Theory of Bonding and Structure of Materials
The main topic of this Thesis is total-energy calculations of metallic systems, with and without adsorbates or impurities.
In this Effective-Medium Theory the expression for the total energy of a system of interacting atoms is divided into three terms. In this thesis all three terms have been re- investigated and improved. Paper VIII; The cohesive term contains cohesive contributions from each of the participating atoms. These so-called Ec-energies have been calculated more accurately, within the generalized-gradient approximation (GGA- II) for exchange-correlation effects. Such Ec results are presented for all the atoms in the first three rows of the periodic table and for a few more embedded atoms. The corresponding Ec energies have been recalculated within the LDA, to be able to compare them with the new results. There are significant changes for several atoms. Papers V-VII; The atomic-sphere-correction term has been reformulated to yield satisfying accuracy even for mixed systems. The new parametrization for the mixed system reduces properly to the one for a single-component system in the proper limit. Applications to, e.g., hydrogen in palladium show that it can give a useful accuracy and also results in agreement with experiment. Also new methods to calculated the one- electron term has been proposed in this thesis. They are based on comparisons with first-principles calculations for certain configurations. For some applications a data-base produced by the CASTEP method is utilized, in others the recursion method is advocated. In this way angular forces can be accounted for. As there are no empirical parameters in the scheme, many different kinds of atoms can be described in many different configurations, and many different compounds can be investigated without prior knowledge about the system. The scheme is not only versatile and flexible. With the improvements, chemical accuracy is being approached in a fast and calculationally simple scheme. Papers I-IV,IX,X; The EMT is applied to a wide range of systems and physical problems in the thesis, including initial stages of oxidation of aluminium, the elastic properties of hydrogen-doped palladium and platinum, chemisorption properties of magnesium, hydrogen-magnesium interactions, intentionally structured catalyst, cohesion of aluminium, bonding properties of YBaCO, and heteroepitaxial growth of magnesium on a palladium surface. These applications tell us that we can count on and with the EMT also in future applications, giving us more information about and a deeper insight into difficult problems concerning bonding, structure and related properties of materials.