The concept of umpolung describes the reversal of the naturally occurring electrostatic polarization of chemical groups. It has now been used to make single mirror-image isomers of nitrogen-containing molecules. See Letter p.445
The action of biologically active molecules depends on the precise spatial arrangement of atoms that interact with biological targets. More than 95% of drug molecules1 contain nitrogen atoms because they improve the cell permeability and water solubility of the compounds, and strengthen their interactions with biological targets. Methods for the spatially selective assembly of nitrogen-containing molecules are therefore of considerable interest for drug discovery. Moreover, biological targets have a particular chirality (handedness). The ability to synthesize just one chiral form — one enantiomer — of biologically active compounds is thus also of great importance, because only molecules of the correct handedness will fit into their targets, in the same way that right-handed gloves best accommodate right hands. On page 445 of this issue, Deng and colleagues2 report a method that solves both problems by reversing the natural electrostatic polarization of groups called imines.
Every chemical group of a given molecule is characterized by a pattern of electrostatic polarization that dictates the group's reactivity towards other molecules. The polarization of carbonyl (C=O) and imine (C=N) groups places partial positive charge at the carbon atom of the group, making it electrophilic — attracted to areas of negative charge (Fig. 1a). Molecules that bear partial negative charge are called nucleophiles, and tend to attack electrophilic carbon atoms, thereby creating a chemical bond.
The 'natural' polarization of a group can sometimes be reversed, so that electrophilic sites become nucleophilic and vice versa. This concept is known as umpolung3, from the German term for reversal of polarity. The umpolung of an imine or of a carbonyl-containing compound, such as a ketone, would place partial negative charge at the carbon atom, rendering the atom nucleophilic (Fig. 1b). The development of synthetic strategies based on umpolung opens up fresh vistas for the construction of biologically active molecules.
Several catalytic strategies4,5,6 have been designed to invert the natural reactivity pattern of carbonyl-containing compounds. An analogous catalytic strategy that allows the enantioselective synthesis of nitrogen-containing compounds from imines is highly desirable for drug-discovery research. Deng and co-workers' ingenious solution to this problem relies on the reaction of transiently formed molecules called 2-azaallyl anions (Fig. 1c) with carbon-based electrophiles.
Their reaction builds on a widely used concept7 first pioneered by the chemist Vadim Soloshonok and subsequently improved by Deng's research group8 and by the chemist Yian Shi and his group9: the use of one enantiomer of a base to isomerize imines to form enantiomerically enriched amine compounds. The proposed intermediate 2-azaallyl anions are similar in reactivity to hydrazone compounds6 that have been used in umpolung reactions of carbonyl-containing compounds, and behave as carbon nucleophiles.
Several non-enantioselective transformations have previously been reported10,11,12,13, and the sole example of a highly enantioselective carbon–carbon (C–C) bond-forming reaction with 2-azaallyl anions was a palladium-catalysed coupling with carbon electrophiles14. Deng and colleagues' C–C bond-forming reaction is rather different. They used a chiral 'phase-transfer' catalyst to shepherd the base from an aqueous solution to the immiscible organic solution in which the reaction occurs, thus enabling the transformation, and also inducing enantioselectivity. The reaction products are modified versions of imines (Fig. 2), and can be readily converted into a variety of other nitrogen-containing compounds.
The most interesting aspect of this work is the clever catalyst design (Fig. 2). It was developed from a quinine compound that was originally derived from Cinchona plants and which has previously been used as a scaffold for other phase-transfer catalysts15. The authors found that the prototypical catalyst delivered a different product to the one they were targeting, but by manipulating the catalyst's groups they were able to redirect the course of the reaction. A large, electron-rich group on the catalyst's nitrogen atom was required for high reactivity and enantioselectivity.
In the previous work by Deng's group8, only highly activated imines could be used in the reaction, but the new catalyst allows a wide variety of imines to participate with nearly equal facility. Furthermore, the reaction proceeds with remarkable enantioselectivity, and yields the amine products with high fidelity. It is also easy to set up and tolerates air and moisture from the atmosphere. It remains to be seen whether the substrate scope can be further extended by manipulating the catalyst's structure, to allow the use of simpler imines and less reactive electrophiles than those reported in this paper. It should also be noted that the catalyst is not currently commercially available, but it seems to be uncomplicated to synthesize.
Deng and co-workers' findings illustrate the power of catalyst development for organic synthesis, and provide a straightforward route to chiral amines. Their method also adds to the arsenal of established umpolung strategies for carbonyl derivatives, and is complementary to other such methods.
About this article
Control of chemoselectivity in asymmetric tandem reactions: Direct synthesis of chiral amines bearing nonadjacent stereocenters
Proceedings of the National Academy of Sciences (2018)