Skip to Content

Palladium Does What It Wants

Organic chemists do a heck of a lot of palladium-catalyzed coupling reactions. And there are lists of catalysts that go all the way down the page for them, for good reason. Some couplings are robust enough to give you product under all sorts of conditions, while others are ridiculously picky, and need just the right sort of liganded Pd to work well. (Then there are solvent effects, additives, and more).

Complicating all this is the way that commercial palladium catalysts themselves vary in quality.(I’ll insert a nonpaid endorsement for both Strem and Frontier Scientific for your palladium-coupling needs in general). Everyone who does these reactions knows that if you’re using good ol’ palladium “tetrakis” that it should be bright, pale yellow, and not orange or (God help you) brown. But some of the other catalysts can be a bit harder to evaluate. Here’s a paper on palladium acetate, and it has some details that may surprise many of its readers.

The authors, from Johnson Matthey, report that the traditional prep for palladium acetate can give material that’s contaminated with both polymeric Pd acetate and with a complex that has a nitro group substituting for one of the acetate ligands. The polymer has been less of a problem in recent years, they say, but many samples have the nitro species in them. So how different are all these species? The authors did the head-to-head comparison, with pure samples of each (and obtaining those is not necessarily straightforward).

In a series of Buchwald-Hartwig couplings (with morpholine and other amines), all three catalysts were effective. But the polymeric acetate was the best of the bunch, performing consistently at low catalyst loadings. This is particularly interesting, given that the stuff appears to be almost completely insoluble. In Suzuki couplings, lower-temperature runs showed the nitro-containing species to be best, probably because of its greater solubility. As the temperature increased, all three Pd catalysts were effective. Meanwhile, some test Heck reactions showed unexplained effects: with 2-bromothiophene, the nitro catalyst was definitely worse, but with 3-bromopyridine, all three worked fine. Finally, they looked at the formation of fancier palladacycle catalysts (such as third-generation BrettPhos). In this case, the pure Pd acetate complex was by far the best. The polymeric material did not give product at all, and the nitro complex gave a number of impurities.

Well, this sounds like the palladium chemistry we all know and love, doesn’t it? Each reaction has its own ideas about what it wants to do, for reasons that are generally unclear, and you can’t necessarily guess what might happen when you change some variable. If it weren’t for the way that these reactions do such useful bond formations, we’d never put up with this crap, but here we are. So add this one to your Puzzling Palladium file, and good luck with your next tricky coupling.

21 comments on “Palladium Does What It Wants”

  1. Martin says:

    To the list of variables I would also add whether the reaction flask has ever seen Pd or other metal catalysts before and how/how well it was cleaned.

  2. Tetramethylsilane says:

    The most maddening thing for me in Palladium Chemistry is why once palladium has mirrored out in a reaction which isn’t finished, that adding more catalyst will almost never get the reaction going again.

  3. John Wayne says:

    This is a great place to get on one of my more usual soap boxes: understanding of reaction mechanisms. Financing for basic physical organic chemistry dried up a while ago, and this slow gaining of understanding in such an important class of reactions is the price we pay. The more financially fortunate research groups (insert your own favorites here) have the bandwidth to perform some mechanistic work, but it probably isn’t as much as we should be doing.

    I’d like to thank the people at Johnson Matthey for taking the time to publish this work, and a special thanks to their managers for not nixing the whole thing. Basic science for the win! (steps down from soap box)

  4. Devils Copper says:

    “Organic chemists do a heck of a lot of palladium-catalyzed coupling reactions”


  5. anon says:

    ohhh, the pun in the first sentence – outstanding 🙂

  6. passionlessDrone says:

    When the cold fusion guys were on 60 minutes a few years ago they insisted that the quality of palladium was critical toward getting results or not. Are they quacks? I have no idea, but it does seem like not all palladium is created equally.

    1. dave w says:

      I don’t think the “cold fusion” folks are -all- quacks: it seems like some of them (e.g., Rossi) may be operating in “fake it till you make it” mode, but others (such as Storms and McKubre) appear to be honest investigators… as for C.F. itself, I think it’s a real but elusive phenomenon, and the exact conditions are not well understood… but it does appear that something about the specific characteristics of the palladium (I suspect more related to the crystal structure than the exact chemistry) may be a significant part of the “required mojo”…

  7. Quinnybray says:

    It’s worth mentioning that the Pd3(OAc)5(NO2) problem has been specifically highlighted before and can cause problems during the synthesis of other Pd catalysts. See here:

  8. Dominic says:

    I’d just like to point out my old lab mates published on “what’s actually in your palladium acetate” earlier this year:

  9. tt says:

    A further complication is that with bisphosphines, it’s often the mono-oxide doing all the chemistry, so if you screen bisphosphines that are nice and new, you often see either no conversion, or very inconsistent run-to-run variation. See:

  10. DRP says:

    I’d like to point out that my old lab mates published on “what’s in your palladium acetate” a few months ago. They found that most samples came in the form of Pd3(OAc)5OH (they also found some of the nitro) and that if you expose pure Pd3(Oac)6 to any OH source (eg ethanol used to stabalise DCM) then the corresponding Pd(OAc)5OR is formed almost instantly

  11. Ted says:


    I made Buchwald catalysts (on scale) at my last o-chem stop. One of our customers had trouble with a batch. I eventually traced it to residual copper, a carryover from the ligand synthetic process. Trace amounts, but big problems, depending on the coupling.

    A couple of months after we had solved that problem, I was sent complementary ‘replacement’ batches of the ligand from a supplier (SAFC, Strem? – I don’t remember) I had purchased reference material from. They came out of the blue, but included a note to the effect that “it came to our attention that some previous lots didn’t meet our high standards…”

    I checked both lots – turned out to be with/without copper…


  12. milkshake says:

    Nice thing about Pd(II) compounds is that their 1H-NMRs are pretty sharp.

    Pd(OAc)2 can be re-crystallized from benzene. (Dubious PdCl2 I like to re-crystallize from refluxing acetonitrile, to get the yellow PdCl2.2MeCN which seems better behaved than PdCl2)

  13. Stuart says:

    Anyone thinking of using Pd2(dba)3 (or currently having problems with it) may find this paper helpful.

  14. Stephen says:

    shouldnt the first sentence be “Organic chemists stille do a heck of a lot of palladium-catalyzed coupling reactions”. Missed your chance Derek!

  15. SLN says:

    Detailed structural and mechanistic elucidation of these Pd catalyzed transformations would be of value, but perhaps the use of even 1 to 5 mole % Pd complex will not be sustainable or economical for commercial scale manufacture. Lipshutz et al. have shown that these Pd catalyzed coupling reactions may be accomplished efficiently in ppm quantity, and can be performed in water – this is not a typo! See Science, Vol. 349, pp. 1087-1091, September 4, 2015. And it seems that Novartis is quickly embracing and validating this green technology. See F. Gallou et al., Green Chemistry (Perspective), Royal Society of Chemistry, October 2015.

  16. Me says:

    Still doesn’t trump my 2 all-time faves in this area: the ‘palladium-free’ Suzuki’s that only worked with KOAc (traces of Pd) and the ‘Iron-catalysed’ couplings that Bolm published a paper on, showing that they were dependent upon source of iron (purity issues again, with both Cu and Pd being the culprit) – meaning that low loadings and water (I used aquaeous acetates for years with Pd) are nothing new.

  17. Me says:


    This site works much better when viewed in Mozilla. (Previously internet explorer)

  18. Jim says:

    I like both Strem and Frontier products, and I am not a paid endorser but a chemist that has been burned too many times by cheap imitations from overseas chemical sellers. We had a very valuable starting material and our purchasing department cheapened out and bought a pinacol ester for cheap. We went through 15K of starting material before we went to Frontier for the compound. The compound worked flawlessly!

    We now avoid using our purchasing department, so maybe that’s the problem. Bad reagents.

  19. I had the pleasure of seeing a talk by Valentine Ananikov on so-called palladium “cocktail catalysis,” or “whatever you start with in the flask will become some god-awful mixture of monomers, clusters, and nano-particles, at least one of which will probably make the chemistry go.” This mechanistic picture gives a pretty good explanation for why palladium has been able to be applied so ubiquitously to coupling reactions.

    More pertinently, he had a nice OM paper talking about the sketchiness of so called “Pd2(dba)3″…the punchline is that the true purity is probably not written on the bottle.

  20. Thomas Colacot says:

    Thanks Derek for reviewing our paper. We undertook this work to educate its users both from academia and industry. The good news is that we can supply high purity Pd Acetate. The byproducts also will be available in the future.

Comments are closed.