backtop


Print 15 comment(s) - last by hyvonen.. on Nov 30 at 7:32 AM

Nanowires may allow many more transistors to be placed on computer chips in the future

Engineers and researchers predict that in the next five to ten years the dimensions of silicon transistors will have been scaled to their limits and will be unable to get any smaller. Without a new breakthrough in creating smaller transistors, Moore's Law will fall flat.

A group of engineers and researchers working together from IBM, Purdue University, and the University of California at Los Angles has learned to create nanowires coated with materials that make for efficient transistors. The nanowires have very sharply defined layers at the atomic level that allow the wires to be efficient transistors.

Eric Stach, associate professor of materials Engineering from Purdue said, "Having sharply defined layers of materials enables you to improve and control the flow of electrons and to switch this flow on and off."

The team of researchers says that electronic devices are often constructed of heterostructures. The term heterostructures means that the structure contains sharply defined layers of different semiconducting materials like silicon and germanium. According to the researchers, the challenge in the past has been the capability of producing nanowires with the requisite defined layers.

The team has detailed its findings in a paper published in the November 27 edition of the journal Science. The transistors that the team have developed are not made on flat pieces of silicon. These nanowires are grown vertically making them have a much smaller footprint, which in turn allows for many more of the nanowires to be placed on the same piece of silicon.

Stach said, "But first we need to learn how to manufacture nanowires to exacting standards before industry can start using them to produce transistors."

The researchers used a transmission electron microscope to view the nanowire formation. The nanowires were formed by heating tiny particles of a gold-aluminum alloy in a vacuum chamber. After the alloy was melted the researchers introduced silicon gas and the alloy bead absorbed the gas becoming supersaturated with silicon. This caused a silicon wire to grow from the alloy bead producing a silicon wire that was topped with a mushroom-like gold-aluminum alloy bead.

At that point, the researchers reduced the temperature of the chamber enough to allow the alloy bead at the top of the wire to solidify, thereby allowing germanium to be deposited on the silicon precisely creating the required heterostructure needed to create a transistor. The heterostructure allows the formation of a germanium gate in each transistor allowing devices to switch on and off.

"The cycle could be repeated, switching the gases from germanium to silicon as desired to make specific types of heterostructures," Stach said.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Moore's Law keeps on truckin'...
By Boze on 11/27/2009 3:02:07 PM , Rating: 2
How in the world did that man so accurately predict computing power??




RE: Moore's Law keeps on truckin'...
By amanojaku on 11/27/2009 3:23:50 PM , Rating: 3
The same as any other credible scientist. He analyzed a data set, then made a theory. He reanalyzed a new data set and compared the results against the first and determined it was a law.

Later, he reanalyzed a new data set and found the results had different short term and long term periods, so he altered the law into its final form.

That's one of the things I like about science. It admits to the possibility of error so that it can facilitate a self-refinement process. If done correctly, you eventually get to perfection.

http://en.wikipedia.org/wiki/Moore%27s_law


RE: Moore's Law keeps on truckin'...
By Shig on 11/27/2009 3:48:59 PM , Rating: 2
/Salute Moore's Law

We haven't even begun making processors that take advantage of all three dimensions.


By Shining Arcanine on 11/28/2009 12:28:23 AM , Rating: 2
He made a conjecture, not a theory.


RE: Moore's Law keeps on truckin'...
By Motamid on 11/27/2009 11:24:55 PM , Rating: 3
I would call Moore's Law more of a self fulfilling prophecy than a Law. It started out as the observation of a trend which soon became the bar set by processor manufacturers to attain. In fact there have been points in time where this bar has been exceeded.


By AnnihilatorX on 11/28/2009 8:19:27 AM , Rating: 2
Easy
Growths from bacterial growth to technological advances are based on exponential growth. Moore's law is basically the exponential law of growth with a time factor of 18 months.


RE: Moore's Law keeps on truckin'...
By DOSGuy on 11/28/2009 3:07:08 PM , Rating: 2
It's a self-fulfilling prophecy. Gordon Moore is co-founder of Intel. At the time, Moore's law was an observation, but once "Moore's Law" was coined it became an expectation. Moore was in a position to establish Intel's policy on the matter, so it became a requirement. The law became self-fulfilling because Moore had the power to keep his law going.


Doesn't matter
By amanojaku on 11/27/2009 12:39:45 PM , Rating: 5
quote:
Nanowires may allow many more transistors to be placed on computer chips in the future
No matter how many transistors we pack into a chip... Adobe will still find a way to make Flash slow. ;-)




RE: Doesn't matter
By modus2 on 11/27/2009 2:17:56 PM , Rating: 1
The real issue being of course if it can run Crysis....


RE: Doesn't matter
By themaster08 on 11/30/2009 5:53:14 AM , Rating: 2
Crysis has nothin' on Flash.


RE: Doesn't matter
By AnnihilatorX on 11/28/2009 8:20:16 AM , Rating: 2
Heard they are accelerating flash with GPUs with Flash 10.1?


RE: Doesn't matter
By hyvonen on 11/30/2009 7:25:38 AM , Rating: 2
Correct.

But they should be accelerating flash with better coding instead.


Time we ditch focus on bulk Silicon
By AnnihilatorX on 11/28/2009 8:32:04 AM , Rating: 2
Bulk silicon has served us well for 50 years, superseded any expectation for it to fail. It has done well, but I think it's time to move on to exploring alternatives, like the Silicon-Germanium hybrid heterostructures, or even more exotic form of semiconductors.

Semiconductors with huge potential: higher charge mobility (higher switching speed >250Ghz) as well as ability to synthesize quantum structures (dots, well), such as Gallium arsenide, GaInAs; unfortunately fell out of development effort because of Silicon's abundance and affordability. It is used, especially on high speed optical circuits due to direct band gap nature and electron mobility. However, I still believe there are much higher potential that can be explored. Look at what we can make Silicon to do in 50 years.

I still believe single electron quantum dots to be the holy grail of transistors.




By Orac4prez on 11/30/2009 1:37:33 AM , Rating: 2
Silicon is cheap (comparatively) and the complexities of fabricating chips and the mechanisms for controlling manufacture, stopping current leakage, heat dissipation, etc have been developed over time. The level of sophistication and knowledge to produce these high quality chips cannot be underestimated. Many of the solutions are specific for silicon and could not be readily used for other materials. Sadly, it is not possible to manufacture novel chips in an economic manner yet with any other material. There is a lot of research going on on alternative technologies and some of these will undoubtedly be utilised for specialised applications. Intergating these technologies into existing chips and devices will be attended by a range of engineering, chemical and physical problems which need to be solved.

Just because silicon has been around for a long time is no reason to drop development of silicon devices. There will be many more iterations of the current technologies before anything else comes along. There are many people currently actively looking at silicon replacements, but things like quantum dots are still at the scientific curiosity stage.


By hyvonen on 11/30/2009 7:32:14 AM , Rating: 2
quote:
...fell out of development effort because of Silicon's abundance and affordability.


And that's the key. Research on other options is being done in universities and corporate research labs, but a total overhaul of process technology is a tad expensive - as long as bulk silicon can be "tweaked" (strain, HKMG etc.) economically to enable continued scaling of size, cost and performance, that's what companies will do.


"So if you want to save the planet, feel free to drive your Hummer. Just avoid the drive thru line at McDonalds." -- Michael Asher











botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki