Many synthetic organic chemists will be familiar with the Chan-Lam reaction, where an organoboronic acid or ester is reacted with an amine in the presence of a copper salt. It’s an appealing way of making C-N bonds, because it’s complementary to Pd-catalyzed processes like the Buchwald-Hartwig reaction. The Chan-Lam transformation uses the “wrong” starting material from a palladium perspective, the organoboron compound instead of an organohalide, so you’re actually coupling two nucleophilic partners head-to-head, which always feels a little strange.
But if you’re familiar with this reaction, you’re also familiar with a lot of foul vocabulary to use while running it, because it’s always been a capricious beast. Chan-Lam aminations are flat-out notorious for suddenly performing poorly when you vary the starting materials in what seem to be trivial ways, and for stopping dead in their tracks at some point in the reaction, with no hope of revival. They’re also a lot harder to run with the (often easier to use) boronate esters than they are with the boronic acids themselves, for reasons that no one’s completely sure about, either.
Well, until now. This new paper from the Watson group Strathclyde University in partnership with GlaxoSmithKline (edit – corrected the credit) seems to have given this reaction the sorting-out it’s long needed. Building on work by Shannon Stahl’s group at Wisconsin on the mechanism of the related copper-catalyzed ether formation, the GSK folks have done a thorough investigation of the reaction intermediates and catalytic cycles. It’s a mess – for example, potassium acetate was found, in the Stahl work, to inhibit the ether formation, but for the aminations it inhibits some of them and actually enhances some others. You can get a mess of byproducts, too, as aficionados of the reaction already know: starting from an arylboronic species, you can end up with your desired product, mixed in with the plain aryl-H from protodeboronation, the aryl-OH from oxidation, and the aryl-O-aryl from oxidative coupling, none of which you probably were in the mood for. If you really bear down and make sure that the acetonitrile solvent and reaction conditions are strictly anhydrous, you can shut down most of that phenol and diaryl ether crap, but your reward is an increase in aryl-aryl product through homocoupling.
Careful control experiments, spectroscopy, mass spec work, and even X-ray crystallography of intermediate species led to several insights. One is that the dissociation of the copper acetate paddlewheel dimer species (which is how that one starts off) is a key event in the reaction. It has to form a monocopper species that has an OH, an OAc, a molecule of the amine coupling partner, and an acetonitrile solvent molecule all coordinated to the copper atom, and the equilibria that produce this active species are key. Another thing to note is that the reason that boron pinacol esters are so temperamental is that the pinacol they produce is a vicious inhibitor of the reaction, likely through formation of pinacol/copper complexes that shut down the whole copper-catalyzed mechanism.
The mechanism goes through a redox cycle with copper, and another big factor in the success of the reaction is the re-oxidation of Cu(I) to Cu(II) species. That takes place with molecular oxygen (this is definitely not a reaction that you want to run under strict inert gas conditions). The copper-one species seem to be responsible for a lot of the side reactions, so coming up with conditions where it doesn’t hang around is a critical factor. As it happens, adding a tertiary amine (like triethylamine) to the reaction does two good things at once – it reacts with the acetic acid generated from the copper acetate starting reagent and prevents it from going back to the inactive paddlewheel species, and later on, the triethylammonium acetate provides some acid to help re-oxidize the copper (I).
The group then solved the pinacol problem mentioned above by adding plain old boric acid to the reaction, which complexes with it so that the valuable copper species doesn’t, so those two things are the key: add triethylamine and boric acid to your Chan-Lam reactions, and they will be much happier, and so will you. The same additives also make the related copper-catalyzed etherification and thioetherification reactions fly right, for the same reasons, and you can get the experimental conditions for all of these from the Supplementary Information file for free, if you need them. And if you ever run one of these reactions, believe me, you need them. Finally, this reaction is ready to move out of the foul-language category!