Charge based electronics is dominated by semiconductor devices because of the excellent field control semiconductors provide on electrical as well as optical properties. Adding the spin degree of freedom to semiconductor electronics will greatly enhance the capability of information technology. Expected merits of such spintronic devices are non-volatility, increased processing speed, decreased power consumption, and increased integration density. Semiconductor materials in spintronic devices will provide several orders of magnitude larger spin-coherence times compared to metals and the possibility to manipulate magnitude as well as orientation of spins via spin-orbit interaction by gate-controlled effective local magnetic fields (Rashba fields). Furthermore, semiconductors or hybrids of ferromagnetic metals and semiconductors will enable the integration of non-volatile magnetic storage and logical processing. Beyond the pathway of spintronics based on spin-valves, semiconductor nanostructures will also facilitate isolation and addressing single spins for quantum computing. Major challenges in this exciting field include creation and detection of spin polarization in non-magnetic semiconductors, transport of spin-polarized carriers across hetero-interfaces, functionality at room temperature, and robustness of spin coherence under environmental conditions.
At present, different strategies are pursued to establish and measure spin polarization in non-magnetic semiconductors. Concepts for all-electric spin polarizers and detectors are either based on ferromagnetic contacts or exploit spin-orbit engineering. While proofs of concept are provided, it is still unsettled which method has the best potential for room-temperature application. Here, material science on a fundamental level is needed. Related projects in this part of our Cluster are devoted to high-quality ferromagnet-semiconductor contacts, fundamental properties of nanostructures consisting of diluted ferromagnetic semiconductors, as well as spin-orbit engineering with semiconductor nanowires. Similar questions involving carbon nanotubes, graphene, and molecular systems are addressed in area C of the Cluster.
In the group “Epitaxial Nanostructures” (Hansen) and “Scanning Probe Methods“ (Wiesendanger) hybrids of semiconductors and ferromagnetic metals will be studied. Such hybrids combine advantages of both material classes, i.e., the high room-temperature spin polarization in ferromagnets and the control over the electronic properties of semiconductors. One challenge in this field is the optimization of the interface with respect to efficient spin injection from the ferromagnetic metal into the semiconductor. Using molecular beam epitaxy (MBE), we fabricate epitaxial hybrids of Heusler alloy electrodes on AlGaInAs semiconductor heterostructures (Merkt, Hansen, Heyn). We use the Heusler alloy Ni2MnIn, for which theory predicts 100 % spin polarization. The metamorphic InAlGaAs heterostructures allow for precise control of the barrier between metal contact and the charge-carrier channel in the semiconductor. Furthermore, the system features high spin-orbit coupling, and the strain in the metamorphic heterostructures system can be precisely controlled. Two-dimensional charge-carrier systems are employed to provide low disorder and a large spin-coherence time. The samples' structural properties are controlled in situ with electron diffraction and Auger spectroscopy, and studied ex situ with transmission electron microscopy (Weller). Grazing-incidence X-ray diffraction at the synchrotron radiation sources DORIS III and PETRA III at DESY will be employed, e.g., for high-resolution strain mapping of the films (Röhlsberger). The macroscopic magnetic properties will be studied with SQUID magnetometry (Nielsch). Neutron scattering (Schreyer) will be employed in order to locally resolve the magnetization in the depth of the Heusler films on the nanometer scale. By using point-contact Andreev reflection (Merkt, Meier) the spin polarization will be determined quantitatively. We will apply our expertise in spin-polarized cross-sectional scanning tunneling spectroscopy (Wiesendanger, Wiebe) to study the electronic and magnetic properties at the interfaces between the magnetic and the semiconducting layers on atomic length scales. In order to adjust the orientation and size of the magnetic domains inside the magnetic layers, we will explore lateral patterning by focused ion beams (Oepen). The spin-injection efficiency will be determined from lateral transport experiments between ferromagnetic contacts in non-local geometry while spin accumulation will be detected at biased contacts via the Hanle effect (Hansen).
Investigations of the magnetization dynamics in ferromagnetic nanostructures of hybrids will be performed in the “Semiconductor Physics” group (Heitmann, Mendach) on the picosecond time scale with submicron spatial resolution. We employ a recently established combination of Kerr-microscopy and microwave spectroscopy to image spin dynamics in ferromagnetic nanostructures from, e.g., permalloy or lanthanides (Schreyer) as well as in ferromagnet/semiconductor hybrids, such as rolled-up spinwave resonators (Hansen, Heitmann). The possibility to excite spin dynamics for selected frequencies and to observe their evolution paves the way to the novel field of spin-wave optics. An important tool for the design of our nanostructures and the understanding of our measurements will be micromagnetic simulations (Möller, Hinze).
The atomic-scale properties of diluted magnetic semiconductors (DMS) will be studied by scanning probe techniques (Wiesendanger, Wiebe, Schwarz) and DMS nanostructures will be modeled in the groups of Pfannkuche, Lechermann, Lichtenstein, and Potthoff. An auspicious route pursued in our Cluster is the combination of DMS with quantum confinement in heterostructures. Quantum wells modulation doped with Mn and featuring ferromagnetism in a two-dimensional hole system with low disorder will be fabricated and investigated (Hansen, Wurstbauer). Quantum dots doped with magnetic ions will be studied with spin-resolving scanning probe techniques (Wiesendanger, Wiebe) and by optical single-dot spectroscopy (Heitmann, Mendach).
The exchange interaction in DMS will be investigated in the Wiesendanger group on an atomic length scale. In addition, the surface of III-V semiconductors will be in situ manipulated by STM-controlled substitution of group III elements by magnetic species. These experiments will constitute fundamental insight into the synthesis of magnetic nanostructures atom-by-atom. Precisely controlled density and location of the magnetic dopants in DMS will enable us to study the exchange mechanism leading to the magnetic phase on the atomic scale. Indirect exchange between magnetic ions in III-V semiconductors will also be studied with magnetic exchange force microscopy (MExFM). If small quantities of magnetic atoms are evaporated on the surface of doped III-V semiconductors, the indirect magnetic interaction between ions relatively far apart from each other can be studied in dependence on concentration and type of the non-magnetic dopant by MExFM. Since this technique is sensitive to forces, even the insulating regime of semiconductors at low temperature and low dopant concentration is accessible.

The interplay between the correlated electronic structure, long-range ferromagnetic order at low temperatures, geometric disorder, and spin-dependent transport in DMS provides a major challenge for theoretical approaches. In order to model universal properties of DMS, "diluted" Kondo-lattice-type systems with a small fraction of disordered magnetic moments will be studied (Lechermann, Lichtenstein, Potthoff). Generally, local and uncorrelated disorder is assumed but also clustering of magnetic moments has to be considered. An inhomogeneous dynamical mean-field theory (DMFT) recently developed in the Potthoff group will be combined with novel ideas to treat electron correlations and disorder on equal footing within a formalism based on dynamic functionals. Material-specific aspects are included in a second stage using supercell local density approximation (LDA) calculations as the starting point for a realistic DMFT approach within coherent potential approximation.
In the Hansen group AlGaInAs based metamorphic heterostructures that contain high mobility low-dimensional hole systems, are grown by MBE and modulation doped with Mn. In this way a DMS is combined with a precisely tailored low-dimensional hole system, for instance for spin injection without conductance mismatch. The interaction of magnetic Mn ions with the high-mobility hole gas leads to unexpected and complex magnetotransport properties. We observe in these systems anomalous and planar Hall effects, weak localization and metal-insulator transitions exhibiting strongly correlated insulating phases with ultra-high resistance values in coexistence with the quantum Hall effect. They are unique in the prospect of a semiconducting model system featuring magnetic order together with strong spin-orbit interaction and low disorder. The structural and magnetic properties of the MBE grown heterostructures will be studied with the same techniques employed for the investigation of ferromagnet-semiconductor hybrids. The magnetotransport will be studied at millikelvin temperatures and in high magnetic fields.
Functionality and control of nano-spintronic devices will be modeled by advanced non-equilibrium transport methods (Pfannkuche, Chudnovskiy). Employing complementary analytical and numerical approaches based on the Keldysh path integral, device modeling stretching over several length scales is aimed at in order to account for the specific properties of DMS. Key issues to be investigated include spin decoherence and magnetic dissipation in spintronic devices out of equilibrium as well as the interplay of charge and spin transport with coherent magnetization precession and switching. 
The experimental methods at hand in the Wiesendanger group will also be used to study self-assembled III-V semiconductor quantum dots doped with Mn atoms. The aim is to tailor through the lateral confinement both the electronic and the magnetic properties in a fundamental way. MBE grown InAs quantum dot samples will be supplied by the Hansen group. By depositing Mn adatoms onto cleaved samples it is possible to use the tip of the cross-sectional scanning tunneling microscope to substitute a single Mn for Ga. This allows us to integrate a single magnetic atom or a well-defined configuration of magnetic atoms into a quantum dot. The advantage here is that one has control on the atomic level for tailoring the material.
The optical properties of magnetically doped self-assembled quantum dots will be investigated in the “Semiconductor Physics” group (Heitmann, Mendach). We will integrate single InAs-quantum dots into flexible semiconductor cantilevers or rolled-up microtubes. Such arrangements recently opened-up the possibility to change the dot strain state in a deterministic fashion in situ during optical experiments. In case of Mn-doped InAs quantum dots this will allow us to address the interplay between strain state, confining potential symmetry, and the interaction of the Mn impurity with charge carriers in the dots. 
In the group “Nanostructure Physics” (Merkt) spin-orbit engineering will be employed in Y-branches of semiconductor nanowires. These nanostructures are fabricated from AlGaInAs heterostructures grown by MBE (Hansen, Heyn). As a consequence of the intrinsic spin-Hall effect a charge current is accompanied by a transverse spin current. Spin accumulation at the opposite edges of nanowires is expected to be strong in the one-dimensional quantum limit. Quantum point contacts at the arms of the device allow us to constrict transport to the lowest subband. Y-shaped three-terminal devices first proposed by Kiselev and Kim act as spin filters that split an unpolarized input current into two oppositely spin-polarized currents at the output terminals. Comparing the output currents of a second filter stage the filter efficiency can be determined. Thus an all-electrical spin detector is implemented. External electric or magnetic fields will be used to measure important spintronic quantities like the temperature dependent spin-precession and spin-coherence length. The experimental setup will require previous design optimization via electronic circuit simulation supplied by the Möller group.

In conclusion, in the research area B groups that have demonstrated successful cooperation within the SFB 508 and the GrK 1286 will undertake a joint effort to advance material science in spintronics. Together with the unique experimental methods at DESY and GKSS the fundamental electronic and magnetic mechanisms at work will be studied mainly in view of the role of strain, quantum confinement, and disorder and with the visionary goal of electric field or strain control of the spin in semiconductors.