The race for inexpensive, highly efficient solar cells may have gained another contender in the form of silicon microwires. Efforts to develop ultra-thin wires that convert sunlight into electricity are not new to the solar power field, but a new method for growing the wires has roughly doubled their conversion efficiency and may hold the key for even larger gains.
“All wires thus far have had 1 or 2 percent efficiency [at the array level] with fundamental questions about whether they could ever go higher,” says Nathan Lewis, a chemist at Caltech who coauthored the study, which appears in Science.“We’ve demonstrated 3 percent efficiency and shown that there is no fundamental reason they can’t perform at over 10 percent.”
Silicon nanowires, or in this case slightly larger-diameter microwires, are typically grown from a silicon substrate with the help of tiny gold droplets. Under high temperatures, a single wire will quickly sprout from each droplet like a blade of grass. Gold is an excellent catalyst for wire growth, but it also introduces impurities that are generally believed to inhibit electron transport within the wires, reducing their overall efficiency.
Using copper instead of gold as the catalyst, Lewis and colleagues achieved roughly twice the efficiency of prior efforts in an array of wires. They believe the results are due to higher silicon purity and increased electron transport capacity compared to prior efforts that relied on gold catalysts.
In what they are calling a proof of concept study, the researchers kept the “packing fraction” of their array at 4 percent. Packing is a measure of how much of the surface of an array has wires protruding from it. A packing fraction of 4 percent means that 96 percent of the surface of the array has no wires and therefore is incapable of capturing sunlight and converting it into electricity. Lewis says that simply increasing the packing fraction to 15 to 20 percent will result in a fourfold increase in efficiency.
Some doubt it will be that simple. “If it’s that easy, why haven’t they done it?” asks Ray LaPierre a professor in the engineering physics department at McMaster University in Ontario. LaPierre says increasing the packing fraction is technically feasible through a technique known as “photolithography,” but this would likely be prohibitively expensive for commercial solar cell production.
Another potential problem that LaPierre thinks may inhibit higher efficiencies is the electron transport capacity of the wires. When photons of sunlight are captured by a wire, they produces electrons that must then escape from the material to produce an electric current. Electrons, however, are easily trapped along the surface of the wires, reducing their overall efficiency. Thin-film solar cells have to overcome this challenge as well, but the problem is especially acute in thin wire cells because they have a much larger surface area per volume than planar films.
The wires that Lewis and colleagues grew, however, are 1.6 micrometers in diameter, three orders of magnitude thicker than typical solar cell nanowires. The thicker microwires have a lower surface area to volume ratio that, according to modeling conducted by the group, boosts the electron transport capacity of the wires.
Matthew Beard, a senior scientist at the National Renewable Energy Laboratory in Golden, CO, says the relatively high surface area of the wires could be a plus for converting solar power into hydrogen fuel. The high surface area and low cost of raw materials of the silicon microwires means they could be used directly as electrodes to hydrolyze water into hydrogen.
Still, Beard says microwire solar technology will have a tough time competing as a source of power against currently available thin films that are relatively inexpensive and already achieve 10 to 12 percent efficiencies. But Beard adds that silicon, the raw material for the wires, is more readily available than metals such as cadmium and telluride that make up today’s most efficient thin films. “This technology has a long way to go, but potentially can compete, as silicon is more abundant than those materials and potentially cheaper,” he says.