Here’s a brief description of the second experimental project generated by the Ideas Factory. As before, this is not yet an officially announced and approved EPSRC project; various administrative steps remain to be completed, including a more formal costing. Again, we anticipate that this project will receive somewhat less than half of the £1.5 million available for the ideas factory.
We propose a scheme to revolutionise the synthesis of nanodevices, nanomachines, and, ultimately, functional materials via the 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. Our proposal rests on the parallel development of novel surface-bound, reconfigurable nanoscale building blocks (molecules, functionalised clusters, nanoparticles) and a prototype computer-controlled matter manipulator best described as a nanoscale conveyor belt. We focus on the generation of two major and immensely challenging functionalities for positionally-assembled nanomachines: switchable energy transduction and conformationally-driven motion. Our archetypal system comprises the following units: an energy harvester, a switchable/gateable link, and an optical or mechanical output. By arranging, configuring, and triggering these fundamental units our long-term goal is no less than the fabrication of an autonomous, abiotic nanomachine.
The project is a collaboration between Ras Raval (Chemistry, Liverpool), Lee Cronin (Chemistry, Glasgow), Philip Moriarty (Physics and Astronomy, Nottingham), Jeremy Baumberg (Physics, Southampton), Guenter Moebus (Materials, Sheffield), Robert G. Jones (Chemistry, Nottingham), Ashley Cadby (Physics, Sheffield), and Tom Grimsey (Fine Art and Sculpture, University of Brighton).
Software Control of Matter 
“Our archetypal system comprises the following units: an energy harvester, a switchable/gateable link, and an optical or mechanical output”
I don’t know how useful a novel optical output would be, unless the light is being used to drive a computer or something. Tuneable quantum dots have already been invented.
I’ll offer the advice that the bigger the initial building blocks chosen are, the quicker a profitable/useful industrial process will be generated. On the flipside bulkier parts enable a smaller product library than, say, hydrocarbon molecules (and are more expensive); are less likely to produce a nanomachine manufacturing system with good parts closure. Something to keep in mind if more funds and research follows this project.
I’d be very interested to know the reservoir molecules being considered and the substrate of the reservoir surface (assuming it is a surface).
I suggest that the question to be answered, with Phillip’s points well taken, is where does the economy of scale begin, and the overall benefit become more important. To use marketability as the primary motivation, while, perhaps a correct approach of one is building autos, might be less of a motivation when considering the overall market for this project.
The fact is, as I have found it, if a project has enough ‘Jazz’ to catch the eye of even a small portion of the market, it will be successful. For myself and the motley crew I work with, what has been demonstrated here thus far is initiatives with incredible foresight, broadly mixed participation bases, and well found knowledge, there is no reason to doubt their success in the long term.
So my respectful counsel remains, go for the best of the smallest.
Please refer to my comment in ‘The end – or the beginning’
I have to admit it’s taking me some time to understand this proposal.
Surface-bound molecules should be easier at cryogenic temperatures.
And optical outputs don’t sound useful, but I suspect they mean plasmonics, not optics: energy packets of photon quantities (and sometimes qualities) but confined to sub-wavelength “waveguides”. That energy could be used to drive actuators coupled to the outputs.
And perhaps plasmons could be delivered, switched, and used more easily than other forms of energy. Here’s a weird analogy, and I’d appreciate a physicist telling me if there’s anything to it. Electrical transformers trade between current and voltage. They need AC, not DC: you have to have a varying magnetic field. Electricity at the nanoscale is basically DC; even GHz frequencies are slow. But light is more like AC, and may be easier to interconvert. My impression is that plasmons don’t actually use the magnetic field, but I’m thinking the rapid variance/resonance of the electric field, well-coupled to propagate from one bit of matter to the next, might be analogous to the propagation of magnetic fields through the transformer core to the next bit of wire.
Back to the proposal: Does the precision along the actuator-driven trajectories only mean that while moving on the X axis, the Y deviation will be controlled/predictable with sub-nm precision? Or does it mean the system will take sub-nm steps along the axis of motion? The latter would be very very cool, since it would imply that combining actuators could yield a very rich trajectory space. But I’m guessing it’s the former–still useful for driving one reaction per mechanical system, but making each system less general-purpose.
I could imagine a “conveyor belt” being implemented with a line or field of “spikes” that change some property upon being “actuated”, making them less or more attractive to the item being moved, similar to the water droplet that was recently moved uphill by the hydrophobic/philic switching of the surface (by laser, IIRC). This might allow precision steps, but each step would be the size of the spike spacing–hard to get finer control unless the item covered several spikes. The switching-power system could then be used to modify individual spikes.
For mechanical, you could again use an array of “spikes” that bent back and forth. If several spikes bent toward a point, the concentration of matter would increase surface forces in that area. Swing spikes in toward the zone of attraction on one side, away on the other side, and you move the “hot spot”, and the item moves with it.
Or perhaps it’s an optical-tweezer-like phenomenon: you don’t change the surface properties, just the radiation flux coming through different points. IANAPhysicist, but I’m wondering… optical tweezers use induced dipoles, surface forces use induced dipoles, perhaps sub-wavelength light emitters create high local forces? I don’t know, but I suspect a plasmon must create a strong dipole as it goes past, which might attract nearby dielectrics.
I still don’t know where the block-joining fits in with this. Of course precise block-joining and bond-forming is very cool however it’s done. Even if it’s just a conformational change in a single object–a C turning into an O–it’s still useful for research. Especially if you can put different molecules in the “jaws of the vise.”
So that’s as far as I can get with this proposal. I will say that switchable energy transduction coupled to actuation will be *very* useful at some point. For the past few months I’ve been thinking that structure was progressing very fast while actuation was rather lacking. When we get the ability to build arbitrary intricate structure–whether by lamp-worked buckytubes in SEM or Rothemund’s DNA–we will need a way to add multiple addressable nanoscale actuators, and we don’t really have one yet. So this proposal sounds like it should provide at least one very useful enabling technology, as well as getting others thinking about other ways to do switched actuation.
But I don’t yet understand this proposal enough to know how tightly integrated the actuation is with the motion, and the motion with the chemistry. Without knowing that, I can’t say whether it’s merely useful, or fully game-changing.
Chris
More thoughts: How will the switchable energy transduction be switched? Can the actuators drive the switches? That could be a pathway to computation.
Can the items you’re moving around drive the switches? It’d be *very* cool to e.g. bring two blocks together, then when they bond, that electronic change directs the plasmons to a different set of actuators that moves the blocks somewhere else, and when they get there the system switches again to load two new reactants… errors are detected and shunted in a different direction…
Is there a plasmon analog to frustrated total internal reflection (FTIR)? (Perhaps surface enhanced Raman spectroscopy (SERS) is the analog?) And if there’s nothing on the surface, then the plasmons come out the other edge…?
Can you do interferometry with plasmons? Use single plasmons to detect whether two channels are the same or different?
OK, I’ll make a strong guess at this point that the computer-controlled linkages are some kind of photochemistry. Hm. That’s great if you can deliver the energy in near-field. But if a photon escapes, it could cause linkages at whatever random place it couples to. Unless you can keep the blocks separate until you’re ready to join them. Hm, use low-frequency light to move them, so as not to trigger the photochemistry; then send higher-frequency light when you want them to join. But can you send multiple photon frequencies over the same plasmonic “wiring”? My guess is “yes–sometimes”; but IANAP.
I’m wondering how literally to take the phrase, “thereby achieving structures not accessible by mimicing natural assembly strategies alone.” Does “natural assembly” mean “pure self-assembly,” or “any strategy found in biology?” The former would be good–the latter could be *extremely* interesting.
I’m glad the building blocks are reconfigurable.
I’ll stop here and hope someone will be able to give me some hints…
Chris
Chris, there’s a lot in this proposal and the description is admittedly very compact. “natural assembly” refers both to strategies both of self-assembly and classical synthetic chemistry; I’m not sure whether this excludes all biological strategies, but it should be clear that this is not an explicitly biomimetic proposal. Plasmon physics is indeed very subtle, particularly for surfaces with complex nanoscale structure; SERS has some analogies to attenuated total reflection spectroscopies in that it gives you surface sensitivity for normally bulk spectroscopic techniques, but the physics is a little different and leads to astonishingly high enhancements in sensitivity. Jeremy Baumberg knows lots about this (indeed he has a spin-out company for sensors based on surface-enhanced Raman).
Hi, Chris.
Thanks for your long list of comments – the feedback is very helpful. I’m going to attempt in the following to address some of your questions while retaining a level of “mystique” about the proposal. There are many people, and thus many ideas, involved in the project so I don’t want to provide information that others in the consortium might feel uncomfortable about releasing. I therefore apologise if my answers are sometimes vague and nebuluous – I know that this may be frustrating.
In addition, it’s worth noting that, as pointed out by Andrew Turberfield in a previous post, the sandpit generated many other exciting ideas which will very likely result in other collaborative proposals to be submitted through the “standard” Responsive Mode (i.e. investigator-driven) route to funding. (I’ll state for the record that I am a very strong supporter of Responsive Mode funding!).
Surface-bound molecules should be easier at cryogenic temperatures
We ultimately plan to work in a range of environments. Cryogenic temperatures are not essential for the type of experiment we’re envisaging: it’s a question of the appropriate choice of molecule/nanoparticle, functional groups, surface, and – in liquids – solvent.
And optical outputs don’t sound useful, but I suspect they mean plasmonics, not optics…
Our initial goal relates to energy transfer through a complex which has been assembled via computer-control of the trajectories and interactions of the component parts. Optical detection of the output need not necessarily be plasmonics (e.g. nanoparticle waveguide)-driven.
Does the precision along the actuator-driven trajectories only mean that while moving on the X axis…the latter would be very cool…
Playing my cards close to my chest, the key ingedient is, as you suggest, “the richness of the parameter space”. This links in with a comment you make later – we want to explore and exploit a trajectory and position “phase space” which is very different to that used by Nature (although we’re not ruling out natural assembly strategies).
Can the actuators drive the switches? That could be a pathway to computation.
I don’t think I’m giving too much away to state that we certainly envisage that the actuators can drive the switches!
Can the items you’re moving around drive the switches?
Yes. Triggered interactions are of interest.
OK, I’ll make a strong guess at this point that the computer-controlled linkages are some kind of photochemistry.
Near-field photochemistry is one way of doing it but there are others…
OK, I’m probably not being very helpful at this point! You raise some very interesting points and I’m disappointed I can’t cover them in detail. I’ll keep you informed about developments in the project.
Thanks again for your interest.
Best wishes,
Philip
“(P.Moriarty wrote:)
In addition, it’s worth noting that, as pointed out by Andrew Turberfield in a previous post, the sandpit generated many other exciting ideas which will very likely result in other collaborative proposals to be submitted through the “standard” Responsive Mode (i.e. investigator-driven) route to funding. (I’ll state for the record that I am a very strong supporter of Responsive Mode funding!).”
About all I know about funding is that at the end of almost every scientific paper there is a thank you to the NSF for funding grant 23534-78. How is this Idea’s Factory different standard funding-experimental feedbacks? Is it because such a high value target (modular chemistry) has been identified, that it is deemed a priority to try to “force” the target rather than wait for researchers to stumble upon it? Or maybe it is just the think-tank like atmosphere..
The reason I ask is that there may be many more “targets” in the future that could benefit from such a strategy, if it works. I wonder if UK or other governments will adapt this model for other applications (Konarka/Nonosolar printable solar cells would have been nice to have a decade ago; still waiting on a good energy storage technology).
Most science funding (at least in the countries I know about) is associated with individual scientists or pre-existing groups of scientists – either what is called in the UK responsive mode (i.e. scientist has idea, writes proposal which is peer reviewed and considered by a discipline-based committee), or bigger block funding associated with a group of scientists or a centre. I think that, for example, NSF does most of its business this way. In the UK, one also has “managed programs” (which Philip M. doesn’t at all like)in which some area is selected and proposals, often from larger collaborations, are invited in that particular area. We are likely to see in the UK in the near future much more funding attached to societal challenges rather than subject areas (and alternative energy is going to be a big one, certainly). The Ideas Factory concept I think is unique in that forming the collaborations and developing the ideas is a big part of the event, which of course carries quite a lot of overhead in terms of time and money. It’s been used mainly in areas that are so interdisciplinary that it’s difficult to imagine the collaborations forming without outside help. I believe that it’s a method that is currently unique to the UK; it’s an experiment that’s been going a couple of years now.
Richard – What has been heartening about this whole process is the apparent non-competitive collaborative nature of it. Each project was judged on its merits at the end of the day with the diversity of the proposals clearly taken into account. One benefit there is in the UK is given the size of the country, the distances which need to be traveled for collaboration are not prohibitive and thus do not lap up large portions of the project budgets.
The whole concept of the Ideas Factory – Sandpit has rapidly become a model I wish to implement within my own sphere of influence. The notion of bringing minds together on neutral territory, while clearly, with an agenda, has an appeal worth exploring.
While there is much to support the potential collaborative nature of the Highway of Light, I would suggest that getting all the players in a room plied with ‘Heroic’ quantities of Caffeine is a bonus.
You say – ‘The Ideas Factory concept I think is unique in that forming the collaborations and developing the ideas is a big part of the event, which of course carries quite a lot of overhead in terms of time and money. It’s been used mainly in areas that are so interdisciplinary that it’s difficult to imagine the collaborations forming without outside help. I believe that it’s a method that is currently unique to the UK; it’s an experiment that’s been going a couple of years now.’ – I suggest it should become a standard.
I posted awhile ago in ‘Getting going. . .’ – ‘While the house looks familiar, I am reminded that the house I was once associated with was in fact near Bristol, not Southampton. I mention this because it was back in the early 70’s when there resided an Ideas Factory of a different sort, where work on projects was progressive and whenever something was needed, it would just appear, usually by the ‘Lorry Load’.’ – It is interesting how history repeats.
This sounds vague but strikes me as being more MEMS-like than “nanotechnology”. However, even if this is unsuccessful for true “nanotech” application, the results of this research could certainly be used to make semiconductor and other micro-structured materials and devices without the expensive fabs and vacumn process equipment that are used today. This would be revolutionary in and of itself given that a new fab is nearing the size of a football field (American) and costs $2-3 billion.
Philip, 1) Thanks for telling us what you can; 2) You have in fact been very helpful. I think I was mis-reading a key point of the proposal. I had thought that “the generation of two major and immensely challenging functionalities” meant that those functionalities (the switch and the actuator) would be required as components of the proposed mechanism. Instead, it sounds like those are your first planned products.
No wonder I was having trouble figuring out how to make a conveyor belt out of switched optical _or_ mechanical output! I hope that my speculations were at least fun to read.
Well, so it sounds like you’re planning to build a system that can put together enough building blocks to create a powered, computing, moving device. The “conformationally driven motion” that I had thought was part of the fabrication system… really means that the device you build will be able to swim or crawl away, right?
Building a nanomachine with multiple functions and composed of multiple parts is certainly a bold goal at this stage. Building a system with various degrees of freedom/trajectories and sub-nm precision that is capable of chemically joining blocks is also bold and interesting. And it sounds like you’re planning a system that has sufficient physical integrity and stiffness to transmit physical force and motion internally. As proofs of concept, these will be extremely useful.
As enabling technologies for exponential molecular manufacturing, they sound good, but I’d need to know somewhat more in order to say for sure. If the conveyor belt actuators are large (e.g. piezo or MEMS), and the product nanosystems are clearly less functional than the system that built them (fewer DOF or less precision with no easy path to improvement), then I’m not sure how useful this would be for my goal of recursive NEMS-building-NEMS. On the other hand, if the system lends itself to recursive construction, then this could be huge. If you’re expecting that the actuators you build will be able to modulate the switches you build, and that the actuators will be able to transmit their motion through built structure, then that doesn’t sound too far away from ratchets and levers…
Chris
Richard, I want to thank you for your generosity in sticking around and answering such a diversity of questions even after the process is completed and you were probably expecting to get back to your work. Of the questions in this comment, one is perhaps off-topic and the other is perhaps not very innovative (so I could learn the answer on my own). Please don’t feel obligated to answer either one.
Last week I sketched out an idea for a low-cost, ongoing, creative process for interdisciplinary generation, evolution, and evaluation of ideas. Basically, it throws scientists together in very small groups by phone/web for an hour or two, during which they are asked to develop a new interdisciplinary idea and rate a previous idea from a different team. As long as you’re talking about the Ideas Factory’s focus on forming collaborations and developing ideas, I’d be interested in your initial brief reaction to my idea.
When I mentioned FTIR, I was actually thinking less of comparisons with other forms of spectroscopy and more of Jeff Han’s amazingly cool MultiTouch display/input (search multitouch on Youtube). With MultiTouch, you bounce light through a sheet of plastic edgewise; when you touch the plastic, the light couples to your finger, illuminates it, and creates quite a bright spot that can be picked up by a camera looking at the back side of the sheet.
So I was wondering if there was a way to send a plasmon along a waveguide, and when something touched the waveguide, create a “bright spot” at that point. Of course the properties of the spot might be used for spectroscopy, but they might also be used for switching or position sensing. It sounds like this is sort of what SERS does–the particle couples the plasmon out of the surface, creating a very “bright” spot right under the particle, which makes the technique very sensitive.
Thanks,
Chris
Kurt9 – there may be some MEMS like aspects to the fabrication strategy but the building blocks are definitely nanoscale, whether molecules or clusters.
Chris – I’m nervous about speculating too much about plasmons because it’s a bit of physics I’m conscious that I have virtually no reliable intuition for. However my understanding is that close to a metallic nanoscale protruberance you get a very big enhancement of electric field. It’s not entirely like classical evanescent waves, because the length scale over which the latter decay is characterised by the wavelength of the light (i.e. half a micron or so), while in the case of SERS the lengthscale over which the field is enhanced is comparable to the lengthscale of the protruberance. I believe this is the principle that underlies the way apertureless scanning near field optical microscopy works – you shine a laser at a silver coated AFM tip, and there’s a huge enhancement of field intensity right at the tip, which decays over a distance comparable to the tip radius, which can be very much less than the wavelength of light.
Your thoughts about on-line ideas generation were interesting, and relevant to reflections I’ve been having about the ideas factory process. I’m all in favour of not leaving home – scientists end up having to travel far too much, which is bad both for our families and for our own research groups. So we really ought to try and get better at doing things remotely using all the web and telecommunications tools at our disposal. On the other hand, though, the process of the Ideas Factory was intensely social; it’s not easy and there are lots of problems related to things like trust and the clash of big personalities in getting very different scientists to work together – the more so since the culture of science does tend to be individualistic and competitive. It seems more possible to work through these issues face to face, though that isn’t always that easy or comfortable either!
“(C.Phoenix)
No wonder I was having trouble figuring out how to make a conveyor belt out of switched optical _or_ mechanical output! I hope that my speculations were at least fun to read.”
Don’w worry Chris. I’d never heard of Plasmonics before your posts. Now I am aware of another useful nanotech process.
Phillip, I am not sure how useful plasmonics is… yet. My impression is that it’s still in the stage of searching for useful abstractions and rules of thumb; it has a lot of possibilities, but people don’t yet know the shape of the field or the structure of useful simplifications/theories, and experiments are still closer to observation/exploration than to engineering (though I’m sure they’re rapidly moving toward engineering).
In April 2004, I noted that Google does not know the phrase “plasmon logic.” Today “plasmonics” has 66,000 hits, but “plasmon logic” still doesn’t exist.
Hm, here’s a wild question: Has anyone observed self-interference in plasmons? Something like a two-slit experiment or a SQUID? Could this be the basis of a computing device–perhaps even a quantum computing device? (Of course, if plasmons don’t preserve quantum state, then this question falls apart, but my impression is that they can–e.g. I think I’ve heard that a polarized photon going through sub-wavelength tunnels in silver may stay polarized, and Wikipedia says plasmons are fundamentally quantum although they’re usually not treated that way.)
Back to this nanomachine experiment: If the fabrication machinery may contain MEMS-like aspects, is this intended to be scaleable to grams of production, or is this more a pure-research tool?
And a followup question about online work, scientist collaborations and issues of ego/trust/etc: Is that something that improves when it’s just two scientists talking off-the-record (with a facilitator)? If the answer is “it’s very individual,” then the iterative structure of the process could allow the facilitators to quietly weed out the scientists who couldn’t easily cooperate in the setting of the exercise–though I suspect they’d remove themselves after one or two unproductive sessions.
Chris
Over on our blog, Philip said, “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 [the Responsible Nanotechnology] blog) that the paragraph represents what is going to be achievable in two years.”
http://crnano.typepad.com/crnblog/2007/01/ideas_factory_m.html#comment-28082941
I had been assuming that the published description was describing the project to be funded, and (based on earlier comments on this blog) that the results were expected to be achieved in 3-5 years.
Could someone please clarify the relationship of the published description to the plan and goals of the funded project, and the timelines involved?
Perhaps the final two sentences of the description starting with “Our archetypal system…” were intended to describe the post-project purpose of the manufacturing system, not the goals of the funded project? If so, then I did misunderstand that point and would correct our posting in that case.
Many thanks,
Chris
Chris, the precise timeline is to be decided by the team in the formal proposal, but 3-5 years is certainly the range we expect, depending on how much can be done in parallel, to see at least proof of principle of the main thrusts of the proposal.
“I don’t know how useful a novel optical output would be, unless the light is being used to drive a computer or something. Tuneable quantum dots have already been invented.”
Came across my own comment in a recent Google search and it makes me shudder now. I’ve since learned novel (silicon) optical “utilities” are a big driver in improving biomedical treatments, among other uses. I was so brainwashed conquering-the-world (diamond) aerospace advances were all that mattered that I never considered this research could be used to do something productive.
“Computer-directed actuators will be used to drive (with sub-nanometre to sub-Angstrom precision)…”
I’m curious about this qualifier. It portends the velocity of the scale-up from research to industrial utility. I don’t know much about the cost of various feedstock molecules and “treating” them so they can function in the nano assembly line. But in general, the smaller the molecule the cheaper the feedstock (I would think). Of course, it is way easier to grap hold of a 50x10x10nm diamond shard than it is an adamantane cage.
I really like your blog.. very nice colors & theme. Did you
design this website yourself or did you hire someone to do it for you?
Plz answer back as I’m looking to create my own blog and would
like to know where u got this from. many thanks