Customer/Supplier Partnering to Provide Statistical Models to Predict New Hardware and Process Interactions and Limitations

摘要

As the semiconductor industry migrates to 200 mm wafers and < 0.5 μm and beyond manufacturing technologies [1], a tremendous responsibility is placed on the equipment suppliers to produce the reliable hardware, software and process(es) required to ensure integration success and continued enhanced technology and manufacturing yields. Traditionally, the customer characterizes their equipment and process(es) to meet their technology needs based on supplier's empirical recommendations, with this rarely resulting in complete shared data or joint learning activities. This paper describes an advanced sputter etch hardware and process characterization effort which produced results comparing the "traditional supplier's" characterization methods (single factor) versus the customer's statistical manufacturing characterization methods (Passive Data Collection and Design of Experiments). The results produced in this study, provided the supplier with a customer based statistical model to predict the behavior and limitations of their advanced sputter etch module, not to mention, an integration and manufacturing tool to predict process settings to achieve the desired technology results. These manufacturing statistical tools and methods, with partnering between a customer and supplier, could become the basis of all future supplier characterizations and support efforts to meet the growing costs and time-to-market needs of both the supplier and customer.