Ideas Factory: Manipulating Molecules

The second to-be-funded proposal that last week's Ideas Factory generated is a system for manipulating and joining molecules and nanoparticles under direct computer control.

"...positional assembly of molecules and nanoscale building blocks. Computer-directed actuators will be used to drive (with sub-nanometre to sub-Angstrom precision) the elements of a nanosystem along pre-defined and entirely deterministic trajectories, thereby achieving structures not accessible by mimicing natural assembly strategies alone. Linkages and bonding between the building blocks will also be initiated, modulated, and - in some cases - terminated by direct computer control."

Pretty cool. They plan to build a system that uses what they describe as a "nanoscale conveyor belt" to move the building blocks. Responses in the comments following the article hint that they expect this eventually to work in a variety of mediums and temperatures, and that they want a number of different degrees of freedom --n ot just one-dimensional motion. So they plan to move building blocks into position, then bond them by some unspecified -- possibly photochemical -- mechanism.

They don't say much about the mechanisms. But they do plan to build a system that can gather energy (perhaps photons), switch the energy into different pathways, and then output the energy to do something useful. They want to build "an autonomous, abiotic nanomachine." Comments indicate that the energy-switches are supposed to be sensitive to events at the molecular level. So whatever they build should be, in a sense, programmable.

The "conveyor belt" sounds MEMS-like, but the products are not. The intended products are, in fact, quite advanced functional nano-constructions. It sounds like the products are intended to transmit both energy and mechanical force through multiple components in order to create sensor-controlled motion. The most advanced mechanical molecule I've heard of so far has approximately the complexity of a pair of fireplace tongs. So this would be a rather large step forward.

Simply setting the goals they set, for the manufacturing system and the products, is an important step that we expect will inspire other work on advanced nanomachines and computer-controlled mechanical/covalent manufacturing systems.

Once their system is working, it sounds very useful for building and testing a wide range of nanomachines. It should pretty quickly settle a number of questions about how to engineer molecular mechanical systems -- how to predict the properties that various structures will have. And the nanomachines that are built may well be an enabling technology for more advanced molecular manufacturing.

Chris Phoenix

Tags: nanotechnology nanotech nano science technology weblog blog

Comments

[Hal]

I am confused about why their proposals are so vague and secretive. They're saying the bare minimum possible about how they're going to do it, while promising these extravagant improvements in technology. Is this customary in science funding?

I can understand a desire to keep one's ideas secret in order to prevent copying by rivals, but in practice it would seem that the only people qualified to evaluate a proposal for funding would be the same ones who would be in a position to steal the ideas. I imagine that at this meeting, the proposals were described in detail and the specific technologies were laid out explicitly. Apparently this was done under conditions of sworn secrecy by all concerned. Yet one would again imagine that the other members of the group would be the very rivals from whom the proposers would want to keep their ideas secret.

It's odd how science, probably the institution in our society which most prides itself on openness, is required to use secrecy as such a key part of its development process.

[John B]

How is this different from the http://crnano.typepad.com/crnblog/2006/12/technology_in_a.html post you folks have effectively hand-waved your way through approving?

[Philip Moriarty]

Hal,

Your comments re. secrecy and the progress of science struck a chord with me. As someone with an avowed loyalty to fundamental academic research, where the underlying ethos should be dissemination of information as opposed to (in some cases, pathological) protection of IPR, I very much take your points on board. However, I need to clarify a couple of points:

(i) The paragraphs posted on the Ideas Factory website represent a generic and "broad brush" overview of the concepts and ideas developed during the week in Southampton. At the moment the Ideas Factory groups are preparing detailed and fully costed proposals which will then be refereed by the Director and the mentors. The deadline for submission of the applications is about one week from now. I for one do not feel comfortable posting detailed information until the proposals have reached their final stage of development and undergone the full peer review "cycle". Following confirmation that the proposals can go ahead, I will be more than happy - with the agreement of my colleagues - to post more detailed information.

(ii) An important aspect of the Ideas Factory which needs to be stressed is that the project descriptions posted on the Software Compilation of Matter website are not the only outcomes of the event. I certainly hope to be involved in at least one other proposal (to be submitted outside the Ideas Factory scheme) based on ideas discussed during the sandpit.

(iii) In the case of the "Directed Reconfigurable Nanomachines" project, the single paragraph description describes our ultimate goals. There appears to be some misunderstanding (certainly with respect to recent posts on this blog) that the paragraph represents what is going to be achievable in two years.

----

John B.,

There's a very big difference between what's proposed in the "Directed Reconfigurable Nanomachines" project and the Drexler-Burch nanofactory simulation. The fundamental difference, although there are many others, is that we are not suggesting that we will build nanomachines from individual atoms.

[Phillip Huggan]

The innovative thing about "Directed Reconfigurable Nanomachines" is that it is building from the bottom up using some principles of mechanical engineering (some rigid molecules I assume). It won't lead to a Drexler nanofactory. Drexlerian ideas won't lead to a Drexler nanofactory without an electrostatic motor blueprint.

[NanoEnthusiast]

Philip M, when you are able to, I would like to know just what kind of building blocks you'll be using and how you plan on making them.

With regards to the nanofactory animation, the main similarity I see is the use of a conveyor belt. I wonder just who came up with the idea of using a conveyor belt at the nanoscale, was Drexler the first?

When the projects are complete, I imagine there will be a lot of interest in seeing to what extent proposal 2 and proposal 3 can be combined into a new project. Then we will have better understanding of the ultimate feasibility of the nanofactory animation.

[Richard Jones]

Hal, it is generally the case that scientific proposals are regarded as confidential, and it is also entirely customary for some scientists to be utterly paranoid that they are going to get scooped to the big result by rivals stealing their ideas (not least the rivals who refereed their grant proposals). There are certainly IP reasons for confidentiality (as suggested, with disapproval, by Philip M above) - too much detail at the proposal stage certainly could compromise a future patent, even more in the UK than in the USA, due to differences in patent law between the two jurisdictions. Scientists also quite often have a superstitious reluctance to talk too much about what they are trying to do until it at least shows serious signs of working, simply for fear of the embarassment of not being able to meet expectations. Science is an odd mixture of competitiveness and collaboration, openness and secrecy, and different bits of science have different customs. In very competitive bits of biomedical science, for example, it is usual to say nothing at all about what you are doing until a peer-reviewed paper is published, whereas physicists are often quite happy to talk about results as they come in. Obviously, our aim in the Ideas Factory was to steer things to the collaborative end of the spectrum. In terms of the information we gave out, the summaries were written by the participating scientists and represent what the scientists themselves were comfortable about releasing. This seems right to me; after all, the ideas generated belong to the scientists, not to me or to EPSRC.

[Chris Phoenix, CRN]

NanoEnthusiast, I doubt that Philip Moriarty's proposal has much to do with any nanofactory design. I suspect that "conveyor belt" is at best a metaphor, not a description of a moving strip of material with cargo. If I had to guess, it might involve pushing objects adsorbed on a patterned surface in order to displace them by (perhaps large) fractions of a "bump."

Phillip Huggan, given the high power transmission rates possible in strong rotating rods, I do not think the nanofactory requires built-in motors. It might have a shaft coming out the side that is split up and distributed, sort of like the overhead power feeds in old steam-engine-driven factories.

Consider: My "Primitive Nanofactory" draws 2E5 watts, thus requiring 2E11 1-microwatt Drexler motors. But it has 3.5E13 computer/drivers. So the power would have to be distributed in the last few stages anyway.

And 2E5 watts is e.g. 1000 volts at 200 amps--pretty hefty wires and bulky insulation--while a half-millimeter shaft should suffice to carry the mechanical equivalent. So mechanical distribution might be a preferable design even after nano-motors become available, because mechanical power transmission should be more volume- and material-efficient than electrical.

Chris

[Phillip Huggan]

" (C.Phoenix wrote:) ...given the high power transmission rates possible in strong rotating rods, I do not think the nanofactory requires built-in motors. It might have a shaft coming out the side that is split up and distributed, sort of like the overhead power feeds in old steam-engine-driven factories."

Rotating nanorods have to be turned by something. Some form of energy we can control has to turn the nanorods we can't control directly.
SPMs utilize PZT, a piezoceramic. What will MNT use? Drexler basically swept this under the table by saying "I hope we can use an electrostatic motor". This fundamental open question has been exposed by R.Jones and others; it is the primary reason why I think diamondoid MNT is only 1/3 to 2/3 likely. If we are stuck using PZT for MNT, it represents a limitation to MNT industrial output, making MNT subservient to mining industry productivity advances.

[Tom Craver]

??? But the post of Chris' you quoted already addresses your objection - bring a shaft out the side of a nanofactory and power it with a conventional electric motor, if necessary, to provide the necessary power. Probably NOT necessary, but if it were...

[Phillip Huggan]

"??? But the post of Chris' you quoted already addresses your objection - bring a shaft out the side of a nanofactory and power it with a conventional electric motor, if necessary, to provide the necessary power. Probably NOT necessary, but if it were..."

That just sounds like a conventional SPM tool-tip save for the addition of a yet-to-be-invented 360 degrees of freedom pivot around the tool-tip's circumference (powered by an un-MNTable electric motor as opposed to an un-MNTable PZT tube). Our SPMs *still* use PZT, and lasers, and a whole bunch of components that may be very difficult to assemble from diamonds and/or other ceramic diamondoids.

How do you make a conventional electric motor from diamond? We are to have a conventional motor for each SPM? That isn't as crazy as it sounds, a few months back I read of a prototype tens-of-nm sized gas turbine motor developed at some American university. IFAIK, electric motors need metals. There is no closure (and thus no industrial productivity acceleration) unless a mechanosynthetic means for creating an electric motor is ultimately forwarded (ie, a diamond-boron motor or boron-nitride motor or something).
Diamonds have a limited operating temperature range. Other industrial processes such as smelting (needed for existing electric motors and many power generation/transmission components) will still be required post-MNT.

[Phillip Huggan]

I can be more specific: Chris's proposal sounds like an AFM that has a high aspect ratio rotating tip (capable of handling a moeity obviously). Instead of PZT (transducing electricity into mechanical motion), an electric motor is to be used?

Does the UHV chamber enclosed the electric motor? How big an electric motor? I don't know much about electric motors, but a PZT crystal seems like an easier transducer.

[Tom Craver]

I'm not sure why this is even coming up - there are already MEMs electric motors made with silicon, and MNT motors should only get better.

But yes, I think you could make a motor using carbon conductors and an electromagnet in place of permanent magnets.

[Chris Phoenix, CRN]

An electromagnetic motor can obviously be made without metal. For small sizes, electrostatic motors will be better. For driving a nanofactory with a single big motor, electromagnetic is probably better.

I don't know why SPM actuators are in the same conversation with large motors driving large shafts. You did understand, I hope, that it would be one large shaft to power the whole nanofactory--split into smaller shafts. (This does not imply that every small shaft needs to be error-free.)

Chris

[Phillip Huggan]

When I Google "electrostatic motor", all I get is a very macro-sized VandeGraff generator. How do we even know a nano-sized prototype is likely to work? Are all the diamond MNT eggs in this one basket?

"An electromagnetic motor can obviously be made without metal."

What atoms is the electrostatic motor to be made of if not metals? Bulk undoped diamond is an electrical insulator (ignoring some minor effects hydrogen passivation has on a diamond surface).

"For driving a nanofactory with a single big motor, electromagnetic is probably better."

But how do you drive a *big* diamond (or feel free to suggest another substrate) shaft electromagnetically without using Rare Earth magnets and coils of wire? And this is ignoring the requirement for closure in the electricity source itself (existing solar cell designs use heavy elements and existing wind-turbines use Rare Earth magnets).

I lack knowledge of some basic MEMs industry practices so maybe these are stupid questions. If MNT relies on Rare Earth magnets closure may require harvesting Nobellium and other elements from space. I don't have time at the moment to research constructive input; just enough time to be the Devil's Advocate for now.

[Phillip Huggan]

Another add-on to the above: if graphite is to be used (as a conductor), is the graphite to be a mechanosynthetic product or will it be brought in the nanofactory in bulk form from conventional graphite mineral deposits?

If magnets are an absolute necessity, it might do to substitute plastic magnets (possibly MNT-able?) instead of Rare Earth magnets. In either case working with weaker carbon atom bond strengths will be required and simulations will prove whether weaker-than-diamond C bonds are within the MNT library.

[Chris Phoenix]

Some buckytubes are excellent conductors, and even graphite is pretty good. I would expect the carbon-based conductors to be built by the nanofactory rather than fabricated from minerals.

Large electromagnetic motors, as far as I know, don't use any permanent magnets--both the rotor and stator use electromagnetic coils.

In my Primitive Nanofactory paper I address carbon-based solar cells. CVD diamond (which can be conductive) has been used in thermionic cells (I suspect buckytubes would also work), and a sun-concentrating lens (Fresnel or water-filled) could certainly be built out of diamond.

Phillip, a few days ago you wrote: "Drexler basically swept this under the table by saying "I hope we can use an electrostatic motor". This fundamental open question has been exposed by R.Jones and others; it is the primary reason why I think diamondoid MNT is only 1/3 to 2/3 likely."

I do not see availability of motors as an open question. It is one of the more certain things about a nanofactory.

Your discussion on this topic seems to repeatedly take the form, "I don't see an easy way to do X. I can see a hard way to sort-of do it. That hard way may not even work. Therefore MNT is unlikely."

Meanwhile, there is an easy way to do X that you didn't know about. I'm starting to feel like I'm debating with a skeptic who'd rather make his case for skepticism than understand how the system might actually work. We won't get anywhere that way.

Chris