Zetta technology major achievement - complete SEM lab built!

Rxke · 2004-03-17 02:19:13

Wow. Double wow...
[http://www.mic.dtu.dk/research/Nanointe … allery.htm]These guys built something really incredible...

Now predictions for feasibility of a nanofactory are only 5 years!

In a nutshell: they built a complete mini-lab under a SEM (Scanning Elecron Microscope) capable of building stuff molecule by molecule, in one package....

This makes [http://www.crnano.org/overview.htm]molecular manufacturing a whole lot closer...

dicktice · 2004-03-17 09:38:02

Talk about Inner-space! Keep up the good work Rxke--I for one promise to: watch this "space."

SBird · 2004-03-17 11:04:58

Again, I hate to be a wet towel but this is something of an area of expertise for me. (I work on biologically assembled nanotech in grad school)

This is cool work but it is not going to create a molecular assembly factory in 5 years. Nanotech has been rife with overoptimistic predictions of its progress rate that have consistently been wrong.

Take for example MEMS - back in the early 90's, they started making gears and stuff with silicon fab technology and they predicted all of these marvellous micro-scale robots that would run aroud, cleaning and doing other stuff for us (sound familiar?) Well, 10+ years later, there are a handful of useful applications for MEMS and none of them involve those little gears. (tiny cantelievers are used as deaccelleration sensors for airbag deployment)

So far nanotech has been great at making tiny junk - like wires and particles. Unfortunately, the ability to wire these into something useful has basicaly been non-existent. Partially, it's because there's no real science for self-assembly of these things and also, we have only a vauge idea of what sort of nanoscale structures will be useful once built. My guess is, that like MEMS, the sorts of structures that we think will be useful and the sort of structures that will end up actually being useful are two completely different things.

Rxke · 2004-03-17 11:43:35

While i also feel sceptical about the 5 years prediction, i *do* think this kind of work will yield usable results in 10-15 years. Not necc. self-assembling stuff, but manufacturing of truly micron-sized 3D electronics and stuff, that replaces macro-scale hardware used in state of the art hardware, like space-probes, and the military.
And OTOH one can't help but dream, grin...

I personlly think the downplaying of MEMS is too harsh. Those tiny (Really tiny) accellerometers have more applicattions than using them for mundane stuff like air-bag triggering. Now you make it sound like they're some kind of dull stuff, but the potential is huge.
The gyroscopes in ISS ( or Hubble etc.) for instance, are heavy, error-prone and very expensive. You could switch to MEMS-derived gyroscopes, by using just 3 of those acc. sensors.
Pack, say 50 or more for *massive* redundancy in a device that fits on a cube the size of a sugarcube, and you can clearly see the advantages in launchcosts, redundancy, etc...
The claim of MEMS-derived 'real-world' roving bots was merely a popularisation thing, the gears etc won't work (van der Waals forces, or simple dust, static charges and all that)

... With self-contained labs as described in the article, you can do some pretty impressive stuff, i'd think. If only for the fact that this 'package' makes experimenting a lot less convoluted. I'd bet there will, in the coming years, grad students like you (and i envy you, really do! I'm student too, but in a completely diff. area, though i read what i can about thes things) will get a chance to 'play' with stuff like the abovementioned labs...
We're only in the 'playing around with bits of strings and tweezers' stage now, but it's going fast. (I tend to think in 5 years increments, and am easily awed, heehee)
The bio-side of the picture becomes more important by the year, too, but is often overlooked (why?) despite the fact that the biotech industry arguably pours much more money into these things than the 'gold and nanotube' boys. Maybe because they don't talk about assemblers, but more about 'non-sexy' stuff like biolabs on a chip?
Designing a self-assembly plant is still way off, if at all possible because no-one has really looked into it in depth (Drexeler's on a one-man road, but that's understandable, for who will want to put in months of his life in designing abstract machinery...)
Still, developments like the 'biolab on a chip' promise such an interesting new market, further researh is all but certain.

I remain optimistic, no: thrilled.

Rxke · 2004-03-17 13:11:11

Gene-extracting lab fits on a chip... [http://www.spacedaily.com/news/clone-04c.html](Spacedaily)

I *swear* i didn't read it untill only minutes ago, creepy how reality bites back...

Related to the initial post: Charge doping of molecules one atom at a time using STM ([http://www.spacedaily.com/news/nanotech-04m.html]Spacedaily)

SBird · 2004-03-17 16:26:35

I'm not saying that MEMS is useless - far from it. However, it's taken about 15 years to even see the most basic applications start appearing and these are quite different from the applications that were originally thought up.

For example, the lab-on-a-chip stuff (I haven't messed with these, they're too expensive for anything but companies and the biggest academic labs still) have actually been commercialized for a few years now and show great promise but were only possible after people realized that the low Reynolds number flow regimes at those size scales make it impossible to make normal valves and pumps. On the other hand, the same flow conditions make possible a whole host of extremely efficient filtration and pumping sustems that are radically different from what we are accustomed to.

As far as accelerometers, the MEMS stuff isn't very sensitive. Those accelerometers are great for detecting whether you just wrapped your car around a telephone pole but not much good for detecting stuff like a degree per day rotation like the Hubble gyroscopes do. (I don't understand why the Hubble doesn't use laser gyros - they're more accurate and don't wear out). OTOH, for things like vibration detection and impact detection, MEMS sensors are great. I'm assuming the hundreds of sensors they're installing in the Shuttle to watch for debris impact are going to be MEMS based.

Things like being able to dope single atoms with an STM tip and the like are nothing but dumb stunts. It's a great way to make a computer if you don't mind waiting a few thousand years for it to be built. Modern lithography can churn out thousands of chips a day. A standard fab line can churn out literally several billion transistors a day.

In contrast, these manipulation techniques are lucky to be able to make a single transistor in several hours of work. There is no way to parallelize this process, either. The only way that nano is going to give us computers is with some sort of massive self-assembly technique. The biological nanotech I'm working on is trying to expand on this approach but practical applications are still years off.

I agree that nano will give us some useful products in afew years but this is nothing new. Photographic film is an applied nanotech system using silver halide nanoparticles. Chinese rose-tinted pottery uses the quantum confinement effects of the plasmon absorption of nano-sized gold particles to generate it's distinctive color and that's been around for about 500 years.

The reason that bio people tend to be underawed by nanotech is that they have been working on studying a fully integrated nanotech system that naturally evolved 3.5 billion years ago. The particle and wire stuff coming out of the labs these days is kinda silly in comparison.

My take is that nanotech will have a number of applications in the next 10 years. Nanotech will start to ahve an impact on semiconductors in 10-30 years. Nanotech will start to become commonplace with artificial living organisms in the 20-100 year timeframe.

Rxke · 2004-03-25 14:21:44

bit of a belated reply...

MEMS... I read about .1 G MEMS, so that's a bit better than being able to sense "for detecting whether you just wrapped your car around a telephone pole" although Hubble needs more than that precision. So MEMS are out.

... About the doping of single molecules or even atoms I'm not so sure... As you rightly pointed out, massive self assembly is needed to get macro-scale results, but these STM techniques may build one day a nano- fabricator that does just that... building copies of itself (with external programming, of course, and so iteratively build up to massive assembly... CRN has a paper on a simple approach of such a device, [http://www.jetpress.org/volume13/Nanofactory.htm]Design of a primitive nanofactory (80-odd pages)

Of course, as long as there is no fabricator, it can't be built... The author stated it could take 20 years to get there, but in his
[http://crnano.typepad.com/crnblog/2004/ … rk_so.html]recent blog article he now says he's concerned it could be done in as little as five...

New Mars Forums

Announcement

#1 2004-03-17 02:19:13

Re: Zetta technology major achievement - complete SEM lab built!

#2 2004-03-17 09:38:02

Re: Zetta technology major achievement - complete SEM lab built!

#3 2004-03-17 11:04:58

Re: Zetta technology major achievement - complete SEM lab built!

#4 2004-03-17 11:43:35

Re: Zetta technology major achievement - complete SEM lab built!

#5 2004-03-17 13:11:11

Re: Zetta technology major achievement - complete SEM lab built!

#6 2004-03-17 16:26:35

Re: Zetta technology major achievement - complete SEM lab built!

#7 2004-03-25 14:21:44

Re: Zetta technology major achievement - complete SEM lab built!

Board footer