Automatic defect classification for integrated circuits

Abstract

Automatic inspection has become an essential part of manufacturing technology for integrated circuit chips. While initial detection of defects is the most critical function of inspection, automatic classification of detected defects is becoming increasingly desirable. Currently, defects detected automatically are reviewed manually and classified by operators. This process is often more time consuming than the initial inspection, and usually only a sample of the detected defects is reviewed and classified. By automating the process, one could improve the accuracy of defect statistics since operator subjectivity would be eliminated and significantly more defects could be included in the sample. Defect classification, however, is much more difficult than defect detection. To date only a few efforts have been reported in the literature, and as yet no commercial ("off the shelf") automatic defect classification (ADC) system is available. Semiconductor manufacturing is an application area that is becoming increasingly important and involves many complex processing steps. Controlling these processes is a critical problem, and has a significant impact on the yield of the manufacturing line. The key to better process control is reliable process measurement. The classification of defects provides valuable process diagnosis information. The hope is that machines can perform this task more reliably than humans. However, there are many problems in automating defect classification, and many of these are related to the central problems in artificial intelligence, such as knowledge representation, inferencing, and dealing with uncertainty. In this paper we pay special attention to the issues arising in the Automatic Defect Classification ( ADC) of integrated circuits. We first discuss technical and system requirements, followed by an outline of the technical challenges to be overcome to develop flexible and powerful ADC tools which can be quickly customized on a user level for diverse applications.