The IBM researchers are also investigating ways to improve the efficiency of the new solar cells, with the goal of reaching about 12 percent in the laboratory--high enough to give manufacturers confidence that they could be mass produced and still have efficiency levels of around 10 percent, says David Mitzi, at IBM Research, who led the work. Beard recommends targeting 15 percent efficiency in the lab, and Mitzi says this should be possible by improving other parts of the solar cell besides the main CZTS material, or by doping the semiconductor with other trace elements (which is easy with the ink-based process).

What's more, commercial cells will likely use different materials for conducting electrons. The experimental cells used indium tin oxide, which is limited by the availability of indium. But Mitzi says several other conductors could work as well.

One key next step is to completely replace the selenium in the solar cells with sulfur. For the record-efficiency cell, the researchers replaced half of the selenium used in a previous experimental cell. If all of the selenium could be replaced, the cells could, in theory, supply all of the electricity needs of the world. (Provided there are suitable means for storing and redistributing power for use at night or on cloudy days.)

The new type of solar cell will have several competitors, Beard says. For example, non-crystalline silicon is cheaper to make than crystalline silicon, and the efficiency of the resulting cells is improving. Researchers are also finding ways to use less expensive grades of crystalline silicon, and large-scale production has decreased the overall cost of producing such cells, making it difficult for new solar materials to gain a foothold.