Take a long good look at this reaction, recently reported in a respected journal. Have you ever seen anything like it? “Unusual formation…” — an understatement of the century, if you ask me. What the Hell is going on here? A porphyrin-ish type material spitting out a C7H5 unit (a toluene minus a handful of hydrogens). If I was teaching a course right now, I would be tempted to use this on the final exam and ask the students for a plausible mechanism. Hehe.

Unusual ring contraction

Org. Lett. 2013, ASAP (DOI: 10.1021/ol303527v)

No but seriously, I have read this paper over and over, and it leaves me with very little doubt that they indeed assigned the correct structure for this highly unexpected product. I mean, it is difficult to argue with X-ray data. Let us just for now assume that the authors got the characterization part perfectly right.

Scheme 3, however, is in my humble opionin deeply unsatisfactory. There are at least two things that annoy me:

Bad mechanism

  1. The proposed mechanism requires 1 equivalent of water under strictly anhydrous conditions (excess POCl3, EtN3 under N2). That is not pretty at all.
  2. That 1,2-shift — the key step — just does not make any sense! Why would that electron-deficient bond do that? The way they have drawn it, it is impossible to draw a resonance structure in which electrons push that bond over (as you can do in for instance the pinacol rearrangement).

One could always argue that the driving force for the overall reaction lies in the lower energy of the product. Fine. Still, that does not justify the abuse of curly arrows! Now, let us fix this one, shall we?

I have spent a reasonable amount of time coming up with a decent alternative. Before showing off, I would love it if you would go ahead and post your ones first. @Heterocyclist56 and @jamesashchem, I dare you!

All of you: Feel most welcome to draw a better mechanism, upload it to any image sharing page (such as but not necessarily restricted to imgur.com) and post the link to it down below.

Go, go gadget arrows!


Support Synthetic Remarks

You know it is the right thing to do

Bitcoin  (BTC)  12KejdK6zwYhPr1S6JB1d3khGCPAbsUDMA
Litecoin (LTC)  LdK1zKCGJcovPogME7QRkTpk2zThN9aHcK
Ethereum (ETH)  0x7095c468393C31B705Ad2fA79a6cf87E90E2d823
Ripple   (XRP)  rfUfHjTUcRoCcuiaMTb9PVZAJu76YzaHmb


12 Responses to Let’s fix this mechanism, shall we?

  1. James says:

    Good catch. Here’s my contribution. I was thinking that it makes more sense if you do the rearrangement from the other resonance form:

    I’m not sure how the final C-C bond breaks. The mechanism requires a proton source. If not water, it must come from adventitious HCl at the very least. Maybe attack at the carbocation by Cl or NEt3, protonation of nitrogen, followed by formation of yet another carbocation that gets trapped?

  2. James says:

    Scratch that, actually, the orbital overlap wouldn’t allow it.

  3. FA says:

    My attempt at mechanism outline (text only):

    Number the carbons atoms of the oxaporphyrine ring starting from the
    migrating alpha-position of the furane ring =1, clockwise around the
    ring so that the benzylic carbon of the C7H5 unit that is eliminated
    = 20.

    1. Attach electrophile (H+ or POCl3) to nitrogen (lower right hand side
    of compound 3 as drawn). I much prefer a mechanism with POCl3 reacting
    at nitrogen rather than carbon.

    2. Add nucleophile (e.g. Cl-, or triethylamine, toluene, but probably
    not water) at C-20. (I really don’t like the idea of electrophilic
    attack in this position). An aromatic pyrrole ring is formed.

    3. Attach electophile (H+ or POCl3) to C16. This is typical pyrrole

    4. 1,2-migration of the C-20 — C-1 bond (ring contraction, new bond formed between C-19 — C-1). A benzylic carbocation is formed at C-20.

    5. Add nucleophile (e.g. Cl-,..) at C-20.

    6. Add electrophile (e.g. H+ or POCl3) to the para position of the leaving C7H5 unit. (Or possibly form a pi-complex of the electrophile with the phenyl ring)

    7. E2 reaction; detach the electrophile at C-16 (see #3) along with the
    C7H5 unit. This gives compund 1.

    I have not read the paper. Has the byproduct derived from the C7H5 unit been isolated and characterized?


    • drfreddy says:

      I like all you bullets.

      As far as I understand it, no characterization of the curious leaving group was made.

      At least I am not alone in the community thinking that this reaction is really weird!

  4. Hop over to my Facebook page to see an alternative scheme. https://on.fb.me/XXVPUE

    First, starting with a more attractive tautomer of the porphyrin, an electrophile (El, not sure what) attaches at the phenyl-bearing carbon. Then migrate the furan (I like this better than migrating the pyrrole, but both are possible. If you want to migrate the pyrrole, you can play around with the tautomers to have a 1H-pyrrole migrate rather than the fulvene-like pyrrole as the authors show).

    Eventually, it seems likely that water is needed so you can fragment (retro-aldol style)and kick out the final product. I could see this happening in the workup if there were no water around to make it happen earlier.

  5. drfreddy says:

    After all your fantastic contributions, all good candidates — and all much better than the one published — I’m gonna need more time before I can show you what I had in mind (currently under revision).

  6. milkshake says:

    I don’t think any of the preceding proposals is fully right. POCl3 is a darn oxygen-loving electrophile. My expectation is it will grab onto furan oxygen and fragment the furan ring – The furan fragmentation will give rise to a vinylogous version of Villsmeyer reagent that will, as an excellent electrophile attack one of the neighboring pyrroles. Pyrrole migration followed by aromatization will lead to eventual excision of benzaldehyde equivalent in the form of PhCHCl2. Here is an idea:


Leave a Reply

Your email address will not be published. Required fields are marked *