Home » Still In The RBF

Napyradiomycin A1   

13 June 2009 15,714 views 56 Comments


Snyder, Tang, Gupta. JACS, 2009, 131, 5744. DOI: 10.1021/ja9014716. Article PDF Supporting Information Group Website

Okay, it’s taken me far too long to get around to this article – I actually wrote the Chemdraw back when this article was in ASAP… but other stuff got in the way, like a little Nature paper from Baran.  However, it’s a damned nice paper, so it’s impossible for me not to go back and look at it properly.  The target here is an halogenated natural product, isolated from Streptomyces bacteria – so it’s not too surprising that this family possess antibacterial activity.  The headline, though, is in vivo activity against MRSA and VRSA strains – a guaranteed grant-winner!  This brought quite a bit of attention, and a racemic synthesis by Tatsuta back in 2002.  However, Snyder has set his sights a little higher, with an enantioselective synthesis in this paper.

The starting point for this post is a bit of chemistry on flaviolin; a prep published when flash chromatography was but a dream… This prep took eight steps to complete flaviolin, but Snyder managed to shorten it to only two – neat work.  The first step I’m covering, and keeping the chemistry nice and flat, is a  Knovenagel condensation (bear in mind the 1,3-diketone tautomer), followed by a sigmatropic electrocyclic rearrangement to give the the desired tricycle.


Now it was time to get asymmetric, doing a dichlorination using a bit of Lewis acid, chlorine gas (eeeeh…), and some BINOL-type axially-chiral ligand.  The prep for this is quite interesting, as they had to use four equivalents of the BINOL-thing, stew it up with borane-THF and isolate the intermediate, which is presumably the chiral-borane adduct. Addition of the substrate at this point, and chelation of the borane is controlled by pi-stacking interaction, and allows selective trans- delivery of chlorine across the double bond.  Neat, but that’s a lot of 9 in the pot – a point addressed by Snyder, who stated that firstly that the ligand could be recovered, and the excess was required to prevent chlorination of the aryl moieties.

They then selectively functionalised the allylic chloride by displacing with acetate, and with retention of stereochemistry by using potassium acetate and 18-c-6, which presumably presents a source of naked acetate.  The acetate group was removed to free the hydroxyl, and the remaining phenol protected – setting them up for a Tot.Syn. favourite – a Johnson Claisen.  This bad-boy does a fantastic job of rearranging the allylic alcohol into a hetero-quaternary center, and providing a methyl ester as a functional handle.  The yield of product was unfortunately low, but if you consider the congested transition state here, it’s not too surprising; at least they were able to recover the SM.


I’ll stop about here, as completion of the target didn’t take much more effort.  The ester was reduced and Wittiged, whilst the final stereocenter was imparted using base and NCS.  Unfortunately this resulted in the enantiomer of the target, but that doesn’t matter a damn.  Good stuff!

1 Star2 Stars3 Stars4 Stars5 Stars (11 votes, average: 4.18 out of 5)
Loading ... Loading ...


  • European Chemist says:

    Am I the first to comment? Yeay!

    Tot. Syn.,

    don’t think I’m trying to tarnish what’s another excellent post, but… there’s a carbonyl group missing in the Knoevenagel intermediate (instead of an alcohol). And the cyclisation is NOT a sigmatropic rearrangement, but a 6-pi-electrocyclisation.

    The synthesis is nice in that it is probably the first example of an asymmetric, highly enantioselective chlorination of a double bond. But what an activated double bond that is!… and how many equivalents of chiral information are needed… and the chlorine gas… still a long way to go before convincing anyone to use that prep in another research group.

    • beaver says:

      I agree with you, it is a 6-pi-electrocyclization.

      • Tot. Syn. says:

        Yep, it is. I knew it wasn’t a sigmatropic rearrangement, but I couldn’t remember exactly what it was. I need to read Flemming again.

        I’ll sort-out the errant carbonyl later.


  • DC says:

    Newbie question: why does acetate displace chlorine with retention? What is the mechanism? Sn1? But “naked” acetate appears to favour Sn2? Thanks guys!

    • NGP says:


      This is likely a neighboring-group participation effect whereby the aliphatic chlorine helps expel the allylic chloride creating a chloronium ion that is then attacked from the front face by acetate giving overall retention. i.e. anchimeric assistance. See Quarterly Reviews of the Chemical Society 1964, 18, p. 45.

      Also, is enolic chlorine accurate terminology or would allylic chloride be better?

      • Tot. Syn. says:

        It would indeed :). I’ve changed it; thanks! And I agree entirely with your explanation for retention of stereochemistry. However, it’s also worth considering the weird case of chlorination with thionyl chloride, as it can often displace secondary alcohols with rentention. If I remember right, it’s all about a paired ion effect.

        • milkshake says:

          SOCl2 can deliver Cl from the same face if done just with SOCl2 alone. SOCl2+pyridine goes with inversion. Lepore group devised a weird kind of tosyl group that displaces with Cl with retention, because of anchimeric assistance for Cl-

  • Cascade says:

    The retention is an outcome of elimination of the “activated” chlorine (thanks to the vinyl ether) followed by chloronium formation (Neighbouring Group Participation) which in turn is opened up by the “naked acetate” .

  • krest17 says:

    Extremely inefficient – after key intermediate (tricycle) the need 12 steps to form 4 bonds. I would like to see how Baran can do it :-)

  • DC says:

    Thanks guys!

  • Farmer oleg says:

    This is a phenomenal synthesis! Extremely difficult target.

  • milkshake says:

    I rather dislike the asymmetric C=C chlorination that is being showcased here. These chromenes are outstanding substrates for Jacobsen asym epoxidation with Mn-salenes. There has to be some more efficient functional group manipulation, to get there from the corresponding epoxide

  • Madforit says:

    Nice target anyway..Does anybody knows HOW a LiOH/THF/H20 saponofication of a methylester can cleave a secondary TBS PG.!!?p.s probably a neighbouring group partecipation between the newly formed coo- and the protected OHgroup(1-3 relationship).Cheers

    • chemystery says:

      Possibly the silyl is transferred to the carboxylate anion formed (or even to the tetrahedral intermediate alkoxide) – cf Brook rearrangement, resulting in a labile silyl ester, which is readily hydrolysed?

  • ch3mical says:

    Could be do to your addition of reagents. If LiOH is added last your water may be acidic enough to cleave a sensitive TBS initially.

  • forget epoxides says:

    milkshake–epoxides will not work. the position next to the aromatic ring will be opened with nucleophiles such as water or chloride, leaving then an alcohol next to a quaternary. a double inversion would then be needed to put in the chloride, and these are well known to be low yielding. this system is pretty unique, and the fact that this type of reaction (asymmetric chlorination) has not been solved is significant. for synthesis to go forward, we should forget epoxides and go for the throat.

    • krest17 says:

      Hmm – not obvious for me. I would say that milkshake is right. This is not just aryl, this is EWG-type of aryl and this epoxide will be conjugated with ketone, but sterics are not on our side, so ???

      • ,,, says:

        Agree with forget epoxides (at least partially). You’de have to open epoxide first and that will go alpha to quinone ring. Then quite a bit of manipulations would be required, particularly challenging would be “neopentyl-like” alcohol transformation into chloride.

        • . . . says:

          you know that “forget epoxides” is right based on the fact that when the alpha chloride is replaced with acetate, it goes to the alpha position with retention of configuration (through anchimeric assistance). It is obvious that nucleophiles will open at that position.

          • ,,, says:

            Well, that wasn’t the point. The most important thing is elaboration of the resulting alcohol alpha to tert-centre. The part you’re talking about is obvious, so don’t misinterpret my previuos post, …

          • . . . says:

            sorry–that was to krest17 since it seemed non-obvious to him/her.

          • krest17 says:

            I just want to say that you may be to strong with your judgments.

            Citation from literature: “The attack of the nucleophiles at C-4 is not evident since a developing positive charge at this position should be electronically disfavored in an SN1 process. However, displacement reactions by an SN2 mechanism are known to be accelerated by neighboring carbonyl groups, and this would lead to epoxide opening at C-4.” from
            Karsten Krohna*, Helga Markd, Hans Peter Kraemerb, and Walter Frankc – “rac-Altersolanol A and Related Tetrahydroanthraquinones, Total Synthesis and Cytotoxic Properties” Liebigs Annalen der Chemie, (11), 1033-41; 1988.

            By the way: Burns, Christopher J.; Gill, Melvyn; Saubern, Simon, Australian Journal of Chemistry, 44(10), 1427-45; 1991 – again opening of very similar epoxide in the total synthesis of 6-demethoxyaustrocortirubin with AcOH led to the mixture both regioisomeric products.

            So, it is not so obvious until you try on real system.

          • ,,, says:

            to krest17

            OK, so you offer SN1, but you still get stuck with further elaborations. Unless you get your beta-chloride in the right position after your SN1, you’re likely to be in trouble.

          • krest17 says:

            I would try B. List chiral acids – might work. Also normally such a king of reactions going with the effect of “chiral memory” – so, I think there is a good chance to get Cl in a right place

          • krest17 says:

            Should not be a probelm anyway – the system is cyclic

          • ,,, says:

            Here we go, first chiral epoxidation, now chiral acid. By the way, have you seen any examples of SN1 installation of halohydrin in these kind of systems with gamma tert-centre?

          • antiaromatic says:

            maybe it’s just me, but don’t you think that if the epoxide idea was so obvious a disconnection that it would’ve been tried? maybe it just doesn’t work on the system…

            Sometimes, at least IMHO, if what you see isn’t the obvious operation, it probably is because it didn’t work. I mean, look at the acetate deprotection with Sm and I2. That is far from what I would call standard, and I can’t even imagine how many conditions they had to run through to find one that works if that’s what they ended up with.

          • krest17 says:

            Sorry for acid – surely chiral epoxide, it was early morning :-)) About “examples of SN1 installation of halohydrin in these kind of systems with gamma tert-centre” – Steroids, 72(1), 95-104; 2007, Tetrahedron, 63(37), 9221-9228; 2007 and many more – by the way opening with BiCl3 might work for this case as well. About suggestion of “antiaromatic” – I think the selling feature of this paper is chiral chlorination, which is amaizing – if they would go via “Jacobsen rout” – what is the aim then?

          • ,,, says:

            to krest17

            Tetrahedron 2007 you cited is a completely different situation: tert-center is on opposite way from the desired attack direction in this case (and outcome and the process itself are obviuos and boring). Why would you give this reference if it doesn’t represent the case at all? (haven’t checked the steroids yet)

          • krest17 says:

            Surely it is different situation, different molecule, and different system – but it is an example of installation of halohydrin with gamma tert-centre, I thought that what you asked. Anyway, I think is it time to stop this hypothetical discussion, not to abuse other people to much with it. I believe that any transformation can be done – just matter of time and efforts.
            Thank you very much for discussion – I truly enjoyed it.
            By the way – I am he, not she.

          • ,,, says:

            this was very unethical of you. The difference between current paper and examples you’re giving is immense (BTW including Steroids paper). This is not a contest in any way. I’m just trying to avoid other people getting confused with you representation. Cheers

  • Madforit says:

    to ch3mical:
    TBS is shielded by other functional groups in the molecule(quatern.stereocenter), TBAF in THF hardly remove that PG…i don t konw what it is, slightly acidic water cannot remove a secondary TBS..Thanks anyway man, cheers.

    • synthon says:

      one thing people tend to forget with TBS groups is that, while more acid stable than TMS, they are in fact more base labile

  • Pilky01 says:

    I like NGP’s answer ‘anchimeric assistance’, but im also a bit partial to a tight ion pair. I think both are plausable. I think the concept of a tight covalent pair will also help Madforit’s chemistry. LiOH is a pretty strong base because the lithium ion can double up and activate any Lewis base type moities by acting as a Lewis acid. (in your case the 1,3 related oxygens to form a nice 6 membered ring) Removal of a silyl protecting group with something as harsh as LiOH is not a huge suprise to me. In general LiOH>NaOH>KOH in terms of basicity, and that reflects the Lewis acidity of the metal cation, so i would try saponifying with KOH see what happens. I hope your molecule can take a good heating!

    • LiqC says:

      There’s hydroxide and there’s water, that is quite capable of taking off this TBS, there’s not much of lithium’s Lewis acidity left in here. I’d guess there could be some intramolecular business, too. Perhaps try TMSOLi or Me3SnOLi?

  • Madforit says:

    Thank you so much guys…

  • Heiko says:

    to the TBS cleavge:
    there was a publication last year in which they cleaved a TBS-ether with Cs2CO3 simply by heating it in THF i think…
    maybe its a comparable situation here…

  • The Next Phil Baran says:

    They trollin’, they hatin’. Loves it.

  • ,,, says:

    To Madforit

    It might be a little off what you’re looking for, but if you could start with something like benzyl ester instead of methyl, you would be able to avoid nucleophilic/acidic reagens and simply hydrogenate (if you don’t have unsaturation, of course :) )

  • Fluorine says:

    KOH, up to 10 % in dry MeOH at RT has always worked well for me, sometimes even in stoichiometric amounts

    anyway the Siether is susceptible to base hydrolysis, the ester should be blown of within minutes, so how long did you stir it?

  • Madforit says:

    Thank you fluorine and …i have H2 sensitive group in the molecule so the benzylester is not possible to put in.
    to fluorine: i tried for the first time the saponification on my molecule so i think the stirring was about 30 minutes,then TLC.In the meantime the silicon migrate on another free OH group..a nightmare..four different products!!the beauty of the total syntheses.Cheers,sorry for the english!!

    • cascade says:

      Just so that you know, not all benzyl esters are hydrogenalized at atmospheric pressure. I would reccomend trying that step with the benzy ester.

    • Tok says:

      Benzyl esters usually come off like a shot, much faster than other hydrogen sensitive functional groups. Check out the transfer hydrosilylation of a benzyl ester in the presence of a disubstituted olefin and benzyl ether in CEJ 2008 p.4293 scheme 15.

  • Matt says:

    Hi guys, I have a question, there is beta silicon effect, but is there any known case of beta halogen (iodine) effect. Thanks

    • Liquidcarbon says:

      Uhm… makes alpha proton acidic?..

      • Matt says:

        I know but does halogen, especially iodine stabalize beta carbocation ?

        • The Next Phil Baran says:


          This seems reasonable to me since halogens can stabilize an alpha carbocation (bromonium ion for example)and I am guessing that a four membered ring intermediate would be more stable than a three membered ring, so yeah why not. The size of an iodine atom should make this even more favorable. I would like to see what would happen with a geminal iodide next to a carbocation or beta to it, that would be cool.

          • ,,, says:

            just not to mix up: you’re confusing alpha and beta positions – in this case 1,1 and 1,2 relation

        • aa says:

          not in the way you are thinking. beta silicon effect is hyperconjugative and inductive. see jorgensen mid 80′s jacs papers on this… any stabilization from halogen would be iodonium/bromonium etc as in halogenation of alkene.

  • Chem-is-try says:

    On what basis is the stereochemistry after the reduction with potassium triphenyl borohydride determined? How could they predict the hydrogen will be beta?