"The biggest problem is the risk associated with transitioning from pilot line to successful commercial production," says Ken Zweibel, who heads up thin-film research at the National Renewable Energy Laboratory (NREL) in Golden, CO. "There's a tremendous amount of manufacturing development implicit in that -- and it's always underestimated by outsiders. It's incredibly subtle to make square miles of novel semiconductors every year at 95 percent yield."

But while the solar-cell industry is maturing, the ultimate technology for the industry remains in doubt. Billy Stanbery, CEO of a CIGS solar-cell startup called Heliovolt, in Austin, TX, says he thinks profits from second-generation solar technologies such as the one his company is developing will in turn be a major source of funds for research into yet another generation of solar technologies.

In fact, such research is already happening, and could get a boost from new funding for basic solar research in next year's federal budget. Third-generation solar technology will be designed to break through a major problem with today's new technologies, which, although they are cheaper and easier to make than crystalline silicon cells, are not much more efficient at converting sunlight to electricity.

Nanostructures could help change this, says Arthur Nozik, a senior researcher at NREL. One promising method, he says, uses nanostructures such as quantum dots, known for their unusual electronic and photonic properties, to divide the energy from each photon into multiple electrons, which can then generate current. This work has the potential to more than double current solar cell efficiency, he says, which could help solar power finally emerge from being a niche fossil-fuel alternative, albeit a booming one, to a mainstay of energy production.