As a practicing drug discovery scientist for thirty-odd years (and some of them have been pretty odd, as the usual joke runs), I just cannot tell you how high a percentage of that time has gone into the manipulation of amine-containing compounds, or those made from them. There are non-nitrogen-containing drugs: steroid hormones and their derivatives are prominent examples, then some non-steroidal anti-inflammatories. You’ve got the anaesthetic gases, the phosphonates for osteoporosis, some stuff like dimethylfumarate, valproic acid, and a few other lipid derivatives. But at that point you’re going to be lifting the seat cushions and hunting under the couch for further examples, because most everything else has a nitrogen somewhere. Famously, Barry Sharpless used to tell people that he was willing to taste a small amount of any nonreactive compound that didn’t have a nitrogen in it, although I believe he reserved to right to dodge some marine natural products at his discretion. Primary/secondary/tertiary amines, aromatic and non-aromatic heterocycles, nitriles, amides, sulfonamides and more – it’s hard to escape Nitrogenland. And that’s just in pharma! When you get into industrial and specialty chemicals, you’re going to hit some high-volume amines and nitriles in the plastics industry, just for starters.
So there are a lot of synthetic chemistry reactions that involve introducing and modifying N-containing functional groups, and the most wide-ranging and reliable of them see heavy use in the drug business. And by “heavy use” I mean at the here-comes-the-rail-car production level all the way down to carefully dinking around milligram (or sub-milligram) amounts of new analogs in high-throughput synthesis plates. For all that, though, there are plenty of reactions that we don’t have that could be very useful, and it looks like one of those has just been published in this new paper from a team at the Leibniz Institute for Catalysis in Germany.
They’re looking at a process that not too many people have explored. It’s common to take a nitrile and reduce that CN triple bond all the way down to an amine, and this is done on very industrial scales indeed. But what if you stop at the first stage of that reduction, at the imine (CN double bond), and then find a way to exchange that nitrogen with one from another amine entirely? This sort of amine-cross-coupling has been described, but is not generally used (and it’s fair to say that there haven’t been many general ways to do it). Until now, perhaps. The paper describes a process using nickel triflate with phosphine ligands as a catalyst under hydrogen pressure to send the reaction along to all sorts of newly substituted secondary and tertiary amines.
Now, if you draw that out, you can imagine taking an R1CN and and R2NH2 system and getting a whole list of different products out of it – plain reduction of the nitrile to the primary amine, different secondary amines (two R1 groups or and R1/R2), different imines that stalled out and didn’t reduce, tertiary amines. The authors screening a number of catalyst combinations and conditions to arrive at one that seems to deliver the defined cross-coupling products quite reliably, though. It’s reliable enough for that pathway that you can take a nitrile and 1.5x of added primary and secondary amines and get around 90% yield of the cross-coupled product. If you take the nitrile in the presence of 5 equivalents of N15-labled ammonia (labeled ammonium acetate as the amine partner), you get the expected reduced nitrile structure, but with 90% incorporation of the labeled nitrogen.
And since there are a lot of amines out there, there are a lot of stuctures you can make rather quickly from a nitrile intermediate. The paper contains a long list of examples, including many drug structures where substituted piperazines and piperidines are brought in, in the presence of many other functional groups. Boronic esters, aryl halides, azoles, carboxylic esters, ketones, hydroxys, thioethers, and azetidines all survive the reaction conditions just fine, and chiral amine partners retain their chirality. The paper demonstrates scale-up to gram levels, but you can bet that process chemists will be looking at this to see how far up the reaction can be pushed, because dumping two things together under hydrogen pressure is operationally pretty attractive. There’s the 4% nickel catalyst load that you’re going to have to clean out of there, true, and the best solvent is the (not particularly cheap) trifluoroethanol, but these are early days, honestly. On the bench scale I would expect this reaction to be picked up quickly. Bring on the nitrile, and bring on the amines!
