A recent press release, describing a paper by Princeton theoretical physicists Rechtsman, Stillinger,and Torquato, begins with the stirring words “It has been 20 years since the futurist Eric Drexler daringly predicted a new world where miniaturized robots would build things one molecule at a time. The world of nanotechnology that Drexler envisioned is beginning to come to pass….” The mention of Drexler has ensured that the release got a mention on the Foresight Institute’s blog, Nanodot, but Christine Peterson disarmingly appeals for help in understanding what on earth the release is talking about. Fair enough, in my view; whatever one thinks of the Drexler reference, this is one of the worst written press releases I’ve seen for some time.
A look at the original paper, in Physical Review Letters, (abstract here, preprint here, subscription required for published paper) gives us more of a clue. The backstory here is the fact that collections of spherical particles in the size range of tens to hundreds of nanometers can (if they’re all the same size) spontaneously self-assemble to form ordered arrays, often called “colloidal crystals”. The gem-stone opal is a natural example of this phenomenon; it’s formed from naturally occurring silica nanoparticles, and its iridescent colours are a result of light diffraction from the crystals. It is these striking optical properties that have raised research interest in synthetic analogues; for some sets of parameters it’s predicted that these materials might have an “optical bandgap” – a range of wavelengths of light that can’t get through the crystal in any direction. This would be useful, for example, in making highly efficient solid state lasers. The problem is that most systems of simple spheres form close-packed crystal structures – of the kind you get when stacking oranges. But it would be useful if one could make colloidal crystals with different structures, such as the diamond structure, which have more interesting potential optical properties. In principle one might be able to do this by tinkering with the interaction potentials between the particles. Close packed structures occur because the particles simply attract each other more and more until they touch, at which point they resist further compression. What this paper shows is that you can design potentials to produce the crystal structure you want – perhaps you need the particles to attract to each other up to a certain distance, then softly repel until they get a bit closer, and then start to attract again until they touch. This is an elegant piece of statistical mechanics. Of course, having designed the potential theoretically you still need to design a system that in practise has these properties. One can imagine how to do this in principle, perhaps by having colloids that combine a tunable surface charge with a soft polymer coating, but such a demonstration needs a lot of further experimental work.
Is this really “turning a central concept of nanotechnology on its head” ? Of course not. It’s a nice step forward in theoretical methods, but it’s absolutely in the mainstream of a well established research direction for obtaining interesting ordered structures by colloidal self-assembly. And as for the next sentence – “If the theory bears out – and it is in its infancy — it could have radical implications not just for industries like telecommunications and computers but also for our understanding of the nature of life” – I can only hope the authors are cringing as much as they should be at what their publicists have put out for them.
Updated with link to preprint Tuesday 20.50.
There has been requests for the paper on the web.
I believe one can find it for free here:
http://www.arxiv.org/pdf/cond-mat/0508495
I will post my thoughts later!
An amateur mathematician.
Thanks very much for that. I’ve updated the main post with the link to the preprint.
Hi Everyone,
Clearly the paper is not about Drexler’s vision for Nanotechnology.
However, I believe for those of us who are enthusiatic about using Mechanical Engineering ideas in Nanotechnology, there is an important lesson to learn. That is of SYMMETRY.
It appears that for the first generation of assisted molecular systems, constructing highly ordered lattices and grids are the order of the day at around the 100 nm scale, and this paper is a large incremental step in that direction.
Lets suppose then that we are able to build lattices like in the above preprint. It is then possible to have simple molecules being attached to various points of these lattices leading to novel material and medical purposes.
IMHO, this should lead to radical materials for use in Space Elevators for example, or high density information storage or targeted medical systems.
Finally, Drexlerian Nanotechnology should not be ruled out yet, maybe someone could say solve Schrodinger N body equation for the coloumb potential effieciently, just like it is done for the Bethe ansatz. There are some very interesting ideas coming through regarding Random systems which could revolutionise Numerical Physics!
An amateur mathematician.
The author mentions colloids. I’m not aware of any of these smaller than around 1nm. Does this mean the “optimization process” desribed in the article will not function at finer resolutions?
The Space Elevator ribbon needs a minimum 45-50GPa tensile strength material, and it needs to be relatively “flexible”. These crystalline structures would probably fracture too easily, I’m guessing.
If a way of reversing the process could be discovered, it might permit an efficient hydrogen storage mechanism. I’ll be keeping an eye on the experiments based on this principle being conducted at NYU in the months ahead.
These colloidal crystals are weak mechanically – in fact they are so weak that if you weren’t careful you wouldn’t notice they were solids at all (think very lightly set custard). There’s nothing to stop the general principle of self-assembly being used for very small particles, but it’s difficult to see experimentally how you would tune the potential for particles smaller than a few tens of nanometers.
Incidentally, I think the experimentalist involved in trying to realise this concept in practise is Paul Chaikin, who is also at Princeton.
Actually, Paul Chaikin recently moved to NYU, where David Grier has begun a huge effort in soft condensed matter physics.
Thanks for that correction (my excuse is that Chaikin hasn’t moved his Princeton webpage yet). And I see David Pine’s moved there from UCSB too. Pine, Grier and Chaikin – that’s a pretty powerful combination!
Future nano-ethics committees should consider recommending mandatory caning for all instances of over-hyped nano press releases.
It appears that my vision has been shot down! I will try to think of suitable materials other than colloids
I was wondering if there are similar efforts in the UK regarding Soft condensed matter physics?
An amateur mathetician
Well, I think the UK has quite a good claim for having started the field of colloidal crystallisation, with the work of Ron Ottewill in Bristol maybe 15 years ago. Ron’s retired now, though his former student Paul Bartlett is still doing neat work in the area there. The Edinburgh physics department is perhaps strongest in this area, with the experimentalists Peter Pusey and Wilson Poon and the theorist Mike Cates making another powerful combination. Of course, if you want an introduction to the field here’s a highly recommended textbook on the subject…
I’m not a scientist but, when I read the release, I wondered what the hype was, too.
Thanks for giving me a “considered” view.
I’m adding a link to your post on the new NanoBlog I’m co-authoring.
~ Alex