Nanotechnology Gets Finer

An anonymous reader writes "ZDNet reports on a new level of detail found in nanotech construction." From the article: "Japan's NEC Electronics has developed a technology to make advanced microchips with circuitry width of 55 nanometers, or billionths of a meter, the Nihon Keizai Shimbun business daily reported Sunday. Finer circuitry decreases the size of a chip and cuts per-unit production costs. It also helps chips process data faster."

Will "top down" beat "bottom up"?

[janneH]

Bottom up construction has been a central tenet in some parts of the nanotechnology community. The idea that putting things together by controlling the position of individual atoms/molecules during fabrication will allow enormous breakthroughs in computing and other fields. But at least in the silicon based semiconductor business, the top down approach keeps marching mercilessly toward the bottom. This while bottom up synthesis/fabrication is still stuck at proof of concept. Might "top down" make it to the bottom - before the "bottom up" makes it to the top?

Re:Nanotechnology?

[GroeFaZ]

The term has, over the last years, become something of a catch-all phrase for all things below 100 nm, also including fairly ordinary chemistry, unfortunately. Originally, the term was invented by Norio Taniguchi, but broadly popularized by Eric Drexler with the famous book "Engines of Creation" (available for free as in beer at http://www.foresight.org/EOC/index.html [foresight.org]). "Engines" was over the top in some respects and often criticized, but even ardent opponents of Drexler's vision of nanotech like the recently deceased Richard Smalley admit they have been brought into nanotechnology by this very book. Back in the days of "Engines", nanotechnology was strictly confined to the not yet developed "mass-manufacturing of devices to atomic precision and specification".

Note that Drexler himself has presumably ceded the term to its current usage and has called Intel's 90nm chips "nanotechnology", although it bears no resemblence whatsoever to Engines-style nanotech. He prefers "zetatech" (mega, tera, peta, exa, zeta) nowadays because of the quantity of atoms involved, but I think it's rarely used. Molecular Manufacturing is the preferred term for what used to be Nanotechnology. Let's see how many more rearguard action Nanotechnology has yet to fight before it becomes reality at last.

Re:how small is a nanometer?

[GroeFaZ]

I prefer an analogy I came up with for myself, being sick of all the "width of a hair" anal-ogies I so often read. Maybe it's just as useless, because in one or the other direction, you'll always have to face distances that are far from what is important in everyday life. Ok, here it goes:

The moon [wikipedia.org] has a minimum distance to Earth of around 360.000 km.
The International Space Station [wikipedia.org] has a minimum orbit to Earth of around 350 km.
The pillars of the Millau Viaduct [wikipedia.org] are 340 meters tall.
If we take the minimum distance to the moon as our reference meter, then the ISS would orbit Earth at around 1 millimeter, the mentioned bridge would have pillars of slightly less than 1 micrometer, and finally a ruler [wikipedia.org] of 35 centimeter length or (a little less than) the circumference of a compact disc [wikipedia.org] would be 1 nanometer.

Re:Will "top down" beat "bottom up"?

[GroeFaZ]

I think conventional silicon semiconductors might never see bottom-up fabrication, for a couple of reasons:
a) There is too much money invested in the traditional top-down process, and
b) the industry will not abandon a proven concept for at best marginal improvements in a dying technology. As we know, silicone is doomed to fail as keeper of Moore's Law, because you can only reduce features to so such and such dimensions before tunneling effects kick in, heat ablation becomes an insurmountable problem, and the statistics of impurity induction fails in practice. These limits are hard-coded in the laws of Physics as we understand them, and cannot easily, as of today, be engineered around, if at all.
c) Silicone and especially silicone in semiconductors (thus including statistical impurities of other elements) is not a rigid, defined atomic grid, which is pretty much a requirement for a bottom up fabrication. Bottom up directs every atom or molecule to a specific, well-known place where it then remains, which simply doesn't apply in a material that's almost a liquid, constantly rearranging its atomic structure, especially at temperatures of a working CPU.

Of course there are other materials that could be used as semiconductors, like diamond, which will make a far superior material in every respect. But as long as there is so much money in silicone and as long as diamond wafer fabrication remains in its infancy, silicone will be the way to go. But eventually, the semiconductor industry will have to make the jump to diamond or some other material, to maintain Moore's Law of transistor density.

Picometer or smaller???

[Anonymous Coward]

Picometer or smaller???

Atom-atom spacing is on the order of angstroms (.1 nm). 100 picometers is an angstrom. In other words, with the current chemistry we can do today, we _are_ at the bottom.

The interesting goal we now face is not getting smaller, but getting bigger-- being able to exert order on larger and larger scales in interesting ways, i.e. self-assembly of these units into larger, more complicated devices.

Re:Plenty of Room at the Bottom

[Doc Ruby]

One approach to nanotube quality control is to make them cheap and dirty, then separate them chemically or mechanically (centrifuge, phoresis etc). Especially with different electromagnetic properties by which to separate them. Doping nanotubes for different chirality, especially heterogenous chirality in a single tube surface, is one of the more compelling avenues for nanocomputing research. Tubes a few dozen nanometers in diameter and dozens of centimeters long (10K:1 ratio), which is a pretty long wire. Solution processing is yielding plenty of results for nanotubes [google.com], and the "fundamental production" problems you predict don't even prohibit Si/DNA coupling techniques (another nanotechnique).

I don't know why you're so pessimistic. Even if those avenues were hitting real obstacles, or faced implicit physical contradictions, the field is extremely young. Especially in shrinking engineering, even small gains create new tools which enable breakthroughs. The actual limits to microengineering we now face, heat dissipation, parallelization, silicon featuresize and others, are the reason nanoengineering is seeing so much investment. We are already seeing nanoRAM announcements [slashdot.org] even here on Slashdot, and even today we saw buckyball films [digitalworldtokyo.com] announced for PEM-type electronics. I see no "sound barrier" for nanotech yet - to the contrary, I see nanotech slipping past the micro "barrier" ever more quickly.

Re:Is there a limit?

[rbrander]

" It won't come to a screeching halt at any obvious point, but expect to see smaller improvements spread further apart."

Nearly 10 years back, before the word "blog" existed, I did a little web article called The End of Moore's Law - Thank God! [cuug.ab.ca] that used the info in two excellent Scientific American articles which hypothesized a slow levelling off of the Moore's Law exponent around ... well, a year or two ago, actually, rather than a few years from now. But close enough.

The second Sci. Am. article stressed that it was an economic decision and drew parallels both to aviation (aircraft grew in size rapidly until the 747) and to trains (the biggest-ever locomotive was designed in the 50's)

In both cases, you wound up with the entire market being needed to pay the costs of the last generation of development. Presumably, the "Last Fab" will require a consortium of Intel, IBM, AMD, Motorola, etc - and make chips for all of them to pay off the $10 Billion construction cost.

Nano...

[Spy der Mann]

Oh, did I mention that you gain less and less from going smaller
because more signal is wasted as heat.

Unless of course, you're optical transistors, nanotubes, spintronics and all that nano stuff that hasn't been applied to electronics yet.

Nice press, but these chips ain't cheap

[smilindog2000]

Finer circuitry decreases the size of a chip and cuts per-unit production costs... NOT!

Moore's Law is showing it's age... The cheapest transistors in the world are not build in 65nm. They are built in 180nm, a much older process.

In China, you can get 8-inch 180nm (.18u) wafers for $600. Today, a 90nm 8-inch wafer is more than 4X more expensive, and you cannot yet buy 65nm wafers. The cost per transistor is actually higher! And people wonder why we're taking our time to move to finer geometry processes!