Skip to Content

Analytical Chemistry

Spectra of An Actual Transition State?

I don’t spend too much time on physical organic chemistry here on the blog, which in a way is a shame. The readership would dwindle, although probably not as much as when I talk about patent law and intellectual property. But physical organic is an area I’ve always enjoyed, intellectually, even though it was sometimes hard to infer that during my graduate school classes. (I doubt if I have the patience to be much good at it in a lab setting).
But there’s a new paper out in Science from a team at Stanford’s SLAC, home of some of the brightest and hardest X-ray beams that ever fried a target sample. (Here’s the press release from Stanford). Working with the University of Stockholm, they claim to have actually detected X-ray spectral data (K-edge absorption) from the transition state in the catalytic oxidation of CO to carbon dioxide. This was done on the surface of a ruthenium catalyst, with extremely fast and precise heating from an optical laser to get things going.
For any non-chemistry types reading down this far, try imagining a chemical reaction as a journey from one valley to another, through a high mountain pass. “Elevation” in this landscape, is how much energy the system has, and an irreversible reaction features things going, overall, into a lower valley/energy state. The absolute peak of the mountain transit, though, is the transition state for the reaction. It’s a real thing, but it only lasts for one molecular vibration before it heads off down one slope or another. It’s the highest-energy species in the whole path because it features all sorts of half-formed and half-broken bonds, the sort of state that molecules generally avoid ever getting themselves wrenched into. But since getting up to and over that particular hump is such a big part of any reaction, anything that stabilizes the TS will speed a reaction up, sometimes immensely, which is just the sort of thing we’d like to learn more about how to do on demand.
As that press release says, quite honestly, this was “long thought impossible”, and my prediction is that there will be quite a few people who won’t accept that it’s been done. My x-ray fu is not strong enough, personally, to be able to offer an informed opinion. Even if this report is accurate, it’s surely right on the edge of what’s possible with some of the best equipment in the world, so you really have to know this area at a high level to critique it thoroughly. But what’s reported is both plausible and interesting.
What they saw was that the oxygen molecules began to change first. The the electron distribution began to change in the CO molecules, followed by a productive collision (some of the time) to form the transition state itself. And one of the interesting things about that was how many times it apparently collapsed back to the original molecules, rather than going on to product. This is going to be subtly different for every reaction, or so theory tells us, but if we are finally able to physically investigate such things we may find ourselves revising a few theories a bit. This particular reaction, taking place between two small molecules, has been modeled extensively at all levels of calculation, and the results seen fit very well, so it’s not like we’re going to be packing big swaths of human knowledge into the dumpster. But anything we can learn about transition states (and how to make them selectively happier and unhappier) is the key to chemistry as a whole.
Here’s a YouTube video from the Stanford team on what they’ve been up to. We’ll see the rest of the analytical and theoretical chemistry community reacts to this work.

13 comments on “Spectra of An Actual Transition State?”

  1. redfiona99 says:

    I can’t get through the paywall to read the article but I would have thought the main thing they’d have to do was make sure what they were detecting wasn’t damaged crystals because of the sheer amount of energy going through them.

  2. anon the II says:

    I’m trying to understand what this “fu” stuff is. “Fundamental Understanding” works. Wikipedia offers a lot of possibilities but none seem to work. The urban dictionary only offers the obvious.
    My guess is that you never really see the transition state but only a lot of things near and on either side of it and the result is an interpolation. I enjoy thinking with energy diagrams. For example, is the Fosbury Flop an example of tunneling?

  3. Derek Lowe says:

    The “fu” is a perhaps-too-obscure slang, which developed out of “kung fu”. So you hear people refer to someone having the “Google-fu” if they can conjure up good search results quickly, and so on.

  4. Daen says:

    Derek: Shame the ruthenium catalyst work wasn’t in the microwave band, or you could have legitimately referred to your Ku Ru fu.

  5. NC says:

    Not that I’m in any way an x-ray x-pert, but #2’s idea seems to be exactly what the authors describe in the paper.

  6. Semichemist says:

    Derek: Shame the ruthenium catalyst work wasn’t in the microwave band, or you could have legitimately referred to your Ku Ru fu.

    Slow clap

  7. Imaging guy says:

    I think there is a similar paper in today’s issue of Nature.
    “Direct observation of bond formation in solution with femtosecond X-ray scattering”
    Nature 518, 385-389 (19 February 2015) doi:10.1038/nature14163

  8. Some idiot says:

    With one hand…

  9. Some idiot says:

    Seriously though, if this stuff is real (and I do not have anything like the background to evaluate this myself) then this is a real biggy…
    I am always very, very cautious about accepting structures or mechanisms which are dished up via computational methods without any cross check with actually measured physical data/properties. This could be a fantastic way to cross check a lot of good computational work.
    I would also like to stress that I am not a “computational chemist basher”. Instead, I have observed that people have a tendency to believe as an absolute truth models when presented as illustrative 3D models (because it looks like something that is so real it could be true), so I tend to approach any model (even hand-drawn structures) with healthy but open scepticism. Anything that can offer orthogonal data to such models will be a huge benefit for all (modellers included)…..!

  10. M Bower says:

    #9 My PhD advisor used to say that there were three kinds of logic: deductive logic, inductive logic, and seductive logic, among the last of which he counted anything presented as a graphics display.

  11. Some idiot says:

    #10 That’s a good one! I must remember that… Thanks! (-:

  12. SAR screener says:

    It’s from a few years ago, but the only other example I know of involving an apparent crystal structure of a transition state didn’t end well.

  13. cookingwithsolvents says:

    Somebody make this work with my understanding of the uncertainty principle, please.
    I don’t think that the standard data modeling programs take this into account since they are for calculating ground-state atom scattering factors. My XAS knowledge (fu, as it were) is medium but I’ll certainly be asking the experts I know what they think about it….

Comments are closed.