A hydrogen bond that reaches across eleven carbons to set chirality

Asymmetric radical chemistry is one of those fields that looks, from the outside, solved by now. Chiral catalysts control all sorts of reactions, and radicals are among the most useful intermediates in synthesis. The catch is that almost all of the well-controlled radical reactions work on resonance-stabilized radicals: benzylic, allylic, next to a carbonyl. The moment you move to an unstabilized alkyl radical, one sitting between two ordinary alkyl groups with nothing to delocalize the electron, the lifetime collapses and the coupling barrier drops almost to zero. There is no time for the catalyst to choose which face to react on, and stereochemistry slips away. A paper in Nature Catalysis on 10 July, from Guosheng Liu’s group in Shanghai working with a joint institute at East China Normal University, reports a strategy that finally controls these radicals.

The idea is called ligand-mediated radical orientation, or LMRO. The team took the classic bisoxazoline, or Box, ligand that copper chemists have used for chiral catalysis for decades, and added a hydroxyl group. The new Box-OH ligand forms a dynamic hydrogen bond with a polar functional group on the substrate, a phosphonyl or ester group that may sit eight to eleven carbons away from the radical centre. That weak, remote interaction pulls the substrate into a particular orientation at the copper centre, the way a magnet pulls a needle without touching it.

The mechanism, worked out with density functional theory, is the clever part. In the usual copper radical coupling, the radical adduct forms an intermediate Cu(III) species that collapses irreversibly, so stereochemistry is decided before the catalyst can choose. The Box-OH hydrogen bond lowers the energy of the radical-adduct step, which turns that formation reversible. Once reversible, the subsequent reductive elimination from Cu(III) becomes the step that decides which enantiomer forms. The ligand has not changed the reaction; it has rewritten the energy profile so the catalyst’s usual tools work again.

The results follow from that. The team applied the method to three hard cases: the asymmetric phosphonyl-cyanation of unactivated terminal alkenes, the trifluoromethyl-cyanation of the same, and the asymmetric cyanation of unactivated C(sp3)-H bonds. The enantiomeric excesses reach up to 95 per cent from the alkene reactions and 96 per cent from the C-H activation reactions. The remote functional group can be an acetyl group, an amide or an ester, and the hydrogen bond still reaches it. The effect disappears when the hydroxyl is removed, which pins the mechanism to the ligand.

Why this matters is the range it opens. Unstabilized alkyl radicals are the kind of intermediate you meet in making complicated aliphatic molecules, the scaffolds of pharmaceutical synthesis, and until now the choice was between resolving them later by chromatography or accepting a racemate. A catalytic step that sets the stereochemistry as the C-C bond forms, using a remote hydrogen bond as the directing force, removes that choice. The C-H variant is the more exciting half, because an inert aliphatic site on a complex molecule can now be functionalized with chirality already in hand.

The honest limits are the usual ones. The method leans on a remote polar group to do the directing, so a substrate with nothing hydrogen-bondable on the molecule will not work. The hydrogen bond is weak and dynamic, which is what makes it reversible, but it also means the reaction is not universal. The cyanation product is a nitrile, a useful handle but not the final functional group.

None of that blunts the central result. Asymmetric radical chemistry has been stuck on resonance-stabilized radicals for twenty years. A ligand that reaches across eleven bonds with a hydrogen bond, and uses that weak interaction to put the stereochemistry back where the catalyst can control it, is the kind of idea that looks obvious in retrospect and took a decade to find. Synthetic chemists now have a new lever on the unreactive half of the radical world.