There's a limit on the conversion efficiency of a conventional solar cell. No matter how it's tweaked, it can only convert 31 percent of the light that hits it into usable electrical current. That's because there's a broad spectrum of wavelengths in sunlight, and some of it has more energy than the active material in the solar cell can handle. High-energy light hits the active material in a solar cell and knocks loose electrons that have a similarly high energy--then these electrons rapidly lose that excess energy as heat.

Physicists know that if they could capture "hot electrons", they could more than double the efficiency of solar cells. The problem is that they lose their energy in a picosecond. Now, researchers have for the first time demonstrated that it's possible to capture hot electrons while they're still in their high energy state, before that heat loss happens.

Careful design at the nanoscale is key. Instead of a conventional bulk semiconductor, the researchers used quantum dots, because these nanomaterials can confine electrons over a longer timescale. "Nanomaterials can keep electrons electrons hot for a longer period of time, so that you can get them out," says Xiaoyang Zhu, professor of chemistry at the University of Texas, Austin.

The confinement is great--until you want to get the hot electrons out. "The electron likes to stay inside the nanomaterial, so you need to make an extremely strong interaction with another material" that will conduct the electrons out of the quantum dot, Zhu says. His group coated the quantum dots with a very thin layer of an electrical conductor, and were meticulous about the quality of the interface between that material and the quantum dots.

So now it's possible to get hot electrons out, but one major problem remains. Those hot electrons require new device designs that prevent them from simply losing their energy to heat once they enter the metal wire of an electrical circuit. "We hope to inspire people to work on the engineering," says Zhu.

This research was published this week in the journal Science.

Silicon solar cells built on a nanostructured substrate (top left) have a surface patterned with nanoscale domes (top right). The scale bar in both electron-microscope images is 500 nanometers. The diagram shows the layers of the device, from bottom to top: a quartz substrate, a reflective layer of silver, a transparent conducting oxide, the active layer of amorphous silicon, and another oxide layer. Credit: ACS/Nano Letters

The accumulation of dust on the surface of a solar cell can block light and cut into cell efficiency. Researchers at Stanford have demonstrated that solar cells patterned at the nanoscale with domed structures absorb more light and, as a bonus, are self-cleaning.

The nanoscale patterning is not just on the surface of the cell but is applied to every layer. The cells are built on a substrate patterned with nanoscale cones. The bottom layer is a film of silver 100 nanometers thick that acts as an electrical contact and a light reflector; atop this is a film of amorphous silicon sandwiched between transparent conducting layers. Though the substrate is jagged, the accumulation of layers results in domed structures that happen to resemble the mushroom-like structures other researchers have been developing for self-cleaning surfaces. An added layer of hydrophobic molecules makes the cells nearly superhydrophobic: water droplets roll along the surface, pulling dust away with them.

These nanodome structures not only repel water, but help trap light. Because they're so small--about 500 nanometers in diameter--the nanodomes interact with light in a cool way, absorbing 94 percent of all light from the infrared to the ultraviolet. A flat solar cell made from the same materials absorbs only 65 percent of light in the same broad spectrum. So far the overall power conversion efficiency of the cells is 5.9 percent. The lead researcher, Stanford materials science professor Yi Cui, says these patterning techniques could be applied to other solar materials. This work is described online in the journal Nano Letters.

Solar companies presenting business plans to investors at a National Renewable Energy Laboratory (NREL) conference this week devoted particular attention to how they hope to compete with Chinese manufacturers. The audience at the NREL Industry Growth Forum in Denver consisted largely of venture capitalists and partners from private equity firms.

Stellaris, a company that assembles solar modules in Lowell, MA, has already received $6.1 million in funding to develop techniques for packaging silicon and thin-film cells. The company, represented at the conference by CEO James Paull, is seeking further financing in 2010.

Paull said that while European companies' cell-to-module costs are 70 cents per watt, China's are half that. "Solar modules have become a commodity, and China is dominating," he said. Like most of the other presenters, Paull didn't reveal too much about his company's technology. But he said that Stellaris hopes to save costs by adding passive plastic concentrators to silicon and thin-film cells and by reducing cell sizes.

An executive from a large European solar company expressed skepticism, however, that the US will ever be able to catch up with Chinese solar manufacturers. The executive, who manages his company's operations in China, said his company had explored manufacturing in California and Texas but that the labor costs were much too high. That said, he was at the conference looking for new solar technologies to buy up--an area where the US does still have an edge.