Geoffrey A. Landis
Presented at SPS-97: Space and Electric Power for Humanity, 24-28 Aug., 1997, Montreal, Canada.
The status of space solar cell technology in 1997 is reviewed. High-efficiency dual- and triple- junction cells have continued to improve in performance, with an efficiency of 26.9% AM0 recently reported, and are approaching technology readiness. The first operational use of these cells in space should occur in late 1997. Thin film cells have also advanced in performance. New array technologies are being developed, and will be demonstrated in space in the next few years. An alternative solar conversion system, solar dynamic power, has recently been tested in the first end to end test under (simulated) space conditions.
Solar power satellites require improvements in five solar cell parameters :
Improvements have been made in each of these parameters; however, not all of these have been achieved in the same cell type. A series of review articles [2,3,4] have presented updates on the status of solar cell technology; this paper discusses some of the advances in technology made since the last of these reviews. For more details and references about technology status, the reader should consult reference . Solar cell efficiencies are also tabulated in the journal Progress in Photovoltaics ; the reader should be cautioned that efficiencies listed are measured under terrestrial (Air Mass 1.5) solar illumination conditions, and efficiencies under space (AM0) conditions are lower.
There are three approaches to large-area photovoltaic arrays in space. The conventional approach is to make flat-plate arrays from individual crystalline solar cells. The solar cell material used on spacecraft in the past is silicon (Si). Gallium arsenide (GaAs) solar cells, with improved efficiency, have now replaced silicon in many space applications. More efficient two and three junction cells, which use the solar spectrum more efficiently, are under development. An alternative cell material, indium phosphide (InP), is also under development. This cell has a considerably higher tolerance to radiation.
An alternative approach is to use mirrors or lenses as solar concentrators to focus light onto small, extremely high efficiency solar cells. Such an approach has yielded the highest conversion efficiencies achieved to date.
A third approach is to use thin-film, integrally-connected solar cells, adapting technology which has been developed for use in low-cost terrestrial solar arrays [6,7] Thin-film materials used include amorphous silicon (a-Si), cadmium telluride (CdTe), and copper indium diselenide or indium/gallium diselenide (CIS or CIGS). This approach has the potential for low weight and low cost, and has been demonstrated to have extremely high tolerance to radiation ; but is unlikely to achieve the high efficiency of single-crystal technologies. Thin-film technology has not yet been used for primary power in space, although individual solar cells have been tested in space, confirming the high tolerance to radiation. A large pay-off in terms of specific power (power per unit mass) could be possible with thin film cells if the technology to make these cells on lightweight, flexible substrates such as polyimide can be developed .
Since the previous review of cell technology , several advances have been made.
Conventional Si and GaAs solar cell technology has not made significant gains in efficiency. High efficiency Si cells are now beginning to be used for space missions, with 18% (AM0) silicon cells , for example, being used as primary power on the upcoming Lunar Prospector mission. Most new missions, however, now use GaAs/Ge solar cells for power.
Higher efficiency can be achieved by using solar cells with two or more junctions of different bandgap, to more efficiently divide the solar spectrum. These can either be manufactured as monolithic cascades, with both sub-cell junctions grown on the same substrate, or as mechanically-stacked cascades, where the sub-cell elements are manufactured separately and joined with adhesive. The technology for high-efficiency monolithic dual-junction solar cells using GaInP high-voltage cell on a GaAs substrate cell has advanced toward commercial maturity. The best cell efficiency has continued to improve slightly, with a recent cell achieving 26.9% under space (AM0) illumination conditions  (it should be noted that this result has not yet been confirmed with high-altitude testing).
Commercial manufacture of dual-junction cells is now is underway, with 22% efficiency achieved in production [12,13]. The NASA "Lewis" and "Clark" satellites, scheduled for launch this year, will have fully functional demonstration panels of cells of this technology.
Substantial advances have been made toward development of triple junction GaInP/GaAs/Ge cells, where the germanium bottom cell scavenges infrared photons that are otherwise lost to the GaAs cell. A joint Air Force/NASA "manufacturing technology" program is focused on bringing these cells to commercial readiness . The best efficiency has improved to 25.5% (AM0), with 24.3% efficiency achieved in large-area cells.
These multi-junction III-V cells are not likely to be manufactured at the low cost required for SPS use. One option is to use the cells in solar concentrators, where a lens or mirror is used to decrease the required cell area. Concentrator technologies have not yet been used for primary power application in space. The recent "PASP-Plus" satellite  successfully tested a concentrator system and verified that it operated as expected in the space environment. The first mission to use a concentrator array for primary power, the Deep-Space 1 mission, will launch next year. The concentrator array on this mission, SCARLET, has chosen GaInP/GaAs high efficiency dual junction cells, and is now being qualified for flight. Testing of the completed modules under space (AM0) conditions shows a measured efficiency of 19% , including the effect of the lens efficiency of 90%. A more advanced SCARLET-II array builds on the SCARLET technology with lower cost, easier fabrication, and simplified assembly and testing .
An alternative technology for low-cost solar arrays, with lower efficiency but considerably lower price, is the use of thin-film solar cells.
Advanced multi-junction amorphous silicon cells manufactured on flexible substrates  have been tested at NASA Lewis and achieved world record performance under space conditions. 12.0% AM0 efficiency was measured on small area multi-junction cells, and 8% AM0 measured on foot-square modules. These are being produced commercially on flexible metal substrates; a version that will be deposited on flexible polymer substrates is under development. In a separate experiment, a panel of flexible amorphous silicon cells will be demonstrated for power on the upcoming NASA "Clark" satellite.
The technology of other thin-film technologies continues to improve, but commercial-scale production is not yet available. If performance under terrestrial (AM1.5) sunlight is extrapolated to space (AM0) conditions, the best copper-indium/gallium selenide ("CIGS") thin-film cells would be expected to produce 14.2% efficiency; the best cadmium-telluride (CdTe) thin-film cells would produce about 12.9% efficiency. However, neither cell type is currently being made and tested under space conditions for space applications, although several corporations are now working on developing space versions. Neither one is made commercially on flexible substrates. ISET has manufactured CIS cells on flexible substrates, achieving an AM0 total-area efficiency of about 7.5% on a 1-cm by 1-cm cell .Table 1. Achieved efficiency of single-crystal solar cells under space (AM0) conditions
World solar cell manufacturing capability continues to increase toward 100 MW/year, with no slow-down in sight. Only a small fraction of a percent of this capability is used space applications; the vast majority of solar array applications are terrestrial applications for consumer electronics and remote power use.
Inflatable technologies for solar array deployment have been reaching maturity , and deployment of a large-area inflatable structure was demonstrated from the space shuttle. Other alternative low-mass technologies are also under the process of being developed, such as the AEC-Able "Ultra-Flex" array, baselined for the Mars-2001 mission. Another innovative lightweight array, which uses planar mirrors to increase the solar concentration by a low factor (roughly 1.5), has been designed by Astro Aerospace, and will fly on the NASA "Clark" satellite later this year .
An alternative approach to production of solar power in space is dynamic conversion of solar energy using a heat engine. This is known as solar dynamic (SD) conversion. A lens or mirror focuses sunlight onto a heat receiver, which provides heat to a thermal conversion unit. SD power systems can incorporate thermal storage as part of the system, instead of requiring batteries. A prototype full-scale Brayton-conversion SD module has recently been tested at NASA Lewis. This is the first end-to-end test of this type of power conversion system under simulated space conditions, and included all the system elements, including concentrating mirrors and the heat-rejection radiators. The test demonstrated a system conversion efficiency of 17%, for a specific power of 4.2 W/kg. Results are detailed in . General characteristics of one module are shown in Table 2 .
Solar array technology made incremental improvements in 1997. Dual and triple junction cells gained in efficiency and came closer to commercial production readiness. There were also gains in the efficiency of flexible amorphous silicon solar arrays. World solar array manufacturing continued to increase, and new technologies for light-weight, flexible solar arrays made improvements and were demonstrated in space. An alternate technology, solar dynamic power, was tested for the first time in an end-to-end system test under space conditions.
Table 2. 30-kW Brayton-Cycle Solar-Dynamic Power Module Characteristics
Average power to distribution system 30 kWe Mass of components Concentrator (mirror) 1573 kg Receiver 1752 kg Brayton Conversion Unit 790 kg Interface structure 602 kg Beta gimbal 190 kg Electrical equipment 290 kg Heat rejection 1356 kg Sun/shade time 55 min/26 min. Thermal power from receiver 96 kWth at 740 C Receiver efficiency 0.85 Power Conversion Unit efficiency 0.31 Pointing requirement of concentrator 0.1° Thermal load to radiator (at T from -9 to 177 C) 66 kWth Radiator area 15.3 m2