Skip to Content

Song of Experience

I noticed this paper when it came out, and filed it in my “Useful Lab Tricks” folder. Organic chemists who have used triphenylphosphine-driven reactions will want to do so, too. For the non-chemists in the crowd, triphenylphosphine (TPP) is a cheap, crystalline substance that is pretty easy to oxidize to triphenylphosphine oxide (TPPO), a noticeably lower-energy species. There are a number of reactions that take advantage of those thermodynamics, setting things up so that the TPP-to-TPPO reaction is part of the mechanism and letting that drive the whole thing forward. It’s not exactly the most atom-efficient way to go, but it gets the job done at the bench. (Process chemists will make a little mark next to your name if you suggest doing this on scale, though, so be warned).

The oxide is an even more crystalline compound, but there are problems. It’s soluble to some extent in a number of common organic solvents, and there tends to be a lot of it around in a reaction that’s using triphenylphosphine in this fashion. The classic Wittig reaction throws off a lot of triphenylphosphine oxide, as does the Mitsunobu reaction, and you can end up trying to purify your product out of a big mound of white stuff. Sometimes it’s straightforward, and sometimes it’s a real pain. It can form very nice crystalline complexes with other organic molecules, which can help you out if you want to grow those, but you can lose material that way if you don’t.

What this new paper has found is a straightforward way of causing all the TPPO to fall out as an even more definitively insoluble mass. Treatment with zinc chloride apparently forms a complex between the two that crashes out of most everything, and I can tell you that I’ve had a few occasions where that knowledge would have come in pretty handy. I wondered at the time how this was discovered, but then I came across this story (courtesy of Chemjobber on Twitter).

The inventor of this trick is 85-year-old Donald Batesky, who used to work at Kodak back in the 1960s doing custom synthesis for non-catalog orders, work that will expose you to a lot of different kinds of chemistry, for sure. After Kodak, he worked for Aldrich (along with some other ex-Kodak chemists) doing contract synthesis for them in rented lab space in Rochester. So it’s safe to say he knows his way around a synthetic lab; by this point he’s seen it all. And now he’s on his third lab career at the University of Rochester, helping out as an assistant in the chemistry department. That’s where he remembered some metal-complex work from back when and hit on the zinc chloride idea.

Long experience at the bench, if you’re any good at all, gives you a fund of knowledge that’s hard to pick up from the literature. What solvents to try first for a crystallization, what reactions have exothermic inductions that you have to watch out for, which reducing agents have the easiest workups, how to get rid of metal contamination in the final product, when to deoxygenate rigorously and when not to worry about it so much, ways to get rid of Common Byproduct X or Pesky Solvent Y. Donald Batesky has just contributed another one of those to the trade, and good for him. This guy is clearly a chemist (he’s already stipulated that he wants to be buried in a lab coat), and I would very much enjoy sitting down and talking with him. It’s always worthwhile to listen to people who are really good at their work or to watch them do it, and I’m glad to hear that he’s still sharing his knowledge with his co-workers at Rochester.

46 comments on “Song of Experience”

  1. drsnowboard says:

    Surely the students must call him ‘The Don’ ..

  2. myma says:

    The Don is a very good man.

  3. John Wayne says:

    Great post Derek. Please post more tricks like this. Us chemists have to stick together and continually prove our value to the companies we work for; nobody else is going to do it for us.

  4. anon says:

    I thought everyone knew and this was common knowledge. Zinc and copper salts are pretty good at removing phosphines and its oxides.

    1. Anon says:

      Not me! If it was known how come we did not see the publication? This procedure has lot of values for those who use TPP. Just sayin’

      1. anon says:

        Because Lewis basic sites in the product will compete with the phosphines and phosphineoxides to coordinate to these metal ions. It is not a universal solution. More like trial and error for your specific conditions, reagents and products.

        1. Barry says:

          I remember an early paper on the diasterioselectivity of ketone reductions reported higher ratios for equatorial delivery (axial product) than others. Turned out the author had used CaCl2 to remove water from the work-up, and it (somewhat selectively) removed equatorial alcohol product from the mixture as well. I expect depending on what you’re making, you could lose some to MgCl2 or ZnCl2, too.
          But–some days–I might not mind losing a couple of percent of product to be rid of all the Ph3PO/TPPO

      2. tlp says:

        Looks like science is approaching sort of a singularity point, where new literature appears on such enormous and ever-accelerating rate that old literature just has no chance of being read. So things are getting reinvented all the time. Barry Sharpless made a great point about that in his review on SuFEx click, having a Reference to a paper from 1927 right in the abstract (PMID: 25112519).

        This is in no way supposed to diminish Donald Batesky’s awesomeness!

    2. b says:

      I first came across using ZnCl2 to complex Ph3PO a long time ago in this paper: Alberts, A. H. Wynberg, H.; Strating, J. Synth. Commun. 1972, 2, 79. http://www.tandfonline.com/doi/abs/10.1080/00397917208081647

    3. Nick K says:

      It’s news to me. I just wish I’d known this when faced with multimole Wittig and Mitsunobu reactions.

  5. imarx says:

    Very cool story, but the fact that basic nitrogens or even anilines prevent this method from working will limit its usefulness to the med chemist. Hopefully someone can figure out a way around that issue.

    1. paul says:

      Adding some acid would probably prevent the amine coordinating to the zinc. Just have to hope the amine salt doesn’t precipitate out instead. Then again, if it did that you can just remove your phosphine by crystallisation of the desired product.

      1. milkshaken says:

        I don’t think there is easy way to fix this problem, in the old days before chromatography, chlorozincate complexes were one favorite method of isolating alkaloids from complex mixtures (plant extracts) in crystalline form. (I used this method as a kid, to isolate about a gram of nicotine from cheap bulk tobacco)

        1. NJBiologist says:

          Good grief–a gram of nicotine? What were you trying to do, kill every insect in the county?

          1. milkshaken says:

            I had no clue. After isolating the chlorozincate complex I decomposed it to free base but knocked over the distillation flask and spilled it all over the bench, so now I know what nicotine freebase smells like (fishy and rotten)

  6. HGMoot says:

    Perhaps in cases where precipitation is not optimal, extraction with aqueous ZnCl2 might do the job?

  7. Kevin says:

    Early on in grad school, I learned the trick of using dppe in Mitsunobu and Staudinger reactions on lab scale. that trick was actually published in the literature: Tetrahedron Letters, 1998, 39(42), 7787-7790.

  8. A Nonny Mouse says:

    Won’t let me reply, but here is an early paper on the complex

    http://pubs.rsc.org/en/content/articlelanding/1975/f2/f29757101640#!divAbstract

  9. A Nonny Mouse says:

    And already patented by BASF!

    US6011181

  10. Wavefunction says:

    What’s “Kodak”?

    And at first I read “He wants to be buried in a lab”: that’s not a bad idea either.

    1. anon says:

      Once upon a time they were doing basic chemistry and research. Lots of polymer science! And did you know that Super Glue was discovered in Kodak labs? I am sure they still into research it but more like OLED’s etc.

      1. Wavefunction says:

        Oh yes, Kodak used to be terrific (in the very early days R. B. Woodward was an important consultant). My comment was simply intended as additional sarcastic affirmation of the dark epitaph to the grand age of American industrial innovation.

        1. NJBiologist says:

          Ash, I found a t-shirt for you… [link in handle]

          1. Wavefunction says:

            Thanks! I’ll wear that for sure, along with the “This Used To Be Us” shirt gifted by Thomas Friedman.

      2. Richard Blaine says:

        Kodak sold off its OLED business lock-stock-and-barrel to LG about 2009, so no more OLED research there.

  11. KtheKnight says:

    In reply to A Nonny Mouse
    This finding was made by chemists in the process division of Boots in Nottingham, they preferred MgCl2 to get rid of large amounts of TPPO. In 1994, Boots divested its prescription drugs division (the inventor of Ibuprofen) to BASF, whose pharma division (Knoll AG) was mainly interested in the anti-obesity drug sibutramine of Boots being in clinical trials at this time.

    1. Ken says:

      Example of the MgCl2 in action on scale – OPRD 2103, p 666.

      1. Mister B. says:

        Small typo here:
        Org. Process Res. Dev. 2013, 17, 666−671
        Abbvie research 🙂

        Thank you btw, it was a nice reading!

        1. Harrison says:

          To bring this full circle, in 2001 Abbott buys Knoll. In 2013 Abbott spins off AbbVie. And that is how you have a AbbVie chemists publishing on Boots chemistry

  12. David says:

    Yet another case of people living long lives with fulfilling work. 🙂

    > ” I would very much enjoy sitting down and talking with him.”
    I’d be interested in you doing such interviews as a (regular?) podcast.

  13. Okemist says:

    MgCl2 works as well.

  14. Larry says:

    Great Post! I cringe when I think of the demise of Industrial research labs and the loss of institutional knowledge when the denizens are scattered to the wind like a diaspora. I was part of one and the knowledge base was beyond reproach. All the files are now in some mountain storage never to be seen again.

    1. Wavefunction says:

      Thanks for the link, what a wonderful man. The following statement caught my eye and made me feel morose: “This discovery was entirely Don’s invention, based upon some knowledge he had of metal salt extractions from the ’50s and ’60s”. And Batesky himself says “I did 11,000 literature searches and made 3,000 chemicals in 15 years, which is one per working day.”

      All the lay offs from industry are getting rid of people like Batesky. Their knowledge is irreplaceable. It’s worth reiterating: industrial chemistry is not like theoretical physics; experience is everything, young geniuses can’t replace old veterans like Batesky, no matter what their pedigree and publication record is like.

      1. Hap says:

        As someone else put it here before, many of the people who are taking up the jobs sent elsewhere don’t have pedigrees – only for the few jobs kept here do the pedigrees of applicants matter.

        Overall, cheap matters, as does now. Later and future don’t matter. Why spend time making and keeping knowledge that can’t be monetized? It’ll be someone else’s problem then. When your “thought leaders” act like Mickey Mantle on a thirty-year bender, knowing how it ended for him, then you can assume it isn’t going to go well for anyone.

  15. Gratified says:

    A Gr8 lab hack for getting rid of XS phosphine is to saturate your work up with Et2Zn. Solution works but neat is best. Last lab hack you’ll ever need!

    1. Pennpenn says:

      I am (unfortunately) not enough of a chemist to follow that at the moment, but honestly the words “last lab hack you’ll ever need!” feel like they’re implying “because your lab will be a smoking crater shortly afterewards).

      1. DRP says:

        that is exactly what is implied with this “procedure”.

        1. Tim says:

          Diethylzinc? Wikipedia notes that it is “highly pyrophoric” (ignites spontaneously in air). “Diethylzinc reacts violently with water and easily ignites upon contact with air.” “Because of its high reactivity toward air, it was used in small quantities as a hypergolic or ‘self igniting’ liquid rocket fuel—it ignites on contact with oxidizer, so the rocket motor need only contain a pump, without a spark source for ignition.” The Library of Congress experimented with it for deacidification of wood paper, but terminated the effort. “It has also been established that tight or loose packing of books; the amount of alkaline reserve; reactions of DEZ with degradation products, unknown paper chemicals and adhesives; phases of the moon and the positions of various planets and constellations do not have any influence on the observed adverse effects of DEZ treatment.”

  16. Me says:

    Whole thread reminds me of one of the senior managers who was in his 60s at the time back in my first role. We had a natural product derivative ready for candidate selection, but we couldn’t get a crystal of it. Everyone had taken a few 100 mg and tried everything they could think of. Strategic technologies had had a look at it and could do something on a very small scale. This guy put on a labcoat for the first time in about 10 years, wandered into the lab, picked up a vial of the material, poured in a couple of solvents, held it up to the light, ‘hmm’-ed a little at it, put it down and walked out. Next day huge needles had crashed out of solution.

  17. Wheels17 says:

    I worked at Kodak as an engineer in a production organization from the ’70s through the ’00s. The scientific and technical capabilities that were available were simply amazing. And we took it for granted. It is truly sad that management couldn’t figure out how to leverage that capability.

  18. Scott says:

    “Process chemists will make a little mark next to your name if you suggest doing this on scale, though, so be warned”

    A mark next to your name to Gibbs-smack you when they meet you, for making their life miserable, I take it? (Or a firmer response, since that sounds positively nasty to sort out of your continuous-flow system)

  19. Richard Blaine says:

    Met Don in Kodak Research Labs while he was renting a lab to make Aldrich stuff, as mentioned. Regarding that lab coat he wants to be buried in – his always strongly of phenol (or maybe cresol), so I doubt any microbes could get with in 3 feet of him. (You could tell when he had been in the freight elevator recently.) Maybe that’s why he’s still going strong in the synthetic lab at 85?

    When he shut down his lab, he gave me a big bottle of 9-bromoanthracene and a bunch of solvents, including methylcyclohexane. (“Good for tricky recrystallizations,” Don said. He was right, of course.)

    Don told me that he was on his third wife, and expected to outlive her as he had the previous two! Certainly his synthetic career is equal to three ordinary chemists.

  20. Richard Blaine says:

    Also, back in the glory days of Kodak chemistry, Don’s custom synthesis department used to send out a newsletter asking the company’s many chemists for alternative routes to customers’ requested molecules, and for improvements to current methods. There were also reports on previous suggestions. As a newbie back then, I was amazed how creative and knowledgeable the Kodak’s chemists were. Nearly all gone now…

  21. kjk says:

    “Long experience at the bench, if you’re any good at all, gives you a fund of knowledge that’s hard to pick up from the literature.”

    This is why “diligent and dumb” automation like Transcriptic (and others) is more likely to succeed near-future than fancy AI stuff (after a short stint (2-4 months?) of “gloves on time” to familiarize ones-self). Flexible automation is an experience amplifier, allowing much more rapid iteration to find out “what X is best for Y”.

  22. Kent G. Budge says:

    I look forward to a complementary “Song of Innocence.”

    I suspect it will involve surprised high-pitched shrieking.

Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.