As NASA will tell you, there are no electrical outlets in space. Satellites rely on arrays of solar cells to generate the electricity needed to power their instruments and other onboard electronics. This aerospace technology is more efficient for generating electricity, more resilient to extreme temperatures — and much more expensive — than the solar cells used by utility companies or consumers to power applications on Earth.
The good news is that government researchers are intent on driving down the cost of this space-qualified aerospace technology to make it more available for terrestrial applications.
The III-V Standard
Solar cells are made from materials that generate a voltage and an electric current when exposed to sunlight. This phenomenon is called the photovoltaic effect. According to the Department of Energy, the photovoltaic effect was first observed in 1839 by French scientist Edmond Becquerel. Solar cells are rated according to their efficiency, i.e. their ability to absorb light and convert it into electricity. The higher the efficiency, the more electricity a solar cell can produce.
“State-of-the-art spacecraft solar arrays use an aerospace technology called triple junction III-V semiconductors,” said Tim Peshek, a research electrical engineer at NASA’s Glenn Research Center, Cleveland, Ohio. III-V materials are so named because they combine an element from group III of the periodic table with an element from group V. Typical III-V compounds include gallium arsenide (GaAs) and indium gallium phosphide (InGaP).
“Triple junction solar cells comprise three types of semiconductors grown on top of each other, typically germanium on the bottom, GaAs in the middle and InGaP on top,” he added. “Each layer absorbs light above a particular threshold energy of the solar spectrum, which spans wavelengths from mid-infrared through visible to near-ultraviolet. By using three complementary materials, we can create devices that absorb light across the entire solar spectrum, which makes them much more efficient than single-material solar cells.”
III-V vs. Silicon
According to Peshek, approximately 30 to 32% is the typical efficiency for III-V semiconductor solar panels. Solar industry data, by comparison, places the typical efficiency at about 18-20% for terrestrial silicon-based solar panels. And while silicon solar panels can be installed for about $2-3 per watt, III-V spacecraft solar panels will set you back $100 to $300 per watt. A solar panel comprises multiple solar cells.
“The standard process for growing III-V cells for space applications is called Metal Organic Chemical Vapor Deposition [MOCVD],” said Nancy Haegel, center director for materials science at the Department of Energy’s National Renewable Energy Laboratory (NREL) Golden, Colorado.
MOCVD is a high precision but time-consuming process. It relies on a chemical reaction between a substrate and special “precursor” molecules in a gas vapor to deposit high-quality, thin-film epitaxial layers on a substrate.
“III-V solar cells are grown carefully, layer by layer, typically on a substrate of GaAs or germanium,” said Haegel. “GaAs is an outstanding solar cell material, but it does not tolerate manufacturing defects well. Its component materials are also more expensive and less abundant than silicon.”
By contrast, she added, silicon solar cells used in utility and consumer applications are typically fabricated directly on silicon wafers cut from mass-produced ingots of pure silicon, the second most abundant element in the Earth’s crust.
Driving Down III-V Costs
NREL is working to reduce III-V production costs by increasing the rate at which solar cells are grown. To reach these higher growth rates, the agency is revamping an older solar cell manufacturing process called hydride vapor phase epitaxy (HVPE). Test results from this updated approach, known as “dynamic HVPE,” indicate that solar cell growth times could shrink dramatically, from the current several hours to just a few minutes.
“HVPE is really promising because it uses a starting material with a much lower cost than that used by MOCVD,” said Haegel. “It also has the potential for much higher growth rates.” To be fair, she added, MOCVD vendors are also trying to bring down their costs by using higher throughput and faster growth techniques.
NREL is also focused on re-using GaAs substrates to reduce III-V production costs. “We’d like to be able to grow a thin film on a GaAs substrate, lift the film off, then use the substrate again, potentially several more times, to grow more solar cells,” explained Haegel.
Building on Silicon
Government researchers are also looking for ways to capitalize on silicon’s maturity and economic scale factors. NREL, for example, is working on what it calls “next generation” solar cells.
“NREL’s silicon Group is working to improve the performance of silicon solar cells by optimizing the design and performance of the contact layers of the cell,” said Haegel.
And innovators at NASA Glenn have developed a technique to bond a multi-junction III-V solar cell to a silicon substrate. This approach, NASA claims, could couple the high efficiency of III-V semiconductors with the maturity and low cost of silicon.
Neither NREL nor NASA expects III-V solar cells to compete economically with terrestrial, high-volume silicon solar cells anytime soon.
A more realistic near-term target for the technology, cautions Haegel, would be intermediate markets such as transportation, aerospace or portable charging systems — markets where the benefit of mobility and remote power could entice people to pay perhaps $10-20 per watt.
“Our goal is to lower the cost of growing III-V solar cells by a factor of 100 through lower cost materials and higher growth rates,” Haegel said. “If we can reduce each of these elements by a factor of 10, we could start writing a very exciting new chapter in renewable energy.”
If you’d like to generate more power for your career, please check out the latest opportunities at careers.northropgrumman.com.