Doctor of Philosophy (PhD)
The process of nucleation is an essential part of understanding and controlling phase changes in a wide array of systems. In the past theories such as Classical Nucleation Theory have been used as a tool to aid experimentalist in the study of phase mechanisms. However, recent studies have shown in detail that theories such as this are not reliable, given that it can mispredict nucleation rates by several orders of magnitude. As a result newer methodologies must be developed into order to improve upon these deficiencies. In this study we use atomistic simulations to examine the non-ideal deviations from classical theory observed in both simple and complex systems. In addition to this we present new algorithms that can be used to improve the rate at which the nucleation properties of these simulations can be sampled. Lastly we apply these new methods to study an atmospherically relevant system that involves the nucleation of water in the presence of multiple charged ionic species. From these studies it was found that the deviation of more realistic systems from the classical theory can be attributed to both the creation of loosely bound clusters as well as the formation of highly ordered stacking in surface induced systems. The algorithms presented in this work have been shown to quickly and accurately replicate previously published data with very little increase to the computational overhead. Finally the application to the atmospherically relevant system showed an interesting trend where the nucleation rate was more heavily correlated to the number of water molecules that could be successfully bound to the ion pair instead of the quality of the bond.
Document Availability at the Time of Submission
Release the entire work immediately for access worldwide.
Loeffler, Troy David, "Advanced Monte Carlo Methods for the Study of Nucleation" (2016). LSU Doctoral Dissertations. 4280.