<?xml version="1.0" encoding="UTF-8"?><?xml-stylesheet type="text/xsl" href="../nsu_article.xsl"?><!DOCTYPE nsuarticle PUBLIC "-//NPG//DTD NSU//EN" "../nsu_article.dtd"><nsuarticle type="news"><articleidlist><articleid type="uid">991118</articleid><storyno>-1</storyno><articleid type="doi">10.1038/nsu991118</articleid><storyno>-1</storyno></articleidlist><pubfm><pubdate><dayofweek name="Friday"></dayofweek><day>12</day><month>November</month><year>1999</year></pubdate><category>technology</category></pubfm><fm><title>It's a small, kinky world</title><aug><fnm>Philip</fnm><snm>Ball</snm></aug></fm><body><p>They are the ultimate in electronic miniaturization: tube-shaped molecules of carbon, each scarcely wider than a filament of DNA, able to conduct electricity and to be bent, cut and moulded into circuit wiring or even into new electronic devices. They are called carbon nanotubes; and US physicists have now reported that they can be manipulated into genuine electrical circuits.</p><p>The idea that integrated circuits could be wired up using conducting carbon nanotubes has been much vaunted since their discovery in 1991. It was quickly appreciated that, since they have essentially the same structure as graphite, another form of pure carbon, nanotubes should, like graphite, be able to conduct electricity. Detailed calculations verified that some carbon nanotubes could indeed act like tiny metallic wires.</p><p>At present, wires for integrated circuits are carved from thin films of metals or semiconductors deposited on a silicon chip. But the standard lithographic techniques used by the microelectronics industry struggle to create features thinner than about one-fifth of one-thousandth of a millimetre &ndash; 0.2 micrometres. Carbon nanotubes can be made 100 times thinner than this, and so might in principle permit a far higher density of wiring &ndash; and so greater miniaturization.</p><p>The problem is finding a way to position objects this small. One solution, developed by researchers at Stanford University last year, is to grow the nanotubes directly onto the metal electrodes of the devices used in the circuit (see <weblink url="http://www.nature.com/">Nature</weblink> <weblink url="http://www.nature.com/cgi-taf/DynaFixer.taf?RQID=395878A0">395</weblink> <weblink url="http://www.nature.com/cgi-taf/DynaFixer.taf?RQID=395878A0">,878</weblink>; 1998)<bibr rid="b1">1</bibr>. Nanotubes can be grown from a carbon-rich vapour, like icicles condensing from winter air, and the Stanford method uses a catalyst to promote growth at the electrode.</p><p>Now Alan Johnson and colleagues from the University of Pennsylvania report in the journal <emphasis>Applied Physics Letters</emphasis><bibr rid="b2">2</bibr> that they can assemble circuits from preformed nanotubes. To manipulate components, the researchers use a device called the atomic force microscope (AFM). Originally conceived as an instrument for taking high-resolution images of the atomic-scale structure of surfaces, the AFM has also proved immensely useful for pushing molecules around on surfaces. It consists of a very fine needle tip attached to an arm which can be moved with great precision. Johnson and colleagues have used the tip as a kind of finger to gently nudge nanotubes around on an oxidized silicon wafer.</p><p>They were able to place nanotubes on top of one another, cut them into shorter sections, and sweep away the fragments and unwanted tubes. The researchers decided to investigate in detail what happens when nanotubes cross, so that one lies kinked over the other. They laid down metal contacts at the ends of the tubes, so that they could apply a voltage and measure the current.</p><p>They found that this simple tube junction acted like a device called a 'tunnel junction', in which electrons leak through a poorly conducting region between two conductors. They suggest that the deformation of the upper tube, as it passes over the lower one, causes this localized decrease in conductivity by disrupting the atomic structure of the tube.</p><p>Kinks in nanotubes, then, might be used to modify their electronic properties, and perhaps to 'build' device-like structures into the very wires themselves. If methods can be found for generating such kinks in a controlled way &ndash; and the AFM will surely be a candidate tool for making them as well as moving them about &ndash; then we might look forward to microelectronics shrunk to molecular dimensions and based not on silicon but on carbon.</p></body><bm><refgrp><bib id="b1" arturl="http://www.nature.com/"><refau><snm>Kong</snm>, <fnm>J.</fnm></refau>, <refau><snm>Sog</snm>, <fnm>H.</fnm> <inits>T.</inits></refau>, <refau><snm>Cassell</snm>, <fnm>A.</fnm> <inits>M.</inits></refau>, <refau><snm>Quate</snm>, <fnm>C.</fnm> <inits>F.</inits></refau> &amp; <refau><snm>Dai</snm>, <fnm>H.</fnm></refau> <atl>Synthesis of individual single-walled carbon nano-tubes on patterened siliconwafers.</atl> <jtl>Nature</jtl>  <vol>395</vol>, <spn>878</spn> <pubyear>1998</pubyear></bib><bib id="b2" arturl="http://ojps.aip.org/journals/doc/APPLAB-home/top.html"><refau><snm>Lefebvre</snm>, <fnm>J.</fnm></refau>, <refau><snm>Lynch</snm>, <fnm>J.</fnm> <inits>F.</inits></refau>, <refau><snm>Llaguno</snm>, <fnm>M.</fnm></refau>, <refau><snm>Radosavljevic</snm>, <fnm>M.</fnm></refau> &amp; <refau><snm>Johnson</snm>, <fnm>A.</fnm> <inits>T.</inits></refau> <atl>Single-wall carbon nanotube circuits assembled with an atomic force microscope</atl>, <jtl>Applied Physics Letters</jtl>   <vol>75</vol>, <spn>3014</spn><epn>3016</epn> <pubyear>1999</pubyear>.</bib></refgrp></bm></nsuarticle>
