Document vVbN1OKa4ZqXr9r1B5EnJZvM8

GDCh Trifluoromethoxylation Hot Paper Communications Angewa InternationalEdition Chet= International Edition: DOI : 10.1002/a nie.201806296 German Edition: DOI: 10.1002/a nge.201806296 Check for updates Radical Trifluoromethoxylation of Arenes Triggered by a VisibleLight-Mediated N--O Bond Redox Fragmentation Benson J. Jelier+, Pascal F Tripet+, Ewa Pietrasiak+, Ivan Franzoni, Gunnar Jeschke, and Antonio Togni* Abstract: A simple trifluoromethoxylation method enables non-directed functionalization of C--H bonds on a range of substrates, providing access to aryl trifluoromethyl ethers. This light-driven process is distinctly different from conventional procedures and occurs through an OCF3 radical mechanism mediated by a photoredox catalyst, which triggers an N--O bond fragmentation. The pyridinium-based trifluoromethoxylation reagent is bench-stable and provides access to synthetic diversity in lead compounds in an operationally simple manner. Late-stage C--H diversification can be used to elaborate molecular complexity without extensive redesign of synthetic routes' However, this strategy relies on the development of functional group transfer reagents with sufficient reactivity to enforce direct transfer of the modality while delicately balancing chemoselectivity. The taming of highly reactive oxygen, nitrogen, and fluorine species into bench-stable reagents suitable for the modification of complex carbon frameworks thus remains a key synthetic challenge?' The inclusion of oxygen-based trifluoromethoxyl fragments into arenes has been a particularly challenging synthetic endeavor, and even though its prominence in biologically active molecules and material applications is well established, a paucity of operationally simple methods persists, especially for direct C--H substitution?' Progress has largely been impeded by the low reactivity of anionic OCF3, the propensity for (3-fluoride elimination in classic transitionmetal-mediated cross-couplings, and the ensuing complicated procedures for the generation of C(sp2)--OCF3 or O--CF3 ether bonds.[4] Thus the synthesis of complex molecules containing an OCF3 substituent has been relegated to commercially available but costly building blocks in unidirectional, multistep syntheses?'51 Although substantial advances were made by Ritter in his seminal paper in 2011,[6] which have since been complemented by Ngai,[7] Qing,[8] Tang,N and [*] Dr. B. J. Jelier,I.1 P. F. Tripet,H E. Pietrasiak,H Prof. G. Jeschke, Prof. A. Togni Department of Chemistry and Applied Biosciences Swiss Federal Institute of Technology Vladimir-Prelog-Weg 2, 8093 Zurich (Switzerland) E-mail: @inorg.chem.ethz.ch Dr. I. Franzoni Department of Chemistry, University of Toronto Toronto, M5S 3H6 (Canada) [] These authors contributed equally to this work. Supporting information and the ORCID identification number(s) for 0 the author(s) of this article can be found under: https://doi.org/10.1002/anie.201806296. Liu,[1] no practical method has been developed to access this privileged class of ethers (Scheme 1). Herein, we report an operationally simple method for the trifluoromethoxylation of arenes that is enabled by photoredox catalysis. Togni (2008) -- Direct, Electrophilic Radical O-Trifluoromethylation of Phenols (ON F3C--I-0 easily oxidized CF3 predominately C-trifluoromethylation [CF3] Ritter (2011) - Silver-Mediated Trifluoromethoxylation with Anionic OCF3 x [OCF3], [Ft], AgPF6 base, -30 C, THF/acetone = SnBu3 or B(OH)2 OCF3 Known Radical Precursors: F3CO-OCF3 O F2 + F ji,F Arguello (2010) highly reactive F-OCF3 limited utility specialized handling toxic gases Navanini (2012) Ngai (2018) -- Direct Homolysis of a Neutral N-OCF3 Benzimidazole NO2 ( OCCFF3, (O CF3 N F3C N 3 step synthesis CF3 UV Homolysis (402 nm) D. = 53.1 kcal.morl I .NR2 + OCF3 intrinsic competing N radical This work -- SET induced N--O Bond Fragmentation of a Cationic Pyridinium OCF3 1 step synthesis N* bench-stable catalytic operationally simple CN r SET blue light, [Ru] = -12.5 kcal.mol' 0CF3 CN selective 0 radical Scheme I. Advances in late-stage arene trifluoromethoxylation. Our research group has had a longstanding interest in trifluoromethylation, and while we have been able to trifluoromethylate a wide range of N, O, and S nucleophiles with our hypervalent iodine reagents,[11] the trifluoromethylation of phenols represents a formidable challenge owing to the innate C character of the phenoxyl radical (Scheme 1).[12] Hence, rather than forging an O--CF3 bond, we sought a methodology that was distinct from the conventional approach. Intrigued by the reactivity of an OCF3 radical, we hypothesized that a well-designed reagent bearing a redoxactive N--O bond could liberate the elusive OCF3 radical. This inspired us to seek direct access to a precursor capable of liberating an OCF3 radical that would not impose as daunting a synthetic challenge as CF3O--FE13a1 or CF3O-OCF3,[13b,c] 13784 Wiley Online Library 2O18 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2018, 57, 13784-13789 Communications Angewandte Chemie which are known to generate the desired radical but only serve limited value as reagents (Scheme 1). Towards this end, in 2013, we reported a general method for the trifluoromethylation of oxygen centers such as hydroxylamines that gives access to putative reagents such as N-trifluoromethoxyphthalimide (1).[14] This method is simpler to carry out than existing trifluoromethylation strategies and has been adopted, for example, by Ngai and co-workers for the trifluoromethylation of N-hydroxyanilines.[7,15] In this regard, oxyphthalimide 1 seemed particularly predestined for redox-neutral fragmentation of the NO bond, akin to a number of reports for the scission of NOAlkyl[16] and NSCF3[17] bonds. A very recent publication from Ngai and co-workers invoking a similar hypothesis prompted us to disclose our own results in this area.[18] Phthalimide 1 and derivatives thereof were prepared in good yields using our previously reported method.[14] Similarly, the N-trifluoromethoxysaccharin analogue 2 was obtained through an eight-step synthesis. However, once subjected to a SET process driven by a photocatalyst, both reagent scaffolds disappointingly yielded the amino radical in excellent yields in a process reminiscent of Sanfords amination (Scheme 2).[19] We rationalized that the propensity of the NO bond fragmentation to give an N-centered radical could be shifted to predominately afford the O-centered radical through a key change in design. Inspired by Stephensons decarboxylative trifluoromethylations[20] and Goulds and Kochis extensive investigations into reductive NO bond fragmentation of N-alkoxy heterocycles, we replaced the neutral amino moiety with a cationic ammonium group.[21] optimized oxidation/trifluoromethylation sequence provided rapid access to a large library of N-trifluoromethoxypyridinium derivatives (3 a-m) in moderate to good yields upon isolation, using simple conditions with commercially available reagents (Scheme 2).[22] We found that in particular 3 a-c enabled the trifluoromethoxylation of benzene in the presence of a simple photocatalyst within less than 30 min of blue light irradiation in high-intensity visible-light photoreactors. The chloro derivatives (3 b, 3 c) were more challenging to prepare and were sensitive to heat and moisture, precluding their use as robust reagents. Conversely, 3 a can be prepared on multigram scale as a free-flowing white powder; it is thermally stable up to 240 8C in the solid state, can be manipulated entirely outside of a glovebox, and stored under ambient conditions for up to one year without appreciable decomposition (see the Supporting Information). Reagent 3 a and the 4-phenyl analogue 3 k were structurally characterized, and the connectivity about the NOCF3 bond was verified, interestingly exhibiting a sharply acute OCF3 bond angle of 112.58 with respect to the pyridinium core (Scheme 3). Cyclic voltammetry measurements of selected reagents under stan- dard conditions revealed a cathodic wave at a negatively shifted potential of 1.2 V to 0.19 V vs. Fc0/+ (Scheme 2).[23] The absence of the coupled anodic wave on the return scan, even at sweep rates exceeding v = 100 V s1, indicated that the one-electron reduction of the pyridinium cations was chemically irreversible and rate-limiting, providing an impetus for undertaking extensive screens with this class of reagents. A thorough evaluation of reaction conditions led us to identify Scheme 2. Conception and development of a cationic OCF3 transfer reagent. optimal conditions for the radical trifluoromethoxylation of arenes. Reagent 3 a (1.0 equiv) reacts with benzene (10 equiv) in acetonitrile with [Ru(bpy)3](PF6)2 (5 mol %) as a competent and economically feasible photoredox catalyst to deliver the desired product PhOCF3 within 30 min in 78 % NMR yield. The OCF3 radical would be considered as highly electrophilic[24] and deactivating based on the purported nonconjugation of the oxygen p electrons.[25] Nevertheless, overtrifluoromethoxylation of the substrate was evident, neces- Angew. Chem. Int. Ed. 2018, 57, 13784 -13789 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org 13785 Communications Table 1: Representative scope of the arene C(sp2)H trifluoromethoxylation.[a] Angewandte Chemie [a] Arene (5.0 equiv), [Ru] = [Ru(bpy)3](PF6)2 (5 mol %), and reagent 3 a (1.00 equiv) were irradiated in MeCN in a 350 W blue LED reactor at ambient temperature for 1 h. Major isomer shown (o/m/p). Yields determined by 19F NMR analysis with an optimized D1 delay and an internal standard. Yields of isolated products given in parentheses. New compounds, including all regioisomers, were isolated by preparatory HPLC in > 95 % purity. sitating the addition of additional equivalents of substrate to prevent the generation of complex mixtures and yield attritions. With optimized reaction conditions in hand, we sought to explore the scope of the C(sp2)H trifluoromethoxylation process. As shown in Table 1, a variety of arene building blocks decorated with common functional groups, including halides, nitriles, ketones, amides, acids, sulfonamides, imides, esters, boronic esters, and phosphonates, are well tolerated under the reaction conditions but the corresponding products were isolated in moderate yields and low selectivity. Notably, benzylic moieties and aldehydes were tolerated whereas substrates bearing unprotected amines or 13786 www.angewandte.org 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2018, 57, 13784 -13789 Communications Angewandte Chemie Scheme 3. Early mechanistic considerations for the radical trifluoromethoxylation of arenes by photoredox catalysis. allylic positions preferentially underwent dimerization or direct oxidation when in direct contact with reagent 3 a. As for aryl fluorides, separation of individual aryl trifluoromethyl ether regioisomers from each other as well as the starting material can be an imposing task.[26] Nevertheless, we developed procedures for the isolation of each regioisomer of nearly every new compound in high purity through careful flash chromatography and preparative HPLC purifications with the exception of highly volatile products for which authentic samples were available (see the Supporting Information). Importantly, unreacted equivalents of starting materials could typically be recovered. To demonstrate the synthetic utility, we used our general conditions for the latestage functionalization of biorelevant molecules, including the breast cancer drug Femara (5 aj) and the acylalanine fungicide Metalaxyl (5 al), which is used in crop protection. The latter gave rise to a diastereomeric pair of meta- atropisomers, which were cleanly separated and isolated. Additionally, we functionalized phenytoin (5 ak), an essential WHO medication for the treatment of seizures, and Procymidone (5 am), a fruit fungicide, which would be arduous to prepare with traditional methods. A preliminary mechanistic proposal for the catalytic radical trifluoromethoxylation of arenes through an aromatic homolytic substitution mechanism is outlined in Scheme 3, and is supported by photo-induced EPR spectroscopy, control experiments, UV/Vis spectroscopy, kinetic isotope experiments, single-crystal X-ray diffraction (SC-XRD), and DFT calculations.[27,28] We surmised that the process would begin with the excitation of the photocatalyst under blue light irradiation.[29] Thus, in the key step, the excited Ru(bpy)32+* (Ered = 1.2 V vs. Fc0/+) undergoes oxidative quenching to afford Ru(bpy)33+.[29b] Upon SET reduction of the pyridinium cation (Ered = 0.24 V vs. Fc0/+), the ensuing fragmentation Angew. Chem. Int. Ed. 2018, 57, 13784 -13789 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org 13787 Communications Angewandte Chemie process could afford either an N- or an O-centered radical.[21d] Based on DFT calculations, the 4-cyanopyridinium radical favors the formation of the desired trifluoromethoxyl radical; this process is driven by the elimination of a thermodynamically favored neutral pyridine over a pyridinium radical cation. Upon addition of the trifluoromethoxyl radical to the arene, the catalytic cycle is completed by oxidation and deprotonation of the cyclohexadienyl radical by the RuIII photocatalyst and 4-cyanopyridine, respectively. The CH/D bond cleavage is likely not to be involved in the rate-limiting step as evidenced by a kH/kD KIE competition experiment (1.0:1.0). Lastly, in the presence of adventitious water, reagent 3 a can decompose to the corresponding pyridone, which was confirmed by NMR and SC-XRD analysis (see the Supporting Information). The occurring NO cleavage pathway (see above) distinguishes our process from a recent report by Ngai and co-workers, who employed high-energy near-UV light for direct NO homolysis of their neutral benzimidazole reagent.[18] When we conducted a similar UV homolysis experiment (DGNO = 53.2 kcal mol1) with radiation of much higher energy, we observed an inferior yield of 30 %. However, even with an NO bifurcation barrier computed to be D(DG) = 3.1 kcal mol1, we expected that the efficiency of our reaction could be diminished by the competing background reaction of arene pyridination under standard photoredox conditions (Scheme 3). Thus, even under our optimized conditions, reagent 3 a produces about 15 % of the undesired N-phenyl-4-cyanopyridinium cation.[30] We also found that the release of a highly oxidizing trifluoromethoxyl radical can further be quenched by a number of species, including ground-state or excited [Ru] in various oxidation states, or by sacrificial oxidation of the cyclohexadienyl radical (see the Supporting Information).[31] To support the radical nature of the reaction, we carried out a radical scavenging experiment with 1,4-cyclohexadiene,[32] light-dark experiments,[28d] and extensive EPR analysis, which all indicated a radical process (see the Supporting Information). Whereas the common nitrone spin traps DMPO and PBN[33] were unsuccessful in trapping the highly reactive oxygen radical, we obtained the first preliminary evidence for a spin-trapped trifluoromethoxyl radical by advanced-pulse EPR measurements from direct excitation of reagent 3 a in the presence of POBN at 440 nm at 193 K in a MeCN matrix; this process ostensibly proceeds through an intermolecular charge-transfer photoexcitation (Scheme 3). In summary, we have designed highly electrophilic NO pyridinium reagents capable of undergoing facile singleelectron reductions to afford a trifluoromethoxyl radical under catalytic conditions. While this initial proof-of-principle study has provided a method suitable for late-stage diversification, this class of pyridinium reagents has the potential to expedite access to trifluoromethyl aryl ethers as oxidants for directed, transition-metal-mediated routes to solve regioselectivity. We envision that these reagents will spark further interest in NX bond homolysis as a general route for the transfer of other emergent moieties. Acknowledgements This research was supported by ETH Zrich, NSERC (postdoctoral fellowship to B.J.J.), and by kind gifts from Syngenta Crop Protection AG (Stein, Switzerland). I.F. thanks the Collaborative Research and Training Experience program for a postdoctoral fellowship. We thank Dr. A. Poblador-Bahamonde and Carmine Chiancone (both U. de Genve) for access to their computational facilities. We became aware that the laboratory of M.-Y. Ngai (State University of New York at Stony Brook) was engaged in related studies towards radical trifluoromethoxylation. We are grateful to Prof. M.-Y. Ngai for collegial discussions during the preparation of our manuscripts. Conflict of interest The authors declare no conflict of interest. Keywords: late-stage functionalization organofluorine compounds photoredox radical mechanisms trifluoromethoxylation How to cite: Angew. Chem. Int. Ed. 2018, 57, 13784 - 13789 Angew. Chem. 2018, 130, 13980 - 13985 [1] Selected articles on late-stage functionalization: a) T. Cernak, K. Dykstra, S. Tyagarajan, P. Vachal, S. Krska, Chem. Soc. Rev. 2016, 45, 546; b) W. Lu, L. Zhou, Oxidation of CH Bonds, Wiley, Hoboken, 2017; c) T. Gensch, M. N. Hopkinson, F. Glorius, J. Wencel-Delord, Chem. Soc. Rev. 2016, 45, 2900; d) J. Douglas, M. Sevrin, C. Stephenson, Org. Process Res. Dev. 2016, 20, 1134; e) M. Yan, J. Lo, J. Edwards, P. Baran, J. Am. Chem. Soc. 2016, 138, 12692. [2] a) K. Mller, C. Faeh, F. Diederich, Science 2007, 317, 1881; b) E. Gillis, K. Eastman, M. Hill, D. Donnelly, N. Meanwell, J. Med. Chem. 2015, 58, 8315; c) H. Davies, D. Morton, ACS Cent. Sci. 2017, 3, 936. [3] a) P. Jeschke, E. Baston, F. Leroux, Mini-Rev. Med. Chem. 2007, 7, 1027; b) T. Smith, X. Yang, H. Wu, B. Pouw, R. Matsumoto, A. Coop, J. Med. Chem. 2008, 51, 3322; c) O. Marrec, T. Billard, J. Vors, S. Pazenok, B. Langlois, J. Fluorine Chem. 2010, 131, 200; d) A. Tlili, F. Toulgoat, T. Billard, Angew. Chem. Int. Ed. 2016, 55, 11726; Angew. Chem. 2016, 128, 11900. [4] a) Synthesis of ArOCF3 by decarboxylative fluorination: M. Zhou, C. Ni, Z. He, J. Hu, Org. Lett. 2016, 18, 3754; b) direct trifluoromethylation of alcohols with unstable CF3-oxonium salts: T. Umemoto, K. Adachi, S. Ishihara, J. Org. Chem. 2007, 72, 6905; c) B. Jelier, J. Howell, C. Montgomery, D. Leznoff, C. Friesen, Angew. Chem. Int. Ed. 2015, 54, 2945; Angew. Chem. 2015, 127, 2988. [5] a) F. Leroux, P. Jeschke, M. Schlosser, Chem. Rev. 2005, 105, 827; b) F. Leroux, B. Manteau, J. Vors, S. Pazenok, Beilstein J. Org. Chem. 2008, 4, 13; c) T. Besset, P. Jubault, X. Pannecoucke, T. Poisson, Org. Chem. Front. 2016, 3, 1004. [6] C. Huang, T. Liang, S. Harada, E. Lee, T. Ritter, J. Am. Chem. Soc. 2011, 133, 13308. [7] a) K. Hojczyk, P. Feng, C. Zhan, M. Ngai, Angew. Chem. Int. Ed. 2014, 53, 14559; Angew. Chem. 2014, 126, 14787; b) K. Lee, J. Lee, M. Ngai, Synlett 2016, 27, 313; c) K. Lee, Z. Lei, C. MoralesRivera, P. Liu, M. Ngai, Org. Biomol. Chem. 2016, 14, 5599. [8] For a silver-mediated oxidative trifluoromethylation of phenols, see: a) J. Liu, C. Chen, L. Chu, Z. Chen, X. H. Xu, F. Qing, 13788 www.angewandte.org 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2018, 57, 13784 -13789 Communications Angewandte Chemie Angew. Chem. Int. Ed. 2015, 54, 11839; Angew. Chem. 2015, 127, 12005; b) J. Liu, X. Xu, F. Qing, Org. Lett. 2015, 17, 5048. [9] S. Guo, F. Cong, R. Guo, L. Wang, P. Tang, Nat. Chem. 2017, 9, 546. [10] For a Pd-catalyzed trifluoromethoxylation limited to sp3 centers, see: a) C. Chen, P. Chen, G. Liu, J. Am. Chem. Soc. 2015, 137, 15648; b) X. Qi, P. Chen, G. Liu, Angew. Chem. Int. Ed. 2017, 56, 9517; Angew. Chem. 2017, 129, 9645. [11] a) J. Charpentier, N. Frh, A. Togni, Chem. Rev. 2015, 115, 650; b) X. Yang, T. Wu, R. Phipps, F. Toste, Chem. Rev. 2015, 115, 826. [12] For trifluoromethylations of phenols with trifluoromethyl iodane transfer reagents, see: a) K. Stanek, R. Koller, A. Togni, J. Org. Chem. 2008, 73, 7678; b) G. Filippini, M. Nappi, P. Melchiorre, Tetrahedron 2015, 71, 4535. [13] a) For CF3OF: V. Francesco, M. Sansotera, W. Navarrini, J. Fluorine Chem. 2013, 155, 2; b) although commercially available at one time from Air Liquide, CF3OOCF3 has been discontinued: R. Syvret, B.-A. Camplon, G. Cooper (Air Products and Chemicals, Inc), EP 1757581A1, 2007; c) W. J. Pelez, G. A. Argello, Tetrahedron Lett. 2010, 51, 5242. [14] a) Trifluoromethylation of hydroxylamines: V. Matousek, E. Pietrasiak, L. Sigrist, B. Czarniecki, A. Togni, Eur. J. Org. Chem. 2014, 3087; b) of pyridones, see: A. Liang, S. Han, Z. Liu, L. Wang, J. Li, D. Zou, Y. Wu, Chem. Eur. J. 2016, 22, 5102. [15] P. Feng, K. N. Lee, J. W. Lee, C. Zhan, M. Y. Ngai, Chem. Sci. 2016, 7, 424. [16] a) M. Zlotorzynska, G. Sammis, Org. Lett. 2011, 13, 6264; b) G. Lackner, K. Quasdorf, L. Overman, J. Am. Chem. Soc. 2013, 135, 15342; c) J. Zhang, Y. Li, F. Zhang, C. Hu, Y. Chen, Angew. Chem. Int. Ed. 2016, 55, 1872; Angew. Chem. 2016, 128, 1904; d) C. Wang, K. Harms, E. Meggers, Angew. Chem. Int. Ed. 2016, 55, 13495; Angew. Chem. 2016, 128, 13693. [17] a) X. Shao, C. Xu, L. Lu, Q. Shen, Acc. Chem. Res. 2015, 48, 1227; b) S. Mukherjee, B. Maji, A. Tlahuext-Aca, F. Glorius, J. Am. Chem. Soc. 2016, 138, 16200. [18] A new N-OCF3 reagent, albeit with a distinctly different mechanism and structure, was reported during the final stages of this study; see: W. Zheng, C. Rivera, J. Lee, P. Liu, M. Ngai, Angew. Chem. Int. Ed. 2018, https://doi.org/10.1002/anie. 201800598; Angew. Chem. 2018, https://doi.org/10.1002/ange. 201800598. [19] L. Allen, P. Cabrera, M. Lee, M. Sanford, J. Am. Chem. Soc. 2014, 136, 5607. [20] a) J. Beatty, J. Douglas, K. Cole, C. Stephenson, Nat. Commun. 2015, 6, 7919; b) J. Beatty, J. Douglas, R. Miller, R. McAtee, K. Cole, C. Stephenson, Chem 2016, 1, 456. [21] a) T. Bockman, K. Lee, J. Kochi, Chem. Soc. Perkin Trans. 2 1992, 1581; b) I. Gould, D. Shukla, D. Giesen, S. Farid, Helv. Chim. Acta 2001, 84, 2796; c) E. Lorance, W. Kramer, I. Gould, J. Am. Chem. Soc. 2002, 124, 15225; d) E. Lorance, W. Kramer, I. Gould, J. Am. Chem. Soc. 2004, 126, 14071; e) E. Lorance, K. Hendrickson, I. Gould, J. Org. Chem. 2005, 70, 2014. [22] a) See the Supporting Iniformation for reaction optimization of the trifluoromethylation of pyridine N-oxides; b) unbeknownst to us, the first trifluoromethylation of pyridine N-oxides was reported in a Chinese patent and involved the use of Umemotos reagent under forcing conditions. Limited examples of the thermal oxidation of electron-rich arenes were demonstrated; see: T. Umemoto, M. Zhou, J. Hu (Faming Zhuanli Shenqing), CN, 105017143 A 20151104, 2015. [23] See the Supporting Information for reduction potentials. There are no agreed upon conventions for reporting CV values. The redox couple Fc0/+ is reproducible and has precedence in photoredox chemistry; see the following tutorial: D. AriasRotondo, J. McCusker, Chem. Soc. Rev. 2016, 45, 5803. [24] a) F. De Vleeschouwer, V. Van Speybroeck, M. Waroquier, P. Geerlings, F. De Proft, Org. Lett. 2007, 9, 2721. [25] a) W. Sheppard, J. Am. Chem. Soc. 1963, 85, 1314; b) I. Serfaty, T. Hodgins, E. McBee, J. Org. Chem. 1972, 37, 2651; c) G. Olah, T. Yamato, T. Hashimoto, J. Shih, N. Trivedi, B. Singh, M. Piteau, J. Olah, J. Am. Chem. Soc. 1987, 109, 3708; d) H. Bhm, D. Banner, S. Bendels, M. Kansy, B. Kuhn, K. Mller, U. ObstSander, M. Stahl, ChemBioChem 2004, 5, 637. [26] a) D. Nagib, D. MacMillan, Nature 2011, 480, 224; b) K. Yamamoto, J. Li, J. Garber, J. Rolfes, G. Boursalian, J. Borghs, C. Genicot, J. Jacq, M. van Gastel, F. Neese, T. Ritter, Nature 2018, 554, 511. [27] Owing to difficulty and unreliability in computing transition states for these radical species, DFT calculations were performed solely from a thermodynamics perspective, which is warranted by NO fragmentation being ostensibly a barrierless process (Refs. [21d,e]). See also the Supporting Information. [28] a) M. Tsao, C. Hadad, M. Platz, J. Am. Chem. Soc. 2003, 125, 8390; b) M. Majek, A. Jacobi von Wangelin, Acc. Chem. Res. 2016, 49, 2316; c) J. Chen, X. Hu, L. Lu, W. Xiao, Chem. Soc. Rev. 2016, 45, 2044; d) a classic light - dark experiment was performed (see the Supporting Information) but was considered to be inferior to quantum yields and luminescence quenching; see: M. Cismesia, T. Yoon, Chem. Sci. 2015, 6, 5426. [29] a) C. Prier, D. Rankic, D. MacMillan, Chem. Rev. 2013, 113, 5322; b) J. Douglas, M. Sevrin, C. Stephenson, Org. Process Res. Dev. 2016, 20, 1134; c) when the reaction mixture is irradiated with blue light, there is an immediate color change from redorange to a vibrant green, which fades back to orange. This observation is consistent with: A. Lewandowska-Andralojc, E. Polyansky, J. Phys. Chem. A 2013, 117, 10311. [30] a) B. Feng, D. Wan, L. Yan, V. D. Kadam, J. You, G. Gao, RSC Adv. 2016, 6, 66407; b) M. Krks, ACS Catal. 2017, 7, 4999. [31] See the Supporting Information for detailed DFT calculations. [32] J. Fossey, D. Lefort, J. Sorba, Free radicals in organic chemistry, Wiley, New York, 1995. [33] DMPO = 5,5-dimethyl-1-pyrroline N-oxide, PBN = N-tert-butyla-phenylnitrone, POBN = a-(4-pyridyl-1-oxide)-N-tert-butylnitrone. Manuscript received: May 31, 2018 Accepted manuscript online: June 21, 2018 Version of record online: August 7, 2018 Angew. Chem. Int. Ed. 2018, 57, 13784 -13789 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.angewandte.org 13789