[extropy-chat] Re: NANOTECH: Lasers can help

Chris Phoenix cphoenix at CRNano.org
Fri Apr 23 03:50:34 UTC 2004

Robert J. Bradbury wrote:
> The Univ. of Michigan has announced [1] the ability to do nanoscale milling
> (down to 20nm) using ultra-short laser pulses.  

I knew that intense light could interact non-linearly with materials, 
and had previously speculated (privately, alas) that this could be used 
to deposit energy in sub-wavelength regions.  But this still made my jaw 
drop; it even took me a while to connect this with my predictions.

We're just starting to learn that light and electrons can be made to 
dance in very small spaces.  Michal Lipson has done some amazing work 
with 50-nm light guides, including coupling them to normal fiber optics 
at high efficiency and low alignment sensitivity. 

> So a top-down milling machine methodology might allow
> the production of a scaled-up nanomanipulator arm which in turn could
> be used to produce increasingly smaller versions of itself.

I have speculated along these lines for a while.  Note that EBID 
(electron beam induced deposition) can be used to make 3D structures 
with 20 nm feature sizes.  (And there's a group in Denmark(?) that has 
already integrated EBID with piezo manipulators in an SEM.)  But I think 
this laser method allows you to use better materials; EBID can use lots 
of precursor gases, but as far as I've heard, the results are all 
amorphous or metallic-with-slag.  Hm... diamond is a dielectric; can 
this laser milling produce diamond shapes?  Can it leave features 
smaller than 20 nm between 20-nm holes?

I don't think you want "increasingly smaller" versions.  Just take 
whatever arm you have, and go straight to mechanosynthesis.

A possibly interesting exercise, inspired by a comment from Eric Drexler 
when I suggested something like this to him:  Imagine you can make 
arbitrary shapes with a technology like this.  Design your 
proto-assembler that can do mechanosynthesis.  Then ask: What 
functionality do you need that macro-scale piezos don't have?

One further note, before our brains recover their normal certainty about 
the limits of physics:

Google still doesn't contain the phrase "plasmon logic."  I predict that 
within five years, plasmon logic will be a very exciting field of study. 
  Now that we know that photons can be converted to plasmons and back, 
we can make the plasmons jump through all sorts of hoops.  We can affect 
them with electric fields and with nanoscale arrangements of matter.  We 
can pause them, shunting them to and from holding cells (this has been 
demonstrated--another way of "freezing light."

And my intuition says we should be able to make plasmons interact with 
each other one-on-one, just as though they were fermions--or change 
conditions a bit, and let them pass through each other like bosons. 
(Prof. Lipson has built a 10 micron ring resonator out of her 50-nm 
waveguides, sensitive to the difference between 1583 and 1584 nm light.) 
  Essentially, this would be the basis for an optical computer, but 
implemented at the nanoscale, reconfigurable using nearby electrodes, 
and running on single photons that (my intuition further claims) can 
probably be coupled directly to electronics.


Chris Phoenix                                  cphoenix at CRNano.org
Director of Research
Center for Responsible Nanotechnology          http://CRNano.org

More information about the extropy-chat mailing list