Org Prep Daily

August 11, 2008

The Road to Boredom (and back)

Filed under: industry life — milkshake @ 6:41 am

Derek Lowe of the Pipeline wrote a very insightful column in the last Chemistry World issue, about the narrow repertoire of synthetic methods used in medchem projects. His take is that the need for cranking out compounds for testing (=as many as possible) drives the chemists towards fool-proof reactions with a good functional group compatibility and building block commercial availability. The narrow choice often leads to insoluble series with poor oral availability and PK problems and insufficient structural diversity. More serious still is the turn-off effect on the chemists:

…For one thing, skills do need to be kept sharp, and running a variety of different chemistries is the best way to do that. And safety seminars aside, medicinal chemists do not actually have limitless capacities for boredom. Running yet another long line of palladium coupling reactions or amine displacements begins, after a while, to feel like working at a sawmill. The blade takes longer to cut through some of the logs than others, but the boards all come out looking about the same.

I think the problem of poor solubility and ugly design became more serious with the advent of kinase projects; the kinase slit-like binding sites typically extend far enough to accommodate ligands consisting of multiple aryl rings put together like beads on a string. Gleevec, Lapatinib, Dasatinib have this sort of structure – but then again, various ugly molecules are used in cancer treatment; I am unsure if this design will work for other therapeutic areas also.  (I have been making kinase compounds for a good part of this decade and I still like the smaller + more compact molecules better).

I met a number of chemists in the industry that already gave up on reading journals – their argument was that there were too many journals and too many articles and most of it was not very useful to them and whenever they needed to find a reaction the Scifinder and Beilstein-Crossfire search engines did adequate job. (Which is true). But I noticed that one typically loses his chemistry interest about the same time when he stops reading the literature; I don’t know what is the cause and the effect here.

More than 90% of synthetic chemistry is routine stuff – and frequently frustrating one, too. There are easier and healthier professions to chose from. I believe that most people got into synthetic chemistry because they experienced a sense of wonder, and they just kept coming back for more. Mixing up obscure, dangerous chemicals to obtain shiny crystals at the end is a pretty awesome and esoteric way to make living. Designing your own experiments and figuring out their problems, inventing tricks to make the chemistry work, trying your nutty ideas (to see if they translate into good compounds) is enormously gratifying experience. If the curiosity is quenched and the excitement is taken away all what’s left is shaking the sep funnels, putting flasks on the rotovap and analyzing fractions as they come off the column…

___________________________________________________________________________________________

There are various non-chemistry reasons why someone stops enjoying his particular project (personal-life problems, the lack of support, evil boss, problems with biology, the bureaucracy and politics, etc.) and the way to go about those difficulties is obvious. Instead I have few chemistry-related suggestions what to try, when the boredom and frustration takes over and the project becomes unsufferable:

1. Scaling-up. When chemistry does not work try some easy reliable procedure. Pushing through bulk material such as a commonly-used building block or an essential reagent that you made before helps to boost your self-confidence in times when your other chemistry failed you. Making a flask full of bright-yellow crystals is deeply satisfying and when you filter them, dry them and put them in a  bottle with a pretty label, when the NMR is clean, at least you don’t feel like a hack anymore. Besides if the material is useful to your group passing it selflessly around will make you popular.

2. Fluorine chemistry: Too many chemists opt to buy the fluorine-containing pieces without worrying how these molecules are made. There is a growing number of organofluorine building blocks available, companies like Apollo have a fat catalog full of them – but many important pieces are not available and easy chemistry to prepare them exists. Medicinal chemists shouldn’t be exhorted to work with F2 or anhydrous HF – but reagents like Deoxyfluor, Selectfluor, (PhSO2)2NF, NEt3.3HF are commercial, affordable, and are not too difficult to use (if the necessary precautions are taken). Many people are simply unaware that metabolically-stable difluoromethoxy group is readily introduced on a phenol OH by alkylating it with difluorocarbene (which is produced from chlorodifloroacetic acid) or that ArCF2CO2Et is easily made from Aryl iodides, BrCF2CO2Et and activated copper metal in DMSO under mild conditions, that CF3 group can be introduced  on aryl iodides with equal ease with catalytic CuI and sodium trifuloroacetate, that difluorocarbene in presence of PPh3 produces a Wittig reagent (that converts aldehydes to useful 1,1-difluoroalkenes), that trifluoromethyl anion generated from TMSCF3 adds to imines and aldehydes with ease etc.  There is a whole continent of a strange organofluorine chemistry and a medicinal chemist would do well to make himself familiar with these methods because apart from the nastiness of some HF-generating reagents these reactions tend to be well-behaved and predictable.

3. Metallation: Thanks to Schlosser and others there is a body of literature on position-selective lithiations of heterocycles, like pyridines and pyridimines, and there is also older literature on use of ortho-directing groups in benzene metellations. Very often these reactions are done with common reagents like BuLi, LDA, LiTMP, at simple conditions like THF -78C.  The functional group tolerability is not as great but this aryl CH-lithiation can be a powerful way of accessing simple building blocks with a special substitution pattern – a great help when changing the centerpiece ring in the series and run into the availability problem. 

Mg and Zn chemistry: Knochel is developing beautiful systems, for halogen-metal exchange, and many of his transmetallation reactions using secBu-Grignard and zinc reagents has been promptly adopted by process groups but medicinal chemists are lagging behind.  One can easily generate Grignards with ester ot even nitro group in the molecule, by transmetallation, and zinc reagents can provide much cleaner Negishi coupling than their Suzuki boronic acid counterparts.

Schwarz reagent, BBN-H borane: hydrometallation and hydroboration is a great way to access commercially-unavailable building blocks for Pd(0) catalysed cross-coupling reactions. BBN boranes do Suzuki-like sp3-carbon coupling with aryl, vinyl halides.

4. Cyclopropanation: Cyclopropyl substituents are of a great interest to a medicinal chemist on their own but they are great also for ring-opening reactions (they behave as a 3-carbon analog of C=C bond). In my opinion the enormously easy-to-use zinc promoted reactions (CH2I2 with ZnEt2)  and titanium-promoted cyclopropanation reactions (Cp2TiCl2 with Grignard) are neglected in medchem projects.

5. Asymmetric methods: There are so many of good ones and yet any medicinal chemist will always look first into buying the chirality – perhaps this has to do with the inconvenience of setting up a chiral column for analysis of the product ee. (One does not have this sort of excuse with chiral auxiliaries).  And some of the reactions are exceptionally easy to run, with commercial catalysts and and great functional group tolerance: Noyori Ru-TsDPHEN transfer hydrogenation of ketones, CBS reductions, Sharpless allyllic epoxidation, dihydroxylation and aminohydroxylation, Jacobsen asym epoxidation and epoxide hydrolysis kinetic resolution. There are Rh and Ru-phosphine C=C hydrogenation that go at modest pressures (50psi) and thus can be set on a Parr shaker. There is a whole field of organocatalysis, with simple catalysts like proline, doing great feats on simple aldehyde substrates – several stereocenters at once.  I think its mostly the laziness that medicinal chemists do not adopt these reactions more frequently even as most of them get trained to use them in grad school; as soon as they join pharma they learn to weed out the chiral centers from their molecules. Nature is chiral and even as it takes more time and effort to employ the asymmetric methods the methodology investment can be well worth the trouble – with a functional-group tolerant method like Noyori transfer hydrogenation, the development work needs to be done only once – and from then on the chemistry is just as easy to perform as a borohydride reduction. 

6. Making your own heterocycles. When you buy a boronic acid piece and slap it onto your molecule you make your final compounds faster – but often it is worth looking into how a particular ring system is made. The chemistry used may be ancient or completely new (Padwa is my hero), one can make everything from TOSMIC, etc. You would be amazed how many 5-membered rings are easily available from aldehyde or carboxylic acid.

7. Screening the ligand/solvent/base etc. Even the robust Pd(0)-catalysed reaction can become finicky for a particular class of substrates. There is the base, Pd-source, ligand, and solvent to choose  (apart from the temperature, concentration and time) and if you get one parameter wrong the reaction usually fails or stops at low conversion. Sometimes little time spent on methodology optimization pays off handsomely.

You should care about the chemistry methodology and do things not just to crank out the final compounds to fill up the testing queue. Your boss perhaps lost all his chemistry interest already and maybe he is unnerved about the project progress and pushes people hard –  but while you try not to get fired you don’t necessarily want to think like your boss (and end up wretched). If you continue to look at your research project with curiosity and do things also for the sake of your chemistry interest you are likely to be more original because thinking about the methodology will suggest new directions in your medchem project.  You may get acused of playing with chemistry and going off-tangent but you will likely remain more content and productive and you will continue to live your life in the lab – which could be a good or bad thing depending on the marriage situation.

28 Comments »

  1. 9-BBN/Suzuki is an awesome way to alkylate pretty much anything.

    Comment by excimer — August 11, 2008 @ 8:59 am

  2. What is important in the lab? “To crush my synthetic targets, to see them in Stage I testing, to hear the lamentations of my manager who made big noises about doing it another way.”

    When you need research in the worst way possible – Uncle Al’s way. “8^>)

    Comment by Ubcle Al — August 11, 2008 @ 11:24 am

  3. Jeeziz milkshake, TL;DR

    Comment by Gus Musselmann — August 11, 2008 @ 2:34 pm

  4. meh. It’s a living.

    Comment by jc — August 11, 2008 @ 6:12 pm

  5. Milkshake and Derek Lowe and my Heroes… Oh, and also Paul Docherty…

    You guys amaze me…

    Javaslinger

    Comment by Javaslinger — August 12, 2008 @ 12:21 am

  6. I had a postdoc tell me, “If you feel like a total hack, go recrystallize an old, nasty bottle of NBS; you’ll feel like a hero in no time.”

    Comment by Jose — August 12, 2008 @ 11:29 am

  7. Jose-

    I agree. Nothing makes me feel more awesome than making huge colorless crystals of NBS from the dirty yellow turds Aldrich once claimed to be NBS, ten years ago when the bottle was purchased.

    Comment by excimer — August 13, 2008 @ 10:18 am

  8. Huge colorless crystals of NBS? Awesome! I should make some… How did you do it? Crystallization from water? Doesn’t NBS hydrolyze in water? What is that orange-yellow stuff in old NBS? Btw. Does anyone know how to get rid of Pd(0) traces in the presence of an aldehyde, which is the desired product?

    Comment by Danny — August 13, 2008 @ 11:43 am

  9. Off-topic: Is there a website to get PDF files for the European Patents (EP) ?

    Comment by pmgb — August 13, 2008 @ 6:28 pm

  10. http://www.wipo.int/pctdb/en/

    Comment by milkshake — August 13, 2008 @ 6:40 pm

  11. Another option; the number search here works better than anywhere else, IMO.

    http://ep.espacenet.com/quickSearch?locale=en_EP

    Comment by Jose — August 13, 2008 @ 6:54 pm

  12. Thanks.

    Comment by pmgb — August 13, 2008 @ 7:12 pm

  13. Danny: red colour from old-NBS is from Br2 present. I guess even traces of Br2 can give some orange-red colour. Crystallization from water is the choice but I had always poor recoveries (around 60-70%) so I guess some NBS is destroyed.

    To get rid of Pd, SiO2 filtration, active carbon, expensive fancy specialties (e.g. JM fibers)… See http://pubs.acs.org/cen/coverstory/83/8336chiral3.html

    Comment by Vasili — August 14, 2008 @ 5:29 am

  14. Pd removal – very commonly done with resins if you can afford the price tag. Works like magic. Polymer Laboratories sells many different kinds. Removal down to <50 ppm easily.

    Comment by SciChick — August 14, 2008 @ 6:32 pm

  15. A Recent OPRD paper discussed palladium removal to < 100 ppm by just washing w/ 20% aq. NaHSO4/toluene

    Comment by Gus Musselmann — August 14, 2008 @ 6:53 pm

  16. Comment by RATM — August 15, 2008 @ 2:42 am

  17. Thanks, guys a lot for your help. Removal of Pd(II) should not be a problem, but colloidal Pd(0) could be. My aldehyde is pretty expansive homemade stuff, VERY lipophilic, base sensitive, and probably could react with NH2, SH, o-NHPhSH, active groups present in metal scavenging resins. But I am going to try thiourea resin, although it supposes to be weaker Pd(0) scavenger compare to thiol resin. Or guanidine resin in acetate form, to suppres basicity.

    Comment by Danny — August 15, 2008 @ 10:59 am

  18. Off-topic: How to deprotect a benzyl group on an amide N (benzo[b][1,4]oxazin-3(4H)-one system) ? I tried these conditions(i) 10% Pd/C, H2, 60 psi, EtOAc, rt, 12 h; (ii) HCO2NH4, 10% Pd/C, methanol, reflux, 12 h and (iii)TsOH, toluene, reflux, 12 h but could not get the debanzylation product. I found a paper that used MsOH (1 h), toluene, 6 h, reflux to cleave a N-(1-phenylethyl) unit [06TL1813]. Any other suggestions?

    Comment by pmgb — August 18, 2008 @ 5:11 pm

  19. Hi pmgb – I had to first translate your IUPAC name as “N-benzyl-benzomorpholine with a carbonyl next to N”. So you have N-benzyl N-aryl lactam kind of product. I think hydrogenation over Pd will not work (based on my bitter experience with N-benzyl beta lactams). PMB- would have been better choice because one could either oxidise PMB away with CAM or cleave it acidolytically with triflic acid (in TFA solution) or with TFA and thioanisole and heating.

    Since you have made the benzyl-protected material already, I would recommend NH3(liq) with THF at -40C, with Na metal – you probably want to titrate the reaction by adding small bits of Na and try not to exceed 2 equivs of Na at first, and as you get close to 2 equivs the reaction should slow down and the blue color should persist for longer time – a sign that your benzyl is mostly gone and now the sodium electrons are munching on somehing else.

    The yields are typically unimpressive – but I knew some old-fashioned peptide chemists that treated their oligo-peptides with Na/liq NH3 to burn off benzyls and tosyls – and apparently they got some material out of the resulting mess. Again, be careful with sodium excess, rather use less then stoechiometric amount, then sample the mix and TLC, then add more Na as needed, etc.

    Sorry I don’t know of any better solution – and please don’t curse me if your molecule gets over-reduced in the process.

    Comment by milkshake — August 18, 2008 @ 9:32 pm

  20. Thank you very much Milkshake.

    Comment by pmgb — August 18, 2008 @ 10:14 pm

  21. Theodora Greene’s “Protective Groups in Organic Synthesis”:

    “Benzyl groups, as well as other alkyl groups, can be converted to various
    carbamates as a variation of the von Braun reaction. The carbamates
    can then be cleaved by conditions that are outlined in the section on carbamates.”

    There are several examples of such deprotection, e.g. a-Chloroethyl chloroformate, CCl3CH2OCOCl, Allyl chloroformate…

    What about FRESHLY prepared Pd black in AcOH (ev. HCl addition)?

    Comment by Danny — August 19, 2008 @ 8:57 am

  22. Not on amides, Danny

    Comment by milkshake — August 19, 2008 @ 3:51 pm

  23. Excellent post milkshake. Curiousity and creativity are surprisingly rare these days. Med Chem projects can be run on semi-autopilot mode in many companies, without anyone having to really think HARD about what they’re doing and why.

    Comment by BCP — August 19, 2008 @ 8:14 pm

  24. I guess now it’s time for you to dedicate a full post to those fluorination methods. I guess that many people would like to get some more detailed procedures for those transformations from you.

    Comment by Me-Et-Pr-Chemist — August 21, 2008 @ 2:30 am

  25. I was going to post a nice procedure with Deoxyfluor – but a colleague thought it was too soon; the piece relates to our ongoing project in a fairly obvious way and this could become a patent claim issue should these things end up to be useful. There are all sorts of sensitivities whenever one gets close to anything of current interest in a medchem group – the stuff that wasn’t either dropped or patented/published already. I would hate to get into a situation where I was asked to take down the post or the site.

    There is a procedure in Org Prep Daily, I think late 2006 or early 2007, for coupling 2,6,-dichloroiodobenzene and bromodifluoroacetic acid Et ester with copper in DMSO that worked very nicely.

    Comment by milkshake — August 21, 2008 @ 12:59 pm

  26. Reg the removal of benzyl group……have you tried aq HBr/heat. Works very well for systems like the ones you are dealing with……

    Comment by moody blue — August 25, 2008 @ 6:18 pm

  27. Any chance of getting more detail or a reference for the CF3 introduction via an aryl iodide using CuI and sodium trifluoroaceate? Some preliminary tinkering hasn’t panned out yet.

    Comment by bones mcghee — September 24, 2008 @ 4:56 pm

  28. The paper reference is: J. Chem. Soc. Perkin I, 1988, 921-6, and the Chem Letters reference cited there.
    The same system works for the introduction of CF2CF3 group also but the yields are better for CF3. I don’t know if there are any more recent papers on this methodology – you would have to do a lit search.

    Comment by milkshake — September 24, 2008 @ 6:38 pm


RSS feed for comments on this post. TrackBack URI

Leave a reply to BCP Cancel reply

Blog at WordPress.com.