Matching automated CD SEMs in multiple manufacturing environments

Increasingly stringent critical dimension design rules for semiconductor manufacturing have driven manufacturers of automated CD SEMs to develop systems with improved line width measurement repeatability and reproducibility (Allgair et al., 1998). However, in a multiple tool manufacturing environment, the technical performance of CD SEMs is as much a function of consistent and tight operational controls as it is a function of the fundamental capability of the system. Efficient and strict methods for the characterization and monitoring of the measurement performance of the systems, and in particular system matching, are required to preserve proper operation. We describe a practical CD SEM control procedure, using a standard daily monitor wafer that tracks the major system components that affect CD SEM performance. A statistical analysis of this monitor data is presented which allows system matching to be verified immediately rather than requiring tests that span several days. This procedure tracks tool stability, provides a common CD SEM length reference, and enables the seamless use of multiple CD SEMs within a single manufacturing environment or between separate manufacturing environments, without significantly increasing the tool qualification time. Critical matching of six automated CD SEMs in two separate manufacturing environments is demonstrated using this technique on a variety of layers.