NASA NTRS1995
It is well known that the behavior of III-V compound based solar cells is largely controlled by their surface, since the majority of light generated carriers (63% for GaAs and 79% for InP) are created within 0.2 microns of the illuminated surface of the cell. Consequently, the always observed high surface recombination velocity (SRV) on these cells is a serious limiting factor for their high efficiency performance, especially for those with the p-n junction made by either thermal diffusion or ion implantation. A good surface passivation layer, ideally, a grown oxide as opposed to a deposited one, will cause a significant reduction in the SRV without adding interface problems, thus improving the performance of III-V compound based solar cells. Another significant benefit to the overall performance of the solar cells can be achieved by a substantial reduction of their large surface optical reflection by the use of a well designed antireflection (AR) coating. In this paper, we demonstrate the effectiveness of using a chemically grown, thermally and chemically stable oxide, not only for surface passivation but also as an integral part of a 3- layer AR coating for thermally diffused p(+)n InP solar cells. A phosphorus-rich interfacial oxide, In(PO3)3, is grown at the surface of the p(+) emitter using an etchant based on HNO3, o-H3PO4 and H2O2. This oxide has the unique properties of passivating the surface as well as serving as a fairly efficient antireflective layer yielding a measured record high AM0, 25 C, open-circuit voltage of 890.3 mV on a thermally diffused InP(Cd,S) solar cell. Unlike conventional single layer AR coatings such as ZnS, Sb2O3, SiO or double layer AR coatings such as ZnS/MgF2 deposited by e-beam or resistive evaporation, this oxide preserves the stoichiometry of the InP surface. We show that it is possible to design a three-layer AR coating for a thermally diffused InP solar cell using the In(PO3)3 grown oxide as the first layer and Al2O3, MgF2 or ZnS, MgF2 as the second and third layers respectively, so as to yield an overall theoretical reflectance of less than 2%. Since chemical oxides are readily grown on III-V semiconductor materials, the technique of using the grown oxide layer to both passivate the surface as well as serve as the first of a multilayer AR coating, should work well for essentially all III-V compound-based solar cells.