Home » Still In The RBF

Oseltamivir phosphate (Tamiflu) Pt. 4   

9 June 2008 15,058 views 24 Comments

Zutter, Iding, Spurr and Wirz1. JOC, 2008, ASAP. DOI: 10.1021/jo800264d. Article PDF Supporting Information Group Website ResearchBlogging.org

Again, I’m stretching that brief to include Tamiflu, basically because I find it such an interesting target. The more click-happy readers will have noticed that this paper comes to us from The Professionals at Roche, who make this bad-boy their business. And quite a synthesis it is too. Retrosynthetic analysis won’t get us anywhere here, so lets leap in with the forward.

Now that’s what I call a reduction! 5% Ru-Al2O3 with a fair old pressure of hydrogen resulted in an all-cis configuration in a respectable yield. You’ll also note that whilst this reduction wasn’t asymmetric, the product is meso. So next up (after demethylation) was a highly efficient desymmetrisation, using pig-liver-esterase, resulting in the desired product in excellent enantiomeric excess.

So far, so smart, but many a synthesis trips up at this point with a stumbling series of functional group transformations to get to the target. However, the folks at Roche used an interesting and effective strategy. A bit of diphenylphosphoryl azide and base resulted in a Curtius rearrangement in which the isocyanate was trapped by the neighbouring hydroxyl group to give an oxazolidinone.

Boc-protection followed by base led to a ‘transcarbonation’ (or something) type process forming a bridged carbonate. This compound is then set perfectly to do an elimination/decarboxylation process, installing the required unsaturation. Completion of the molecule required a more familiar approach of displacement of the hydroxyl group (as it’s triflate analogue) using azide and then reduction, and a few more less interesting steps. However, the synthesis and strategy are compelling, and well described in this paper. Read!

[1] This dude has an awesome name: Beat Wirz! ‘spec!

Zutter, U., Iding, H., Spurr, P., Wirz, B. (2008). New, Efficient Synthesis of Oseltamivir Phosphate (Tamiflu) via Enzymatic Desymmetrization of a meso-1,3-Cyclohexanedicarboxylic Acid Diester. The Journal of Organic Chemistry DOI: 10.1021/jo800264d

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)


  • Klug says:

    The PDF links to another article from ACIE.

  • Tot. Syn. says:

    Cheers – fixed!

  • milkshake says:

    Behold the power of desymmetrizaton! To me this clever FG manipulation scheme is incomparably more beautiful than a synthesis of some polyketide

  • ch3mical says:


  • the dude says:

    Hi Tot. Syn.

    not being Travis ‘T-Bone’, TheEdge, Stewie, or Walter, but being capable of speaking German, I just wanted to let you know that the first name of the Wirz dude is actually pronounced “Beh-at” and not beat. It is a common Swiss male name and derives from latin, beatus, happy. The name Beatrice comes from the same source.


    the dude
    (or duder, his dudeness or il duderino, if you are not into the whole brevity thing)

  • Madforit says:


  • ch3mical says:

    Tot. Syn.

    What’s going on with post dates?

  • Tot. Syn. says:

    Please don’t tell you’re confused by international dates… http://en.wikipedia.org/wiki/Calendar_date
    Remember this is a UKian website (even if nearly 50% of the traffic is USian)!

  • ch3mical says:

    yea im dumb

  • a. d. says:


  • PotStirrer says:

    Wow, I love this synthesis. Right on the mark there, Milkshake.
    On another note, does anybody have some good tricks for suppressing Wurtz coupling when forming a Grignard of an allyl chloride? I want to then react it with paraformaldehyde, but am losing more than 50% of my SM to the Wurtz.

  • milkshake says:

    I dont know if its any help but about decade ago I was working with BrMgCH2CH2CH2MgBr, which is made from 1,3-dibromopropane and Mg turning – and this bis Grignard is exceptionally difficult to make because it likes to Wurtz to cyclopropane and BrMg(CH2)6MgBr. In THF the Wurtz products dominated. I was desperate and even started messing with Riecke Mg.

    But eventually I found that I could make the bis-grignardo-propane reagent with only few% of the dimeric BrMg(CH2)6MgBr in ether as a solvent, at room temp, with a decent yield. The trick was to use a fairly high dilution (0.05M if I remembre correctly), and syringe-pump the dibromopropane over several hours into Et2O+Mg in a big flask – and use a large excess (more than fivefold) of Mg turnings that were freshly activated by dry-stirring the Mg turnings with a big-ass stirbar egg at max speed, with added few drops of Br2 under dry Ar overnight. (Dry stirring crushes the Mg turnings which creates fresh surface which gets temporarily protected as MgBr2. Anh ether is then canula-transferred from the still to this mix and the Br2 color promptly disapears. It is important to work under Ar because activated Mg reacts with nitrogen)

  • Potstirrer says:

    Thanks Milkshake. I was reading the JOC paper (don’t have the reference with me at home right now) that describes the stirring under inert gas activation trick just yesterday. They claimed that argon vs nitrogen makes no difference, however. I tried that reaction today at relatively high dilution but the thing just won’t initiate until most of the chloride is in there and then it all goes to Wurtz. I’ve seen procedures where people add a mixture of aldehyde and halide together. Do you think I can suspend paraformaldehyde with activated Mg to immediately react with any Grignard as it forms?

  • milkshake says:

    I dont know. But people were doing aldehyde+halide Grignards on ultrasonic bath and claimed that sonicating allowed them to do it all one pot, Barbier-like

  • Joe says:

    Forming a grignard reagent from an allyl halide is best done by the jaworsky method not too different from what milkshake did. This minimizes wurtz-coupling. See JOC, 1942, 326 for procedures and refs – worked well for me!

  • TWYI says:

    Agree milkshake. I have lost count of the number of times I have seen people ‘activating’ thier magnesium turnings under a nitrogen atmosphere then wondering why simple Grignards are going in 15% yield.

  • antiaromatic says:

    As inconvenient as it is, I have found that it typically just is not worth purchasing Grignards, except for the simplest of them (Me, Et, etc.), as they go bad extremely quickly and often times aren’t as pure as you’d like. I have had no less than 5 bottles of vinyl MgBr give me strange products that never show up if I make the Griganrd myself.

  • dchemkid says:


  • InfMP says:

    I was just planning my first grignard. thanks for the advice (I had planned to order vinylMgbr). hopefully ill find a org syn or something?

  • antiaromatic says:

    There definitely is an orgsyn prep for vinyl mgbr, but it deals with using vinylbromide as a neat liquid. If at all possible, I advise against this, as vinyl bromide is exceedingly toxic and exceedingly volatile. You can buy vinyl bromide as a 1.0 M solution in THF that you can use to prepare your Grignard. The only catch is that you need to use a dry ice/acetone reflux condenser or cold finger while preparing it because the vinyl bromide is too volatile to really cool efficiently under normal running water. In my case, flame drying the turnings in the flask and stirring the turnings with a stirbar under vacuum for ~ 2 hr prior to starting was sufficient. The Grignard won’t initiate immediately b/c the solution is relatively dilute (you usually use much more concentrated solutions to initiate), so give it about 30 minutes with gentle warming with your hands, and you should be set!

  • InfMP says:

    wow. that the fastest reply ever. 2 min.

  • milkshake says:

    Aldrich iPrMgCl and sec-BuMgCl actually do not spoil that easily if you tape over the septum and make the hole smal. You keep them in cabinet and they are fine. One reason why Grignards spoil is that people store them in fridge – which usually causes various magnesium complexes to precipitate out and form a thick layer at the bottom, one which is hard to re-dissolve.

    But I agree that vinyl and allyl Grignards do not store well and thse commercially available ones are often full off dark nasty stuff. Maybe it is something related to C=C

  • antiaromatic says:

    I agree with milkshake. The vinyl Grignards (bromide and chloride variants) sold by Aldrich show an astonishing variability in both color and opacity. The one bottle that I got that worked well was light tea colored. Every other bottle I got was completely black and opaque. Within 24hr of puncturing the SureSeal, you are almost guaranteed to see precipitate forming on the walls of the container (stored at r.t.) that absolutely do not redissolve.

  • mr_pish says:

    Reckon the ALdrich Sure-Seals might be to blame…..have never bought vinylMgBr from anyone else but Aldrich secBuLi & tBuLi can’t be stored at all and are often rubbish when delivered. The Acros seals are way way better and you get a nice bottle afterwards.