Nickel-catalyzed asymmetric hydrogenation of oximes

The asymmetric hydrogenation of oximes to chiral hydroxylamines is a long-standing problem due to the labile N-O bond and inert C=N bond. It has now been enabled by a chiral nickel catalyst with excellent results in which the weak interactions between catalyst and substrate play crucial roles.

Like Comment
Read the Paper

Transition metal-catalyzed asymmetric hydrogenation provides a convenient and practical method to prepare optically pure molecules from prochiral unsaturated compounds. This methodology has been widely applied to the industrial synthesis of key intermediates of pharmaceuticals and other bioactive compounds as recognized by the Nobel prize awarded to Knowles and Noyori. However, the efficient asymmetric hydrogenation of oximes for the preparation of the corresponding chiral hydroxylamines has remained a great challenge over a half century despite these products being vital chiral molecules (Fig. 1a). Most efforts targeted at the catalytic reduction of oximes to hydroxylamines produce only primary amines as a result of the cleavage of a weak N-O bond (Fig. 1b). This is because that in addition to the intrinsically low reactivity of the C=N bond of oximes, the N−O bond can be readily ruptured under reduction conditions due to the repulsion between the lone pairs of both the N and O atoms of the N−O group. A compromised strategy to overcome this drawback is the use of the relatively stable oxime ether as the substrate. Nevertheless, only two catalytic systems have been applied in the hydrogenation of oxime ethers without the cleavage of the N-O bond (reported by Oestreich and Cramer groups).1-2

Fig. 1 | Reduction of oximes to chiral hydroxylamines.  a, Representative chiral/achiral compounds bearing N-O skeletons. b, Previous work concerning the metal-catalysed hydrogenation of oximes to amines. c, This work: Ni-catalysed asymmetric hydrogenation of oximes to chiral hydroxylamines via weak attractive interactions. The reaction gives yields of up to 99%, with 99% e.e., and with a substrate/catalyst ratio (S/C) of 1,000.

Recently, it has been recognized that the weak attractive interactions (CH/π‧‧‧HC/π) between the catalyst and substrate, which are usually present in biological transformations, play an important role in reducing the reaction barrier and securing the generation of chirality in the asymmetric catalytic reactions. We have also developed several metal-catalyzed asymmetric hydrogenations of poorly activated unsaturated compounds which were greatly promoted by weak interactions.3-10 The DFT calculations of these catalytic cycles all indicated that the weak attractive interactions play a key role in stabilizing the transition state. Thus, we envisioned that we could try to use the weak interactions between the selected catalyst and oxime to lower the barrier of reduction of the C=N bond. In recent years, we have turned our attention to transition metal-catalyzed asymmetric hydrogenations based on earth-abundant metals, such as cobalt and nickel, which are inexpensive and environmentally friendly compared with rhodium, iridium and other noble metals. In this work, we report a nickel-catalyzed asymmetric hydrogenation of oximes for the preparation of the corresponding chiral hydroxylamines with excellent results (Fig. 1c).

Fig. 2. | Selected chiral products.

To accomplish this work, we explored a series of reaction conditions, and found that Ni(OAc)2∙4H2O/(S,S)-Ph-BPE was the best catalyst system for this hydrogenation. The ligand (S,S)-Ph-BPE not only has four phenyl groups which could provide CH/π‧‧‧HC/π interactions with the oximes, but also has a compatible backbone with rigidity and flexibility which could contribute to forming a favorable reaction space suitable for the very small structure of the oximes. In addition, we also studied the effects of the acids on the reaction and found that the steric hindrance and chirality of the acid have little effect, while its acidity was of great importance. The acid is required to dissociate the product from the hydrogenated complex. However, the acid must not be too strong, otherwise the protonation of the starting reactant oxime will interfere with its efficient coordination to the Ni catalyst. With the optimized conditions in hand, we next evaluated the free oximes (31 examples) as well as oxime ethers (18 examples) and found that most of the tested substrates were converted to their corresponding chiral products with good yields and excellent enantioselectivities (Fig. 2).

Fig. 3. | Computed mechanism. a, The catalytic cycle. b, Energy profiles. c, Transition states.

In order to understand the mechanism of chirality generation in this reaction, it is important to consider first the results of the acid-less hydrogenations that demonstrated practically perfect enantioselection in one catalytic cycle. These data clearly indicate that acid is not involved in the stereospecific stage – it is necessary for recovering the catalyst from a stable catalyst-product complex. Next, a computed catalytic cycle for substrate (E)-1a is shown in Fig. 3a-b. The optimized structures of TS1(S) and TS1(R) are compared in Fig. 3c. The TS1(S) is stabilized by numerous non-covalent intermolecular interactions, especially the CH‧‧‧HC/π interactions. Such interactions in TS1(R) are notably fewer in number. Even so, it is not completely possible to evaluate quantitatively the strength of each of the non-covalent interactions, because it depends not only on the corresponding interatomic distance, but also on the orientations of the respective bonds which in turn are expected to be strongly affected by possible conformational changes. Nevertheless, if we would look just at the numbers of the interatomic distances in the range 2.0-3.0 Å, we can conclude that the entire difference in their total numbers (TS1(S) has 3 C-H‧‧‧H-C interactions more than TS1(R)) is due to the better stabilization of the phenyl ring of the substrate; this is in line with the better suitability of this ligand for aryl-alkyl substrates.

For more details, especially on product derivatization and further discussions of the mechanistic studies, please see our article: https://www.nature.com/articles/s41557-022-00971-8

 References

  1. Mohr, J. & Oestreich, M. B(C6F5)3-catalyzed hydrogenation of oxime ethers without cleavage of the N-O bond. Chem. Int. Ed. 53, 13278–13281 (2014).
  2. Mas-Roselló, J., Smejkal, T. & Cramer, N. Iridium-catalyzed acid-assisted asymmetric hydrogenation of oximes to hydroxylamines. Science 368, 1098–1102 (2020).
  3. Chen, J. et al. Pd(OAc)2-catalyzed asymmetric hydrogenation of sterically hindered N-tosylimines. Commun. 9, 5000 (2018).
  4. Zhang, J. et al. Chemo- and enantioselective hydrogenation of ɑ-formyl enamides: an efficient access to chiral ɑ-amido aldehydes. Chem. Int. Ed., 58, 11505–11512 (2019).
  5. Chen, J. & Gridnev, I. D. Size is important: artificial catalyst mimics behavior of natural enzymes. iScience 23, 100960 (2020).
  6. Li, B., Chen, , Zhang, Z., Gridnev, I. D. & Zhang, W. Nickel-catalyzed asymmetric hydrogenation of N-sulfonyl imines. Angew. Chem. Int. Ed., 58, 7329–7334 (2019).
  7. Hu, Y. et al. Nickel-catalyzed asymmetric hydrogenation of 2-amidoacrylates. Chem. Int. Ed. 59, 5371–5375 (2020).
  8. Liu, D. et al. Ni-catalyzed asymmetric hydrogenation of N-aryl imino esters for the efficient synthesis of chiral α-aryl glycines. Commun. 11, 5935 (2020).
  9. Hu, Y. et al. Cobalt‐catalyzed asymmetric hydrogenation of C=N bond enabled by assisted coordination and non‐bonding interaction. Chem. Int. Ed. 58, 15767–15771 (2019).
  10. Zhang, J. et al. Asymmetric hydrogenation of γ-branched allylamines for the efficient synthesis of γ-chirogenic amines. Sci. 1: e10021. (2021).

Wanbin Zhang

Professor, Shanghai Jiao Tong University