AVS1996 Session SS+NS-WeA: Adatom Diffusion During Growth

Wednesday, October 16, 1996 2:00 PM in Room 204C

Wednesday Afternoon

Start	Invited?	Item
2:00 PM	Invited	SS+NS-WeA-1 A Chemical View of Bonding at Metal Surfaces P. Feibelman (Sandia National Laboratories) Despite the delocalized nature of electrons in metals, elementary chemical ideas, including coordination, valence-saturation and directional bonding, are indispensable guides to the systematics of adatom-metal potential energy surfaces. Drawing from first-principles total energy calculations, as well as atomic resolution microscopy data, I will illustrate the application of "chemical thinking" in the analysis of representative examples, culminating with a recent study (done in collaboration with Stefanie Esch and Thomas Michely, Forschungszentrum Juelich, Germany) of the interaction of adsorbed O with steps on Pt(111). In this case, the first-principles calculations show that O adatoms are attracted to step edges, gaining 0.2 to 0.3 eV per nearest neighbor step-edge Pt atom, and that they favor "fcc-like" over "hcp-like" sites by ~0.4 eV. These preferences account for the very different appearances, in scanning tunneling micrographs, of O saturated, A- and B-type island edges on Pt(111). Work supported by the U. S. Department of Energy under Contract DE-AC04-94AL85000.
2:40 PM		SS+NS-WeA-3 Hydrogen Promotion of Surface Self-diffusion on Rh(100) and Rh(311) G. Kellogg (Sandia National Laboratories) Field ion microscope observations show that the rate of surface self-diffusion for individual atoms on the (100) and (311) planes of Rh is significantly increased by exposure of the surface to hydrogen. On Rh(100) admission of hydrogen at partial pressures in the 10\super -9\ Torr range causes the onset temperature for migration of a Rh adatom to decrease from 290K to 240K. Once mobile, the mean-square displacement of the adatom remains constant as a function of time. On Rh(311) the mean-square displacement increases monotonically as a function of time after admission of similar pressures of hydrogen. An increasing diffusion rate indicates that the enhancement of the diffusion rate due to hydrogen is coverage dependent. A coverage dependent diffusion rate implies that mechanism of hydrogen promotion involves more than a simple lowering of the activation barrier of surface diffusion by the attachment of a hydrogen atom to a rhodium atom. The reason for the different behavior on the two different crystal planes along with possible mechanisms of hydrogen promotion will be discussed. *Work supported by the U. S. Department of Energy under contract DE-AC04-94AL85000.
3:00 PM		SS+NS-WeA-4 Effects of Hydrogen in Ni(100) Submonolayer Homoepitaxy K. Haug (Lehigh University); Z. Zhang (Oak Ridge National Laboratory) Effects of hydrogen in Ni(100) submonolayer homoepitaxy are investigated by classical and semi-classical potential total energy calculations. Evaporation/condensation and peripheral diffusion mechanisms for small Ni islands being formed on the Ni(100) surface in the presence or absence of a H-adatom are examined. This study is motivated by recent experimental investigations of vacancy island and adatom island migration on metal surfaces and connections to the available experiments are made.
3:20 PM	Invited	SS+NS-WeA-5 A Method for Extending the Time Scale of Molecular Dynamics Simulations of Surface Growth A. Voter (Los Alamos National Laboratory) During the growth of surface layers, many of the important dynamical events (e.g., surface diffusion and reorganization) occur on time scales much longer than the few nanoseconds accessible to molecular dynamics (MD) simulation. When the nature of these infrequent events is understood, transition state theory (TST) can be used to recover the dynamics of the system for much longer time scales than are available from direct MD. TST is often used as the basis of time-dependent Monte Carlo simulations, in which atom positions are mapped onto a lattice, and a catalog of TST rate constants for all possible events is used to propagate the site occupations on the lattice. However, the true dynamics of surface growth processes often violate assumptions built into this type of model, exhibiting unexpected transitions that are highly concerted, and/or evolving to states that defy mapping onto a lattice. Simulations of surface growth should allow for these uncatalogable events if the system is to evolve correctly. A new method is being developed with this goal in mind. The method extends the time scale of MD without any advanced knowledge of the accessible states of the system or the transition paths between them. The method works in continuous space, without any underlying lattice. Application of this method to metallic surface diffusion using embedded atom method potentials will be presented, demonstrating that time scale enhancements of two orders of magnitude or more (relative to direct MD) are possible.
4:00 PM		SS+NS-WeA-7 Adatom Mobility during Homoepitaxial Growth of Platinum(111) A. G\um o\lzh\um a\user, G. Ehrlich (University of Illinois, Urbana-Champaign) The mobility of adatoms during the crystal growth of platinum(111) has been explored using helium-cooled field ion microscopy. Pt(111) islands were formed by field evaporation in uhv. Thereafter platinum atoms were deposited on top of the islands and their behaviour has been observed at temperatures between 20 K and 150 K. After deposition with the surface at 20 K, adatoms are found at both fcc and hcp sites of the close-packed (111) plane. Above 50 K, Pt atoms occupy only fcc sites. However the surface of Pt(111) islands is not uniform: an empty zone separates the central region of the island from its edges, and atoms initially deposited in the center have to overcome an energy barrier to reach the island edges. In the central region, diffusion of single Pt adatoms, in the absence of any electric fields, is already noticeable at temperatures below 75 K. Implications of the existence of the empty zone on the mobility of Pt atoms will be discussed. (Supported under DOE Grant DEFG02-91ER-45439. AG thanks the Alexander von Humboldt - Stiftung for a Feodor Lynen Fellowship)
4:20 PM		SS+NS-WeA-8 Dislocation Motion Caused by Exchange with the Surface Adatom Lattice Gas A. Schmid, N. Bartelt, J. Hamilton (Sandia National Laboratories); C. Carter (University of Minnesota); R. Hwang (Sandia National Laboratories) The dynamics of dislocation motion in single monolayer films of Cu on Ru(0001) were studied by time-resolved scanning tunneling microscopy. Random dislocation climb of edge dislocations was observed. The dislocation motion involves absorption and ejection of atoms and is interpreted as a signature of the equilibrium exchange between the solid film and the adatom lattice gas covering the surface. From time resolved measurements, we find the edge dislocations make 1-D random walks with a diffusion constant of 0.0077 nm\super 2\/sec. This leads to an estimate of 1 atomic exchange every 30 seconds. Through this interaction with the adatom gas, surface dislocations play a key role in determining the equilibrium state of the surface, thus affecting structural properties such as adatom gas density and step edge structure. These factors are known to have important implications in surface structure and chemistry. Perhaps more importantly, they are crucial in defining evolution and response of thin films to further growth and external forces.
4:40 PM		SS+NS-WeA-9 Effects of Dimer Shearing in Heteroepitaxy and Dimer Sliding Along Step Edges on Metal (100) Surfaces Z. Shi, Z. Zhang (Oak Ridge National Laboratory) A novel atomic process, shear motion of a dimer belonging to a compact cluster, has been introduced recently as an important mechanism for cluster diffusion and dissociation on metal (100) surfaces \super 1\. Consideration of this mechanism could give rise to a large jump in critical island size, from 1 to 8, in submonolayer homoepitaxy. In this study, we extend to the case of metal (100) heteroepitaxy. We propose a scheme to selectively induce a jump in critical island size from 1 to 3 or 1 to 8 by a proper choice of the lattice mismatch, and list a few specific systems in which such jumps are likely to be observable. We further consider the feasibility of dimer shearing near surface defects, and find that the energy barrier for dimer sliding along a step on several fcc(100) surfaces is always lower than the energy barrier to split the dimer at the step. While the barrier for dimer sliding along the step is expectedly higher than the barrier for single adatom diffusion along the step, it turns out to be even lower than the barrier for adatom diffusion on a terrace. These results will be discussed in connection with recent experiments. \super *\Managed by Lockheed Martin Energy Research Corp. for the U.S. Department of Energy under contract DE-AC05-96OR22464.\super 1\ Z. P. Shi, Z. Zhang, A. K. Swan and J. F. Wendelken, Phys. Rev. Lett., in press.