Search

Multijunction III-V concentrator cells of several different types have demonstrated solar conversion efficiency over 40% since 2006, and represent the only third-generation photovoltaic technology to enter commercial power generation markets so far. The next stage of solar cell efficiency improvement, from 40% to 50%-efficient production cells, is perhaps the most important yet, since it is in this range that concentrator photovoltaic (CPV) systems can become the lowest cost option for solar electricity, competing with conventional power generation without government subsidies. The impact of 40% and 50% cell efficiency on cost-effective geographic regions for CPV systems is calculated in the continental US, Europe, and North Africa. We take a systematic look at a progression of multijunction cell architectures that will take us up to 50% efficiency, using modeling grounded in well-characterized solar cell materials systems of today's 40% cells, discussing the theoretical, materials science, and manufacturing considerations for the most promising approaches. The effects of varying solar spectrum and current balance on energy production in 4-junction, 5-junction, and 6-junction terrestrial concentrator cells are shown to be noticeable, but are far outweighed by the increased efficiency of these advanced cell designs. Production efficiency distributions of the last five generations of terrestrial concentrator solar cells are discussed. Experimental results are shown for a highly manufacturable, upright metamorphic 3-junction GaInP/GaInAs/Ge solar cell with 41.6% efficiency independently confirmed at 484 suns (48.4 W/cm2) (AM1.5D, ASTM G173-03, 25 °C), the highest demonstrated for a cell of this type requiring a single metal-organic vapor-phase epitaxy growth run. Copyright © 2012 John Wiley & Sons, Ltd.

Concerns about the changing environment and fossil fuel depletion have prompted much controversy and scrutiny. One way to address these issues is to use concentrating photovoltaics (CPV) as an alternate source for energy production. Multijunction solar cells built from III–V semiconductors are being evaluated globally in CPV systems designed to supplement electricity generation for utility companies. The high efficiency of III–V multijunction concentrator cells, with demonstrated efficiency over 40% since 2006, strongly reduces the cost of CPV systems, and makes III–V multijunction cells the technology of choice for most concentrator systems today. In designing multijunction cells, consideration must be given to the epitaxial growth of structures so that the lattice parameter between material systems is compatible for enhancing device performance. Low resistance metal contacts are crucial for attaining high performance. Optimization of the front metal grid pattern is required to maximize light absorption and minimize I2R losses in the gridlines and the semiconductor sheet. Understanding how a multijunction device works is important for the design of next-generation high efficiency solar cells, which need to operate in the 45%–50% range for a CPV system to make better economical sense. However, the survivability of solar cells in the field is of chief concern, and accelerated tests must be conducted to assess the reliability of devices during operation in CPV systems. These topics are the focus of this review.

The performance of multijunction solar cells has been measured over a range of temperatures and illumination intensities. Temperature coefficients have been extracted for three-junction cell designs that are in production and under development. A simple diode model is applied to the three-junction performance as a means to predict performance under operating conditions outside the test range. These data may be useful in guiding the future optimization of concentrator solar cells and systems. Copyright © 2008 John Wiley & Sons, Ltd.

Dual-junction Ga 0.5 In 0.5 P/GaAs solar cells on Ge substrates have rapidly gone from small, high-efficiency laboratory cells, to large-area, high-efficiency cells manufactured at Spectrolab in high volume. Over 500,000 of these dual-junction (DJ) cells with 27-cm 2 area have been produced, with average AM0 load point efficiency of 21.4%. The next step in the evolution of this type of multijunction solar cell has been taken, with the development of triple-junction (TJ) Ga 0.5 In 0.5 P/GaAs/Ge cells. The addition of the germanium third junction, plus several significant improvements in the device structure, have led to a measured efficiency of 27.0% (AM0, 28°C) at Spectrolab on large-area (>30 cm 2 ) TJ cells. The TJ cell is now in production at Spectrolab. Ga 0.5 In 0.5 P/GaAs/Ge cells are viable not only for non-concentrating space applications, but also for terrestrial and space concentrator systems. Efficiencies up to 32.3% at 47 suns under the terrestrial AM1.5D spectrum have been achieved.

The effects of 1 MeV electron and 3 MeV proton irradiation on single-junction p + n InGaP solar cells are studied. The cell radiation response is monitored through measurements of current vs voltage (both in the dark and under simulated solar light (1 sun, AM0)), quantum efficiency, and capacitance vs voltage. Analysis of the data is used to determine the basic mechanisms for the cell response. High current injection causes the irradiated cells to partially recover due to a reduction in the junction dark current, which induces recovery in the cell open circuit voltage. The degradation data measured under electron and proton irradiation are correlated in terms of displacement damage dose, from which a characteristic degradation curve for this cell technology is determined.

An analysis of the radiation response of state-of-the-art InP/Si, InGaP, and dual junction (DJ) InGaP/GaAs space solar cells under both electron and proton irradiated is presented. The degradation data are modeled using the theory of displacement damage dose. For each technology, a characteristic curve which describes the cell degradation in any radiation environment is determined, and the characteristic curves are used to compare the radiation resistance of the different technologies on an absolute scale. The radiation data are used as input to a code which predicts the end-of-life (EOL) performance of a solar panel in earth orbit. The results show that in orbits outside the earth's radiation belts, the high-efficiency DJ InGaP/GaAs solar panels provide the highest EOL specific power. However, in orbits which pass through the belts, the radiation hard InP/Si panels provide the highest specific power by as much as 30%.

The effect of 1 MeV electron and 3 MeV proton irradiation on the performance of n+p InP solar cells grown heteroepitaxially on Si (InP/Si) substrates is presented. The radiation response of the cells was characterized by a comprehensive series of measurements of current versus voltage (I–V), capacitance versus voltage (C–V), quantum efficiency (QE), and deep level transient spectroscopy (DLTS). The degradation of the photovoltaic response of the cells, measured under simulated 1 sun, AM0 solar illumination, is analyzed in terms of displacement damage dose (Dd) which enables a characteristic degradation curve to be determined. This curve is used to accurately predict measured cell degradation under proton irradiation with energies from 4.5 down to 1 MeV. From the QE measurements, the base minority carrier diffusion length is determined as a function of particle fluence, and a diffusion length damage coefficient is calculated. From the C–V measurements, the radiation-induced carrier removal rate in the base...

High efficiency Inverted Metamorphic (IMM) multi-junction solar cells have been under development at Spectrolab for use in space and near space applications This paper reviews the present state-of-the-art of this technology at Spectrolab with an emphasis on performance characterization data at in-flight operating conditions. Large area IMM3J and IMM4J solar cells with 1X AM0 efficiency greater than 32% at 28 °C have been fabricated and characterized. Degradation factors after exposure to 1 MeV electron irradiation for both IMM3J and IMM4J technologies is presented. A coupon utilizing large area, IMM solar cells has been assembled and subjected to thermal cycling. Pre-and post thermal cycling data have been collected. Preliminary temperature cycling data indicate that a small coupon populated with strings of these cells suffered no degradation.

The annealing behavior of electron, proton, and alpha particle irradiated, epitaxial n+p InP solar cells has been characterized using several techniques. Current–voltage measurements were made under simulated 1 sun, AM0 solar illumination and in the dark. The radiation‐induced defect spectra were monitored using deep level transient spectroscopy and the base carrier concentration profiles were determined through capacitance–voltage measurements. The irradiated cells were annealed at temperatures ranging from 300 up to 500 K. Some cells were annealed while under illumination at short circuit while others were annealed in the dark. These experiments produced essentially the same results independent of illumination and independent of the irradiating particle. An annealing stage was observed between 400 and 500 K, in which the radiation‐induced defects labeled H3 and H4 were removed and the carrier concentration recovered slightly. Concurrently there was a small reduction in the junction recombination current...

The degradation of epitaxial, shallow homojunction n+p InP solar cells under 1 MeV electron and 3 MeV proton irradiation is presented. The data measured under 3 MeV proton irradiation are analyzed in terms of displacement damage dose which is the product of the particle fluence and the calculated non-ionizing energy loss (NIEL)[1]. A characteristic proton degradation curve is derived from which the cell degradation under any energy proton irradiation can be calculated. The data measured under 1 MeV electron irradiation is also analyzed in terms of displacement damage dose. The electron irradiation-induced degradation is correlated with the proton degradation curve by determining electron to proton dose ratios for each of the photovoltaic (PV) parameters. A comparison of the characteristic degradation curves for InP and GaAs/Ge solar cells, which was determined previously, shows InP to be intrinsically more resistant to displacement energy deposition. The base carrier concentration was measured during the irradiations, and significant carrier removal was observed. When analyzed as a function of displacement damage dose, the reduction in carrier concentration under both the 1 MeV electron and the 3 MeV proton irradiation is shown to follow the same degradation curve. From this common degradation curve, a characteristic carrier removal rate is calculated for InP under any irradiation. The junction dark current was also measured during both irradiations, and the data were fit to a three-term diode dark current equation. From the fits, the diffusion current is determined as a function of particle fluence. Changes in the diffusion current under electron and proton irradiation are shown to correlate in terms of displacement damage dose in the same way as the cell maximum power. The junction recombination current is also determined from the dark current data, and the results show the energy level of the dominant radiation-induced recombination center to be approximately the same in both the electron and proton irradiated samples. In addition, the dark current analysis indicates that the relative changes in the hole and electron lifetimes are essentially the same under both the electron and the proton irradiations. Based on these results and the overall correlation between the electron and proton damage, a detailed description of the mechanism of the radiation response of InP is developed which describes the cell degradation under any particle irradiation.

Results are presented for 1 MeV electron‐irradiated, two terminal, monolithic InP/Ga0.47In0.53As tandem solar cells. These highly efficient prototype cells show radiation resistance that is comparable to single junction InP cells. A current mismatch between the subcells does not occur until high fluence levels, that is, near 3×1015 e−/cm2. This value for the onset of current mismatch and the measured remaining absolute efficiency of 9.4% at 1×1016 e−/cm2 are excellent results reported for a tandem cell designed for space applications.

This paper proposes the use of a monolithically integrated back scattering layer to improve the efficiency of ultra-thin GaAs single-junction solar cells. The device consists of a double-heterostructure GaAs/InGaP single-junction solar cell integrated with a lattice-matched ZnSe layer etched as the back scattering layer coated with a highly reflective Au film and flip-chip bonded on a carrier substrate. ZnSe surface was roughened to form different textures by wet etching. Without anti-reflective coating, 300×300 µm2 devices with a 300 nm GaAs active region and a reflective ZnSe/Au back scattering layer, demonstrate a short-circuit current density of 17.2 mA/cm2, an open-circuit voltage of 0.915 V, and a power conversion efficiency of 11.8% under 1 sun solar radiation (AM 1.5G, 0.1 W/cm2). External quantum efficiency (EQE) measurements show EQE over 65% at 400 nm and ∼30% at 850 nm.

Information on the temperature of a packaged III–V multijunction solar cell mounted on a heat sink, operating under concentrated light is often not readily available. Availability of such information would facilitate the design of different receiver module configurations in a concentrating photovoltaic system (CPV). To this end, a heat transfer model is developed from finite difference techniques to predict the temperature from various parts of a concentrator cell assembly (CCA). The CCA consists of a solar cell mounted on a direct-bonded copper ceramic substrate with bypass diode. Temperatures of the solar cell with applied conformal coating are modeled as well as the temperature difference, ΔT, between the various layers within the CCA. Isotherm contour plots are generated for the cell under different conditions. It is found that the solar cell temperature in the CCA without conformal coating is 32 °C when illuminated at 50 W/cm2 with the CCA back surface temperature at 25 °C. When the CCA is bonded to a surface with thin bondline of a silicone-based thermal adhesive of 2 W/m K under the same intensity and back surface temperature, the cell rises to 37.3 °C. Further, the effects of the thermal adhesive thickness as well as the adhesive thermal conductivity on the solar cell temperature are examined. An effective thermal resistance of the CCA is determined to help in the design of a CPV system. The results from the model are validated against conservation of energy where the heat input from solar radiation on the solar cell is equal to the heat rate by conduction minus the converted electrical power of the cell.

Summary form only given. Power-by-light schemes to provide electrical power to remote opto-electronic modules typically involve the use of multi-segmented photovoltaic (PV) cells or commonly known as laser power converters (LPC) illuminated by an optical fiber to produce output voltages between 2 and 12 volts, with up to several hundred milliwatts of electrical power. Typically these photovoltaic devices are based on a GaAs single junction cell grown on semi-insulating GaAs substrates and they are further segmented and connected in series depending on the output voltage required. Theoretically, these cells should have efficiencies near 60% but currently these cells have efficiencies only in the range of 40 to 45%. Optical losses due to high grid metal coverage and cell isolation trench can reduce efficiency by nearly 10%. In order to eliminate the losses in the multi-segmented cells and to improve the efficiencies further we have designed a vertical multi junction device. The vertical multi-junction approach has several advantages over the multi- segmented approach. The advantages are; a) reduction in optical losses due to trench formation, b) lower metal obscuration losses, due to lower current density in the device, c) reduction of sheet conduction losses, and d) simple process to fabricate the devices. We have successfully designed and fabricated a new vertical multi-junction TV laser power converter with an efficiency of over 53%, a considerably higher efficiency than the regular multi-segmented 2V GaAs laser power converter and this is the highest efficiency reported so far for a photovoltaic device.

A major step forward has been made towards cost reduction of terrestrial PV. World-record multijunction III-V solar cells have been integrated into a commercial Concentrator Photovoltaic (CPV) system.. A dense array of high-efficiency solar cells in the receiver of a high-intensity (~500x) concentrator system has been identified as a viable, cost-effective system. Concentrator Ultra Triple Junction (CUTJ) cells have been developed for use in the Solar Systems CS500 solar electric power generator. The cell is designed for efficient conversion of the specific solar spectrum delivered to the system receiver while minimizing cell cost. Cells are optimized for maximum active area in a Solar Systems dense-array cell module. Solar Systems modules using CUTJ dense-array cells have demonstrated module efficiencies of over 35%. Field testing of CUTJ dense-array cells in a CS500 CPV dish unit at the Hermannsburg solar power plant in Australia was initiated in December 2005. A full multi-junction receiver in a CS500 dish has delivered over 30kW with an efficiency of almost 30%. Following qualification, these systems are slated for entry into the terrestrial market in 2006.

Metamorphic III-V semiconductor materials offer access to bandgaps that span key portions of the solar spectrum, enabling new bandgap combinations in multijunction solar cells, and increasing both theoretical and practical efficiency limits for terrestrial concentrator cells. Experimental results are given for the quantum efficiency of metamorphic GaInAs solar cells with bandgap from 1.1 to 1.4 eV, and for metamorphic GaInP with both ordered and disordered group-III sublattices. Variable intensity Jsc vs. Voc measurements are used to compare recombination components due to n = 1 and n = 2 mechanisms in metamorphic and lattice-matched GaInAs, GaInP, and 3-junction solar cells. A record efficiency metamorphic GaInP/ GaInAs/ Ge 3-junction solar cell has been produced with 38.8% efficiency independently confirmed (241 suns, AM1.5D, low-AOD, 25°C), essentially equaling the performance of a lattice-matched 3-junction cell with 39.0% efficiency, the highest efficiency yet demonstrated and verified for a solar photovoltaic conversion device. With the combination of high-quality metamorphic materials that are increasingly less controlled by recombination at dislocations, and the higher efficiency limits afforded by freedom of lattice constant selection, practical terrestrial concentrator cell efficiencies well over 40% are expected in the near future.

Results are reported for the temperature dependence of the IV parameters as a function of concentration level for triple junction (3J), GaInP/GaAs/Ge concentrator solar cells measured outdoors. Each IV parameter was measured in the range from 1 to 950 suns. The temperature coefficient of Voc, dVoc/dT, showed an abrupt decrease at low concentration but then decreased slowly at mid concentration for which a steady value was reached at higher concentration. In this latter case, the generation of minority carriers from high injection conditions overcomes the effects of the dark saturation current. The characteristics of the temperature coefficient of Jsc, dJsc/dT, however, are primarily dependent on high injection conditions coupled by the influence of the subcell that limits the current in the 3J cell stack. It is suggested the subcell that limits the device current is also influenced by chromatic aberration effects from the Fresnel lens. Values for dVoc/dT and dJsc/dT as a function of the concentration ratio are presented.

This paper discusses the approach taken by Concentrating Technologies (CT), Spectrolab, and Arizona Public Service (APS) to demonstrate a High Concentration Photovoltaic (HCPV) module using Spectrolab's GaInP/GaInAs/Ge triple-junction solar cells. This module is currently connected to an inverter, feeding electricity into the grid at the Solar Test & Research (STAR) facility of APS in Tempe, AZ. Although the module output is small, under 1 kW AC, it is the world's first demonstration of a grid-connected utility module using the same triple-junction solar cell technology that have been used to power satellites and other spacecrafts. This paper also discusses the next steps required to increase module efficiency and enhance its reliability. The economics of the MJ receivers in the context of this module are also discussed.

Results are reported for the effects of chromatic aberration on the electrical performance of triple junction (3J), GalnP/GaAs/Ge concentrator solar cells for indoor measurements. A Fresnel lens was moved away from the solar cell such that the concentration level increased, and then decreased while the lens-to-cell distance continued to increase. From such movement, the FF exhibited a series of minima and maxima in the region where light is apparently focused as determined by the naked eye. These effects are due to chromatic aberrations that affect the FF, and to some extent, the subcell short circuit currents in the 3J stack. Aberrations can drive the 3J device in a top cell-or middle cell-limited mode, and make the top cell FF become more predominant than the middle cell FF, or vice versa with the 3J FF influenced more by either subcell FF. It is shown that the multijunction celt power output is also affected, and the gain in power depends on the size of the solar cell relative to the size of the beam waist, and the focal length of blue light relative to that of red light.

The photovoltaic characterization of triple-junction InGaP2/GaAs/Ge solar cells is presented. Measurements made using a single light source solar simulator are compared with other measurements made using a multi-light source solar simulator that provides a close match to the air mass zero (AM0) solar spectrum. The output spectrum of the solar simulators has been measured, and two methods for calibrating the simulator output intensity haven been employed. The spectral response of the solar cells has been characterized through quantum efficiency measurements. These data are analyzed to determine the effect of the simulator spectrum on the measured photovoltaic response, and in particular, areas where spectral mismatch between the simulator and AM0 can lead to inaccurate performance predictions are highlighted. In addition, the effects of the different calibration techniques on the measured data are studied. Exploiting the capabilities of the multi-source, close matched simulator, the response of each of the three sub-junctions are studied individually, and the interplay between the spectral response of the sub-junctions and the incident spectrum is investigated.

The use of the appropriate characterization tools along with the use of single cell assemblies that are robust, are necessary for proper evaluation of device performance and stability tests for GaInP/sub 2//GaAs/Ge triple junction (3J) concentrator solar cells. Certain discrepancies exist between the photovoltaic (PV) values from continuous outdoor illumination measurements at high concentration levels and those from the High Intensity Pulse Solar Simulator (HIPSS) measurements, which are addressed. The series resistance inferred from the current-voltage (IV) curve, however, is similar from both measurement techniques. Outdoor tests of concentrator cells yielded efficiencies near 29% (31% from HIPSS) for an average concentration of 432 suns. Both types of measurements revealed that the fill factor (FF) limited the performance of the cell at high concentrations due to the role of the series resistance. The formation of heat from the nonuniformity in the Fresnel optics also limited the device performance. Before the cell stability can be evaluated, cells are first screened for defects from the 1-sun, indoor, IV measurements. Then, the good cells are incorporated into small assemblies for further test by forward bias injection (FBI). The stability outcome of the FBI determines whether or not the cell can be used for outdoors tests. Outdoor IV measurements are then conducted. Stability test results showed that no degradation occurs in the 1-sun IV data before and after light soak of the concentrator assemblies.

Using the energy bandgap of semiconductors as a design parameter is critically important for achieving the highest efficiency multijunction solar cells. The bandgaps of lattice-matched semiconductors that are most convenient to use are rarely those which would result in the highest theoretical efficiency. For both the space and terrestrial solar spectra, the efficiency of 3-junction GaInP/GaAs/Ge solar cells can be increased by a lower bandgap middle cell, as for GaInAs middle cells, as well as by using higher bandgap top cell materials. Wide-bandgap and indirect-gap materials used in parasitically absorbing layers such as tunnel junctions help to increase transmission of light to the active cell layers beneath. Control of bandgap in such cell structures has been instrumental in achieving solar cell efficiencies of 29.7% under the AMO space spectrum (0.1353 W/cm/sup 2/, 28/spl deg/C) and 34% under the concentrated terrestrial spectrum (AM1.5G, 150-400 suns, 25/spl deg/C), the highest yet achieved for solar cells built on a single substrate.

This paper discusses a dense array receiver design that utilizes Spectrolab's high efficiency triple junction cells. The system involves the use of a reflective dish optics and dual tracking system, and at 300-Sun concentration, the peak power is projected to exceed 2 kW. The details of the cell and module design are discussed in this paper.

As part of a continuing program to determine the space radiation resistance of InP/In/sub 0.53/Ga/sub 0.47/As tandem solar cells on Ge substrates, n/p In/sub 0.53/Ga/sub 0.47/As solar cells fabricated by Research Triangle Institute (RTI) were irradiated with protons of various energy. The cells were grown with a 3 /spl mu/m n-InP window layer to mimic the top cell in the tandem cell configuration for both 1 sun, AMO solar absorption and radiation effects. The results have been plotted against proton fluence and "displacement damage dose" which is the product of the nonionizing energy loss (NIEL) and the particle fluence. A characteristic radiation damage curve is then obtained for predicting the effect of irradiation by any particle of any energy on these cells.

Radiation damage results are presented for two-terminal InP/Ga/sub 0.47/In/sub 0.53/As tandem solar cells irradiated with 1 MeV electrons and 3 MeV protons. The performance of the irradiated tandem cells are compared with each other which are, subsequently, described in terms of the component cells. The current mismatch appearing in the tandems are almost nonexistent compared to previous electron-irradiated cells. This is explained in terms of the thickness of the InP top subcell for a top cell-limited tandem. At high fluences, the proton-irradiated cell showed very good radiation resistance compared to a shallow homojunction InP cell. The increase in the recombination current appears to be the major degradation mechanism as revealed by the degradation of the open circuit voltage for the proton-irradiated cells. Annealing of the tandems showed partial recovery for the PV parameters in general but the short circuit current and the fill factor decrease at elevated temperatures for the cell irradiated with protons.

This paper presents the radiation results on p/n In/sub 0.53/Ga0/sub 0.47/As bottom cells. The main conclusion is that In/sub 0.53/Ga0/sub 0.47/As solar cells display the same radiation tolerance, regardless of the cell polarity. This result is, in general, not expected. Previous results involving damage coefficients for both Si, GaAs, and InP show that the polarity of the device does indeed matter. This paper therefore gives device fabricators the liberty of choosing either type of polarity depending on their needs. >

Two-terminal, monolithic InP/Ga/sub 0.47/In/sub 0.53/As tandem solar cells grown by the atmospheric-pressure metalorganic vapor phase epitaxy (APMOVPE) technique have been studied under successive fluences of 1 MeV electrons. Illuminated I-V results obtained at 1 sun, AM0 (25/spl deg/C) are described. Besides the high performance of these novel devices under particle irradiation, the current mismatch between the top and bottom subcells began to be noticeable at a fluence of 3/spl times/10/sup 15/ cm compared to 3/spl times/10/sup 14/ cm/sup -2/ as demonstrated in previous cells. This is a dramatic improvement in the radiation resistance. The cells exhibited beginning-of-life (BOL) efficiencies as high as 22%, and it is suggested that additional slight improvement in the radiation tolerance of the tandem cells can be achieved by further adjusting the device structure. >

InP/Si solar cells have been irradiated with 3 MeV protons to very large fluences where carrier removal, instead of decreases in minority carrier diffusion length, dominates the radiation response. In this regime, radiation-induced expansion of the base depletion width causes I/sub sc/ to increase above the pre-irradiation level before falling catastrophically to zero. Current-voltage (IV), capacitance-voltage (CV), quantum efficiency (QE), and electrochemical capacitance-voltage (ECV) profiling measurements are presented.

The radiation response of single and dual-junction p/sup +/n InGaP/GaAs solar cells is studied. The degradation mechanisms of single-junction InGaP cells are identified, and characteristic degradation curves in terms of displacement damage dose are calculated. The response of dual-junction cells is presented, and the response of each sub-cell is discussed. The cell response is compared with those of other technologies. The effect of current injection on irradiated InGaP cell is presented.

We identify a failure mode due to a photoactive back contact for Ge concentrator solar cells. This problem manifests itself as a leveling off and subsequent decrease of open-circuit voltage (V/sub oc/) as the concentration increases above /spl sim/20 suns. Correction of this problem yields a much improved Ge cell for which V/sub oc/ increases in an almost ideal n=1 manner from 0.2 volts at one sun to 0.4 volts at 1400 suns. This cell's fill factor remains at or above its one-sun value up to 500 suns, confirming that this cell is fully suitable for high-concentration use. We show that solving the back-contact problem can significantly improve the high-concentration performance of GaInP/GaAs/Ge three-junction solar cells.

This paper presents key results of an on-going program to develop terrestrial concentrator modules incorporating high-efficiency multi-junction (MJ) photovoltaic (PV) cell technology. This program evolved from a successful space concentrator module development. Indeed, space mini-concentrators, comprising a stretched-membrane line-focus Fresnel lens and a prism-covered triple-junction solar cell, have been found to perform exceptionally well in the terrestrial environment. In outdoor tests at both ENTECH and NREL, space mini-concentrators have achieved overall module (combined lens and cell) operational solar-to-electric conversion efficiency levels from 25 to 29% under a variety of conditions. The long-term goal of the present program is to incorporate MJ cell technology into existing field-proven modules and two-axis sun-tracking arrays. With twice the efficiency of present silicon-based concentrators, these next-generation MJ concentrators should offer unprecedented economics.

Ga/sub 0.5/In/sub 0.5/P/GaAs/Ge solar cells have been fabricated at Spectrolab under the Multijunction Solar Cell Manufacturing Technology (Mantech) program, sponsored by the US Air Force and NASA. The cells were irradiated with increasing 1 MeV electron fluences, and the degradation of their PV parameters was characterized using light I-V and external QE. Analysis of the PV parameters of the GaInP top subcell showed little degradation, and was not a limitation for triple junction (3J) cell performance. Furthermore, the radiation degradation of the Ge subcell PV parameters was almost negligible. The GaAs subcell I/sub sc/, however, did limit the device performance as is traditionally documented. The final-to-initial maximum power ratio (P/P/sub 0/) of 3J cells was near 0.833 at a fluence of 1/spl times/10/sup 15/ e/sup -//cm/sup 2/, and matches Spectrolab's presently established value of 0.83 for standard production of space-qualified dual-junction cells.

Results from Spectrolab-grown Ga/sub 0.5/In/sub 0.5/P/GaAs/Ge structures optimized for the AM1.5D spectrum are described along with progress toward developing next generation multijunction solar cells for high concentration ratios (X). The epitaxially-grown layers were processed into triple junction cells both at Spectrolab and NREL, and I-V tested vs. X. Cells were tested with efficiencies as high as 32.4% near 372 suns. The FF limited the performance with increasing X as a result of the increased role of the series resistance. The V/sub oc/ vs. X showed its log-linear dependence on I/sub sc/ over 1000 suns. Based on cell improvements for space applications, multijunction cells appear to be ideal candidates for high efficiency, cost effective, PV concentrator systems. Future development of new 1 eV materials for space cells, and further reduction in Ge wafer costs, promises to achieve cells with efficiencies >40% that cost $0.3/W or less at concentration levels between 300 to 500 suns.

Next-generation solar cell approaches such as AlGaInP/GaAs/GaInNAs/Ge 4-junction cells, lattice-mismatched GaInP/GaInAs/Ge, concentrator cells, and improved 3-junction device structures hold the promise of greater efficiency than even today's highly successful multijunction cells. Wide-bandgap tunnel junctions, improved heterointerfaces, and other device structure improvements have resulted in several record-efficiency GaInP/GaAs/Ge cell results. Triple-junction (3J) cells grown in this work have demonstrated 29.3% efficiency for space (AMO, 1 sun). Space concentrator 3J cells have efficiency up to 30.0% at low concentration (AMO, 7.6 suns), and terrestrial concentrator cells grown at Spectrolab and processed at NREL have reached 32.3% (AM1.5D, 440 suns).

This paper describes progress toward achieving high efficiency, multijunction solar cells for cost effective application in terrestrial PV concentrator systems. Small area triple junction GaInP/sub 2//GaAs/Ge solar cells have been fabricated with an efficiency of over 32% when measured by NREL under an AM1.5D spectrum at 47 suns concentration. Small changes to the device design can achieve similar efficiencies at concentration ratios of about 500 suns, resulting in cell costs of $0.5 to $0.6/W today, at production volumes of approximately 50 MW/year. This makes them highly cost effective in existing concentrator systems, compared to flat plate technologies. The future development of new 1 eV materials for space cells, in conjunction with further reduction in Ge wafer costs, promises to achieve solar cells of >40% efficiency that cost $0.4/W or less at these concentration ratios.

Both n/p and p/n InP thick emitter (0.3 /spl mu/m) space solar cells with and without Ga/sub 0.5/In/sub 0.5/P windows are studied. Both polarity cells are considered for possible growth on Ge. While not achieving the high efficiencies of thin emitter InP cells, the thicker emitter cells may provide an advantage in radiation hardness. The Ga/sub 0.5/In/sub 0.5/P window has little effect on cell performance, for either polarity.

The Naval Research Laboratory has presented numerous reports on the radiation hardening of the two-terminal InP/Ga/sub 0.47/In/sub 0.53/As tandem solar cells for space use. To date, most work has been done on n/sup +/p devices. However, in order to facilitate the heteroepitaxial growth of these tandems on substrates such as Si or Ge, the polarity of the subcells may need to be reversed. With this in mind, spectral response measurements on irradiated homoepitaxial p/sup +/n InP solar cells were carried out for several fluences of 1 MeV electrons. The quantum efficiency data degraded at high wavelengths but no significant degradation was observed at low wavelengths. An important material parameter in solar cell devices is the minority carrier diffusion length (L) for which L/sub P/ was measured in the present study. This parameter was measured as a function of electron irradiation using the steady-state short-circuit current method. The degradation of L/sub P/ was plotted as a function of electron fluence and the data were fit to the standard semi-empirical model which relates /spl Delta/L to the damage constant, K/sub L/. The fit yielded values for K/sub L/ and the unirradiated minority carrier diffusion length, L/sub 0/.