group will develop a general theory for such structures, using powerful and well-established semi-classical methods, and their recent extensions, incorporating fully all interesting quantum coherence effects [5]. Circuit theories and the underlying semi-classical methods will be extended by the Delft and Basel and Pisa groups to describe the full statistics of electron and spin transfer, and applied to study electron noise.

(2b) FN Structures: Groningen will study the behaviour of spin accumulation, spin currents and spin decoherence using mesoscopic devices with varying material parameters. The non-local spin accumulation will be explored theoretically (Basel and Delft groups) and experimentally (Groningen) by studying long normal wires covered by a series of identical isolated F-electrodes. The Basel group will study transport and noise in multi-terminal magnetic structures.

(2c) FS Structures: The Basel and Delft groups will apply the theory of (2a) to study proximity effect in the ferromagnet, or a spin imbalance in the superconductor. Groningen will study spin injection into a superconductor experimentally, using multiterminal spin valve geometries.

Objective 3: Spin-Dependent Transport in Confined Geometries

(3a) Quantum Dots: Delft will use the technique of electron spin resonance (applying microwave irradiation) to study spin dynamics, spin decoherence and spin manipulation in quantum dots. Delft will also search for signatures of spin entanglement in noise measurements. These experiments will be guided by theoretical analyses to be done in Basel, Pisa and Delft. Zürich will exploit the steep potential wells and precise pattern transfer of their AFM nanolithography method to construct smaller and more complex structures. Theoretical descriptions of such complex systems will be developed and compared to experiment by the Aachen, Karlsruhe, Budapest, Pisa, Delft, München and Weizmann groups. The theoretical studies will use, i.a., Keldysh real-time and numerical renormalization group methods. The Weizmann group will also incorporate interactions into the Keldysh framework following the Kamenev-Andreev scheme, but extending it to include spin; they will describe interference effects using a variant of real-time perturbation theory developed by Aachen and Karlsruhe.

(3b) Quantum Wires: The Weizmann group has extended the cleaved-edge overgrowth method to construct two wires that are parallel with atomic precision. Orsay will study spin transport in carbon nanotubes, which will be connected to ferromagnetic contacts using two methods: either by depositing them over prefabricated ferromagnetic contacts using Kasumov’s aligned laser technique, or by directly using the ferromagnetic contacts as catalysts for the growth of the carbon nanotube. Theoretical support can be expected from Aachen, Karlsruhe, Budapest, Pisa and München, using bosonization and Luttinger-liquid theory.

(3c) Quantum Point Contacts: Poznan will study spin-dependent transport in point contacts within the framework of a realistic tight-binding model analysed with the nonequilibrium Green function technique. Budapest will extend their previous theories of point contacts and short wires.

(3d) Magnetic Single-Electron Transistors: The standard methods that have been developed for nonmagnetic systems will be developed further by Aachen, Karlsruhe, Delft, München and Poznan to include effects related to ferromagnetism.

(3e) Hybrid multilayers: The Kubo and Landauer formalisms will be used to study various aspects of electronic transport in hybrid multilayers.

(3f) Thin ferromagnetic films: The mean-field approach will be extended by Karlsruhe and Aachen to the case of thin films with weak links.

Objective 4: Nanoscopic Magnetism:

Aachen will study itinerant magnetism in nanograins using a microscopic 2-band Hubbard model in a confined geometry. DMRG and DMFT methods will be used to calculate static and spectral properties, which will be compared to experimental data of tunnelling spectra. München will investigate the possibilities for observing of macroscopic spin tunnelling in nanograins. Pisa will use exact diagonalization studies and renormalization group to study the interplay between ferromagnetic and superconducting correlations in metallic nanograins.