We first present an extended introduction of the various methods used to extract electronic and structural information from the K pre-edge X-ray absorption spectra of 3d transition metal ions. The K pre-edge structure is then modelled for a selection of 3d transition metal compounds and analyzed using first-principles calculations based on the density functional theory (DFT) in the local density approximation (LDA). The selected compounds under study are presented in an ascending order of electronic structure complexity, starting with the Ti K-edge of rutile and anatase, and finishing with the Fe K-edge of the cyanomet-myoglobin. In most cases, the calculations are compared to polarized experimental spectra. It is shown that DFT-LDA methods enable us to reproduce satisfactorily the experimental features and to understand the nature of the electronic transitions involved in the pre-edge region. The limiting aspects of such methods in modelling the core–hole electron interaction and the 3d electron–electron repulsion are also pointed out.

Complex intermetallic compounds such as transition metal (TM) aluminides are promising alternatives to expensive Pd-based catalysts, in particular for the semi-hydrogenation of alkynes or alkadienes. Here, we compare the gas-phase butadiene hydrogenation performances of o-Al₁₃Co₄(100), m-Al₁₃Fe₄(010) and m-Al₁₃Ru₄(010) surfaces, whose bulk terminated structural models exhibit similar cluster-like arrangements. Moreover, the effect of the surface orientation is assessed through a comparison between o-Al₁₃Co₄(100) and o-Al₁₃Co₄(010). As a result, the following room-temperature activity order is determined: Al₁₃Co₄(100) < Al₁₃Co₄(010) < Al₁₃Ru₄(010) < Al₁₃Fe₄(010). Moreover, Al₁₃Co₄(010) is found to be the most active surface at 110°C, and even more selective to butene (100%) than previously investigated Al₁₃Fe₄(010). DFT calculations show that the activity and selectivity results can be rationalized through the determination of butadiene and butene adsorption energies; in contrast, hydrogen adsorption energies do not scale with the catalytic activities. Moreover, the calculation of projected densities of states provides an insight into the Al₁₃TM₄ surface electronic structure. Isolating the TM active centers within the Al matrix induces a narrowing of the TM d-band, which leads to the high catalytic performances of Al₁₃TM₄ compounds. Graphical abstract

Clusters,i.e.polyhedral geometric entities, are widely used to describe the structure of complex intermetallic compounds. However, little is generally known about their physical significance. The atomic and electronic structures of the Al₁₃TM₄complex intermetallic compounds (TM = Fe, Co, Ru, Rh) have been investigated using a wide range ofab initiotools in order to examine the influence of the chemical composition on the pertinence of the bulk structure description based on 3D clusters. In addition, since surface studies were found to be a relevant approach to address the question of cluster stability in complex phases, the interplay of the cluster substructure with the 2D surface is addressed in the case of the Al₁₃Co₄(100) and Al₁₃Fe₄(010) surfaces.

The growth of complex intermetallics as surface alloys is investigated by annealing Al thin films deposited on Cu(111) substrate in ultrahigh vacuum. Already at room temperature, the large lattice mismatch between Al and Cu results in interfacial intermixing. Upon annealing, various phases are formed by diffusion depending on the thickness of the Al films and the annealing temperature. The surface structures are characterized by scanning tunneling microscopy, low-energy electron diffraction, and x-ray photoelectron spectroscopy. Three different superlattice phases are identified as well as the complex Hume-Rothery $\gamma {\text{-Al}}_4{\text{Cu}}_9$ phase. The epitaxial relationships between the surface phases and the Cu(111) substrate are determined. We further investigate the electronic structure of the $\gamma$ phase by density functional calculations. Experimental valence bands are compared to calculated density of states and simulated STM images are used to identify possible bulk planes appearing as surface termination.