Comments on: Feynman’s Path to Nanotech (part 4) http://www.foresight.org/nanodot/?p=3163 examining transformative technology Wed, 03 Apr 2013 18:23:47 +0000 hourly 1 http://wordpress.org/?v=3.0.4 By: J. Storrs Hall http://www.foresight.org/nanodot/?p=3163#comment-859515 J. Storrs Hall Sat, 11 Jul 2009 02:04:33 +0000 http://www.foresight.org/nanodot/?p=3163#comment-859515 Whether stiction proves to be a major concern depends on a lot of details. One major one is does the material you're working in get smooth with wear, or undergo major erosion? Good tolerance (actually, low surface roughness) makes the friction go down but the stickiness is always there. The key is making surfaces slippery enough to slide even with the stickiness. A 10^-4 tolerance should make standard designs work, e.g. wheels with axles and bearings. Bearings can run dry as far as heat is concerned but not if wear is an issue. Watches etc use jeweled bearings, which just means sleeve bearings in hard covalent crystal solids. It seems likely we could make spheres by various methods and use them in ball bearings. The key is being able to pick them up and put them in the race! Whether stiction proves to be a major concern depends on a lot of details. One major one is does the material you’re working in get smooth with wear, or undergo major erosion? Good tolerance (actually, low surface roughness) makes the friction go down but the stickiness is always there. The key is making surfaces slippery enough to slide even with the stickiness. A 10^-4 tolerance should make standard designs work, e.g. wheels with axles and bearings. Bearings can run dry as far as heat is concerned but not if wear is an issue.
Watches etc use jeweled bearings, which just means sleeve bearings in hard covalent crystal solids. It seems likely we could make spheres by various methods and use them in ball bearings. The key is being able to pick them up and put them in the race!

]]> By: Michael Kuntzman http://www.foresight.org/nanodot/?p=3163#comment-859512 Michael Kuntzman Thu, 09 Jul 2009 18:56:10 +0000 http://www.foresight.org/nanodot/?p=3163#comment-859512 Right now photolithography is at the 45nm node, and the next nodes are under development. Assuming a tolerance of 10^-1 or better, that makes errors on the order of 5nm or less. This will obviously improve as time goes by, at Moore's law rates. For a 500um device, where the typical part features would probably be on the order of 5-50 um, that gives us tolerances of 10^-3 - 10^-4 or better, even without surface improvement. So would I be correct to expect that for a 500um device, stiction should not be a major concern? Right now photolithography is at the 45nm node, and the next nodes are under development. Assuming a tolerance of 10^-1 or better, that makes errors on the order of 5nm or less. This will obviously improve as time goes by, at Moore’s law rates.

For a 500um device, where the typical part features would probably be on the order of 5-50 um, that gives us tolerances of 10^-3 – 10^-4 or better, even without surface improvement.

So would I be correct to expect that for a 500um device, stiction should not be a major concern?

]]>