11-11-2007, 06:14 PM
You may want to google hafnium as Intel has unleashed 45nm sized transistors based on hafnium. See: Sam
http://www.intel.com/pressroom/archive/releases/20071111comp.htm?iid=pr1_releasepri_20071111m
You may want to google hafnium as Intel has unleashed 45nm sized transistors based on hafnium. See: Sam
http://www.intel.com/pressroom/archive/releases/20071111comp.htm?iid=pr1_releasepri_20071111m
It's 'Hafnium-Based High-k Metal Gate Transistors' but not 'Hafnium based transistor'.
Yeah, I'd think you'd have to dope hafnium with tons of semiconductor to make a transistor out of it! :)
I had little information when I posted. Sam
That's why for the next generation they're moving to wholenium.
Edited: 12 Nov 2007, 3:16 p.m.
Sounds like the kind of bread I'm supposed to be eating these days...
Not to belabor the point nor to lecture anyone, but there can be four ways of looking at materials- metals, semimetals, semiinsulators, and insulators. Hafnium is a relatively obscure material, even among chemists, even still among materials chemists, and so most educated technical professionals would not know much about it. I only knew that it was a metal, and nothing about its conductivity, reactivities, nor mechanical properties.
I am now curious, however, why hafnium gates would be superior to gold or platinum or some such rather inert but highly conductive metal. I suspect it may have to do with the involvement or lack thereof of hafnium with regard to the formation of trap states, or even barrier formation, if a few atoms, say 10^12 or 10^14 or so, get loose and wend their ways into the semiconductor bulk regions near the interface.
Today, I listened to someone speak about tantalum catalytic chemistry. While not nearly as ignored as Hf, Ta is not really a common object of chemists or materials people's attentions. That's twice in one day for me with obscure metals...
... I now wonder if they'll be the next big thing, like silicon was initially, gallium arsenide at one point, the Atkins' diet a little while ago, or reality TV shows. 8)
IIRC, Hafnium is used in the tips of cutting torches.
I suggest not running Hafnium based gated xistors too hot!
B^)
Ren
dona nobis pacem
I thought two hafnium gates (arranged as a complimentary pair) are regarded a wholenium device.
:-)
Edited: 13 Nov 2007, 1:23 p.m.
The hafnium is used in the form of hafnium oxide and is used as the insulator between the FET gate and transistor body. It's acting as an insulating dielectric in this case, replacing silicon dioxide that's been used for 40 years. Why? Because we've gotten down to five atomic layers with silicon dioxide and that's just too thin and causes too much leakage, not to mention the issues with trying to maintain a uniform film of five atomic layers. One atom either way makes a 20% difference in device characteristics. Hafnium oxide has a higher dielectric constant, so it goes on three times thicker and is easier to control, resulting in less device variability.
Quote:
The hafnium is used in the form of hafnium oxide and is used as the insulator between the FET gate and transistor body. It's acting as an insulating dielectric in this case, replacing silicon dioxide that's been used for 40 years. Why? Because we've gotten down to five atomic layers with silicon dioxide and that's just too thin and causes too much leakage, not to mention the issues with trying to maintain a uniform film of five atomic layers. One atom either way makes a 20% difference in device characteristics. Hafnium oxide has a higher dielectric constant, so it goes on three times thicker and is easier to control, resulting in less device variability.
Thank you for this clear and succinct explanation.
...bt
Quote:
Not to belabor the point nor to lecture anyone, but there can be four ways of looking at materials- metals, semimetals, semiinsulators, and insulators.
"Everything can be a fuse"
(c) Jeff Bronfield, HHC2007
:-)
Ditto!
(Thanks for explaining the use of Half Knee Yum Ox Hide in xistors)
Ren
dona nobis pacem
Heeheehee...
... funny you should bring this up:
I was reading Digital Electronics Troubleshooting by Joseph Carr last night (one of the old TAB books) and on one page about protecting CMOS circuits while working on them, the author mentions that they suffer breakdown easily and that the worker ought to work on and ground himself to a sheet of Al foil or some such equivalent thing.
But then he says that that too is dangerous, if only to the person. Unless he puts a 1k ohm resistor in line between himself and the metal sheet, he might become a 5000 W fuse!
(Only 5000 W?!!)