Skip to Content

Chemical News

Indistinguishable From Magic?

What would you say if I told you that a reaction rate could be influenced by the shape of the container the reaction is being run in? Those of you who do larger-scale reactions will have no trouble believing it, since mixing effects start to become prominent in larger vessels. But that’s not where I’m going here: what if I told you that a particular simple reaction ran five times slower, compared to the bulk liquid, when it was confined between two walls that are six microns apart? How about that system still running five times slower than the one where it’s being confined between two walls that are eight microns apart?

Right. But we’re going to have to get our heads around this, because this paper demonstrates that it’s exactly what happens, and for a good reason, too. This is a followup to some work that I blogged about here, from the Ebbesen group. They’ve been investigating the intersection of organic chemistry with physics, in the form of the coupling of vibrational modes with the vacuum state and inside optical cavities. In this case, it’s a Fabry-Pérot cavity (two parallel reflective surfaces), and the reaction is the deprotection of trimethylsilyl phenylacetylene, using TBAF in methanol. That’s a simple one, and well-suited to the experiment, since there’s a prominent C-Si band in the vibrational (IR) spectrum that can be targeted. The liquid reactants are mixed and injected into the optical cavity, and the reaction is monitored by FTIR. For controls, they use the bulk reaction, and also in another cavity whose spacing is not tuned to that particular vibrational peak.

A Fabry-Pérot cavity has a regular series of resonance modes, depending on both the cavity size and the refractive index of the material inside it, and this one gets tuned so that one of these lands right on top of that C-Si peak. In this situation, you have vibrational strong coupling, VSC, which splits that original peak into two hybrid states, separated by Rabi splitting.(Dang it, I told you that this was physics, but fear not, we are not going to go on to Consider the Hamiltonian). That splitting is proportional to the strength of the original absorption, so the more concentrated you make the sample, the bigger an effect you should see. As long as the starting material(s) and the product have slightly different refractive indices, you can monitor a product peak by IR without affecting anything about the reaction itself. The thing is, all this VSC and Rabi splitting happens whether there’s any IR hitting the system or not – it’s a consequence of an interaction between the optical and vibrational zero-point energies, fundamental properties of the molecules themselves.

This system works exactly as predicted. The reaction rate changes noticeably, going down about fivefold inside the cavity, and it depends on the Rabi splitting just as expected. Running the reaction at different temperatures and breaking down the thermodynamics shows some interesting stuff: the transition state of the reaction has clearly changed (both the enthalpy and entropy terms show large differences). Normally, this reaction goes by attack of the fluoride on the silicon, forming a pentavalent species, but under VSC conditions, it looks like the carbon-silicon bond starts breaking first, a dissociative mechanism rather than an associative one. There’s actually a higher energy barrier in this new landscape, which is why the reaction slows down.

What’s also weird to think about is that this effect depends on the orientation of each individual molecule with respect to the cavity. At any given instant, only a given fraction of them are really experiencing the strong coupling, so the rate change of the whole sample is actually an averaged-out value of what (for the molecules involved) must be a much greater effect. A number of things might be possible in other systems:

While we found that in our case, the reaction was slowed down, it is possible that depending on the chemical landscape, reaction rates can also be accelerated. When the reaction leads to multiple products, it is likely that the product ratios are modified under VSC, providing a way to optimize the yield of a given product. Site-selective chemical reactions are another possibility that should be explored. Finally, chemistry under VSC has the advantage that it works in the dark and at room temperature.

Picture the landscape of organic chemistry as we know it, all the reactions and catalysts that we have now – and then picture being able to twang the rubber-sheet landscape of all their thermodynamics, altering ground state and transition state energies just by running things between properly sized gaps. I’m imagining a (rather expensive) flow machine that sends its reaction mixtures through precisely formed optical cavities, each of which is tunable to the vibrational modes of the particular bonds in the particular molecules you’re trying to affect. Not all of those are going to be accessible (the Rabi splitting has to be large), but enough of them surely are so that entirely new reactions are possible. What a thought – Clarke’s third law, for sure.

Update: here’s another post on this work, worth a look since it’s hard to find comment on this area of research.

39 comments on “Indistinguishable From Magic?”

  1. Curious Wavefunction says:

    Interesting. This reminds me of how chemistry can change when done on the surface of liquids, or how the phenomenon of dewetting in protein cavities or between two hydrophobic surfaces can displace water and change binding affinity. Basically what seems to be happening in these systems is that you are acquiring control over the behavior of individual molecules, creating a difference in the behavior of these molecules and the rest. It’s a very cool domain of scale to explore further.

  2. Me says:

    Plenty of papers make me think ‘WTF?!?’

    Sometimes it’s bad WTF, sometimes it’s good WTF.

    This one falls firmly in the latter.

  3. Anon says:

    “What would you say if I told you that a reaction rate could be influenced by the shape of the container the reaction is being run in?”

    I would expect *any* reaction that occurs principally at the surface of a reaction vessel to be heavily influenced by shape — or at least surface area to volume ratio — whether that is due to interaction with the solid surface, or exposure to UV light, for example.

    Why do you think leaves are flat???

    1. Anon says:

      PS. Conversely, if being close to the solid surface of the vessel slows the reaction then the reaction will be faster in bulk liquid. For example, the solid surface my polarize reagents so that they are less reactive, or perhaps they just cool down more quickly near the surface…

      Why do you think warm-blooded animals living in cold climates are *not* flat?

    2. Marlon Blando says:

      all leaves are not flat. you’ve left out all the gymnosperms. that’s a whole lot of non flat leaves there partner

      1. Anon says:

        Well, besides any bulk material required to store water (e.g., in cactuses), at least the part that does photosynthesis is always flat.

        1. Matt says:

          I believe Lithops conduct photosynthesis in the interior of the leaf. Biology hates absolute statements.

    3. Gerben van Straaten says:

      This is a reaction occuring in the bulk though and the container is nowhere nearly small enough for surface effects to take over

  4. Anon says:

    Incredible.

    Can’t help but appreciate the irony of advocating performing chemistry in the dark after so much chemistry performed in the light!

  5. Tuck says:

    Sounds like this could be a major factor in living systems too… Lots of properly-sized gaps in an organism.

    1. anon says:

      But none with mirrored surfaces.

      1. kjk says:

        Heard of structural coloration?

        https://en.wikipedia.org/wiki/Pollia_condensata

        If biology can do this to visible light, why not infrared?

        1. anon says:

          Yes, of course I’m familiar with the phenomenon of interference colors, which is a lot more common than many people realize. But that’s a far cry from a naturally occurring optical cavity with high enough Q to produce a useful enhancement. And in the IR you have to contend with absorption by (or at least the highly variable refractive index of) the organic compounds from which the interfering structures would be made.

          So, possible? Perhaps. But likely? I seriously doubt it. My point was that the mere existence of micrometer-scale gaps is not sufficient to produce this effect.

          1. kjk says:

            You are right that the role in biology opens many questions. We would need to find gaps with photonic properties, do we even know of any? What kind of Q-factor is needed, and how high a Q-factor biology can build with structural coloration (i.e by making a dielectric mirror)? Would this mechanism even be a useful for biology? I generally put the odds near 50:50 for unknown “does biology do this” questions because biology is 99.9+% unexplored.

  6. rhodium says:

    The orientation effects makes we wonder what liquid crystal solvents could do.

    1. Barry says:

      Dick Zare showed thirty years ago that vibrational energy distributes through a molecule faster than bonds break. So we would expect no special orientation effect. (of course, Fermi reminds us that science moves forward on the unexpected results)

  7. Goat sea says:

    Flask shape screens, stir bar screens, stir plate screens… 5 years is hardly enough for a PhD anymore!

    1. Special Sauce says:

      What about position in the cryocool?

  8. Ir(wtf)bpy says:

    How long until vibrational modes becomes the new histogram?

  9. Magrinhopalido says:

    Very cool!! I am eager to see how far this can be extended.

  10. DCStone says:

    The first thing that came to my mind was, “Zeta potential” – depending on the material (eg glass or silica) defining the container and the nature of the solution (pH for aqueous being an obvious consideration), things happen at the surface that can easily influence out a micron or two.

  11. Eugene says:

    Casimir effect.

    1. Derek Lowe says:

      Yeah, that one weirds me out too, I have to say. But from what I understand, experiment matches theory on that one pretty much perfectly, so I guess we have to learn to deal!

      1. Li Zhi says:

        I’m definitely not well informed on this topic! So fwiw:
        There is an alternative “explanation” for the Casimir effect. While convenient, invoking vacuum energy is not required. (Although the fashion police may object.) See Jaffe’s 2005 paper. This, imho, seems to be a case of us wanting a “gee whiz!” explanation rather than the more boring one (relativistic van der Waal interactions). Other comments even more pedantic: Theories fit facts, not the other way around. Any one who believes some theory, observation, structure,…, is “perfect” has left the realm of Science.

    2. ³{ⁿ≤ says:

      Oh let the sun beat down upon my face, stars to fill my dream

  12. Nick K says:

    Spooky. I must read up on Rabi splitting before I try reading the paper.

      1. Nick K says:

        Thanks for the link. I still find it profoundly weird that the effect still occurs without light photons being present.

        1. kjk says:

          Photons ARE still present, the thermal radiation in the infrared.

  13. Jose says:

    Wasn’t some HC Brown borane chem only reproducible with a football-shaped stirbar, as it ground semi-insoluble particles small enough to react?

    A postdoc also told me of some voodoo reactions that only worked with the *red* septa and failed with the tan ones, presumably due to traces of plasticizers……

    the more we know the more we realize we don’t know anything!

  14. Jose says:

    Wasn’t some HC Brown borane chem only reproducible with a football-shaped stirbar, as it ground semi-insoluble particles small enough to react?

    A postdoc also told me of some voodoo reactions that only worked with the *red* septa and failed with the tan ones, presumably due to traces of plasticizers……

    the more we know the more we realize we don’t know anything!

  15. Joy Craig says:

    I never forgot these words from a wise developmental biology professor on the difference between what happens in the organism (or cell, or organelle) and what we can recreate in a test tube:
    Every extract is a lie.

  16. Nick K says:

    Have these effects been shown in a molecule encapsulated in a fullerene?

  17. DrOcto says:

    Can’t help but surmize that this might be a retardant effect of the surface itself. It would be interested to know what the cell is made from, and if other materials give the same reduction in rate, for example silicon dioxide would not have been my first choice. Flow chemists already know that mixing is lower at the walls of their vessel, so it could be a diffusion/mixing effect. Or otherwise some local binding effect that accentuates the high surface area to volume ratio.

    1. Derek Lowe says:

      Taking the same cavity and making the walls a tiny bit closer or further apart seems to wipe out the effect, from what I can see.

  18. Ursa Major says:

    “what if I told you that a particular simple reaction ran five times slower, compared to the bulk liquid, when it was confined between two walls that are six microns apart? How about that system still running five times slower than the one where it’s being confined between two walls that are eight microns apart?”

    Surface effects are huge but only extend a very small distance so I would expect this much more than there being any measurable difference between a 100 mL round-bottom flask and a vat in a factory (just stir the damn thing properly!). But then I’m not a bench or process chemist…

  19. I wonder if it would be possible to orient the molecules using a gold-thiolate self assembled monolayer, and see an enhancement of the effect? If it is really dependent on orientation, this should make a dramatic change. Can you can still have a reflective cavity with gold thiolate on the inner surface?

    1. Rethinking this – you can’t get a high mole fraction of the molecules on the surface compared to the bulk, even in a micro cavity. So, what if you attach a cyclodextrin to the surface via gold-thiolate, and let it act like an enzyme binding site, orienting the molecule to the VSC? You’d want to use an effect that enhances the rate of reaction. The cyclodextrin-binding moiety could orient the molecule without changing the vibrational frequency of the bond you’re breaking.

  20. Li Zhi says:

    Doing it in the dark, finding size (and fit) matters, my gosh the INNUENDO! Someone had to lower the level of discourse, right? Anyhoo, this is neat. I mean, we’re talking microns, not nanometers! Practically macroscopic (arguably macroscopic?)

Comments are closed.