Home » Still In The RBF

Welwitindolinone & Fischerindole   

27 March 2007 6,917 views 43 Comments


Baran, Maimone and Richter. Nature, 2007, 446, 404-408. DOI: 10.1038/nature05569.

There’s actually not too much more to add to this synthesis of three further related natural products – using chemistry familiar from the previous synthesis, they were able to complete 11-epi-fischerindole G in only six steps from carvone oxide. With this target already complete, a simple oxidation with DDQ returned fischerindole I in one step and in a great yield.

However, using xenon difluoride for fluorohydroxylation of the indole was certainly less simple; the disconnection to electrophile addition, addition of water, rearrangement and then elimination of the “electrophile” was inspired. The molecule then underwent a [1,5] sigmatropic rearrangement to give the spirocyclobutane of welwitindolinone A, as a single diastereomer. Damn nice!

I’m also glad to read they admit that the selectivity of the latter process was unknown to the group, and that they intend to study this further.


Folks, I’ve had to freeze comments and delete the most inflammatory/libellous – by using this blog, and certainly writing commnets, you’ve read the AUP, and agreed to it. It states that the discussion should remain civil, and frankly we’d gone beyond that considerably. Please remember that, especially if you don’t use your real name as your “Name”, I take the flak and indeed the harm to my career prospects when commenters are libellous. I’ll open the comments again shortly, and allow you to re-write your comment then – with your real name attached – if you still think it’s a good idea.
To paraphrase the great Michael Winner “Calm-down dear…. It’s only a synthesis paper!”

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)


  • HR says:

    nice endgame. but VERY costly… (check the price of xenon difluoride)

  • RM says:

    I eat xenon difluoride for breakfast!

  • TB says:

    Baran’s a damn genius and the spiro-oxindole is a great scaffold. After seeing so many 30+ step syntheses of molecules using conventional disconnections that spend half the effort on changing the oxidate state, it is nice to see a beautifully planned and extraordinarily efficient synthesis. I also like Baran’s general concept that the oxidation level of the molecule was gradually increased over the course of the synthesis. I discovered this blog a month ago and have told all of my colleagues – this really lets me keep in touch with my synthetic roots. Fantastic job PD.

  • milkshake says:

    2,3-disubst indoles can be gently oxidized to 3,3-disubst oxindole with 2->3 substituent migration. This trick for making spirooxindoles has been demonstarted couple times before (Danishefsky) although I don’t remember a substrate so complicated as this one. I wonder how may things they tried – until they finaly arrived to XeF2.

  • JB says:

    Reading this paper blew my mind. Baran has just raised the bar on organic synthesis in the field of total. This synthesis is not only elegant and efficient (minus the cost of xenon difluoride), it is just flat-out ingenious. I’m sure most people reading this blog (excellent website by the way!!) has purchased or has read a copy of Nicalou’s Classics in Total Synthesis. This synthesis is of the caliber deemed ‘classic.’ If you don’t agree with me, let me put it in more perspective. John Porco Jr. of Boston University has published a Nature article (doi:10.1038/446383a) which HIGHLIGHTS Baran’s synthesis and the importance of building molecules without protecting groups. Imagine having a Nature paper written on YOUR OWN Nature paper. It just boggles my mind.

  • HR says:

    With regard to the reaction-workup-column-another reaction sequence, he did mention that the column is not actually necessary so perhaps we could give him benefit of the doubt? However, does rotavap-another reaction sequence still count as a one-pot (or one step, if you will)?
    It has also come to my attention that he uses brsm yield every now and then, which can make a step much more impressive than it actually is. Furthermore, he never gave a full account of the various reagent screenings (which I assume he would have done, eg. the XeF2) that he did before finding the best reagent for some of the key steps. That way, we can better assess the cost efficiency/productivity (and the time spent?) of the synthesis.
    Nevertheless, there is a lot of very inventive and unconventional methods used that we would not even dare to think of when planning a synthesis, which is what impresses me mostly.

  • Lurker says:

    Haven’t gotten through all the supporting info yet…

    Aldrich sells starting material “A” in the supporting information (p-menth-1-en-9-ol). Since I’m at home, no $$ information given, so I can’t judge whether it’s cheaper to make it than buy it.

    Going from A->B, it’s 50/50 to me to count that as a single step or two steps. It sounds like, in practice, they did it as a single run through without characterizing the intermediate cyclobutanone. The column, with the slight decomposition, was likely to aid in method development and to publish it (just my guess).

    Going from B->C, I count two reactions. They run a column in between and then at the end, too, it appears.

    Going from C->7, looks debatable. It’s one column, but there’s a workup and solvent switch. I’d count it as two reactions myself, but again, that’s just me.

    Baran published Welwitindolinone A and Fischerindoles I and G in 2005 as a JACS communication DOI:10.1021/ja056171r (sorry, don’t know how to link to the right doi). If I remember correctly, the welwitindolinone synthesis is significantly shorter than Wood’s synthesis (which is nothing against Wood’s synthesis – it’s a nice piece of work itself).

    IMO, even if a step or 2 were glorified a bit, it doesn’t take away from the elegance of the approach and the sheer brilliance of a couple of the individual steps. And, for many of those compounds, that’s a small truckful he was able to prepare.

  • JamesB says:

    My $0.02: If you intentionally isolate an intermediate (column, distillation, crystallisation etc.) then that is a step. Stripping off solvent does not count. A solvent switch does not a new reaction make.

    I say “intentionally”, so if you are a jammy git and your product drops out of the reaction, I’ll let you call it one reaction.

  • WillisWill says:

    What happens to the Xenon? I can’t read the paper, but I presume they use a stoichiometric amount. Wikipedia (anything that supplants Paquette is allright in my book!) tells me xenon oxides (di, tri, tetra) are very powerful oxidizing agents and explosive in either the presence of organics (trioxide) or temp>-30 (di, tetraoxides) Is XeO known?

  • Russ says:

    WillisWill: I haven’t read anything on the subject, but in my imagination, the byproducts of that reaction are neutral Xe gas and fluoride anion (picture one more arrow, breaking the other Xe-F bond, in TotSyn’s scheme above).

  • pi* says:

    I don’t mean to be a crossposting tard, but I just posted this in the earlier post…

    It is a damn good synthesis. Probably one of the nicest in some time, better that the Jacs standard (which is usually pretty high).
    Nature hardly ever has any chemistry, let alown synthesis. The entire issue was primarily chemistry (and it came out 2 days before the big ACS meeting in chicago…). I think that the issue (and to include the synthesis) was kind of a hats off to synthesis.
    Yeah it isn’t going to change anything, you didn’t discover anything new. but certainly at the top of the game in a very difficult field.
    The paper on breaking bonds using mechanical strain is pretty typical nature fair, and I’m sure would have gotten in even without a special chemistry issue…this will open up a new field of chemistry.

  • TheEdge says:

    And TotSyn gets his first flame war. How cute.
    The menthenol is $17/gram as a mixture of diastereomers, FWIW. I’m sure we’ve all started with more expensive reagents than that.
    I think he does some really novel and cutting edge chemistry, but he overstates some simple stuff, which brings the paper down. Honesty above everthing. If the material didn’t require purification to be used in the next step, why not just “purify a small amount for characterization”?
    What we’re all missing is the reproducibility question. Do any of his really slick transformations work in anybody else’s hands? Anybody here have any experience with doing his chemistry?
    Dave Evans has been out on the lecture circuit with the oxidation level theory for about a year, and Baran actually cites him, so let’s give credit where credit is due.

  • aa says:

    “Dave Evans has been out on the lecture circuit with the oxidation level theory for about a year, and Baran actually cites him, so let’s give credit where credit is due.”

    I’m sure many of you have seen Baran give a talk, and one thing I appreciated was the way he credited various “philosophies” of total synthesis directly to the profs who “invented” them ie atom economy, step economy, oxidation state change, careful retrosynthesis etc. To me, acknowledging your debt to your mentors and predecessors is much more impressive than claiming “I invented this all by myself!” Of course, these ideas have been out there for a while, but to see them summarized in one placed, and applied with such elegance to complex targets was pretty incredible.

  • Hap says:

    TE: lots of stuff is not pure enough for characterization and publishing, but will easily undergo the next reaction in a sequence. I’m reading a book contg. Woodward’s papers and he/his students seem to do this a fair amount – part of the motivation is to avoid material loss (still valid now), and part of it is laziness/workload optimization (why spend time do large scale recrystallization when I can plug it through and purify later?). I don’t think that’s a problem.

    I’m not a fan of brsm, though (it depends on how much work is needed to recycle the starting material and whether it decomposes or does something else bad in the process).

  • HR says:

    aa: Indeed, if we look carefully at the paragraph where he talks about his “philosophies”, we will see that each point is referenced to another paper or talk by his predecessors. In fact, during his talk at my school, he put up the corresponding reference right after each point on his “philosophies” slide. So, clearly the element of deference is there.
    willyoubemine: He did mention during his talk at my school that it’s not easy to write proposals for such works, as most of the steps are not precedented. In his illustration, he would draw starting material and product, propose a transformation which is pretty reasonable (or at least not unlikely to happen) and a question mark on top of the reaction arrow, ie. finding the correct reagent to make the reaction work.

  • TheEdge says:

    Hap: I’m fine with not purifying SM. I do it, too. I just take issue with people who say stuff doesn’t need to be flashed, but report yields and procedures in which they do. I bet RBW didn’t do that. If you claim your reaction works without purification, you should provide yields that support that claim.
    I don’t have a problem with brsm yields, as long as the actual yield or amount of recovered SM are listed along side it. Some reactions just don’t go to completion.

  • monkey want a phd too! says:

    >>>I’m not a fan of brsm, though (it depends on how much work is needed to recycle the starting material and whether it decomposes or does something else bad in the process).>>>

    Neither am I. This as much as anything detracts from some very nice work. The reader is left to wonder (or dig through the SI to find out) just how much SM was recovered.

    If you have 50% BRSM and recover 75% SM, then your isolated yield is only 12.5% and you’re going to be recycling quite a bit to ‘realize’ the 50% yield. (Homework: develop an equation to estimate how many reactions would one need to run to realize 40% isolated yield in the above scenario, then do 49%. – short answer, too many)

    So why not list the isolated yield or the percent of SM recovered when disclosing a BRSM yield? It just looks to be hiding something… perhaps a truly dismal yield? And perhaps you’d even be running so many reactions to realize your “50% BRSM” that you’d even consider using a protection/deprotection strategy if it would allow you to avoid the recycle…. just a thought…

    Baran would do well to just put this info on how much he recovers up front, in the scheme. Then people would not ding him on this in the comment sections of blogs, or worse, in other places where people discuss science. At least it doesn’t look to be sweeping anything under the rug. Same goes for creative ‘step accounting’.

    I agree with “spot to spot” as well. By the opinions of some on this board, it would appear that if I dump pet ether in to my reaction, the product crashes out, and I filter it before adding dissolving it for the next transformation, that’s a step. But if I throw the same reaction mixture on the rotovap, then add solvent for the next step, it isn’t? Give me a break. Just because there is no point in full characterization of every intermediate in a synthetic sequence doesn’t mean individual steps did not make the uncharacterized intermediates.

  • The Next Phil Baran says:

    I love how one “step” would actually require the installation of three 5 million dollar reactors, the utilization of three times a thousand litres of non-disposable, polluting solvent in industry and yet its passed off here as one fluid motion. Then the reaction solvent needs to be purified with a few thousand gallons of water, which then needs to be repurified before its used again or dumped. We synthetic chemists consider a day with one less column a treat because they require an extended amount of time. I mean setting up reactions, extractions and evaporation of solvents is not that bad but a column is a commitment. A column free step is a big deal to a researchers day, but a synthesis will not likely be used in industry if it needs a column chromatography step because it requires too much solvent and hence money either. Phil Baran should not be lying to himself about his actual yields he should be displaying them to show where his reactions need improvement. THe best part of this paper is the constraint he is placing on himself as a researcher: no protecting groups; but i do not know one book about the philosophy of synthesis that does not list the addition and removal of protecting groups as being the least efficient aspect of synthesis. For those who are keeping score, this requires two additional reactions be added to your synthesis.
    Phil has not reinvented the wheel but he is trying to improve on it. Protecting groups have allowed a lot of theoretical research to occur allowing access to compounds that are not presently accessible but if our field is to be applied more generally and in a way become more efficient we need to seek to simplify what we do as chemist and protecting group minimization or exclusion is definitely a method of challenging the status quo.

  • cjdquest says:

    lucagh Says: >>>as far as reproducibility – he is the only young chemist publishing full papers. Anyone else do it??

    Stoltz Et Al JOC 2006, 71, 8357
    Wolfe J. et al JACS 2005, 127, 16468-16476

    Gee, there are two I came up with off the top of my head. Why don’t you go try and find some more. Sounds like you need practice with web of science and scifinder.

  • Cat Herder says:

    I agree with Baran and many of the posters on this thread that protecting groups are often the hallmark of intellectual laziness (yes, there are clearly situations where they are needed or even desirable). Even though there is a certain element of salesmanship in the paper, this seems to be the case for many if not most tot. syn. manuscripts.

    Basically, the only way an author can satisfy his/her critics with respect to the issues being thrown about at this point (brsm yields, what constitutes a step, etc.) is full disclosure and objective reporting. For instance, yields can and should be given as both “traditional” and brsm (where appropriate) and each should be labeled appropriately (I hate it when the yield is brsm and they don’t tell you that). Also, # steps should always be described from commercial material, not known materials. Furthermore, and this is probably a bit more controversial, I think crude spectra should be provided if the material is claimed to be used in the next step without purification. I don’t care if someone purified the crap out of 10 mg and got a nice clean NMR. No one repeating the work will do that so data should be given about what to *actually* look for or expect. Similarly, negative results should also be published in the supporting information, so if they tried setting the spiro center with a different reagent and got crap, we know that too. There is nothing wrong with having a reaction fail in synthesis – it happens all the time (go look at I think Chembark or maybe Chemblog for the poll about this).

  • milkshake says:

    I bet the most furious comments here (about Baran deceiving the synthetic community about merits of his work published in this Nature article) came from people associated with groups that are currently pursuing a total synthesis of certain tetra- and pentacyclic isonitriles…

  • TheEdge says:

    Well, milkshake, I think that plays a role. But I think part of it is that a lot of us hold Science and Nature to a much higher standard than we hold other journals. I think it’s because so little synthetic organic chemistry, and total synthesis in particular, is published there. The few articles that are published are the face we show the rest of science and the science-minded public. They need to represent where our field is and where it is going as accurately and honestly as possible. The American public have a hard time understanding what science can and cannot understand/achieve, and I’m sure other scientists have a hard time understanding what chemistry can and cannot understand/achieve. If we overstate what is possible, then we do everyone a disservice.

  • Problems Reading? says:

    I guess it is no surprise that you have problems with reading, since you also have problems with counting.

    Skeptic did not say anything about removing solvent counting as a step. Now read carefully:

    Step 1:
    Running a reaction, doing a workup (ie your intermediate is in a separatory funnel, being extracted, washed etc, dried, filtered, concentrated. Maybe you ran a column, maybe you can carry it on crude.

    Step 2:
    The crude product you have made now becomes the substrate in a second reaction and so forth.

    Baran routinely counts this as one step. That is a fact, in black and white in his supporting info, and he is wrong and deceptive in doing so.

    Now, if you are going to argue against what I have written here, please do me a favor and actually try and refute the point, rather than just typing about something random and unrelated.

  • milkshake says:

    I just showed the Nature article to my boss; he was Overmann graduate/Danishefsky postdoc so he has seen things. He is not a religious man.
    His immediate reaction was: “Xenone what?”

    The second reaction came a minute later after much headshaking: “These steps couldn’t have happened all on their own. God’s hand was involved”

  • king kamehameha says:

    Tis is a very ambiguous statement: “These steps couldn’t have happened all on their own. God’s hand was involved”. Also, idols are more appropriate for religions or singing competitions, not science. lets look at hard facts (supporting infos etc) Baran’s synthesis is very creative, but there must be a great pressure at Scripps to oversell.
    P.S. Baran did not steal any girlfriens and I couldn’t care less about indoles
    P.S.S. I also spend some time in the Overman group (one N please)

  • rb says:

    what a bunch of haters. bottom line, Baran got his paper into Nature, and the rest of us didn’t. And you know what, check out the rest of his independent papers, all in JACS or ACIE. This guy is the real deal, definitely very creative. Anybody (well not anybody, but you know what I mean) can go “swern, wittig, ozonolysis, Evans”, aka the “lather, rinse, repeat, method of synthesis” But how many can go oxidative enolate coupling, reductive Heck, xenon difluoride on that ass? Not many.

  • TWYI says:

    Well said rb.

    As with any other profession, being at the very top of your field unfortunately seems to bring as many critics as plaudits.

    I am 100% the latter.

  • king kamehameha says:

    “Lovers”, “haters”, it’s really becoming a soap opera. Are you guys all teenagers here? :))

  • Baran’s grammar sucks as well – from the Nature abstract

    As a consequence, molecules that have previously required twenty or more steps to synthesize racemically in milligram amounts can now be obtained as single enantiomers in significant quantities in ten steps or less.

    Should, of course, be “fewer”.

  • Tot. Syn.'s other half says:

    Being an inorganic chemist i was also what the XeF2? But I think this paper has encouraged a good debate, even if nothing else can be agreed upon.

  • pi* says:

    to baran’s other skills, you can add:
    Got synthesis paper into nature or science.

  • .... says:

    I am confused by this discussion. In my opinion the paper is excellent and belongs in Nature. The only other recent Science/Nature total syn. that comes to my mind are Norzoanthamine and the Tetracyclines. I feel that Baran’s paper is on the same level as those outstanding papers. I’m curious to know what those who are so upset by Baran’s Nature paper feel about earlier examples of total syn in Nature/Science.

  • ZAL says:

    I think this is an excellent paper, that obviously (being a product of human ingenuity) has its weak spots, described more or less gracefully in some of the previous comments. The point is, in my opinion, that we all synthetic organic chemists should be happy to see such a work (as well as the highlights by Bergman and Toste in the same issue on other topics) published in Nature, which should help to increase the visibilty of our field at least a little bit. In the last years 3-4 I recall papers of Rawal, Enders, Hoveyda, Stoltz and Blackmond published on Nature (strangely almost all about organocatalysis): all excellent contributions, but nothing mind-blowing, not better than some JACS or ACIE papers on similar topics. Maybe that’s enough, even for Nature, and we should become a little bolder and try submit more, when we feel we have something very good in hand!

  • TheEdge says:

    The tetracycline paper contained excellent chemistry, and was an incredible synthesis, but the feature that got it into nature was the extensive biological testing. The norzoanthamine paper contained good chemistry, but it was a land war in asia for the latter portion, and the feature that got it into nature was the amazing deuterium-isotope driven step.
    Baran’s work is excellent, but I think the reason it is in Nature (as opposed to JACS or ACIE) is this revisitation of the idea that an investigation of a compound’s innate reactivity allows us to generate functionality more rapidly than we do with the current approach to synthesis. And he’s absolutely right. And the idea (although it has been around since Woodward) deserves some press. The problem some of us have with this particular paper is that we feel he oversells his chemistry. If somebody at the top of his game does that, then what can we expect from those who are not?

  • European Chemist says:

    If I may comment…

    Of course, some of the bitter comments on this paper come from the fact that, for a Tot Syn article in Nature, you would expect a synthesis all with 80%+ yields and with at least 10 steps using strange reagents (Baran ONLY uses one XeF2 reaction, tssk tsssk). As someone mentioned previously, this only means that we definitely should submit more often to these journals instead of sticking to JACS and ACIE.

    As to Baran. His work thus far is clearly outstanding. The work reported in this Nature article is clearly outstanding.But I agree that
    a) the trend of “borsm” or “brsm” yields is not a good one – but Baran is not a lonely sinner on that one
    b) he should be clearer on the number of steps. But so many syntheses in JACS start off from “readily available compounds” which actually were described in some 1978 Tet Lett in 9 steps from “readily available” trimethyl-amino-oxo-carvone oxide, and so on…
    c) the work itself (and I mean essentially the oxidative enolate coupling) has not been used by other groups so you really don’t know wether it works in someone else’s hands.

    I remember seeing some Baran synthesis in JACS where the final steps included a decarboxylation of a pyrone intermediate for which my only reaction was “how the HELL did he think of this???” (maybe it was Chartelline A or something like that). We’ve got to give him credit, the man has something new to give to the field.

    Now, one of my personal tenets is that “if something is really good, it will become apparent by itself”. This should hold for scientific publications, and overselling your (inherently good) chemistry will only end up pissing off some persons.

    Last but not least… what should really make you think is that oxidative enolate coupling was there before some of us were born.
    If you take a really good look at the work described in this Nature paper, it involves
    a) a nice use of that methodology – on which the group already capitalised through loads of JACS, ACIE and OL papers-
    b) a standard method for isonitrile synthesis – reductive amination + amine formylation + dehydration –
    c) the realisation that a simple chiral pool building block could be used to install chirality – the “carvone connection” and
    d) TWO, and TWO only, clever solutions for problems: the irradiation of compound 12 to make Ambiguine and the use of XeF2 to make Welwitindolinone A.

    So, you should all be thinking: I could have done this myself!!! I could have come up with this idea!!!
    And these are always the best ideas in science: the ones that look sooooo simple that each one of us beats him/herself up for not thinking of them first.

  • willyoubemine says:

    european chemist:
    I remember seeing some Baran synthesis in JACS where the final steps included a decarboxylation of a pyrone intermediate for which my only reaction was “how the HELL did he think of this???” (maybe it was Chartelline A or something like that). We’ve got to give him credit, the man has something new to give to the field.

    the pyrone decarboxylation was Houamine, and that is an old school solution to that D-A type problem

    the Chartelline C synth is nice, albeit with some ridiculously unexplained items (i.e. the decarboxylation, the Br to Cl exchange by washing at rt with brine?)


  • willyoubemine says:

    thanks totallysynthetic for existing. this is great discusison, that otherwise probably wouldnt exist.

  • TheEdge says:

    I agree whole-heartedly with European Chemist. Thank you for laying out your arguement so elegantly.
    And I second willyoubemine in his/her thanks for totsyn.

  • .... says:

    The key-step in haouamine was a pyrone-Diels-Alder across an alkyne followed by decarboxylation to make an aromatic. This is standard, well-established chemistry for the preparation of aromatics.

    The cute bit was the recognition that there are no coupling methods (Suzuki, Stille, etc) that are capable of making that highly strained biaryl bond.

  • pi* says:

    This is the full paper, its an article…not a communication

  • pi* says:

    Both science and nature usually publish a commentary to explain to the lay folk what is going on in the article, so that doesn’t say anything about it. its the rule not the exception.
    As much as I like synthesis, unless you do something *new* its not going to get into nature, there will probably be another nature tot synth in 2009 or something

  • milkshake says:

    The only reason for NOT publishing this paper in Nature is that lots of biologists will read it and the Baran paper will thoroughly re-inforce their ingrained belief about synthetic chemistry: “Total synthesis of polycyclic natural products is easy, any group can be quickly put anywhere in the molecule because XeF2 is magic and organic molecules are built from LEGO”

    I have only seen Baran upclose once, at a seminar – and he came across like thoroughly likable guy – but looking back I have this sneaky suspicion now that I liked him because he made us to. It was his unguarded flash of eyes and quickness of movement and unnatural flight of ideas that gaved him away. In that crowded classroom I was in front row and he was standing right next to me as he was giving his seminar – and I swear that in the most animated moments of his talk I could smell a faint whiff of SO2…