Computational materials science based on density functional theory (DFT) allows to predict the atomic structure and stability of binary metal alloys and intermetallic compounds. This includes properties like mixing enthalpy, substitutional ordering and aging processes. For this, we calculate the formation enthalpies of a set of geometrically fully relaxed ordered structures via DFT and use them to extract effective cluster interactions by use of the cluster-expansion (CE) method.
In our CE Hamiltonian, the pair interactions are treated in reciprocal space permitting for a prediction of the coherency strain of any arbitrary alloy structure. As an example, the picture to the left shows the direction-dependence of the elastic energy in a Cu_50-Mn_50 superlattice, whereby the distance from the middle of the cube to the surface of the shape represents the absolute value of the so-called constituents strain-energy. Using the ECI's in Monte Carlo (MC) simulations gives access to finite temperatures.
Right now, we are working on Cu-Mn and Cu-Pd which show interesting 1- and 2-dimensional long-period superlattices and novel low-energy structures.

When a surface is created by a cut through the bulk, the truncation of chemical bonds leaves the surface atoms in a non-equilibrium. As a consequence they search for and eventually assume new positions. When this leads to a new lateral periodicity the surface is called "reconstructed" with properties much different from a mere bulk-truncated surface. An impressive example is that of the Ir(100) surface whose reconstructed phase exhibits a unidirectional 5-fold periodicity compared to (100) bulk layers, i.e. a (5x1) reconstruction.

We are interested in the positions of the surface atoms. They can be imaged by Scanning Tunnelling Microscopy (STM) as demonstrated for Ir(100)-(5x1) in the atomically resolved image on the left (top panel). As obvious, there are linear atomic chains which protrude pairwise from the surface (bright yellow) and are separated by a less protruding chain (dark red). We complement such real space images by reciprocal space techniques, i.e. Low-Energy Electron Diffraction (LEED) which is sensitive also to atoms located deeper in the surface and thus gives access to its complete structure. Diffraction spots related to the lateral periodicity appear as demonstrated for two orthogonal domains of Ir(100)-(5x1) on the left (lower panel). We evaluate the intensities of the diffraction spots (considering multiple diffraction processes) and succeed to retrieve the atomic positions with a precision in the picometer range for atoms down to about 1 nm below the surface.

We use the lateral superlattice of the Ir surface as a template to produce new nanostructures by deposition of other atoms. In separate activities we combine STM and LEED to resolve the positions and chemical identity of atoms in the surface of binary alloys, i.e. aluminides which are used in aircraft industry.

On the left you can see the dispersion of the first and second image-potential state. These spectra were measured with angle-resolved two-photon photoemission.

For pulsed laser deposition (PLD), high intensity laser pulses are focused on a target which leads to ablation of a cloud of particles - atoms, ions and clusters (upper image). Growth of thin films by PLD is a process far from thermodynamical equilibrium, with an instantaneous flux of incoming particles that is more than 4 orders of magnitude higher and high average particle energies compared to thermal deposition (TD). In general, this leads to different film properties, for example smoother films or better stochiometry.

We are interested in the initial growth of ultrathin films during PLD for metallic substrates and adsorbates. Our aim is to identify the role of the growth parameters particle flux, particle energy, substrate temperature etc. at coverages below one atomic layer by comparing our results for PLD to TD and to theoretical simulations (molecular dynamics, rate theory etc.) from other work groups.

Scanning tunneling microscopy (STM) provides real space images of islands and defects formed on crystalline surfaces. An example is given at the left: an STM image of 0.05 atomic layers of Co (20nm x 20nm) on the Cu(001)- surface, deposited by PLD. We obtain information about the nucleation kinetics on the surface by quantitatively analyzing the correlation between islands (white) and defects (dark, incorporated Co atoms) and the growth parameters.