Network Behavior in Thin Film Growth Dynamics
StatisticsView Usage Statistics
Understanding patterns and components in thin film growth is crucial for many engineering applications. Further, the growth dynamics (e.g., shadowing and re-emission effects) of thin films exist in several other natural and man-made phenomena. Recent work developed network science techniques to study the growth dynamics of thin films and nanostructures. These efforts used a grid network model (i.e. viewing of each point on the thin film as an intersection point of a grid) via Monte Carlo simulation methods to study the shadowing and re-emission effects in the growth. These effects are crucial in understanding the relationships between growth dynamics and the resulting structural properties of the film to be grown. In this dissertation, we use a cluster-based network model with Monte Carlo simulation method to study these effects in thin film growth. We use image processing to identify clusters of points on the film and establish a network model of these clusters. Monte Carlo simulations are used to grow films and dynamically track the trajectories of re-emitted particles. We treat the points on the film substrate and cluster formations from the deposition of adatoms / particles on the surface of the substrate as the nodes of network, and movement of particles between these points or clusters as the traffic of the network. Then, graph theory is used to study various network statistics and characteristics that would explain various important phenomena in the thin film growth. We compare the cluster-based results with the grid-based results to determine which method is better suited to study the underlying characteristics of the thin film. Based on the clusters and the points on the substrate, we also develop a network traffic model to study the characteristics and phenomena like fractal behavior in the count and inter-arrival time of the particles. Our results show that the network theory of the growth process explains some of the underlying phenomena in film growth better than the existing theoretical and statistical models.