Traditional photovoltaic solar cells have an inherent limit on the efficiency at which they can convert sunlight into energy. This limit—based on the bandgap of the material used and known as the Shockley-Queisser limit—is about 33.7 percent for standard solar cells. It is essentially due to any material's inability to respond to all wavelengths of sunlight; so what if there was a way to change the wavelengths that actually reach the cell to those it converts best? MIT researchers have unveiled the best-yet version of that idea, known as solar thermophotovoltaics.
These modified solar cells place an absorber/emitter device above the cell itself. Sunlight is absorbed by this layer, it heats up—a lot—and emits light tuned directly to the bandgap of the PV cell beneath it. That means that much more of the energy in the sunlight can turn into electricity. According research in Nature Nanotechnology by graduate student Andrej Lenert and colleagues, this idea offers the benefits associated with both solar thermal power and traditional photovoltaics, and the ability to harness much of sunlight's spectrum and thus achieve extremely high efficiencies.
In theory, these devices could climb all the way toward 80 percent efficiency and beyond, though for now we'll have to settle for a mere 3.2 percent. Still, that is more than triple the efficiency of previous efforts, which have peaked at around 1 percent.
Among the reasons for the huge gap between potential and reality is heat. The new device's absorber-emitter reached a temperature of 962°C; at those temperatures, the devices are difficult to optimize and operate. The 3.2 percent achieved is a result, the investigators say, of the specific materials and design of the absorber-emitter: the outer layer uses an array of multiwalled carbon nanotubes, and the emitter portion is a photonic crystal layer made of silicon and silicon dioxide .
"Our device is planar and compact and could become a viable option for high-performance solar thermophotovoltaic energy conversion," they wrote in the Nature Nanotechnology. And it also has the potential to aid in energy storage, since heat is an easier stored form of energy than electricity. The prototype has reached 3.2 percent, but the group thinks 20 percent, which would put it in range with standard PV modules, is well within reach. In an e-mail, Lenert told me that "efficiencies beyond this level will require improvements in low-bandgap cells, as well as even better control of the thermally-driven spectral conversion process using wavelength and angular selective surfaces." The research center at MIT is pursuing those and other angles to bring this idea into popular use.