Lower Cost and Higher Efficiency Solar Cells for 1-sun Applications

点击翻译按钮获取中文摘要

There are currently two broad technologies used for converting sunlight to electricity; Si solar cells used at 1-sun (in a flat panel format) and III-V based multijunction cells used in conjunction with concentration optics and tracking systems. As single junction devices, Si cells are limited theoretically to ~30% efficiency and have reached their practical limit of ~25%. Multijunction technology is likely to produce cells exceeding 50% efficiency under concentration this decade; however the cell costs are prohibitively high to be used in 1-sun applications. Therefore, the greatest imperative in Si cell technology is to reduce the cost, while maintaining high efficiency. There is also a tremendous opportunity to develop a multijunction cell technology that is sufficiently cost effective to be implemented in 1-sun applications. Based on recent research results, we have made significant progress in both directions and continued research efforts are likely to yield breakthroughs. Since the Si material is a significant fraction of the cell cost, there is substantial incentive to use thinner wafers than the current ~180 µm. However, the Si wafer in use today serves a dual purpose. It provides mechanical support for wafer handling and since Si is a poor optical absorber its thickness is required for optimal absorption. However, the optical path length can be increased by at least a factor of 50 using advanced light trapping strategies, with the potential to achieve efficiencies comparable to conventional thick cells, while virtually eliminating the material cost. We have fabricated 10 µm thick single-crystal Si-membrane solar cells, incorporating simple light trapping methods, yielding a device efficiency of ~10%. The design (via numerical simulation), fabrication (via low cost methods), optical characterization (corroborating designs) and implementation of more advanced light trapping strategies are expected to yield cells with thicknesses in the range of 2-10 µm and efficiencies exceeding 20%. Thin Si cells with advanced designs have the potential to dramatically reduce cost, while maintaining efficiency. In multijunction solar cells, each sub-cell is designed to optimally capture energy from a portion of the solar spectrum. An alternate approach to Ge substrates is to use much less expensive Si substrates for the bottom cell, leveraging the tremendous infrastructure of Si cell technology. There are two impediments with this approach, firstly that the Si cell becomes the current limiting cell for the triple junction device, compromising the energy conversion from the top cells and secondly that it has not proven possible to grow materials on Si substrates with sufficiently high quality. In our recent work, we have found solutions to both problems, with the invention of areal current matching to match the current of the Si cell to the top cells and the development of wafer bonding methods to join separately optimized Si and III-V cells in a hybrid integration approach, yielding cells with 25.8% efficiency, a world record for a Si-based multijunction solar cell. We believe that cells with efficiency in excess of 30% can be achieved with this method. However, our long term objective is to make multijunction cells that are sufficiently inexpensive for use in 1-sun applications, while exceeding the present limits of conventional 1-sun cells. We propose to make high quality III-V solar cells using a much less expensive process than the MOCVD process currently used and combine them with Si bottom cells via wafer bonding and areal current matching. This ambitious approach would be transformative, providing cells with significantly higher efficiencies than currently available, suitable for widespread deployment.