Probabilistic zero-dimensional model with calibration and uncertainty quantification for capacitively coupled plasma reactor simulations

Zero-dimensional (0D)/global models are used frequently in the modeling of plasma process reactors. The principal advantage is their simplicity and fast computational time. However, these models exhibit significant limitations in accurately simulating capacitively coupled plasmas (CCPs), primarily due to the lack of explicit model representations of certain physical processes, such as capacitive power coupling, which impacts the accuracy of the simulations. Here, we discuss an approach to test the fidelity of a typical 0D model through a formal uncertainty quantification, followed by the Bayesian calibration of key model parameters, to represent a CCP reactor under a range of operating conditions. The results show that the 0D model's predictive capabilities can be improved through this approach, but they also reveal the inadequacies that limit the 0D model's validity over broader parameter spaces. Consequently, relying on 0D models to predict capacitive plasma reactor behavior over a broad operating range is not possible, and this work underscores the need for higher-fidelity (multidimensional) models that can better capture the complexities of CCP discharges across wide operating conditions.