ICMCTF2003 Session TS3-1: Computational Studies in Thin Films Science
Monday, April 28, 2003 10:30 AM in Room San Diego
Monday Morning
Time Period MoPL Sessions | Abstract Timeline | Topic TS Sessions | Time Periods | Topics | ICMCTF2003 Schedule
| Start | Invited? | Item |
|---|---|---|
| 10:30 AM |
TS3-1-1 Multi-scale Modeling of Thin Film Deposition
G.H. Gilmer (Lawrence Livermore National Laboratory); J.D. Torre (Commissariat a l'Energie Atomique-Saclay, France); P. O'Sullivan, F.H. Baumann (Bell Laboratories); T.D. de la Rubia (Lawrence Livermore National Laboratory) In this paper we discuss results from atomisitc and continuum simulations to predict thin film microstructure evolution. We are primarily concerned with physical vapor deposition, and deal with issues related to the deposition of dense films of uniform thickness on foreign substrates. Deposition on structured substrates is also treated, including breadloafing and the filling of vias and trenches. Multi-scale modeling is used in the sense that information obtained from molecular dynamics and first principles calculations provide atomic interaction energies, surface and grain boundary properties and diffusion rates for use in Monte Carlo and continuum models. Texture formation during the deposition of polycrystalline films onto foreign substrates will be described based on anisotropies in crystal growth rates, and their effects on the growth of grains of different orientations. Results on stress in thin films observed in molecular dynamics simulations will be included. |
|
| 11:10 AM |
TS3-1-3 In Situ High-temperature Scanning Tunneling Microscopy Studies of Surface Dynamics on Epitaxial TiN(001) and TiN(111) Layers
S. Kodambaka, V. Petrova, S.V. Khare (University of Illinois); A. Vailionis (Stanford University); I. Petrov, J.E. Greene (University of Illinois) We used in situ high-temperature (1000-1250 K) scanning tunneling microscopy to study time- and temperature-dependent two-dimensional (2D) island coarsening/decay (Ostwald ripening) kinetics and temporal fluctuations about the equilibrium island shapes on atomically smooth TiN(001) and TiN(111) surfaces and hence develop an atomic-scale understanding of early stage growth kinetics on epitaxial TiN surfaces. From the coarsening/decay results, we found that 2D TiN(001) islands follow surface-diffusion-limited decay kinetics with an activation barrier Ea = 2.6±0.6 eV, while TiN(111) islands exhibit size-dependent detachment-limited decay kinetics with Ea = 2.3±0.6 eV and = 3.3±0.4 eV for the decay of large and small islands, respectively. The equilibrium TiN(001) island shape is a rounded square bound by <110> steps, while the TiN(111) island shape is a truncated hexagon bound by alternating <110> steps which form [100] and [110] nanofacet ledges. From the analytically derived Legendre transformation of the equilibrium island shapes, in combination with quantitative adatom island decay [in case of TiN(001)] or shape fluctuation measurements [in case of TiN(111)], using exact theoretical approaches valid for anisotropic islands, we determined absolute orientation-dependent step energies β(φ) and step stiffnesses γ(φ), at all φ. Specifically, for the high-symmetry <110> and <100> steps on TiN(001), we obtain β<110> = 0.21±0.05 eV/Å and γ<110>= 0.9±0.2 eV/Å with β<100> = 0.25±0.05 eV/Å and γ<100>= 0.07±0.02 eV/Å while β1 = 0.24±0.05 eV/Å and γ1= 1.6±0.5 eV/Å with β2 = 0.31±0.06 eV/Å and γ2= 0.08±0.01 eV/Å for the two alternating S1 and S2 <110> steps on TiN(111). |
|
| 11:50 AM |
TS3-1-5 Binding Energetics and Stacking Faults in Homoepitaxial Growth
T.W. Michely (RWTH Aachen, Germany) In the talk two different fundamental topics in epitaxial growth are addressed on the basis of atomic scale scanning tunneling microscopy investigations. The binding energy of a pair of adatoms is a first order estimate for the bond strength in growth processes on the surface. It is thus a basic parameter governing the morphological evolution during growth and determines its temperature and energy scale. Therefore, it is unsatisfactory that the three different methods for determining the dimer binding energy - direct field ion microscopy (FIM) measurements, island density analysis based on scanning tunneling microscopy (STM) and density functional theory (DFT) calculations - yield conflicting results. Analyzing as examples the experiments and calculations for the fcc(111) surfaces of Ir, Al and Pt, the reasons for this conflict are explored. Stacking fault formation during growth far from equilibrium may occur on fcc(111) surfaces, which supply regular and faulted adsorption sites. In homoepitaxy on Ir(111) stacking fault islands are readily identified with scanning tunneling microscopy by their distinct island shape. Using FIM results for the occupation probability of stacking fault sites by adatoms and small clusters, a detailed understanding of the temperature dependence of stacking fault formation is achieved. Stacking fault related defect structures, healing of stacking faults and their propagation during multilayer growth are discussed on the basis of detailed atomistic observations. The contributions of Carsten Busse, Peter Feibelman, Henri Hansen and Celia Polop to this work are acknowledged. |