Complex 3D molecules are ubiquitous in daily life, with functions in everything from high-performance materials to smart medicines. Just as 3D shape often reflects function on the macroscopic scale, so, too, does it determine microscopic behaviour. When building 3D molecules for applications, chemists must therefore develop synthetic routes that ensure that each atom is correctly positioned in the final product. However, even if such precision is achieved, some molecules can be produced as mirror-image isomers (enantiomers), and their properties might vary widely, affecting their use in applications. In a paper in Nature, Wendlandt et al.1 report a reaction that not only enables the synthesis of single enantiomers of molecules, but does so using a reaction mechanism that was thought to be intrinsically lacking in enantioselectivity.
A rich arsenal of synthetic methods is available to connect molecular fragments in a highly predictable manner, and substitution processes are among the most powerful of these. Known as SN1 or SN2 reactions, these processes share a common requirement for an electron-rich species (a nucleophile) and an electron-deficient species (an electrophile). However, the two reaction types proceed by distinct mechanisms.
In SN2 reactions, the approach of the nucleophile towards the reactive centre of the electrophile leads to the extrusion of a chemical group, known as the leaving group (Fig. 1a). The leaving group on the electrophile is then replaced — substituted — by the nucleophile. The bonds around the reaction centre behave much like the spokes of an umbrella turning inside out on a stormy day. This highly orchestrated mechanism ensures that the 3D geometry in the tetrahedral framework of bonds around the reaction centre is inverted, and the ‘information’ about that geometry, encoded by the framework (the stereochemical information), is not lost. In other words, only one enantiomer of the product is formed.
By contrast, the SN1 reaction involves the initial extrusion of the leaving group — independently of the nucleophile — to generate a charged species called a carbocation (Fig. 1b). Because this intermediate is planar, it can be intercepted by the nucleophile from either face, to generate the reaction product as a one-to-one mixture of enantiomers, thus losing the stereochemical information in the electrophile. Wendlandt et al. overcome this long-standing stereochemical limitation by reporting an enantioselective SN1 reaction.
The authors’ method can be thought of as emulating biometric facial recognition: they use a small-molecule catalyst to discriminate between the two faces of the transient carbocation. Much as algorithms efficiently analyse structural features of faces, the catalyst identifies one face of the cation and thus directs the nucleophile to the other (Fig. 1c). This facial discrimination manifests in the high enantioselectivity of the reaction. The formation of one-to-one mixtures of enantiomers is prevented, but, in contrast to the SN2 reaction, the stereochemical information in the starting material is irrelevant. The reaction is therefore said to be enantioconvergent.
The small-molecule catalyst used by Wendlandt and colleagues is combined with a Lewis acid (a molecule that accepts electron pairs from other molecules), to generate a negatively charged complex that can be thought of as the active form of the catalyst. This type of combination is common in nature, and often facilitates reactions that would be impossible using the catalyst alone. In the present study, the active catalyst activates electrophiles known as propargyl acetates by removing the leaving group (the acetate), forming a carbocation intermediate. This positively charged intermediate interacts with the negatively charged active catalyst to form an ion pair; such ion pairing has previously been used in other types of enantioselective catalytic reaction2. The active catalyst recognizes one face of the planar cation, shielding it so that only the opposite face can interact with an approaching nucleophile. The products of this SN1 reaction are thus obtained predominantly as one enantiomer.
Not only does Wendlandt and co-workers’ study transform an SN1 reaction into a catalytic, enantioconvergent process, but it also constitutes a powerful synthetic route for preparing molecular motifs called quaternary carbon centres3,4, which are notoriously challenging to make enantioselectively. Quaternary carbon centres have four different carbon-based substituents attached to a central carbon atom, and are commonly found in biologically active, naturally occurring compounds, such as morphine or various steroids. A great deal of structural diversity could be generated by varying each of the four substituents, making quaternary carbon centres valuable starting points for drug discovery. The authors’ study is a breakthrough in that it allows readily accessible racemic mixtures (one-to-one mixtures of enantiomers) of starting materials to be quickly processed to make structurally complex molecular scaffolds containing these motifs.
Wendlandt et al. exemplify their reaction using substrates that contain structural groups specifically chosen to stabilize the intermediate cation. However, the underlying concept is general, and will certainly be translated to related classes of reaction. The reported products are striking because the carbon atoms attached to the central atom in the quaternary centre represent a diverse range of electron orbitals (sp, sp2 and sp3). This means that the groups attached to the central atom have markedly different geometries and reactivities. Future work in which the four groups are varied will therefore produce highly versatile libraries of molecules that could be used in a wide range of reactions for synthesis, and allow the exploration of a large amount of ‘chemical space’ — the vast array of all possible molecules.
Nature 556, 438-439 (2018)