Density Functional Description of Metal-Metal and Metal-Ligand Bonds

Abstract

Problems related to surfaces (such as the catalytic processes which often involve metal-ligand interactions) present considerable challenge to computational scientists. From the point of view of solid state physics, difficulties arise because of the low symmetry, compared to bulk materials, of the relevant model systems. From the point of view of chemistry, difficulties arise from the sheer number of atoms needed for clusters or supermolecules to model realistically systems of interest. These difficulties are compounded when transition metal atoms are involved because of the large number of valence electrons, the importance of electron correlation, spin polarization effects, etc. Taken together these impose severe computational restrictions for the treatment of these many-electron systems. As we will show in this contribution, density functional theory (DFT) is a practical first-principles approach for the study of the electronic structure of large and complicated systems and also a very useful tool for the study of interactions between ligands and metals in various states of aggregation (atoms, clusters, and infinite surfaces). DFT has been widely used in solid state physics since the introduction of the Xα method by Slater in the early 50’s[1]. About twenty years later, in the early 70’s, the first DFT method generally applicable to finite systems of interest for chemistry was devised by Slater and Johnson[2]. Over the following twenty years, DFT methods in chemistry have gradually evolved to include many of the standard features of ab initio methods: basis sets, algorithms for geometry optimization, etc.... In parallel, more sophisticated treatments of exchange and correlation were developed. Indced, the use of non-local functional allows, in many cases, quantitative predictions of total energy differences (binding energies, ionization potentials, etc.). With these advances, DFT has gained much popularity recently and there has been an explosive growth in the number of DFT applications to chemistry in the last two or three years[3]. In electronic structure theory, the study of surfaces is, in a sense, at the interface of physics and chemistry: insight can be gained both from few-atom systems and from infinite ideal surfaces. We think that it is profitable to take both points of view and to study systems of any size within a single conceptual framework and with consistent numerical methods. In our work, we use the conceptual framework of DFT and the numerical methods implemented in the program deMon. We approach the subject from a “chemical” point of view and the applications will follow the reverse order of historical development of DFT: from systems involving very few atoms, to clusters, and finally to infinite surfaces and the bulk.