Skip to main content

Life in the Drug Labs

Solvents and More Solvents

I’ve been doing a solvent screen on a particular reaction recently, and it has me thinking about the number of times I’ve done that sort of thing before. Synthetic organic chemists spend most of their time using relatively few reaction solvents – there’s a lot of dichloromethane, a lot of tetrahydrofuran (and some diethyl ether), and some polar stuff like acetonitrile, DMF and DMSO. Methanol and ethanol get plenty of use, too. Once in a while you’ll run something in acetone or ethyl acetate (those are too reactive for a lot of chemistries), or go the other polarity direction and use toluene. (You’ll see that one more often if you’re doing process chemistry, since it’s easier to deal with on scale).
Then there are the solvents I think of as cousins to the main ones. The lesser-used ethers, like dimethoxyethane (DME) and dioxane are an example. They’re in between THF and diethyl ether, and if a reaction calls for one of them, it’s probably a good sign that someone, for better or worse, has been chopping the conditions pretty finely. (For a while, you saw a fair amount of methyl t-butyl ether, MTBE, around, because it was a popular gasoline additive at one point and became available cheaply in bulk). Benzene/toluene/xylene, dimethylformamide/dimethylacetamide, acetone/methyl ethyl ketone and dichloromethane/dichloroethane are all “cousin” sets in my mind. They’re pretty similar – a bit more polar, a bit less, a bit lower or higher-boiling. Diglyme, to pick another example, seems like an almost exact mimic of DME, just a lot more syrupy and able to be taken up to higher temperatures.
Methanol and ethanol, already mentioned, are in this relation to each other. As you go up, though, isopropanol is not so useful, because it doesn’t dissolve as many things, and n-butanol doesn’t either, and reeks as well. Its only distinction is that it lets you go up to higher temperatures. (I’ve never used n-propanol; it’s just not as commonly available). On the chlorinated spectrum, chloroform is a different beast than dichloroethane, notably more polar, but when you go up to carbon tetrachloride, the meter swings back around in the other direction.
On the really, really polar end of things, as mentioned, DMF and DMSO get most of the action. But there’s NMP, which I’ve tried a few times, and DMPU (but that one mostly as an additive). Those last two are standing in for HMPA, which is about as polar as a solvent gets, but is also notably toxic. You’ll see it all the time in the 1970s synthetic literature, but then its reputation caught up with it. If those other solvents won’t cut it, but HMPA will, you are really getting near the end of your synthetic tether. Back in grad school, I tried a reaction or two in neat HMPA, but I wasn’t happy about it.
The amines are more often used as reagents than as solvents, but you certainly do run reactions in pyridine once in a while – a long while, with any luck. I don’t think I’ve ever run a reaction in straight triethylamine or the like, though, come to think of it. That would indeed be a fishy way to go. On the other end of the pH scale, I’ve done a few reactions in straight glacial acetic acid, of course, which will certainly clear your sinuses. And as for stinky solvents, I can say that I did have a reaction in grad school that ran in neat ethanethiol. Never again. That’s a sign of desperation, and I was nothing if not desperate at that point. (It did work, unfortunately, so I had to run it more than once).
There are also a set of solvents that get broken out for high temperature reactions. Decalin is the nonpolar member of that tribe, and collidine is pyridine’s representative up in this territory. I’ve used benzonitrile and nitrobenzene for this purpose when I needed some polarity with more stability than DMF/DMA at high temperature, and less reactive potential than DMSO. Sulfolane is, in my mind, sort of the end of the line in this regard. I think I’ve only used it once, and if you’re up to sulfolane, you’re really having to roast something. I’ve heard of people running reactions (like recalcitrant Claisen rearrangements) in diphenyl ether at high temperature, but I’ve never had the pleasure.
And finally, there are a few zebras, weirdo one-offs. I’ve tried reactions in trifluoroethanol, which is a funny solvent in several ways, but it’s never done any magic for me yet. I’ve used cyclohexane, but it (and di-isopropyl ether, which is basically impossible to store due to peroxide formation) are to me relics of the age when recrystallization was the main way to purify compounds. Nitromethane I think I’ve used only once or twice as a solvent (as opposed to a nitroaldol reagent in something else), and I’ve used N,N-dimethylaniline once or twice (and not so happily, because it’s a pain to get rid of). Probably my least favorite solvent was an Arbuzov reaction in straight trimethyl phosphite, because trimethyl phosphite is so relentlessly revolting.
Anyone have any others to add to the list? (Note – I’ve left everything aqueous off, because to an organic chemist, water is Over There and everything else is Over Here.)
Update: here’s an excellent collection of short descriptions of “Solvents of the Week”. Any working organic chemist will enjoy them!

96 comments on “Solvents and More Solvents”

  1. Processator says:

    Did I kiss acetonitrile? Very important solvent in my opinion. Few years ago there was a major shortage of it.

  2. Martin says:

    Carbon disulfide used to be more common in the past too. I’ve done a few small scale diphenylether reactions in the past. Usually distill off under vacuum on Kugelrohr

  3. Douglas Kell says:

    Actually solvent MIXTURES (widely used in TLC) can be great for reactions using biology. Salter GJ, Kell DB: Solvent selection for whole cell biotransformations in organic media. CRC Crit Rev Biotechnol 1995; 15:139-177.
    Harder to recycle, of course, but VERY effective. Under-studied area.

  4. Hap says:

    Trifluorotoluene was supposed to be a higher-boiling replacement for dichloromethane (though it depends what you’re doing as to whether that’s a bug or a feature). Fluorobenzene was supposed to be a replacement for benzene, but it’s probably too pricey for common use. Someone here had suggested 1-chlorobutane as a recrystallization solvent.

  5. Cato the Elder says:

    Doing reactions in nitrobenzene is the worst… but there are a couple of reactions where that really is the best solvent

  6. BB says:

    Depends if we count things that are both a solvent and reagent. I have used neat thionyl chloride as a solvent/reagent more than once, same with TFA.

  7. oldnuke says:

    Good ole TriButyl Phosphate (TBP) in dodecane for those old uranium, thorium and plutonium chemists. 🙂 Magical stuff.

  8. fluorogrol says:

    Ionic liquids, anyone?
    Only gone down that road once, and I was less than impressed to open the bottle of the particular ionic ‘liquid’ prescribed by the prep to find that it was very much an ionic solid.
    That it required significant heating to become, well, solvent may have contributed to the dark brown mess that I got out of the reaction for my troubles.

  9. anonymous says:

    Do inorganic solvents count? Have used POCl3 as solvent for the Meth-Cohn quinoline synthesis, and for converting a hydroxypyridine to the chloropyridine. Doing the latter on large scale once (200-300 mL?) the quench get out of control, spent all the next morning cleaning it off the ceiling of my hood.

  10. Anonymous says:

    Cyclohexane was the solvent of choice in my previous polymer plant. Seemed to be just the right polarity to keep polystyrene, polybutadiene, and polyisoprene (and their various combos) in solution. When they wanted polarity, they would add diethyl ether.

  11. Slurpy says:

    I’ve used formic acid as a solvent, that was new to me.

  12. Falanx says:

    #8, Surely the prep made it clear that just because it’s used as a liquid, doesn’t mean it’s supplied as such? Most ionic liquids/deep eutectic solvents are solids, and become liquids when mixed and given a gentle thermal kick over the energetic barrier. They’re certainly the Next Big Thing in metal finishing.
    #3 solvent *mixtures* are the only way the paint coatings industry exists. PPG et al ship 100 million litres of mixed solvents every year for their paint curing reactions to happen in/after it’s flashed off.

  13. fluorogrol says:

    @12 Falanx
    No, it said nothing of the sort. OK, it didn’t describe it as a room-temperature ionic liquid, but still.
    Not a big problem, obviously, just added another level of annoyance to a last-chance-saloon set of conditions for a stubborn reaction.
    Oh, and in this case I wouldn’t describe the thermal kick involved as ‘gentle’.

  14. PG says:

    Propylene Carbonate, nicely polar with no acidic protons. I’ve done DNA synthesis in it instead of MeCN because I needed the high BP and viscosity to shoot droplets of nucleotide phosphramidites out of an ink-jet head. Electrochemists use it so it’s available in an anhydrous grade. Non-toxic to boot.

  15. Ryan says:

    Tetralin is a useful, high-boiling aromatic. Less unpleasant than di- or trichlorobenzene, higher boiling than xylenes or mesitylene.

  16. Didymium says:

    Hexamethyldisiloxane, aka bis-TMS ether. Less polar (seemingly) than hexane despite being an ether. Actually sort of a stealth fluorocarbon in my opinion.

  17. proceschemist says:

    In my job, most used solvennts are:
    Acetic acid
    Ethyl Acetate
    Methyl Acetate
    (and water, obviously)
    MeTHF is more expensive than THF but has some fancy features: you can dry a rection system by distilling his azeotrope with water – and, being asymmetric, can wash out some orientating effects in reactions with Grignard reagents and lithium alkyls.

  18. A Nonny Mouse says:

    Ethoxyethanol for me.
    Hexafluoro isopropanol great for making 4-substituted quinolines from anilines.
    Propan-1-ol is good when removing a benzyl group in peptides- minimal removal of Fmoc groups (which can be quite extensive with ethanol and methanol- just occurs to me that I have never tried ethoxyethanol!).

  19. Didymium says:

    Hexamethyldisiloxane, aka bis-TMS ether. Less polar (seemingly) than hexane despite being an ether. Actually sort of a stealth fluorocarbon in my opinion.

  20. JJ says:

    1,2-DCE is good if need higher BP and/or a little better solubility

  21. PierrikL says:

    When I was an undergraduate, I used diphenylether on a 1 to 2 L scale to prepare p-tertbutylcalix[4]arene. I had to bring the mixture to reflux as quickly as possible (to favor the calix[4]arene instead of the calix[8]arene). But I really didn’t like the smell! Moreover, it solidified readily on the top of the reactor.
    Also, diethylcarbonate can be a useful alternative to toluene, and was used in some palladium-catalyzed direct arylation reactions of 5-membered heterocycles.

  22. nitrosonium says:

    what about PPA as a solvent? yuck

  23. processchemist says:

    1,2-DCE is highly toxic, 1,2-dichloropropane is a safer alternative.

  24. Curt F. says:

    I’m not a polymer chemist but I think one of those folks would probably want to add hexafluoroisopropanol (HFIP) to the list. You could view it as a distant cousin to trifluoroethanol I suppose, but I think it dissolves a lot that TFE won’t.
    MTBE is still very popular among analytical biochemists, because unlike the classic 2:1 chloroform:methanol mix, it sits on top of the aqueous phase, simplifying pipetting. It’s safer than most other ethers and dissolves many different types of lipids amazingly well.
    n-Butanol is used in some fairly obscure methods for DNA extraction/concentration, because water, but not DNA, is appreciably but not completely soluble in it, so it soaks up excess water from your aqueous phase.
    Analytical biochemists have used a bunch of weird solvents for low temperature extraction. I’ve seen papers that use “Freon”, presumably R-11, as an extractant, usually in conjugation with an ion pairing agent like triethylamine. People have used dimethyl ether (another DME, but much lower boiling than dimethoxyethane) as an extractant too. I’ve never used DME or freon but I can understand the appeal: solvent removal is apparently as simple as “remove your sample from the ice bath and wait a few minutes”.
    The weirdest solvent system I’ve used personally was for a protocol where cells are centrifuged through a non-aqueous layer to cleanly remove them from the extracellular aqueous medium. The bottom of the centrifuge tube had a blend of methanol, water, some acid, and TFE to keep that phase denser than everything above it. Above that phase was 1-bromododecane, which is less dense than most cells but more dense than most culture media. You put your sample of culture on top of the bromododecane layer and centrifuge. Cells get pulled through the bromododecane layer and are instantly lysed when they reach the TFE-containing layer below. Culture medium, now cell-free, remains on top of the bromodecane. Made for tricky pipetting but I haven’t seen anything else that let you extract just the cells, not the medium, in as short a time.

  25. Falanx says:

    #13. Wow, that was helpful of them. Which IL was it you were suffering to use?

  26. Mark Thorson says:

    Have you ever done a blog posting about polymorphs? Solvents aren’t just used for reactions — they are also used to prepare polymorphs. For example, when Schricks was providing sapropterin, it was supplied in tablets as the acid salt with antioxidants, yet it still had to be kept deep frozen, had a short shelf life, and couldn’t be shipped in the summer. BioMarin now supplies it (at a much higher price, of course) and they claim it’s stable for six months at room temperature. They have a patent (U.S. Patent 8003126) on particular solvent combinations for generating a shelf-stable polymorph of sapropterin.
    I am amazed that just the choice of solvents can result in a product with such different properties. Apparently, it is unknown just how this form differs from the standard stuff, but they do have some powder x-ray diffraction method that distinguishes it. I think BioMarin has another patent that describes how they explored the polymorph space for sapropterin, preparing a large number of polymorphs by using a variety of solvents and solvent mixtures. It was actually one of the earliest ones in the series which they wound up using in production. I’d be interested in reading about how someone goes about selecting solvents for making a variety of ploymorphs — you don’t just try them at random, do you?

  27. MA says:

    In regards to sulfolane, it’s nice in heavily oxidative environments (like nitration/nitrolysis). Sometimes used in combination with nitronium tetrafluoroborate for certain (niche application) aromatic nitration reactions.

  28. Falanx says:

    #25 Oh, and butan-1-ol and propan-1-ol are regularly used in paint and depositable solid lubricants solvent systems, too

  29. oldnuke says:

    Another “hot” solvent these days seems to be supercritical carbon dioxide (for the green folks out there).

  30. David P says:

    MTBE still gets used quite a bit, it is less prone to peroxide formation so has a longer shelf life. The process group would use it all the time for that reason.
    Like #21, I had the pleasure of using diphenyl ether for a reaction but he neglected to mention just how hot you have to get it for a reflux (bp 259C)! That it was solid at RT was a major pain.
    At the other end of the scale, I am running a reaction in liquid ammonia today…

  31. Ted says:

    Was this in Kazoo? Are you my old labmate?
    I remember the hissing sound that a liter or so of mixed POCl3/H3PO4/HCl foam made as it flowed, in apparent defiance of gravity, in three parallel streams, straight up out of the flask and into the ceiling of the hood, spreading out across the ceiling and dripping back down like an amber, smoky rain…
    @ Derek: NMP becomes a lot more appealing of a solvent when you realize you can dilute it with a little water on workup and extract products with MTBE. Following that with a single saline back wash will usually give you an NMP-free solution.

  32. LittleGreenPills says:

    I have not used them as reaction solvents, but the mention of n-butanol and di-isopropyl ether bring back memories.
    I used n-butanol for extractions and found that it formed a nice azeotrope with water. It always seemed strange to watch it collect in the trap and separate into two layers.
    As for the di-isopropyl ether, as an undergrad I picked up a crusty bottle of it… It had a lot of grey solid in the bottom and thankfully was stoppered with a cork. I was asking my professor what it was, and it instantly started to fizz and hiss. It shot the cork off with a nice pop. Both of us were left standing there very shaken, and I was holding the bomb, I mean bottle by the loop at the top. All he said was I am glad you did not drop that and lets gently put that down.

  33. fluorogrol says:

    @25 Falanx
    Memory is fading/has been efficiently repressed, but I think it was [bmim][Br].

  34. Chrispy says:

    Is it possible to influence the chirality of a reaction through the use of a chiral solvent? It seems like it would work, but I’ve never seen it done.

  35. Processing says:

    As a process chemist making a rather polar API, I have come to quite like 1-propanol and NMP for my coupling reactions. And recently optimzed a borylation process that worked best in ethylene glycol. Go figure. Then there is the really weird one going on in my fumehood right now that uses hydrazine hydrate as solvent (at reflux no less).

  36. IUPAC police says:

    just want to make sure we all know there is no such thing as n-butanol. there is n-butylalcohol or 1-butanol thank you. that is all

  37. Simon Higgins says:

    An interesting alternative solvent for electrochemistry, published by Meyer’s group in the late 80’s, was 1,2-difluorobenzene. It has quite a high dielectric constant (14 or so) but is non-coordinating compared with things like acetonitrile. I wonder if anyone ever tried it as a solvent for synthesis? A bit pricy I should imagine.

  38. Simon Higgins says:

    An interesting alternative solvent for electrochemistry, published by Meyer’s group in the late 80’s, was 1,2-difluorobenzene. It has quite a high dielectric constant (14 or so) but is non-coordinating compared with things like acetonitrile. I wonder if anyone ever tried it as a solvent for synthesis? A bit pricy I should imagine.

  39. anon says:

    Orthodichlorobenzene had its niche uses.
    Xylenes were nice to perform oldschool alcoholate preparation (R-OH + Na at boiling point) if you didn’t want to go neat.
    Triglyme was a good way to have a high bp polar one, but it’s a horror to dry properly…
    Fluorous ones had their moment of glory but sadly the funky phase-switching is not as popular anymore

  40. Anon says:

    Molten/fluid polyphosphoric acid is a cool way in helicenes synthesis! Wonder if it has other uses though

  41. CTChemist says:

    @17, MeTHF is indeed a great solvent and suprised others did not mention this as well. It’s also not miscible with water, so makes work ups easier than if THF was used. Great solvating properties as well. I’m surprised it’s not become more popular with chemists outside process. I think I saw an Ellman paper, though, where he took advantage of the higher boiling MeTHF and CPME to prepare some more difficult-to-form sulfinylimines. Couple of useful papers on MeTHF: Synlett 2003 2353; OPRD 2007, 11, 156.

  42. Anonymous says:

    @26: I did some polymorph work in the past and found it interesting as well about solvent choice and temperature. Even though I’m an organic chemist, physical chemical properties are fascinating.

  43. another process guy says:

    “relics of the age when recrystallization was the main way to purify compounds.”
    Come on Derek . . . for a process chemist, crystallization is the BEST way to purify and isolate compounds.
    If a process chemist has to do chromatography, he has failed at his job!

  44. Anonymous says:

    I’ve had the, uh, pleasure of working with condensed trimethylamine at elevated pressure/temperature as the solvent for some otherwise intractable salts about a year ago. That Parr bomb still smells…

  45. Fluorine Chemist says:

    The weirdest ones I’ve came across were perfluorocarbons (perfluoro-methylcycloheyane and perfluoro-perhydro-phenantrene). Also, CFCs like CFCl3 and CF2ClCCl2F used to be popular solvents in fluorine chemistry 30 years ago but they have been replaced with acetonitrile.
    Anyone had any experience with new bio-based “green” solvents like Cyrene or gamma-valerolactone?

  46. Mark says:

    Liquid ammonia got me a long way no one does Birch reductions anymore I guess.

  47. milkshake says:

    molten neat pyridine hydrochloride at 140C – surprisingly nice and gentle for demethylation of aryl methyl ethers.
    Trifluoroethanol works really well for amine additions to epoxides and acrylates – the kind of reaction catalyzed by protonation, but where you don’t wish to use a stronger acid, because of the amine.
    I worked with 1-methoxy-2-propanol for aminolysis of chloropyrimidines: it is a high-boiling cousin of IPA, it is water soluble also (so you can drown the reaction at the end, to crash out the crude product), and it gives less solvolysis problem and has low toxicity compared to methoxyethanol.
    I was briefly an intern in a calixarene lab where they used instead of diphenyl ether its liquid 50:50 eutectic mix with biphenyl; I think the stuff was originally developed as a transformer oil fluid, and they were refluxing their reaction in it and their whole lab was soaked through with the vapors – just one hour spent there and you too smelled from it, like a cheap soap

  48. exGlaxoid says:

    #47, the diphenyl ether eutectic mix with biphenyl is known as Dowtherm A, I believe. I ran a bunch of reactions in it for one project, it boils about 300C, stinks (as noted above), and requires a lot of work to get rid of the solvent. If you are lucky, you could pour the reaction into hexane and get the product to crash out; if not, you often had to chromatograph the whole reaction via a silica gel plug and hope that your compound stuck well. But if you want to cyclize a quinoline or such with the removal of water, that seems to be the trick.
    I liked Freon 112 and 113 and 113a for a few flurochemistry reactions years ago. They were also great for putting out fires in the lab.

  49. Ted says:

    We found 2MeTHF often worked better than THF when initiating certain Grignards at scale.

  50. milkshake says:

    cyclohexane is quite lovely as antisolvent for recrystallizations, it has a better dissolving power than hexane and heptane. My favorite is benzene+cyclohexane mix, as it does not separate at reflux.

  51. simpl says:

    #26 good heads-up on polymorphs.
    – Hydrocarbon fractions (like heptane) as well as cyclohexane, also isomeric mixtures like xylene.
    – our chemical production use a lot of solvent mixtures, which are reused as distillate mixes
    – Acetonitrile and ethyl acetoacetate are perhaps two less mainstream ones
    Otherwise the old favorites

  52. The troll in the basement says:

    As part of my graduate work in Chem E, I am doing a bunch of work in high concentrations of acetic and formic acids with a kicker of HCl, just to make it yet more corrosive (because we weren’t having enough fun yet). I know it probably has too much water for this list, but it is an interesting one, especially if you throw MIBK or toluene on top of it and take the temperatures up to 180 C.

  53. Anonymous says:

    I looked around for the BioMarin patent that had a better description of their experiments in sapropterin (a.k.a. tetrahydrobiopterin or BH4) polymorphs. I thought I’d seen a U.S. patent, but I found this European patent which is substantially the same thing I remember reading.
    Dang, they made a lot of polymorphs. And it was polymorph B that was the good one.
    Interesting reading if you’re interested in that kind of thing. It might be useful as a model for planning a large polymorph research effort. (I almost said “thorough”, but there are so many possible combinations of solvents that you can never be thorough, only large.) I’d be skeptical you could solve a toxicity problem that way, but shelf stability and maybe a solubility problem might be solved with a new polymorph. Or a going-off-patent problem.

  54. Rock says:

    Found this recent paper in Org Process R & D on solvents used in the journal over a 15 year period to be very interesting. DMF, DMA, acetonitrile, NMP and DMSO made up 95% of reaction solvents from process chemists. Still find that hard to believe.

  55. anony-mous(e) says:

    I used to do “melts” with acetobromosugars and mercuric cyanide – not sure which was the “solvent” though… Also, acetylation with pyridine / acetic anhydride was the “go to” method in carbohydrate chemistry for decades…

  56. Triazine says:

    Don’t dismiss water so quickly! I’ve used aqueous cyclodextrins to do a couple regioselective reactions that would not have worked any other way.

  57. OldLabRat says:

    Neat m-anisidine as both solvent and reactant. Along with Dowtherm A, probably the most annoying non-aqueous solvent I’ve used. BTW, if the cyclized product in Dowtherm is MeOH soluble, extracting the Dowtherm with pentane works well. Just pray for no emulsion….

  58. Curt F. says:

    @58. m-anisidine sounds like a horrible solvent. I hope your reaction went quickly to 100% yield.

  59. Matt says:

    Can you get drunk on trifluoroethanol?

  60. LiqC says:

    My outliers:
    Dimethyl acetylenedicarboxylate – well, just a neat reaction. Same for furan + acryloyl chloride
    1-propanol: that one reaction ran quite a bit faster from pushing it into the 80s, or so I convinced myself
    Perfluorodiglyme – for extracting fluorinated products. Note 96C drop in boiling point relative to non-fluorinated diglyme.

  61. KerryF says:

    Hot concentrated sulfuric acid served nicely as catalyst and solvent for nitrile to amide transformations.

  62. Chemystery says:

    I used HMPA as an additive a couple of times as a grad student. It was supposed to increase cis-selectivity in a Peterson olefination. One time, the needle tip touched the cold THF and froze, so no HMPA added. The result – exactly the aame E/Z ratio as previously. Needless to say, that was the last time HMPA went in.
    I’ve also used hexafluoro-isopropanol – which can have some pretty wierd solvating and activating properties, and diisopropyl ether, which is, well… a bit like diethyl ether, MTBE etc. And on the non-polar end, cyclopentane

  63. anonymous process gal says:

    @54 Rock, I think the pie chart in the link has 95% of dipolar aprotic solvents not because they were the solvents exclusively used to run reactions in the OPRD papers, but because a large category of typical reaction solvents was simply excluded from analysis. They say they only looked at dipolar aprotic solvents, solvents with issues (be it toxicity or other safety issues) and “novelty” solvents. So if you decide not to include alcohols, acetone, toluene, iPrOAc etc etc in the analysis, then yes, the pie chart in the link makes sense.

  64. Cymantrene says:

    I use methylcyclohexane, n-heptane, heptane-isomer mixture, even 80-110 petrolether mixture mainly as antisolvents, p.e. adding to MTBE-solution and then distilling off MTBE (if needed)
    n-pentane for extractions, clarification and filtration – has low density, polarity and viscosity, separates well, leaves any colored/polar stuff on silica and flows through 0,5 micron filter even without overpressure.
    iso-octane (2,2,4-trimethylpentane, the very one used for octane-number determination) for a PCC+tBuOOH oxidation mixed up with an aqueous bottom phase. I did not see the point, but it was customers procedure and they adhered to it, not allowed further research (Flavor&Fragrance industry has strange habits)
    In the same production, n-dodecane was used to heat up product to 150-160 °C to decompose peroxydes (hydroperoxydes can be removed by aq. Na2SO3, peroxydes are not that easy)
    CPME for no special reason in an extraction, simply it was given in the customers procedure. It has even lower explosion limit than other ethers, 1,1 v/v% in air. And, the bastard, really gave a puff when the aqueous phase was discharged from below it, which still contained enough CPME to reach that level in the airspace of a 200 L drum and (despite grounding) was charged electrostatically enough to produce a spark. No one got hurt, luckily, but it was quite alarming.

  65. MTK says:

    The best example of a non-common solvent that I ever used was dimethyl carbonate in a radical benzylic bromination.
    Did it on scale too. I can’t remember how much solvent was used but the reaction resulted in about 40kg of the desired benzyl bromide.
    Worked like a champ.

  66. Hap says:

    @61: I did one of those in grad school (p-cyanobenzoic acid, I think, to the terephthalic acid monoamide) in conc. sulfuric acid, and it worked really well, which was unusual.

  67. AlphaGamma says:

    Does trifluoroacetic acid count when used for cleavage in solid-phase peptide synthesis?

  68. Dennis says:

    @67 If TFA doesn’t count for peptide cleavage, it should count when it’s used as a solvent for removal of Cbz groups with H2 and Pd/C?

  69. a. nonymaus says:

    When your life lacks enough excitement, there’s always water’s evil twin – anhydrous HF. High-boiling enough to do reactions in, oxidatively robust in the extreme, what’s not to like?

  70. Doug Steinman says:

    I have used a solvent mix of toluene, ethanol and water successfully in a wide range of Suzuki reactions. Weird combination but works quite well in a lot of cases.

  71. RD says:

    I’ve used neat TEA for a Sonogashira coupling, but like #6 I’m not sure it counts since it’s the base for the reaction as well. It did work much better than adding a “real” solvent though.
    Also I don’t think I saw any mentions of t-butanol. Of course in most labs you have to melt it first, but it’s good for fluorinations with CsF. I’ve also used it for Miyaura borylations and Suzuki couplings. Not too shabby.

  72. Secondaire says:

    Also used neat TEA for the Sonogashira. It works really well for most things, or cut 1:1 with benzene or THF.
    Other things I’ve used (reluctantly) but was never a fan of: cyclohexanol; chlorobenzene, 2-butanone, acetamide (mp: 80 ˚C).

  73. Steve says:

    My two bits on solvent used in last ten years
    Ethylamine is a much more convenient solvent for Birch reductions than liquid ammonia as you can do it on ice.
    Nitromethane: good solvent for Friedel Crafts type reactions as easy to remove
    Methanesulphonic acid: for phosphorous pentoxide dehydrations
    Carbon Tetrachloride – still like it for NBS reactions as you know when the reaction is finished when the succinimide floats to the top.
    Of course the best solvent is no solvent at all!

  74. BAM says:

    I used TFAA as a solvent for the nitration of pyridine. 1,2,4-trichlorobenzene is horrible.
    2-Hydroxyethylammonium formate is cool ionic liquid.
    Heptane for the in situ preparation of 9-I-9-BBN

  75. Baltic says:

    Condensation of phthalic acid derivatives with acetoacetic ester to yield substituted indane-1,3-dione derivatives is typically run in acetic anhydride as solvent. Some DMSO may be added when the starting material is particularly poorly soluble in AA.

  76. AqueousPhase says:

    Just out of interest what are peoples opinion on the use of water as a solvent for organic reactions?

  77. Anonymous says:

    #48 Oh dear, Dowtherm. Great heat transfer agent for when you don’t want to deal with high pressure steam, but boy do you sure know it right away if you have a coil leak. Double bonus points if you happen to get it on you, soap doesn’t even touch the stuff.

  78. Dave says:

    #69: Anhydrous HF – Oh, Hell No!
    Don’t forget that even relatively common solvents, such as Hexane, can add a lot of excitement when they escape, especially in scale.

  79. Sofia says:

    Reactions in NMR solvents can be interesting – if a lil expense to run. Key advantage is that you can check the whole reaction at once via NMR rather than doing work up.
    Had undergrad working with me when one summer when we were doing DMSO-d6 rxns in microscale and found one condition to work. I told them to scale it up, and came back to them having emptied three bottles of DMSO-d6 into a reaction vessel…

  80. Poison Ivy League says:

    Many Davies C-H insertions use a mixture of trifluorotoluene (TFT) and 2,2-dimethylbutane (dmb). These reactions work surprisingly well, however it is non-ideal when the entire world’s supply of dmb is gone (currently back-ordered for months from all suplliers)…

  81. Hap says:

    Can you hydrogenate tert-butylethylene? Alfa has that for $70/250mL (assuming they have it though…)

  82. NC says:

    Have performed the horror of neat triethylamine. Unpleasant. Also, acetic acid as a youngun; spilt a drop on my leg and discovered it feels nothing like vinegar. Rather painful actually…
    Other than the ones you mention I’ve used 1,2-DCE, chlorobenzene, cymene (basically high-temp, spicy-smelling toluene), n-butanol and t-butanol (in mixtures; tBuOH needs slight warming or cosolvating to become liquid in the cold Oxford lab). I used to have a labmate who made helicenes at ludicrously high temperatures; he often used halogenated benzenes. For his really high temperatures I think he went for 1,2,4-trichlorobenzene or something like that.

  83. Yancey Ward says:

    I used to do reactions in molten pyrazole, though it was also a reagent (was making pyrazolyl borates and the carbon analogues).

  84. sp3 says:

    Not sure if anisole was mentioned. Good solvent for cross couplings and has a good azeotrope with water so easy to dry.
    By the way, for the snarky IUPAC police #36, there is no such solvent n-butylalcohol. Try n-butyl alcohol.

  85. Mark Thorson says:

    With regard to chiral solvents, I’ve never heard of that being useful. I would guess that if it were even a little bit useful for reaction chemistry, it would have an even bigger effect as an HPLC mobile phase. Chiral columns are useful for chiral separations in HPLC, but I’ve never heard of a chiral solvent being used for a chiral separation. HPLC is a well-plowed field, and if there were any value to a chiral mobile phase, I think that would be well known by now.

  86. Tocrat says:

    anhydrous hydrazine anyone?

  87. newnickname says:

    I’ve read up to #86 (Mark / chiral solvents) and I think I can add a few more.
    Some, above, did mention perflourinated solvents but I’ll mention some name brand families: the Fomblins, the Geldans and the Krytoxes (perfluoropolyethers). Were any of those characters on The Simpsons? Not cheap, but industrially and easily available. (Lubricants, cleaning fluids, etc.)
    As a grad student, I was freeze-thaw degassing under Ar (LN2 bath) when I realized that I was condensing Ar in my reaction flask (a tin hydride reduction in toluene). On the last cycle, I just replaced the cold bath with a hot bath and let it evaporate off so I will claim that a tiny part of my reaction was done using Ar(liq) as co-solvent.
    Bergman used liquid xenon for some reactions. Easier to condense than Ar but also more expensive. Be prepared to capture and reuse the gas.
    How about scCO2, scMeOH and other accessible supercritical phases? I worked on a project using scMeOH but I was helping the students do it so I only claim borderline doing it myself. Likewise, scCO2 as both extraction and reaction solvent. Although it’s “water” and Derek wanted to preclude water, supercritical and subcritical water can do some non-watery things.
    I’ve always wanted to try things like polysiloxanes as solvents. I’ve used them as co-solvents and to prevent foaming. “D5”, decamethylcyclopentasiloxane, is a dry cleaning solvent (so you know you can actually buy it and not have to roll your own) and it breaks down to SiO2, CO2, formic acid, etc. at high temperature. (There are other D#s available.) If your product is heat stable, that could be an easy workup.
    The good old “Chemists Companion” book has a section on Molten Salts that predates the current craze for Ionic Liquids. The old ionic liquids – molten salts were actually low-melting inorganic salts and their mixtures. If I had my CC handy, I’d look some up. Organic reactions in such molten salts goes back 80+ years. (I’m thinking of refs from the 1920s but I can’t look things up right now.)
    Anyone ever done constructive organic chem in molten metals (like Wood’s metal, Rose’s metal, etc.)? I guess that isolation could be tricky.
    I’ve done flash vacuum pyrolysis so can I claim “low pressure gas” as one phase of my solution? Technically, I guess my reactions were unimolecular and pretty far removed from the nearest neighbor carrier solvent (hexane, e.g.) molecules.
    There are some reactions that occur in their own “melt”. I.e., you have a stable solid, melt it (as in a capillary or on a microscope slide on a hot stage) and, voila, new product! One that I’m thinking of didn’t do the same thing when dissolved in a hot solvent. Does “liquified self” count as a solvent?
    (Please see Pterin entry next)

  88. newnickname says:

    Pterin, sapropterin: Everyone here knows how to pronounce those.
    I was watching PBS “Cancer: Emeperor of All Maladies” on-line. In Episode 1, Magic Bullets, at around 12’15” – 13’00”, the narrator says, “Though doctors elsewhere had experimented with chemical cures, Farber was the first to try aminopterin, a rare and possibly dangerous compound. …”
    ah-min-OPT-uh-rin. Ugh. I’ve done pterin chemistry (making folate analogs, a la, E C Taylor) and I think everyone here would pronounce it “amino” (like amino acid) “terrin” (like terror but terrin): ah-meen-o-terr-in. Unfortunately, every on-line talking dictionary says “ah-min-OPT-uh-rin” so I don’t really know what’s going on. What IS the correct (acceptable) pronunciation?
    (The story of Yellapragada Subbarow and the pterins is yet another great story.)
    I keep hearing other chemicals mispronounced on radio, TV and the internet, including drug ads. I’m not remembering a single one as I type this.
    Does this bother anyone else out there? Other examples of botched pronunciations? Derek has had topics on botched STRUCTURES in advertisements and other visual media. I’m asking about AUDIO.
    Thank you.

  89. Nick K says:

    Strangely, very little mention here of liquid ammonia. There is an odd Org Syn prep of a cysteine dipeptide in which a CBz protecting group is removed by hydrogenolysis with Pd/C in liquid ammonia. Sodamide or potassium amide in this solvent used to be quite popular for enolate chemistry in the 50’s and 60’s.

  90. Mark Thorson says:

    I remember noticing that mispronounciation of aminopterin too. I notice people who mispronounce “giga”, but I try not to let it bother me. I have to admit being guilty of mispronouncing “piezo”, but that doesn’t come up very often.

  91. Joe Q. says:

    In the past, I had good results using neat morpholine as a solvent for the Sonogashira coupling. Also neat trifluoroacetic acid as a solvent for the “hydrolysis” of tert-butyl esters.
    Now in a process-like role, we avoid chlorinated solvents, and end up using toluene wherever humanly possible. DMF, ethyl acetate, isopropanol and MEK where polar solvents are needed. We avoid aliphatic hydrocarbons due to the potential for static charge build-up and detonation.

  92. Mark Thorson says:

    Just one of the obscure bits of knowledge some of us enjoy.
    The guy who ran the electron microscope lab at a former employer mentioned that the plastic he used for embedding specimens had a solvent that smelled like bananas. I knew immediately what solvent that was.
    I warned him that this solvent is also the honeybee alarm pheromone. People have been attacked by bees for wearing perfume that contained isoamylacetate.

  93. sepisp says:

    The hate against ionic liquids is unjustified since they’re very new solvents and require procedure development. They’re salts. You can’t expect them to function like molecular solvents. They can’t replace common solvents, but in the things they’re good at, there are no or only bad replacements within molecular solvents. They’re generally nontoxic enough to stick your hand (not a metaphor) but then dissolve things like high polymers that no not-awful molecular solvent can. If you’re worried about trivial things like having to heat it, then your problem is probably too easy to be solved by any exotic method like ionic liquid solvent.
    Also, most often when people sell “ionic liquids”, they only refer to a very specific set of imidazolium ionic liquids.
    That being said, 2-methoxyethanol, propylene carbonate, sulfolane and dimethyl phthalate are probably the most exotic molecular solvents I have used.

  94. Hano3s says:

    Done a Birch reduction in liquid ammonia; lovely blue colour would appear as you slowly dissolved the hammered-to-a-foil and precleaned-in-ethanol bits of lithium in it.
    Also used anisole in synthesizing roxindole (doi:10.1016/S0960-894X(99)00135-3), can still remember the greasy candy-like smell, but never used it since.
    Most extreme solvent I used probably has to be bromine. Tried to brominate a thiazole, couldn’t get it to run in CCl4, with a few equivs of Br2, neither in acetic acid, so ended up dissolving it in neat bromine and slowly spooning in iron powder. Thing started to reflux and neatly (2 spots on TLC) gave 60 % product with 40 % of starting material, no more, no less.

  95. Jonathan says:

    Don’t forget ethylene carbonate. Uniquely high dielectric. Worked for us in quaternising certain tertiary amines when nothing else worked

  96. simpl says:

    e of autoimmune diseases), and it worked. And my lipid profile is so clean it squeaks. But I have quality fat-burning genes, I guess.
    ric mixtures like xylene.
    – our chemical production use a lot of solvent mixtures, which are reused as distillate mixes
    – Acetonitrile and ethyl acetoacetate are perhaps two less mainstream ones
    Otherwise the old favorites

Comments are closed.