"This is pretty simple stuff, but it's made possible by sophisticated materials," says Zettl. Earlier this month, researchers at Japan's Meijo University established that carbon nanotubes arranged in forestlike, vertical arrays are the blackest materials ever tested, absorbing almost all the light that falls on them. "This material has the ideal properties for collecting energy from the sun," says Fréchet.

Zettl and Fréchet say that, in theory, these thermal surface-tension effects should be scalable. The Berkeley group started with millimeter-scale machines not only because they were convenient to test in the lab, but also because manipulating objects this size in liquid poses particular challenges. Turbulence is a huge factor at the millimeter scale, says Fréchet. The light-powered mechanism could potentially be used to move nanoscale objects through microfluidic devices employed for medical diagnostics. At the nanoscale, says Fréchet, "surface tension beats gravity." The researchers also hope to scale up their work to make actual boats. Lenses mounted on the back of a large boat should focus sufficient sunlight onto the absorbent nanotubes to propel it. They also hope to make large nanotube-embedded rotors for generators powered by the sun.

"Now they need to see if this will operate in a real environment," says Dean Alhorn, lead engineer on NASA's solar-driven satellite NanoSail-D. NASA's satellite, which was tested this summer, uses a reflective material to absorb the momentum, rather than the heat, from photons; this technology should work well in the vacuum of space but hasn't been practical on Earth. Alhorn says that Zettl and Fréchet have solidly demonstrated that it's the light, not something else, causing the tiny boats to move. However, he notes that the machines have only been tested in tubs of water inside the lab. Alhorn says that the researchers will have to answer the question, "How much force can they generate, versus opposing forces like waves in the real world?"

Indeed, the Berkeley researchers say that their next step is further engineering their devices. "This could be very efficient because nanotubes absorb light so well, but we have to see if this is a viable thermodynamic system," says Zettl.