Graphene just got easier to see. The one-atom-thick material which is the favourite substance to succeed silicon for making electronic components, can now be identified quickly and cheaply instead of by the lengthy and expensive methods of either atomic-force microscopy (AFM) or scanning electron microscopy (SEM).
The Journal of the American Chemical Society (JACS) reports that Jiaxing Huang, Assistant Professor of Materials Science and Engineering at the North Western McCormick School of Engineering and Applied Science, has come up with a way to identify graphene by using dye - more particularly fluorescein dye.
"If one cannot examine these materials, quality control is going to be difficult," says Huang, "so we thought, how about we just put dye everywhere? That way, the whole background lights up, and wherever you have graphene will be dark. It's an inverse strategy that turns out to work beautifully."
The material was viewed through a fluorescence microscope (FQM) which are freely available and cheap to use.
The quality of the image obtained using FQM was as good as those obtained using AFM and SEM.
"When (graduate student) Jaemyung first showed me the FQM images of graphene materials, I was tricked by the vivid details and thought they were SEM or AFM images", says Huang, "it's a simple and dirt-cheap method that works surprisingly well."
Graphene is a mono-layer of carbon atoms that resembles nano-chicken wire. Effectively it is unrolled carbon nanotube which is a candidate for combining fast nano-carbon semiconductors with the semiconductor industry's standard planar process.
MIT says the technology has the capability of increasing processor speeds by 100 times delivering practical systems in the 500 to 1,000 gigahertz range.
Key to the usefulness of graphene is the fact that electron mobility in graphene is 100 times faster than electron mobility in silicon.
Comments (1)
I hope that the power consumption of graphene circuits will also be 100x lower as well, otherwise the increased speed will be unusable in most cases.
Silicon CMOS is around 50W/cm2 right now which is difficult enough to manage. 5kW/cm2 would be slightly difficult to deal with, to say the least -- I doubt if even solid diamond heatsinks could cope with this :-)
Ian
Posted by Ian Dedic | January 14, 2010 11:23 AM
Posted on January 14, 2010 11:23