We study bulk and surface collective modes in a metal superlattice using a formalism bringing out the classical nature of the excitations. Appropriate limits are investigated and earlier erroneous mode assignements for semiconductor superlattices are corrected. Explicit calculations are performed for a structure having metals of very different electron densities. Our study indicates that measuring the collective mode frequencies in superlattices can yield important geometrical information.

The adsorption of a Be atom on a metal surface has been studied with several complementary methods. Self-consistent calculations within the Kohn-Sham scheme show the covalent part of the chemisorption to be unusually weak and indicates the physisorption to be unusually strong, a result that is supported by a careful evaluation of the dielectric-response expression for the van der Waals energy. The unusual behaviour of Be originates from its unique electron structure, with a weakly bound, (1s)2(2s)2 core and a low-lying 2p-orbltal, empty in the ground state of the free atom and broadened and partially filled in the chemlsorbed Be. This results in an unusually wide and shallow adsorption potential well for Be on jellium. On typical free-electron-like metal surfaces, the (100), (111) and (110) faces of Al and Li, the covalent chemisorption is still small but favourable geometry may allow a substantial additional stabilization in some positions, where the atom can benefit from the electrostatic attraction from the metal ions. This effect is estimated by perturbation theory.

In an earlier report the van der Waals interaction between an atom and a metal surface was calculated taking into account the finite extension of the atom. This was found to make the atom to response less well to density fluctuations of short wavelengths. Since those fluctuations are more important close to the surface, the ordinary divergent expression for the van der Waals potential, when the atom-metal separation goes to zero, was replaced by another which gave finite values of the interaction even at distances very close to the surface. The previous treatment is extended in this paper to include also the inability of the metal itself to respond to short wavelength density fluctuations. As can be expected, the improved metal response affects the physisorption potential close to the surface. Oure results are conveniently parameterized for a number of atom/metal configurations in order to make them of easy use in many other applications.

The interaction of H2 with metal surfaces gives rise to elastic and inelastic back-scattering, physisorption, dissociative and possibly associative chemisorption. Theoretically, the interaction is the simplest example that displays this complete range of properties. Experimental study is greatly aided by the availability of the isotopes H2, D2 and HD with their well separated masses and well-spaced rotational spectra. H2-metal systems are therefore particularly worthy of study and it would seem feasible through judicious interplay of theory and experiment to understand the interaction in detail. This note constitutes a summary of current theoretical ideas and their relation to experiment.

Decomposition of water at surfaces with formation of adsorbed hydroxyl intermediates is discussed in two cases with partially oxygen-covered metal surfaces: (i) on a free-electron like metal, like Zn, (ii) on a transition metal, like Pd(100), in ultra-high vacuum and (iii) in the case with a titanium dioxide electrode in an electrolyte. The decomposition occurs differently in the three cases with different reaction products. Making use of available data from electron-structure experiments together with a qualitative analysis of the rearrangement of the molecular-orbital energy-level diagrams for the adsorbed species, we propose the differences in molecular-orbital coupling to the surface, which in turn relate to identified differences in the substrate electron structure, to be responsible for the differences in formation of reaction products.

P-polarized light has the unique feature of probing the surface region of a metal which gives rise to effects not taken into account in a classical (Fresnel) treatment. Previous calculations of this non-local effect have been restricted to free-electron metals. Within a simple framework we will here consider the modifications because of the presence of bound electrons. The somewhat surprising result is that a system with both free and bound electrons will still be possible to analyze in free-electron terms, provided we plot the experimental results with a normalization which removes all dependence of the angle of incidence.

Corrections to the conventional dipole expression for the attractive interaction between an atom and a metal surface are calculated and discussed in several models: (i) the electrostatic image model, primarily used to define the multipole corrections and to give a simple, although accurate, estimate thereof; (ii) a model that includes both $s$ and $d$ electrons in the metallic dielectric response together with an atomic polarizability accounting for multipole contributions; and (iii) a model with a realistic treatment of the coupling between density fluctuations in the metal and on the atom (the conventional methods overestimate it). The models show what factors are needed to avoid a singular interaction at the dynamical image plane. While the point dipole of model (ii) allows polarization response at all wavelengths, and leads to singular behavior, model (iii) shows how the finite extent of the atom limits the ability of the system to respond to polarizing fields of short wavelengths. The latter "saturation" of the response competes with the multipole contributions and reduces their influence on the interaction potential over the whole range of distances, and leads to a finite potential also for shorter distances, where the saturation is particularly important. The models are illustrated with numerical calculations for helium on noble metals. With a proper description of the repulsive interaction the resulting physisorption potential is in agreement with experimental findings. We also apply our results for helium to give some brief comments for another interesting atom/substrate system: ${\mathrm{H}}_{2}$ on noble metals.

The attractive interaction between an atom and a metal surface is calculated, taking into account the d-electron contribution to the metal response, together with an atomic polarizability going beyond the dipole approximation. A numerical calculation is presented for helium on noble metals, giving a slightly deeper physiosorption well compared to the “standard” treatment.

We calculate multipole corrections to the standard van der Waals (dipole) potential in atom-surface scattering. The quadru- and octupoles give at most a 15% deeper potential at relevant physisorption distances, and the degree of metal screening is shown to have very little effect on this conclusion.