The objective of this proposal is to demonstrate the feasibility of a process for producing ultra-high conversion efficiency (37-41%) AM0 multijunction solar cells from dislocation-free non-lattice matched heterostructures using a proprietary wafer bonding / layer transfer process. The overall cell fabrication process will yield a two terminal, series-connected four junction InGaP/GaAs/InGaAsP/InGaAs/InP/Si solar cell. Specifically, as a proof of principle, we will transfer of Ge layers < 500 nm thick and 50 mm in diameter onto Si substrates, and then use the resulting Ge/Si substrates as epitaxial growth templates for high bandgap InGaP/GaAs tandem cells. We will also transfer < 500 nm thick InP layers and 50 mm diameter onto Si substrates and use the InP/Si substrates as epitaxial growth templates for low bandgap InGaAsP/InGaAs tandem cells. Device active region structure will be characterized via electron microscopy, X-ray diffraction, and minority carrier lifetime will be characterized via time-resolved photoluminescence. The dark and AM0-illuminated current-voltage characteristics of an InGaP/GaAs dual junction cell on Ge/Si templates will be determined. The results obtained will be used to estimate the overall four junction cell efficiency potential and guide development of a commercial prototype four junction cell process in Phase II.
Benefits: A commercially viable wafer bonding and layer transfer process that enables low-resistance, optically transparent interfaces between non-lattice matched materials could offer significant advantages to the development of next generation solar cells. Specifically, the ability to optimize the bandgap energies using dislocation-free sub-cell materials, which have up to now been constrained by lattice-matching requirements, could enable a conversion efficiency improvement from ~28% for todays lattice-matched 3J designs to over 40%. In addition, by enabling the use of lightweight, inexpensive non-lattice matched substrates (e.g., silicon), the process could dramatically improve the specific power of the cell while at the same time reducing its manufacturing costs. As a demonstration of the potential benefits of the project, a process for creating a 4 junction cell offering 40% efficiency at AMO, 25°C is described. Multi-junction cells like this one could offer immediate benefits to space power applications, and, longer term be incorporated into concentrator-based terrestrial designs. In addition, such a process would find immediate application in many other compound semiconductor markets particularly those that would benefit from the integration of non-lattice matched materials via conductive interfaces or those in which substrate costs represents a significant fraction of total device costs. Such markets include HB-LEDs, photodetectors, RF power amplifiers for cellular handsets, etc.
Keywords: multi-junction, solar cells, photovoltaics, high specific power, wafer bonding, layer transfer
