Org Prep Daily

April 29, 2011

Neglected lab solvents

Filed under: procedures — milkshake @ 9:34 am

Trifluoroethanol as a solvent dramatically accelerates epoxide opening with amines and also conjugate addition of amines to enones and acrylates. Thermal Diels-Alder is another good reaction to try in TFE.

1-methoxy-2-propanol behaves as a higher-boiling version of iPrOH. It has a reasonable volatility and it is miscible with water (so it could be pulled off on a rotovap – but also drowned if needed). It has no apparent toxicity issues (unlike methoxyethanol) . This obscure solvent was introduced as an eco-friendly paint thinner but it turns out to be quite useful for direct non-catalyzed SNAr of halogenated heteroaromatics with amines, a reaction that benefits from protic media but where the solvolysis poses a problem. I also used this solvent for high-temp cyclizations done in a microwave reactor.

tert-BuOMe: I suppose the minty reek turns people off but MTBE is a great ether replacement in column chromatography, salt precipitation and recrystallizations. Ortho-litiations have sometimes improved regioselectivity in MTBE,  and since this solvent is more resistant to cleavage by organolithiums, it is recommended for room-temperature lithiations with BuLi. MTBE typically does not have radical scavenger stabilizers and  the formation of peroxides is negligible – so when you are running a high dilution experiment or evaporating a large volume of fractions from column that contain only miligram quantities of your desired material, MTBE is a good solvent choice.

Acetonitrile: Some people use MeCN only for reverse phase HPLC and it is a pity – MeCN is a wonderful solvent for silylations, acylations, alkylations. Straight acetonitrile is also a pretty good solvent for re-crystallizing very lipophilic compounds.

Benzene is rather useful for recrystallizing stuff, especially with cyclohexane as an anti-solvent. (Benzene+cyclohexane mixtures have a nearly constant boiling point and do not freeze in the fridge). I have seen many examples where a compound could be successfully re-crystallized only from benzene – maybe this has to do with an empiric fact that benzene frequently co-crystallizes with compounds (so in such cases the solvate formation  changes the properties of the crystal structure).

Methanesulfonic acid: Old-fashioned cyclizations (Skraup, Isatin, Fisher Indole etc) are typically performed in neat concentrated sulfuric acid. Replacing sulfuric acid with MeSO3H frequently improves the purity and yield. The quench is also less exothermic with methanesulfonic acid.


  1. Please remove benzene from this list before someone tests it. It is a very nasty carcinogen.

    Comment by Spiro — April 29, 2011 @ 3:11 pm

  2. but it is not. People have exaggerated fears of benzene, especially in Western Europe, so they never use it – because they were told early in school that benzene is the Antichrist.

    Old-time synthetic chemists were washing their hands with benzene and eluting columns with it while working on the bench. It was the most common solvent used in paints and glues. So when it was finally recognized that benzene accumulates in bone marrow and chronic exposure can give you leukemia – and since benzene was so cheap and widespread in the industry – the campaign to phase it out was vigorous and benzene thus became an evil banned solvent. The argument was also that benzene can be fully replaced by toluene (which is not actually true in case of crystallizations). There are solvents that are more hazardous than benzene (CS2, CCl4, propionitrile) and they are still being used in the lab to run reactions. Benzene is not a solvent suitable for industry scale-up but in the lab it is perfectly useful with common-sense precautions. You just don’t pour it on yourself and you work with it in the hood. I would not recommend running a column in benzene and I would not use it for extraction-workup because there is too much splashing around – but in my experience a half-liter scale recrystallization from benzene can be easily done in the hood without much exposure. And I am prepared to work with benzene if I can get a improved product purity or extra ten percent yield.

    Comment by milkshake — April 29, 2011 @ 5:46 pm

  3. One underused solvent is water. I’m seeing so many procedures that perform the most trivial acylations and sulfonylations in MeCN or DCM, with organic tertiary amine base, followed by extractions – why?

    Schotten and Baumann are your friends.
    My standard recipe:

    For aliphatic amines, use 2M NaOH, start on ice. If acid chloride is liquid, add as is, otherwise dissolve in whatever you like. DCM works for me.

    For anilines, use 1M Na2CO3 at rt. Also use that for sulfonylations.

    90% of the time your crystalline product is isolated by filtration. If you’re unlucky, you can always go back to extractions.

    Also, no solvent is a good solvent sometimes.

    Comment by LiqC — April 30, 2011 @ 6:34 am

  4. Good to see you back. I and no doubt many others enjoy your posts. Keep it up.

    Comment by Young Padawan — May 1, 2011 @ 6:20 pm

    • there is not much to write about. These days I am sending out resumes, answering e-mails, updating the research summary, working phones, asking favors, despairing. Too much free time, no lab and no literature.

      Comment by milkshake — May 1, 2011 @ 8:39 pm

  5. Milkshake! I second the happiness of #4 at having you in the blogosphere. We also spent some time together at the same workplace a few jobs ago.

    For neglected solvents, I nominate linear hydrocarbons NOT named hexane(s). Heptane is cheaper, safer, and can be a hexane replacement in nearly all applications. Meanwhile, pentane is the savior of organometallic chemists everywhere, and is nearly as useful as ether for crashing out salts or drying out trace water.

    Comment by See Arr Oh — May 1, 2011 @ 11:29 pm

    • I spoke with a guy who had to use eucalyptol as a solvent for recrystallization. (His intermediate was making a co-crystal with eucalyptol and this helped him to purify it from some pesky isomeric impurities). He said he grew exceedingly tired of the eucalyptus smell…

      I wonder if dimethylcarbonate has any useful technical applications as a solvent – it is cheap, low-toxicity, it has a reasonable boiling point and a better hydrolytic stability than EtOAc.

      Comment by milkshake — May 2, 2011 @ 12:42 am

      • Do you know how he identified eucalyptol for making co-crystals? Literature or a good guess?

        Comment by JH — May 2, 2011 @ 11:39 am

        • It has been a long time, I only remember that he was making some intermediates with free phenol OH and had problem with regiochemistry. I bet it was not his first solvent choice.

          Also Brian Stoltz when he was working for prof. Wood on his Staurosporine thesis had some remarkable scale-up conditions that you don’t see often: He cyclized bis o-toluidine amide of oxalic acid to 2,2′-bi indole by using excess of molten tBuOK as a solvent and base, in a pressure vessel at 300 C. I think he found some lit precedent for that. Right the next step, it was Rh-catalyzed carbene insetion into the biindole, a cascade that assembled the whole polycyclic aglycone core in one step. It was a pretty finicky cyclization because the diazo carbene precursor liked to fall apart. Brian optimized it up to a 35% yield, by using tBuCOMe (pinacolone) as a solvent. I remember someone asked: why pinacolone? Prof Wood’s answer was that when you see your student going into a stockroom, loading a cart with every low-molecular liquid on the shelf and then testing them one by one as a solvent, you can safely guess that the methodology optimization has run into difficulties…

          Comment by milkshake — May 2, 2011 @ 3:01 pm

      • Milkshake

        You mention dimethyl carbonate which obviously has some uses at GSK. It is one of only 7 green solvents they prefer. t-Bu, iPr and Pr acetate are the other 3 esters they consider most acceptable.

        Comment by petros — June 8, 2011 @ 8:35 am

  6. Surprisingly wikipedia knows something about my question. The article on cocrystals says that eucalyptol was one of the first compounds to form cocrystals with aromatics without making pi-pi interactions, it appears that many phenols crystallize with it. Surely it has to do with the hydrogen bond acceptor ether group and possibility for close contact between methyls and aromatic pi orbitals.

    Comment by JH — May 3, 2011 @ 5:37 am

  7. I actually do like a somewhat more common (but def under used) solvent, dimethoxyethane. I also like MTBE a lot too! And i ❤ benzene, its so easy to work with/remove, its versatile and and its relatively low heat capacity makes it easy to heat up/cool

    Comment by ckellz — May 4, 2011 @ 10:01 pm

  8. Good to see you back, Milkshake! I agree with #7 about DME. I used it very successfully as a solvent for the alkylation of a hindered cyclohexanone enolate. THF or even THF/HMPA failed to give more than 10%, but with DME I obtained 87% yield. The only drawback was the need to distil it freshly from potassium to dry it. Sodium didn’t work.

    Comment by Nick K — May 5, 2011 @ 10:57 am

  9. There once was a kilogram scale synthesis where final product crystallized most perfectly only from diisopropyl ether. Which as solvent is real ouch regarding peroxides, but if it was needed to get clean product, then it was needed.

    And that may depend on lab and traditions, but people also tend to neglect MEK – methylethyl ketone. Dissolving properties similar to acetone, boiling point higher, only partially miscible with water – can even be used for extractions for extreme cases.

    Comment by Design Monkey — May 16, 2011 @ 10:34 am

    • I don’t have experience with iPr2O, I only heard that process people avoid it for that reason. But I did Finkelstein halide exchange with NaI in MEK couple times, and I have also done an aqueous extraction with methyl isobutyl ketone once (but I hate the smell of MIBK).

      Comment by milkshake — May 16, 2011 @ 3:08 pm

  10. Milkshake, I have a technical question. I want to repeat an experimental procedure from the literature where a nitrile group is reduced with hydrogen at moderately high pressure (280 psi) and temperature (120C) using Raney Nickel. I know there are better ways now to do this, but for other reasons, I need to follow the patent prep.

    The problem is that there are many many forms of Raney Nickel out there: Aldrich has 4 versions (4200, 3202, 2800, 2400) all priced about the same. What kind works best, or is most active? I don’t see any info online to choose one from another.

    Comment by partial agonist — May 21, 2011 @ 1:32 pm

    • I think Ra-Ni 2800 is the commonly used one, equivalent to the historic W-2 grade. All commercial active-form Ra-Ni catalysts are attenuated by their aging in storage. If you need a higher-activity catalyst than W-2 the best thing is to make the catalyst freshly, by digesting the Al-Ni alloy powder with aqueous NaOH and storing the catalyst in the fridge – but the high activity of the freshly-made Ra-Ni lasts only for some weeks and then it goes down to the W-2 level.

      The nitrile reduction with Raney Ni can be tricky because the intermediate aldimines do not hydrogenate fast enough to avoid side reactions: With Ra-Ni one typically obtains a mixture of a secondary and primary amine (secondary amine formation can be exclusive when you hydrogenate a dinitrile that can close a ring to piperidine or pyrrolidine), and with aliphatic nitriles you can generate C-C-linked oligomers as well. Also, acids deactivate Ra-Ni even more – when you hydrogenate nitriles on Ra-Ni in the presence of formic acid, you can obtain aromatic aldehydes from benzonitriles.

      Raney nickel and “skeletal nickel” catalyst is the same thing (Raney Ni is a trademarked name). Some grades are doped with a trace metal like molybdenum. To get a good recommendation which catalyst would be best suited for your procedure, you would need to contact a tech support of the catalyst manufacturer – or to find a good literature synthetic precedent where the catalyst grade is clearly specified.

      But maybe you can start with Ra-Ni 2800 from Aldrich – and give it a good water wash before the use. And if there are too many side reactions with your nitrile hydrogenation, I would then try to use the freshly-made Ra-Ni from digesting the Al-Ni alloy with NaOH.

      (The digestion takes only few hours to perform and you can do it in a large beaker in a rotovap water bath… Use a mechanic overhead stirrer. The important thing is to wash the catalyst thoroughly with DI water so as not to have it alkaline, I think I used a large chromatorgaphic column for the washing at the end. Even more active Ra-Ni grade can be obtained by washing under hydrogen atmosphere but I remember that I skipped the hydrogen part and the catalyst I got was still pretty good. The alloy digestion procedure is in Orgsyn: Organic Syntheses, Coll. Vol. 3, p.176 (1955); Vol. 29, p.24 (1949). Watch out for fire, Ra-Ni should never go dry because it is really pyrophoric – more so than Pt-C).

      I should mention that I have seen a fairly large-scale hydrogenation of a complex benzonitrile substrate to benzylamine done with Pd-C in straight TFA as a solvent at RT at 50 psi on a Parr shaker – and it was surprisingly clean. (One would think that in TFA with Pd the benzylamine product would suffer hydrogenolysis on benzyl-N bond but that was not the case, and TFA performed actually far better than other solvents).

      Comment by milkshake — May 21, 2011 @ 7:16 pm

  11. Thanks! I’ll give the 2800 a try first, then if that fails, try the 4200 (since I have an unopened bottle), then go the fresh-made route (since I don’t have any alloy on hand).

    Comment by partial agonist — May 22, 2011 @ 10:04 am

    • I think Raney cobalt might be a good choice too. (There is also an excellent CoCl2+NaBH4 system that generates H2 and Co-boride in situ and gives very high yields of primary amines under mild transfer hydrogenation conditions. But I suspect that the primary amine is not what you are after)

      Comment by milkshake — May 23, 2011 @ 4:41 am

  12. The Cobalt borohydride method was the one I was thinking of when I mentioned that there’s better ways to do this— Not using a pressure bomb would certainly be an easier way to go.

    Comment by partial agonist — May 23, 2011 @ 7:50 am

  13. Hi Milkshake,

    can you tell me if gramine functionality is a no-no in a drug, perhaps due to its alkylating powers? I’m involved in a medchem project now which uses indole scaffold, and I just love the Mannich. Already got a crystal structure of our protein target with a gramine derivative. I’d hate to present it one day to a more medchem-savvy audience and hear “what the hell are you thinking kid?” I have a sort of an exit strategy of reducing it do indoline, or switching to indazolone scaffold that can be O-alkylated…

    Comment by LiqC — May 23, 2011 @ 5:15 pm

    • I don’t think gramin is a strong electrophile (sure it can eliminate to azafulven species and alkylate strong nucleophiles if you heat it long enough but this is not something that you would expect to happen in vivo). I would rather worry more about the metabolism: indoles with elecron donating substituents get oxidized fast. And the aminomethyl will probably be also a weak metabolic spot so you may have a PK problem like a very high clearance in vivo. I have seen this kind of problem with aminomethyl-substituted pyrroles made by Mannich.

      In your case I would happily churn out few dozens more Mannich derivatives if the chemistry is clean and pleasing. I would worry about PK later – once you get a decent understanding of SAR in your series and perhaps few more crystal structures. People make snotty comments on someone-else series, it feels great to put down a more junior colleague during his presentation.

      One way to fix the metabolic liabilities of gramin-like compounds is to add some electron withdrawing groups on the indole and also on the amine part. Maybe once you get good potent compounds you can also move away from indole later on, replace it with indazole, benzofurane, 7-azaindole, naphtalene etc or perhaps switch to beta-aminoethyl analogs

      Comment by milkshake — May 23, 2011 @ 5:47 pm

      • If I take up a gramine made from indole, Boc-piperazine, and formaldehyde in TFA for 15 min, and then try to acylate the piperazine, I get an ugly mixture of target compound, diacylpiperazine, and high-MW stuff with two or more indoles, meaning that retro-Mannich happens in acidic conditions, so it might happen in your stomach as well. I know it doesn’t happen on the second step because saponification of an ester happens smoothly (aq NaOH, MeOH, 50°C, o/n).

        Luckily, there is an EWG on the benzene ring of the indole. Don’t know if it’s enough, but this is something that my successors will have to figure out.
        Thanks for the proline tip.

        Comment by LiqC — May 25, 2011 @ 3:36 am

        • yeah, I think dissolving your compounds in straight TFA does not look like a great idea – simple indoles and pyrroles can crap up in TFA in couple different ways. (I think you are probably getting azafulvene oligomers in your case). I would use mono-Alloc piperazine and deprotect the product by dissolving it in 25% Et2NH in THF and adding few mol% of Pd(PPh3)4 under Ar – this deprotection method is very fast (15 min, RT) and the sideproduct (allyl-NEt2) is volatile.

          Mono-Alloc-piperazine is quite likely commercial – but if not you can prepare mono-alloc piperazine by dissolving a large excess (10 equivs) of anhydrous piperazine in boiling benzene, then adding a diluted Alloc-Cl solution gradually by a syringe through the condenser. After complete addition you let the sludge cool down, the piperazine excess and piperazine .HCl will crash out, so you remove it by filtration, concentrate the filtrates and distill the residue on highvac. The reason for doing this monoprotection in benzene is that piperazine is not soluble in benzene at room temp but dissolves nicely at reflux, so you can get rid of the piperazine excess easily

          At this point you need not worry about what happens in stomach. (Besides, HCl in your belly is pretty diluted, only about pH=1)

          Comment by milkshake — May 25, 2011 @ 3:47 am

    • if you are making gramin derivatives you can maybe try cyclic imines for Mannich – you can make them either by mild oxidation of cyclic amines or – better- by DIBAL reduction of lactams to hemiaminals.
      You should also check out azomethine ylides derived from benzaldehydes and proline:

      Comment by milkshake — May 23, 2011 @ 6:21 pm

  14. #10: Just to add to Milkshake’s comprehensive post on Raney nickel, I’ve found that no commercial RaNi ever beats freshly-prepared catalyst. The high pressure isn’t necessary. I do a lot of benzonitrile reductions in a Parr shaker at 2-3 bar. The best solvent is ethanol saturated with ammonia, which seems to suppress the side reactions mentioned by Milkshake. A couple of tips: the dissolution of the AlNi alloy in aqueous NaOH is VERY exothermic, so add it in small batches to avoid a foam-over. Don’t stint on the catalyst as it isn’t expensive and the reduction is much faster with plenty of it.

    Comment by Nick K — May 27, 2011 @ 3:23 pm

  15. Great post. I know you have your own ideas for what you want to do with this website, but a lot of people benefit from your knowledge. That comment you left on new reactions, for instance, about getting rid of Ph3P=O . A lot of people have trouble with Ph3P=O but don’t know about that trick. Any possibility of taking more snippets like that and turning them into blog posts?

    Comment by James — May 31, 2011 @ 11:46 am

    • I don’t know where is this personal chemistry web page going. I am still looking for a job, and there is a good chance that the job I get will be with a smaller custom synthesis lab/company. If that happens, I may not be able to post synthetic procedures in the future, even procedures for intermediates. For a contract lab that is trying to keep its customers, against the race-to-the bottom onslaught from Shanghai and Bangalore, a good procedure and a good purification method is what makes all the difference. Thank you for the support but again, I don’t know how much I will be able to write in the future even when I am back working in a lab.

      Comment by milkshake — May 31, 2011 @ 9:12 pm

      • Milkshake,

        I’ve followed your blog since 2006. It’s helped keep me sharp on chemistry I wasn’t always exposed to – and your posts on interviewing tips were really helpful to me. I’m fortunate to have a job in pharma – for now. I sure hope you can find a position you enjoy – you’re a really good chemist. In any shape or form that this blog takes in the future, thank you for having it in the past. It’s always been an enjoyable ready.

        Hoping you are employed soon.

        Comment by Tom — June 1, 2011 @ 7:38 pm

  16. Hexafluoroisopropanol (HFIP) is TFE on steroids. It has higher ionizing power, lower nucleophilicity, and is more acidic (pKa 9.3) yet still much milder than neat TFA. It is also supposed to be a great solvent for peptides. It has some odd but possibly useful behavior on workup (miscible with water, miscible with ether, immiscible with pentane, immiscible with some slightly salty aqueous phases, and capable of forming a fluorous phase.)

    Comment by Matt B. — June 2, 2011 @ 7:52 pm

    • If I remember correctly HIPF is not fully miscible with water, it forms a bottom phase. And few years back it was pretty expensive, to be used neat as a bulk solvent – I don’t know what the price is now. I used HIPF as a co-catalyst with DABCO for Baylis-Hillman, it had a rather dramatic effect on this reaction which otherwise can be quite slow, and the exact ratio of used DABCO-HIPF played a role.

      By the way, you don’t need TFA to cleave Boc groups and amide N-trityls, neat formic acid can do the job as well. I think peptide people liked HIPF as an additive because it is supposed to break peptide strong aggregates that can mask N-terminus during the synthesis of long peptides on solid phase

      Comment by milkshake — June 2, 2011 @ 8:13 pm

      • It is still kind of pricy, but as with all things fluorinated the price is more affordable from Oakwood/Fluorochem (99% purity is $165/kg vs $95 for TFE). The ionizing power drops slowly with dilution (supposedly it clusters around polar solutes) so you can often blend it with DCM to stretch the supply or improve solubility. I used it for mechanism studies, so the cost was tolerable since there really was no alternative. It is miscible with water, but it readily salts out at fairly low salt concentrations.

        Comment by Matt B. — June 5, 2011 @ 12:10 am

        • Ah, that makes sense. Oakwood is a really good company

          Comment by milkshake — June 5, 2011 @ 1:19 am

  17. Agree with 3 LiqC about water … useful when doing reductions with sodium borohydride.

    Someone once suggested doing a Friedel Krafts in toluence, everyone laughed but his SM was an anisole, the difference in reactivity was such that it ran like a charm.

    Comment by gyges — June 6, 2011 @ 11:08 am

  18. Sigh ‘toluence’ = toluene

    Comment by gyges — June 6, 2011 @ 11:08 am

  19. I love benzene. It’s one of the most useful solvents I’ve got lying around. The idiots keep trying to have it banned and removed, but those of us who do real chemistry desperately need it at times. Much more useful than toluene.

    Comment by Paul — June 6, 2011 @ 9:28 pm

  20. *sigh* I wish you could give me a list of underused solvents for casting thin films. (Our standard set is chloroform, PhCl, ODCB, toluene…sometimes THF…)
    You make me miss making molecules. Best of luck finding a job!

    Comment by psi*psi — June 7, 2011 @ 1:07 am

    • fairly cheap and useful halogenated solvents with a reasonable boiling point are dichloroethane, trichloroethylene, tetrachloroethylene and trifluoromethylbenzene.

      But seriously, if you are considering a high-boiling unhealthy stuff like o-DCB, you should also try natural (R)-limonene – It is inexpensive, it is a great solvent comparable to xylene, it is optically active, and your lab will smell like oranges.

      Comment by milkshake — June 7, 2011 @ 5:26 am

    • Psi – if you’re talking about spin coating, try o-dimethoxybenzene. Worked great for me, back in the day.

      Comment by James — June 7, 2011 @ 7:18 am

      • isn’t veratrole (o-dimethoxybenzene) supposed to be a solid at room temperature, with b.p. above 200C? Are you sure you are not talking about 1,2-dimethoxyethane?

        Comment by milkshake — June 7, 2011 @ 8:46 am

        • Crap. thanks for the correction Milkshake. I meant p-dimethoxybenzene.

          Comment by James — June 7, 2011 @ 8:52 am

          • … p-dimethoxybenzene has a melting point 60 C, boiling point 212 C …

            Comment by milkshake — June 7, 2011 @ 8:59 am

          • It was definitely a dimethoxybenzene, not diethoxy. Had a minty smell. Must have been the meta then. Not that I have a lot of credibility left on this particular subject!

            Comment by James — June 7, 2011 @ 11:23 am

          • veratrole’s definitely liquid in our lab! we use it as an internal standard for NMR yields.

            Comment by artemis — June 8, 2011 @ 1:36 pm

        • Melting point 22-23 °C

          turn the heating up when you do the work

          Comment by anon — June 9, 2011 @ 12:19 am

  21. Hi Milkshake- please send me an email at gmail dot com… not a joke. Have a serious potential position to tell you about.

    Comment by Daniel Plainview — June 7, 2011 @ 1:42 am

  22. Milkshake,

    I’ll agree with everyone else and say how nice it is to see a post from you. You’re expertise has helped me look good many times over the past 5 years. Hope you find a good position soon (a Pipeline mention is gotta help!). If you can’t, the rest of us are really in trouble!

    Best regards

    -yeah, I’ve been writing a lot of cover letters too

    Comment by Z — June 7, 2011 @ 5:39 pm

  23. Hi Milkshake,
    i am trying to react substituted 4-hydroxyquinoline with difluoro p-nitro benzene (Cs2CO3, DMF-CH3CN 1:1, RT).
    The product formation is quite fast and reaction is clean, the main problem is the formation of the N-Arylation product (Pirone tautomer) ca 1:1 with the desired one; is there any trick to avoid N arylation instead of OH.
    Thank you so much.

    Comment by madforit — June 8, 2011 @ 3:42 am

    • I’m not Mikeshake, but I have seen cases where related N vs. C alkyations are tipped toward O-alkylation in acetone/K2CO3 as opposed to the same reaction in DMF. The example I am thnking of was alkylation instead of arylation though.

      Comment by partial agonist — June 8, 2011 @ 4:01 pm

  24. Can you access the 4-chloroquinoline and react it with a nitrophenol?

    The only trick I know, how to alkylate O preferentialy over N in hydroxypyridines, is to use very hard electrophiles, for example oxonium salts, alkyl triflates, and also Mitsunobu reaction with alcohols. But this will not help you since there is no obvious way of modifying the arylation in this way (by replacing fluoronitrobenzene with a harder electrophile). So I don’t know

    Comment by milkshake — June 9, 2011 @ 1:20 am

  25. Thanks both of you guys,
    i’ll try Acetone-K2CO3 for sure. I am starting from Chloroquinoline and hydrolizing the Chloro to Hydroxy because of the poor reactivity as an electrophile. The other way should be Buchwald arylation between phenol and chloroquinoline, but i have to find out the right cat-ligand couple.
    Thank you again.

    Comment by madForIt — June 9, 2011 @ 5:56 am

    • For Buchwald arylation I would try peroxide-free THF with cesium carbonate solid and Pd2(dba)3 with Xanphos (1:1 Pd-diphosphine ligand) at reflux. If there is no reaction I would heat it in a pressure flask to 100C. The other catalyst to try would be simple Pd(PPh3)4. These two catalyst worked reasonably well with 4-chloropyridines for me. (Buchwald type of hindered biaryl phosphines may not be suitable for strongly coordinating substrates such as these)

      Comment by milkshake — June 9, 2011 @ 6:14 am

  26. Dear milkshake,
    I’m running a Suzuki coupling on a chloro-iodo-purine-like substrate and I’m only recovering protodehalogenation product and protodehalogenation + suzuki at the chloro position product. Any tips on how to get this iodine to react and avoid this problem? I am avoiding alcohols as solvents but there’s 2M aq carbonate in the mixture, plus rx is always properly degassed.

    Comment by tuky tuky — June 11, 2011 @ 8:42 am

  27. You should look up publications from group of Dr. Michal Hocek in Prague, they have been doing lots of medicinal chemistry on purines there at the Institute of organic chemistry and biochemistry, and I remember that they were publishing on purine Suzuki couplings. I think he also co-wrote some review on purine chemistry few years back. Maybe you should send him an e-mail – I bet he would be delighted that you contacted him.

    From my limited experience with somewhat similar heteroaromatic substrates, I think that you do have a problem with competing protolysis of ArPdX (the produced Pd(II) probably gets recycled back to Pd(0) by oxidative homocoupling of 2 equivs of boronic acid).
    You should test few different diphosphines – sometimes it helps to modify the catalyst, I have seen examples where the unwanted reductive dehalogenation in Suzuki was limited by a careful ligand choice. One good catalyst to try is PdCl2(dppf). But I would also try the anhydrous conditions with Cs2CO3 powder in THF or dioxane, with simple Pd(PPh3)4, and with added activated 4A molecular sieves. I have actually done anhydrous Suzuki once before (on a pyrone substrate that was falling apart in water) and it worked well but it was somewhat slower than usual aqueous Suzuki so it needed higher reaction temperature). The other thing you can do is to trasform your boronic acid by KHF2 to K[ArBF3] salt, that ones are supposedly very nice for anhydrous Suzuki. Another possibility is to use Stille coupluing with ArSnBu3.

    A procedure for anhydrous Suzuki is here:
    By the way, in this original procedure I would replace the microwave heating with conventional heating on oil bath, in a pressure vial – dioxane heats up with great difficulty in microwave (it absorbs microwaves poorly because because it has no dipole).

    Comment by milkshake — June 12, 2011 @ 1:15 am

    • cheers mate! I think I will tosylate the pyrrole N first, that seems to help too. Then play with anhydrous conditions…

      Comment by tuky tuky — June 13, 2011 @ 5:57 pm

  28. About 1-methoxy-2-propanol being useful for SNAr of halogenated heteroaromatics with amines – is there a literature refference for this?

    Comment by Baltic — June 12, 2011 @ 8:30 am

    • Electrophiles like chloropyridines, and chloropyrimidines do get activated by protonation, sometimes it actually helps to add a drop of methanesulfonic acid to the reaction – this works even in the presence of amine excess. I don’t have any reference but you can take my word for it – I did it quite a few times. You can use 2-propanol too but in many cases the reaction temperature is above bp of 2-propanol so you would need a pressure flask. So I was looking for some higher-boiling secondary alcohol and I ended up using 1-methoxy-2-propanol because we had a bottle of it (its cheap), the bp is reasonable so you could pull it off on rotavap. Other higher alcohols are typically not miscible with water so you can’t precipitate the product by drowning the reaction mix in water, and primary alcohols are too nucleophilic, they have a solvolytic liability with halopyridines.

      Comment by milkshake — June 12, 2011 @ 7:30 pm

  29. I can’t say I know any appropriate employers, but as a postbac science student, I spent long nights reading your entertaining anecdotes in this and other blogs. You’d make a damned awesome teacher: your descriptions of chemical reactions leap to life, rather, in more ways than one. Best wishes to you.

    Comment by Northwest AJ — June 21, 2011 @ 3:38 am

  30. We use MTBE as a common substite for diethylether for in extractions an recrystallisations. Most times its even superior in regards to solubilty, recrystallisation (higher bioling point/temp range) etc.
    Although toxic, chloroform and MeOH (most times mixtures have synergetic effects) are are sometime solvents of chioce for really polar stuff like some alkaloids, amino acid derivates and so on. You can even filtrate an crude amino acid over plain silica by eluating with Chloroform/MeoH/aqueous ammonia 80:35:4 (or similar). I´ve done that in a enzymatic kinetic resolution to remove enzym junk (enzym and substrate was required at a 1:1 or more mass ratio) before doing RP-18-chromatography. DCM doesnt substitute nice in this case.

    Comment by Matthias Goehl (@enzy84) — August 12, 2011 @ 3:55 pm

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Blog at

%d bloggers like this: