Home » Still In The RBF


3 May 2007 13,662 views 35 Comments


Feng, Tian, Yuan, Mu and He. Org. Lett., 2007, 9, 2019. DOI: 10.1021/ol070597o.

Bear with me on this one – I guess you’ve seen this target before. I was shown this paper by a few of the chaps in our group, and was just astounded with the transformation, and even more by the mechanism.

Awesome! But the mechanism contains some rather worrying steps. I’ve capture the entire thing directly from the supp. info., cause even I can’t be bothered redrawing that:


(The supp. info., BTW, is here.)

I’m particularly concerned by the Fe-mediated addition of carbon dioxide to formaldehyde. Any thoughts?

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)


  • milkshake says:

    I wonder if iron carbonyls are involved in any way.

  • milkshake says:

    Also, it should be NaHCO3 in your first scheme, you fot a typo

  • willyoubemine says:

    i mean, it has to be true, right? its not like phenol is an uncharacterized compound. But dang, that is crazy. 0.8% yield of phenol.

    a little surpring there werent furans, or pyrans or phenol-formylated products…why only phenol? that is some crazy s though.

    Semantic note: Does that really count as a “Total Synthesis”

  • valleygirl says:

    Of course it does, they like, totally synthesized it.

  • Tot. Syn. says:


    Cheers for that – shows I can mess-up even the simplest of schemes…! I was thinking along the lines of a iron-carbonyl type system, but without details on the “Fe Powders” they used, I’m kinda lost. Tricky stuff…

  • Jose says:

    “The fact that we caught trace
    formic acid and formaldehyde in the initial stage of the
    reaction confirms the suggested mechanisms.”
    Hmmm, that is a pretty bold statement!
    And why would the pyrocatechol react further?

  • Ψ*Ψ says:

    This is the first target you’ve ever posted that I can look at without getting a headache.

  • rosko says:

    I’m most bothered by the “tetraketones” (the middle and right structures in the third line of the actual mechanism, ignoring the formation of CO2 and the oxidation of iron). I didn’t think that compounds with four carbonyl carbons in a row could ever exist without being destroyed through immediate nucleophilic attack by water–or virtually anything else with a lone pair.

  • provocateur says:

    At 200 c, I will beleive anything!

  • Jose says:

    The supp info is a joke…shouldn’t they at least mention the particle size for the “Fe powders?” 200 deg C is not that big a deal, and 1.8 MPa is only 260 psi (18 atm). Chemglass has glass pressure vials that are tested to 150 psi. If were talking diamond anvil pressures, a hexa-carbonyl compound might be a plausible intermediate. The ferrocyclic CO2-formaldehyde thing- aren’t trimolecular processes essentially forbidden (or maybe extremely rare outside of enzymes)?

  • TheEdge says:

    I’m with Jose. I’m most sketched out by the fact that they only see one carbon products and the phenol by Mass Spec, and then draw a mechanism with many, many intermediates. Why can’t they see anything else? They clearly have the ability to see down to a very low threshold.

    Rosko: FWIW, I don’t think their mechanism is changed if the internal ketones exist as hydrates. I’m not saying I agree with their mechanism, just that everything is easier to draw if they ignore this particular issue.

    Re trimolecular processes: I think they’re banking on some sort of metal template, as in the trimerization of alkynes using Co and Ti (Sato, F. et. al. J. Am. Chem. Soc. 2002, 9682.) Maybe I’m just giving them too much credit. Maybe I’m just happy that they made some attempt to explain what they saw.

  • Totally elegant says:

    The elegant total synthesis of phenol is now complete.

    Classics in Total Synthesis III ~ full steam ahead!

  • ZAL says:

    I think the value of the paper obviously is not in the preparation of phenol, but just in showing that starting from simple inorganic materials in the presence of a metal abundant in the earth crust and under certain conditions it is possible to form organic materials (if, as I assume, we all agree that the really did observe phenol). Of course, then, drawing a mechanism should imply having some solid evidence for it: I did not read the article carefully, so I’ll let you decide if this is the case.

  • dtb says:


    Actually, the formation of simple hydrocarbons from inorganics has been demonstrated before under a variety of conditions:


  • Tot. Syn.'s other half says:

    After reading the paper they seem to have very liitle evidence to support their mechanism (react-ir) might have been useful to spot some of those feCO)n species? The experimental has a lot to be desired….not sure i believe it

  • Mike says:


    (i) Wouldn’t it have been trivial to corroborate some bits of their proposed mechanism by taking 13C-labelled glycolic acid, glyoxylic acid, catechol and o-benzoquinone and seeing whether they are incorporated into the phenol under these conditions?

    (ii) Does figure 1 really show a “self-catalysis” reaction? Note that the y-axis doesn’t go down to zero. That looks pretty close to a straight line through the origin as far as I’m concerned.

    (iii) Given a yield of 0.8% and an apparent sudden stop to the reaction, what happens to the rest of the bicarbonate? Is it all converted to sodium carbonate?

    (iv) They say that they checked the bicarb and the iron for contamination by organics… what about the water??

  • ZAL says:


    thanks for the links: I honestly suspected that something like that had already been reported, although I had never actually read a paper on this topic (but now I can!). My post was more intended to be an answer to, for example, #12 (funny anyway!).

  • bink says:

    What about 0.8% of phenol as a contaminate in the bicarb?… or this mysterious “Fe powder” (sounds suspicious – sort of like “mishmetal” to me…). I think maybe someone spilled some phenol on the benchtop or in the reaction vessel…

    Also, I would think that from their mechanism you would expect to see some of those intermediates in there… especially the catechol compound! In one mechanism they have what is essentially an oxidation to the catechol… then a reduction to the phenol?!?!

    I know that you can get “organic” compounds from “inorganic” compounds, but this is just strange.

    Let me shoot up some heroin so that this makes sense…

  • bink says:

    Not only that, but catechol is a great binding ligand for iron. You would definitely think it would stick around instead of the phenol.

  • walkerma says:

    The reverse reaction, Fe-mediated elimination of CO2 from glycolic doesn’t seem that far fetched so the forward reaction is mechanistically feasible.

    Two other points, 0.8% might be too high to be considered a contaminate in NaHC03….and multiple carbonyl compounds really aren’t that unstable…you can purchanse keto-malonate (tricarbonyl) cheaply.

  • walkerma says:

    glycolic acid…

  • milkshake says:

    The other possibility is that the reaction proceeds by some weird oxalate-glycolate-formaldehyde condensation to provide salicylate or p-hydroxybenzoate; those could reversibly decarboxylate under the reaction conditions.

  • milkshake says:

    I think their proposed mechanism is crap. I would expect a slow reduction to methanol together with some C2 products. If there is a trace of glyoxylic acid on the way, then it can do the analog of the well-known formose reaction (which incidentaly works best in mildly alkaline aqueous media at elevated temperature in the presence of 2(+) charged cations like Ca, Mg, Pb or Zn.) Eventually a C7 piece forms that cyclizes to a hydroxylated/unsaturated cyclohexanecarboxylic acid that aromatizes by dehydratation+reduction sequence with decarboxylation in the final step.

    These conditions are quite close to what one can achieve with the common Biotage cavity microwave reactor in a glass pressure vial. At phenol yields 2mg/10mL the produced phenol should be observable by NMR and HPLC (even though iron can mess up things) I think I will give it a try and report back the results.

  • milkshake says:

    There is a clear methanol signal in 1H-NMR(D2O+water) of the reaction mixtures done at 180C and 190C. (The methanol signal is higher in the 190C experiment). Also there was toluene as the main product identified on HPLC in the experiment done at 180C. There was a clear toluene smell when opening the vial also.

    I will do two more runs at 180C and do writeup on the experiment on Org Prep Daily.

  • provocateur says:

    did anybody tell me u r awesome….

  • provocateur says:

    the previous comment was for milkshake….this man actually did it!

  • ddd says:

    I wonder if you could use the same strategy to make brevitoxin

  • Jose says:

    Milkshake, you’re the man! Brilliant, brilliant. Toluene DNE phenol, but I’ll be happy with anything aromatic….

  • milkshake says:

    There was a (relatively) large toluene peak in one out of 4 experiments, in the other 3 there was just a tiny blip on HPLC baseline (one among many) where toluene should have been. My best guess is the one exceptional experiment worked only because there were some iron metal particles above the reaction mixture level and they were probably throwing off white sprarks and glowing red under the microwaves. If you do such thing in anaerobic mix in the presence of CO and MeOH, you get something like Fisher-Trops alkane formation and alkane-to-arenes reforming proces, hence the observed toluene. So instead of examing a hydrothermal process, I have probably simulated chemical refinery processes by a short series of poorly-controlled experiments. The only thing I have proven is that microvawe reactor is a poor substitute of autoclave, especialy when metal particles are present.

    I believe that the observed methanol is probably real product of CO2 reduction but I have not proven it beyond reasonable doubt that it is not an artefact. The samples from all experiments apart from a number of tiny peaks also contained some very broad hump of UV-absorbing (255-310nm) polar substances which might be some hydroxylated/carboxylated conjugated polymeric stuff that might concievably serve as a reservoir of from which phenol forms – but this is a speculation. I have not seen any phenol under my conditions.

  • blatnoi says:

    They also say in the paper that if you change the acidity, the yield goes up drastically, but they don’t mention what it actually becomes and what acidity is required. And I couldn’t find anything in the two page supporting info. I guess they want to save it for a future paper, but so far this one gets the official ‘blatnoi thumbs down’. This sound like the kind of research project that you would design if your funding agency told you to pick two instruments, and you chose ‘GC/MS’ and ‘autoclave’.

  • Warren says:

    Like my PhD advisor Burke always said, “Heat it until something happens.”

  • NaHCO3 —-> Phenol?…

    Ob das wirklich so “einfach” geht? Sieht auf jeden Fall mal interessant aus, hier gibt’s den Mechanismus.
    Quelle: Feng, Tian, Yuan, Mu and He. Org. Lett., 2007, 9, 2019. DOI: 10.1021/ol070597o.

  • Hahaha.I have used instead of Fe a mixture of Pd, Pt, Fe,Co, Ni, C, and I have obtained Hydrochinone ( 21%)
    And your mechanism is completely wrong; First you obtain CO which polymerases to C6O6, and than, + H2 you obtain C6H5OH

  • Tot. Syn. says:

    I think I should point out that “wannabe milkshake” above tried wanted us all to think he was Milkshake. But it was obvious he was a fraud. Anyway, here’s the IP trace:

    “ is from Romania(RO) in region Europe”

  • Albert says:

    Trolls on Totally Synthetic? The world has gone mad…