Home » Still In The RBF

Structure Revision of Hexacyclinol / Total Synthesis pt. III: Synthesis   

27 July 2006 10,637 views 32 Comments

Hexacyclino_4.jpg

Porco, Su, Lei, Bardhan and Rychnovsky. ACIEE, 2006, Early View. DOI: 10.1002/anie.200602854.

So the dust has settled – and it looks like Rychnovsky was right. But lets disregard the controversy for now, and look at the synthesis. John Porco’s approach to this reassigned structure bears relations to his work on the similar natural product, panepophenanthrin.

Hexacyclino_5.jpg

This work hinges on the fact that both targets are actually dimeric; however, in the case of Hexacyclinol, a further cyclisation has occured to deliver the product. Thus, they created the monomer in four steps from the literature, and noticed spontaneous cyclisation immediately, using TREAT-HF to perform the deprotection of the TES group. This funky reagent is Et3N.3HF, and is apparently a far safer fluorinating agent than the more common HF.py.

So with the monomer in hand, they left it sitting around neat for 3 days, and acheived a 87% yield of dimer (my kind of reaction). The final cyclisation was promoted by K-10 clay, acting as a Lewis acid, delivering hexacyclinol in 99%. Nice! And in far fewer steps than La Clair’s “synthesis”…
Hexacyclino_6.jpg

Additional:

For more about TREAT-HF, have a look here.

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)
Loading ... Loading ...

32 Comments

  • movies says:

    Missed a double bond in your D-A precursor!

    I love this synthesis. It’s so concise and takes advantage of the symmetry of the molecule. It reminds me of Chapman’s synthesis of carpanone (JACS 1971, 93, 6696-6698).

  • movies says:

    There is a double bond missing in the product of the D-A too.

  • Mitch says:

    Also missing in the same precursor is the alcohol group that performs the cyclization. ;)

    Mitch

  • Mitch says:

    Seems like Chemical Forums staff to the rescue… Or at least to the nag. :P

    Mitch

  • Tot. Syn. says:

    And a down H too. Sorry, posted the wrong version of the file…

  • secret milkshake says:

    Porco’s synthesis is a total sweetness.

    HF/py is more like 9HF.py by stoechiometry. The stuff is brutal – we were doing transformation of aminopyrimidines into fluoropyrimidines at one time, with tBuONO and neat HF/py as a solvent (50-500mL of it each time). You have to use plastics and the quench (by dumping it on ice) is dramatic, the produced HF fog is not healthy. I heard that NEt3.3HF does not eat glass.

  • Tot. Syn. says:

    I’m used to using TBAF, which is a nice reagent for removing those pesky Si protecting groups. However, in my undergrad project, we had a b*stard of a peptide coupling, and eventually had to use the acyl fluoride to couple with the amine partner. That was made using the acid and cyanuric fluoride – which is pretty far up the Seriously Nasty scale of things. Had to use DAST too…

  • secret milkshake says:

    There is a much better reagent for generating acyl fluorides in situ, TFFH (Me2N(+)=CF-NMe2/PF6-). (Hunig’s base, R.T., 15 min). You can buy THHF from Aldrich or make it in one pot from tetramethylurea.

  • Tot. Syn. says:

    Ashutosh: I’d remove the statement saying that you’ll mail Nature articles to those without access – not a good idea – their legal team is pretty good, I know one or two of their employees read this site… Also, the picture of La Clair stolen from the ACS website is a poor idea too.

  • Tot. Syn. says:

    Secret Milkshake: Thanks for that. I’ve drawn out the reagent:

    TTFH

  • TPG says:

    La Clair has posted again on tenderbutton: http://blog.tenderbutton.com/?p=165#comments

  • secret milkshake says:

    thank you, Tot Syn for drawing the structure – I can post the improved prep for TFFH from tetramethylurea (about half a page) that we published in Org.Prep.Proc.Int in old times, when it was a tolerable journal.

    La Clair is realy nasty and threatening everyone who would repeat his “explanation” from Gordon – his story that involved Brazilian spies, American embassy in Berlin and US army genereal, re-synthesis of FIFTY grams of material and some dead human guinea pigs. I think JJ must be getting his inspiration from Ron Hubbard.

  • Desilylating says:

    Secret Milkshake, could you post the reference for TFFH? Thanks!

  • secret milkshake says:

    OPPI Volume 29, No.4, 1997, page 497-8
    The reagent was developed by Carpino, we just improved the scaleup. Dimethyl urea + oxalyl chloride to make the chloroformamidinium chloride, which is decanted and then heated with KF + KPF6 in acetonitrile in the same flask to get TFFH in 86-90% overall Y.

  • TPG says:

    In addition to the very concise nature of Porco’s synthesis, I like the thoroughness of the supporting information. I was wondering about one note, though, at the end of the opening paragraph: “The ArthurTM Suite Reaction Planner (Symyx Technologies, Inc.) was used for experimental procedure planning.” Has anyone used this software before?

  • Tot. Syn. says:

    La Clair has taken an interesting approach to his detractors and nay-sayers; one of complete confrontation and confusion. Reading his comments on Tenderbutton hasn’t helped me understand the situation any better, and has only made him look more like a nut-job. I don’t know anyone who was at the GRC, but the story posted by Mitch at Chemical Forums… and now taken down again… seems to be one of complete paranoia and delusion. Why is he being so confrontational about this? Mitch, you can post that story if you like – it’s not like you stated it was fact, just rumour… post it here under a pseudonym if you want; all you need to do is insert the odd “alleged” to avoid allegation of liable.

  • evgeny says:

    When I heard of that software, I thought, ‘finally, something that will tell me what to do in a few seconds after I imput my target!’. Then I went to their website and saw their demo. Basically, it’s the electronic version of a lab notebook. You draw your reaction and enter the equivalents and it calculates the molecular weights for you and the grams you need to add for each. Then if it knows the density, it tells you how much you need to add by volume. I’m not sure if it would save time, or waster it for a synthetic organic chemist, since a lot of the reactions you do will probably have materials that are not in the program’s database and you’ll have to enter the structure and density by yourself. It can generate a name though…

    Still, it’s probably faster to sit down with a lab book and a calculator for a couple of minutes.

  • Tot. Syn. says:

    We have to use an electronic labbook at work. It’s shit. I mean *really* shit. It uses MS Word templates with Excel sheets for the quantities, and ISIS Draw for the schemes. However, the they’re not dynamically linked, so Excel doesn’t know the molecular weight until you tell it yourself.
    To make matters worse, it’s all on the network (read NotWork), so at times you can’t save. Or edit your structures. Or open your labbook. Durrrr….

  • evgeny says:

    That’s what this ArthurTM Suite was all about, except that it’s supposed to work well. I’m still waiting for a program that will come up with a retrosynthetic analysis, or a reaction scheme, when I draw a natural product and press ‘OK’.

  • Tot. Syn. says:

    Argh! No! That’d put us out of a job – they could just get the trained chimps in instead. Nah, it could perhaps spot the major disconnections okay, but you’d end out with it doing loads of aldols and then reducing the carbonyls with CBS conditions and burning excess hydroxyls off et c. It’d never see the crazy disconnections that Woodward was known for, and probably wouldn’t see the dimerisation to get hexacyclinol for instance.

  • TPG says:

    I think Corey tried to develop such a program in the late sixties, he talks about it quite a bit in The Logic of Chemical Synthesis. There is a Baran group seminar presentation on computers in organic synthesis at http://www.scripps.edu/chem/baran/html/meetingschedule.html
    From talking to a former grad student at BU, I found out that in addition to the E-notebook features of the Arthur program, it also links up to a database of known transformations, allowing you to design a scheme from chemistry already worked out by the group or company. The downside is that you need to manually create the database.

  • Tot. Syn. says:

    I love ChemDraw’s NMR prediction functions (even if the chemical shifts are way-off… damn solvent effects), and the stuff pioneered by Rychnovsky et al is fantastic. CamSoft’s eNotebook is a significant step forward from what I’m using just now, and the database software (ChemFinder) is superb (I’ve just set one up back in academia), so you could say I’m a big fan of computers in organic chemistry. But when you need software to tell you to do an RCM, or macrolactonisation, I despair. I know that the computer could “memorise” every reaction known to man, but it still couldn’t have “style”. We all know the difference between an ugly synthesis and a sweet synthesis, but the computer wouldn’t. Sure, it could count the steps and arrive at a Longest Linear Sequence number, but that’s not the be-all/end-all.

  • Disillusioned Hamster says:

    ChemDraw’s NMR prediction function is not a good thing. I’ve met some people whose idea of analysing/assigning a spectrum is to print out the ChemDraw prediction. I’d much rather see people taking the time to go back and record extra 2-D data and think if they need to fully assign a spectrum, rather than just writing down the order of peaks the software gives – as you say, some of the shifts are way off.
    I’ve never used an electronic notebook – I imagine there isn’t a terminal by every fumehood so how does recording ones work on computer square with the recommendation that observations should be permanently recorded straight away rather than scribbled on a bit of paper and written in the lab book later?

  • Tot. Syn. says:

    I agree – you can’t assign a spectra with ChemDraw’s NMR function, but you can use it as a guide to what’s going on with your spectra. You know that feeling – you’ve just done a particularly interesting transformation, and just got the crude proton back. It’ll be a couple of days before you’ve got the HMQC and COSY back, so it’s always reassuring to see the ChemDraw agrees somewhat. Even though the chemical shift is often way off, the splitting and even the fine structure can often be spot-on.
    You’re right about the recording of some of the reaction data – I use a PC in an attached office to write my labbook. That works pretty good, but it’s not ideal. Of course, you can’t take TLCs out of the fumehood, so recording them in the report can be a little tricky. Rf’s are great, but there can often be subtleties, such a shape, time to develop (great for differentiating acetylenes and olefins) et c.

  • TPG says:

    I’ve seen several people obtain permanent images of TLC’s by placing them onto a copier or scanner (a blank transparency employed, of course, to protect the glass surface). I tried this once, placing several plates in a sheet protector to cover both sides and then sealing it shut before placing it on the scanner. It worked out fine, though I tend to prefer simply drawing the plate in my notebook and adding any comments about appearance next to the diagram.

  • second gunman says:

    One of the subtleties of Porco’s synthesis (maybe not so subtle, since he comes right out and states it) is that both panepohenanthrin and non-La Clair hexacyclinol may arise as pure artefacts of the isolation (concentration, heating, methanolysis, etc.) of the desmethyl, allylically rearranged analog (a.k.a. panepoxydon) of the simple epoxy cyclohexenone shown in Tot. Syn.’s artwork, above.

    Thus, hexacyclinol is not a natural product, nor is panepophenanthrin. And this whole mess, as well as La Clair’s apparent fraud, is founded in an artefact of isolation!

  • Ashutosh says:

    Thanks for the reminder. I have removed the picture and the statement.
    Ashutosh

  • SJB says:

    Totally off tangent, but isn’t the picture at http://totallysynthetic.com/images/splash_jacs.jpg of a different molecule as well (fr1828*7*7)?

  • [...] This concise synthesis of various members of the Mitorubin family by John Porco might be less controversial than my last post on his work, but it’s no less impressive. The centerpiece (for me at least) is a pair of copper-mediated transformations, appending the second ring and completing the core structure: [...]

  • bingo cards generator…

    pooling rye uttered!…

  • [...] written loads about this in four posts (1, 2 – which contains my opinions, 3, 4), and frankly nothing has changed, other than the fact that some fancy DFT calculations have [...]