Credit: © 2008 ACS

Carbon nanotubes excel at carrying electrons, and so have inspired a wide variety of device designs. However, nanotubes come in two flavours — semiconducting and metallic — and these designs usually require a flavour guarantee. So far this has proven difficult to deliver. Now, Yuta Sato and colleagues1 at AIST have made an observation that could help.

The researchers followed a standard technique (density-gradient centrifugation) to generate nanotube bundles that were mostly metallic or mostly semiconducting. At this point, optical or Raman spectroscopy is usually used to characterize the tubes. Instead, the authors examined individual tubes using a transmission electron microscope. The microscopy data told them the nanotube chiral angle and diameter, and could even distinguish between rotated versions of identical tubes.

They found that tubes from the metallic bundle had a concentration of high chiral angles (greater than 20°), whereas tubes from the semiconducting bundle did not. This may point to a chirality dependence of the tube separation technique, the growth technique, or of the tube type itself. Furthermore, chiral angle characterization on an individual tube-by-tube basis using electron microscopy, gave different results than absorption spectroscopy. This may have resulted from the presence of defects in individual tubes, and points to the need for additional research to reconcile these approaches.