Phase-Field Models for Simulating Physical Vapor Deposition and Microstructure Evolution of Thin Films.

摘要

The focus of this research is to develop, implement, and utilize phase-field models to study microstructure evolution in thin films during physical vapor deposition (PVD). There are four main goals to this dissertation. First, a phase-field model is developed to simulate PVD of a single-phase polycrystalline material by coupling previous modeling efforts on deposition of single-phase materials and grain evolution in polycrystalline materials. Second, a phase-field model is developed to simulate PVD of a polymorphic material by coupling previous modeling efforts on PVD of a single-phase material, evolution in multiphase materials, and phase nucleation. Third, a novel free energy functional is proposed that incorporates appropriate energetics and dynamics for simultaneous modeling of PVD and grain evolution in single-phase polycrystalline materials. Finally, these phase-field models are implemented into custom simulation codes and utilized to illustrate these models' capabilities in capturing PVD thin film growth, grain and grain boundary (GB) evolution, phase evolution and nucleation, and temperature evolution. In general, these simulations show: grain coarsening through grain rotation and GB migration such that grains tend to align with the thin film surface features and GBs migrate to locations between these features so that each surface feature has a distinct grain and orientation; the incident vapor flux rate controls the density of the thin film and the formation of surface and subsurface features; the substrate phase distribution initially acts as a template for the growing microstructure until the thin film becomes sufficiently thick; latent heat released during PVD increases the surface temperature of the thin film creating a temperature gradient within the thin film influencing phase evolution and nucleation; and temperature distributions lead to regions within the thin film that allow for multiple phases to be stable and coexist. Further, this work shows the sequential approach for coupling phase-field models, described in goals (i) and (ii) is sufficient to capture first-order features of the growth process, such as the stagnation of GBs at the valleys of the surface roughness, but to capture higher-order features, such as orientation gradients within columnar grains, the single free energy functional approach developed in goal (iii) is necessary.