Engineering topics
Publications and source records attributed to Jain, R. K..
Indium phosphide is considered to be a strong contender for many photovoltaic space applications because of its radiation resistance and its potential for high efficiency. An overview of recent progress is presented, and possible future research directions for indium phosphide space solar cells are discussed. The topics considered include radiation damage studies and space flight experiments.
Ga(0.47)In(0.53)As solar cells were processed by OMVPE and their characteristics determined at proton energies of 0.2, 0.5, and 3 MeV. Emphasis was on characteristics applicable to use of this cell as the low bandgap member of a monolithic, two terminal high efficiency InP/GaInAs cell. It was found that the radiation induced degradation in efficiency, I(sub SC), V(sub OC) and diffusion length increased with decreasing proton energy. When efficiency degradations were compared with InP it was observed that the present cells showed considerably more degradation over the entire energy range. Similar to InP, R(sub C), the carrier removal rate, decreased with increasing proton energy. However, numerical values for R(sub C) differed from those observed with InP. The difference is attributed to differing defect behavior between the two cell types. It was concluded that particular attention should be paid to the effects of low energy protons especially when the particle's track ends in one cell of the multibandgap device.
The highest AMO efficiency (19.1 percent) InP solar cell consisted of an n+pp+ structure epitaxially grown on a p+ InP substrate. However, the high cost and relative fragility of InP served as motivation for research efforts directed at heteroepitaxial growth of InP on more viable substrates. The highest AMO efficiency (13.7 percent) for this type of cell was achieved using a GaAs substrate. Considering only cost and fracture toughness, Si would be the preferred substrate. The fact that Si is a donor in InP introduces complexities which are necessary in order to avoid the formation of an efficiency limiting counterdiode. One method used to overcome this problem lies in employing an n+p+ tunnel junction in contact with the cell's p region. A simpler method consists of using an n+ substrate and processing the cell in the p+ nn+ configuration. This eliminates the need for a tunnel junction. Unfortunately, the p/n configuration has received relatively little attention the best cell with this geometry having achieved an efficiency of 17 percent. Irradiation of these homoepitaxial cells, with 1 Mev electrons, showed that they were slightly more radiation resistant than diffused junction n/p cells. Additional p/n InP cells have been processed by some activity aimed at diffusion. Currently, there has been some activity aimed at producing heteroepitaxial p+nn+ InP cells using n+ Ge substrates. Since, like Si, Ge is an n-dopant in InP, use of this configuration obviates the need for a tunnel junction. Obviously, before attempting to process heteroepitaxial cells, one must produce a reasonably good homoepitaxial cell. In the present case we focus our attention on homoepitaxially on an n+ Ge substrate.
InP p(+)nn(+) cells, processed by MOCVD, were irradiated by 0.2 MeV protons and their performance and defect behavior observed to a maximum fluence of 10(exp 13)/sq cm. Their radiation induced degradation, over this fluence range, was considerably+less than observed for similarly irradiated, diffused junction n p InP cells. Significant degradation occurred in both the cell's emitter and base regions the least degradation occurring in the depletion region. A significant increase in series resistance occurs at the highest fluenc.e. Two majority carrier defect levels, E7 and E10, are observed by DLTS with activation energies at (E(sub C) - 0.39)eV and (E(sub C) - 0.74)eV respectively. The relative concentration of these defects differs considerably from that observed after 1 MeV electron irradiation. An increased carrier concentration in the cell's n-region was observed at the highest proton fluence, the change in carrier concentration being insignificant at the lower fluences. In agreement with previous results, for 1 and 1.5 MeV electron irradiated InP p(+)n junctions, the defect level E10 is attributed to a complex between zinc, diffused into the n-region from the zinc doped emitter, and a radiation induced defect. The latter is assumed to be either a phosphorus vacancy or interstitial. The increased, or enhanced carrier concentration is attributed to this complex acting as a donor.
Indium phosphide solar cells are more radiation resistant than gallium arsenide and silicon solar cells, and their growth by heteroepitaxy offers additional advantages leading to the development of lighter, mechanically strong and cost-effective cells. Changes in heteroepitaxial InP cell efficiency under 0.5 and 3 MeV proton irradiations are explained by the variation in the minority-carrier diffusion length. The base diffusion length versus proton fluence is calculated by simulating the cell performance. The diffusion length damage coefficient K(L) is plotted as a function of proton fluence.
An overview of the current status and future directions of InP space solar cell research is provided. The scope of the paper does not allow us to discuss other recent major developments in InP cell modeling, contacts, and characterization, or developments in other solar cell materials. Solar cells made from InP and related materials are not expected to be used in the near future for terrestrial applications, but significant Air-Mass1.5 (AM1.5) cell efficiencies are given for comparison. This paper deals with the developments in single-junction cells, multijunction tandem cells, and space flight testing, including radiation effects. Concentrator InP solar cells are also discussed, since they offer the possibility of simultaneous thermal and current injection annealing. These cells also promise cost effectiveness and the concentrator elements may provide cells with extra protection from space radiation. The concluding section addresses the steps to be taken in the future and provides guidelines for further research and development.
The performance of indium phosphide solar cells with lattice matched wide band-gap In(0.52)Al(0.48)As window layers was calculated using the PC-1D computer code. The conversion efficiency of p(+)n InP solar cells is improved significantly by the window layer. No improvement is seen for n(+)p structures. The improvement in InP cell efficiency was studied as a function of In(0.52)Al(0.48)As layer thickness. The use of the window layer improves both the open circuit voltage and short circuit current.For a typical In(0.52)Al(0.48)As window layer thickness of 20 nm, the cell efficiency improves in excess of 27 percent to a value of 18.74 percent.
The comparative modeling of p(+)n and n(+)p indium phosphide solar cell structures is studied using a numerical program PC-1D. The optimal design study has predicted that the p(+)n structure offers improved cell efficiencies as compared to n(+)p structure, due to higher open-circuit voltage. The various cell material and process parameters to achieve the maximum cell efficiencies are reported. The effect of some of the cell parameters on InP cell I-V characteristics was studied. The available radiation resistance data on n(+)p and p(+)p InP solar cells are also critically discussed.
Significant new developments in InP solar cell research are reviewed. Recent accomplishments include monolithic multibandgap two junction cells (three and two terminal) using InP as the top cell and lattice matched GaInAs and GaInAsP as the bottom, low bandgap component. Concentrator cells include the three terminal multibandgap cell and n + p cell using an InP substrate. The review also includes small scale production of ITO/InP cells and results for n + p InP and ITO/InP cells in space on board the LIPS 3 satellite.
Significant new developments in InP solar cell research are reviewed. Recent accomplishments include monolithic multibandgap two junction cells (three and two terminal) using InP as the top cell and lattice matched GaInAs and GaInAsP as the bottom, low bandgap, component. Concentrator cells include the three terminal multibandgap cell and an n + p cell using an InP substrate. The review also includes small scale production of ITO/InP cells and results for n+p InP and ITO/InP cells in space on board the LIPS 3 satellite.
Calculations are made to study the dependence of heteroepitaxial InP solar-cell efficiency on dislocation density. Effects of surface recombination velocity and cell emitter thickness are considered. Calculated results are compared with the available experimental results on representative InP solar cells. It is shown that heteroepitaxial InP cells with over 20 percent AM0 efficiency could be fabricated if dislocations are reduced to less than 100,000/sq cm.
Indium phosphide (InP) solar cell efficiencies are limited by surface recombination. The effect of a wide-bandgap lattice-matched indium aluminum arsenide (In0.52Al0.48As) window layer on the performance of InP solar cells was investigated using a numerical code PC-1D. The p(+)n InP solar cell performance improves significantly with the use of a window layer. No improvement is seen for n(+)p InP cells. Cell results are explained by the band diagram of the heterostructure and the conduction-band energy discontinuity. The calculated I-V and internal quantum efficiency results clearly demonstrate that In0.52Al0.48As is a promising candidate as a window layer material for p(+)n InP solar cells.
Minority carrier diffusion length is one of the most important parameters affecting the solar cell performance. An attempt is made to estimate the minority carrier diffusion lengths is the emitter and base of InP/GaAs heteroepitaxial solar cells. The PC-1D computer model was used to simulate the experimental cell results measured at NASA Lewis under AMO (air mass zero) spectrum at 25 C. A 16 nm hole diffusion length in the emitter and a 0.42 micron electron diffusion length in the base gave very good agreement with the I-V curve. The effect of varying minority carrier diffusion lengths on cell short current, open circuit voltage, and efficiency was studied. It is also observed that the front surface recombination velocity has very little influence on the cell performance. The poor output of heteroepitaxial cells is caused primarily by the large number of dislocations generated at the interfaces that propagate through the bulk indium phosphide layers. Cell efficiency as a function of dislocation density was calculated and the effect of improved emitter bulk properties on cell efficiency is presented. It is found that cells with over 16 percent efficiencies should be possible, provided the dislocation density is below 10(exp 6)/sq cm.
Metalorganic chemical-vapor-deposited heteroepitaxial indium phosphide (InP) solar cell experimental results were simulated by using a PC-1D computer model. The effect of emitter parameter variation on the performance of n(+)/p/p(+) heteroepitaxial InP/GaAs solar cell was presented. The thinner and lighter doped emitters were observed to offer higher cell efficiencies. The influence of emitter thickness and minority carrier diffusion length on the cell efficiency with respect to dislocation density was studied. Heteroepitaxial cells with efficiencies similar to present day homojunction InP efficiencies (greaater than 16 percent AM0) were shown to be attainable if a dislocation density lower than 10(exp 6)/sq cm could be achieved. A realistic optimized design study yielded InP solar cells of over 22 percent AM0 efficiency at 25 C.
The magnitude of the photon-induced current in Ag-Al2O3-Al metal-oxide-metal junctions has been studied as a function of photon energy and angle of incident radiation. Photocurrents were theoretically analyzed on the basis of a modified vacuum photoemission model (Jain, 1975; Slayman et al., to be published). Optical constants previously reported in the literature (Irani et al., 1971; Ehnrereich et al., 1963) were used to calculate the true spatial generation rate in Ag and Al as a function of the angle, polarization of incident radiation, and film thickness. Results were found to be in very good agreement with experimentally determined values for a tunable dye laser with a KDP doubling crystal pumped by a Q-switched Nd:YAG laser with a LiIO3 doubling crystal. The system provided risetimes of 50 ns or less and peak powers of 10 W. Under short circuit conditions, the photoresponse to incident power was linear up to available power densities of 10 kW/sq cm. Quantum efficiencies of about 0.1% at zero-bias, near 3.8 eV under P polarization, were typically observed.
Explore the source record for details and available documents.
There are no author-identified significant results in this report.
There are no author-identified significant results in this report.