NEURAL-NETWORK MODELS FOR THERMALLY BASED MICROELECTRONIC MANUFACTURING PROCESSES

This paper presents artificial neural networks (ANNs) to model thermal ly based microelectronic manufacturing processes. The specific process es chosen are the chemically vapor deposited (CVD) epitaxial depositio n of silicon in a horizontal reactor and ''pool boiling'' as applied t o vapor-phase soldering. In the CVD processes, an analytic model is us ed to generate data under simulated production conditions. Part of the data sets are used to train the neural network models. These models, referred to hereafter as physiconeural models, are then used to predic t the output as a function of input parameters for the other part of t he data sets. For pool boiling, an empirical correlation is used to tr ain the ANN model. A comparison of these predictions with the physical model's computational results for the CVD process, and the experiment al data for the pool boiling shows good agreement. These results show the effectiveness of the artificial-neural-network technique for model ing complex processes. Further work is in progress to exploit fully th e potential of neural models, singly or in conjunction with physical m odels for run-to-run and real-time process control.