Monolithic, tandem two-junction InxGa1-xP/InxGa1-x As-on-Ge Solar Photovoltaic Cells are being developed. The cells are designed to be used with solar concentrators to generate electric power aboard spacecraft. The theoretically achievable energy-conversion efficiency of these cells under air-mass-zero (AM0) conditions and 10× concentrated solar illumination is about 33 percent. In comparison with other known dual-junction photovoltaic cells, the proposed cells would feature highest conversion efficiency and the highest specific power. The cost per unit of power of these cells would be less than the costs per unit power of GaAs and InGaP/GaAs cells. With modifications, the cells might offer similar advantages in terrestrial applications.

To approach the theoretical efficiency, it is necessary to configure a tandem cell with an optimal combination of bandgaps. In the present case, this translates to putting the InxGa1-xP cell (which has a bandgap of 1.65 to 1.7 eV) on top and the InxGa1-xAs cell (which has a bandgap of about 1.1 eV) on the bottom. Although the InxGa1-xAs and InxGa1-xP are lattice-matched to each other, they are lattice-mismatched to all available substrate materials. Heretofore, the lack of a lattice-matched substrate material has impeded the realization of tandem cells with optimal bandgaps.

The Proposed InxGa1-xAs/InxGa1-xP-on-Ge Tandem Cell Structure is expected to afford unprecedented 1-Sun AM0 energy-conversion efficiency as high as 33 percent.

The approach taken in the present development effort is to form the tandem cell structure on a Ge substrate by organo-metallic vapor-phase epitaxy (OMVPE), using a lattice-grading technique to overcome the lattice mismatch. In addition, Ge offers the advantage of being inexpensive and rugged.

The figure depicts the proposed cell structure. The first layer deposited on the Ge substrate would be a buffer layer of GaAs, which is lattice-matched to Ge. Then a stack of InxGa1-xAs layers with increasing In contents would be deposited. The final InGaAs layer would have a total In content of about 23 atomic percent, as needed to obtain a bandgap of 1.1 eV for the lower cell. The subsequent lattice-matched layers (deposited without additional buffer layers) would include (1) InxGa1-xAs tunnel-junction layers, InxGa1-xAs layers of the bottom cell, more tunnel-junction layers, an AlInP window layer (bandgap 2.1 eV), InxGa1-xP layers for the top cell, and a top window layer of AlInP. An antireflection coating and a contact grid would be applied to the top window layer, and a contact layer would be formed on the bottom of the substrate.

Because of limitations of time and short-term unavailability of Ge substrates, GaAs substrates were used in initial experiments on the feasibility of OMVPE deposition of lattice-mismatched InxGa1-xAs. The experiments proved successful in that the uppermost InxGa1-xAs layers had the desired 1.1-eV bandgaps and the densities of dislocations in these layers were below the detection limit of transmission electron microscopy (107/cm²).

For other experiments to demonstrate the feasibility of the lower-cell design and part of the upper-cell design, InxGa1-xAs cells with InxGa1-xP window layers, without antireflection coatings, and with 19-percent front grid coverage, were fabricated on GaAs substrates. In the experiments, the best of these cells, with bandgaps of 1.07 and 1.15 eV, exhibited AM0 conversion efficiencies of 10.94 and 11.5 percent, respectively under 1-Sun illumination, and about 12.5 percent under 10-Sun illumination. Estimation of the effects of including antireflection coatings and reducing front grid coverage to a typical value of 5 percent leads to extrapolated 1-Sun AM0 conversion efficiencies of 17.5 and 18.4 percent, respectively. These efficiencies are almost as good as the 1-Sun AM0 efficiencies (about 20 percent) of the best Si cells.

This work was done by Richard W. Hoffman, Jr., Mark A. Stan, and Navid Fatemi of Essential Research, Inc., for Lewis Research Center. No further documentation is available. Inquiries concerning rights for the commercial use of this invention should be addressed to NASA Lewis Research Center, Commercial Technology Office,

Attn: Tech Brief Patent Status Mail Stop 7-3, 21000 Brookpark Road, Cleveland, Ohio 44135. Refer to LEW-16697.