Self-consistent Monte Carlo simulations of plasma processing reactors

When the fractional ionization of a plasma exceeds 10$\sp{-5}- 10\sp{-4}$, electron-electron (e-e) collisions become important. These collisions cause the electron energy distribution (EED) to approach a Maxwellian distribution. Electron cyclotron resonance (ECR) reactors for etching and deposition have a high plasma density and fall into the category of devices for which e-e collisions must be considered. In this thesis, a self-consistent Monte Carlo (MC) simulation for low-temperature partially ionized plasmas is presented. In this simulation, the effects of electron-electron collisions are taken into account. Electron-electron collisions are treated as being functionally equivalent to electron-neutral collisions. That is, instead of having an electron collide with an individual electron, the electrons collide with an energy-resolved electron fluid. The modified null-cross-sectional technique is employed, making the MC simulation computationally tractable. The model is used to study ECR reactors and is a hybrid MC fluid model. The MC simulation generates details of the EED and the fluid model generates the ambipolar fields. The MC and fluid models are iterated to obtain a converged solution. The model has been utilized to investigate electron swarm parameters in ECR reactors for Ar and N$\sb2$ plasmas at different pressures and different input microwave powers. The parameters investigated are the EED, electron impact rate coefficients, average electron energy, plasma potential, and power deposition. The results are in general agreement with experiment.U of I OnlyETDs are only available to UIUC Users without author permissio