Enantioselective Alkylation of Imines Driven by Visible Light

Oct 16, 2018 - −40 blue LEDs. 21. 99. 89. aReaction conditions: 1a (0.10 mmol), 2a (0.15 mmol), metal salt .... observed R-configuration in the prod...
1 downloads 0 Views 3MB Size
Article pubs.acs.org/JACS

Cite This: J. Am. Chem. Soc. 2018, 140, 15850−15858

Copper(II)-Catalyzed Asymmetric Photoredox Reactions: Enantioselective Alkylation of Imines Driven by Visible Light Yanjun Li, Kexu Zhou, Zhaorui Wen, Shi Cao, Xiang Shen, Meng Lei, and Lei Gong* Key Laboratory of Chemical Biology of Fujian Province, iChEM, College of Chemistry and Chemical Engineering, Xiamen University, Xiamen, Fujian 361005, China

Downloaded via KAOHSIUNG MEDICAL UNIV on November 21, 2018 at 09:52:26 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Asymmetric photoredox catalysis offers exciting opportunities to develop new synthetic approaches to chiral molecules through novel reaction pathways. Employing the first-row transition metal complexes as the chiral photoredox catalysts remains, however, a formidable challenge, although these complexes are economic, environmentally friendly, and often exhibit special reactivities. We report in this Article the development of one class of highly efficient asymmetric/ photoredox bifunctional catalysts based on the copper(II) bisoxazoline complexes (CuII−BOX) for the light-induced enantioselective alkylation of imines. The reactions proceed under very mild conditions and without a need for any other photosensitizer. The simple catalytic system and readily tunable chiral ligands enable a significantly high level of enantioselectivity for the formation of chiral amine products bearing a tetrasubstituted carbon stereocenter (36 examples, up to 98% ee). Overall, the CuII−BOX catalysts initiate the radical generation, and also govern the subsequent stereoselective transformations. This strategy utilizing chiral complexes comprised of a first-row transition metal and a flexible chiral ligand as the asymmetric photoredox catalysts provides an effective platform for the development of green asymmetric synthetic methods.



INTRODUCTION Visible-light photoredox catalysis has been developed into a powerful tool to construct carbon−carbon or carbon− heteroatom bonds in organic synthesis.1 Through high-energy intermediates such as radicals and radical ions, unique reactions that are unavailable under thermal conditions can be accessed. Significant advances have been achieved in this field by employing ruthenium(II), iridium(III) complexes, or organic dyes as photoredox catalysts.2 In contrast, the use of first-row transition metal complexes such as copper species has been much less frequently reported.3−7 Beside the aspects of relatively shorter excitedstate lifetimes and weaker visible-light absorption, an inherent drawback of copper complexes as photoredox catalysts is that the low reduction potentials of CuII → CuI might impede the closure of a photocatalytic cycle.4 Very recently, an appealing strategy involving light-accelerated homolysis (CuII−X → CuII−Y → CuI + Y•) has been developed to address this problem, and opens new avenues for copper-based photocatalysis.8 On the other hand, asymmetric catalysis promoted by visible light is emerging as an attractive synthetic strategy for chiral organic molecules. However, achieving a high level of enantioselectivity in the photoredox reactions remains a remarkable challenge.9 In the limited success that has been reported, the solutions typically rely on the dual catalysis involving a ruthenium or iridium-based photocatalyst and a © 2018 American Chemical Society

chiral cocatalyst, which often contains one or more precious metals.10 Bifunctional catalysis employing a chiral-at-iridium or rhodium complex as both the asymmetric catalyst and the photocatalyst has also been developed by Meggers and other groups recently.11,12 Although use of a single chiral complex of a first-row transition metal as the asymmetric photocatalyst would be economically attractive and environmentally friendly, it has been scarcely explored.13−15 An impressive pioneering work by Fu et al. disclosed light-promoted enantioselective C− N cross-couplings by a copper(I) catalyst containing a chiral phosphine ligand.13 Even more recently, a nickel(II)-DBFOXcatalyzed enantioselective photoredox reaction of α,β-unsaturated carbonyl compounds and α-silylamines was developed in our laboratory.14 These studies revealed that chiral complexes of first-row transition metal exhibit special features and reactivities in photochemical reactions beyond the role as inexpensive alternatives to ruthenium or iridium-based photocatalysts. Therefore, the development of new chiral photocatalysts or applications based on the first-row transition metals would be of great value and in high demand. Herein, we disclose the development of one class of copper(II) bisoxazoline complexes (CuII−BOX) as asymmetric/photoredox bifunctional catalysts in the light-induced enantioselective alkylation of imines (Scheme 1). Without Received: August 28, 2018 Published: October 16, 2018 15850

DOI: 10.1021/jacs.8b09251 J. Am. Chem. Soc. 2018, 140, 15850−15858

Article

Journal of the American Chemical Society Scheme 1. Strategy of Using a Single CuII−BOX Complex as the Chiral Photoredox Bifunctional Catalyst

Other metal salts were also tested in this reaction, and it was found that the cuprous salt Cu(MeCN)4BF4 provided similar results (entry 11), while Ni(OTf)2 and Fe(OTf)3 failed in generating the product (entries 12 and 13). Ligand screening experiments showed that the chiral BOX ligands with different substituents R1, R2, and R3 remarkably affected the conversion and enantioselectivity (entries 14−19). For example, L2 (R1 = iPr) only provided trace amounts of the product, while L3 (R1 = Bn) resulted in 37% conversion and 50% ee (entries 14 and 15). Ligands L4−6 with two side chains (R2, R3) gave a slightly improved the enantioselectivity (entries 16−18). Such sidechain effects of chiral BOX ligands have been intensively investigated by Tang et al. and others.19 According to these results, a sterically more demanding ligand (L7) with adamantyl-modification was designed and used in the reaction. This gave the best result, a quantitative conversion in 6 h, with 85% ee (entry 19). Other oxazoline-type ligands L8−10 led to both lower reaction rates and lower enantioselectivities (entries 20−22). The chiral diphosphine ligand (L11) failed to deliver any product (entry 23). In all cases, the catalytic outcomes were strongly dependent on the light source. The reaction under irradiation with a 30 W red LEDs lamp did not proceed (entry 24), while the same process with a 15 W UV light provided lower conversion (entry 25). Ultimately, the enantioselectivity was further improved to 89% ee by increasing the concentration of 1a to 0.030 M and reducing the temperature to −20 °C (entries 26−29). Substrate Scope. With the best catalyst and reaction conditions in hand, we next evaluated the reaction scope. A range of alkyl trifluoroborates bearing substituents with different electronic and steric properties were examined (Scheme 2). Primary benzyl trifluoroborates with electronwithdrawing substituents on the phenyl ring gave products 3a−i with better enantioselectivity (87−94% ee), while those with an electron-donating substituent resulted in products 3k,l with reduced ee values. Halogen and ester substituents were well tolerated, offering the opportunity for further functionalization of the chiral sultam products. Naphthalen-2-ylmethyl trifluoroborate (product 3m) and thiophen-3-ylmethyl trifluoroborate (product 3n) were also compatible under the standard conditions. Moreover, a secondary benzyl trifluoroborate provided product 3o as a 1:1 diastereomeric mixture with moderate ee values (59% and 59% ee). The two diastereomers could be readily separated by silica gel chromatography. A tertiary alkyl trifluoroborate also afforded

the need of any other photosensitizer and under very mild conditions, the copper-catalyzed photochemical reactions afforded a range of chiral amine products containing a tetrasubstituted carbon stereocenter with good to excellent enantioselectivities.



RESULTS AND DISCUSSION Initial Experiments. In our primary studies on catalyst selection, we found that several chiral copper(II) bisoxazoline complexes (CuII−BOX) exhibited obvious absorption in the visible-light region. In particular, absorption in the range of 400−550 nm, attributed to metal−ligand-charge-transfer (MLCT), was one of the key features required for a visiblelight photocatalyst. In combination with their readily tunable chiral environment, redox-active metal center, and the established photoactivities of their reduced states−copper(I) species,4 we assumed that these chiral copper(II) complexes might be potential candidates for the development of asymmetric photoredox catalysis based on the first-row transition metals. Accordingly, we developed a model catalytic system using in situ generated copper(II) bisoxazoline complexes as the catalyst, benzyl trifluoroborate 2a as the radical precursor, and an α-carbonyl imine 1a as the coupling partner.16 The αcarbonyl adjacent to the CN bond of the imine substrate was thought to be able to act as a directing group for the asymmetric induction.17 Such a photochemical reaction would lead to the development of an inexpensive method for radicalbased enantioselective alkylation of imines under mild and convenient conditions.18 In an initial experiment, the mixture of substrate 1a, 2a, premixed copper salt Cu(BF4)2·H2O (10 mol %), and ligand L1 (11 mol %) in chloroform was stirred at 25 °C in argon under irradiation with a 24 W blue LEDs lamp. The desired product (3a) was produced in quantitative yield and with 74% ee (Table 1, entry 1). In the absence of the blue LEDs lamp, stirring in the dark at room temperature (entry 2) or 80 °C (entry 3), removing the ligand (entry 4), or the copper salt (entry 5), the reaction failed to proceed. Addition of another photosensitizer such as [Ru(bpy)3](PF6)2 dramatically accelerated the reaction (2 h, 100% conversion, entries 6 and 7), implying a light-induced pathway. Water was well-tolerated in the system (entry 8), while protic additive such as methanol led to the lower reaction rate and slightly reduced enantioselectivity (entry 9). 15851

DOI: 10.1021/jacs.8b09251 J. Am. Chem. Soc. 2018, 140, 15850−15858

Article

Journal of the American Chemical Society Table 1. Initial Experiments and Optimization of Reaction Conditionsa

entry

metal salt

ligand

T (°C)

light source

t (h)

conv. (%)b

ee (%)c

1 2 3 4 5 6d 7d 8e 9f 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26g 27g 28g 29g

Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O none Cu(BF4)2·H2O none Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(OTf)2 Cu(MeCN)4BF4 Ni(OTf)2 Fe(OTf)3 Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O Cu(BF4)2·H2O

L1 L1 L1 none L1 L1 none L1 L1 L1 L1 L1 L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 L11 L7 L7 L7 L7 L7 L7

25 25 80 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 25 0 −20 −40

blue LEDs none none blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs blue LEDs red LEDs UV lamps blue LEDs blue LEDs blue LEDs blue LEDs

6 12 12 12 12 2 12 6 15 6 6 21 21 6 6 6 6 6 6 6 6 6 6 6 6 3 3 12 21

quant. 0 0 0 0 quant. 22 quant. 93 92 quant. 0 0