Home » Still In The RBF

ent-Malbrancheamide B   

16 March 2009 13,820 views 31 Comments


Simpkins, Frebault and Fenwick. JACS, 2009, ASAP. DOI: 10.1021/ja900688y. Article PDF Supporting Information Group Website

Sometimes, the sweetest syntheses come in the smallest packages, and with only one-and-a-bit pages, this is a very short paper.  However, Nigel Simpkins (now Haworth chair of Chemistry at Birmingham) manages to pack in quite a lot of cool stuff into that modest space.  A quick examination of the structure reveals a similarity to the stephacidin family, which has appeared in two previous posts here, one by Robert Williams, the other by Phil Baran.  However, there doesn’t appear to be a ‘family name’ – they’re just bicyclo[2.2.2]diazaoctane containing natural products.  They might be lacking in family identity, but the three stereocenters provide quite a challenge.

Quite a bright man...

Quite a bright man...

Simpkins begins the synthesis with a piece of classic stereochemical control – Dieter Seebach’s self-reproduction of chirality.  In fact, Simpkins uses a system identical to Seebach’s original paper – key to this it the presence of proline at the core of the starting material, which is condensed with an aldehyde (pivaldehyde, in fact) to generate a new stereocenter via 1,3-control.  Then, top quote Seebach: “The original centre is destroyed by deprotonation to give a chiral, nonracemic enolate. Attack at this enolate … is subject to asymmetric induction by the acetal centre. After this diastereoselective reaction, the auxiliary centre can be removed … to give the product of overall substitution with retention of configuration.” To reinforce what he means by self-reproduction – he means stereocenter A informs creation of stereocenter B.  If A is destroyed, it can be recreated with overall retention by B.  And it works very well for Simpkins!


Removal of the pivolyl N,O-acetal thingy (check wikipedia – that’s what it’s called… acutally, it’s Seebach’s acetal) to form a O-benzylhydroxamic acid, then coupling with a chloroindole derivative gave them the starting point for formation of the bicycle.  First, removal of the SEM protecting group (2-(TMS)ethoxymethyl) was done using rather exotic conditions – carbon tetrabromide in isopropanol.  Simpkins explains that these were used as TBAF and HF didn’t touch it – quite surprising, but at least it let them move on.  This deprotection reveals an enol, or more realistically, an ?-keto amide – which immediately snaps shut into the diketopiperazine.  Clearly, this little beastie looks a little fragile – that acetal centre with the N-O bond looks particularly labile.  So guess what happens when a bit of tee-em-ess triflate is added…


The hydroxyl group was removed by this treatment, with Simpkins suggesting that a formal carbocation is left as the reactive intermediate which reacts with the isoprene group and the indole to form both remaining rings in one process.  This penultimate intermediate brings convergence with the Williams synthesis, completing a rather neat piece of work, but I’m a bit confused by the cyclisation.  Y’see, when I first looked at the reagents, I expects that elimination of the alcohol had occured, followed by IMDA, then isomerisation of the resulting cyclohexene to provide aromaticity in the indole.  But Simpkins shows the carbocation intermediate as compound 5 in the paper…  Anyone like to help me?

1 Star2 Stars3 Stars4 Stars5 Stars (5 votes, average: 3.40 out of 5)
Loading ... Loading ...


  • gyg3s says:

    If I understand your question correctly …

    Alcohol elimination leaves a carbocation which reacts with the isoprene alkene to form the first six membered ring – well, the bicyclo[2.2.2]diazaoctane but the alkane part of it. But with a “CC(C)” (in smiles notation) carbocation hanging off it. The indole then reacts with the carbon cation in the usual manner.

    Or looking at it from the other direction; disconnect the product between the indole ’2′ carbon and the carbon in the alkane ring bearing the two methyl groups leaving a +ve charge on that carbon. Second disconnection is between the carbon next to the carbocation (the carbon bearing the two methyl groups)and the carbon that is next to the nitrogen and the carbonyl carbon, leaving a carbocation and splitting up the bicyclo species. The remaining carbocation is where the alcohol came from.

    Of course, you most probably knew all of that ‘cos I misunderstood your original question; apologies if that’s the case.

    • Tot. Syn. says:

      Thanks for writing; I probably wasn’t clear enough with my question… What I’m wondering is why Simpkins suggests an ionic pathway, rather than the intramolecular (pericyclic) cycloaddition I anticipated. If rather than forming a persistant carbocation, the alcohol did an elimination to form a diene. His carbocation is drawn adjacent to an nitrogen (surely the lone pair would get involved), and homo-allylic to the indole alkene (which must promote elimination). I realise probably a bit of both, but the distinction is important.

      • Brane says:

        A 4+2 cycloaddition is unlikely given that both the diene and dienophile would be electron rich. The enamide would always protonate back under strong acid so ionic mechanism could take place. That’s my guess

        • milkshake says:

          Nah, the diene is not so el rich, and besides you don’t need much electronically differentiated partners in intramolecular DA. The main reason why the DA mechanism is out is that with N-substituent R = OBn, the diene could never stay anywhere close to planar. And dienophile C=C is not anywhere near to synchronous 4+2 Always take time to build a reasonable model, you will immediately see all potential strains, clashes and repulsive interactions in 3D. (Works before a marriage, too).

  • ch3mical says:

    But, but you should always use the lone pair.

  • milkshake says:

    Its hard to tell the pathways apart just based on the desired product (intramolecular DA should also gove a mix of diastereomers, because the diene can be cis or trans on the newly crated C=C). For me the cationic pathway feels more natural because I used to play with oligoprenoid cationic cyclizations, and indole is such a great cation sponge. A proof would be isolation if 3,3-spiro-disubstituted isoindole compounds. Alas, those spiroisoindoles are not very stable and like to rearrange to 2,3-disubst indoles…

    CBr4 in isopropanol probably generates traces of HBr. I have seen a “magic” procedure for selective deprotection of TBS from primary alcohol in the presence of secondary OTBS, by dissolving the material in a CCl4+methanol mix and buzzing it for few minutes on a sonicator bath

  • ,,, says:

    Regarding the mechanism…
    Check out Baran’s work (stephacidin): he has never been able to do DA with similar intermediate.

  • gyg3s says:

    … I’m wondering is why Simpkins suggests an ionic pathway, rather than the intramolecular (pericyclic) cycloaddition …

    OK. Get you.

    Maybe he’s over egging the pudding by describing the carbocation route as, well, a carbocation. As you point out, the nitrogen’s lone pair is going to get involved almost immediately forming an ‘imminium’ species. Maybe the necessary loss of the proton to form the required (for subsequent DA) ‘homoallylic indole’ (or enamine) doesn’t have a chance to happen before the cyclisation occurs. I don’t know the pKa of that particular proton, it would be interesting to find out in a model system (anyone got access to Scifinder or Beilstein online?).

    I wonder if, “… says” (5:52) would give us a link to Baran’s work?

    As I write, I get the feeling I’m either telling you what you already know or just echoing the thoughts that you had when reading and providing the original post.

    (All very interesting, though).

  • Pilky01 says:

    The lone pair of nitrogen is involved in the elimination of the hemi-aminal (amide varient) to give an N-acylimminium (which he eludes to in the text but represents it as a resonance carbocation 7 in the diagram) This can then undergo an aza prins reaction to form the 6 membered ring, leaving a tertiary carbocation which is picked up at the 3 position of the indole (to make the spirocycle indoliminum species that milkshake mentioned) this then undergoes ring expansion (by tertiary alky migration) to give an indole cation intermediate which aromatises through proton loss.
    Personally i didnt see the IMDA as a possibility until i read it here, its nice to see a different perspective. I still think the aminal unit (even as an amide) will eliminate much more quickly than that required for diene formation. Once you get the N acylimminium it should be more reactive than usual because its alpha to another electrophilic centre (the carbonyl), making the aza prins cyclisation pretty quick.
    What do you think to this?

  • Pilky01 says:

    sorry i didnt mean that to read like a fact! feel free to shoot it down.

  • Jose says:

    Baran’s first non-JACS, non-ACIE, non-Nature paper!


  • InfMP says:

    yeah, it’s an invited paper, so it doesn’t really count.
    We got on that issue too, and we wouldn’t normally send to elsevier, but it would have been rude to refuse.

  • TWYI says:

    Baran also has a nice Heterocyles paper out too, another invited special issue I believe

  • John Wood says:

    anyone make vinigrol yet?

    • UBChem says:

      I would have made vinigrol if I didn’t have to graduate before my boss moved. But no, no ones made it.

  • InfMP says:

    BY THE WAY: in the abstract to that paper by baran, he says “a cheap organic base”
    Aldrich 25 mL = 700$

  • Smitty says:

    I see where 25mL costs $480 (are you talking about N,N,N,N-tetramethyl-t-butyl-guanidine?), which is still way too much to be called cheap, but can be made in one orgsyn-tried step in 75% from tetramethylurea (250mL, 91.50, or 100g, 49.30, depending on the quality) and t-butyl amine (100ml, 49.30 or 1L, 279.50).

    • milkshake says:

      The OrgSyn procedure is nice but uses triphosgene which can be costly- depending on supplier. Alternatively one can use oxaloyl chloride to form the tetramethyl chloroformamidinium salt intermediate. Just add 1.2 eqivs of oxalyl chloride to tetramethylurea in toluene at room temp, then raise gradually to 70C on oil bath with stirring and gas outlet, then cool and decant off the product.

  • Tot. Syn. says:

    I love how random pricing can be on supplier websites, especially the bigger ones like Sigma-Aldrich. I need to build a thio-furan soon, so I was ordering some ethyl mercaptoacetate – have a look at the pricing per gram:

    5G 8.00
    25G 53.90
    100G 9.10
    500G 29.50

    I guess the disconnect will be to do with packaging (this stuff *really* hums), but it’s only one example…

  • Ian says:

    In what world is 500 bucks for 25mL considered cheap?

  • Ian says:

    sorry 700 bucks for 25mL :shock:

    Fine if Scripps is paying I guess.

  • Pete says:

    Would have made vinigrol UBChem? Interesting…What was the key step might I ask?

  • UBChem says:

    Pete… you can read my thesis through SUNY Buffalo libraries.

  • InfMP says:

    yeah my pricing was in CDN, sorry

  • optional says:

    To Check sweet stuffs on Vinegrol, check out the work of two Ottawa Chemistry profs. Louis Barriut and I forgot the other guys name

  • aa says:

    the other u ottawa prof who has an approach to the vinigrol skeleton is alex fallis. i believe his key steps revolve around some diene-transmissive diels-alder reactions. unfortunately prof. fallis is winding down his lab, but he may still have some tricks up his sleeve.

  • chinstrap says:

    A nice article by Ken Miller of the realated versicolamides is free of charge in nature chemistry ASAP. They propose a DA biosynthetic pathway but with a slightly different disconnection than malbrancheamide.

  • InfMP says:

    I am really hoping someone finishes vinigrol soon. It’s offensive to me that such a small molecule can thwart so many people.