This Small Business Innovation Research (SBIR) Phase I project aims to substantially reduce the cost of multijunction solar cells for terrestrial concentrator photovoltaic (CPV) solar power generation. The presently favored type of solar cell for CPV applications is the inverted metamorphic multijunction (IMM), which offers the highest conversion efficiency. However, the thickest and most costly single region of this cell is the buffer layer, which provides a transition in the lattice constant between subcells, but does not contribute to current generation. Plus, the metamorphic subcell, which is grown on the buffer layer, is adversely affected by defects which originate in the buffer layer. In this project, the buffer layer will be replaced with a nanopatterned layer, which will allow the metamorphic subcell to form in a short distance with few defects.
The broader/commercial impact of this project will be the potential to provide a new technique to fabricate IMM solar cells, allowing a significant cost reduction. IMM solar cells are a key component of terrestrial CPV solar power generators and account for a substantial fraction of their cost. A reduction in the cost of IMM cells should therefore lower the cost of solar-generated electricity. This approach is expected to achieve a 50% reduction in growth time and a 30% reduction in the bill of materials (BOM) of IMM cells.
The objective of the proposed work was to substantially reduce the cost of high-efficiency, multijunction solar cells by replacing the thick buffer layer in with a nanopatterned SiO2 layer. By using nanometer patterned openings in a SiO2 covered substrate , it is expected to virtually eliminate the thick buffer layer and greatly reduce threading dislocations or defects in the metamorphic subcell, thereby reducing the cost. The reliability of the solar cells could also be improved with a reduction of the dislocation density. Transfer of a nano-pattern appears complete over a 100 mm diameter wafer. PS-b-PMMA chemistry appears to be a good candidate for this approach. Etching chemistries are adequate for pattern transfer but additional experiments are needed to optimize times and powers. The feasibility of this technology over large 4-inch GaAs substrate areas was demonstrated for the first time, which is attractive for high volume manufacturing. Single crystal material was demonstrated in the epitaxial overgrowth materials. The quality of the material could be improved with additional surface preparation to ensure a high quality interface prior to growth. The potential of this technology to demonstrate higher performance devices was demonstrated. This technology has wide applications for other devices that are epitaxially grown in material systems such as GaAs, GaN, and InP.
