Right now, most of the solar panels on the market are only able to produce electricity from a very small bandwidth of the solar spectrum. Some high-efficiency photovoltaic cells can convert multiple parts of the visible spectrum into sunlight, but they’re expensive. Researchers at the Lawrence Berkeley National Laboratory have demonstrated devices that can convert virtually all of the sun’s light into electricity, using methods already used by the solar industry.

The research is being conducted by the Berkeley Laboratory’s Solar Energy Materials Research Group, which is led by Dr. Wladek Walukiewicz. The group published its research, “Engineering the Electronic Band Structure for Multiband Solar Cells,” in the Jan.11 issue of Physical Review Letters.

It’s not the first multiband photovoltaic breakthrough for Walukiewicz and Yu.

In 2002, they found that by adjusting the amounts of indium and gallium in an alloy, they were able to produce semiconductors that respond to different spectrums of light. Though the resulting device responded to the full spectrum of light, the manufacturing process was too complex.

Now, Walukiewicz and his team of researchers have produced a device capable of responding to most of the light spectrum using technologies already used by the photovoltaic industry. The device is a multi-band semiconductor made of gallium arsenide nitride. By adding nitrogen to gallium arsenide, the scientists were able to develop the intermediate energy band that, along with two other layers of semiconductors, is able to respond to most of the solar spectrum.

Ultimately, the research could produce photovoltaic cells that are close to 50 percent efficient at converting sunlight into electricity, Walukiewicz said. He said the cells would be ideal in concentrated photovoltaics.

“So the amount of material used could be relatively small,” he said.

Whereas it took multi-junction technology 20 years to get over 40 percent efficient, this process could reach the commercial phase much quicker. Walukiewicz said the technology needs to reach 20 percent to 25 percent efficiency to become commercially viable.

“That could be done within 2 to 3 years,” he said. “It depends to a large extent on how much money is put into it.”

Pictured: Lawrence Berkeley Laboratory Doctors Walukiewicz and Yu, courtesy of the Berkeley Laboratory.