CONTINUOUS FLOW CHEMISTRY – INDUSTRY AND ACADEMIA PERSPECTIVES CHIMIA 2019, 73, No. 10 823 doi:10.2533/chimia.2019.823 Chimia 73 (2019) 823–827 © Swiss Chemical Society A Chemoselective and Scalable Transfer Hydrogenation of Aryl Imines by Rapid Continuous Flow Photoredox Catalysis Rowan L. Pilkingtona, Nikolai P. Rossouw a, Dean J. van Asa, and Anastasios Polyzos*ab Abstract: The chemoselective reduction of diaryl imines in the presence of competitively reducible groups is uniquely accessed through precise control of reaction and irradiation time by continuous flow visible light photoredox catalysis. The method enables the mild and efficient transfer hydrogenation of diaryl imines in the presence of sensitive functionality including halides, ester, ketone, and cyano groups. The flow protocol is efficient, rapid (>98% conversion within 9 min) and readily scaled to deliver multigram quantities of amine products in high purity. Keywords: Flow chemistry · Imines · Photoredox · Radicals · Transfer hydrogenation Rowan L. Pilkington was born in England Anastasios Polyzos was awarded his before moving to Australia in later life. PhD in 2005 from La Trobe University He studied Science (Chemistry) at the and appointed to Research Fellow at the the University of Melbourne, working in Australian national science agency, CSIRO his final year under the supervision of Dr. in the same year. In 2008 he pursued post- A. Polyzos on photoredox carbonylation doctoral research at University of Cambridge in flow. He has completed an MSc in the under guidance of Professor Steven V. Ley same group, with research interests in the FRS. In 2011 he returned to Australia to development of novel photoredox processes lead the flow chemistry and catalysis group and their application in flow processing. at CSIRO in Australia. He established an independent research career with appointment to Senior Lecturer Nikolai P. Rossouw was born in Richards within the School of Chemistry, University of Melbourne in 2015. Bay, South Africa, before immigrating His research interests include the development of new methods and to Australia. He graduated with a BSc enabling technologies for organic synthesis, photocatalysis, C–H (Chemistry) from the University of reaction discovery, and the development of sustainable industrial Melbourne, working on photoredox process chemistry. Anastasios currently serves as Director of the thiolation in flow and synthesis of organic Australian Research Council Industrial Transformation Training photoredox catalysts with Dr. A. Polyzos. Centre for Chemical Industries. He is continuing a MSc with the Polyzos laboratory developing natural product total The ubiquity of aryl amines in pharmaceutical, agrochemical synthesis in flow. and industrial products has fostered the development of methods enabling their straightforward and scalable production.[1] Dean J. van As grew up in Melbourne, Innumerous protocols for the synthesis of aryl amines exist, [2] Australia. He obtained an MSc from the encompassing SNAr of aryl fluorides, Ulmann, Buchwald- chemistry department at the University of Hartwig and Chan-Evans-Lam coupling,[3] and hydrogen Melbourne in 2015 under the supervision of borrowing.[4] One of the most robust and well-established Dr. W. Wong. He is currently completing his methods is reductive amination by heterogeneous catalytic PhD studies at the University of Melbourne hydrogenation of imines on platinum and palladium metal. under the supervision of Dr. A. Polyzos in Whilst catalytic hydrogenation is scalable and generally photoredox catalysis. His areas of interest economically viable, chemoselectivity is challenged by the involve the development of new visible light presence of multiple unsaturated bonds or other reducible photocatalytic methods for the synthesis groups on the intermediates and products.[5] The development and functionalization of amines in batch and flow. of methods for reductive amination under mild conditions with requisite chemoselectivity and functional group tolerance remains a focus in industrial, and academic laboratories.[6] These requirements become progressively important when reductive amination is required in the late-stage functionalisation of targets with increasing complexity, including bioactive lead molecules and natural products. Mild hydride transfer agents *Correspondence: Dr. A. Polyzosab, E-mail: [email protected] aSchool of Chemistry, The University of Melbourne, Parkville, VIC 3010, Australia; bCSIRO Manufacturing, Clayton, VIC 3068, Australia 824 CHIMIA 2019, 73, No. 10 CONTINUOUS FLOW CHEMISTRY – INDUSTRY AND ACADEMIA PERSPECTIVES Scheme 1. Continuous flow Our Previous Work: processing enabling the scalable, chemoselective photoredox transfer hydrogenation of imines. Ir(ppy)2dtbbpy(PF6) (1.5 mol%) FG = functional group. N NEt3 (5 eq.) HN Ar Ar MeCN, N2 Ar Ar 14 W blue LEDs H FG over-reduction at extended reaction times: challenge for scale up + e + e I PC + H I PC H + H + H N HN HN Ar Ar Ar Ar Ar Ar This Work: Ir(ppy) dtbbpy(PF ) (0.5 mol%) FG 2 6 FG NEt3 (5 eq.) N HN Ar Ar Ar Ar 7 - 9 mins H MeCN, N2 14 W blue LED photo/flow reactor Chemoselectivity and scalability with continuous flow photoredox including NaCNBH and NaBH(OAc) offer opportunities for 4 3 NEt chemoselective reduction of imines, however a dependence on 2 H NEt2 r NA acidic conditions, solvent polarity and imine electrophilicity R impedes application to a broader selection of substrates.[7] H NEt H V Ar VI Furthermore, the requirement for stoichiometric hydride transfer 3 IV reagent and the formation of cyanide by-products (in the case of III NaCNBH4) may limit the translation of these methods beyond NEt2 laboratory preparation.[6] Ir(II) We recently disclosed a visible light photoredox method for Ir(III)* the catalytic transfer hydrogenation of diaryl-ketimines (Scheme r NA R NEt2 1).[8] The process utilises triethylamine as a highly economical Blue Ar single source of electron, proton and hydrogen atom. The H LEDs Ir(III) reaction was characterised by high product yields, fast reaction II times and visible light as the solitary energy source. Based on [9] Ar N literature reports, our mechanistic hypothesis (Scheme 2) was R centralised on the direct single electron reduction of the diaryl- ketimine moiety (I) to the corresponding α-aminyl radical anion I Ar species (II) by the photoredox catalyst, followed by subsequent Scheme 2. Proposed mechanism of photoredox-catalysed transfer protonation and hydrogen atom transfer (HAT) from triethylamine hydrogenation. radical cation (IV) and triethylamine (III) respectively to afford the hydrogenated diaryl amine (VI). flow reactors compared to standard reaction glassware,[11] and Through the course of these investigations, we observed that the continuous nature of flow chemistry could be leveraged to imines bearing an aryl-iodo substituent underwent competitive efficiently scale the photocatalytic chemoselective reduction.[12] protodehalogenation with extended reaction times. To address this We initiated the study by assembling a flow platform limitation, we reasoned that the different rates of single electron consisting of a commercially available syringe pumping system transfer (SET) to the imine moiety relative to the aryl halide, (Syrris Asia) fitted with an injection port and sample loop (any could be exploited to furnish chemoselectivity that was dependent commercially available system is suitable for this application). on reaction time. Continuous flow photochemistry methods[10] A bespoke tubular photoreactor was fitted, consisting of two 7 offer the temporal and spacial resolution necessary to enable the W 447 nm LED arrays and a coiled perfluoralkoxy alkane (PFA) kinetically favoured reduction of the imine whilst minimising tube (3 mL, 0.75 mm i.d.). The system was pressurised by a back unwanted protodehalogenation. This selectivity mode can be readily pressure regulator, rated to 40 PSI (Fig.1) and the solvent reservoir achieved by establishing a finely-tuned set of reaction conditions was continually sparged with nitrogen. that impede the time-dependent over-reduction of the aryl halide The photoredox transfer hydrogenation of the aryl iodide and can be extended to other reducible substituents. Furthermore, bearing imine 1a was investigated and studies to establish improved photocatalytic activity can be expected since greater conditions for the quantitative production of iodoamine 1b over light permittivity is effected over the smaller pathlength in tubular the protodehalogenated product 1c were undertaken. Following CONTINUOUS FLOW CHEMISTRY – INDUSTRY AND ACADEMIA PERSPECTIVES CHIMIA 2019, 73, No. 10 825 Fig. 1. Reaction time screening for substrate 1a. Conversions determined by 1H NMR analysis of the crude reaction mixture. Performed on a 0.10 mmol scale. optimisation of residence time (Fig. 1) excellent conversions flow reactor configuration was modified to accommodate the (>97%) of the substituted imine 1a were achieved with rapid processing of larger volumes of reagent and solvent . The sample TR as low as 7 min. Incomplete conversion of 1a was observed loops were removed and the reagents were delivered from a at TR below 6.5 min, indicating that the transfer hydrogenation single flask under an N2 atmosphere. FlowIR™ infra-red (FTIR) was incomplete at faster flow rates. Slower flow rates resulting spectroscopic monitoring was utilised to monitor the reaction in TR more than 7 min increased protodeiodination of 1b to 1c. mixture in-line to provide