MIMO stochastic minimum variance control strategies for improving plasma characteristics in Reactive Ion Etching

In this paper, we propose to use multi input multi-output (MIMO) stochastic minimum variance control strategy to control the plasma characteristics in reactive ion etching (RIE) process. Given the RIE model in the state-space representation a stochastic adaptive minimum variance tracking controller (SMVAC) is designed and tuned to control the key plasma parameters. For the stochastic adaptive minimum variance approach a system identification scheme based on the minimum variance principle is integrated with a control strategy derived from minimization of a per-interval performance index. This control strategy performs very well, has good accuracy, as demonstrated in simulation environment, and can easily be applied to MIMO RIE system. The principal idea is that by controlling appropriate key plasma parameters (the concentrations of the reactive radicals and ions and ions energy) it is possible to improve the etch performance of the reactive ion etchers, namely their selectivity, uniformity, anisotropy and etch depth. The proposed real-time control strategies reveal superior accuracy and performance when compared to the results available in the plasma literature.