Tolerance of Three-Stage CIGS Deposition to Variations Imposed by Roll-to-Roll Processing: Phase II Annual Report, May 2003--May 2004

Abstract

Co-evaporation of CuIn x Ga 1-x Se 2 (CIGS) via two-stage and three-stage processes has been used in a variety of laboratories throughout the world to produce small-area devices with efficiencies greater than 15%. Thus, these deposition methods have come to be viewed as laboratory standards for the formation of CIGS absorbers used in respective photovoltaic devices. Although quite successful and relative easy to implement on a small R&D scale, scale-up to a commercially viable level proves to be rather challenging, as a number of conditions are encountered during continuous manufacturing that differ from the laboratory process. Such differences include both those imposed by continuous processing of moving substrates, and those implemented to decrease costs and increase throughput. It is therefore beneficial to understand the tolerance of the established laboratory processes to variations in deposition procedures. Research under this program consists of four basic parts to examine the tolerance of the established laboratory process to variations in deposition procedures: 1. Setting up the National Renewable Energy Laboratory (NREL)-developed three-stage CIGS laboratory process in a bell jar. (Phase I) 2. Characterizing the GSE roll-to-roll production chambers and device finishing steps in terms of the variables important to the laboratory processes. (Phase II) 3. Using the bell jar system to step incrementally from the NREL process to the conditions experienced by a sample during manufacturing, and characterizing the resulting films and devices. (Phase II and III) 4. Applying the process sensitivity information gained from the bell jar system to the production systems to realize improved device performance, yield, and process robustness. (Phase II and III) Work during Phase II progressed in three major areas: production system characterization, quantification of process sensitivities in the bell jar, and application of results to the production roll-coaters. To form a complete picture of conditions in the production systems, metals flux profiles as a function of Se pressure were documented, Se impingement rates were derived from sensor data, and thermocouple data was used to estimated temperature profiles. In the bell jar, the impact of the following process variations on device performance were quantified: CIGS cool-down rate, cool-down Se flux, venting temperature, shortened time between venting and CdS, maximum Cu ratio, and final Cu ratio. Related to production systems, the following topics were investigated: process robustness and device performance as a function of maximum and final Cu ratio, real-time sensing of maximum Cu ratio, effect of processing delays between CIGS and CdS, transparent conducting oxide stability with exposure to damp heat, and improved junction formation via thioacetamide treatment. Phase III tasks will build on this year's results: In the bell jar, several process sensitivity investigations will be concluded, and some new ones begun. These investigations will include use of less expen...