The nature of metal-adsorbate interactions revealed by density functional calculations on complexes and clusters and by a new embedded cluster approach

The nature of metal-adsorbate interactions in complexes and surfaces is the main concern of this thesis. The cluster-model approach has been improved and a new embedded cluster approach developed within the density functional framework. The Linear Combination of Gaussian Type Orbitals-Model Core Potential-Kohn Sham-Density Functional (LCGTO-MCP-KS-DF) method has been used to study the mechanism of adsorbate interactions with the transition metals aiming to contribute to the molecular level understanding of important reactions catalysed by metals.;A comparative study of the mononitrosyls of the first-row transition metals (M- NO) allowed us to describe the bonding mechanism and to explain the bending of the molecules for high multiplicities from a molecular orbital picture. Furthermore, a scheme is presented from which one can predict qualitatively the properties of MNO systems such as geometries, multiplicities, backdonation, charge transfers and relative bond lengths. We have also shown that such a scheme can be used to rationalize the binding in other systems such as MCO, MCN, MN2 and MO2, where M is a first-row transition metal. We have investigated the interaction of NO and its dimer with the nickel dimer. We have addressed here the difficulties that DFT meets in describing the geometry, binding energy and electronic structure of the NO dimer in the gas phase. We have also discussed the nature of its ground state electronic structure and shown that the recently developed correlation functional that involves the Laplacian of the electron density and the kinetic-energy density has improved the description of the NO dimer with respect to the more commonly used GGA exchange-correlation schemes.;N2, one of the products of NO reduction, has also been studied. The structure and the frequencies of the newly discovered Fe(N2) n (n = 1,2,3,4,5) complexes have been described.;Modeling the metal surface has also been discussed extensively. The cluster model approach has been shown to be useful in describing the geometry and qualitative features of the adsorbate-surface interactions.;A new embedded cluster approach has been developed in this thesis. A localized electronic subspace DFT approach has been adopted. The long range interactions due to the extended surface are taken into account from an embedding potential surrounding the cluster. The DFT extension to systems with fractional electron numbers has also been used in this embedding technique to assure that the embedded cluster has the same chemical potential as the metal surface. (Abstract shortened by UMI.).