conducting lines are, the more efficiently they’ll collect charge from the silicon. If too much of the cell’s surface is shaded by the lines, however, the cell can’t absorb enough light. The thinner the lines are, the closer they can get without causing this problem, but the printing process can’t make them thinner than about 120 micrometers.
The Suntech researchers developed a way to chemically treat the silicon wafer in narrow bands. These treated areas attract silver, which forms metal lines just 20 micrometers wide. In addition to resulting in thinner lines, the process makes it possible to save material costs by using wafers of silicon so thin that screen-printing equipment might break them as it stamped the lines on their surface. It also replaces a treatment used in conventional manufacturing that reduces the cells’ efficiency by damaging the surface of the silicon. The best modules made with the new technology convert about 18 percent of the energy in light into electricity–as opposed to 13 percent for the company’s original solar panels. Next year Suntech intends to roll out a newer version of the technology, which preliminary tests suggest will improve efficiency by another one or two percentage points. The improvement might seem modest, but increased efficiency has a big impact on the cost of the resulting electricity. As a rule of thumb, a percentage-point improvement in efficiency can cut costs by over 6 percent.
Suntech is also funding collaborations with universities, including New South Wales and Swinburne University of Technology in Melbourne, to develop solar cells that get around a fundamental limitation of today’s photovoltaics: they can’t absorb all the wavelengths in sunlight, and they can’t convert all the energy in many of those wavelengths into electronic charge. One key investment is in plasmonics, which makes use of the fact that metal particles deposited on a cell’s surface can guide light energy so that it bounces back and forth within the cell instead of being reflected back out. Exploiting this effect could enable researchers to reduce the amount of active semiconductor material in a solar cell by orders of magnitude, or even to make cells out of materials far cheaper than purified, crystalline silicon (see “Light-Trapping Photovoltaics,”May/June 2010). “Those concepts
will probably not find their way into commercial products in the next 10 to 20 years,” Wenham says. “But they will eventually.”
On the Verge
In spite of the rapid growth of Suntech and the solar industry worldwide, solar power still contributes a vanishingly small portion of the total electricity produced each year. In the United States, it’s slightly above 0.1 percent. “It’s a rounding error,” says Nathaniel Bullard, an analyst for Bloomberg New Energy Finance.
It’s hard to project the course of the still tiny industry. For one thing, all predictions of when solar power might reach grid parity are rife with uncertainties, Bullard says. To take just one example, consider that today the solar panels themselves account for less than half the total cost of the technology. The costs of installation, additional equipment such as inverters, sales and marketing by installers, and, crucially, financing will also need to come down. What’s more, when it comes to grid parity, the price that photovoltaics manufacturers charge for their products is actually more significant than the money it costs to make them–and that will depend on the market. If demand for photovoltaics remains high, in part because government incentives in Germany and elsewhere prop it up, then solar panels could remain expensive enough to keep the price of solar energy well above that of electricity from the grid.
It’s also not yet clear what technology is best suited for widespread use of solar power. Ten years from now, the solar panels most people buy might not even be made of silicon. Switching would be hard for Suntech. While it has the expertise to change direction, its low manufacturing costs depend on investments in equipment and agreements with silicon suppliers. Meanwhile, rival companies have a head start on the technologies that use other materials. First Solar, based in Tempe, AZ, makes thin-film solar cells made of cadmium and tellurium for even less per watt than the Chinese companies making silicon cells. Admittedly, First Solar’s technology converts only about 11 percent of sunlight into electricity; that relatively low efficiency translates into higher installation costs and limits the applications it’s good for. Still, thin-film solar is accounting for a steadily greater share of the overall market, from 3 percent in 2003 to more than 15 percent today.
Yet for all this uncertainty, Shi remains convinced that silicon-based solar power is on the verge of becoming competitive without government subsidies. The idea that solar energy will have to wait for a breakthrough to reach grid parity is “crap,” he says. He adds: “We’re not talking about rocket science. We’re talking basic engineering.”
Kevin Bullis is Technology Review’s energy editor.