The application of AI techniques to the control and interpretation of C-V measurements

Abstract

Pattern-recognition and knowledge-based techniques are applied to help advance the interpretation of capacitance-voltage (C-V) curves. This is implemented by integrating instrument control software with an expert system shell to intelligently sequence tests to enhance conventional measurements. A prototype system is able to correctly identify a number of process faults, including a leaky oxide, as described in examples. In this instance some useful information could be obtained and a warning issued to the operator about the possible inaccuracy of some of the other parameters. Moreover, further analyses are disabled on account of the sample being inappropriate for their assumed equivalent circuit models. Other examples illustrate the improvements to be gained from measurements simply by recognizing the important factors in a single C-V measurement.<>