Continuous, Automated Manufacturing of String Ribbon Si PV Modules: First Annual Report, 21 May 1998 - 20 May 1999

This report describes the first year of a 3-year PVMaT Phase 5A2 program. Evergreen Solar will soon be expanding into a multi-megawatt facility and the PVMaT work will be used to further the objective of a high-throughput automated production line in every aspect of PV module making from producing silicon ribbon to making a finished module. The project has four task areas for this first year:crystal growth; cell manufacturing; modules; and factory layout and automation. The vast majority of the work this first year has been in the crystal growth and cell manufacturing areas. Evergreen Solar has its own unique technology in each of these areas. In crystal growth, a key goal of this PVMaT project has been developing and deploying an improved string material. The high-temperaturestring materials are used to stabilize the edges of the growing silicon ribbon in our String Ribbon silicon sheet growth. The result has been one of the major successes of this first year. Significant cost reductions and yield improvements have emerged from this improved string material. In addition, some of the groundwork for automation of the String Ribbon crystal growth process has been laidand shows much promise. A method for controlling the edge meniscus height was developed, and a patent has been filed based on this discovery. In the cell manufacturing area, the focus has been on reducing the number of processing steps and on design and construction of high-speed processing equipment. The possibility of eliminating all pre-diffusion etching and going directly from growth todiffusion has been demonstrated on an R&D scale. Unique designs for high-speed drying equipment and for a high-speed contact and AR-coating application machine have been developed. In the latter case, the basic concepts underlying various aspects of the machine design have been successfully tested for viability. The integration of the different components of this machine into a smoothly workingwhole is now well under way. In addition, nonvacuum methods of hydrogen passivation have been investigated, but no production-viable process has emerged from this work. In the module area, initial work has shown that the unique properties of the backskin developed under Evergreen's earlier PVMaT Phase 4A1 contract can be expected to make the lay-up process simpler and involve less labor. Forfactory layout and automation, we have enlisted the Fraunhofer Manufacturing Institute at Boston University to aid us with the layout and process flow of our new factory. In general, the project is on schedule, with no significant technical barriers to reach our ultimate objectives.