he tantalizing possibility of making nano-scale electronic devices fuels the continuing excitement in the area of carbon nanotubes and explains the appearance in the current Hot Ten list of four papers (#2, #3, #4, and #5), of which paper #4 appears for the first time. Carbon nanotubes are long, thin cylindrical molecules that get their name from their diameter, which is about one nanometer (10^-9 m), and they can be up to 100,000 nanometers long. Some conduct electricity as if they were ultrafine wires, although this depends very much on the way the carbon atoms are stashed. A slightly different arrangement of carbons in the tube wall can result in its being a semiconductor.

   Paper #2 appeared earlier this year in the March/April list and was reported on in some detail since it announced the "large-scale" production of SWNT's (single-walled carbon nanotubes) by using a simple electronic discharge between to carbon electrodes (see Science Watch, 10(2):7, March/April 1999). "Large scale" in this context refers to gram quantities, although the authors point out that it could easily be upgraded to give much higher yields. The work originated from groups at the universities of Montpelier and Nantes in France and Pennsylvania in the United States, and the team was led by Patrick Bernier. Their success came about through the incorporation of an yttrium-nickel catalyst in the carbon anode.

   While this work offers the possibility of making SWNT's readily available, if their potential is to be realized then there needs to be a reason to make them. The new paper, #4, reports a possible application in the form of a nano-scale transistor. This device, made up by Sander Tans, Alwin Verschueren, and Cees Dekker at the Delft University of Technology in The Netherlands, consists of the nanotube, which is supported on a silicon oxide layer on top of a silicon gate, bridging a source and drain electrode.

   When a voltage is applied to the conducting substrate, then "holes" flow down the tube and the device behaves as a p-type transistor. Remarkable as it appears, this research has now been verified by Phaedra Avouris at IBM, Yorktown, Paul McEuen at UC Berkeley, and Hongjie Dai at Stanford University, says Dekker. In addition, the resistance of the nano-transistor can be varied by a thousand-fold or more, a feature that has so far eluded most other nano-scale devices.

   Paper #4 is one in a series of reports in this area by Dekker's group; the first, coauthored with Richard Smalley of Rice University, Houston, Texas, was published in Nature (386:474-7, 1997) and made the #2 in chemistry in the March/April Science Watch. (This was followed by another, equally innovative, paper from Smalley's team describing the electronic structure of carbon nanotubes–see #3 on the current list. That particular issue of Nature also contained paper #5, describing similar work being done at Harvard University by Charles Leiber's group.

Dekker, too, has continued to be productive in this field–witness his paper in Nature on electron correlations in nanotubes and on their behavior at room temperature (394:761-4, 1998). However, it is paper #4 which has attracted attention recently, and it was in this that he speculated about the possibility of designing a metallic-semiconductor junction that consisted entirely of carbon atoms, simply by varying the arrangement of carbon atoms in the tube itself.