No abstract

No abstract

No abstract

Epitaxial Grain Growth (EGG) is an orientation-selective process that can occur in polycrystalline thin films on single crystal substrates. EGG is driven by minimization of crystallographically anisotropic free energies. One common driving force for EGG is the reduction of the film/substrate interfacial energy. We have carried out experiments on polycrystalline Ag films on Ni(OOl) substrates. The orientation dependence of the Ag/Ni interfacial energy has been previously calculated using the embedded atom method. Under some conditions, EGG experiments lead to the (111) orientations calculated to be interface-and surface-energy-minimizing. However, when Ag films are deposited on Ni(OOl) at low temperature, EGG experiments consistently find that (111) oriented grains are consumed by grains with (001) orientations predicted to have much higher interface and surface energy. The large elastic anisotropy of Ag can account for this discrepancy. Strain energy minimization favors growth of (001) grains and can supersede minimization of interfacial energy if sufficient strain is present and if the film is initially unable to relieve the strain by plastic deformation. II. INTERFACE-ENERGY-DRIVEN GRAIN GROWTHConsider a polycrystalline film of thickness h in which all the grains are columnar, as shown in Fig. 1. The mean grain size of the film is 7, the mean surface energy is yj, and ~yi is the mean film/substrate interfacial energy. The growth rate of a grain of radius r can be expressed as2

Microstructural and texture evolution during grain growth in polycrystalline thin films was investigated. Grain growth in thin films is a coarsening process driven by the reduction of grain boundary energy, surface energies, and strain energy density. Because crystal properties can be anisotropic, grain growth in thin films is an orientation selective process. Surface and interfacial energy minimization or reduction favors the growth of grains with low combined surface and interfacial free energy. For the films and substrates investigated in this thesis, surface and interfacial energy promotes the growth of (111)-textured grains. Thin films on thick substrates are usually subjected to a non-zero state of strain, arising from differential thermal expansion between the film and the substrate, from densification and from intrinsic strains. For elastically deformed fcc metal films, strain energy density promotes the growth of (001)-textured grains. In plastically deformed films, strain energy density can favor the growth of (011)-textured grains; this results from the orientation dependence of the yield stress of grains in thin films. (111) grains are predicted to maximize the yield stress, and (011) grains are predicted to have low yield stress. An analytic model for texture evolution during grain growth in thin films can be developed by equating the magnitudes of the orientation-dependent driving forces, for pairs of orientations. The analytic model can be used to generate texture maps that define which orientations are expected to grow preferentially as a function of the processing conditions, i.e., the deposition temperature, the grain growth temperature, and the film thickness. Experimental texture maps can be generated and used to test the validity of the analytic model.Computer simulations of grain growth have been carried out using a front-tracking simulation method. Interfacial energy, elastic and plastic strain energy density, and grain growth stagnation are accounted for in the simulations. Materials parameters characteristic of Ag/(001)Ni were used. The main result of the simulations is to validate the analytic model for texture evolution during grain growth. The computer simulations also provide insights into the coupling between yielding and grain growth.Grain growth experiments in Ag/(001)Ni/(001)Ag/(001)MgO, Ag/SiO2/MgO, Ag/SiO 2 /Si, Ni/SiO 2 /Si, and Al/SiO 2 /Si were carried out. Both the thickness and the thermal strain were systematically varied, and an experimental texture map was constructed for each system. The dependence of texture evolution on strain and thickness was found to be consistent with the trends predicted by the analytic model in all of these systems. While the texture map for Ag/(001)Ni was found in quantitative agreement with the model, with no adjustable parameters, no single set of fitting parameters was found for Ag/SiO 2 /Si and for Ag/SiO 2 fMgO. Possible origins of this discrepancy are discussed.Additional experiments are proposed that could provide a better under...