Carbon nanotubes hold the promise of flexible electronics, smart sensors and unique medical devices.
 
But before they can be put to wide commercial use, scientists have to figure out a way to sort through the tangle of different types of nanotubes that results when the tiny structures are made.
 
A team of DuPont Co. and Lehigh University researchers has taken a big step forward by refining a process to sort nanotubes using a surprising material: DNA.
 
The research, led by DuPont research associate Ming Zheng, built on a 2003 project that first showed that nanotubes could be sorted using a sequence of DNA, the substance that carries genetic instructions for living organisms.
 
In the latest work, published earlier this month in the peer-reviewed journal Nature, the research team identified more than 20 DNA sequences that reacted with 12 types of nanotubes, sorting them at an 80 percent to 90 percent purity level.
 
"People have tried many different molecules to interact with nanotubes, and DNA stands out as one of the best," said Zheng, a nine-year DuPont employee with degrees in electronics, physics, chemistry and microbiology.
 
Carbon nanotubes -- cylinder-shaped structures that are only a few atoms wide -- are a key structure in the field of nanotechnology, the manipulation of tiny materials measured in billionths of a meter.
 
Current production techniques create a mixture of nanotubes with varying properties. Semiconducting nanotubes must be separated from the metallic variety in order to use them in electronics.
 
Beyond their conductivity, the nanotubes also vary in size and chirality, or symmetry, and the structures must be uniform to get predictable results in electronic applications.
 
"If you don't have control over the nanotube's structure," Zheng said, "you basically don't have control over what kind of properties you're going to get from the nanotube."
 
Other researchers have turned to chemistry to sort nanotubes -- for example, a DuPont and Cornell University research team published a paper earlier this year detailing a method for isolating semiconducting nanotubes using a fluorine-based solution.
 
The latest research used the fundamental building block of biology, DNA, which shares a complementary structure with nanotubes that allows the two materials to bond. A major challenge for the scientists was the massive number of possible DNA sequences.
 
The research team -- Zheng and Xiaomin Tu of DuPont, and Lehigh chemical engineering professor Anand Jagota and student Suresh Manohar -- devised a systematic method to search the DNA library, eventually identifying more than 20 useful sequences out of 350 tested.
 
"I think it's an impressive study," said R. Bruce Weisman, a chemistry professor at Rice University in Houston who studies carbon nanotubes. "These people have developed the most refined method for separating nanotubes into their different structural forms."
 
Weisman said the DNA method has demonstrated the highest quality separations, but other sorting methods are more cost-effective.
 
"It's a process which is really only feasible on a small scale at the moment," Weisman said.
 
Zheng said the cost of DNA may prohibit scaling up the sorting technique to produce nanotubes on an industrial scale, for use in products like circuits, solar panels and electronic displays.
 
But he said the sorting process could be used to isolate a "seed" nanotube, which could potentially be cloned and reproduced on a wide scale -- an active area of nanoscience research.
 
"I think that's probably a more realistic solution for scaling up," Zheng said.