Alta Devices, a start-up in Santa Clara, Calif., presented research at the 37th IEEE Photovoltaic Specialist Conference, in Seattle, this week that claims its thin-film gallium-arsenide cell can convert 27.6 percent of the sunlight striking the cell into electricity, under standardized conditions. Since the paper was submitted, the company says it has upped the efficiency to 28.2 percent. That beats the previous record of 26.4 percent for a solar cell with a single p-n junction, which was the first improvement in years over 26.1 percent. Both numbers, according to Alta, were independently confirmed by the National Renewable Energy Laboratory.
The efficiency was measured on a laboratory-made solar cell. Efficiency tends to decrease once the cells are packaged into usable modules. “We assume we will ultimately be able to achieve modules that are around 26 percent, and that’s plenty to be competitive with fossil fuels,” says Christopher Norris, CEO of Alta.
The key to achieving the record was photon recycling. When the photons in sunlight are absorbed in a photovoltaic material, they split into an electron and a hole. The electrons that pass out of the cell can be used as electricity, but many of them are lost in the semiconductor when they recombine with a hole to produce either waste heat or a new photon. By carefully growing a high-quality single crystal of gallium arsenide, the company managed to ensure that more than 99 percent of the recombinations would result in new photons. Those photons could then create a new electron-hole pair and give the electron another chance to be captured as electricity. The Alta team also improved the reflectivity of the metal contacts on the back of the solar cell, so that any photons that exited the cell would be sent back in for possible reabsorption.
The theoretical maximum conversion efficiency for a solar cell with a single junction is 33.5 percent. “We can see a path to 30 percent with our same design right now,” says Norris. Adding a second junction could also increase the energy output.
The more efficient a solar cell is, the faster it pays back the cost of manufacturing and installing it. But efficiency and cost have been at odds with each other in solar cell design. Gallium arsenide is naturally better at converting light to electricity than the chief contenders, such as silicon and cadmium telluride, but it tends to be more expensive.
The most efficient materials are single-crystalline semiconductors, but those are usually pricier. Low-cost materials, such as amorphous silicon, cadmium telluride, and copper indium gallium selenide, are less efficient; CdTe cells are around 12 percent efficient. Alta solves this problem by using only a small amount of a high-quality material—a thin film of gallium arsenide about 1 micrometer thick.
“That is the whole trick. Don’t use much gallium and don’t use much arsenic,” Norris says. He says an Alta module should cost about the same per watt as a CdTe module but produce three times the energy.
The company cut down on the material cost by using a process called epitaxial liftoff, developed by Eli Yablonovitch, an engineering professor at the University of California, Berkeley, and a cofounder of Alta. Technicians start with a GaAs wafer as a seed layer and grow a thin-film photovoltaic device structure on top of that. They peel off the thin film, attach it to a metal backing, and finish processing it into a solar cell. The process leaves the original wafer, which they can reuse for the next batch of solar cells.
Alta is working on a pilot production line to produce samples of its solar cells sometime this year and expects to have early commercial shipments by late next year, Norris says. The company has raised US $72 million to develop its production process.