DNA Sorts Carbon Nanotubes

Single-walled carbon nanotubes (SWNTs) show great promise as components of nanoscale electronic devices, but most commercial applications have been stymied by the difficulty in isolating nanotubes of identical chirality from a synthetic mixture. Now, Xiaomin Tu and Ming Zheng of DuPont Central Research & Development, together with Suresh Manohar and Anand Jagota of Lehigh University, have shown that the unique molecular properties of DNA can be exploited to sort SWNTs (Nature 2009, 460, 250).

SWNT synthesis produces a mixture of nanotubes with nonuniform diameters and chiralities and, therefore, heterogeneous physicochemical properties. Having previously shown that a particular DNA sequence could form an ordered structure on SWNTs, Zheng and colleagues reasoned that they might be able to find a DNA sequence to purify each type of SWNT in a synthetic mixture. The problem was identifying the correct DNA molecules among an unfeasibly large number (10¹⁸) of possible 30-nucleotide sequences.

To reduce the DNA library to a more manageable size of 350 oligonucleotides, the researchers devised a sequence-pattern-expansion scheme that considered all possible DNA sequences composed of mono-, di-, tri-, and tetranucleotide repeats. They added each DNA oligonucleotide to a random mixture of SWNTs. Then, they used ion-exchange chromatography to separate the 350 solutions into fractions, which they analyzed spectroscopically for the presence of specific DNA-SWNT hybrids.

The study yielded more than 20 sequences that could together purify all 12 major chiral semiconducting SWNTs. To explain the SWNT-sorting ability of DNA, the authors propose a model in which the recognition sequences form stable hydrogen-bonded DNA barrels around specific SWNTs. The ordered structure minimizes interactions with the ion-exchange chromatography resin and causes early elution.

"This is a very impressive study that reports the most selective method yet found for isolating specific structural forms of SWNTs from mixed samples," R. Bruce Weisman of Rice University says. "The main limitation is the small scale and high expense." Zheng tells C&EN that the major obstacle to scaling up the method is the high cost of DNA, which could decrease if oligonucleotide suppliers shift their business model to meet the demands of the nanoelectronics industry.