Your news round-up from the big world of the
incredibly small. Feb 2002 edition 
It's
been a while since the last Smalltalk, so I've put together a roundup
of the best stuff from the last 6 months in the world of the small
and weird. Some big nanonews that the popular press picked up on last
August was that remarkable plastic sculpture of a bull the size of a
red blood cell. While small, it's still about 100 times too big to be
called nanotechnology but this doesn't mean it's a load of bull.
Japanese
researcher Professor
Satoshi Kawata showed that he can produce mechanically useful 3D
objects like springs too, so we can now test complex little objects
to see how they behave. Things made from silicon can be made smaller,
but the silicon process is like stamping things out of steel and only
being able to stack and weld 5 layers total thickness. The 3D
capability we've just acquired will fit in nicely with the way that
nanotechnology is going: Building real structures rather than the
thin, flat circuits of today's microworld.
Small tubes were also big in August, as professor Reza Chadiri and his colleagues at The Skaggs Institute for Chemical Biology, a part of the renowned Scripps Research Institute discovered what some consider to be the first practical application of nanotechnology in medicine. They created "cyclic peptide nanotubes", which needs a bit of explaining and translation into English. Them things are molecules shaped like rings. The rings have chemical groups on them that make them stick to particular bits of other rings, and also to specific bits of a target organism's cells. So when they meet a target cell, they flock to the same place and stack up like a packet of lifesaver mints with all the holes aligned. This forms a kind of miniscule hypodermic needle stuck in the cell, and all the contents ooze out. If the cell is an anthrax bug, it is now a dead anthrax bug. This is a novel kind of antibiotic that bugs will find it hard to become resistant to.
Many laboratories are now experimenting with ways of producing electronics using just a few atoms per component. Hewlett Packard ("HP") are very active, attacking the problem on many fronts. They have already demonstrated a working memory made up of a few molecules per memory cell, but it only holds a couple of bytes. They'll improve on this, no doubt, and plan to store about 16,000 bytes in a memory chip the size of a bacterium. Strung together in bulk, these can make far bigger memories than we currently have and use less power into the bargain.
But
where HP really excel is in making the computer chips themselves.
They have now figured out - and patented - ways to make molecules
that behave like computer logic, the tiny little grids of lines
needed to hook them up, and recently a method of breaking the grid up
in strategic places so they can make computers out of it all. The
image on the left shows how they trap a layer of molecules in a grid
'sandwich'. Miraculously, the parts build themselves given the right
conditions. They've not got the breaking up bit working on a small
scale yet, and the rest of it is still pretty crude, but they're
reckoning on showing something that works in 2005. To make it work,
they have developed a way of designing a computer processor chip for
each grid, taking into account any defects that the grid might have.
So even if they can't make the grids 100% operational, they can still
produce computers that work perfectly. As all the electronics
self-assembles and automatically works around dud parts, they might
be able to do it more cheaply than anyone currently imagines.
One
thing you'll be seeing this year is the so-called "3D"
memories from Matrix
Semiconductor Inc., which are a new twist on the silicon chip.
Ordinary boring chips are limited in the number of layers of silicon
that can be piled on top of one another as per the bull article
above. The silicon surface goes kinda lumpy as it gets higher, making
it difficult to get the upper layers in the right place - so most
chips are flat. 5 layers is the limit, but IBM discovered that they
could literally polish the last layer down flat, and start over again
with a new layer. Matrix took this a step further and can put down
six or more layers of completed circuitry on the one piece of
silicon, shown here on the right in their press photo. As the cost of
making and packaging the purified silicon base is most of the cost of
the chip, this means six or more times as much electronics can be
piled on for pretty much the same cost.
Matrix see no reason why they shouldn't make chips with ten or even more layers at similar cost, and that's what they're going to make for the French company Tompson. But as initial memories will only be able to be written to once, the idea is to get the cost down so low that these memories are disposable. In digital cameras, it would compete with film on price. With MP3 players it would compete with audio cassette tape. Digital quality, no moving parts, and not many dollars - it could revolutionise consumer electronics.
One new toy that I can't wait to see is Microvision's latest display. This is one of those things that projects an image into your eye, and a screen appears in front of you as if by magic. Microvision have managed to get the technology refined to the point where they use three LEDs to provide red, green and blue light, and a microscopic mirror to flick the light around. The whole thing could mount in your sunglasses frame or the side of your cellphone when the finished article rolls off the production line sometime around 2004. OK, so it might make you look slightly scary with a glowing eye and all, but I can't wait to try one!
Dr.
Nadrian Seeman continues his amazing work, creating the molecular
equivalent of a four-stroke engine, entirely built from and fuelled
by DNA. While it can go round and round continuously, it has to be
"stepped" a quarter turn at a time, which requires it to be
flushed with the right sequence of DNA. By flushing it with different
sorts of DNA, it can be made to run forward, in reverse, or to step
forward in lesser or greater amounts.
It is possible that in the relatively near future, tiny little molecular motors based on this technology will both power and control generations of molecular machines. Exactly what these machines will be doing we can only but guess at.
http://olliver.family.gen.nz/launchpad 24th February 2002 vik@olliver.family.gen.nz
Back to the Launchpad Main Page