On Fri, 17 Aug 2001 hal@finney.org wrote:
> Robert is right, but I don't think this small technical error
> invalidates Whitesides' argument. His lead-in sentence was "The
What is the argument? I don't have the URL.
> similarity between flagellar and electrical motors is, however,
> largely illusory." While he was wrong about ATP, his basic point was
> correct that unlike macro-scale motors, the flagellar motor is not
> magnetic. It is apparently electrostatic, and/or may rely on changes
> in shapes of proteins; the details are not known.
What's the argument? That you can't make NEMS motors? That's nonsense,
we've got MEMS motors already -- which happen to be electrostatic -- and
we've got linear actuators made from telescoping concentric buckys --
which -- gosh -- also happen to be electrostatic. Even Nanosystems already
said you need to go electrostatic, not electromagnetical, as that doesn't
scale.
So you can most assuredly make electrostatic motors with graphenes, and
the darn thing would have rpms in GHz range, if appropriately sized.
> Also, the proton gradient itself is presumably maintained by molecular
> pumps powered by ATP, so there is a sense in which Whitesides'
> statement can be interpreted as being correct.
Again, what is the argument?
> I think it would be relevant here to list the specific criticisms that
> Whitesides made based on this misconception of submarine size, in
> order to show that they would not be relevant. It seemed that the
> main point made by Whitesides with regards to the size of the
> submarine was the difficulty in navigation due to Brownian motion.
Well, bacteria manage to navigate nicely using flagellar propulsion within
animal cells, so I don't see why an array of electrostatic motors with
bucky or sapphire fibers sticking out wouldn't make for a rather speedy
nanosub.
> However he is wrong about the Brownian effects since the subs would be
> much larger than he assumes. Also, Freitas suggests that the nanobots
> would be able to travel at 1 cm/sec (100 times faster than the 100-200
> microns per second of sperm). This is fast enough to negotiate
> capillaries and small blood vessels without much difficulty.
You can of course stick and crawl, or burrow through tissue, too.
> Your references rely pretty heavily on Freitas. Another article I like
> is Merkle's hydrocarbon metabolism,
> http://www.zyvex.com/nanotech/hydroCarbonMetabolism.html, which is the
> best article I've seen to address the "grabbing atoms" issue. He shows
> specific molecular structures for reactive tools which can be used to
> add or remove hydrogen and carbon atoms, as well as to build more of
> the tools.
The problem with this is that you use reactive moieties attached to rigid
rods which imply bulk (which he sidesteps by showing them as quadratic
blocks), which have to touch down at a surface (one hemisphere of space is
hence unaccessible). Several of the moieties much be present at the site
simultaneously, which is difficult to steric hindrance which he refers to
as the fat fingers problem. Plus you have to cycle the rods, retracting,
recycling and protruding them again in a rapid oscillation, which implies
you have to minimize bulk, or be slow and burn a lot of juice.
This is unelegant. What would make more sense is a nanolithoprinter, using
a flexible hollow (e.g. graphene) duct to transport monomers and a head
with an aperture aligning them and activating them with a current pulse,
at least initially (when you've got a cluster bristling with radicals you
can just firehose your allenes at it, and they will stick). That may not
give you atomic control, but when do you really need atomic control? I see
no showstoppers even if you can deposit only a tiny subset of all possible
structures, as long as the repertoire is sufficiently rich.
Once you have got your computronium, you can burn ridiculous amounts of
simulation time to fit the optimal paradigm, and optimize it into
ridiculous dexterity, here you're only constrained by raw physics, and
local physics looks so far extremely friendly to hackers.
-- Eugen* Leitl <a href="http://www.lrz.de/~ui22204/">leitl</a>
______________________________________________________________
ICBMTO : N48 10'07'' E011 33'53'' http://www.lrz.de/~ui22204
57F9CFD3: ED90 0433 EB74 E4A9 537F CFF5 86E7 629B 57F9 CFD3
This archive was generated by hypermail 2b30 : Fri Oct 12 2001 - 14:40:10 MDT