Chiral γ-amino alcohols are crucial building blocks in numerous pharmaceuticals and bioactive molecules, including antidepressants such as (S)-duloxetine, (R)-fluoxetine, and (R)-atomoxetine. Existing synthetic methods for these compounds, including nucleophilic substitution, Ru-catalyzed asymmetric hydroamination, Michael addition-asymmetric transfer hydrogenation, and asymmetric hydrogenation of β-amino ketones, each present limitations. Asymmetric hydrogenation of prochiral β-amino ketones offers a highly atom-economic and environmentally friendly approach, but achieving high enantioselectivity remains a significant synthetic challenge. Previous attempts using Ru and Rh catalysts often suffered from drawbacks such as high catalyst loading, long reaction times, poor stability, narrow substrate scope, and low reactivity/enantioselectivity. While iridium catalysts have shown promise in asymmetric hydrogenation of various unsaturated compounds, their application to the challenging β-amino ketones has been less explored. This research aimed to address these limitations by designing and synthesizing a new series of iridium catalysts with novel ligands to achieve efficient and enantioselective hydrogenation of β-amino ketones, leading to the synthesis of a wide range of chiral γ-amino alcohols, including key intermediates for important pharmaceuticals.

The literature extensively documents the enantioselective hydrogenation of β-amino ketones using chiral Ru- and Rh-phosphine complexes. Achiwa's (2S,4S)-MCCPM-Rh catalyst, for instance, achieved (R)-fluoxetine synthesis with 90.8% ee. However, this catalyst exhibited lower enantioselectivity with other substrates. Noyori and colleagues reported efficient protocols using chiral [RuCl2(diphosphine)(1,2-diamine)] catalysts for γ-tertiary amino alcohols. Zhang's group demonstrated the enantioselective hydrogenation of β-secondary-amino ketones with Rh-duanphos catalysts, achieving high yields and ee values. While Rh and Ru catalysts have shown some success, they often require high catalyst loading, extended reaction times, and exhibit poor stability or limited substrate scope. Ir catalysts have been successfully applied to various asymmetric hydrogenations, but their use in β-amino ketone reduction has been limited. Zhou's SpiroPAP-Ir catalyst demonstrated high TON in simple ketone hydrogenation. Zhang and Zhong reported asymmetric reductions of simple ketones using Ir-PNN complexes, but these were limited to symmetric vicinal diamines. This study sought to overcome these limitations by exploring the potential of unsymmetrical vicinal diamines derived from amino acids as a more versatile and tunable ligand platform for Ir catalysts.

The researchers designed and synthesized a series of tridentate ferrocene-based phosphine ligands ((RS,CP,RC)-L1-L7) incorporating modular and tunable unsymmetrical vicinal diamine scaffolds derived from commercially available (R)-Ugi's amine and chiral amino alcohols. The catalytic performance of these ligands was evaluated in the iridium-catalyzed asymmetric hydrogenation of β-amino ketones, using in situ catalyst preparation by mixing [Ir(COD)Cl]2 with the ligands in iPrOH under 30 atm of H2. N-Boc-3-(methylamino)-1-(2-thienyl)-1-propanone (1a) served as the model substrate due to its synthetic challenge. Reaction optimization involved investigating the effects of various alkyl and aryl substituents within the ligand's unsymmetrical vicinal diamine scaffold and sulfonamide groups, assessing their impact on reactivity and enantioselectivity. The influence of different bases (LiOtBu, NaOtBu, KOtBu, LiOH, NaOH, KOH, K2CO3, Na2CO3) and solvents (iPrOH, MeOH, hexane, CH2Cl2, toluene, THF) was systematically studied. The optimal conditions were then applied to a broad range of N-Boc-substituted β-amino ketones with diverse heteroaromatic and aromatic rings and varied amino groups, including alkyl, benzyl, and those bearing electron-withdrawing/donating substituents. The substrate scope was further expanded to include β-amino ketones with different N-substituents relevant to the synthesis of (S)-duloxetine, (R)-fluoxetine, and (R)-atomoxetine, such as alkoxycarbonylamino groups and β-tertiary-amino ketones. The study also explored the hydrogenation of β-dialkyl amino ketones, both as free amines and hydrochlorides, with various aromatic and heteroaromatic rings and cyclic/acyclic amino groups. Gram-scale reactions were performed to assess the scalability and industrial potential of the optimized catalytic system. Enantiomeric excesses were determined by HPLC analysis, and the absolute configuration of a key product was confirmed by X-ray crystallography.

The researchers successfully developed a highly efficient iridium-catalyzed asymmetric hydrogenation system for β-amino ketones. The newly designed ferrocene-based phosphine ligands, particularly (RC,SP,RC)-L6, exhibited exceptional performance. The optimal reaction conditions involved using NaOtBu as the base and toluene as the solvent. A wide range of substrates, including those with various heteroaromatic and aromatic substituents and different amino groups, were successfully hydrogenated with excellent yields (up to 99%) and exceptionally high enantioselectivities (>99% ee). The system showed remarkable tolerance to sulfur atoms, unlike many other catalysts, and exhibited excellent reactivity and enantioselectivity even for challenging β-dialkyl amino ketones. The method successfully produced chiral γ-amino alcohol intermediates of (S)-duloxetine, (R)-fluoxetine, and (R)-atomoxetine. Gram-scale reactions with low catalyst loading (S/C up to 50,000) were achieved, showcasing the potential for industrial applications. The turnover number (TON) reached as high as 48,500. The use of the enantiomer of (RC,SP,RC)-L6, namely (SC,RP,SC)-L6, enabled access to the opposite enantiomers. The study demonstrated the synthesis of a diverse array of chiral γ-amino alcohols, including those relevant to potential analgesic and antidepressant agents.

The findings of this study significantly advance the field of asymmetric catalysis. The developed iridium-catalyzed system overcomes the limitations of existing methods for the synthesis of chiral γ-amino alcohols, offering a highly efficient and versatile approach. The use of the novel ferrocene-based phosphine ligands with unsymmetrical vicinal diamine scaffolds provides a highly tunable platform for optimizing catalytic performance. The broad substrate scope and high enantioselectivity achieved in this study highlight the system's potential for the synthesis of a wide range of valuable chiral molecules. The gram-scale synthesis of key pharmaceutical intermediates demonstrates the method's practicality and industrial relevance. The success in hydrogenating both β-secondary and β-tertiary amino ketones, including those containing challenging heteroaromatic substituents, underscores the method's robustness and versatility. The ability to access both enantiomers by employing enantiomeric ligands further enhances the practical utility of this catalytic system.

This research successfully developed a highly efficient and versatile iridium-catalyzed asymmetric hydrogenation system for the synthesis of chiral γ-amino alcohols. The novel ferrocene-based phosphine ligands demonstrated exceptional catalytic activity and enantioselectivity across a broad range of substrates. Gram-scale synthesis of key pharmaceutical intermediates was achieved, highlighting the method's industrial potential. Future research could focus on exploring the application of these ligands in other asymmetric transformations and further optimization of the catalytic system for even broader substrate scope and improved efficiency. Investigating the detailed mechanism of the catalytic reaction and exploring alternative ligand designs could lead to additional advancements in this area.

While this study demonstrates high efficiency and broad substrate scope, further investigation into the limitations is warranted. The specific interactions between the ligand and the substrate, and how these affect enantioselectivity, could be further elucidated through computational studies. Exploring the long-term stability and robustness of the catalysts under various industrial conditions should be conducted. Finally, while gram-scale reactions were demonstrated, scaling up to larger production levels requires further investigation to ensure consistent yields and enantioselectivity.