Modelling nanoscale kinetics of radiation damaged surfaces

Abstract

Materials in nuclear reactors and satellites experience continually damaging radiation which leads to their degradation over time. Currently, a materials safe working lifetime within these environments is estimated with a large, costly, safety margin. The work of this thesis aims to improve the usefulness of an optical technique known as reflection anisotropy spectroscopy (RAS), which once fully characterised could allow materials to be actively monitored in such environments. The intrinsic optical anisotropy of the Cu(110) surface has been exploited to study nanoscale kinetics of ion bombarded surfaces. Within the Cu(110) RA spectrum the 2.1eV peak is particularly sensitive to surface defects and largely unaffected by the bulk of the substrate. Using the Poelsema-Comsa model (which assumes defects scatter surface electronic states within a patch centred on the defect) it can be demonstrated that at finite temperatures the decay of the 2.1eV peak contains information relating to the diffusion of surface defects. A kinetic Monte Carlo simulation has been created to model the destruction of this peak and allows further understanding of the diffusion processes involved.

The decay of the 2.1eV peak with ion bombardment has been successfully modelled for a range of temperatures using experimental RAS data for comparison. Through a novel way of analysing RAS data, it has been shown that the total scattering cross section per ion impact decreases with bombardment time, which it is believed to be due to surface diffusion. This could give a novel way of measuring surface diffusion directly from RAS measurements.

Clustering of ion induced surface defects has been analysed and the results found are consistent with STM images of the same surface obtained 30 minutes after bombardment. While molecular dynamics calculations have previously attempted to predict the surface topology and defect clustering nanoseconds after impact, using a kinetic Monte Carlo simulation improves on this, demonstrating that diffusion on long time scales (currently inaccessible using molecular dynamics calculations) play an important role in predicting nano-surface topology.

2.1eV peak recovery after surface damage by ion bombardment was also investigated. The peak was found to recover at finite temperatures, which is also seen in experimental data. It was concluded that the surface diffusivity values in the literature are too high and a new value for diffusivity has been calculated by comparing simulation and experimental data.