Program Organizer: Dr. David E. Jesson, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6030; Robert Q. Hwang, Sandia National Labs, Livermore, CA 94551; David J. Srolovitz, Dept. of Materials Science and Engineering, University of Michigan, Ann Arbor, MI; Chris Palmstrom, Chemical Engineering and Materials Science and Engineering 151 Amundsen Hall, University of Minnesota, Minneapolis, MN 55455
Monday, AM Room: Orange County 4
February 5, 1996 Location: Anaheim Marriott Hotel
Session Chairperson: Andrew Zangwill, School of Physics, Georgia Tech., Atlanta, GA 30332
8:30 am Invited
DYNAMICS OF ISLAND GROWTH DURING SUBMONOLAYER EPITAXY: G. S. Bales, Computational Materials Science Department, Sandia National Laboratories, Livermore, CA 94550
Under ideal conditions, the size distribution of islands which grow during the early stages of epitaxial growth obeys a simple scaling form. The distribution is specified completely by the average size and a "universal" scaling function. Moreover, the total number of islands has a simple power law dependence on the ratio of surface diffusion rate D and deposition rate F. These simple scaling rules apply to infinitely flat surfaces with no defects or impurities, fixed critical island size, and values of D/F greater than about 106. In this work, a formalism is presented for understanding the scaling behavior when the above conditions are not satisfied. The shape of the distribution function and the value of the number density become dependent on a limited set of coverage dependent "dynamical state variables" which reflect the various competing processes which occur during growth. A systematic study using Kinetic Monte Carlo simulations and self-consistent rate equations identifies the important variables. The results are compared to experimental data. Work supported by the Office of Basic Energy Sciences of the Department of Energy.
MORPHOLOGY OF THIN FILMS AT INITIAL STAGES OF GROWTH: MONTE CARLO SIMULATIONS: Ashok Challa, Center for Solid State Science, Arizona State University, Tempe, AZ 85287- 1704; Timothy S. Cale, Center for Solid State Electronics Research, Arizona State University, Tempe, AZ 85287-6206
The growth mechanism and resulting atomic morphology at initial stages of epitaxial growth of thin films by vapor phase deposition is investigated with Monte Carlo simulations using a solid-on-solid model. The initial stage of film growth is clearly divided into two stages: a nucleation stage and a growth stage. Thin film growth can be characterized as 2D layer-by-layer growth or 3D island growth. We have developed a 3D island growth model to study the growth morphology of the thin films at sub-monolayer coverages. The model incorporates the potential barriers near edges of the island that suppress the diffusion of adatoms from upper layer to lower layer and vice versa. The role of these potential barriers on the island morphology is investigated. We also examine the influence of reversible and irreversible attachment and restructuring of adatoms at the island edges on the 3D island growth morphology. The island morphologies obtained from the simulations are compared to the experimentally observed morphologies.
9:20 am Invited
ROUGHENING AND SMOOTHING KINETICS OF Ge(001): David G. Cahill, S. Jay Chey, Joseph E. Van Nostrand, Department of Materials Science, University of Illinois, 1101 W.Springfield Ave., Urbana, IL 61801
We use in- situ scanning tunneling microscopy to characterize the rough surface morphologies of Ge(001) produced by MBE growth and low- energy ion etching at low temperatures. For crystal growth, a regular pattern of multi- layer growth mounds develops over a wide range of temperatures, 60- 200deg.C. A similar pattern of etch pits is formed during ion- sputtering by 240 eV Xe ions at 250 and 275deg.C. In both cases, crystal growth and etching, the lateral length scale of the roughness evolves directly from the submonolayer island separation. We attribute the roughening mechanism to a diffusion bias for surface defects-- adatoms during growth, and surface vacancies during sputtering. We have also studied the relaxation of multilayer roughness by thermal annealing at 225, 250 and 275deg.C. Despite the relatively low annealing temperatures and short lateral length scales studied in our experiments, the data are consistent with the Herring- Mullins (continuum) description of surface smoothing.
9:50 am lnvited
EFFECTS OF CRYSTAL MICROSTRUCTURE ON EPITAXIAL GROWTH: Jacques G. Amar, Fereydoon Family, Physics Department, Emory University, Atlanta, GA 30322
In the past, simulations of epitaxial growth have used solid- on- solid (SOS) models to simulate the crystalline structure of both the substrate and the growing crystal. These models have produced results in the early stages of growth in good agreement with experiments for a number of different quantities, including the island density and island- size distribution. For multilayer growth, however, there exists a competition between microscopic effects such as the Ehrlich-Schwoebel step barrier and the crystalline microstructure. Therefore, the crystal structure and geometry are important in determining the evolution of epitaxial morphology. We present the results of large- scale realistic kinetic Monte Carlo simulations of multilayer epitaxial on fcc(100) and bcc(100) surfaces. The influence of crystal structure on the formation and coarsening of mounds and facets will be discussed. We also discuss and compare our results with recent experiments.
10:20 am BREAK
ISSUES IN METAL EPITAXY
Session Chairperson: R. F. C. Farrow, IBM Research Division, Almaden Research Center, 650 Harry Road, San Jose, CA 95120- 6099
10:40 am Invited
THE ROLE OF PLACE EXCHANGE AND ADATOM INCORPORATION IN METAL EPITAXY: R. J. Behm, J. A. Meyer, Abt. Oberflachenchemie und Katalyse, Universitat Ulm, D- 89069 Ulm, Germany
In simple models, epitaxial growth is limited to processes on top of the respective surface layer, and the substrate serves as a passive support. In a rapidly growing number of cases, however, it turned out that the situation is less ideal. Incorporation of diffusing metal adatoms into the surface layer and place exchange with metal atoms in that layer can play an important role in metal epitaxy. These processes can strongly affect both nucleation and growth behavior of adislands and hence roughness and morphology of the resulting film (i) by modifying the density of diffusing adatoms, (ii) by providing nucleation sites for subsequent heterogeneous nucleation and (iii) by creating an adlayer of mixed composition. These processes are identified from recent scanning tunneling microscopy data, which also allow an estimate of the kinetic barriers for the respective processes.
11:10 am Invited
THE STRUCTURE OF CLOSE- PACKED SURFACES: C. B. Carter, Department of Chemical Engineering and Materials Science, Amundson Hall, University of Minnesota, Minneapolis, MN 55455; R. Q. Hwang, J. C. Hamilton, Sandia National Laboratories, Livermore, CA 94551
Several different structures have been observed for the close- packed 111- fcc surfaces of pure materials and when layers of a second element are grown on a 0001- hcp substrate. We will review both the experimental observations of the phenomenon and computer modeling studies. The strain may be relaxed locally in different ways that involving threading dislocations, misfit Shockley partial dislocations, and both rotation and buckling of the surface layer. We will then describe and use the cubic- hexagonal (CH) model for the structured 111 surface to show the relationship between the different observed structures which include trigon arrays, node networks and the herringbone pattern. Real surfaces tend to show mixtures and distortions of these different structures. We will relate the different terminologies which have been used to describe the same phenomenon and relate these concepts to the structure of 111 heterojunctions in thin films.
11:40 am Invited
STM STUDIES OF INTERMIXING AND ALLOY FORMATION IN ULTRATHIN METAL FILMS: D. D. Chambliss, IBM Research Division, 650 Harry Road, San Jose, CA 95120- 6099
Intermixing can play an important role in determining the morphology of epitaxial thin films. Conversely, the structure of the substrate can control the stability of alloys formed when deposited materials react with each other or with the substrate. Using the scanning tunneling microscope (STM) we have found that ultrathin films of Au on Cu(100), Fe on Cu(100), and Ag and Co on Mo(110) demonstrate many aspects of this interplay. Au/Cu(100) forms a surface alloy analogous to bulk Cu3Au, modified to relieve misfit strain. Fe/Cu(100) tends to intermix at the monolayer level, despite the miscibility gap for bulk Fe and Cu. This intermixing leads to substantial roughness of the Fe/Cu interface. Ag and Co, though immiscible in bulk, form an intermixed "quasialloy" phase when codeposited on Mo(110), to establish an epitaxial fit with the substrate. This metastable phase is replaced at high temperatures by a phase separation between pure Ag and a Co- Mo surface alloy.
STM STUDIES OF THE STRUCTURE AND GROWTH DYNAMICS 2- D METAL ALLOYS: A. K. Schmid, J. L. Stevens, R. O. Hwang, Sandia National Laboratories,Livermore, CA 94551- 0969
Strain has been recently shown to play a key role in determining not only the
structure of thin metal films, but also their composition. This is especially
true in multi-
films and superlattices. To investigate the effects of strain on such films, we
have chosen to study monolayer binary mixtures of Ag/Cu, Ag/Co, and Co/Cu on a
Ru(0001) substrate. These metals all exhibit a substantial miscibility gap in
their bulk binary phase diagrams. However on the Ru(0001) substrate, strain
induced by the lattice mismatch gives rise to mixing behavior substantially
different than the bulk. In this talk we will describe our STM experiments on
the structure and the dynamics of formation of these alloy combinations. New 2-
phase diagrams are found for the various combinations of metals, which are
strongly driven by strain relief. Furthermore, growth morphology of these
alloys differ significantly from those of the single component films. These
differences point to a variety of new atomistic processes not found in the
growth of the single component case.This work is supported by the Office of
Basic Energy Sciences, Divsion of Materials Science of the US DOE under
|Search||TMS Annual Meetings||TMS Meetings Page||About TMS||TMS OnLine|