The Promise of Data-Driven Methods for Characterization, Diagnostics and Control of Plasma Processing of Complex Surfaces

Data-driven methods can create unprecedented opportunities for characterization, diagnostics and process control of non-equilibrium plasmas (NEPs), which are increasingly used for treatment of heat and pressure sensitive materials in surface etching/functionalization, environmental, and biomedical applications. Some of the main challenges in modeling and control of NEP applications arise from their inherent complexity and variability. Firstly, the behavior of NEPs are highly nonlinear and spatio-temporally distributed, which are hard to model due to their mechanistic complexity. Secondly, the NEP effects on complex surfaces are generally poorly understood. And thirdly, NEPs exhibit run-to-run variations and time-varying dynamics, whereby the same NEP treatment may be carried out under similar conditions, but yield different results. In this talk, we will discuss how advances in machine learning and data-driven optimization methods can be leveraged for: 1. modeling and simulation of NEPs to enhance understanding of plasma-surface mechanisms; 2. real-time diagnostics to “soft sense” plasma and surface characteristics; 3. plasma process control to realize reproducible and effective control of NEP processes; and 4. active learning-guided design of experiments to explore parameter space of NEP processes systematically.