Organic and (especially) medicinal chemists spend a lot of effort messing around with carbon-nitrogen bonds. It’s a pretty rare drug that has no nitrogen atoms in it, and because there are so many amines available, C-N substitution is a great way to explore a lot of structural diversity very quickly. If, of course, you have a reliable way to couple them.
If you want to substitute aryl rings with new nitrogen substituents via amine couplings, for example, the old-fashioned way is the copper-catalyzed Ullman reaction. That can work really well when the constellations are in the right position, but it can also be a trackless wilderness of copper salts and additives, trying to find the right combination – and it often requires a goodly amount of the copper reagents, so you’ll have to make sure you’ve cleaned that off of your products. Another copper-catalyzed reaction is the Chan-Lam coupling to boronic acids, which has long had a reputation for poor yields and “stalling out”, although more recent work may have tamed that.
The last 20 or 30 years have seen an explosion of palladium-catalyzed methods (the Buchwald-Hartwig reaction), which couple amines to aryl halides, arylboronic acids, and so on. That one also can be optimized to work very well, but the number of experimental variations on the reaction are simply beyond counting. Those didn’t emerge because people were bored, either: it can be very difficult to figure out what’s most likely to work. That’s why this coupling is a prime candidate for automated high-throughput reaction screening if you have a really important step to optimize: run a few dozen (or a few hundred) variations and see what works.
A recent area of work has been oxidative coupling, with redox photochemistry or outright electrochemistry as the electron-removing step. Those are quite interesting and can furnish some very neat transformations, but it’s safe to say that they’re still developing techniques, and there’s definitely not a “one size fits all” recipe here any more than there is with most of the reactions above. This new paper fits into this mechanistic category, but via a route that I certainly never would have thought of: how about doing C-N oxidative coupling at room temperature in water, without light or photocatalysts, without metal catalysts or oxidizing reagents, and without an electrochemical cell?
Weirdly, this group finds that these reactions can be accomplished simply via spraying the solution into microdroplets. We’re into the world of surface nano chemistry here, where things are very different from the bulk solution. With microdroplets, there’s less and less bulk solution involved as compared to the surface layer. There have been reports of accelerated reactions taking place in microdroplets, often revealed under mass spec analysis conditions, and it’s becoming increasingly clear that one of the big factors is that these surface layers are experiencing startlingly high electric fields which might be as high as one billion volts per meter. That corresponds to about three volts for a five-micron droplet, and under these conditions, hydroxy radicals and free solvated elections can be formed, because the potential for taking an electron off of a hydroxide ion is 2.72V. Hydrogen peroxide, for example, has been shown to form spontaneously in water microdroplets.
The authors demonstrate diarylamines forming new C-N bonds by coupling onto other aryl substrates, and mechanistically it looks like there are some radical-radical couplings as well as radical attacks on neutral/nonradical molecules as well. They’re sending these microdroplets right into the mass spec, but a key finding is that the yield of the products increases with longer microdroplet flight time, which argues that that’s where the chemistry is taking place (and not in the gas phase inside the mass spec). They can actually see most of the key radicals and intermediates in the mass spectra themselves, and the reaction rates are estimated to be about six orders of magnitude faster than the same reactions in the bulk water phase.
This area of research is certainly worth keeping an eye on; it really does feel a bit like getting something for nothing in terms of reactivity and rates. Scaling something like this up will be interesting, but surely not impossible, either, and it could provide very energy-efficient synthetic routes with no metals and no nonaqeous solvents in the waste stream. The process chemists will surely find a way if those advantages are waiting!
