B.R.S.M. When all you have is a hammer everything looks like a nail


Conditions You’ll Probably Never be Desperate Enough to Try

Or: Somewhat Like Cooking

We all know what it’s like to be desperate for a reaction to work; to need a result so badly, or to have a supervisor breathing down our necks, telling us that something must be possible. I’m sure all bench chemists can remember searching the literature for a procedure to prepare a given compound, only to find one, read it and then discard it for being too dangerous or just plain ridiculous. We can all probably think of a few procedures that we're just not paid enough to follow. For example, before Christmas I wanted to make some of this carboxylic acid.

Easy, right? Surely you just oxidise the 2,5-hexadienol that you can get in one step from allyl bromide, unprotected propargyl alcohol and indium? Nope, there are numerous preparations reported, and they all use either tin, or worse, nickel tetracarbonyl. A thing that myself and Derek Lowe definitely won’t work with.[1]

Yup, chemists love to talk about dangerous procedures. But what about ridiculous ones? I once heard an academic tell a story about his time as postdoc with Steve Ley. In his first few months in the group, another postdoc whom he referred to as ‘Crazy Norman’ had been working out the last few steps in a natural product synthesis. The final reaction was the selective reduction of one of three or four double bonds in the molecule by hydrogenation. Norman performed this reaction once on a small scale, got some NMRs and then left and moved on. After he was gone, Ley set another postdoc the task of making a little bit more of the natural product, by repeating Norman’s final step. But they couldn’t. The particular selectivity that Norman observed (and had the NMR data to prove) was never seen again. His successor tried everything – different catalysts, solvents and temperatures. New glassware, clean stirrer bars, old glassware and dirty stirrer bars. They stirred clockwise and anticlockwise, by the window and in the dark. It was rumoured that someone in the lab might have been doing some sulfur chemistry on the day when Norman did his reduction, so they tried even tried wafting some thiols at the flask full of catalyst and solvent, to maybe poison it just a little bit. Eventually, they rang up Norman to ask if he remembered doing anything – anything at all – unusual when he ran the reaction. He thought for a while and then replied, in total seriousness, “I think I might have washed the glassware with Vimto”.[2]

Image from americansweets.co.uk

Give that man a Tet. Lett.! Here are a few other examples of ridiculous things you’ll never try:

1.  Michael Additions, Catalysed by Genomic Salmon Testes

Tetrahedron, 2012, 68, 3086

Obviously the actual catalyst is the DNA, as it says in the title of the paper. The idea of DNA as a catalyst isn't new, but I'm not sure why such an exotic source is required. I mean, isn't it in, like, everything? Initially I imagined PhD students being sent out with nets and scissors to 'acquire' fresh catalyst, but it turns out that Aldrich sells it. I still don't think I'd try this.


2. Heterocycle Synthesis Catalysed by Animal Bone Meal

Tetrahedron Letters, 2011, 52, 3492–3495

“Animal bones were collected from nearby butcher shops. All of the attached meat and fat were removed and cleaned from the bones. The bones were then washed several times with tap water and left in open air for several days to get rid of odors. Later, they were transferred to the oven at 80 ºC for drying. The dried bones were crushed and milled into different particle sizes in the range 45–200 μm then calcined for 2 h at 800 ºC… The resulting material was denominated ABM. The catalysts obtained were characterized by X-ray diffraction”

The fun doesn't stop there as the ABMs don't do an awful lot until they're doped with Lewis acids such as ZnCl2 and CuCl2, but then you can use them to do all kinds of tricky transformations like the formation of 2-phenylbenzimidazoles from o-phenylenediamine and benzaldehyde.


3. Asymmetric Reductions Over Finely Divided… Carrots

J. of Mol. Cat. B: Enzymatic, 2006, 41, 103

Appreciated carrot

From the abstract:

“Reduction of (+)-and (−)-camphorquinones (1a, 1b) by various vegetables (carrot, potato, sweet potato, apple, Japanese radish, cucumber, burdock and onion) gave α-hydroxycamphor selectively. Using burdock, (+)-camphorquinone was reduced to give (−)-3S-exo-hydroxycamphor (4a) as major product in high stereoselectivity with high yield. Moreover, 1,2-cyclohexanedione (1c) and 2-methylcyclohexanone (1d) with various vegetables afford enantiomerically pure trans- and/or cis-alcohol, respectively. Various vegetable reduction gave a new idea of a biotechnological process.”

This isn’t a process that uses isolated enzymes, or cell cultures or anything fancy. Because, as the authors assert, ‘It is known that various enzyme is included in vegetables’. You literally put some sliced carrot in a flask with your substrate and shake. I find this statement about a control experiment quite amusing:

“Furthermore, for consideration of the influence of a microbe, we tried under the sterilization condition (a carrot is sterilized with 70% ethanol and an ultraviolet lamp for 6 h). Consequently, yield and stereoselectivity did not change.”

Doesn't the thought of finding a carrot under your TLC lamp make you smile?[3]

Many thanks to David Lindsay for bringing this to my attention!


4. Michael Additions to Nitrodienes… in Heineken

Org. Lett., 2008, 10, 4557

Admit it – if you had a reaction that worked great in H2O/EtOH 5% v/v you’d obviously try running it beer. I know I would. If I was refereeing this, I would definitely have suggested that a screen of different beers was carried out. Incidentally, did anyone else see that rather bizarre J. Chem. Ed. paper at the start of the year, titled ‘Beer as a Teaching Aid in the Classroom and Laboratory’? Apparently, ‘Beer was chosen as a teaching tool to maximize students’ class participation and systemize and enhance their knowledge of chemistry.’ Hmmm, okay. Still, a great abstract image:

Taken from pubs.acs.org

References Etc

1. You can also get a flavour of its nastiness from the Wikipedia page:

“Its LC50 for a 30-minute exposure has been estimated at 3 ppm, and the concentration that is immediately fatal to humans would be 30 ppm. Some subjects exposed to puffs up to 5 ppm described the odour as musty or sooty, but since the compound is so exceedingly toxic its smell provides no reliable warning against a potentially fatal exposure.”

There's not enough money in the world...


2. There’s another great Crazy Norman story, where he apparently staggered into the lab late one morning, clutching a half empty bottle of Bell’s Whiskey, and claiming that the alcohol content couldn’t be what it said on the bottle as he wasn’t anywhere near drunk enough. He then proceeded to waste the entire day GC-ing and NMR-ing the whiskey to try and get some proof of this, before writing an angry letter to the manufacturers. Whoever received it must have been pretty intimidated by the (nonsensical) GC traces and Cambridge University return address as they sent him a grovelling apology along with several bottles of whiskey by way of recompense.


3. My favourite ridiculous paper of recent times has to be this absolute gem from Angewandte on fluorescence spectroscopy of bananas. These are easily the best figures I've ever seen in a chemistry research paper:


Comments (39) Trackbacks (2)
  1. your piece on conditions you’ll never try actually has special meaning to me as i’ve spent a postdoc in NZ trying to carry out organocatalytic asymmetric reactions with (and you should cross your legs here) salmon testes DNA. this can also be had fron aldrich. the interest in using DNA is that its abundant and cheap, much cheaper than other catalysts in the aldrich catalogue.

  2. I once attended a conference where each speaker tried to one-up the other for nasty conditions. One ran his reaction in beer, which led the next to call back his lab and tell them to run it in coffee (works OK), then the next guy ran his in WHISKEY. (still, 40% yield or so!)

    Would you say, in fact, that this is “Just Like Cooking?”

    • Well, it’s nice to have a robust reaction (I think I started writing something about this once, but gave up for lack of any direction), but once you’ve found something works in an open flask, using off the shelf reagents, handled by a hungover undergrad, then that’s usually taking things far enough. I guess this is only like cooking if your aim when preparing a meal is to make the most outlandish thing you can think of that you can still eat.

    • Barry Sharpless is quite proud of his Cu-AAC click reaction because it works in both beer and whiskey

      • a professor, after a long days’ lecturing, retires to his office for a well deserved dram of his 15 year old single malt whiskey only to find it two fingers lower than he last left it. “strange”, he says, blaming his befuddlement on his increasing years.
        before leaving for home he drops by his lab and sees his eager students hard at work and glimpses a brilliant amber liquid stirring away in a flask. intrigued he inquires only to find out his postdoc has been following Sharpless’ protocol.

  3. I was looking for preparation of 2-acetyl-3-hydroxyfuran aka isomaltol. Turns out the best way to make it on a large scale is to reflux lactose with a mix of ethanol+AcOH+NEt3+piperidine for 1 day, upon which the glucose piece turns itself into the acetylfuran, with galactose stil hanging on. You can burn off the galactose by steam distillation with diluted phosphoric acid but the yields and purity is crappy. Much better is to use almond emulsin aka defatted almond meal. You need only a reasonable amont of this magic almond powder (400 mg for 2.9 g of the substrate) and after 4 days standing in water at rom temp you are rewarded with 79% yield of isomaltol… So you mess up lactose with almonds and destroy 10 chiral centers to obtain a nonchiral product with molecular weight one third of the starting sugar./
    Syn Comm 29(6), 989-1001 (1999)

    • That’s awesome! I’ve never heard of almond meal ever being used for anything chemical. Where does one obtained the defatted version?

      • I suppose by grinding fresh almonds with some gentle organic solvent that does not denature proteins too much (i.e. ether) but I have not done it myself. Emulsin is an ancient and crude preparation that most chemical supply companies by now replaced by more purified (and considerably more expensive) glycosidases.

      • You used to be able to buy it from Sigma although I don’t see it listed now – I suspect almond flour from the local health store would suffice.

        It’s a good source of mannosidase and a heck of a lot cheaper when doing prep scale reactions. Slightly different selectivity to jack bean mannosidase which also works well (and is still available from Sigma) in meal form.

        Tet. Asym., 2004, 2821

      • I recommend the almond-oxynitrilase work of Don Deardorff’s group at Occidental College. His avocados are pretty awesome, too.

        Tetrahedron: Asymmetry 2005, 16, 1655-1661

      • If its the same as almond flour, then its really good for converting aldehydes to cyanohydrins in an enantioselective manner. I shared a lab with a chemist who used it with neat HCN (he’d distill it and freeze it and thaw it out everytime he needed a bit)

        • in small volumes it is much better to use TMSCN in place of HCN – as a strong silylating agent it produces HCN in situ instantly with any OH source – moisture, equiv of iPrOH or MeOH, etc. It has much higher boiling point and does not polymerize so it can be stored at room temp. By the way, Schlenk storage flasks with a 14/20 or 24/40 sidearm that allows you to distill stuff into them (under Ar or highvac) are absolutely awesome and they cost only about $30 apiece

  4. This entry should have been titled “Chemists go bananas”.
    Anyone try enzymatic oxidation of 2,5-hexadienol?
    Did washing the glassware with Vimto work?
    I believe salmon sperm DNA is used routinely in molecular biology.

    • I’m not aware of anyone trying enzymatic oxidation. It’s not something I have ever done. As far as I recall there was no precedent for any kind of oxidation of that alcohol in the literature when I searched. When I tried to go to the acid in one step (Jones’ reagent; PhI(OAc)2, TEMPO, CH2Cl2-H2O; PCC, DMF) there were so many hard to remove by-products that it was quite hopeless, and although I could get the aldehyde with MnO2 or Dess-Martin trying to make the acid from that was again quite tricky. I eventually got enough acid to have a go at the next step, which turned out not to work as I’d hoped either. I changed routes before I solved this problem.

      Regarding the selective hydrogenation: history does not relate if the experiment was ever repeated with the Vimto washed glassware – I asked the same question myself.

      • there is an old Corey procedure that takes allylic alcohols directly to methylester of beta substituted acrylic acids, by using MnO2 (5+ equivs) in anhydrous MeOH with a catalytic amount of NaCN (about 0.1 equiv if I remember correctly). Supposedly it proceeds through cyanohydrine oxidation to acyl cyanide. In my hands it worked well but depending on the grade of used MnO2 the reaction was sometimes quite slow and or accompanied by E/Z isomerisation of the conjugated C=C in case of farnesol as a substrate.

        • Yeah, the so-called Corey-Ganem-Gillman oxidation. I think the original procedure actually calls for 5 equivs and the catalytic version came later (possibly an Istvan Marko invention). I didn’t actually try this as using any amount of any cyanide at my institution is a bureaucratic nightmare.

      • this is late but have you tried the Oppenauer Oxidation

  5. Possibly apocryphal story from (now deceased) emeritus prof about student who tried and tried to recrystallize compound to no avail. One night frustrated, he pissed in the flask and went home. Next day, lo and behold, crystals. Experimental writeup: “co-crystallized with uric acid”.

  6. There are also other papers using carrots for asymmetric reductions of ketones e.g. this one: http://pubs.acs.org/doi/suppl/10.1021/ol8029214

    I discovered it a while ago when i had a reduction that just didnt want to work selectivly and i was searching for all kind of conditions. It was high on my to do list anyway since i havent found a solution for this problem yet and after reading this post yesterday in the evening i desided to finally set it up today.

    Reaction is running right now, will maybe post if it worked on sunday…

  7. @gippgig

    Salmon sperm is also used to produce protamine, which is used clinically as an antidote for heparin.

  8. I’ve used salmon sperm before, although that’s more orthodox in Biochemistry than Organic Chemistry. It seems to be a very popular source of “generic” DNA; I’ve always wondered why, given that DNA’s quite easy to extract from E. coli…

  9. I’ve actually tried the carrot reduction mentioned by Ed above, though if I remember correctly, it was a paper from Indian authors (ca. 2004?). Was trying to obtain a chiral 1-indanol from the 1-indanone. Despite multiple attempts and following the procedure, I got nothing but orange tinted starting material. Maybe Indian carrots are different than North American carrots, or the indanone wasn’t a good substrate?

    I believe they used acetophenone as the test S.M.

  10. This paper — http://www.jbc.org/content/203/1/167.full.pdf — explains the particular choice of salmon testes: “… ripe salmon testes consist almost entirely of masses of spermatozoa …” (probably in contrast to other animals that don’t stock up on spermatozoa and release them all at once) and since spermatozoa are fairly small compared to their nuclei, it stands to reason that this might be a particular DNA-rich tissue among easily sourcable ones.

  11. Outrageous. I remember seeing some good “additives” during polycarbocycle syntheses, things like “pieces of clay plate” and “cracked gasoline”.

    I think the award goes to the labmate I had in grad school who proposed to do a McMurry coupling with some kind of catalyst doped on…wait for it…wool. Not exactly well-received.

    • A McMurry on wool? I would have figured all those low valent Ti/K/etc. species would beat the crap out of the support, and that drying it out would be unfun.

      We saw the carrot paper awhile ago, and one of the other chemists wondered why rutabagas weren’t on the list.

      • “I would have figured all those low valent Ti/K/etc. species would beat the crap out of the support”

        That’s what you’d think, but someone apparently got it to work. It was some kind of asymmetric McMurry coupling, and I think the guy’s downfall was more the lousy selectivity rather than the wool.

  12. Salmon sperm because fish sperm is these big bags inside the fish with a high concentration of DNA.
    Salmon: http://www.wordpress.tokyotimes.org/archives/salmon_sperm02.jpg
    Cod: http://www.aftenposten.no/migration_catalog/article5828135.ece/BINARY/w380/sild+torsk.jpg
    Failed to find a herring picture.

  13. Here is a paper of my former boss where they report reaction in Rattus norveg. urine and other bio-media 🙂

  14. I had a similar target molecule to your acid. In the end I oxidized homoallyl alcohol to the aldehyde and then performed a Wittig reaction with (2-Ethoxy-2-oxoethylidene)triphenylphosphorane to give the acid ester.

  15. Then there’s the (in)famous Tour paper that describes the synthesis of wonder-material(?) graphene from… dead cockroaches and dogshit. http://pubs.acs.org/doi/abs/10.1021/nn202625c

    • I had the privilege of hearing Tour speak about this last fall. It really was rather remarkable, if not a little bit incredulous.

  16. So… did the Vimto work? Enquiring minds want to know!

    • I asked the same question, but the academic telling the story didn’t know. I’m trying to think what I’d do in the situation. Probably try it, I guess!

  17. It doesn’t make sense to say “genomic salmon testes”, as in “genomic salmon testes DNA” it is the DNA that is genomic. 😛

  18. I believe salmon are tetraploids, i.e. have four sets of chromosomes unlike us diploid (two chromosome) humans. That would imply that the sperm has two sets of chromosomes with DNA as most other cells. I believe it has much more DNA than E. coli as well.

Cancel reply