Recipe generation for large-area meniscus coating using Kalman filter estimation

Material deposition is one of the repeated, and important, steps in electronics manufacturing. Meniscus coating is a low-waste technique that may be used in applications where spin coating is found. The process exhibits time-varying behavior and thus utilizing feedback control methods to compensate for this behavior is desirable. This paper describes the large-area meniscus coating process, explains modeling techniques used and presents results from closed-loop experiments. Feedback in the form of Kalman filtering and EWMA filtering is contrasted, with Kalman Filter techniques improving performance considerably. These process control techniques require no additional hardware and therefore very little cost is associated with implementation of these feedback algorithms