Atomic-scale manufacturing now a reality

May 23, 2018, University of Alberta
Atomic-scale manufacturing now a reality
Credit: ACS

Scientists at the University of Alberta have applied a machine learning technique using artificial intelligence to perfect and automate atomic-scale manufacturing, something which has never been done before. The vastly greener, faster, smaller technology enabled by this development greatly reduces impact on the climate while still satisfying the insatiable demands of the information age.

"Most of us thought we'd never be able to automate atomic writing and editing, but stubborn persistence has paid off, and now Research Associate Moe Rashidi has done it," said Robert Wolkow, professor of physics at the University of Alberta, who along with his Research Associate has just published a paper announcing their findings.

"Until now, we printed with about as efficiently as medieval monks produced books," explained Wolkow. "For a long while, we have had the equivalent of a pen for writing with atoms, but we had to write manually. So we couldn't mass produce atom-scale devices, and we couldn't commercialize anything. Now that has all changed, much like the disruption following the arrival of the printing press for those medieval monks. Machine learning has automated the atom fabrication process, and an atom-scale manufacturing revolution is sure to follow."

Doing more with less

This builds on Wolkow's extensive body of work in creating solutions to drive atomic-scale low-power electronics. The physicist has devoted his career to pushing atomic-scale manufacturing forward in response to not only the rapidly changing needs of our but also the changes to our climate. Some estimates predict that if we continue on pace with our current energy consumption habits, by 2025, the information and communication technology industry would not only consume 20 percent of the world's energy but also contribute more than five percent of the .

For Wolkow, this all adds up to an urgent need for a new basis for our electronics, something which he predicts will be powered by atomic-scale fabrication and mass manufacturing, now possible thanks to his new discovery.

"Fabrication at the ultimate small scale not only lets us do things better, but we can also create entirely new functions that conventional technology simply cannot do. Combining that with a practical path to will be game changing. This allows us to create a new, extremely efficient basis for computing using the natural properties of individual atoms."

Explore further: Electrical currents can be now be switched on and off at the smallest conceivable scale

More information: Mohammad Rashidi et al, Autonomous Scanning Probe Microscopy in Situ Tip Conditioning through Machine Learning, ACS Nano (2018). DOI: 10.1021/acsnano.8b02208

Related Stories

Super-resolution laser machining possible

March 5, 2014

A newly discovered natural phenomenon shows that light could be used to pick apart a substance atom by atom, with new avenues for nano-scale diamond devices, as identified this week by Australian researchers in Nature Communications.

Recommended for you

Nanobot pumps destroy nerve agents

August 21, 2018

Once in the territory of science fiction, "nanobots" are closer than ever to becoming a reality, with possible applications in medicine, manufacturing, robotics and fluidics. Today, scientists report progress in developing ...

Flexible color displays with microfluidics

August 16, 2018

A new study published on Microsystems and Nanoengineering by Kazuhiro Kobayashi and Hiroaki Onoe details the development of a flexible and reflective multicolor display system that does not require continued energy supply ...

Twisted electronics open the door to tunable 2-D materials

August 16, 2018

Two-dimensional (2-D) materials such as graphene have unique electronic, magnetic, optical, and mechanical properties that promise to drive innovation in areas from electronics to energy to materials to medicine. Columbia ...

36 comments

Adjust slider to filter visible comments by rank

Display comments: newest first

flashgordon
5 / 5 (1) May 23, 2018
I'm not going to spend money for a subscription; it would be nice to know more details on this. Like, what nano-structures have they built? Have they built anything to a macroscale? Even building to micrometers is good enough at this stage!
Captain Stumpy
3.7 / 5 (3) May 23, 2018
I'm not going to spend money for a subscription; it would be nice to know more details on this.!

arXiv is your friend - or google scholar

https://arxiv.org...7059.pdf

flashgordon
5 / 5 (2) May 23, 2018
I'm not going to spend money for a subscription; it would be nice to know more details on this.!

arXiv is your friend - or google scholar

https://arxiv.org...7059.pdf



indeed, very good, thanks!
TheGhostofOtto1923
not rated yet May 23, 2018
Just wait until we are able to manufacture goods and materials atom by atom with the same speed and efficiency as we now do textiles.
Da Schneib
3 / 5 (2) May 23, 2018
This is the technology that will drive Moore's law to the ultimate in electronic miniaturization; the placement of individual atoms on a device by nanobots in a manufacturing environment. Electronics will come first but ultimately we will be able to create most of the products of our technology this way.

For a pretty good preview folks should read The Diamond Age: Or, A Young Lady's Illustrated Primer by Neal Stephenson. The politics in it are mostly handwaving but the technology and its impacts on society are, if anything, conservative estimates.

This is the power to make any imaginable object from scratch, from the individual atoms on up. It's 3D printing on steroids. This will open up space, but it comes at a price: any insane person will eventually be able to use these techniques to manufacture any destructive device we can imagine and use it for terrorism. The corollary is that we must begin now to ensure mental hygiene for all people, or we are doomed as a race.
Da Schneib
3 / 5 (2) May 23, 2018
And if you doubt it, see this article: https://www.cnn.c...dex.html

This is real. It is here. And it may be the reason for the Fermi Paradox. There is a discussion about mental hygiene over here: https://phys.org/...acy.html

In the face of this, gun control laws are a joke. For that matter it will eventually (most likely within the lifetime of the younger among us) be possible to 3D print a nuclear weapon. What then? Think about it.
Eikka
2.3 / 5 (3) May 23, 2018
Just wait until we are able to manufacture goods and materials atom by atom with the same speed and efficiency as we now do textiles.


Don't forget that in dealing with individual atoms, you have to deal with the chemical energy barriers that link atoms in molecules in order to pick the atom up and place it down. This involves truly massive amounts of energy if done on a bulk scale.

For example, aluminium oxide has an entalphy of -1669.8 kJ/mol so about -64 kJ per gram, which means 64 kiloWatts of heat per gram per second is released to combine aluminium and oxygen atom by atom to "print" something that contains alumina. The resulting object will be pretty hot, and the printer's printing head would be glowing like an arc welder making anything larger than a penny in a hurry.

Trying to print as fast as an automated loom makes canvas would probably result in the machine evaporating itself.
big_hairy_jimbo
not rated yet May 23, 2018
Once manufacturing atom by atom has been fully achieved, along with A.I. and robots, then you can expect the wealthy to eliminate most of humankind, to turn the world into utopia. So live it up now people, as the countdown to our extermination is on. This wont be A.I. thinking it needs to get rid of us, but the elite no longer needing tax paying plebs. The planet will be better for it.
Eikka
3 / 5 (2) May 23, 2018
the placement of individual atoms on a device by nanobots in a manufacturing environment.


The common fantasy about nanobots is that they're like their bigger cousins, only smaller - but people forget that when you go to the nanoscale, your robots are still made of the same atoms, so when you're talking about nanobots you're actually talking about finely controlled chemistry and quantum effects rather than computers and mechanisms. It hasn't got anything to do with robotics anymore.

Molecular machines aren't like a supercomputer shrunk into the size of a virus - they are effectively viruses or at least work in a similiar fashion. This also means they won't be exceedingly smart or capable.

And again, the energy constraints apply. You can't have a manufacturing system that requires enough energy input to vaporize the materials. It would, well, vaporize.
Eikka
3 / 5 (2) May 23, 2018
For that matter it will eventually (most likely within the lifetime of the younger among us) be possible to 3D print a nuclear weapon.


First you need to find a couple kilograms of refined plutonium without anyone noticing, and grind it down to very very fine powder, then ionize it into hot gas in a hard vacuum chamber, and then supply enough power to the device to assemble your nuclear bomb atom by atom.

Without killing yourself in the process of course.
Da Schneib
not rated yet May 23, 2018
@Eikka, you've forgotten quantum tunneling. And nobody's talking about fast manufacturing yet; a potential terrorist can wait a few months to get a gun.

Right now they're using cars: https://www.cnn.c...dex.html Maybe you forgot.
flashgordon
not rated yet May 23, 2018
all they've done is make sure the tip doesn't get blunt. But, they say they can apply this immediately to some parallel S.T.M. technology they've already had in the works; should prove interesting how quickly they get up to speed.
Eikka
1 / 5 (2) May 23, 2018
@Eikka, you've forgotten quantum tunneling.


Quantum tunneling is a very low probability event. You slam a million billion heavy atoms against a barrier, and a handful tunnel through. What's it gonna help?
Da Schneib
5 / 5 (1) May 23, 2018
@Eikka, you've forgotten quantum tunneling.


Quantum tunneling is a very low probability event. You slam a million billion heavy atoms against a barrier, and a handful tunnel through. What's it gonna help?
Those who are not worried about waiting.

@Eikka, nobody has to grind anything up or use gas-phase manufacturing. Pretty sure you don't get how this technology works.
Da Schneib
not rated yet May 23, 2018
all they've done is make sure the tip doesn't get blunt. But, they say they can apply this immediately to some parallel S.T.M. technology they've already had in the works; should prove interesting how quickly they get up to speed.
It doesn't matter if it takes a year or a hundred years, @flash. In the long run anyone can make anything.
Da Schneib
not rated yet May 23, 2018
@Eikka, you should also be aware that nanobots can exert forces using those same charges. Are you seriously arguing that these forces are different between the substrate and the nanobots? How, by magic? Because jebus said so? Enzymes perform such tasks in living cells all the time. They seem to work just fine across the "enormous chemical barriers" you blather about.
Eikka
5 / 5 (1) May 24, 2018
@Eikka, nobody has to grind anything up or use gas-phase manufacturing. Pretty sure you don't get how this technology works.


Well, how else are you going to deliver the materials to the printing head? By magic? Somehow you have to take atoms from one place and lay them down at another.

@Eikka, you should also be aware that nanobots can exert forces using those same charges. Are you seriously arguing that these forces are different between the substrate and the nanobots?


No. I was telling you they're the exact same forces. You spend energy when you pick up an atom of aluminium from a block of aluminium, and then release said energy, or more, when you combine the atom to other atoms at the target location. If the material deposition rate is great, that energy will be great.

Enzymes perform such tasks in living cells all the time.


And when enzymes perform their task rapidly, you get heat, which denatures the enzymes and stops them from working.
Eikka
5 / 5 (1) May 24, 2018
One simple example is brewing beer. You take the grain and you germinate it, and then you pitch them in a pot with water. The amylase enzymes present in the germinated grain convert the starches into sugars, and that happens over a rather precise temperature window at around 65 C give or take a few degrees. Above 78 degrees the enzymes get destroyed.

These enzymes are like your nanobots. Their structure and chemical composition determines what they do, how well they do it, and under what conditions they can perform. If you try to run the process faster, you'll just ruin the beer.

Atomic scale manufacturing is fundamentally electrochemistry, like copper plating a piece of steel by running an electric current - only you're doing it in a more controlled fashion atom-by-atom and so the energies are more concentrated on the tip of your writing implement.
Eikka
not rated yet May 24, 2018
As for electrochemical manufacturing, it's already a thing:

https://advanceme...-copper/

Takes one hour to deposit a 10mm line of copper some micrometers thick. Trouble being that at higher deposition rates, all sorts of chemical side reactions start happening and the deposition efficiency drops.
Eikka
not rated yet May 24, 2018
Pretty sure you don't get how this technology works.


Actually, the technology they're talking about is based on pushing grooves into the substrate material and nudging the surface atoms around using the same scanning probe microscope tip that is used for imaging the surface. It's not a technology that builds macroscopic objects by adding material by itself.

https://www.youtu...X8h8cG1s

What they have is like an atomic-scale equivalent of a chisel used to sculpt the object like an artist makes a statue. The solution they found is a way to automatically sharpen the tip, so a computer can perform the work automatically without making a mess of it.
Da Schneib
not rated yet May 24, 2018
OK, so what, in your estimation, is the neural networking AI for?
Eikka
not rated yet May 24, 2018
One of the points from that TED talk video is that the atomic scale manipulation they do is only possible under vacuum, to keep the surfaces clean, and under very low temperatures or else the atoms simply will not stay put where they place them. They start diffusing around. The structures they're building are not useful as it is.

The advantage is that as you scale down the individual features of say, a CPU, the traditional manufacturing methods do not produce a perfectly even distribution of atoms, so your semiconductors may end up not semiconducting because you didn't land the right kind of atoms in the right areas. This problem gets worse as the feature size shrinks.

With atomic scale manipulation, you essentially sprinkle all the atoms around approximately where they should go, and then nudge them into the right places, fixing all the faulty transistors.

Of course, there being a billion of them, that's going to take the age of the universe if not automated.
Eikka
not rated yet May 24, 2018
OK, so what, in your estimation, is the neural networking AI for?


Replacing the human operator who's trying to nudge the atoms around. As the tip of the probe is wearing out and changing shape continuously anyways, the machine has to adapt to its changing properties. Otherwise the machine would just keep going through the motions whether the tool works or not.

TheGhostofOtto1923
not rated yet May 24, 2018
Many different technologies besides this one.

"Molecular nanotechnology (MNT) is a technology based on the ability to build structures to complex, atomic specifications by means of mechanosynthesis.[1] This is distinct from nanoscale materials. Based on Richard Feynman's vision of miniature factories using nanomachines to build complex products (including additional nanomachines), this advanced form of nanotechnology (or molecular manufacturing[2]) would make use of positionally-controlled mechanosynthesis guided by molecular machine systems. MNT would involve combining physical principles demonstrated by biophysics, chemistry, other nanotechnologies..."
TheGhostofOtto1923
not rated yet May 24, 2018
And look! They have an institute.

"Molecular manufacturing will eventually transform our relationship to molecules and matter as thoroughly as the computer changed our relationship to bits and information. It will enable precise, inexpensive control of the structure of matter."
— Neil Jacobstein, Chairman, Institute for Molecular Manufacturing

"Molecular manufacturing is a future technology that will allow us to build large objects to atomic precision, quickly and cheaply, with virtually no defects. Robotic mechanisms will position and react molecules to build systems to complex atomic specification. The theoretical capabilities and performance of these systems have been analyzed for over fifteen years, molecular machine components are being built now, and molecular manufacturing could mature within the next ten years."

-Doesn't seem to share your pessimism.
Eikka
not rated yet May 24, 2018
Based on Richard Feynman's vision of miniature factories using nanomachines to build complex products


That's a funny story actually. Feynman envisioned a pair of mechanical arms that would be used to construct a pair of smaller mechanical arms, that would be used to construct even smaller mechanical arms (and tools to go along) unitl you reach the atomic scale.

That of course had the distinct problem of tolerances, where the motion of your arms using the larger tools no longer carry along to any precise motion at the atomic scale because of all the play in the system - but it was just a thought experiment anyhow.

-Doesn't seem to share your pessimism.

When have you ever seen a pessimist salesman, or even a realist one?
Eikka
not rated yet May 24, 2018
One of the criticisms towards this mechanosynthesis is exactly that they're proposing to use nanomachines, which are individually hard to make in sufficient numbers so they'd have to make nanomachines make more nanomachines (replicate), but then you've got essentially a virus or a bacteria - an artificial living organism that is subject to mutation and running out of your control.

As good as the error correction mechanism can get, when you're employing them by the kazillions, someone's going to get "cancer" - and this problem hasn't been adequately answered by the nanomanufacturers.

So it's back to the good old chain of diminishingly smaller pairs of mechanical hands doing the work.

Plenty of other criticism here:
https://en.wikipe...chnology

Summa summarum, it's an idea based on science fiction that some people are desperate to make true more by arguing rather than by showing how it's done.
TheGhostofOtto1923
not rated yet May 24, 2018
That's a funny story actually. Feynman envisioned a pair of mechanical arms
Perhaps the experts at the Institute for Molecular Manufacturing understood that Feynmann was being figurative and not literal? Feynman was a funny guy.

"IMM designs of complex molecular machinery are the most advanced done to date, and have survived intense critical review by the research community. Here are some recent examples:
Simple pump, selective for neon
Fine-motion controller for molecular assembly
A molecular differential gear"
http://www.imm.org/about/

-Then of course theres this...

"Nobel Prize in Chemistry 2016 is awarded to Jean-Pierre Sauvage, Sir J. Fraser Stoddart and Bernard L. Feringa for their development of molecular machines that are a thousand times thinner than a hair strand. This is the story of how they succeeded in linking molecules together to design everything from a tiny lift to motors and minuscule muscles."

-Somebody (besides you) is taking it seriously, no?
TheGhostofOtto1923
not rated yet May 24, 2018
Summa summarum, it's an idea based on science fiction that some people are desperate to make true more by arguing rather than by showing how it's done
So explain for me when did the Nobel committee begin awarding prizes for scifi? And 'showing how its done' - isnt that a requisite for consideration in oslo?
TheGhostofOtto1923
not rated yet May 25, 2018
Silence from eikka. Otto wins another one yay.
Eikka
not rated yet May 26, 2018
So explain for me when did the Nobel committee begin awarding prizes for scifi?


Same as how they award prizes to Bush junior and Obama. It's just a political prize.

And 'showing how its done' - isnt that a requisite for consideration in oslo?


Not really.

Silence from eikka. Otto wins another one yay.


People have other things to do than argue with you.
Eikka
not rated yet May 26, 2018
"Nobel Prize in Chemistry 2016 is awarded to Jean-Pierre Sauvage, Sir J. Fraser Stoddart and Bernard L. Feringa for their development of molecular machines that are a thousand times thinner than a hair strand.


"Molecular machines" isn't the same thing as "molecular manufacturing".

You/they are shifting the goalposts.

It's one thing to demonstrate a working molecular machine, which can be made in any old way, and a whole other thing to demonstrate molecular manufacturing which is the nanobots making nanobots making other stuff idea that they're fundamentally trying to sell.

It's basically just like pretending that magic is real by showing regular card tricks.
Eikka
not rated yet May 26, 2018
The problem here is the same as with the AI hype. Since like "intelligence", "manufacturing" is sufficiently vague - meaning everything from people assembling molecules atom-by-atom by hand on an STM, to the sci-fi ideas of magic replicators and "grey goo" that makes you stuff - you can pretend that you're doing/approaching the latter by doing the former and simply calling them by the same name.

It's a gap of imagination the size of inventing the hammer, and immediately jumping to the conclusion that eventually because hammers exist, houses will build themselves by magical flying hammers that make other hammers (and the nails, planks...).

It's the modern equivalence of Alchemy, where basically a bunch of conmen performed chemical parlor tricks to enthusiastic but naive audiences, taking something that works and hyping it up to a ridiculous degree.
TheGhostofOtto1923
not rated yet May 26, 2018
Lessee, eikka whining or nobel committee awarding? Who has more cred?

BTW who was it who invented the laser? Did you whine about that as well? Did that seem like alchemy to you also?
Whydening Gyre
not rated yet May 26, 2018
Eikka,
The are doing rudimentary manufacturing at nano level.
Baby steps, m'friend...
Anonym341970
1 / 5 (1) May 31, 2018
If it were me, my first task would be to start manufacturing gold.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.