NHC-Cu Three-Coordinate Complex as a Promising Photocatalyst for Energy and Electron Transfer Reactions

Herein, we describe a simple three-coordinate complex of Cu(I) with an NHC and 1,10-phenanthroline ligands as an effective photocatalyst for energy (e.g., olefin E/Z isomerization) and electron transfer (e.g., aryl halide dehalogenation) reactions under blue-light irradiation. This complex can be obtained in a one-pot procedure starting from commercially available reagents and green solvents (EtOH, water). We hereby present a study of its activity and mechanistic insight into its mode of operation.


■ INTRODUCTION
−3 The most typical and comprehensive photocatalysts, in terms of their portfolio of catalyzed transformations, are low-spin ruthenium and iridium complexes.A major drawback is their high price, resulting from low natural abundance and toxicity, which translates into limited usefulness in large-scale industrial (i.e., pharmaceutical) applications.−9 McMillin et al. 10 and Meyer et al.'s 11 pioneering research on tetracoordinate bisphenanthroline Cu[I] complexes laid the foundations for the family of homoleptic complexes with the schematic structure [Cu-(N^N)2]+, which are widely used in photocatalysis. 9An important note is that due to steric and electronic factors, the excited-state lifetimes and redox potentials for simple nonsubstituted complexes are very limited.Several strategies were developed to overcome this issue, including the application of heteroleptic Cu[I] complexes. 9In order to maximize the potential of those catalysts there should be no dynamic exchange of ligands in solution, which is ensured by the use of certain biphosphines (such as Xantphos) 12 or bisisonitriles. 13n 2014, Linares et al. showed photophysical and electrochemical properties of a series of new three-coordinate, heteroleptic NHC-Cu[I] complexes containing bidentate N^N ligands (Figure 1). 14,15In general, these compounds absorb light in the blue and near UV range, possessing high redox potentials for both reduction and oxidation and, for some of them, long lifetimes in the excited state as well as high triplet energies (Figure 1).Even though these properties make the reported NHC-Cu(I) complexes perfectly suited for the application in photocatalysis, their use as such has never been explored.Therefore, considering the photophysical properties of reported NHC-Cu(I) complexes along with the fact that they are straightforward to synthesize, nontoxic, and inexpensive, we envisioned that their application as photocatalysts might lead to the development of a general alternative for Ir-and Ru-based complexes.
Figure 1.Selected photoactive copper complexes with their respective ground-state redox potentials (E ox and E red , vs SCE) 14 and experimental emission energies (as triplet energy estimation ̃ET ) 15 ■ RESULTS AND DISCUSSION Energy Transfer.To prove our hypothesis, we decided to check their photocatalytic activity in both energy and electron transfer reactions.Since E → Z photoisomerization of olefins was already documented in Cu(I) photochemistry, 12 we decided to check the effectiveness of NHC-Cu(I) photocatalysts as potential photosensitizers in this type of transformation.The synthesis of [Cu]-1, following literature procedures as well as our newly developed one-pot, green protocol, is presented in the SI.We initiated our studies by exploring the reactivity of trans-stilbene (1-E, E T = 214.3kJ/ mol, E 1/2 = 1.48 V vs Ag/AgCl) under the action of [Cu]-1.The starting conditions were based on previous protocols for stilbene isomerization. 16,17The reaction produced the desired cis-isomer 1-Z in an 86/14 ratio (Table 1, entry 1).
Subsequently, several reaction parameters (catalyst loading, light source, concentration, and time) were optimized (Table 1).Overall, photoisomerization of 1-E under the action of [Cu]-1 at room temperature was complete in only 4 h and lead to the Z/E-mixture in a 95/5 ratio (much faster rate compared to previous protocols 16,17 that typically require overnight irradiation).Control experiments confirmed that the desired transformation cannot take place in absence of light (entry 13) or a photocatalyst (entry 14).
Next, we decided to test the scope of the catalyzed isomerization under the optimized conditions (Scheme 1).In general, variously substituted stilbenes were susceptible to the reaction conditions (56 to 95% of isomerization).Reactions were mostly clean and gave no other products besides the Zisomer and could be observed via 1 H NMR and/or GC after 4 h of irradiation (see the SI).4-Nitrostilbene 3-E and πextended stilbene 12-E were significantly less reactive.On the other hand, for substrates possessing multiple EDG�OMe: methoxystilbenes 9-E, 10-E, and 11-E, we observed that prolonged irradiation resulted in a decreased Z/E ratio.After short reoptimization of the reaction conditions for olefin 9-E (which is a proven anticancer agent named DMU-212), 18 we found that the most optimal time for the reaction is 2 h, giving a 92/8 ratio.Presumably the reason for the back-isomerization is a lower triplet excited-state E/Z energy barrier resulting in a lesser degree of stereochemical control and/or gradual decomposition of [Cu]-1 into photoactive derivatives leading to a different photostationary state (for details see the SI).Cinnamonitrile 13-E and cinnamaldehyde 14-E as well as bromostyrene 18-E and styryl ketones 6-E and 19-E were subject to isomerization to a very limited extent while olefin derivatives 15−17 proved unreactive under developed conditions, presumably because the triplet energy of [Cu]-1 (E T ∼ 245 kJ/mol) is not high enough for these derivatives (15�246.8kJ/mol, 17�256.1 kJ/mol; for details, see the SI).

The Journal of Organic Chemistry
Based on our results and the literature data, 16,17 we propose a plausible reaction pathway for the E → Z photoisomerization of olefins (Scheme 2).Most of the literature data suggests a triplet energy transfer pathway; however, during our studies, we established that the reaction was noticeably slowed down in the presence of a singlet-state quencher, TEMPO, and both singlet and triplet quenchers benzoquinone and azulene (see the SI).On the other hand, addition of a triplet quencher, 1,3cyclohexadiene, does not slow down isomerization of 1-E in the presence of [Cu]-1, which, together with little sensitivity to the presence of oxygen, might suggest the involvement of a singlet state mechanism for this isomerization in agreement with previous propositions by Poisson for Cu II /BINAPcatalyzed E/Z isomerization of cinnamates 19 and BINOLcatalyzed E/Z isomerization of vinyl boronates. 20Still, we have to take into account that singlet quenchers might affect the outcome of a triplet-energy transfer-based isomerization due to the inhibition of the photosensitizer's S 0 → S 1 excitation, which produces T 1 via ISC.To further investigate this problem, we performed reactions in the presence of triplet quenchers (O 2 atmosphere and 1,3-cyclohexadiene) with lower catalyst loading (1%), decreased irradiation power (12.5 W), and shorter time.With this approach, we were able to see a subtle decrease of the rate of isomerization with both additives (see the SI).This observation along with previous reports 16,17 makes us presume that reaction follows the triplet energy transfer mechanism yet is somewhat resistant toward triplet quenching.Photoisomerization of pure cis-olefins 1-Z and 9-Z under standard conditions leads to a very similar Z/E mixture as when the appropriate trans-olefins are used as starting materials (see the SI).This result suggests to us the presence of a specific photostationary state of the reaction and is further evidence for the triplet energy transfer as a reaction mechanism.We have also synthesized [Cu]-2 and [Cu]-3, dipyridylamine analogues of [Cu]-1, and examined that both complexes exhibit a much slower rate of isomerization in every studied case (see the SI).
Additionally, we decided to check the photostability of [Cu]-1 (for the procedure details, see the SI).The amount of remaining catalysts was monitored by 1 H NMR, showing less than 1% of decomposition after 4 h and 20% decomposition after 24 h, which indicates the high photorobustness of the tested catalyst even in the presence of atmospheric oxygen.Encouraged by these results, we decided to investigate the reusability of the photocatalyst by carrying out the isomerization reaction of stilbene 11-E with successive addition of subsequent portions of the substrate to the reaction mixture (for a detailed procedure, see the SI).It turned out that the NHC-Cu photocatalyst not only is stable under photo-irradiation but also can be used at least four times for the photoisomerization.
Electron Transfer.After successful application of the NHC-Cu(I) photocatalyst for energy transfer reactions, we turned our attention toward its application for electron transfer reaction.Inspired by the work of Cibulka et al. 21a and our results, we decided to check the effectiveness of this NHC-Cu(I) ([Cu]-1) photocatalyst for the reduction of aryl halides.To our delight, the photoreduction of 18a (E 1/2 = −2.75V vs SCE) under the action of [Cu]-1 as photocatalyst yields the reduced product (18b) in 34% yield (Table 2).Subsequently, some reaction parameters (concentration, catalyst loading, and time; for details, see the SI) were optimized.Overall, photoreduction of 18a catalyzed by NHC-Cu(I) (2.5 mol %) under blue-light irradiation (405 nm), in the presence of triethylamine (2.1 equiv) and Cs 2 CO 3 (1.1 equiv) at room temperature for 24 h, gives the product 18b in 65% yield.Background experiments confirmed that the application of photocatalyst, light, and base is crucial for the reaction to proceed (Table 2, entries 2, 3, and 5).
With the optimized conditions in hand, we examined a scope and limitations study (Scheme 3).The reduction of aryl halides proceeds nicely for arenes with various EWG functional groups (COMe, CO 2 Me, CN, and COOH) giving the desired products (21b, 23b, 26b, and 27b) in good to superb yields (Scheme 3).Furthermore, the NHC-Cu photocatalyst is also active in the photoreduction of difficult to reduce EDGsubstituted aryl halides (E red <−2.5 V vs SCE), giving the desired product in 41−85% (18b, 19b, 24b, 28b), depending on the halide derivative).However, the reaction proved difficult for CHO substituents for all derivatives (ArI�40%, ArBr�64%, ArCl�41%).The reduction of arenes possessing more than one aromatic ring (34a, 35a, and 36a) gives products with good to moderate yields (45−79%).On the other hand, reduction of halogen-substituted heteroarenes is also possible by the use of the NHC-Cu photocatalyst (37− 39), however, with diminished yields (35−38%).Based on the aforementioned results and the literature data, 21 we propose a plausible radical reaction pathway for the photoreduction of aryl halides under the action of the [Cu]-1 photocatalyst (Scheme 4).In this approach, the first event is photoexcitation

■ CONCLUSIONS
In summary, NHC-Cu(I) complexes, which are wellresearched in conventional organometallic catalysis 22 and as potential OLED materials, 14,15 can also promote photoinduced energy and electron transfer processes when exposed to visible light irradiation.They act as efficient energy transfer catalysts (E → Z photoisomerization of olefins) and electron transfer photocatalysts (photoreduction of aryl halides).Studied NHC-Cu(I) photocatalysts exhibit features superior to those of other metal-based photocatalysts that work under visible-light irradiation as inexpensive and ready to make and showing great potential in future applications in biological or industrial systems.Hence, we believe that NHC-Cu(I) complexes are valuable additions to the visible-light photocatalyst library and that this study will lead to more applications.

■ ASSOCIATED CONTENT Data Availability Statement
The data underlying this study are available in the published article and its online Supporting Information.
* sı Supporting Information 01510.Calculations have been carried out using resources provided by Wroclaw Centre for Networking and Supercomputing (https://wcss.pl),grant no.518.