GOALI: Model-Based Equipment Control in Microelectronics Manufacturing

摘要-EDGAR/TRACHENBERG-9614174该项目致力于联合开发、实施、验证和转让用于下一代快速热处理的先进模型控制系统和用于单晶片加工的等离子刻蚀工具。这是一项由德克萨斯大学、先进微设备公司(AMD，奥斯汀)、制造商以及两家工具开发商和设备供应商--MATRIX集成系统公司(德克萨斯州奥斯汀)和AG Associates公司(加利福尼亚州圣何塞)--联合开展的项目。微电子制造需要先进的系统技术，以实现令人满意的产量，最大限度地提高生产能力并降低0.25微米半导体的成本。改进的数学模型、传感器和控制系统可以提供自动调整工艺条件的能力，以响应工艺环境和市场需求的变化。为了演示关键概念，PI将重点介绍使用单晶片反应堆的最新技术。对于各种商业设备系统，在快速热处理中，将建立基本(理论)数学模型，将操作和设备设计变量与性能联系起来。这些模型将使用位于不同工业现场的商业设备的生产数据进行验证。接下来，他们将研究新传感器和控制策略的应用，以便为每个反应堆设计先进的基于模型的控制器。由此产生的控制系统将用反应堆的实时计算机控制进行测试，使用现场直接和间接测量将进行调查，以改善控制器的性能。对于等离子刻蚀系统，他们将专注于监测终点信号，以确定何时到达终点或设备或晶片是否发生故障。

Abstract - Edgar/Trachtenberg - 9614174 This project involves efforts directed toward the joint development, implementation, validation and transfer of advanced model-based control systems for next generations rapid thermal processing (RTP) and plasma etching tools for single wafer processing. It is a joint program between the University of Texas, Advanced Micro Devices (AMD, Austin), a manufacturer, and two tool developers and equipment suppliers: Matrix Integrated Systems (Austin, TX) and AG Associates (San Jose, CA). Microelectronics manufacturing requires advanced systems technology in order to achieve satisfactory yields, maximize throughput and reduce costs for 0.25 micrometer semiconductors. Improved mathematical models, sensors, and control systems can provide the capability to automatically adjust processing conditions in response to changes in the process environment and market requirements. To demonstrate the key concepts, the PI's will focus on the latest technology using single-wafer reactors. For various commercial equipment systems, in rapid thermal processing, fundamental (theoretical) mathematical models will be developed that relate the operating and equipment design variables to performance. These models will be verified using production data from commercial equipment located at various industrial sites. Next they will research the application of new sensors and control strategies that permit design of advanced model-based controllers for each reactor. The resulting control systems will be tested with real-time computer control of the reactors use of in-situ direct and indirect measurements will be investigated to improve controller performance. For plasma etching systems, they will focus on the monitoring of end-point signals to determine when the end point is reached or if a fault has occurred with the equipment or wafer.