Photo and copper dual catalysis for allene syntheses from propargylic derivatives via one-electron process

Different from the traditional two-electron oxidative addition-transmetalation-reductive elimination coupling strategy, visible light has been successfully integrated into transition metal-catalyzed coupling reaction of propargylic alcohol derivatives highly selectively forming allenenitriles: specifically speaking, visible light-mediated Cu-catalyzed cyanation of propargylic oxalates has been realized for the general, efficient, and exclusive syntheses of di-, tri, and tetra-substituted allenenitriles bearing various synthetically versatile functional groups. A set of mechanistic studies, including fluorescence quenching experiments, cyclic voltammetric measurements, radical trapping experiments, control experiments with different photocatalyst, and DFT calculation studies have proven that the current reaction proceeds via visible light-induced redox-neutral reductive quenching radical mechanism, which is a completely different approach as compared to the traditional transition metal-catalyzed two-electron oxidative addition processes. Transition-metal-catalyzed couplings of propargylic alcohol derivatives with organometallic reagents proceeds via two-electron transformations, which present limitations in scope and selectivity. Here, the authors present visible-light-mediated copper-catalyzed cyanation of propargylic oxalates to form allenenitriles via a one-electron pathway

In this work, we wish to report such a concept-a radicalbased efficient syntheses of allenenitriles from propargylic oxalates and TMSCN under the dual catalysis of photo and copper (Fig. 1c) 60

Results
Optimization of reaction conditions. We began our study on the coupling reaction of propargylic oxalate 1a with trimethylsilyl cyanide (TMSCN) under blue light irradiation in the presence of CuBr and photocatalyst, fac-Ir(ppy) 3  after increasing the catalyst loadings of CuBr and L5 to 15 mol% and 18 mol%, respectively. A wide range of reactive yet synthetic useful functional groups, such as sulfide (2e, easily poisoning Cu catalysis), amide (2f), halogen (2n, 2o, 2p), ester (2k, 2q), ketal (2g, 2s), terminal alkyne (2q), and terminal olefin (2r) were intact under the standard mild reaction conditions. Interestingly, under the standard conditions the propargylic oxalate 1h with a ketone unit was converted to nitrile 2h with the in situ formation of a synthetically useful enol silyl ether entity 63,64 in 65% yield. The thiophene unit in substrate 1t was also accommodated. Furthermore, products incorporating Boc-protected L-proline 2u, pentoxyifylline 2v, Boc-protected tropinone 2w and 2w', and raspberry ketone tetra-O-acetyl-β-D-glucopyranoside 2x, mestranol 2y worked well without affecting the other fragile functionalities. The structure of 2w' was unambiguously established by its X-ray analysis. The reaction could be easily conducted on gram-scales (2q and 2y), demonstrating the practicality of this protocol. Even the reaction of terminal secondary propargylic oxalates 1z and 1A still afforded 1,3-disubstituted allenenitriles 2z and 2A as the products in decent yields and a very high allene/ alkyne selectivity (25:1 and 14:1). 4-Phenylallenenitrile 2J could also be obtained via the current method in 54% yield as the only isomer, and the slightlylower isolated yield may be attributed to its instability. The reaction could be further extended to non-terminal propargylic oxalates, such as 1B, 1C, and 1K. When trimethylsilyl-substituted alkyne 1C was used, TMS-substituted allenenitrile 2C was produced exclusively in 88% yield, which was not readily accessible by other ways 65 and very useful in propargylation reaction 66,67 . For non-terminal propargylic oxalates with R 3 being Ph (1D) and CO 2 Me (1E), dinitrile products 4a and 4b were obtained, which must be produced from the Interestingly, when MgCl 2 or MgBr 2 •6H 2 O replaced TMSCN as the nucleophile, various chloroallene or bromoallene bearing sterically hindered adamantyl (12l or 13l), ketal (12s), ester, or terminal alkyne (13q) could be obtained in decent yields. As a comparison, TMSBr or TMSCl gave inferior results (Fig. 3). respectively. The configuration of endo-5 was unambiguously identified by X-ray analysis. Conjugate addition of 4-methylbenzenethiol with 2a afforded sulfur-substituted tetrasubstituted alkene 6 in an excellent yield 69 . Deuteration of α-H of 2m (R 1 = Me, R 2 = -(CH 2 ) 2 Ph) with D 2 O in the presence of K 2 CO 3 and n-Bu 4 NBr readily yielded d-2m in 96% yield with 96% D-incorporation. Hydrolysis of nitrile group in 2l (R 1 = Me, R 2 = 1-adamantyl) with a base produced allenyl amide 7 in 64% yield 70 . In addition, the ethynyl group in 2q underwent the Cucatalyzed click reaction with anti-HIV drug AZT (Zidovudine) 71 while the allenenitrile unit remained unreacted, offering useful handle for further synthetic elaboration.   Fig. 2 Substrate scope study. a CuBr (15 mol%) and L5 (18 mol%) were used. b Due to the difficulty of separating the two regioisomers, the yield value refers to the isolated yield of a mixture of alkyne and allene; the regioselectivity was determined by 1 H NMR analysis. c The reaction was conducted in 10 mL CH 3 CN.
Mechanistic studies. To probe the reaction mechanism, we conducted a set of mechanistic studies. First, several propargylic compounds with different leaving groups 1F (Boc), 1G (Ac), 1H (CO 2 Me) were prepared. The Cyclic Voltammetry (CV) experiments were performed to measure the reduction potential of these substrates 1d, 1F, 1G, and 1H (Fig. 5a). The half peak potential of redox active oxalate 1d was determined to be E p/2 [1d/1d •-] = −1.71 V vs SCE (Saturated calomel electrode) in CH 3 CN. However, under the same measurement conditions for 1F (Boc), 1G (Ac), and 1H (CO 2 Me), no apparent anodic and cathodic current peaks could be observed in the range of −3.0 to 0 V, suggesting that these were redox-inactive leaving groups. Indeed, when 1F (Boc), 1G (Ac), or 1H (CO 2 Me) were subjected to the optimal conditions, 100% of the corresponding unreacted starting materials were recovered.
Two possible reaction pathways for this transformation based on CV data were proposed as shown in Fig. 6a and Supplementary  Fig. 5. In oxidative quenching cycle ( Supplementary Fig. 5) 72 to produce LCu II CN, which would further react with TMSCN to yield LCu II (CN) 2 . Alternatively, in reductive quenching cycle (Fig. 6a)  To distinguish the two pathways, Stern-Volmer quenching experiments of fac-Ir(ppy) 3 were carried out. As shown in Fig. 5b, the excited state of the photocatalyst fac-Ir(ppy) 3 was efficiently quenched by the CuBr/L5 catalyst. Furthermore, if the cyanation of 1d would be realized via oxidative quenching cycle, considering the redox-potential window of typical photocatalysis (Ir/Ru/ organic-PC etc.) 77 , the Ph-PTZ was selected as another potential photocatalyst for this transformation. The reaction in the presence of photocatalyst Ph-PTZ instead of fac-Ir(ppy) 3 would provide readily radical 10 or 11, the subsequent SET process between the oxidized state of Ph-PTZ + (E 1/2 [Ph-PTZ + /Ph-PTZ] = +0.815 V vs SCE in CH 3 CN) 61 and LCu I CN (E p/2 red [Cu II /Cu I ] = +0.15 V vs SCE in CH 3 CN) would form LCu II CN, which could yield 2d. However, such a reaction only afforded 10% of 2d with 90% of 1d being recovered (Fig. 5c). When 2 equiv of TEMPO were used as the radical trapping agent in the reaction of 1d, the formation of 2d was obviously reduced (16% vs 87%), and the TEMPO-trapped product 14 and/or 15 could be detected by LC-HRMS analysis, which supports the involvement of radical intermediates in the current transformation (Fig. 5d). Furthermore, in order to check the possible triplet energy transfer mechanism, other ruthenium-or iridium-based dyes or organic photocatalysts were tested under standard conditions (for details on photocatalyst screening, see the Supplementary Information): Photocatalysts (Ir(dtbbpy)(ppy) 2 PF 6 , E T = 49.2 kcal/mol and Ir[dF(CF 3 )ppy] 2 (dtbbpy)PF 6 , E T = 60.8 kcal/mol) with its triplet energy similar to that of fac-Ir(ppy) 3 (E T = 57.8 kcal/mol) did not provide 2d at all (Fig. 5e) 78,79 .
To further elucidate the reaction mechanism, density functional theory (DFT) calculations were preformed to survey the reaction of 1B using ligand L4 (For details on DFT calculations, see the Supplementary Information and Supplementary Data 1). As proposed by Fig. 6a, radical intermediate Int1 could be formed from oxalate 1B. Mulliken atomic spin density analysis of Int1 suggests that the single electron distributes on C 1 and C 2 with a similar spin density (0.64 and 0.47, Fig. 6b), indicating Int1 is a combination of resonance forms of allenyl radical and propargylic radical. As an allenyl radical, Int1 reacts with L4Cu II (CN) 2 via a singlet diradical transition structure TS1_a with a free energy barrier of 10.1 kcal/mol, providing a closed-shell propargyl-Cu(III) complex Int2_a reversibly. Subsequent reductive elimination produces the final allenenitrile product 2B with a very low barrier of 1.0 kcal/mol (TS2_a). Furthermore, the concerted radical cyanation process is also investigated. A triplet transition structure TS_a was obtained with a much higher free energy barrier of 30.3 kcal/mol, which indicates that the stepwise pathway via a Cu(III) intermediate is more favorable. On the other hand, the possibility of Int1 acting as a propargyl radical has also been considered. A similar oxidation/reductive elimination process is obtained, but more energy demanding, due to the steric effect caused by the cyclohexyl group with the ligand. Thus, allenenitriles 2B were generated as the only products.
These above results definitely confirmed that the reductive quenching cycle in Fig. 6a was the dominant pathway in the current transformation, which is different from the wellestablished oxidative quenching mechanism 61,75,76 .
In conclusion, we have developed a general and efficient method for the highly selective synthesis of di-, tri-, and tetra-  indicated that propargylic radical and allenyl radical generated via light-induced one-electron process were involved via the reductive quenching cycle. This protocol for allenenitrile syntheses involving one-electron mechanistic pathway is very different from the traditional transition metal-catalyzed twoelectron coupling reactions and will surely overcome the scope limitation of the known protocols and enjoy scopes for the efficient syntheses of differently functionalized allenes due to the powerful catalytic activity of copper 80,81 . Further studies on highly selective allene synthesis via such one-electron process and other photocatalysts are being actively pursued in this laboratory.

Methods
General procedure for the copper-catalyzed cyanation of propargylic oxalates.

Data availability
The X-ray crystallographic coordinates for structures of 2w' and endo-5 reported in this study have been deposited in the Cambridge Crystallographic Data Centre (CCDC) under deposition numbers CCDC 2047907 (2w'), and CCDC-2047908 (endo-5). These data can be obtained free of charge from http://www.ccdc.cam.ac.uk/data_request/cif. The experimental procedures and characterization of the new compounds in this study are provided in the Supplementary Information. All other data are available from the authors upon request.
Received: 27 October 2021; Accepted: 6 May 2022; Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/ licenses/by/4.0/.