Single atom channel for transistor

[thilanliyan]

http://www.xtremesystems.org/forums/showthread.php?t=240479

 

[Idontcare]

Probably not what many people are thinking of in their mind when the phrase "single atom transistor" is used.

That phrase conjures up the idea that a single-atom physically resides between the source and the drain (and not two, or three, or twenty), whereas what these guys are reporting is that they made a channel (the active region between source and drain) comprising a single dopant/implant atom.

That dopant atom of course is embedded in a matrix of surrounding silicon atoms, so the channel forming the transistor involves far more than just a single atom residing between the source and drain.

 

[Shilohen]

Isn't that pretty much unimpressive? Not that it's not a feat, but I'm pretty sure I saw Intel proving that a 3-atom transistor was possible a couple of years ago. Sadly I cannot find the link so I might be living in my own reality. However, I did find this: http://www.eetimes.com/news/semi/showArticle.jhtml?articleID=10810046

Intel is expected to develop chips based on this process by 2009. By then, the company could have high-speed processors running at speeds of 20-GHz or faster, according to analysts.

 

[jvroig]

20GHz. Awesome find. :thumbsup: :thumbsup:

I hope they fired those analysts, by the way.

 

[Idontcare]

Isn't that pretty much unimpressive? Not that it's not a feat, but I'm pretty sure I saw Intel proving that a 3-atom transistor was possible a couple of years ago.

Essentially that is the meat of it. Its an academic exercise, and no doubt some value will come of it. We can't learn about the limits of the frontiers of semiconductor operation unless we build one-off devices that enable us to probe it experimentally.

At the same time there is a big difference between an academic exercise and reducing the experiment to a commercially viable productization pathway. Look at fusion, or splitting water to hydrogen, or the quantum computer...all can be done and it is great that researchers have figured out how to get the experiments functional but until the experiments are reduced to commercially viable products there is little "wow" factor involved for the lay person.

So then they (scientist, publisher, journalist, someone) needlessly hype up the prospects of the technological advance by making outlandish claims like "with this technology you could have 20GHz computers in 3-5yrs! and cure cancer while talking to aliens!".

That's where the value of the scientific breakthru is lost in translation and the masses end up with an artificially inflated sense of expectation...fast forward 10yrs and expectations aren't met by reality and now you have depressed/jaded/cynical masses.

How's your flying car working out for you? Taking a trip to the moonbase for the family vacation this year? Gonna power your car on sunshine and water alone this year?

 

[Shilohen]

I think I was unclear, or rather semantically imprecise in my claim. When I say proved it, I meant demonstrated it. As far as I can remember it was the same year that IBM showed a plate of silicon with "IBM" carved in it using some really low amount of atoms with a tunnel effect microscope... Of course I cannot find that damned link either. Anyone else can concur or was my brain hit by a random tachyon from a future CES?

How's your flying car working out for you? Taking a trip to the moonbase for the family vacation this year? Gonna power your car on sunshine and water alone this year?

Oh god, I don't know if that makes me old of young, but I remember when I was young and the flying car was predicted for the 2000 and trips to moon base around 2010 I think.

EDIT: Ok, I'm not completely mad, I found http://everything2.com/title/atom

Two recent types of microscope, the Microscope Electronique à Transmission (or MET) and the tunnel effect microscope have allowed scientists to actually view individual atoms, making them no longer completely invisible. A special version of the tunnel effect microscope, the atomic force microscope, even allows scientists to deposit and move individual atoms - IBM recently used this to spell out their company name with individual Xenon atoms.

EDIT2: Finally found the IBM link http://www.ieeeghn.org/wiki/index.php/Don_Eigler, the Intel one remains

 

[TuxDave]

Probably not what many people are thinking of in their mind when the phrase "single atom transistor" is used.

That phrase conjures up the idea that a single-atom physically resides between the source and the drain (and not two, or three, or twenty), whereas what these guys are reporting is that they made a channel (the active region between source and drain) comprising a single dopant/implant atom.

That dopant atom of course is embedded in a matrix of surrounding silicon atoms, so the channel forming the transistor involves far more than just a single atom residing between the source and drain.

I was imagining that somehow someone found a way for a single atom to act as the entire transistor and pass electrons in and out of its valance bonds depending on the state of its neighboring atom. Your interpretation was my second guess.

 

[Shilohen]

And here's the Intel link: http://www.hoise.com/primeur/01/articles/weekly/AE-PR-07-01-50.html

One component of the transistor is only three atoms wide.

Of course it's not very precise and does not describe which part of the transistor they're talking about

 

[Shilohen]

I was imagining that somehow someone found a way for a single atom to act as the entire transistor and pass electrons in and out of its valance bonds depending on the state of its neighboring atom.

Now that would be a great breakthrough and nice bridge between modern and quantum computing, although I find that somehow unlikely.

 

[Idontcare]

I was imagining that somehow someone found a way for a single atom to act as the entire transistor and pass electrons in and out of its valance bonds depending on the state of its neighboring atom. Your interpretation was my second guess.

In a sense, at least to the degree that quantum physics allows us to envision atoms existing in a "solid" material, I suppose that is what molecule-based transistors are supposed to be about.

The world of possibilities opens up when you expand the options of transistor-equivalent devices to something beyond your traditional semiconductor materials.

The downside I suppose is that the switching speeds of most the competing transistor alternatives are markedly lower than current device operating speeds. It is difficult to compete with sub-picosecond switching times generated by tens if not hundreds of billions of dollars of R&D research to date invested into optimizing solid-state xtors over the past 60 yrs.