Molecules can exhibit a handedness, known as chirality. This is crucial to many aspects of chemistry and biology because the mirror-image isomers (enantiomers) of a chiral molecule can have distinctly different properties, reactivities and chemical or biological functions. For example, nature often uses just one enantiomer of a family of molecules as building blocks to construct sophisticated structures such as DNA, and in other biological processes. The development of methods for synthesizing chiral molecules asymmetrically — predominantly as a single enantiomer — is therefore one of the most important goals in organic and medicinal chemistry. In a paper in Nature, Hölzl-Hobmeier et al.1 report an approach that can also be used to achieve a seemingly impossible task in asymmetric synthesis: the light-induced, catalytic and apparently irreversible formation of single enantiomers of molecules called allenes from a one-to-one mixture of enantiomers (a racemic mixture).
One modern approach to asymmetric synthesis is to use light to induce the formation of a particular enantiomer of a molecule, a strategy called photochirogenesis. Often complementary to conventional methods of asymmetric synthesis, photochirogenesis is useful for making single enantiomers of molecules that are difficult or tedious to prepare when in their ground states, but more easily made using light-induced (photochemical) reactions that proceed through electronically excited states.
Nevertheless, achieving highly enantioselective photochirogenesis is not a trivial matter, because excited molecules are short-lived and highly reactive, and because it is difficult to precisely control the stereochemistry — the geometrical arrangement of groups in a molecule — of products formed from reactions of excited molecules. The control problem has been overcome using a supramolecular approach2, in which a ‘guest’ molecule is fixed into a particular position and orientation within a chiral ‘host’ environment to enable better stereochemical control of the guest’s reactions. Hölzl-Hobmeier and co-workers have developed a new take on supramolecular photochirogenesis that they apply to allenes.
Allenes are organic molecules in which one carbon atom (designated C2) forms a double bond to both of its neighbouring carbon atoms (C1 and C3; Fig. 1). These molecules can assume a type of chirality known as axial chirality if two different groups are attached at each of C1 and C3. In their study, Hölzl-Hobmeier and colleagues used axially chiral allenes that have a lactam group attached at C1. The enantiomeric form of these allenes is fixed when they are in their ground states, but they spontaneously interconvert between the two enantiomers when excited to a state known as a triplet3.
The lactam group is designed to form pairs of hydrogen bonds with a molecule known as an enantiomeric photosensitizing template (T), which was developed previously by workers from the same research group4. T forms complexes with the allene, within which it can absorb visible light and transfer the energy to the allene, exciting the latter to a triplet state5.
So how does T induce the conversion of a racemic mixture of allenes into a single enantiomer, a process known as deracemization? The process begins with the formation of a complex of T with either of the two enantiomers (which are known as the S- and R-enantiomers, hereafter referred to simply as S and R; Fig. 2). Irradiation of the resulting complex T∙S (or T∙R) by visible light excites T into a triplet state (T*), which then transfers energy to S (or R). The excited molecule S* (or R*) is then released from the complex, regenerating free T for another catalytic cycle. The liberated excited molecules then undergo racemization, and — in the absence of any factors that discriminate between the two enantiomers —eventually relax to form both S and R products in the ground state in equal quantities.
However, the overall deracemization process can be enantioselective depending on two things: one is how strongly T binds to S to form a complex compared with how strongly it binds to R; the other is the relative rate at which energy is transferred from T* to S and from T* to R. For an allene that carries an extremely bulky group known as a tert-butyl, Hölzl-Hobmeier and colleagues’ experiments show that T binds to S about five times more strongly than it does to R. This makes sense in the context of the authors’ computational simulations, which show that S and R stack above T in their respective complexes, but that S stacks more closely to T than R does (see Fig. 4a of the paper1) — which suggests that T∙S is thermodynamically more stable than T∙R.
Moreover, the enantiomeric excess (e.e.) — a measure of the ratio of enantiomers in a sample of a chiral compound, where 100% indicates the presence of just one enantiomer — reported by Hölzl-Hobmeier et al. for the deracemized allene is 96% in favour of R. The rate of energy transfer for T* to each of the two enantiomers can therefore be calculated, and it emerges that the rate of energy transfer to S is about ten times the rate of transfer from T* to R.
The chiral environment generated by T for the allene in the complex therefore has dual, synergistic roles that lead to the extraordinarily high e.e.: when T is in its ground state, there is a thermodynamic preference for it to bind to S rather than to R; and when T is in its excited state, kinetic factors greatly favour energy transfer to S compared with transfer to R. Impressively, the authors demonstrated that they could even use their method to convert a sample of the S-isomer of the tert-butyl-bearing allene (which had an e.e. of 95%) to the R-isomer (which had an e.e. of 96%).
Hölzl-Hobmeier and colleagues did not make a wide survey of which chemical groups could be attached to the allenes without disrupting the enantioselectivity of the deracemization, so this remains to be explored. However, groups that could disturb the hydrogen bonding between T and the allene would need to be avoided or protected (temporarily converted into another group that does not interfere with the hydrogen bonding). Nevertheless, the authors show that 17 racemic allenes bearing a variety of groups (see Fig. 3 of the paper1) can be deracemized to produce single enantiomers of 89–97% e.e. in good to excellent chemical yields (52–100%). These e.e. values far exceed the value (3.4%) obtained for the first reported deracemization of an allene6 in the early days of photochirogenesis research. Another attractive feature of the new method is that it uses a small amount of catalyst (only 2.5 mol% compared with the amount of allene used).
The authors’ findings unequivocally demonstrate that supramolecular photochirogenesis, when appropriately designed, can be a powerful tool for asymmetric synthesis that cannot be achieved using conventional, heat-activated reactions. The new reactions might be limited by the need to append hydrogen-bonding groups to both the substrate and the photosensitizing template, and by the narrow range of compounds to which they are immediately applicable (which include sulfoxides and binaphthyl compounds). Nevertheless, the general concept and methodology, as well as the mechanistic details revealed by this study, will generate much discussion and open up fresh avenues of research.
Nature 564, 197-199 (2018)