Laser deposition & Semi-tech [was: near anything boxes]

From: Robert J. Bradbury (bradbury@aeiveos.com)
Date: Mon Mar 13 2000 - 05:52:00 MST


As Spike pointed out, my posting on Laser deposition got cut off.
Guess, I'm not clever enough to multiplex postings to the extropians
and Slashdot at the same time.... :-(.

At any rate, I'll try to complete the thought --

Spudboy100 wrote about the possibility of building nanotech via
laser deposition (his actual posting was about Laser Deposition
Spectroscopy & sorting matter by atomic weight which seem unrelated).

I don't think this is feasible. Normal light has wavelengths much
bigger than atomic scales. This is why current lithographic
methods are having to move into shorter wavelengths in the EUV,
X-ray, and eventually E-beam.
See:
  http://dailynews.yahoo.com/h/zd/20000302/tc/20000302132.html
or an older
  http://www.eetimes.com/story/eezine/OEG19990618S0014

There is technology in the pipeline [see Nanotechnology, G. Timp (ed.),
in Chapter 10: Atom Optics: Using Light to Position Atoms], but
the methods demonstrated are being used to draw lines, not 2-D
patterns. Now, perhaps by rotating 90 degrees and repeating
you might turn lines into squares, but the method seems limited
and is definitely not precise down to the small molecule/atomic
level.

Of interest when researching this comment, I came across the
fact that things really are getting faster. As:
  http://www.ebnews.com/story/OEG19990505S0009
and
  http://www.semibiznews.com/story/OEG19991122S0026
show, the manufacturers are *dropping* the entire 0.15-micron
chip generation and pushing the timetable to get us to 35 nm
(64 Si atoms wide) by 2014. Nikon expect to have the e-beam
systems in production by 2006 when they will be used to produce
16-gigabit DRAMs.

Not everything is surging ahead however. It looks like DRAMs
will give way to Microprocessors as the technology driver.
My guess would be that programmer productivity is insufficient
to write programs that require 16 gigabytes of memory!!!
Things will shift again however when the microprocessor
architectures move to processor-in-memory and you can put
dozens of them into desktop machines.

Real nanotech (atomic assembly), derived from top-down approaches,
would seem to be on track for circa 2010 at least in the lab.

Robert



This archive was generated by hypermail 2b29 : Thu Jul 27 2000 - 14:05:00 MDT