Real-time plasma etch control using in-situ sensors and neural networks

Recent work has shown that neural networks offer great promise in modeling complex fabrication processes such as reactive ion etching (RIE). This paper explores the use of neural networks for real-time, model-based feedback control of RIE. This objective is accomplished in part by constructing a predictive model for the system, which can be inverted (or approximately inverted) to achieve the desired control. The efficacy of this approach is demonstrated using experimental data from a SiO₂ etch process to simulate real-time control of etch rate, uniformity, selectivity, and anisotropy. In addition, using a residual gas analysis system as a sensor, the approach is further demonstrated using actual experimental data acquired during the etch of a GaAs/AlGaAs metal-semiconductor-metal structure. In the latter case, real-time control of the etch rate of the constituent materials is investigated