There are many competitive processes for making smooth films, says Nicholas Fang, assistant professor of mechanical science and engineering at the University of Illinois at Urbana-Champaign. Norris's method for smoothing metal surfaces is "quite unique," says Fang, and should prove useful for making plasmonic structures, particularly if the molds prove to be durable over the long term. However, surface roughness is only one source of problems with plasmonics, says Fang. Now what are needed are methods for making the edges of the features in these patterned metal films smooth.

Harry Atwater, professor of applied physics and materials science at Caltech, agrees. "When you're making waveguides, the edges are just as important as the surfaces." Atwater is developing plasmonic concentrators for solar cells. Silicon solar cells are usually about 100 micrometers thick; thinner cells would be cheaper, but their performance suffers. Atwater has found that adding a patterned layer of metal that can interact with plasmons makes it possible to collect and concentrate light from wider angles and improves performance in thin silicon solar cells. Techniques for printing plasmonics for solar cells will have to be cheap and scalable because cost per unit area is such an important consideration for photovoltaics. Norris's technique is "a useful idea," says Atwater, but only time will tell whether it can work repeatedly over the large areas required for solar cells.

The future of plasmonics, says Atwater, will probably be in new materials besides metal. Metals like gold and silver, which have been used in plasmonics for about a decade, have an intrinsic electrical resistance that causes plasmons to scatter, no matter how smooth the surface and edges. The carbon nanomaterial graphene, which has a low resistance, might fit the bill. Atwater says scientists will also have to "pull out the metallurgy textbooks" to look for other materials.