| In radio frequency capacitive discharges, ions are accelerated by the electric field in the sheath. During this acceleration, ions can collide with neutrals via charge exchange to create fast neutrals. The particles can then impact the wall with a distribution of energies and angles depending on the power source, gas constituents, pressure, and other reactor parameters. The distributions are important for materials processing, and therefore fast computation models to predict and control these are valuable.;An ideal model consists of a series of simple computational steps that result in accurate ion energy, ion angle, neutral energy, and neutral angle distributions (IED, IAD, NED, and NAD, respectively) at the wall given the input parameters. In addition, it predicts the input driving signals needed to give particular distributions. This dissertation covers three parts of this ideal model, which are developed and validated using one dimensional particle-in-cell (PIC) simulations.;The first topic deals with predicting the collisionless IED. This assumption is appropriate where the mean free path of the ions is large compared to the sheath width. The model gives the IED given the sheath voltage obtained from PIC simulations. The second topic deals with a synthesis process to estimate the input parameters needed for a uniform IED in a collisionless regime. A cumulative distribution function of the IED is used to determine the simplest ion response, which is then chosen as the driving voltage. Then the steps in the first topic are applied and compared to the results.;The last topic deals with IED, IAD, NED, and NAD from PIC in a collisional regime. At higher pressures, the mean free path of the ions is not large compared to the sheath width. Therefore, the ions will collide with the background gas as they accelerate across the sheath. As a result, there will be fewer energetic ions. In addition, fast neutral particles created by charge exchange and elastic scattering can dominate the heavy particle distribution. For computational expediency, low energy neutrals may be neglected with minimal physical consequences. Models have been developed to predict the NED. |
