This Small Business Innovative Research (SBIR) Phase I project will demonstrate the feasibility of using model-based control of temperature and stress for millisecond flash annealing of semiconductor wafers. In response to the demand for ever faster chips that consume less power, the tools for fabrication of integrated circuits (IC's) are required to provide increasingly precise control of the manufacturing process, e.g., control of the transient wafer temperature during thermal processing. The manufacture of an IC involves hundreds of steps. One of these steps is thermal annealing used for dopant activation, a critical step in IC chip manufacturing. Next-generation IC's need very shallow and abrupt junctions that require annealing over shorter time periods and at higher temperatures, which is not possible with current conventional spike annealing equipment. Millisecond annealing with pulsed flash-lamps has the potential to meet the new annealing requirements. However, challenges related to reproducibility and stress-induced wafer damage must be overcome before flash anneal can be used in high-volume manufacturing. The project proposes to develop and commercialize a control software product to help accelerate the transition of flash anneal to the semiconductor industry.
The broader impact/commercial potential of this project is in its ability to help enable widespread adoption of flash anneal during semiconductor wafer processing. There is a significant potential for millisecond flash anneal to complement and eventually replace conventional lamp-based Rapid Thermal Processing (RTP) for some important steps in semiconductor manufacturing. The RTP market has reached close to $308 million in 2008, and is expected to grow significantly in the future. The proposed product will enable flash annealing to meet the formidable process control requirements and economic metrics, and help transition it to high-volume manufacturing. This will benefit customers and will in turn benefit the IC makers and ultimately benefit US consumers with smaller, faster, lower power devices. These innovations will make a significant contribution to advancing production technologies for future microelectronic devices, which will benefit the society at large.
This Small Business Innovation Research Phase I project demonstrated the feasibility of using model-based control of temperature and stress for millisecond flash annealing of semiconductor wafers. Thermal annealing is a process in which the wafer (of diameter 300 mm) is heated very rapidly (a few milliseconds) that allows critical chemical changes to occur that are necessary for the integrated circuits (IC) on the wafer to work. These IC’s increasingly require annealing over shorter time periods and at higher temperatures, which is not possible with conventional spike annealing. Flash annealing, in which a semiconductor wafer is subjected to an intense heat pulse that raises the wafer temperature to about 1000°C within a few milliseconds, has the potential meet the process requirements. However, challenges related to process reproducibility and stress-induced wafer damage must be overcome before flash anneal can be used in high-volume manufacturing. A wafer that is shattered during flash anneal due to excessive thermal stress has to be discarded is a significant loss. However, the resulting interruption to production in a billion-dollar manufacturing fab, in order to clean up the equipment, would lead to much larger losses. A real-time, feedback controller is needed for flash anneal equipment to be a viable production-worthy tool. Development of such a controller is a challenging problem because the process is extremely fast, lasting only a few milliseconds. In this Phase I program, SC Solutions, a leader in providing advanced control solutions to the semiconductor industry, has laid the groundwork towards developing a control software product in Phase II to help enable the transition of flash anneal to the semiconductor industry. SC Solutions is partnering with a leading manufacturer of thermal processing equipment and a pioneer in flash anneal. SC Solutions first developed an accurate physical model for a commercial flash anneal system using a commercial finite element modeling tool. Model predictions for wafer deflection due to thermal stresses during flash heating were compared with experimental data from our industrial partner’s commercial flash anneal chamber. The agreement between experimental data and the model prediction is excellent. The model revealed that the maximum likelihood of wafer damage occurred early in the process. This experimentally-validated physical model was then used in the design of a feedback controller, which demonstrated significant improvements in keeping thermals stresses below specified threshold as compared to the open-loop flash anneal system. All the goals in the Phases I were achieved or exceeded. We believe our results are superior to the results in the literature and a significant advancement in the state of the art in modeling and control of flash anneal systems. There is a significant potential for flash RTP anneal to complement and eventually replace conventional lamp-based Rapid Thermal Processing (RTP) in semiconductor manufacturing. The RTP market has reached close to $308 million in 2008, and is expected to grow significantly in the future. The proposed product will enable flash annealing to meet the formidable process control requirements and economic metrics, and help transition it to high-volume manufacturing. Our industrial partner is a leading U.S. manufacturer of thermal processing equipment for the semiconductor industry, and will directly benefit from improved flash anneal process performance. This benefit to our partner will in turn benefit their customers, the IC makers, and ultimately benefit U.S. consumers with smaller, faster, lower power devices. The proposed innovations are necessary for the commercial manufacturing of future IC designs. Finally, microelectronics affects almost every aspect of our lives. The proposed work makes a significant contribution to advancing production technologies for future microelectronic devices, which will benefit the society at large.
