First-Principles Study of Domain Evolution of IR on IR(111)
- PDF / 546,631 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 29 Downloads / 209 Views
ABSTRACT We have calculated the effective cluster interactions (ECI) which govern the ordering of Ir adatoms on the Ir( 1t1) surface. The computations are based on a tight-binding Hamiltonian in which no adjustable or experimentally determined parameters were introduced. Both atoms adsorbed in 'bulk' sites (i.e. continuing the fcc lattice) and those in 'surface' sites (i.e. producing hcp stacking) are considered. We use this formalism to determine the relative stability of various adsorption sites and cluster shapes at zero temperature. The overall trends are in excellent agreement with the experimental results found by Ehrlich and co-workers. Next, we employ these ECI in Monte Carlo simulations of the kinetics of domain growth and evolution. Specifically we analyze the effect of diffusion barriers and the competition between the ordering tendencies of the system and entropic effects. Typical 'snapshots' in a range of temperatures and coverages are discussed. INTRODUCTION The growth of a crystal by atomic adsorption is one of the key phenomena in solid state physics. It poses profound questions about the nature of the bonding as the coordination number changes and about the driving mechanism of the symmetry breaking that is often associated with the building up of a crystal from lower-dimensional units. A thorough understanding of these phenomena would also lead to the ability to control defect formation in the bulk or growth processes at surfaces or in multilayers, etc. Such insights would have immediate impact in a wide range of technologies. On the experimental side, much understanding has been gained through the work of Wang and Ehrlich, notably for the adsorption of Ir on Ir(l 11) [1-3]. Using a field ion microscope these authors observed a large number of adsorbed clusters of varying sizes, all the way from single atoms to -r13 islands. At temperatures below 460 K the clusters always formed compact structures, with the possible exception of the linear trimer. Although, the latter was occasionally observed it was found to be less stable than the triangular trimer. Moreover, Wang and Ehrlich established that at 100 K only 15 % of the isolated adatoms sit in 'bulk' sites but that this percentage quickly goes to 100 % as the number of adsorbed neighbors increases. This is, of course, as it should be if the lattice is to continue its fcc stacking sequence. Nevertheless, the preferred occupancy of 'surface' sites at low concentrations is unexpected. Thus, the question arose as to the driving mechanism for this site preference. Clearly, the answer must be sought in the electronic structure of the material. There have been relatively few theoretical studies of the energy differences responsible for the preferential occupancy of certain sites and the nature of the diffusion. Piveteau et al. [4] used the tight-binding formalism to calculate the relative stability of the two types of adsorption sites and found good agreement with the results of Wang and Ehrlich, i.e. a cross-over from predominantly 'surface' site occupa
Data Loading...
