Scientists Strive to Replace Silicon with Graphene on Nanocircuitry

Jun 10, 2010
In a technique known as thermochemical nanolithography, the tip of an atomic force microscope uses heat to turn graphene oxide into reduced graphene oxide, a substance that can be used to produce nanocircuits and nanowires with controllable conductivity. Image Credit: University of Illinois at Urbana-Champaign

(PhysOrg.com) -- Scientists have made a breakthrough toward creating nanocircuitry on graphene, widely regarded as the most promising candidate to replace silicon as the building block of transistors. They have devised a simple and quick one-step process for creating nanowires, tuning the electronic properties of reduced graphene oxide and thereby allowing it to switch from being an insulating material to a conducting material.

The technique works with multiple forms of and is poised to become an important finding for the development of graphene electronics. The research appears in the June 11, 2010, issue of the journal Science.

Scientists who work with nanocircuits are enthusiastic about graphene because meet with less resistance when they travel along graphene compared to silicon and because today's silicon transistors are nearly as small as allowed by the laws of physics. Graphene also has the edge due to its thickness - it's a carbon sheet that is a single atom thick. While graphene nanoelectronics could be faster and consume less power than silicon, no one knew how to produce graphene nanostructures on such a reproducible or scalable method. That is until now.

"We've shown that by locally heating insulating graphene oxide, both the flakes and epitaxial varieties, with an tip, we can write nanowires with dimensions down to 12 . And we can tune their electronic properties to be up to four orders of magnitude more conductive. We've seen no sign of tip wear or sample tearing," said Elisa Riedo, associate professor in the School of Physics at the Georgia Institute of Technology.

On the macroscale, the of graphene oxide can be changed from an to a more conductive graphene-like material using large furnaces. Now, the research team used TCNL to increase the temperature of reduced graphene oxide at the , so they can draw graphene-like nanocircuits. They found that when it reached 130 degrees Celsius, the reduced graphene oxide began to become more conductive.

"So the beauty of this is that we've devised a simple, robust and reproducible technique that enables us to change an insulating sample into a conducting nanowire. These properties are the hallmark of a productive technology," said Paul Sheehan, head of the Surface Nanoscience and Sensor Technology Section at the Naval Research Laboratory in Washington, D.C.

The research team tested two types of graphene oxide - one made from carbide, the other with graphite powder.

"I think there are three things about this study that make it stand out," said William P. King, associate professor in the Mechanical Science and Engineering department at the University of Illinois at Urbana-Champaign. "First, is that the entire process happens in one step. You go from insulating graphene oxide to a functional electronic material by simply applying a nano-heater. Second, we think that any type of graphene will behave this way. Third, the writing is an extremely fast technique. These can be synthesized at such a high rate that the approach could be very useful for engineers who want to make nanocircuits."

"This project is an excellent example of the new technologies that epitaxial graphene electronics enables," said Walt de Heer, Regent's Professor in Georgia Tech's School of Physics and the original proponent of epitaxial graphene in electronics. His study led to the establishment of the Materials Research Science and Engineering Center two years ago. "The simple conversion from graphene oxide to graphene is an important and fast method to produce conducting wires. This method can be used not only for flexible electronics, but it is possible, sometime in the future, that the bio-compatible graphene wires can be used to measure electrical signals from single biological cells."

Explore further: Tiny graphene drum could form future quantum memory

Provided by Georgia Institute of Technology

4.8 /5 (32 votes)

Related Stories

A huge step toward mass production of graphene

Mar 10, 2010

Scientists have leaped over a major hurdle in efforts to begin commercial production of a form of carbon that could rival silicon in its potential for revolutionizing electronics devices ranging from supercomputers ...

Light-speed nanotech: Controlling the nature of graphene

Jan 21, 2009

Researchers at Rensselaer Polytechnic Institute have discovered a new method for controlling the nature of graphene, bringing academia and industry potentially one step closer to realizing the mass production ...

A 'huge step' toward mass production of graphene

Jun 09, 2010

Scientists have leaped over a major hurdle in efforts to begin commercial production of a form of carbon that could rival silicon in its potential for revolutionizing electronics devices ranging from supercomputers ...

Unzipping Carbon Nanotubes Can Make Graphene Ribbons

Apr 20, 2009

(PhysOrg.com) -- By "unzipping" carbon nanotubes, researchers have shown how to make flat graphene ribbons. Graphene, which is a one-atom-thick sheet of carbon that looks like chicken wire, has unique electrical ...

Recommended for you

Tiny graphene drum could form future quantum memory

Aug 28, 2014

Scientists from TU Delft's Kavli Institute of Nanoscience have demonstrated that they can detect extremely small changes in position and forces on very small drums of graphene. Graphene drums have great potential ...

Graphene reinvents the future

Aug 27, 2014

For many scientists, the discovery of one-atom-thick sheets of graphene is hugely significant, something with the potential to affect just about every aspect of human activity and endeavour.

User comments : 18

Adjust slider to filter visible comments by rank

Display comments: newest first

Quantum_Conundrum
Jun 10, 2010
This comment has been removed by a moderator.
solar2030
Jun 10, 2010
This comment has been removed by a moderator.
Skeptic_Heretic
not rated yet Jun 10, 2010
Not to continue the 1-up game but in reality the theoretical application is 1 ZHz transition speed.
Quantum_Conundrum
not rated yet Jun 10, 2010
Not to continue the 1-up game but in reality the theoretical application is 1 ZHz transition speed.


Can you provide a link to something official that states that? The highest I've seen claimed was around 100ghz transistor with about a 10ghz core speed for a processor.

If what you have said is true, we'd be looking at computers in the very near future which are as big an improvement over modern top of the line 6-core processors as those are vs counting on your fingers...

One zeta-flops would make a single processor exponentially more powerful than the most advanced supercomputer in the world today.
PinkElephant
5 / 5 (1) Jun 10, 2010
This is very cool (no pun intended), but they'll need to figure out some other way of making the circuits. Literally sketching out a few billion transistors and connections between them with what is essentially a nanoscopic hot toothpick, wouldn't exactly amount to a practical mass-production technique.

Maybe they could combine lithography with this substrate? E.g. deposit a layer of something that preferentially reflects a frequency of light that's efficiently absorbed by graphene oxide. Etch out a circuit on that layer. Then shine bright light of the right frequency, and you end up heating only the exposed parts of graphene oxide...

Or, instead of heat come up with a chemical method to transform graphene oxide to graphene, and use still more conventional lithographic etching techniques...
Skeptic_Heretic
not rated yet Jun 10, 2010
Not to continue the 1-up game but in reality the theoretical application is 1 ZHz transition speed.


Can you provide a link to something official that states that?
No but I can show you the math for it within the current theory.

The reason why silicon isn't as advantageous as carbon is that due to the chemical properties of carbon it will withstand a far greater(about a thousand fold) electrical current before you reach terminal resistance. It's lengthy math and physorg doesn't handle simple arithmatic symbols. If you want to PM me your email address I can walk you through it.
maxcypher
not rated yet Jun 10, 2010
The article used the acronym, TCNL. Can someone expand that for me?
Skeptic_Heretic
not rated yet Jun 10, 2010
TCNL
T.C.N.L
T..C..N..L
T ... C ... N ... L

Ok joke over, Thermochemical Nanolithography.
maxcypher
not rated yet Jun 10, 2010
The quote, "bio-compatible graphene wires can be used to measure electrical signals from single biological cells", has amazing implications for brain augmentation, I think.
Quantum_Conundrum
not rated yet Jun 10, 2010
Maxcypher:

brain augmentation, prosthetics, artificial ears, eyes, organs, you name it.

Graphene also makes the "self replicating robot" concept much more "eco-friendly" and low energy than a metal and silica based robot. (thinking Mars here.)
maxcypher
5 / 5 (1) Jun 10, 2010
Frankly, I consider all those other enhancements to be essentially brain augmentation, but I get your point. Your "self replicating robot" sounds like the perfect place to transmit my so-called mind into a new body. BTW, thank you Sk. Her. for the joke and info.
Sonhouse
not rated yet Jun 11, 2010
The article used the acronym, TCNL. Can someone expand that for me?

Tonight Can Not Last:)
LivaN
not rated yet Jun 11, 2010
Not to continue the 1-up game but in reality the theoretical application is 1 ZHz transition speed.


What would that make the average core speed?
Skeptic_Heretic
not rated yet Jun 11, 2010
Not to continue the 1-up game but in reality the theoretical application is 1 ZHz transition speed.


What would that make the average core speed?

It would depend on the geometry of the processor and transistor total count.

Still writing it out QC.
dirk_bruere
not rated yet Jun 11, 2010
If you want real brain augmentation, then check out this - its amazing!
http://viterbi.us...team.htm
DaveGee
not rated yet Jun 11, 2010
Question.. Comparing this to todays chip manufacturing technologies does this have anywhere to go?

What I mean is... thu various methods chip manufacturers have been able to etch with smaller and smaller methods ... in turn allowing for a greater number of components to be contained within the same patch of silicon.

With this new tech once perfected, do we have anywhere to go in terms of size? I imagine we aren't going to see the same level of 'writing smaller' as we saw with yesterdays methods of etching chips. However I also imagine this would be a great leap (shrink) when compared with traditional 65nm processes.
PinkElephant
5 / 5 (1) Jun 11, 2010
@DaveGee,

Modern processes (as represented by Intel) are already at 32 nm, and headed toward 22 nm starting next year.

Ultimately, you can't go less than 1 atom per circuit element, but long before that you hit the wall of quantum noise. Somewhere in that regime, we switch from bulk material to molecular computing. In terms of molecular computing, then, carbon-based molecules sound as flexible and promising, as they are for life in general :-)

Once you're down to molecule-sized circuit elements, the only way to keep increasing complexity is to start layering your circuits (or in other words, constructing them in 3D rather than 2D.) Of course, with that approach the problem of heat management becomes vastly more difficult. Carbon materials like graphene, nanotubes, and diamond just happen to have both excellent electronic, and excellent heat conducting properties (so the circuit can become its own heat pipe...)
Quantum_Conundrum
5 / 5 (1) Jun 11, 2010
Given the possibility of optical circuitry being developed in other branches of nanotechnology, 3D computer circuitry could have amazing bus capabilities, because you can aim multiple lasers across one another in the same space, or even across the same point, without destroying the data, which is something you cannot do with wires. So these technologies all taken together could allow all new concepts in data storage and retrieval, as well as all new forms of parallel processing.

--=

Regarding Intel, from what I've seen, their latest processors actually have far more "cores" than the main "six" that it is advertized as. They are almost like an entire motherboard integrated onto one chip. They now have video pre-processors, and L2 and L3 caches which exceed the entire RAM and video card of a win95 era PC...all smashed into a single chip, while what we now call "video cards" are up to 480 processor cores at ~350mhz each (~480 pentium 2 machines on one chip), and 1gig of RAM.
Quantum_Conundrum
not rated yet Jun 11, 2010
So the top of the line Nvidia video card you buy today has as much processing power as four hundred eighty pentium 2 era computers combined, though you gotta figure some loss because a few of those cores do nothing but tell other cores what to do, and you also have to be fair and remember they are specialize processors that are specifically designed to handle video and not much else.
Quantum_Conundrum
not rated yet Jun 11, 2010
An idea I just had for video cards is this...

(requires motherboard manufacturers to standardize the distance between PCI express ports.)

Since the video cards often "kill" an extra port due to the size of the cooling unit, wouldn't it make sense for the manufacturer to just go ahead and design a "3d" card which plugs into and uses both PCI Express ports to begin with?

Not sure what they are doing right now anyway, I haven't bought a new video card in several years...