In addition to reducing costs by using less active material, LaPierre’s team can also cut the cost of the substrate that the nanowires are grown on. LaPierre’s team doesn’t require an expensive Group III-V substrate. It has successfully grown its nanowires on substrates made of more plentiful and relatively cheaper silicon. It’s also working on using even lower cost substrates made of glass, which would be ideal for building-integrated PV applications. Flexible substrates such as polymer films and carbon nanotube fabric could be useful for many applications, and could be manufactured with inexpensive roll-to-roll processes.

To further drive down costs, the focus on cheaper substrates will be complemented by an attempt to replace the gold catalysts used to grow the nanowires with aluminum, although more work in this area is needed to achieve the necessary nanowire densities. “We have grown nanowires from aluminum, but gold works much better,” says LaPierre.

Charles Lieber, a professor of chemistry at Harvard University who has created single light-harvesting nanowires made of silicon, says that his team is also pursuing the use of other materials for making nanowires. “But there are many challenges in going from nanowire to photovoltaic,” says Lieber. He adds that comparison of approaches is difficult without data on the energy-conversion properties of each material.

Nathan Lewis, a professor of chemistry at the California Institute of Technology and an expert on nanowire structures, says that it’s too early to say which approach and materials are best. “We know nanowires work in bulk form, but we don’t know if you can make high-purity, high-quality nanowires and control all their electrical properties,” says Lewis. “There’s no theory that one works better than the other. It’s just a question of getting any of them to work.”

It’s still early days for McMaster, which in prototypes has only achieved low efficiencies–“where silicon PV was in the 1950s,” says LaPierre. But he’s optimistic that the higher-efficiency materials and the approach chosen will get results.