Intelligent control of via formation process in MCM-L/D substrates using neural networks

Abstract

Via formation is a critical process sequence in multichip module (MCM) manufacturing, as it greatly impacts yield, density, and reliability. To achieve low-cost manufacturing, modeling, optimization, and control of via formation are therefore crucial. In this paper, a model-based supervisory control algorithm is developed and applied to reduce undesirable behavior resulting from various process disturbances. Sequential neural network process models are used for system identification, and genetic algorithms are applied to update process recipes. Computer simulation results show excellent control of output response shift and drift and reduction process of variation. The supervisory control algorithm is then verified experimentally, and control results again show significant improvement in Film thickness and via yield. It is expected that this system can also improve the other process responses when suitably tuned and improved.