|
Monday, October 30, 2006 Nanotube Computing BreakthroughA method for sorting nanotubes by electronic properties could help make widespread nanotube-based electronics a reality. By Kevin Bullis
The use of carbon nanotubes in ultrafast computers and other electronic devices has been stymied because batches of the material contain nanotubes with varying electronic properties. One nanotube is semiconducting, while the next is conducting. Now Northwestern University researchers have developed a reliable and potentially practical way to sort through this mess, segregating nanotubes into precisely the types needed for high-performance electronics. The advance could speed progress toward nanotube computers and has many nearer-term applications, including high-definition displays, devices for nanotoxicity testing, and solar cells. The new process separates metallic and semiconducting nanotubes. It also segregates them by diameter (another important parameter for reliable computer chips) and eliminates contaminants, such as other forms of carbon. While the researchers expected to be able to sort nanotubes by diameter, the sorting by electronic type came as a surprise, says Mark Hersam, materials-science and engineering professor and one of the Northwestern researchers. "We didn't believe it at first," he says. Carbon nanotubes are appealing candidates for eventually replacing silicon-based computing because of their small size and excellent electronic properties: some are semiconductors--perfect for transistors--and others are metallic conductors and could be useful as wires for connecting transistors. But getting the right electronic type "makes a big, big difference," says Mildred Dresselhaus, professor of physics and electrical engineering at MIT. Placing metallic nanotubes where there should be semiconducting nanotubes would cause the chip to fail. So although researchers have been able to painstakingly create logic circuits using carbon nanotubes (see "Carbon Nanotube Computers"), the methods employed to sort them are "all pretty tedious," Dresselhaus says, and not something that could be scaled up for manufacturing chips with the millions of transistors needed to compete with today's computers. In addition, past methods have failed to completely separate semiconducting and metallic nanotubes, says Richard Martel, chemistry professor at the University of Montreal. Martel calls the Northwestern researchers' new approach, described this month in the new journal Nature Nanotechnology, "a breakthrough in the field." The researchers begin by adding surfactants to a batch of nanotubes. The surfactants latch on to the nanotubes, but differences in the nanotubes' size and electronic properties cause the surfactants to assemble in different concentrations and arrangements, which in turn lead to measurable differences in density. These distinct densities can be sorted out using a well-known process called ultra-centrifugation, which involves spinning the materials at ultrafast speeds--up to 64,000 revolutions per minute.
|
Share »
Digg this
Add to del.icio.us
Add to Reddit
Add to Facebook
Slashdot It!
Stumble It!
Add to Newsvine
Add to Connotea
Add to CiteUlike
Add to Furl
Googlize this
Add to Rojo
Add to MyWeb
Favorite
Print
E-mail
Assembling Nanotubes
07/10/2008
Comments
protn7 on 10/30/2006 at 6:50 PM
19
kitk on 10/31/2006 at 12:19 AM
52
sobisch on 10/31/2006 at 3:00 AM
1
Unfortunately, no information was provided relating to the nature of surfactant combinations used to produce the density differences. I would speculate that because metallic and semiconductor nanotubes should have marked differences in their electronic structure surfactants like LAS with an aromatic moeity might be able to specifically interact with one form of carbon nanotubes. As common in floatation of ores surfactants or other additives with specific interaction might be designed.
I wondered that no reference was made in the article that there is already a way to sort the nanotubes based on alternating current electrophoresis developed by German researchers in 2003.(http://www.sciencemag.org/cgi/content/abstract/sci;301/5631/344)
Maybe both methods could be combined.
Kind regards
T. Sobisch
http://AppliedColloidsSurfactants.blogspot.com
http://www.AppliedColloidsSurfactants.info
billdale on 12/11/2006 at 10:40 PM
15
As for the different means the Germans and Northwest teams use, the team at Northwest is wise to have looked for other avenues to the same end. Each approach can have specific advantages of speed, purity, cost, and other factors. The technique at Northwest may have advantages in the area of less pollution byproduct, for instance. Only as the procedures mature can such matters be evaluated. As an American, I do hope the team here in the US will find their system of great value.
As for the product not being 100% pure and nanoscale circuits requiring complete purity, it may mean that there is simply a certain percentage of circuits that do not pass QC, much as LCD panels suffered years ago. They were initially very expensive due to a low percentage of acceptable product, but they were still able to produce good panels. As processes improved, acceptable yield increased, and costs dropped accordingly. The same may be true of nanoscale circuits-- QC will determine yield and costs.
My own area of interest lies in the use of CNTs for batteries and ultracapacitors. Costs and energy densities need to improve dramatically for some applications, and Northwest's approach gives hope. I do not think 100% purity in each batch will be absolutely necessary as it would be in attempting to make CNT-based computing devices. There is an enormous number of potential apps for nanotubes of every flavor, and not all apps require absolute purity-- very good may be good enough. and therefore patentable and usable.