High-efficiency solar cells come down to earth
After proving themselves in space, advanced multijunction solar cells are finally returning to the earthbound applications for which they were originally designed.
Unlike most space technologies, multijunction solar cells, credited with helping scientists gather much more data than expected on Mars by dramatically extending the extraterrestrial lifetimes of the Spirit and Opportunity space rovers, were initially designed for earthbound applications. After proving themselves in space, the high-efficiency cells are finally becoming cost-effective for generating renewable energy back on terra firma.
More than 20 years ago, Jerry Olson of the U.S. National Renewable Energy Laboratory (NREL) and his colleague Sarah Kurtz developed a type of multijunction cell that uses advanced materials—gallium indium phosphide and gallium arsenide—to increase energy efficiency. Olson stresses that the cells were "always envisioned as being one piece in the puzzle for terrestrial solar photovoltaics." However, satellite manufacturers were the first industry to capitalize on the cells' potential. Nasser Karam, vice president for advanced technology at Boeing Spectrolab, a company that licensed the NREL design and has continued to improve upon it, says that such multijunction cells are widely used for powering the latest generation of satellites.
In December 2006, Spectrolab achieved a record 40.7% efficiency with a triple-junction version of the cell. The rapidly rising efficiency of such cells is part of the reason so many businesses and governments around the world have become interested in their terrestrial applications, says Martha Symko-Davies of NREL. To go from 32% to 40% efficiency in 5 years is a very significant achievement, she explains.
Multijunction cells perform at higher efficiencies than conventional single-junction silicon solar cells, because they convert more of the solar spectrum into energy by breaking it up into chunks, Olson says. For example, the first layer of Spectrolab's record-breaking triple-junction cell is composed of gallium indium phosphide, which converts short-wavelength portions of the spectrum, such as blue and UV. The second layer, made of gallium arsenide, captures the middle part of the spectrum. The third germanium layer does a good job with IR light, Olson says.
The multijunction cells' newfound popularity for terrestrial applications is also attributable to the skyrocketing cost of generating energy, especially outside the U.S. A worldwide shortage of silicon for producing conventional photovoltaics is also a factor, as is the availability of solar concentrators to exploit the full spectrum of solar light more efficiently.
Solar concentrators are not a new technology for amplifying solar energy, but multijunction cells can capitalize on this ability much more fully than silicon-based single-junction cells can, Kurtz says. "If you go to the trouble and expense of putting together the optics of a concentrating system, the tracker, and the other hardware, you'd like to make the best use you can of your investment, so you will naturally want to choose the highest-efficiency cell—about 35% for a multijunction cell compared to 25% for a silicon cell," she explains. Paul Sharps of Emcore Corp., another company developing high-efficiency solar cells with advanced materials, says that his company's designs can absorb the energy equivalent of 1000 suns and that even higher levels are possible.
"Utility-scale solar electricity using concentrating technologies saw a surge of interest" in 2006, according to the Solar Energy Industries Association. In February 2007, Merrill Lynch Equity/Energy Technology predicted that the U.S. market alone for concentrated photovoltiacs would be $1.2 billion, Sharps says. When the multijunction cells are used in conjunction with a concentrator, their high cost—roughly 20 times that of conventional silicon cells—can be offset by the concentration ratio, Olson explains. The cost of concentration systems is also significant, but Sharps says that the U.S. Department of Energy's goal of bringing the total system cost down to below $2 per watt by 2020 is feasible. At this price, solar energy would be cost-competitive with nuclear fission and wind energy, according to a widely cited report produced for the Energy Foundation, a partnership of organizations focused on sustainable energy production.
Because of the way multijunction cells are manufactured, gains in efficiency can be realized quite rapidly. For example, Karam predicts that Spectrolab will be able to produce commercial versions of the prototype cell that achieved the 40.7% record by the end of this year. However, the company's chief scientist, Richard King, stresses that the production version of the chip will probably be 2% less efficient. Even so, this time lapse is much shorter than the gap from prototype to finished product in most other industries, Symko-Davies says.
Meanwhile, Olson and Kurtz received the Dan David Award in March 2007 for their technology's terrestrial applications. The award includes a $1 million prize that they will share with cohonoree James Hansen of the NASA Goddard Institute for Space Studies.
