How might one design a nano-machine?
Save This Post as PDF
Trinh Xuan Hoang is currently a postdoc at Penn State University. He is also a researcher at Institute of Physics, Vietnamese Academy of Science and Technology. Dmitry.
Significant advances in laboratory techniques in tailoring and processing materials at the atomic level have resulted in nanotechnology becoming an increasingly mature field. One of the exciting goals of nanotechnology is the design of powerful nano-machines, i.e. functional entities at the nano-scale that work like macro-world machines. A simple nano-machine would be an entity that is able to switch between two distinct conformations under some kind of external perturbation. In fact, molecular switching of various kinds has been the subject of many recent studies.
In our recent paper, my collaborators and I have asked what the basic principles are underlying the design of a nano-machine that is capable of molecular switching. Our starting point is based on the lessons learned from proteins and liquid crystals. Proteins are amazing molecules that are machineries of life. They perform many functions in our organisms (catalysis, transportation of oxygen, participation in the immune system etc. are just a few examples). Proteins are chain molecules built from 20 kinds of amino acids and have well-defined three-dimensional structures. Proteins can undergo a change in conformation upon binding to a substrate.
Liquid crystals are very useful materials for technological applications. The liquid crystal phase opens up between the crystal phase and the liquid phase, and is very sensitive to external perturbations. This very interesting phase of matter arises primarily from the anisotropy of the constituent molecules.
Back to our design problem, assume that we have a set of spheres and a set of rules for the interaction between them. Like in a “lego” system, there are numbers of ways one can build a machine from the spheres. Our goal is to build a machine that can exist in two well-defined geometries and is able to switch between them. Our design is armed with the insights from the behaviour of proteins and liquid crystals. First, like for proteins, we want our machine to be a chain molecule. This can be easily done by tethering the spheres in a linear chain. We consider also side spheres attached to the main chain in a specified direction, mimicking the side chains in proteins. Second, like for liquid crystals, the molecule should have an inherent anisotropy associated with its building blocks. This arises spontaneously in a chain molecule and can be accentuated by making the consecutive spheres overlap. The entity formed by two overlapping spheres has overt uniaxial symmetry instead of the spherical symmetry of each sphere. For simplicity, we use only two kinds of spheres, one for the main chain and the other for the side spheres (see Fig. below).
Fig. 1: Sketch of a chain molecule. The main chain is modeled as a chain of (blue) spheres. The nearest neighbor spheres along the main chain are allowed to overlap with each other thereby overtly introducing uniaxial anisotropy. 
and 
are the tangent and normal vectors assigned to each sphere, 
, except those at the ends of the chain.
Side-spheres (shown in pink) are attached to the main chain in the negative normal direction. The side-spheres are not allowed to overlap with either the main chain spheres or with each other.
By tuning the sphere radii and the range of attraction between the main chain spheres we showed that a machine built of just 30 spheres has an interesting phase diagram with robust ground state conformations such as the single helix and the twin helix. The single helix and twin helix phases exist between the random coil phase and the compact globule phase. Our Monte Carlo simulations showed that this machine/molecule is able to switch between the single helix and the twin helix conformation by thermal fluctuations (see movie).
The main lesson from our paper is that it behaves a designer of nifty machines to consider which phase of matter it occupies. The power of a machine derives from the choice of the appropriate phase. Taking a lesson from nature, we outline how we can exploit the same phase that has been used so successfully by living matter.
Further details and references can be found in our paper:
Jayanth R. Banavar, Marek Cieplak, Trinh X. Hoang, Amos Maritan. First-principles design of nano-machines. PNAS 106, 9601 (2009); arXiv:0904.4037
Related posts:
Google Wave
Everybody (Terence Tao for one) seems to be excited about forthcoming Google Wave, and so am I. Here is the video demonstrating some of the product’s features:
I think, we are yet to see whether Google Wave is to become ultimate science collaboration tool (I signed up on their site – and hope they’ll get me into beta testing). My current opinion is that Google Wave provides you functionality similar to forums rather than wikis: in collaboration projects, I’ve found that messages tend to group into project categories, not conversations – since it is also good to see also conversations which ended up long time ago, not just recent ones, if the topic of the conversation is the same. And of course, I would love if a collaboration platform naturally would support TeX formulae (embed them as fugures, MathML or in some other – not terribly ugly – way) 
(nobody among big players seems to be interested to satisfy needs of little egghead nerds – scientists)
I fell in love with two features of GWave so far: the possibility to embed waves on arbitrary webpages and the way it shows updates in a document you are collaborating on (25-30 min from the beginning of the video).
Also, did you notice that GWave interface resembles somewhat the one of Microsoft Outlook 2007? 
Update: As Terence Tao explains,
Apparently a LaTeX renderer is already being developed as an API extension to Google Wave.
His screenshot looks impressive enough.
Related posts:
The coming collapse of the middle class
As usual on Saturdays, discussion of physics is forbidden (why? check out Old Testament). Let us talk a bit about global financial crisis instead , namely about work of Elizabeth Warren, professor of Harvard Law School.
Elizabeth Warren was a senior consultant of Clinton’s National Bankruptcy Review Commission – the one which tried to figure out why bankruptcy rate in US is nowadays higher then it was during Great Depression (and continues to grow). As Warren has found, today the total wealth of a family in US is way less than it was even 30 years ago, and naturally nowadays
a family cannot afford not to borrow.
On the other hand, unregulated financial instruments featuring growing interest rates and multiple fees simply kill households – giving rise to increasing rate of household bankruptcies. Who knows how much subprime credits contributed to the global financial crisis takeoff? At least, they contributed to the housing market bubble birth, that’s for certain. As we learned, modern economics cannot exist with banks unable to pay, but can banks really exist if households are unable to pay?
Below is the video of her lecture at UC Berkeley, with the title “The coming collapse of the middle class”, you’ll learn from it what awaits us in the future according to Elizabeth Warren:
Related posts:
Workshop on tests of gravity in Case Western – day 2: aether and modified gravity
Let me finally briefly review the reminder of the second day of the workshop.
Justin Khoury (whom I knew from Perimeter years and who is in Penn now) gave the first talk afternoon – titled “observational hints of IR modified gravity”. His talk followed Nima’s, and the latter almost completely blew me away, so I was somewhat unfocused during Justin’s presentation. Yet, I was able to capture that as such observational hints he wanted to present local bulk flow of matter within 50 
MPc scale, excess power in Lyman 
(about 30%) and small scale CMB anomalies (which I would hardly call anomalies due to lack of statistics there). According to him, all this shows that gravity should be stronger at larger scales – and that’s exactly what many models of modified gravity predict.
Next, Stacy McGaugh has talked about MOND. I should really confess that my attitude towards MOND was always somewhat arrogant – I had no idea why people might be interested in such an ugly theory. This time I got it, I guess: in order to fit all galaxy rotation curves, you only need to pick one and fit the MOND parameter (acceleration 
) – and you’ll automatically fit all others, that is, an almost infinite data set turns out to be one-parametric. Well, I guess, choosing the density of cold dark matter 
will do the job as well.
Nima Arkani-Hamed has asked him about the bullet cluster and Stacy’s answer was like “I find it somewhat fishy that all proponents of cold dark matter model always use this example to defend their models”. Well, that’s the point – they use this argument, because it works.
The first talk after lunch (by Ted Jacobson) was devoted to the discussion of Einstein aether theory. What is the theory possessing such a scary name that would probably create an itchy feeling in guts of every 1 year physics student?
Consider a theory with the action
,
where
,
is a time-like unit vector (since 
is really lagrangian multiplier). The name of the theory involves the word “aether”, since a “4-velocity” vector (velocity of aether) introduces a preferred direction in spacetime (Lorentz transformations along and perpendicular to it are different). It therefore belongs to the class of exciting deeply flawed theories according to Nima’s classification, and not surprisingly the theory messes with BH thermodynamics – no unique Hawking temperature or BH entropy has been found so far and generalized second law seems to be violated (that is, you can construct perpetuum mobile in the universe described by this theory).
Since Ted Jacobson has mostly discussed classical properties of the theory, Nima Arkani-Hamed immediately pointed out that if one turns to the quantum level, the theory above features a non-perturbative mass scale – since it looks similar to non-linear sigma model.
My personal comment to him: actually, this is true that lagrangian multiplier acquires a non-zero VEV due to non–perturbative effects, however, its fluctuations are only suppressed if the vector has many components (namely, inifinite number of them), while in this case 
, and lagrangian multiplier fluctuates strongly.
Levon Pogosyan has talked about CMB and LSS observations – are we able to determine from them whether gravity is modified or not? As it turns out, our data are particularly sensitive to scale dependent modifications of gravity, and future surveys such as LSST might give us more information that just effective EOS for dark energy 
.
Finally, Mark Wyman has discussed N-body numerical simulations of DGP. The result he presented is well understandable: since DGP means stronger gravity at larger distances (and therefore earlier times), in DGP large scale structure is formed faster and its features are stronger than in GR (see the Fig. below)
Related posts:
Susskind’s general relativity – lecture 9
… where Leonard Susskind discusses spacetime – spacelike, timelike and lightlike directions, explains how one gets special relativity from general relativity (post-Newtonian approximation), non-relativistic limit of GR and finally … Einstein equations (hurray!)
Related posts:
Did not find what you were looking for? Try
Loading ...
