Two-dimensional modeling of long-term transients in inductively coupled plasmas using moderate computational parallelism. I. Ar pulsed plasmas

Quantifying transient phenomena such as pulsed operation is important for optimizing plasma materials processing. These long-term phenomena are difficult to resolve in multidimensional plasma equipment models due to the large computational burden. Hybrid models, which sequentially execute modules addressing different phenomena, may not be adequate to resolve the physics of transients due to their inherent iterative nature. In this article, a different modeling approach is described in which a moderately parallel implementation of a two-dimensional plasma equipment model is used to investigate long-term transients. The computational algorithms are validated by comparing the plasma properties for sequential and parallel execution for a steady state case. The physics model is validated by comparison to experiments. Results from the model were used to investigate the transient behavior of pulsed inductively coupled plasmas sustained in Ar. The consequences of varying pulse repetition frequency, duty cycle, power, and pressure on plasma properties are quantified. We found that the electron density, temperature, and source function, and plasma potential, peak beneath the coils during avalanche at the beginning of a pulse, finally attaining a diffusion dominated profile with a small off axis peak. As the pulse repetition frequency decreases, a more pronounced local maximum in plasma potential and electron temperature occurs