<b>Computational modeling of the growth of oxide-metal multiphase thin film with pillar-in-matrix configuration</b>

<p dir="ltr">The interesting physical and chemical properties of multiphase self-assembled oxide-metal thin films with pillar-in-matrix configurations, known as vertically aligned nanocomposite (VAN), have received a great deal of attention in recent years. The growth and morphology evolution processes of such films are complicated and, as a result, our current understanding of them is still limited. Recent research in the field has suggested that the ordering and evolution of the embedded phase configuration in the matrix can be described in terms of the lattice-mismatch elastic strain and the interfacial energy. From the self-assembly point of view, adatom diffusion on the surface, binding energies, as well as the effect of interfaces and elastic stain on the migration and binding energies, all impact the nucleation and growth of the film phases. The growth of VAN systems is also affected by the physical and chemical conditions of the pulsed laser deposition method.</p><p dir="ltr">In this dissertation, the kinetic Monte Carlo (kMC) method parameterized by density functional theory (DFT) has been used to simulate multiphase film growth. In these multiphase films, elastic strain arises due to lattice and thermal mismatch and the heterogeneity of the material properties. Focusing on the growth of Au-CeO₂ system on SrTiO₃ as a model system, DFT was used to obtain the hoping and binding energies of adatoms (Au) and admolecules (CeO₂), and to understand the impact of strain on the binding and hopping energies and the hopping pathways on SrTiO₃, Au and CeO2 surfaces. In addition, the elastic problem resulting from lattice and thermal mismatch between the substrate, matrix and pillar materials is cast in the form of Representative Volume Element (RVE) with average constraints and solved using mesh-free Fast Fourier Transform (FFT). An effective methodology for incorporating the elastic effects in kMC simulations has been developed, whereby the DFT results for the binding and hopping energies were fit to the strain state in an analytical form, while the FFT provides the strain state on the fly during kMC simulations. The resulting DFT-kMC-FFT framework has been used to investigate the growth of Au-CeO₂ pillar-matrix system on (100) SrTiO₃ substrate and analyze the effect of growth conditions of temperature and deposition rates on the growth morphology.</p><p dir="ltr">It was found that the kMC model captures the correct growth behavior of Au pillars in CeO₂ matrix in terms of the dependence of the density and average pillar size on temperature and pulsed laser deposition (PLD) frequency. Without strain, it was found that kMC simulations underestimates the pillar size due to limiting the deposition and diffusion of adatom and admolecuels to be within the topmost monolayer, while considering growth to occur one layer at a time. Upon including the strain effects, the kMC results were found to come closer to experiments in terms of pillar size. The prediction of pillar shape and pillar arrangement also improved, bringing us close to a full understanding of the factors that affect self-organization of pillars.</p>