Until now, solar cells made from such cheap materials have been impractical because of a fundamental contradiction in their design requirements. To be efficient, solar cells must do at least two things well. First, they must absorb light, so they need active materials thick enough that light can't pass through them. But they also need to collect the electrons knocked loose by absorbed photons. For this, extremely thin materials are usually better; otherwise, electrons can get trapped inside the material. One way to reconcile these competing design constraints is to make relatively thick layers of material but to use extremely pure, crystalline materials that lack the defects and impurities that can trap electrons. Such materials work well, but they're expensive, keeping the price of solar panels high.

Nanowires such as those Lieber used for his solar cells offer an alternative. The nanowires can absorb significant amounts of light along their length. At the same time, electrons have only to move a short distance in the nanowire, from one concentric layer of material to another, to be collected. (The layers serve to separate electrons from their positive counterparts, holes, which allows the electrons to be collected.) Since the materials are thin, the chances of an electron being trapped by a defect before escaping from one layer to the next are low, so it's possible to use cheaper materials with more defects.

Lieber demonstrated that nanowires can indeed produce electricity, but a number of challenges remain before they will find their way into commercial solar cells. Lieber has tested only small numbers of nanowire solar cells. For large-scale applications, the nanowires would need to be chemically grown in dense arrays. Atwater and Lewis recently took steps in this direction, publishing in the past month two papers in which they describe growing dense arrays of microscopic wires, but wires without the multiple layers that Lieber's have. Paired with a liquid electrolyte, the wires generated electricity from light. Since it may prove easier to manufacture solid-state solar cells such as Lieber's, however, Lewis and Atwater are working to produce arrays of wires with multiple layers.

The most significant limitation of the work of both groups is the poor efficiency of their solar cells. For example, Lieber's cells converted 3.4 percent of incoming light into electricity. While that's an encouraging number for proof-of-concept solar cells in the lab, it's a far cry from the 20-plus percent efficiency of conventional silicon solar panels. Even with the potential advantage of cheaper materials, wire-based solar cells would probably need to be about 10 percent efficient if they were to compete with existing technology. The researchers' next steps include finding ways to make more dense arrays of wires to absorb more light and, in Lieber's case, to find ways to generate increased voltage from nanowire solar cells.