ChemComm View Article Online

COMMUNICATION

Published on 14 May 2015. Downloaded by Fudan University on 14/05/2015 21:40:58.

Cite this: DOI: 10.1039/c5cc01923k Received 6th March 2015, Accepted 5th May 2015

View Journal

Copper(I)-photocatalyzed trifluoromethylation of alkenes† R. Beniazza,a F. Molton,b C. Duboc,b A. Tron,a N. D. McClenaghan,a a a D. Laste ´coue `res and J.-M. Vincent*

DOI: 10.1039/c5cc01923k www.rsc.org/chemcomm

Using the photoreducible CuII precatalyst 2, trifluoromethylation reactions of alkenes are conducted effectively at low copper loading (0.1–0.5 mol%) on exposing the reaction mixture to sunlight/ ambient light.

The development of effective/selective trifluoromethylation methodologies is currently a very active research area.1 In this domain, metal-catalyzed trifluoromethylation reactions have emerged as powerful tools, particularly using alkenes as substrates.1k,l The combination of electrophilic trifluoromethylated hypervalent iodine reagents,2 such as Togni’s reagent 1 (Scheme 1), and copper(I/II) salts/complexes has proved to be an extremely convenient source of trifluoromethylating species,3 i.e. the electrophilic trifluoromethyl radical or trifluoromethyl cation. Despite their great potential, a drawback associated with most of the copper-based processes reported so far is their relatively low catalytic efficiency, the reactions requiring 10–20 mol% loading in copper to achieve satisfactory conversions. It should also be noted that in none of these numerous reports, was light-dependent reactivity identified. At the same time, photosensitized ruthenium(II)- and iridium(III)-catalyzed trifluoromethylation of alkenes/arenes has been shown to proceed efficiently at lower catalyst loading (0.5–5 mol%).4 Here, Single Electron Transfer (SET) from the strongly reductive photoactivated Ru or Ir catalyst to various trifluoromethylating reagents occurs effectively to generate the reactive trifluoromethyl radical. An interesting example of the photocatalyzed radical hydrotrifluoromethylation of terminal alkenes/alkynes exploiting methylene blue as the photocatalyst (2 mol%), Togni’s reagent and an amine as the sacrificial electron source was recently reported.5 We recently developed photoreducible copper(II) complexes,6 in particular the CuII-DMEDA (DMEDA = N,N 0 -dimethylethylendiamine) complex 2 (Scheme 1),6c for which the photogenerated a

Institut des Sciences Mole´culaires, CNRS UMR 5255, Univ. Bordeaux, 33405 Talence, France. E-mail: [email protected] b Univ. Grenoble Alpes, DCM UMR-CNRS 5250, F-38000 Grenoble, France † Electronic supplementary information (ESI) available: Synthesis and characterization of compounds 4, 5a–e, 6a–b; ES-MS, UV-vis and EPR spectra; cyclic voltammogram of 2. See DOI: 10.1039/c5cc01923k

This journal is © The Royal Society of Chemistry 2015

Scheme 1

Structures of Togni’s reagent 1 and photoreducible CuII complex 2.

copper(I) species was shown to be a highly reactive catalyst for the CuI-catalyzed alkyne–azide cycloaddition (CuAAC) reaction. Under irradiation at 365 nm (Thin Layer Chromatography lamp), the benzophenone chromophore present in the carboxylate counterion mediates a highly efficient Photoinduced Electron Transfer (PET) process from the solvent (typically H-donating solvents such as MeOH or THF) to the CuII ions, as evidenced by the photoreduction quantum yield value of 0.22 determined in MeOH, while a value approaching unity was determined in THF. As a consequence of such an efficient PET process, the CuII to CuI reduction occurs very effectively under sunlight illumination.6c Considering the ease and efficiency of the in situ photocatalytic generation of the reducing CuI species, we applied this system to the trifluoromethylation of terminal alkenes. The screening of the reaction conditions was undertaken, employing alkene 3 as the substrate and complex 2 at a loading of 0.5 mol% (Table 1). Sample irradiation at 365 nm was carried out using a standard TLC lamp, while for daylight illumination the NMR tubes were placed behind a window under direct sunlight exposure. Under sunlight illumination the reaction catalyzed by 2 at 0.5 mol% loading proceeded to full conversion of the alkene 3 in B4 h reaction time (entry 1), thereby providing the trifluoromethylated product 4 in a satisfactory 81% 1H NMR yield (E/Z: 91/9). In sharp contrast, in the absence of light no reaction was observed and Togni’s reagent and the alkene remained intact after 13 h (entry 2). When the reaction was conducted under ambient light (behind a window on a cloudy day) the reaction proceeds well, albeit at a slower rate (entry 3). Under TLC lamp irradiation at 365 nm, the reaction is fast, and full alkene conversion was reached in B1 h. However, analysis of the 1 H and 19F NMR spectra (see ESI†) reveals that more by-products

Chem. Commun.

View Article Online

Communication

ChemComm

Published on 14 May 2015. Downloaded by Fudan University on 14/05/2015 21:40:58.

Table 1 Screening of the reaction conditions for the copper(I)-photocatalyzed trifluoromethylation of alkenes

Table 2

Light sourceb

Time

Yieldc [%]

1 2

TLC Sun

70 min 7h

85 75

3 4

TLC Sun

1h 6h

70 84

5 6

TLC Sun

1h 6h

70 76

7 8

TLC Sun

70 min 5h

75 93

9 10

TLC Sun

2h 8h

49 50

11d

TLC

2h

73

12d

TLC

2h

80

Entry

Cat.

Solvent

Light source

Timea

Yieldb [%]

Entry

1 2c 3d 4 5e 6f 7 8

2 2 2 2 — 2 2 2

CD3OD CD3OD CD3OD CD3OD CD3OD CD3OD THF-D8 CD3CN

Sun — Ambient TLC Sun Sun TLC TLC

4 h (495%) 13 h (o5%) 22 h (90%) 1 h (490%) 4 h (0%) 6 h (34%) 2 h (43%) 2 h (24%)

81 0 68 58 0 9 o5 o5

a

Time at which the conversion of 3, given in parenthesis, was attained (assessed by 1H NMR). Reactions conducted at 25 1C in CD3OD (0.5 mL) in NMR tubes with 0.1 mmol of alkene, 0.12 mmol of 1 and 2 (0.5 mol%). The solutions were deaerated by gentle Ar bubbling for 20 min keeping the tube in the dark (aluminium foil). b Yield of 4 determined by 1 H NMR spectroscopy using the peak of CHD2OD as an internal standard (see ESI). c Reaction in the dark (NMR tube in aluminium foil). d NMR tube placed behind a window of a cloudy day. e No catalyst. f Air-equilibrated reaction mixture.

are formed, leading to a significant decrease in the yield of 4 (entry 4). In the absence of catalyst but under sunlight illumination, no reaction occurs (entry 5), and the alkene and Togni’s reagent rest intact. When the reaction mixture is not dexoygenated, a slower conversion is observed (entry 6). This is consistent with the involvement of CuI species in the catalytic cycle. Indeed, we previously showed that photoreduced 2 is extremely reactive towards O2. It could be assumed that the reduction of O2 by CuI competes with the reduction of an activated form of Togni’s reagent, as proposed in the reaction mechanism (vide infra). Moreover, the reaction is less selective, most probably due to the generation of reactive dioxygen species. In THF-D8 and CD3CN the reactions were sluggish (entries 7 and 8). The low photoreduction quantum yield measured for 2 in CH3CN (0.03) compared to the much better hydrogen-donating CH3OH (B0.30), and possible side-reactions of 1 with THF,7 might partly account for such poor reactivity. In agreement with a photocatalyzed process, the reaction can be switched ON/OFF on-demand by alternating light/dark cycles. The reaction can be instantaneously and completely stopped when protecting the NMR tube from light. The reaction can be subsequently reinitiated when re-exposed to sunlight and then proceeds with similar rates. Overall, these results show that light is required, not only to generate the active CuI species when using 2, but also to allow the reaction to proceed, most probably to favour the oneelectron reduction of Togni’s reagent (vide infra). The activity of 2 was then tested on a range of alkenes (Table 2). Reactions proceeded well at low catalyst loading (0.5 mol%), with reaction times ranging from 5 to 8 h under sunlight illumination, and 1 to 2 h when irradiated with a TLC lamp. As previously shown by Buchwald and coworkers,3h heterocyclic trifluoromethylated compounds 6a–b (entries 11 and 12) can be obtained in satisfactory yields, ensuring that heating is provided to facilitate the intramolecular nucleophilic cyclization.

Chem. Commun.

CuI-photocatalyzed trifluoromethylation of alkenes

Alkenea

Product (E/Z/other isomers)

a Reactions conducted at 25 1C in CD3OD (0.5 mL) in NMR tubes with 0.1 mmol of alkene, 0.12 mmol of 1 and 2 (0.5 mol%). The solutions were deaerated by gentle Ar bubbling for 20 min keeping the tube in the dark (aluminium foil). b Tube placed behind a window under direct sunlight illumination, or irradiated at 365 nm with a TLC lamp (6 W) placed at B1 cm. c Determined by 1H NMR spectroscopy using the peak of CHD2OD as an internal standard (see ESI). d Reactions conducted at 70 1C (water bath) under TLC irradiation.

Scheme 2

Trifluoromethylations conducted on a preparative scale.

Reactions can be effectively conducted in round-bottom flasks on a preparative scale (Scheme 2). Gratifyingly, at 0.1 mol% loading of 2, the compounds 4 and 5d could be obtained in 81% and 76% yields, by simply exposing the flask to light for 14 to 16 h. This corresponds to Turnover Numbers of 810 and 760, respectively. To gain an insight into the possible role of light in the reduction of Togni’s reagent, the reactivity of 2 towards 1 was studied under various illumination conditions. When a catalytic amount of 2 (0.5 mol%) was reacted with 1 in CD3OD, a lightdependent decomposition of 1 into the 2-iodobenzoic acid is observed, the full decomposition requiring B3 h under sunlight irradiation (see ESI†). Concomitantly, CF3D and CF3I were formed in 44%,8 and 8% 19F NMR yields, respectively (see ESI†).

This journal is © The Royal Society of Chemistry 2015

View Article Online

ChemComm

Published on 14 May 2015. Downloaded by Fudan University on 14/05/2015 21:40:58.

Scheme 3 TEMPO.

Communication

Reactivity of 2 and/or Togni’s reagent in the presence of

In the absence of 2, no decomposition of 1 occurs. Thus 2, in the presence of light, is able to very effectively reduce Togni’s reagent photocatalytically, most probably to generate  CF3, which then abstracts H or D from the solvent. The radical nature of the mechanism and the implication of  CF3 is further supported by the experiments represented in Scheme 3. The addition of TEMPO (1.5 equiv.) to a reaction that was run in the conditions of entry 1 in Table 1 fully inhibits the formation of 4, while TEMPO-CF3 is formed in 72% NMR yield (see ESI†). In the absence of alkene the TEMPO-CF3 was generated with a similar yield (71%) and at a similar rate, while no TEMPO-CF3 is formed in the dark. It should be noted that in the absence of 2, 12% of TEMPO-CF3 is still formed, showing that under light illumination the TEMPO reacts directly with 1 to a significant extent. Interestingly, conclusive evidence that 2 rapidly reacts with 1 in its ground-state, i.e. in a light-independent process, was provided by spectroscopic studies. When an air-equilibrated MeOH solution of 2 and 1 (4 equiv.) was left in the dark, a new mononuclear copper(II) species forms rapidly, the complete disappearance of the characteristic signal of 2 occurring in B15 min, while a new signal appears (Fig. 1). This species is transitory, its EPR signal evolving slowly into a third-one, i.e. within B5 h (in the dark). UV-vis studies confirmed this behavior, and the characteristic d–d absorption band of 2 at 638 nm shifted to 732 nm (see ESI†). We hypothesized that the fast change of the EPR signal could be ascribed to the coordination of the carbonyl group of the Togni’s reagent to the CuII, or to the coordination of the carboxylate group of the carboxylato trifluoromethyl iodonium, which may form. To test these hypotheses the following experiments were conducted. When adding an excess (20 equiv.) of ethyl benzoate or benzoate sodium salt onto a solution of 2 the EPR signal remained unchanged (see ESI†). We surmised that the carboxylate of the carboxylato trifluoromethyl iodonium forming in situ would exhibit a ‘‘naked character’’ which would enhance its binding properties. When combining the benzoate sodium salt with 15-crown-5 ether in a 1 : 1 ratio to increase the carboxylate binding properties, the spectrum corresponding to the species forming rapidly in the presence of 1 (red spectrum in Fig. 1) is generated in B15% yield (see ESI†). Interestingly, this yield increases to 50% when the iodobenzoate sodium salt with 15-crown-5 ether is used (see ESI†). Finally, a peak at m/z = 692.0 assigned to a [CuII-DMEDAcarboxylato trifluoromethyl iodonium-3-benzoylbenzoate]+ ion was evidenced in the ESI-MS spectrum after only 2 min of reaction of 2 with Togni’s reagent (see ESI†). Overall, these data

This journal is © The Royal Society of Chemistry 2015

Fig. 1 EPR (X-band) spectroscopic monitoring of the reaction of 2 with 1 in the dark. The spectra were recorded at 100 K. Black spectrum: air-equilibrated MeOH solutions (0.3 mL) of 2 (1 mM). Red, blue and green spectra: airequilibrated MeOH solution (0.3 mL) of 2 (1 mM) and 1 (4 mM) left in the dark for 15, 60 and 300 min, respectively.

strongly suggest that in the reaction conditions and before irradiation, the precatalyst 2 is rapidly transformed into a CuII carboxylato trifluoromethyl iodonium intermediate. Based on these preliminary results, the reaction mechanism depicted in Scheme 4 is proposed. First, it should be noted that a direct reduction of 1 in its ground state by reduced 2 is thermodynamically unfavorable, the redox potentials of the CuII/CuI couple in 2 (E01/2 = 0.30 V vs. Ag/Ag+ in MeOH, the solvent employed for catalytic experiments in this work, see ESI†) being significantly higher than the reported value for the 1/1  couple (E01/2 = 1.1 V vs. Ag/Ag+ in CH3CN).9 The CuI intermediate B would be generated by photoreduction of a CuII carboxylato trifluoromethyl iodonium intermediate A. Photoactivation of intermediate B would then mediate a SET which produces the trifluoromethyl radical which rapidly reacts with the alkene. According to extensive work by Kochi on the oxidation of primary/secondary alkyl radical by Cu(OAc)2,10 we propose that the CuII-carboxylate species C would react with the trifluoromethylated alkyl radical to generate

Scheme 4 Tentative mechanism for the photoredox trifluoromethylation of alkenes from Togni’s reagent and precatalyst 2 (L and X denote the DMEDA ligand and benzoylbenzoate counterion, respectively).

Chem. Commun.

View Article Online

Published on 14 May 2015. Downloaded by Fudan University on 14/05/2015 21:40:58.

Communication

the alkylcopper(III) intermediate D,11 which then undergoes b-hydride elimination to afford the E isomer of the allylic trifluoromethylated alkene. Interestingly, previous DFT calculations conducted from a cupric species such as D validated a base-assisted elimination process to generate the allylic trifluoromethylated product.12 Finally, preliminary results suggest that the generated Cu(I) species could react with 1 to regenerate intermediate B.13 The Agence Nationale de la Recherche (ANR-13-BS07-0006-01, ´gion ‘‘PET-Cat’’), the University of Bordeaux, the CNRS, the Re ´seau National de RPE interdisciplinaire Aquitaine, the TGE Re (FR-CNRS 3443) and the European Research Council (FP7/20082013) ERC grant agreement no. 208702, are gratefully acknowledged for their financial support.

ChemComm

4

Notes and references 1 Selected reviews: (a) T. Umemoto, Chem. Rev., 1996, 96, 1757; (b) G. K. S. Prakash and A. K. Yudin, Chem. Rev., 1997, 97, 757; (c) T. Billard and B. R. Langlois, Eur. J. Org. Chem., 2007, 891; (d) J. Nie, H.-C. Guo, D. Cahard and J.-A. Ma, Chem. Rev., 2011, 111, 455; (e) O. A. Tomashenko and V. V. Grushin, Chem. Rev., 2011, 111, 4475; ( f ) A. Studer, Angew. Chem., Int. Ed., 2012, 51, 8950; ( g) Y. Ye and M. S. Sanford, Synlett, 2012, 2005; (h) T. Liu and ´ and E. Magnier, Q. Shen, Eur. J. Org. Chem., 2012, 6679; (i) Y. Mace Eur. J. Org. Chem., 2012, 2479; ( j) S. Barata-Vallejo and A. Postigo, Coord. Chem. Rev., 2013, 257, 3051; (k) H. Egami and M. Sodeoka, Angew. Chem., Int. Ed., 2014, 53, 8294; (l ) E. Merino and C. Nevado, ˜ o and Chem. Soc. Rev., 2014, 43, 6598; (m) S. Barata-Vallejo, B. Lantn A. Postigo, Chem. – Eur. J., 2014, 20, 16806. ¨h and A. Togni, Chem. Rev., 2015, 115, 650. 2 J. Charpentier, N. Fru 3 Selected examples: (a) A. T. Parsons and S. L. Buchwald, Angew. Chem., Int. Ed., 2011, 50, 9120; (b) X. Wang, Y. Ye, S. Zhang, J. Feng, Y. Xu, Y. Zhang and J. Wang, J. Am. Chem. Soc., 2011, 133, 16410; (c) R. Shimizu, H. Egami, Y. Hamashima and M. Sodeoka, Angew. Chem., Int. Ed., 2012, ´pez, K. M. Engle, S. Verhoog, 51, 4577; (d) S. Mizuta, O. Galicia-Lo K. Wheelhouse, G. Rassias and V. Gouverneur, Chem. – Eur. J., 2012, 18, 8583; (e) H. Egami, R. Shimizu and M. Sodeoka, Tetrahedron Lett., 2012, 53, 5503; ( f ) C. Feng and T.-P. Loh, Chem. Sci., 2012, 3, 3458; ´, Org. Lett., (g) P. G. Janson, I. Ghoneim, N. O. Ilchenko and K. J. Szabo 2012, 14, 2882; (h) R. Zhu and S. L. Buchwald, J. Am. Chem. Soc., 2012, 134, 12462; (i) C. Feng and T.-P. Loh, Angew. Chem., Int. Ed., 2013, 52, 12414; ( j) X.-P. Wang, J.-H. Lin, C.-P. Zhang, J.-C. Xiao and X. Zheng, Beilstein J. Org. Chem., 2013, 9, 2635; (k) H. Egami, R. Shimizu, S. Kawamura and M. Sodeoka, Angew. Chem., Int. Ed., 2013, 52, 4000; (l) R. Zhu and S. L. Buchwald, Angew. Chem., Int. Ed., 2013, 52, 12655; (m) H. Egami, S. Kawamura, A. Miyazaki and M. Sodeoka, Angew. Chem.,

Chem. Commun.

5 6

7 8 9 10 11 12 13

´, Int. Ed., 2013, 52, 7841; (n) N. O. Ilchenko, P. G. Janson and K. J. Szabo Chem. Commun., 2013, 49, 6614; (o) Q. Yu and S. Ma, Chem. – Eur. J., 2013, 19, 13304; (p) X. Wang, Y. Ye, G. Ji, Y. Xu, S. Zhang, J. Feng, Y. Zhang and J. Wang, Org. Lett., 2013, 15, 3730; (q) Y.-T. He, L.-H. Li, Y.-F. Yang, Y.-Q. Wang, J.-Y. Luo, X.-Y. Liu and Y.-M. Liang, Chem. Commun., 2013, 49, 5687; (r) D.-F. Lu, C. L. Zhu and H. Xu, Chem. Sci., 2013, 4, 2478; (s) Y.-T. He, L.-H. Li, Z.-Z. Zhou, H.-L. Hua, Y.-F. Qiu, X.-Y. Liu and Y.-M. Liang, Org. Lett., 2014, 16, 270; (t) J.-S. Lin, Y.-P. Xiong, C.-L. Ma, L.-J. Zhao, B. Tan and X.-Y. Liu, Chem. – Eur. J., 2014, 20, 1332; (u) H. Egami, R. Shimizu, Y. Usui and M. Sodeoka, J. Fluorine Chem., 2014, 167, 172. Selected examples: (a) D. A. Nagib and D. W. C. MacMillan, Nature, 2011, 480, 224; (b) Y. Yasu, T. Koike and M. Akita, Angew. Chem., Int. Ed., 2012, ´pez, 51, 9567; (c) S. Mizuta, K. M. Engle, S. Verhoog, O. Galicia-Lo ´debielle, K. Wheelhouse, G. Rassias, A. L. Thompson M. O’Duill, M. Me and V. Gouverneur, Org. Lett., 2013, 15, 1250; (d) E. Kim, S. Choi, H. Kim and E. J. Cho, Chem. – Eur. J., 2013, 19, 6209; (e) Y. Yasu, T. Koike and M. Akita, Org. Lett., 2013, 15, 2136; ( f ) P. Xu, J. Xie, Q. Xue, C. Pan, Y. Cheng and C. Zhu, Chem. – Eur. J., 2013, 19, 14039; (g) S. Mizuta, S. Verhoog, K. M. Engle, T. Khotavivattana, M. O’Duill, K. Wheelhouse, ´debielle and V. Gouverneur, J. Am. Chem. Soc., 2013, G. Rassias, M. Me 135, 2505; (h) A. Carboni, G. Dagousset, E. Magnier and G. Masson, Org. Lett., 2014, 16, 1240; (i) Q.-Y. Lin, X.-H. Xu and F.-L. Qing, J. Org. Chem., 2014, 79, 10434. S. P. Pitre, C. D. McTiernan, H. Ismaili and J. C. Scaiano, ACS Catal., 2014, 4, 2530. (a) L. Harmand, S. Cadet, B. Kauffmann, L. Scarpantonio, P. Batat, ´coue `res and J.-M. G. Jonusauskas, N. D. McClenaghan, D. Laste Vincent, Angew. Chem., Int. Ed., 2012, 51, 7137; (b) L. Harmand, ´coue `res R. Lambert, L. Scarpantonio, N. D. McClenaghan, D. Laste and J.-M. Vincent, Chem. – Eur. J., 2013, 19, 16231; (c) R. Beniazza, L. Harmand, R. Lambert, F. Molton, C. Duboc, S. Denisov, ´coue `res and J.-M. G. Jonusauskas, N. D. McClenaghan, D. Laste Vincent, Chem. – Eur. J., 2014, 20, 13181. S. Fantasia, J. M. Welch and A. Togni, J. Org. Chem., 2010, 75, 1779. We assume that the amount of CF3D that forms is underestimated due to its very high volatility (boiling point of CF3H = 82.1 1C). S. Mizuta, S. Verhoog, X. Wang, N. Shibata, V. Gouverneur and ´debielle, J. Fluorine Chem., 2013, 155, 124. M. Me (a) J. K. Kochi and J. D. Bacha, J. Org. Chem., 1968, 33, 2746; (b) J. K. Kochi, A. Bemis and C. L. Jenkins, J. Am. Chem. Soc., 1968, 90, 4616; (c) C. L. Jenkins and J. K. Kochi, J. Am. Chem. Soc., 1972, 94, 843. Review: A. Casitas and X. Ribas, Chem. Sci., 2013, 4, 2301–2318. J. Xu, Y. Fu, D.-F. Luo, Y.-Y. Jiang, B. Xiao, Z.-J. Liu, T.-J. Gong and L. Liu, J. Am. Chem. Soc., 2011, 133, 15300. The ESI-MS spectrum of a freshly prepared MeOH solution of [Cu(MeCN)4]PF6 and Togni’s reagent shows intense peaks at m/z = 625.8, 694.8 and 941.7, which were assigned to Cu carboxylato trifluoromethyl iodonium ions (see ESI†). Of particular interest is the observation of the peak at m/z = 694.8, which is assigned to a CuI dicarboxylato trifluoromethyl iodonium cation. When running these experiments from the CuII salt CuOTf2, the MS spectra were very similar, except for the absence of the peak at m/z = 694.8 (see ESI†).

This journal is © The Royal Society of Chemistry 2015

Copper(I)-photocatalyzed trifluoromethylation of alkenes.

Using the photoreducible Cu(II) precatalyst 2, trifluoromethylation reactions of alkenes are conducted effectively at low copper loading (0.1-0.5 mol%...
1MB Sizes 2 Downloads 16 Views