Features

Atwater Team Awarded $2.4 million to Pursue Light-Splitting Approach to Building Ultraefficient Solar Modules

Applied physicists at the California Institute of Technology (Caltech) have suggested a radical change in the way solar cells collect sunlight and convert it to electricity, which could lead to photovoltaic panels with ultrahigh efficiency—above 50 percent. Now the team has received a $2.4 million grant from the Advanced Research Projects Agency–Energy (ARPA-E) of the Department of Energy to develop the sunlight-harnessing technology.

"The goal of our project is to break sunlight, which is inherently white light, up into its spectral color components, and collect each color in its own separate solar cell, thereby making use of more of the solar spectrum," says Harry Atwater, director of the Resnick Sustainability Institute at Caltech and principal investigator on the new Full-Spectrum Photovoltaics Project. "If we can make these modules at the same cost per area as current solar modules, it will absolutely be the future of solar power. You wouldn't want to make solar modules any other way."

The Energy Department's ARPA-E selected the Caltech project as one of 10 funded in the renewable-energy category of its OPEN 2012 program. According to the Department of Energy's press release
"ARPA-E seeks out transformational, breakthrough technologies that show fundamental technical promise but are too early for private-sector investment."

Atwater, who is also the Howard Hughes Professor and professor of applied physics and materials science at Caltech, and Albert Polman, director of the FOM–Institute for Atomic and Molecular Physics (AMOLF) in Amsterdam, the Netherlands, outlined the full-spectrum approach in a commentary article in Nature Materials earlier this year.

"Rather than making solar modules by wiring up identical solar cells that are each trying to absorb the whole spectrum of visible light, our module will actually be a relatively complex optoelectronic integrated circuit where each cell is highly optimized for absorbing light of a particular wavelength and converting it into electricity," Atwater explains.

The approach aims to boost efficiency by addressing two major sources of energy loss in current solar cells—lack of light absorption below a certain energy level, and heat production.

Typical solar modules are made up of wafers of a single semiconducting material, such as silicon, and convert about 15 percent of the energy that comes in as sunlight into usable electricity. Each semiconductor absorbs and converts light most efficiently for only one particular color or wavelength of light. As in the story of Goldilocks, the energy of that particular wavelength has to be just right. The semiconductor is transparent to longer wavelengths (lower energies), so it will fail to absorb light at those longer wavelengths.

At the same time, owing to a process called thermalization, not all of the light at shorter wavelengths (higher energies) is converted into usable electricity. Instead, a large portion of the sunlight energy that enters a solar cell with higher than the ideal energy is lost as heat.

"If the sun were a laser, and you could choose its optimal energy to match with the semiconductor you're using for your solar cells, then the efficiency could be extremely high—more than 70 percent," Atwater says.

Since the sun is not a laser, the researchers plan to incorporate optical components—such as filters and light splitting elements—into their photovoltaic components to split sunlight into its different spectral colors. Individual cells made from different semiconducting materials can then absorb the spectral colors that they best match in terms of energy.

Atwater notes that other teams have developed "multijunction" solar cells that use more than one type of semiconductor to absorb greater percentages of incoming light. But most of these designs involve simple stacked layers of semiconductors grown on one chip, and are usually limited to four layers. The Caltech group plans to include anywhere from six to 12 different semiconductor cells in its new solar modules.

The researchers are exploring multiple design options for how to best split and focus the light. But it is clear that the new components will be relatively small. Whereas today's solar cells are typically about 100 cm2 in size, each of the cells in this new approach will be about the size of an LED—between a few square millimeters and just a single square centimeter. They will then be assembled on a circuit board in the same way that different semiconductor components such as transistors are interconnected on circuit boards in smartphones and tablets.

"In fact, the front of your TV or computer display is a huge optical integrated circuit," Atwater says. "This is proof that we can make integrated circuits at the scale of photovoltaic modules. So why not actually make them?"

The manufacturing infrastructure already exists to assemble large-scale interconnected electronic devices at high speed and low cost. Beyond that, the company cofounded by Atwater, Alta Devices, has devised a way to make thin-film solar cells from highly efficient semiconductors, such as gallium arsenide, in a way that tremendously reduces the cost of making cells from such materials.

"Now that we've addressed the problem of how to make inexpensive compound semiconductor photovoltaics, I think this light-splitting approach is the logical next step," says Atwater. "I think we're at a great moment where we have emerging solar cell and optics technologies that can yield ultrahigh efficiency and also drive down the cost of the whole module."

The Caltech team currently consists of Atwater and five graduate students and postdoctoral scholars who are already starting to build components and design the optics for the new modules. The ARPA-E project is funded for three years.

This year, through the OPEN 2012 program, ARPA-E awarded funding to 66 research projects in 11 technology areas, totaling $130 million.
 

Written by Kimm Fesenmaier

.