A Step Closer to Realizing a High Efficiency Solar Thermophotovoltaic Device
by Joseph Bennington-Castro
Materials Research Society | Published: 30 January 2014
Photovoltaic and solar thermal devices are the most common approaches to harness sunlight for energy. Decades ago, scientists proposed that hybrid solar thermophotovoltaic (STPV) devices could be a way to get the best of both worlds, and achieve high efficiency in a compact form. However, previous attempts at making such devices have only resulted in efficiencies of about 1 percent. Now, researchers at the Massachusetts Institute of Technology (MIT) have fashioned a new prototype STPV device, which utilizes a unique absorber-emitter surface to attain efficiencies of 3.2 percent. With a little scaling, the device should be able to easily reach 20 percent efficiency, researchers say.
"Although it sounds low, our work suggests there is potential to further tailor the energy balance of the device," says Evelyn Wang, a mechanical engineering professor at MIT who directed the new study, published recently in the journal Nature Nanotechnology . "There is significant potential in the future to do better and get closer to the theoretical limit of over 80 percent efficiency."
Photovoltaic (PV) devices generate power by directly converting sunlight into electricity, but the technology is based on silicon semiconductor physics, so each PV cell has a range of wavelengths that it cannot use (the bandgap). What's more, it's difficult to store electrical energy in a low-cost way. Solar thermal devices use sunlight to drive heat engines, but these machines are best suited for utility-scale power plants. STPVs are thought to be able to get around these issues by collecting the whole solar spectrum and converting it to heat, before reemitting light in wavelengths tuned specifically for PV cells. However, because of the high temperatures involved-on the order of 1000 °C-STPVs run the risk of experiencing a lot of radiative losses. This issue has kept previous STPVs in the realm of low efficiencies of 1 percent or less.
The key to the success of Wang and her colleague's STPV lies in the absorber-emitter surface, Wang says. The team first created a one-dimensional photonic crystal by depositing layers of polycrystalline silicon and silicon dioxide on to a 1-cm 2 silicon wafer using plasma-enhanced chemical vapor deposition. This crystal served as the emitter. On the other side of the substrate, they grew an absorber made of multiwalled carbon nanotubes, which are nearly blackbody absorbers. To achieve optimal performance, they varied the emitter-to-absorber area ratio from 1 to 10. They also metalized the sides of the silicon substrate, as well as the inactive area around the absorber, with tungsten to help limit parasitic radiative losses from the device.
Using simulated sunlight and a focusing system to concentrate the light by a factor of 750, the team heated the absorber to 962 degrees Celsius. "The solar concentration is much lower than what was demonstrated in previous work," Wang says. The heat transferred to the emitter, which then glowed with a peak intensity mostly above the bandgap of an adjacent PV cell from MIT's Lincoln Laboratory. In their experimental tests, this new STPV device achieved efficiencies that were threefold higher than previous STPV devices.
In conjunction with the device, the team also developed a model that was in good agreement with their experiments. The model suggests that simply scaling the device to 100 cm 2 could result in efficiencies of 20 percent. "That would make it comparable to efficiencies of typical silicon PV cells," Wang says. Further optimizing the device, such as by using better PV cells and photonic crystals, could get it closer to the theoretical STPV limit of over 80 percent.
"This work represents a great breakthrough in the field of STPV research, establishing the paths for achieving highly efficient, practical STPV systems," says Alejandro Datas, a researcher at the Technical University of Madrid's Institute for Solar Energy, who has also worked on STPV devices and wasn't involved in the current research. "The three-fold efficiency improvement, [with] respect to our previous result (published in 2012), suggests a high learning rate for the STPV development, so that I believe that the efficiencies claimed in this work (20 percent) could be approached in a relatively short period of time."
Read the abstract of the study in Nature Nanotechnology here .