Abstract:
Relaxed atomic geometries and chemisorption energies have been calculated for the dissociative adsorption of molecular hydrogen on vicinal Si(001) surfaces. We employ density-functional theory, together with a pseudopotential for Si, and apply the generalized gradient approximation by Perdew and Wang to the exchange-correlation functional. We find the double-atomic-height rebonded D_B step, which is known to be stable on the clean surface, to remain stable on partially hydrogen-covered surfaces. The H atoms preferentially bind to the Si atoms at the rebonded step edge, with a chemisorption energy difference with respect to the terrace sites of >sim 0.1 eV. A surface with rebonded single atomic height S_A and S_B steps gives very similar results. The interaction between H-Si-Si-H mono-hydride units is shown to be unimportant for the calculation of the step-edge hydrogen-occupation. Our results confirm the interpretation and results of the recent H_2 adsorption experiments on vicinal Si surfaces by Raschke and Hoefer described in the preceding paper.

Abstract:
Non-adiabaticity in adsorption on metal surfaces gives rise to a number of measurable effects, such as chemicurrents and exo-electron emission. Here we present a quantitative theory of chemicurrents on the basis of ground-state density-functional theory (DFT) calculations of the effective electronic potential and the Kohn-Sham band structure. Excitation probabilities are calculated both for electron-hole pairs and for electrons and holes separately from first-order time-dependent perturbation theory. This is accomplished by evaluating the matrix elements (between Kohn-Sham states) of the rate of change of the effective electronic potential between subsequent (static) DFT calculations. Our approach is related to the theory of electronic friction, but allows for direct access to the excitation spectra. The method is applied to adsorption of atomic hydrogen isotopes on the Al(111) surface. The results are compatible with the available experimental data (for noble metal surfaces); in particular, the observed isotope effect in H versus D adsorption is described by the present theory. Moreover, the results are in qualitative agreement with computationally elaborate calculations of the full dynamics within time-dependent density-functional theory, with the notable exception of effects due to the spin dynamics. Being a perturbational approach, the method proposed here is simple enough to be applied to a wide class of adsorbates and surfaces, while at the same time allowing us to extract system-specific information.

Abstract:
The energetics of H$_2$ interacting with the Si(100) surface is studied by means of {\em ab initio} total energy calculations within the framework of density functional theory. We find a direct desorption pathway from the mono-hydride phase which is compatible with experimental activation energies and demonstrate the importance of substrate relaxation for this process. Both the transition state configuration and barrier height depend crucially on the degree of buckling of the Si dimers on the Si(100) surface. The adsorption barrier height on the clean surface is governed by the buckling via its influence on the surface electronic structure. We discuss the consequences of this coupling for adsorption experiments and the relation between adsorption and desorption.

Abstract:
Three different clusters, Si_9H_12, Si_15H_16, and Si_21H_20, are used in density-functional theory calculations in conjunction with ab initio pseudopotentials to study how the energetics of H_2 dissociativ e adsorption on and associative desorption from Si(001) depends on the cluster size. The results are compared to five-layer slab calculations using the same pseudopotentials and high qu ality plane-wave basis set. Several exchange-correlation functionals are employed. Our analysis suggests that the smaller clusters generally overestimate the activation barriers and reaction energy. The Si_21H_20 cluster, however, is found to predict reaction energetics, with E_{a}^{des}=56 +- 3 kcal/mol (2.4 +- 0.1 eV), reasonably close (though still different) to that obtained from the slab calculations. Differences in the calculated activation energies are discussed in relation to the efficiency of clusters to describe the properties of the clean Si(001)-2x1 surface.

Abstract:
We demonstrate how first-principles calculations using density-functional theory (DFT) can be applied to gain insight into the molecular processes that rule the physics of materials processing. Specifically, we study the molecular beam epitaxy (MBE) of arsenic compound semiconductors. For homoepitaxy of GaAs on GaAs(001), a growth model is presented that builds on results of DFT calculations for molecular processes on the beta2-reconstructed GaAs(001) surface, including adsorption, desorption, surface diffusion and nucleation. Kinetic Monte Carlo simulations on the basis of the calculated energetics enable us to model MBE growth of GaAs from beams of Ga and As_2 in atomistic detail. The simulations show that island nucleation is controlled by the reaction of As_2 molecules with Ga adatoms on the surface. The analysis reveals that the scaling laws of standard nucleation theory for the island density as a function of growth temperature are not applicable to GaAs epitaxy. We also discuss heteroepitaxy of InAs on GaAs(001), and report first-principles DFT calculations for In diffusion on the strained GaAs substrate. In particular we address the effect of heteroepitaxial strain on the growth kinetics of coherently strained InAs islands. The strain field around an island is found to cause a slowing-down of material transport from the substrate towards the island and thus helps to achieve more homogeneous island sizes.

Abstract:
The initial steps of MBE growth of GaAs on beta 2-reconstructed GaAs(001) are investigated by performing total energy and electronic structure calculations using density functional theory and a repeated slab model of the surface. We study the interaction and clustering of adsorbed Ga atoms and the adsorption of As_2 molecules onto Ga atom clusters adsorbed on the surface. The stable nuclei consist of bound pairs of Ga adatoms, which originate either from dimerization or from an indirect interaction mediated through the substrate reconstruction. As_2 adsorption is found to be strongly exothermic on sites with a square array of four Ga dangling bonds. Comparing two scenarios where the first As_2 gets incorporated in the incomplete surface layer, or alternatively in a new added layer, we find the first scenario to be preferable. In summary, the calculations suggest that nucleation of a new atomic layer is most likely on top of those surface regions where a partial filling of trenches in the surface has occurred before.

Abstract:
A computational study of the epitaxial Co2MnSi(001)/MgO(001) interface relevant to tunneling magnetoresistive (TMR) devices is presented. Employing ab initio atomistic thermodynamics, we show that the Co- or MnSi-planes of bulk-terminated Co2MnSi form stable interfaces, while pure Si or pure Mn termination requires non-equilibrium conditions. Except for the pure Mn interface, the half-metallic property of bulk Co2MnSi is disrupted by interface bands. Even so, at homogeneous Mn or Co interfaces these bands contribute little to the minority-spin conductance through an MgO barrier, and hence such terminations could perform strongly in TMR devices.

Abstract:
We study the energetics of island formation in Stranski-Krastanow growth within a parameter-free approach. It is shown that an optimum island size exists for a given coverage and island density if changes in the wetting layer morphology after the 3D transition are properly taken into account. Our approach reproduces well the experimental island size dependence on coverage, and indicates that the critical layer thickness depends on growth conditions. The present study provides a new explanation for the (frequently found) rather narrow size distribution of self-assembled coherent islands.

Abstract:
This article reviews the basic computational techniques for carrying out multi-scale simulations using statistical methods, with the focus on simulations of epitaxial growth. First, the statistical-physics background behind Monte Carlo simulations is briefly described. The kinetic Monte Carlo (kMC) method is introduced as an extension of the more wide-spread thermodynamic Monte Carlo methods, and algorithms for kMC simulations, including parallel ones, are discussed in some detail. The step from the atomistic picture to the more coarse-grained description of Monte Carlo simulations is exemplified for the case of surface diffusion. Here, the aim is the derivation of rate constants from knowledge about the underlying atomic processes. Both the simple approach of Transition State Theory, as well as more recent approaches using accelerated molecular dynamics are reviewed. Finally, I address the point that simplifications often need to be introduced in practical Monte Carlo simulations in order to reduce the complexity of 'real' atomic processes. Different 'flavors' of kMC simulations and the potential pitfalls related to the reduction of complexity are presented in the context of simulations of epitaxial growth.