Modelling microwave plasmas for deposition purposes : exploring the freedom in space and chemistry

This thesis deals with the modelling of various microwave produced plasmas that are in use for the plasma enhanced chemical vapour deposition process in the fabrication of optical glass fibres. To that end, the platform plasimo was employed to construct a self-consistent simulation package. Several parts of plasimo have been improved, added or adjusted in order to perform these simulations. The completed simulations are valuable instruments for the analysis and optimization of the fibre production process. Microwave induced plasmas offer a manifold of novel high-tech industrial potentials. Using microwaves enables electrodeless plasma operation and has the advantage that the propagation of the waves can be modified by dielectrics and the resonator geometry like slit and chokes. Modelling microwave induced plasmas introduces several challenges. The constructed models consist of the interplay between three main modules: transport, chemistry and electromagnetic power coupling. The convective transport is modeled by solving the Navier-Stokes equations for the bulk flow, whereas the Stefan-Maxwell equations are applied to describe the diffusion of the various species with respect to the barycentric velocity. The electromagnetic energy coupling module is mainly dedicated to one specific resonator set-up. The fields in the plasma that result from an interplay between the applied field, the cavity shape, the dielectrics and the plasma are calculated self-consistently by using the Maxwell equations in the frequency domain. The resulting electric field gives the leading source term of the electron energy balance. In this resonator, the influence of dimensions and placement of features like slit and chokes are investigated. Other simulations use an approximate electromagnetic power coupling module for an axially symmetric self-guiding wave configuration; the surfatron. With respect to the chemistry, several compositions are investigated like argon and oxygen. To find a solution for the numerical difficulties we constructed a dedicated tool, PyRate, that integrates a set of equations for species concentrations in time. This gives insight in the various time constants, the main particle reservoirs and suitable initial conditions to run the two-dimensional simulations. The PyRate tool is applied to gas compositions like tetrachlorosilane and oxygen. Ultimately, the behaviour of precursors for the deposition of glass layers is investigated.