Development of Hybrid Deterministic-Statistical Models for Irradiation Influenced Microstructural Evolution.

Abstract

Ion irradiation holds promise as a cost-effective approach to developing structured nano--porous and nano--fiberous semiconductors. Irradiation of certain semiconductors leads to the development of these structures, with exception of the much desired silicon. Hybrid deterministic-statistical models were developed to better understand the dominating mechanisms during structuring.

This dissertation focuses on the application of hybrid models to two different radiation damage behavior: (1) precipitate evolution in a binary two-phase system and (2) void nucleation induced nano--porous structuring. Phenomenological equations defining the deterministic behavior were formulated by considering the expected kinetic and phenomenological behavior. The statistical component of the models is based on the Potts Monte Carlo (PMC) method. It has been demonstrated that hybrid models efficiently simulate microstructural evolution, while retaining the correct kinetics and physics. The main achievement was the development of computational methods to simulate radiation induced microstructural evolution and highlight which processes and materials properties could be essential for nano--structuring.

Radiation influenced precipitate evolution was modeled by coupling a set of non-linear partial differential equations to the PMC model. The simulations considered the effects of dose rate and interfacial energy. Precipitate growth becomes retarded with increased damage due to diffusion of the radiation defects countering capillarity driven precipitate growth. The effects of grain boundaries (GB) as sinks was studied by simulating precipitate growth in an irradiated bi-crystalline matrix. Qualitative comparison to experimental results suggest that precipitate coverage of the GB is due to kinetic considerations and increased interfacial energy effects.

Void nucleation induced nano--porous/fiberous structuring was modeled by coupling rate theory equations, kinetic Monte Carlo swelling algorithm and the PMC model. Point defect (PD) diffusivities were parameterized to study their influence on nano--structuring. The model showed that PD kinetic considerations are able to describe the formation of nano--porous structures. As defects diffuse faster, void nucleation becomes limited due to the fast removal of the defects. It was shown that as the diffusivities' ratio diverges from unity, the microstructures become statistically similar and uniform. Consequently, the computational results suggest that nano--pore structuring require interstitials that are much faster than the slow diffusing vacancies, which accumulate and cluster into voids.