Search

This paper presents a method for using the SPENVIS on-line computational suite to implement the displacement damage dose (D(sub d)) methodology for calculating end-of-life (EOL) solar cell performance for a specific space mission. This paper builds on our previous work that has validated the D(sub d) methodology against both measured space data [1,2] and calculations performed using the equivalent fluence methodology developed by NASA JPL [3]. For several years, the space solar community has considered general implementation of the D(sub d) method, but no computer program exists to enable this implementation. In a collaborative effort, NRL, NASA and OAI have produced the Solar Array Verification and Analysis Tool (SAVANT) under NASA funding, but this program has not progressed beyond the beta-stage [4]. The SPENVIS suite with the Multi Layered Shielding Simulation Software (MULASSIS) contains all of the necessary components to implement the Dd methodology in a format complementary to that of SAVANT [5]. NRL is currently working with ESA and BIRA to include the Dd method of solar cell EOL calculations as an integral part of SPENVIS. This paper describes how this can be accomplished.

An analysis of the effects of low energy proton irradiation on the electrical performance of triple junction (3J) InGaP2/GaAs/Ge solar cells is presented. The Monte Carlo ion transport code (SRIM) is used to simulate the damage profile induced in a 3J solar cell under the conditions of typical ground testing and that of the space environment. The results are used to present a quantitative analysis of the defect, and hence damage, distribution induced in the cell active region by the different radiation conditions. The modelling results show that, in the space environment, the solar cell will experience a uniform damage distribution through the active region of the cell. Through an application of the displacement damage dose analysis methodology, the implications of this result on mission performance predictions are investigated.

The Naval Research Laboratory (NRL) has been involved in the development of solar cells for space applications since the 1960s. It quickly became apparent in this work that radiation damage caused to solar cells by electrons and protons trapped by the earth's magnetic field would seriously degrade the power output of photovoltaic arrays in extended missions. Techniques were therefore developed to harden the cells by shielding them with coverglass, etc. Ultimately, however, there is a limit to such approaches, which is determined by the radiation response of the semiconductor material employed. A desire for high efficiency and radiation resistance led to the development of alternative cell technologies such as GaAs, which has since become the technology of choice for many applications. InP cells are currently the most radiation resistant, high efficiency, planar cells known. NRL first sponsored InP solar cell technology in 1986, when Arizona State University was contracted to grow p/n cells by liquid phase epitaxy. NRL's interest in InP cells was generated by the results presented by Yamaguchi and his co-workers in the early 1980s on the remarkable radiation resistance of cells grown by diffusion of S into Zn doped p-type InP substrates. These cells also had beginning of life (BOL) efficiencies approximately 16%(AM0). Related to the radiation resistance of the cells was the fact that radiation-induced damage could be optically annealed by sunlight. Relatively large quantities of 1 x 2 cm(exp 2) diffused junction cells were made and were used on the MUSES-A and the EXOS-D satellites. These cells were also available in the U.S. through NIMCO, and were studied at NRL and elsewhere. Workers at NASA Lewis became involved in research in InP cells about the same time as NRL.

The measured degradation of epitaxial shallow homojunction n(+)/p InP solar cells under 1 MeV electron irradiation is correlated with that measured under 3 MeV proton irradiation based on 'displacement damage dose'. The measured data is analyzed as a function of displacement damage dose from which an electron to proton dose equivalency ratio is determined which enables the electron and proton degradation data to be described by a single degradation curve. It is discussed how this single curve can be used to predict the cell degradation under irradiation by any particle energy. The degradation curve is used to compare the radiation response of InP and GaAs/Ge cells on an absolute damage energy scale. The comparison shows InP to be inherently more resistant to displacement damage deposition than the GaAs/Ge.

An analysis embodied in a PC computer program is presented, which quantitatively demonstrates how the availability of radiation hard solar cells can help minimize the cost of a global satellite communications system. An important distinction between the currently proposed systems, such as Iridium, Odyssey and Ellipsat, is the number of satellites employed and their operating altitudes. Analysis of the major costs associated with implementing these systems shows that operation at orbital altitudes within the earth's radiation belts (10(exp 3) to 10(exp 4)km) can reduce the total cost of a system by several hundred percent, so long as radiation hard components including solar cells can be used. A detailed evaluation of the predicted performance of photovoltaic arrays using several different planar solar cell technologies is given, including commercially available Si and GaAs/Ge, and InP/Si which is currently under development. Several examples of applying the program are given, which show that the end of life (EOL) power density of different technologies can vary by a factor of ten for certain missions. Therefore, although a relatively radiation-soft technology can usually provide the required EOL power by simply increasing the size of the array, the impact upon the total system budget could be unacceptable, due to increased launch and hardware costs. In aggregate, these factors can account for more than a 10% increase in the total system cost. Since the estimated total costs of proposed global-coverage systems range from $1B to $9B, the availability of radiation-hard solar cells could make a decisive difference in the selection of a particular constellation architecture.

Indium phosphide (InP) solar cells were made on silicon (Si) wafers (InP/Si) by to take advantage of both the radiation-hardness properties of the InP solar cell and the light weight and low cost of Si wafers. The InP/Si cell application is for long duration and/or high radiation orbit space missions. Spire has made N/P InP/Si cells of sizes up to 2 cm by 4 cm with beginning-of-life (BOL) AM0 efficiencies over 13% (one-sun, 28C). These InP/Si cells have higher absolute efficiency and power density after a high radiation dose than gallium arsenide (GaAs) or silicon (Si) solar cells after a fluence of about 2e15 1 MeV electrons/sq. cm. In this work, we investigate the minority carrier (electron) base diffusion lengths in the N/P InP/Si cells. A quantum efficiency model was constructed for a 12% BOL AM0 N/P InP/Si cell which agreed well with the absolutely measured quantum efficiency and the sun-simulator measured AM0 photocurrent (30.1 mA/sq. cm). This model was then used to generate a table of AM0 photocurrents for a range of base diffusion lengths. AM0 photocurrents were then measured for irradiations up to 7.7e16 1 MeV electrons/sq. cm (the 12% BOL cell was 8% after the final irradiation). By comparing the measured photocurrents with the predicted photocurrents, base diffusion lengths were assigned at each fluence level. A damage coefficient K of 4e-8 and a starting (unirradiated) base electron diffusion length of 0.8 microns fits the data well. The quantum efficiency was measured again at the end of the experiment to verify that the photocurrent predicted by the model (25.5 mA/sq. cm) agreed with the simulator-measured photocurrent after irradiation (25.7 mA/sq. cm).

Indium phosphide (InP) solar cells are being made on silicon (Si) wafers (InP/Si) to take advantage of both the radiation-hardness properties of the InP solar cell and the light weight and low cost of Si wafers compared to InP or germanium (Ge) wafers. The InP/Si cell application is for long duration and/or high radiation orbit space missions. InP/Si cells have higher absolute efficiency after a high radiation dose than gallium arsenide (GaAs) or silicon (Si) solar cells. In this work, base electron diffusion lengths in the N/P cell are extracted from measured AM0 short-circuit photocurrent at various irradiation levels out to an equivalent 1 MeV fluence of 1017 1 MeV electrons/sq cm for a 1 sq cm 12% BOL InP/Si cell. These values are then checked for consistency by comparing measured Voc data with a theoretical Voc model that includes a dark current term that depends on the extracted diffusion lengths.

When determining the best solar cell technology for a particular space flight mission, accurate prediction of solar cell performance in a space radiation environment is essential. The current methodology used to make such predictions requires extensive experimental data measured under both electron and proton irradiation. Due to the rising cost of accelerators and irradiation facilities, such extensive data sets are expensive to obtain. Moreover, with the rapid development of novel cell designs, the necessary data are often not available. Therefore, a method for predicting cell degradation based on limited data is needed. Such a method has been developed at the Naval Research Laboratory based on damage correlation using 'displacement damage dose' which is the product of the non-ionizing energy loss (NIEL) and the particle fluence. Displacement damage dose is a direct analog of the ionization dose used to correlate the effects of ionizing radiations. In this method, the performance of a solar cell in a complex radiation environment can be predicted from data on a single proton energy and two electron energies, or one proton energy, one electron energy, and Co(exp 60) gammas. This method has been used to accurately predict the extensive data set measured by Anspaugh on GaAs/Ge solar cells under a wide range of electron and proton energies. In this paper, the method is applied to InP solar cells using data measured under 1 MeV electron and 3 MeV proton irradiations, and the calculations are shown to agree well with the measured data. In addition to providing accurate damage predictions, this method also provides a basis for quantitative comparisons of the performance of different cell technologies. The performance of the present InP cells is compared to that published for GaAs/Ge cells. The results show InP to be inherently more resistant to displacement energy deposition than GaAs/Ge.

An analysis embodied in a PC computer program is presented which quantitatively demonstrates how the availability of radiation hard solar cells can minimize the cost of a global satellite communication system. The chief distinction between the currently proposed systems, such as Iridium Odyssey and Ellipsat, is the number of satellites employed and their operating altitudes. Analysis of the major costs associated with implementing these systems shows that operation within the earth's radiation belts can reduce the total system cost by as much as a factor of two, so long as radiation hard components including solar cells, can be used. A detailed evaluation of several types of planar solar cells is given, including commercially available Si and GaAs/Ge cells, and InP/Si cells which are under development. The computer program calculates the end of life (EOL) power density of solar arrays taking into account the cell geometry, coverglass thickness, support frame, electrical interconnects, etc. The EOL power density can be determined for any altitude from low earth orbit (LEO) to geosynchronous (GEO) and for equatorial to polar planes of inclination. The mission duration can be varied over the entire range planned for the proposed satellite systems. An algorithm is included in the program for determining the degradation of cell efficiency for different cell technologies due to proton and electron irradiation. The program can be used to determine the optimum configuration for any cell technology for a particular orbit and for a specified mission life. Several examples of applying the program are presented, in which it is shown that the EOL power density of different technologies can vary by an order of magnitude for certain missions. Therefore, although a relatively radiation soft technology can be made to provide the required EOL power by simply increasing the size of the array, the impact on the total system budget could be unacceptable, due to increased launch and hardware costs. In aggregate these factors can account for more than a 10% increase in the total system cost. Since the estimated total costs of proposed global coverage systems range from $1 Billion to $9 Billion, the availability of radiation hard solar cells could make a decisive difference in the selection of a particular constellation architecture.

The Naval Research Laboratory has been involved in developing InP solar cell technology since 1988. The purpose of these programs was to produce advanced cells for use in very high radiation environments, either as a result of operating satellites in the Van Allen belts or for very long duration missions in other orbits. Richard Statler was technical representative on the first program, with Spire Corporation as the contractor, which eventually produced several hundred, high efficiency 2 x 2 sq cm single crystal InP cells. The shallow homojunction technology which was developed in this program enabled cells to be made with AMO, one sun efficiencies greater than 19%. Many of these cells have been flown on space experiments, including PASP Plus, which have confirmed the high radiation resistance of InP cells. NRL has also published widely on the radiation response of these cells and also on radiation-induced defect levels detected by DLTS, especially the work of Rob Walters and Scott Messenger. In 1990 NRL began another Navy-sponsored program with Tim Coutts and Mark Wanlass at the National Renewable Energy Laboratory (NREL), to develop a one sun, two terminal space version of the InP-InGaAs tandem junction cell being investigated at NREL for terrestrial applications. These cells were grown on InP substrates. Several cells with AM0, one sun efficiencies greater than 22% were produced. Two 2 x 2 sq cm cells were incorporated on the STRV lA/B solar cell experiment. These were the only two junction, tandem cells on the STRV experiment. The high cost and relative brittleness of InP wafers meant that if InP cell technology were to become a viable space power source, the superior radiation resistance of InP would have to be combined with a cheaper and more robust substrate. The main technical challenge was to overcome the effect of the dislocations produced by the lattice mismatch at the interface of the two materials. Over the last few years, NRL and Steve Wojtczuk at Spire have been developing a single junction InP on Si cell, in an ONR-sponsored SBIR program. Both cell polarities were investigated and the best efficiencies to date (approximately 13% on a 2 x 4 sq cm cell) were achieved with n/p cells. Earlier this year NRL began a program with ASEC to develop a two terminal InP-InGaAs tandem cell on a Ge substrate. RTI and NREL are subcontractors on this program. The results of an ONR-sponsored study of the potential market for InP/Si cells will be discussed. Also the technical status of both the InP/Si and the InP-InGaAs/Ge programs will be given. The technical challenges still remaining will be briefly described.

As part of a continuing program to determine the space radiation resistance of InP/ln(0.53)Ga(0.47)As tandem solar cells, n/p In(0.53)Ga(0. 47)As solar cells fabricated by RTI were irradiated with 1 MeV electrons and with 3 MeV protons. The cells were grown with a 3 micron n-lnP window layer to mimic the top cell in the tandem cell configuration for both AMO solar absorption and radiation effects. The results have been plotted against 'displacement damage dose' which is the product of the nonionizing energy loss (NIEL) and the particle fluence. A characteristic radiation damage curve can then be obtained for predicting the effect of all particles and energies. AMO, 1 sun solar illumination IV measurements were performed on the irradiated InGaAs solar cells and a characteristic radiation degradation curve was obtained using the solar cell conversion efficiency as the model parameter. Also presented are data comparing the radiation response of both n/p and p/n (fabricated by NREL) InGaAs solar cells as a function of base doping concentration. For the solar cell efficiency, the radiation degradation was found to be independent of the sample polarity for the same base doping concentration.

It has been found useful in the past to use the concept of 'equivalent fluence' to compare the radiation response of different solar cell technologies. Results are usually given in terms of an equivalent 1 MeV electron or an equivalent 10 MeV proton fluence. To specify cell response in a complex space-radiation environment in terms of an equivalent fluence, it is necessary to measure damage coefficients for a number of representative electron and proton energies. However, at the last Photovoltaic Specialist Conference we showed that nonionizing energy loss (NIEL) could be used to correlate damage coefficients for protons, using measurements for GaAs as an example. This correlation means that damage coefficients for all proton energies except near threshold can be predicted from a measurement made at one particular energy. NIEL is the exact equivalent for displacement damage of linear energy transfer (LET) for ionization energy loss. The use of NIEL in this way leads naturally to the concept of 10 MeV equivalent proton fluence. The situation for electron damage is more complex, however. It is shown that the concept of 'displacement damage dose' gives a more general way of unifying damage coefficients. It follows that 1 MeV electron equivalent fluence is a special case of a more general quantity for unifying electron damage coefficients which we call the 'effective 1 MeV electron equivalent dose'.

Radiation damage results from two-terminal monolithic InP/Ga(0.47)In(0.53)As tandem solar cells subject to 1 MeV electron irradiation are presented. Efficiencies greater than 22 percent have been measured by the National Renewable Energy Laboratory from 2x2 sq cm cells at 1 sun, AMO (25 C). The short circuit current density, open circuit voltage and fill factor are found to tolerate the same amount of radiation at low fluences. At high fluence levels, slight differences are observed. Decreasing the base amount of radiation at the Ga(0.47)In(0.53)As bottomcell improved the radiation resistance of J(sub sc) dramatically. This is turn, extended the series current flow through the subcell substantially up to a fluence of 3x10(exp 15) cm(exp -2) compared to 3x10(exp 14) cm(exp -2), as observed previously. The degradation of the maximum power output form tandem device is comparable to that from shallow homojunction (SHJ) InP solar cells, and the mechanism responsible for such degradation is explained in terms of the radiation response of the component cells. Annealing studies revealed that the recovery of the tandem cell response is dictated by the annealing characteristics exhibited by SHJ InP solar cells.

We report on the development and performance of deep-junction (approximately 0.25 micron), graded-emitter-doped, n(sup +)-p InP solar cells grown by metallorganic chemical vapor deposition (MOCVD). A novel, diffusion-transport process for obtaining lightly-doped p-type base regions of the solar cell is described. The I-V data and external quantum-efficiency response of these cells are presented. The best active-area AMO efficiency for these deep-junction cells on (100)-oriented InP substrates is 16.8 percent, with a J(sub SC) of 31.8 mA/sq cm, a V(sub OC) of 0.843 V, and a fill-factor of 0.85. By comparison, the best cell efficiency on the (111)B-oriented InP substrates was 15.0 percent. These efficiency values for deep-junction cells are encouraging and compare favorably with performance of thin-emitter (0.03 micron) epitaxial cells as well as that of deep-emitter diffused cells. The cell performance and breakdown voltage characteristics of a batch of 20 cells on each of the orientations are presented, indicating the superior breakdown voltage properties and other characteristics of InP cells on the (111)B orientation. Spectral response, dark I-V data, and photoluminescence (PL) measurements on the InP cells are presented with an analysis on the variation in J(sub SC) and V(sub OC) of the cells. It is observed, under open-circuit conditions, that lower-V(sub OC) cells exhibit higher band-edge PL intensity for both the (100) and (111)B orientations. This anomalous behavior suggests that radiative recombination in the heavily-doped n(sup +)-InP emitter may be detrimental to achieving higher V(sub OC) in n(sup +)-p InP solar cells.

The superior radiation resistance of InP over other solar cell materials such as Si or GaAs has prompted the development of InP cells for space applications. The early research on radiation effects in InP was performed by Yamaguchi and co-workers who showed that, in diffused p-InP junctions, radiation-induced defects were readily annealed both thermally and by injection, which was accompanied by significant cell recovery. More recent research efforts have been made using p-InP grown by metalorganic chemical vapor deposition (MOCVD). While similar deep level transient spectroscopy (DLTS) results were found for radiation induced defects in these cells and in diffused junctions, significant differences existed in the annealing characteristics. After injection annealing at room temperature, Yamaguchi noticed an almost complete recovery of the photovoltaic parameters, while the MOCVD samples showed only minimal annealing. In searching for an explanation of the different annealing behavior of diffused junctions and those grown by MOCVD, several possibilities have been considered. One possibility is the difference in the emitter structure. The diffused junctions have S-doped graded emitters with widths of approximately 0.3 micrometers, while the MOCVD emitters are often doped with Si and have widths of approximately 300A (0.03 micrometers). The difference in the emitter thickness can have important effects, e.g. a larger fraction of the total photocurrent is generated in the n-type material for thicker emitters. Therefore the properties of the n-InP material may explain the difference in the observed overall annealing behavior of the cells.

At the last SPRAT conference, the Naval Research Laboratory (NRL) presented results from two experiments. One studied n+p diffused junction (DJ) InP solar cells, and the other studied n+p shallow homojunction (SHJ) InP mesa diodes grown by metalorganic chemical vapor deposition (MOCVD). The former work showed that a DJ solar cell in which the maximum power P(sub max) had been degraded by nearly 80 percent under irradiation recovered completely under short circuit illumination at 450K. The recovery was accompanied by the removal of all but one of the radiation-induced defect levels. The latter work, on the other hand, showed that the radiation-induced defects in the SHJ diodes did not anneal until the temperature reached 650K. These results suggest that an irradiated DJ solar cell, under illumination, will anneal at a temperature 200K lower than an irradiated SHJ cell. This is an unexpected result considering the similarity of the devices. The goal of the present research is to explain this different behavior. This paper investigates two points which arose from the previous studies. The first point is that the DJ cells were annealed under illumination while the SHJ diodes were annealed without bias. The second point investigated here is that the emitters of the DJ and SHJ devices were significantly different.

The effects of temperature and radiation on n(+)p InP solar cells and mesa diodes grown by metallorganic chemical vapor deposition (MOCVD) were studied. It was shown that MOCVD is capable of consistently producing good quality InP solar cells with Eff greater than 19 percent which display excellent radiation resistance due to minority carrier injection and thermal annealing. It was also shown that universal predictions of InP device performance based on measurements of a small group of test samples can be expected to be quite accurate, and that the degradation of an InP device due to any incident particle spectrum should be predictable from a measurement following a single low energy proton irradiation.