Thesis (Ph. D.)--University of Rochester. Materials Science Program, 2012.
There exist many systems in Nature which present overall behavior of great complexity. Yet research in chemistry and physics has shown that the fundamental components of many systems are quite simple. The observed great complexity has its origin in the simple interactions between fundamental components. For example, corrosion of metals in aqueous environments involves elemental particles such as metal atoms, water molecules, and ions in the solution. Each elemental particle interacts with neighboring species according to the rules of Nature. When large amounts of these individual reactions act together, diverse corrosion phenomena of great complexity are generated. Conceptually, if a simulation correctly implements the interactions between species, system behaviors should spontaneously arise from the simulation.
The molecular automation developed in this work takes the approach of Nature to model the local interactions between individual particles, allowing corrosion phenomena to emerge spontaneously as outcomes of the simulation. Atoms, molecules, and ions are simplified to be particles of no size or weight. Each particle interacts with neighboring species based on the local information regarding its immediate neighbors. Species interactions are represented as thermally activated events which occur probabilistically, mimicking the thermal activation mechanism so prevalent in Nature. A set of rules that imitates the process of Nature is developed to govern the reactions between species. Since corrosion is an interfacial phenomena which considers both the solution and the metal phase, this work addresses the applications of molecular automation to the physical chemistry in solutions, the distribution of electrons in metals, and finally the structural features of the electrical double layer. Both two-dimensional and three-dimensional models have been developed. Parallel computation using Nvidia's graphics processing unit (GPU) and compute unified device architecture (CUDA) platform is employed to accelerate the computation. Simulation results presented in this work, including the ionic activity coefficients, corrosion reaction kinetics represented by the Tafel behavior, concentration polarization, pitting, surface deposition, distribution of electrons, and the electrical double layer, etc., are all emergent behaviors that spontaneously evolve from repeated elementary interactions between individual particles.
