News

Concentrating dawn-to-dusk solar energy

MOSAIC award spurs MIT research into concentrator solar cells that can run in shade and full sun with power control and wavelength separation

Denis Paiste MIT News

Lighter, more efficient flat-plate solar cells are the goal of MIT researchers who kicked off a collaborative research effort Dec. 15 with a three-year, $3.5 million award under the Department of Energy’s ARPA-E program. Their aim is to bring the technology to the marketplace.

“We are early on looking for companies to collaborate with us who are interested in finding a way to bring that technology into the marketplace after the three-year project funding,” says principal investigator Jurgen Michel, senior research scientist at the MIT Microphotonics Center and senior lecturer in the Department of Materials Science and Engineering. “The best outcome is a solar cell and companies that will actually make those or take that into further development to make a product.”

ARPA-E’s Micro-Scale Optimized Solar-Cell Arrays with Integrated Concentration (MOSAIC) program has challenging specifications, Michel says. The goal is to reach overall efficiency of greater than 30 percent, which is about 5 percentage points higher than the best efficiency achieved with crystalline silicon solar cells.

Technical challenges

The MIT-led project, “Integrated Micro-Optical Concentrator Photovoltaics with Lateral Multijunction Cells,” aims to develop a three-junction concentrator cell in a flat-plate system just under 1 inch thick. It includes a partnership with Arizona State University. Besides Michel, collaborators include:

Juejun (JJ) Hu, the Merton C. Flemings Assistant Professor in Materials Science and Engineering, who will design and prototype a special microlens to split sunlight into wavelengths from visible to near infrared and concentrate sunlight up to 300 times;

Eugene A. Fitzgerald, the Merton C. Flemings-SMA Professor of Materials Science and Engineering, who will work on solar cells made from indium gallium arsenide;

David J. Perreault, professor of electrical engineering and associate department head, who will work on power management of the solar cells; and

Cun-Zheng Ning, professor of electrical engineering at Arizona State University, who will work on nanopillar semiconductor material with a bandgap gradient that is grown in a single step.

Ning developed a single-growth process for varying the bandgap in nanopillars by varying the temperature in the reactor. “That would be very low cost, but the challenge there is efficiency. For our approach, we have to get our substrate to low enough threading dislocation densities in order to get low-cost, high-efficiency solar cells,” Michel says.

Multiple benefits

The proposed solar system with a mix of cells to maximize collection of light a varying times of day addresses one of the key issues with solar energy, which is its intermittent nature. As more solar systems are deployed, they will have to be integrated with energy storage systems to achieve maximum benefit, according to The MIT Energy Initiative report, “The Future of Solar Energy,” released in May 2015. Without storage, solar systems can provide power only during the day.

Solar has enormous potential over the long-term. According the MITEI report, installing solar on less than one-half of 1 percent of the continental United States could produce all the electricity the country needs today. Solar also can reduce the nation’s carbon dioxide emissions, the report noted.

FOCUS results

Michel previously received a one-year ARPA-E FOCUS grant for research on Spectrum Splitting for High-Efficiency Photovoltaic and Solar Thermal Energy Generation. Two papers are pending publication on that work, including significant reductions in threading dislocations in solar cells and enhanced performance of indium gallium phosphide (InGaP) solar cells. These indium gallium phosphide materials are called III-V materials because their elements come from columns III and V of the periodic table.

“One of our main goals was to lower the cost of III-V semiconductor solar cells, so we’ve been using a silicon wafer with a germanium-on-silicon virtual substrate to grow our III-V cells on top of that,” Michel explains. “We’ve reduced the threading dislocation density to below mid-106 per square centimeter. Once you get down to about 106 per cm2 in threading dislocation density, you get actually high quality III-V semiconductor materials for high performance solar cells on a silicon substrate and that reduces the cost dramatically.” This technology is available for licensing through the MIT Technology Licensing Office.

In the new MOSAIC work, researchers will include the indium gallium phosphide (InGaP) solar cells based on the germanium-on-silicon approach in the FOCUS program and add solar cells made from gallium arsenide (GaAs) and indium gallium arsenide (InGaAs) to cover the whole spectrum of sunlight. These cells will be connected to each other in a parallel layout. The lens will direct specific wavelengths of light to matching solar cells.

The work builds on an earlier theoretical paper that showed that under realistic operating conditions over the course of a year, parallel cells coupled with wavelength separation, or spectrum splitting, outperformed a stacked array, or tandem, solar cell. “We found that if you split your spectrum in the way you spread it out onto separate solar cells, you have an overall gain in power output compared to the other solar cell,” Michel explains.

In the new project design, Michel says, “We can optimize the power point for each of the cells individually, because we have now CMOS control. That means we can respond very quickly to shading, for instance, [as] a cloud moves across a panel.” Under cloudy skies, light absorbed by silicon cells in the structure will maintain power output at about 20 percent power efficiency.

Despite the three-year prototype goal, it probably will take three to five years beyond that to bring to market a solar cell system that will last for 30 years. “If that can be done, then you’d actually have solar cells that would have a much higher output than current solar cells for thin plates, which makes it much easier to handle,” Michel explains. “Also if you have to, for instance, track your cells with the sun, weight is much lower, efficiency is high, and so that could be the next step in solar cell efficiency. We are not the only ones that are working on that. There are quite a few competitors. We just hope that one will be successful at least,” he says. “The best outcome is a solar cell and companies that will actually make those or take that into further development to make a product.”


Electric powerEnergy storageRenewable energy

Press inquiries: miteimedia@mit.edu

We're hiring! Learn more and apply