Regulating Chlorine and Hydrogen Atom Transfer for Selective Photoelectrochemical C─C Coupling by Cu‐coordination Effect at Semiconductor/Electrolyte Interfaces

Abstract

Semiconductor‐based photoelectrochemical (PEC) organic transformations usually show radical characteristics, in which the reaction selectivity is often difficult to precisely control due to the nonselectivity of radicals. Accordingly, several simple organic reactions (e.g., oxidations of alcohols, aldehydes, and other small molecules) have been widely studied, while more complicated processes like C─C coupling remain challenging. Herein, a synergistic heterogeneous/homogeneous PEC strategy is developed to achieve a controllable radical‐induced C─C coupling reaction mediated by the copper‐coordination effect at the semiconductor/electrolyte interfaces, which additionally exerts a significant impact on the product regioselectivity. Through experimental studies and theoretical simulations, this study reveals that the copper‐chloride complex effectively regulates the formation of chloride radicals, a typical hydrogen atom transfer agent, on semiconductor surfaces and stabilizes the heterogeneous interfaces by suppressing the radical‐induced surface passivation. Taking the Minisci reaction (the coupling between 2‐phenylquinoline and cyclohexane) as a model, the yield of the target C─C coupling product reaches up to 90% on TiO₂ photoanodes with a selectivity of 95% and long‐term stability over 100 h. Moreover, such a strategy exhibits a broad scope and can be used for the functionalization of various heteroaromatic hydrocarbons.

A synergistic heterogeneous/homogeneous PEC strategy mediated by Cu²⁺ achieves a controllable radical‐induced C─C coupling reaction between aliphatic compounds with inert C(sp³)─H bond and a wide range of aromatics by using TiO₂ photoanodes. Experimental studies and theoretical simulations reveal that the presence of Cu²⁺ regulates the formation of chloride radicals and stabilizes the heterogeneous interfaces.

1. Introduction

Semiconductor‐based photoelectrochemical (PEC) water splitting reaction has been widely studied for hydrogen production,^{[ 1} , ² , ³ , ^{4 ]} but the overall performance is limited by the anodic oxygen evolution reaction (OER) with sluggish kinetics.^{[ 5} , ⁶ , ⁷ , ⁸ , ⁹ , ^{10 ]} Alternatively, PEC organic reactions,^{[ 8} , ¹¹ , ^{12 ]} such as alcohol oxidation,^{[ 13} , ¹⁴ , ¹⁵ , ^{16 ]} have drawn much attention due to their more favorable thermodynamics and the production of high‐valued chemicals. For example, glycerol is oxidized to 1,3‐dihydroxyacetone by nanoporous BiVO₄ photoanodes in an acidic medium without the addition of oxidizers.^{[ 15 ]} Similarly, the high‐performance PEC oxidation of 5‐hydroxymethylfurfural (HFM) is also reported.^{[ 17 ]} In addition to simple oxidation, PEC activation of C─H bonds provides more possibilities for artificial photosynthesis.^{[ 18} , ¹⁹ , ²⁰ , ^{21 ]} For example, cyclohexanol and cyclohexanone are prepared by the C─H activation of cyclohexane, which enables the construction of C─O bonds and achieves high oxidation selectivity.^{[ 18 ]} The reaction of electron‐rich aromatic hydrocarbons with azole compounds to produce medicinal nitrogen heterocycles has been realized by a PEC method.^{[ 20 ]} The construction of the C─P bond has also been realized, which shows good functional group tolerance.^{[ 21 ]}

Even though photoelectrochemistry has achieved the construction of C─P bonds and C─N bonds, these works only account for a tiny part of organic reactions. The reported semiconductor‐based PEC organic transformations usually show radical characteristics, and the reaction selectivity is often difficult to precisely control due to the nonselectivity of radicals. Therefore, developing new PEC strategies to achieve more complex reactions (e.g., C─C coupling) with high product selectivity still needs to be explored.^{[ 22} , ²³ , ^{24 ]} It is a feasible way to construct C─C bonds by directly forming alkane radicals with C(sp³)─H activation, which can be expanded to a wide range of substrates.^{[ 25 ]} However, as one of the most prevalent but least active bonds in organic molecules, it is challenging to break the C(sp³)─H bond due to the high bond dissociation energies (413 kJ mol⁻¹).^{[ 26} , ^{27 ]}

Semiconductor photoelectrodes with deep valence bands, such as TiO₂, can directly oxidize chlorine ion (Cl⁻) to obtain chlorine radical (Cl ^· ),^{[ 19} , ^{28 ]} an emerging hydrogen atom transfer (HAT) agent for C─H bond cleavage,^{[ 25} , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ^{31 ]} which is promising to generate alkyl radicals for C─C coupling through the HAT mechanism. However, we find that this reaction suffers from poor selectivity and stability. Herein, we develop a synergistic heterogeneous/homogeneous PEC strategy via the incorporation of copper ions (Cu²⁺) for highly selective and stable C─C coupling. Through experimental studies and theoretical calculations, we demonstrate that Cu²⁺ at semiconductor/electrolyte interfaces can effectively regulate the formation of Cl ^· on TiO₂ surfaces and stabilize the heterogeneous interfaces. Interestingly, excellent regioselectivity is achieved owing to the coordination effect of the CuCl_x complex. Taking the Minisci radical coupling (the coupling between 2‐phenylquinoline and cyclohexane) as a model reaction, a high C─C coupling product yield of 90% together with a selectivity of 95% and long‐term stability over 100 h are realized. This strategy exhibits a broad spectrum of applications and can be employed to functionalize diverse heteroaromatic hydrocarbons.

2. Results and Discussion

The Minisci radical coupling reaction utilizes Cl ^· as the HAT reagent to transform alkane into alkane radicals, and then the alkane radicals can be combined with protonated aromatics to obtain C─C coupling products (Figure  1a). The TiO₂ photoanode can directly oxidize Cl⁻ to obtain Cl ^· , which showed a high photovoltage of 1.77 V under 380–800 nm irradiation as the advantage for solar energy utilization (Figure S1, Supporting Information). We screened the conditions for the reaction by using 2‐phenylquinoline (1A) and cyclohexane (2A) as substrates. Since 1A itself has an ultraviolet absorption below 380 nm (Figure S2, Supporting Information), the PEC reaction was carried out under light irradiation with wavelength between 380 and 800 nm. 100 µL of HCl was used as the source of Cl⁻ and proton, and 0.1 M of tetraethyl ammonium tetrafluoroborate (TEATFB) was used as the electrolyte (Figure S3, Supporting Information). Under the PEC condition, the yield of 3A (the target C─C coupling product) reached 70% but only in the first cycle with a freshly prepared photoanode, while the yield decreased to 50% in the second cycle and even lower in the following cycles (Figure 1b). We find that by introducing 5% mol Cu²⁺ into the electrolyte, the yield of 3A was significantly improved to 80% and showed no decay within 9 cycles, which exhibited a long‐term stability of overall 117 h (Figure 1b). Notably, the color of the TiO₂ photoanode turned yellow after photoelectrolysis in Cu²⁺‐free solutions when using quinoline as the reaction substrate, while it was not observed for photoanodes exposed to Cu²⁺‐bearing solutions (Figure S4, Supporting Information). Therefore, the presence of Cu²⁺ plays an important role in the stability of TiO₂. Despite the much‐enhanced C─C coupling behavior, the photocurrent obtained from the linear sweep voltammetry (LSV) measurement showed little difference after introducing Cu²⁺ (Figure S5a, Supporting Information), as well as the steady‐state photocurrent (Figure S5b, Supporting Information).

Figure 1.

The Minisci radical coupling reaction on a TiO₂ photoanode. a) Schematic of the PEC process. b) Reaction performance and stability test for the C─C coupling for 91 h in batch reaction (7 batches) on a TiO₂ photoanode, and with Cu²⁺ for the C─C coupling for 117 h in batch reactions (9 batches). The reaction was carried out in acetone/water solution with 0.1 m TEATFB and 100 mL HCl at 1.2 V_Ag/AgCl under the illumination of 380–800 nm (400 mW cm⁻²) with 13 h for every batch. The photoanode was washed by deionized water and acetonitrile after each batch reaction.

Through the optimization of reaction conditions, we achieved a high yield of 90% and the conversion rate of 95% in a mixture of acetone and water (V/V = 19/1) containing 5% mol Cu²⁺, 0.1 m TEATFB and 100 µL HCl under 1.6 V_Ag/AgCl for 13 h (Table 1 , entry 1). As anticipated, the reaction did not proceed in the absence of light or applied bias (Table 1, entry 2–3). The potential‐dependent product yield is presented in Table S1 (Supporting Information) (entry 1–3), in which the product yield significantly increased with higher applied bias. The high light intensity was utilized to enhance the current density, thereby attaining a greater yield within a fixed time (Figure S6, Supporting Information). Lowering the content of Cu²⁺ would decrease the product yield (Table 1, entry 4). On the contrary, increasing the concentration of Cu²⁺ beyond 10% mol led to no improvement in yields (Table 1, entry 5). When the reaction was carried out directly in the air, the yield decreased to 63% (Table 1, entry 6), indicating of the importance to deaerate the reaction electrolyte. Increasing the temperature was conducive to the reaction below 50 °C, whereas an excessive temperature led to the decrease in yield as the solvent started to evaporate (Table S1, Supporting Information; entry 4–5). The absence of TEATFB would result in a significant decrease in the yield of 3A (Table 1, entry 7), and the yield was also influenced by the type of electrolyte salts (Table S1, Supporting Information; entry 6–9). HCl also plays an important role. It not only provides an acidic environment to protonate 1A, but also acts as the source of chloride required for HAT (Table 1, entry 8). Therefore, the influence of acid quantities on the yield was investigated, and an optimized content of HCl was found to be 100 µL (Table 1, entry 9–10).

Table 1.

Optimization of reaction conditions.

Entry	HCl [µL]	Cu2+ [%mol]	Conversion [%]	Yield [%] a)
1	100	5	95	90
2 b)	100	5	0	0
3 c)	100	5	2	2
4	100	1	72	69
5	100	10	88	85
6 d)	100	5	70	63
7 e)	100	5	33	32
8	–	5	0	0
9	50	5	68	66
10	150	5	78	75

Standard reaction conditions: TiO₂ as the photoanode, Ni foam as the cathode, 1A (0.1 mmol), 2A (0.3 mL), 5 mL acetone/water (V/V) = 19/1, HCl (100 µL), Cu²⁺ (5% mol), TEATFB (0.1 M), 40 °C, 13 h.

^a) Determined by ¹H NMR analysis using 1,3,5‐trimethoxybenzene as the internal standard;

^b) In the dark;

^c) No bias;

^d) In the air;

^e) No electrolyte.

To demonstrate the general applicability of the PEC strategy for C─C coupling reaction, we extended the optimal conditions (Table 1, entry 1) to more substrates (Figure  2 ). We investigated the range of alkanes capable of reacting with 1A. Simple cycloalkanes similar to cyclohexane afforded a good yield under these conditions (4A–6A). The low yield of cyclopentane may be attributed to its low boiling point, which was easy to volatilize under these conditions. Cyclododecane afforded the desired product with 95% yield (7A) by raising the temperature to 50 °C and prolonging the reaction time to 24 h. Notably, 1,4‐epoxycyclohexane provided heteroarenes (8A) by reacting with 1A. Inspired by this, we explored the reactivity of oxygen‐containing substances. To our gratification, this method was found to be broadly compatible with ethers (9A), alcohols (10A–12A), and epoxides (13A–16A) and showed a satisfactory yield. However, the C─O cleavage was observed for the reactions of methanol under the standard reaction conditions. In particular, the reaction with tetrahydrofuran (THF) induced the C─O cleavage by decreasing the acid amount (15A), while the standard reaction conditions resulted in the C─H bond cleavage (14A). It was worth noting that our system can facilitate the reaction between aromatic compounds with a yield of ≈65% (17A–18A). We carried out a high‐dose experiment by using 4.88 mmol of 2‐phenylquinoline to react with 1,4‐dioxane for 60 h, and obtained 0.215 g of the product by separation.

Figure 2.

Substrate scope. Reaction conditions: TiO₂ as the photoanode, Ni foam as the cathode, 1 (0.1 mmol), 2 (0.3 mL for liquid, or 10 equiv. for solid), 5 mL acetone/water (V/V) = 19:1, HCl (100 µL), Cu²⁺ (5% mol), TEATFB (0.1 M), 40 °C, 13 h. Determined by ¹H NMR analysis using vinyl carbonate as the internal standard. [a] The reaction was run for 24 h. [b] The reaction was heated to 50 °C. [c] 25 µL HCl was used.

The substrate and functional group tolerance were further investigated by coupling various heterocycles with 2A (Figure  3 ). We were pleased to see that very high regioselectivity was obtained under the PEC conditions. Quinoline was predominantly alkylated at the C2‐position (19A).^{[ 29} , ^{30 ]} However, the absence of Cu²⁺ in the solution resulted in poor regioselectivity of products, with a yield of only 30% for the C2‐position, in addition to the presence of the C4‐position (Figure S7, Supporting Information). Moreover, alkylation at the C1‐position resulted in excellent yields for isoquinolines (20A) and phenanthridine (21A). When quinoline substituted at C2‐position was replaced by methyl, chlorine, bromine functional groups, or methoxy groups, they afforded the corresponding alkylated products with a satisfactory yield at the C4‐position (22A–25A). Meanwhile, when the substituent group was an electron‐withdrawing group, the yield decreased. On the other hand, quinolones bearing a C4‐substituent underwent alkylation at the C2‐position (26A–28A). In addition, cyclohexane also underwent reactions with pyridine (29A), pyrimidine (30A), benzothiophene (31A), benzothiazole (32A), benzopyrimidine (33A), quinoxaline (34A), pyridazine (35A), and purine (36A) aromatics. The PEC strategy was further applied to complex substrates such as hydrocinchonine (37A) and Fasudil (38A).

Figure 3.

Substrate scope. Reaction conditions: TiO₂ as the photoanode, Ni foam as the cathode, 1 (0.1 mmol), 2 (0.3 mL for liquid, or 10 equiv. for solid), 5 mL acetone/water (V/V) = 19:1, HCl (100 µL), Cu²⁺ (5% mol), TEATFB (0.1 M), 40 °C, 13 h. Determined by ¹H NMR analysis using vinyl carbonate as the internal standard. [a] In standard reaction conditions without Cu²⁺.

The role of Cu²⁺ is twofold as unveiled in Figure 1b, as it not only maintains the stability of TiO₂ photoanodes but also enhances the C─C coupling performance. To comprehend these dual functions, we initially assessed its impact on the physical structure of the photoanode. Scanning electron microscopy (SEM) measurements (Figure S8, Supporting Information) and powder X‐ray diffraction (XRD) measurements (Figure S9, Supporting Information) were used to characterize the surface morphology and crystal phase of photoanodes, respectively. Both the SEM and XRD results showed that the structural integrity of the TiO₂ photoanode remained unaffected when operated in Cu²⁺‐free solutions, although it presented a remarkably decreased performance for the C−C coupling reaction (Figure 1b). The decreased stability of the TiO₂ photoanode derives from the surface passivation, as revealed by X‐ray photoelectron spectroscopy (XPS) measurement (Figure  4a). The peak intensity of the sp²‐C (288.6 eV) displayed an obvious increase for the TiO₂ photoanode used in the Cu²⁺‐free electrolyte when compared with that of other photoanodes. Accordingly, we suspect that the organics in the Cu²⁺‐free electrolyte may deposit on the photoanode surface during reactions, which resulted in the passivation of the active sites on TiO₂ surfaces. Fortunately, such a process can be effectively suppressed by introducing Cu²⁺ into the electrolyte (Figure 4a), which maintained the high activity of the TiO₂ photoanode (Figure 1b). However, there was no obvious change in the binding energy of those Ti elements (Figure S10a, Supporting Information), indicating that the surface structure of the TiO₂ photoanode was not changed after the reaction in the Cu²⁺‐free electrolyte, although it presented a significantly decreased C─C coupling activity. The structures of photoanodes were further explored by Raman spectroscopy as shown in Figure S11 (Supporting Information), where the peaks at 443 and 610 cm⁻¹ were attributed to the E_g and A_1 g of rutile TiO₂, respectively.^{[ 31 ]} The Raman vibration modes of A_1 g and E_g of TiO₂ after the reaction exhibited obvious changes. To gain insight into this change, the area ratios of A_1 g and E_g ($S_{A_{1g}}/S_{E_g}$) were calculated (Table S3, Supporting Information). We found that the $S_{A_{1g}}/S_{E_g}$ of the TiO₂ reacting in the Cu²⁺‐bearing solution was almost the same as that of the fresh TiO₂ photoanode, while the $S_{A_{1g}}/S_{E_g}$ strength ratio of the TiO₂ reacting in the Cu²⁺‐free solution was significantly increased.^{[ 31 ]} The increased $S_{A_{1g}}/S_{E_g}$ area ratio implied that the up‐and‐down vibration modes in the TiO₂ unit cell was inhibited. The constrained up‐and‐down vibrational mode influences the Ti─O bond in the TiO₆ octahedron to converge, and affects the stretching of the shorter four Ti─O bonds. This result again indicated the strong interaction between TiO₂ and sp²‐C.

Figure 4.

Effect of Cu²⁺ on the TiO₂ photoanode. a) Normalized C1s XPS spectra of the fresh TiO₂ (Fresh), the TiO₂ after reaction without Cu²⁺ and with Cu²⁺. b) Normalized IMPS spectra of the fresh TiO₂ (Fresh), the TiO₂ after reaction without Cu^2+, and the TiO₂ after reaction with Cu²⁺. All were performed in standard reaction conditions without Cu²⁺. c) Schematics of TiO₂ surface in the Cu²⁺‐free electrolyte and Cu²⁺‐bearing electrolyte. d) CVs obtained in 0.3 m NaClO₄ by using the fresh TiO₂ (Fresh), the TiO₂ after reaction in the solution with Cu^2+, and the Cu‐modified TiO₂ (Cu/TiO₂). The scan rate was 10 mV·s⁻¹. e) Cu 2p XPS spectra of the fresh TiO₂ (Fresh), the Cu/TiO₂ before and after reactions.

The intensity‐modulated photocurrent spectroscopy (IMPS) (Figure 4b) was used to dissect the difference in the interfacial charge transport efficiency (η_ct) for TiO₂ photoanodes that have been used in the Cu²⁺‐bearing or Cu²⁺‐free solution.^{[ 32} , ^{33 ]} First, the fresh TiO₂ was utilized to assess the charge transport efficiency in Cu²⁺‐free or Cu²⁺‐bearing electrolyte (Figure S12a, Supporting Information), while the presence of Cu²⁺ did not exert any influence on η_ct. For used TiO₂ photoanodes as shown in Figure 4b, the η_ct of the TiO₂ after reaction in Cu²⁺‐bearing solution was 71%, while the η_ct of the TiO₂ after reaction in Cu²⁺‐free solution was only 47.9% (Table S3 and Figure S12b, Supporting Information). The significant difference indicated that the charge‐transfer kinetics of TiO₂ without Cu²⁺ was much lower than that with Cu²⁺, which can be attributed to the deposition of organics onto the photoanode surfaces in the Cu²⁺‐free electrolyte. It can be seen from the above characterization that the introduction of Cu²⁺ can prevent the surface passivation of TiO₂ photoanodes (Figure 4c).

To further rule out the possibility that the Cu²⁺ would be deposited onto the TiO₂ surfaces during the reaction, cyclic voltammetry (CV) tests were conducted on the fresh TiO₂ (Figure S13a, Supporting Information), the TiO₂ after the reaction Cu²⁺‐bearing solutions, and the Cu‐modified TiO₂ (Cu/TiO₂) (Figure S13b, Supporting Information).^{[ 32 ]} As shown in Figure 4d, compared with that of the fresh TiO₂, there were minimal changes after the reaction in Cu²⁺‐bearing solutions. On the contrary, the significant redox peaks associated with Cu oxide reduction (cathodic sweep) could be observed on the Cu/TiO₂. These results demonstrated that Cu²⁺ is not likely to deposit on the TiO₂ surface during the C─C coupling reaction in the Cu²⁺‐bearing solution. It was further supported by XPS measurements as shown in Figure 4e and Figure S10d (Supporting Information). The characteristic peaks of the Cu element could be observed on the Cu/TiO₂, while the signal of Cu on the TiO₂ surface after reaction in Cu²⁺‐bearing solutions was not detected. Cu/TiO₂ was used in standard reaction conditions to test performance. After the reaction (Figure S13c, Supporting Information), the characteristic peaks of the Cu element disappeared in XPS results (Figure 4e; Figure S13d–e, Supporting Information), and Cu oxide reduction (cathodic sweep) could not be observed in the CV test (Figure S13f, Supporting Information). Therefore, compared with the utilization of Cu/TiO₂, the synergistic heterogeneous/homogeneous PEC strategy is more convenient and exhibits greater stability.

The radicals or intermediates during the C─C coupling reaction were detected to further gain insight into the role of Cu²⁺. It showed that both radical quenchers 2,2,6,6‐tetramethylpiperidine 1‐oxyl (TEMPO) and 3,5‐di‐tert‐4‐butylhydroxytoluene (BHT) exhibited significant inhibition on the formation of 3A (Figure  5a), confirming the radical characteristic of the PEC process.^{[ 34} , ^{35 ]} The existence of alkyl radicals was further confirmed through the identification of a radical adduct 39A by using gas chromatography‐mass spectrometry (GC‐MS) (Figure S14, Supporting Information). To further probe the presence of chlorine species, a diallyl sulfonamide 40A was used under the optimized reaction condition with Cu²⁺ (Figure 5b). A cyclochlorinated compound 41A was proved by ¹H NMR and ¹³C NMR. From these results, we can draw a conclusion that the C─C coupling reaction was triggered by the PEC generated Cl ^· , while chlorine (Cl₂) is not active species in our system due to the absence of any specific alkenyl dichloride products.^{[ 34} , ³⁵ , ^{36 ]}

Figure 5.

Detection of radicals or intermediates. a) Radical quenching experiments. b) Chorine radical trapping experiments. c) EPR spectra detected from the photoelectrolysis of TiO₂ under 380–800 nm (400 mW cm⁻²) illumination in acetonitrile and water solution containing DMPO as the spin‐trapping agent. d) Concentration of chlorocyclohexane in standard reaction conditions with Cu²⁺ and without Cu²⁺. e) UV–vis spectra of CuCl₂ in different conditions and possible structures of CuCl_x complex. All were dissolved in a mixture of acetone and water (acetone/water (V/V) = 19:1).

The EPR spectra were further recorded in acetonitrile and water mixtures by using 5,5‐dimethyl‐1‐pyrroline N‐oxide (DMPO) as trapping agent (Figure 5c). In the 0.1 m LiClO₄, only OH ^· was detected. Upon the addition of HCl, the intensity of OH ^· signals decreased, but no significant Cl ^· signals were observed. This may be due to the low concentration of Cl ^· under this condition. Notably, new signals associated with Cl ^· were significantly enhanced, and the signals of OH ^· almost disappeared after adding Cu²⁺ into the electrolyte.^{[ 19} , ²⁸ , ^{37 ]} In addition, we used TEMPO as trapping agent to capture Cl ^· . When the mixture was subjected to high‐resolution mass spectrometry (HRMS) test, the TEMPO‐Cl was detected, which indicated the formation of Cl ^· (Figure S15, Supporting Information). These results confirmed the effective regulation of Cl ^· generation during PEC C─C coupling process with the assistance of Cu²⁺. Cl₂ was generated by dimerization of two Cl ^· , which can react with 2A to produce chlorocyclohexane. We further screen the possibility of Cl₂ generation. As determined by gas chromatography (GC), the presence of Cu²⁺ significant suppressed the yields of chlorocyclohexane (Figure 5d), indicating that the presence of Cu²⁺ would preferably inhibit the formation of Cl₂. Furthermore, we purposely replaced CuCl₂ with CuCl for this reaction. Compared with Cu²⁺, the presence of Cu⁺ exhibited similar product yield and selectivity (Figure S16, Supporting Information). Therefore, Cu⁺ does not affect our proposed mechanism. We further explored the effect of Cu²⁺ on water oxidation reaction (WOR). The oxygen gas was detected by gas chromatography to detect the competition from WOR. The Faraday efficiency of oxygen evolution was relatively low (< 10%) regardless of whether there was Cu²⁺ in the solution (Figure S17, Supporting Information).

The structure of Cu²⁺‐complex was tested by the UV–vis spectroscopy. Figure 5e shows that the absorption peaks of the Cu²⁺‐complex in the acetone were located at 476 and 329 nm. The peak at 329 nm is originated from the n → π transition, while the peak at 476 nm is due to the π → π ^* transition.^{[ 38 ]} The addition of HCl resulted in a new peak at 405 nm and a noticeable color change in the electrolyte from colorless to yellow (Figure S18, Supporting Information), which is consistent with the literature.^{[ 39} , ^{40 ]} Based on UV–vis spectroscopy and previous studies of copper‐chloro complexes in acetonitrile,^{[ 41 ]} we presented a possible structure (inset of Figure 5e) of the Cu²⁺‐complex, in which two Cl⁻ and two acetone molecules are coordinated to the Cu²⁺ center, i.e., [CuCl₂(C₃H₆O)₂].

We further investigated the production of the Cl ^· /Cl₂ at the Cu²⁺‐complex/TiO₂ interface or on the surface of bare TiO₂ by using density functional theory (DFT) calculations (Figure  6 ). For the Cl⁻ oxidation co‐catalyzed by the Cu²⁺ and TiO₂, the whole pathway started from the structure of the adsorbed Cu²⁺‐complex on the (110) facet of the TiO₂ photoanode as depicted in Figure 6a (slab). The adsorption of the Cu²⁺‐complex is a spontaneous process with a large adsorption energy of −1.58 eV, which positions the Cu²⁺‐complex within the Helmholtz layer of the TiO₂ photoanode. Besides, the valence state of Cu sites in the complex almost remains +2 within the catalytic cycle via Bader charge analysis as revealed in Figure S19 (Supporting Information). For the Cl⁻ oxidation process, two HCl molecules were introduced as the chlorine source to elucidate the energy difference in the formation of Cl₂ in these two pathways (Figure 6a). The first Cl ^· was produced by transferring a proton and electron from one HCl molecule, forming a five‐coordinated CuCl₃(C₃H₆O)₂ intermediate (Figure 6b). This process exhibits a lower formation energy of 1.06 eV compared with that of 1.45 eV on the bare TiO₂. Notably, as shown in Figure 6b, there were two resonance structures for the CuCl₃(C₃H₆O)₂ intermediate, displaying the de‐localization of spin electrons. These two Cl atoms coordinated to the centered Cu²⁺ site present a spin population of 0.35 and 0.46, indicating of a more stable Cl ^· . As a result, it required a high energy barrier of 0.70 eV for the formation of Cl₂ at the Cu²⁺‐complex/TiO₂ interface, compared with that of 0.15 eV on the surface of bare TiO₂. These results indicate that the Cu²⁺‐complex/TiO₂ interface facilitates Cl ^· formation while inhibiting Cl₂ production, thereby promoting the HAT efficiency.

Figure 6.

DFT calculations. a) Potential energy surfaces of the Cl⁻ oxidation at the Cu²⁺‐complex/TiO₂ interface and on the surface of bare TiO₂. b) the Cl ^· resonance structure in the Cu²⁺ solution. The isosurface in yellow corresponds to the spin‐charge density, and the numerical values of 0.35 and 0.46 represent the spin population of those two Cl^{−/ ·} species.

The proposed mechanism of C─C coupling reaction via the PEC strategy by using TiO₂ is illustrated in Figure  7 . First, CuCl₂ in the system is coordinated with acetone to form CuCl_x complex (I). Subsequently, the I is adsorbed on the TiO₂ surface through electrostatic interactions. Then, the HCl adsorbed near the Cu site is oxidized by photogenerated holes to form pentacordinate CuCl_x complex (II). The Cl ^· acts as a highly efficient HAT agent, reacting with cyclohexane to generate alkane radical (A′) and release the I for the next cycle spontaneously. Then A′ combines with the protonated aromatics (B) to get the intermediate of the radical cation (B′), and Cl ^· plays a crucial role to rearomatize the intermediate and yields the final product (AB). In addition, the A′ readily undergoes a reaction with Cl₂ to yield chlorocyclohexane as a byproduct. The presence of Cu²⁺ can inhibit the interaction of Cl ^· to form Cl₂ and reduces unwanted formation of alkyl chloride.

Figure 7.

The proposed mechanism for the Cu²⁺ co‐catalyzed Minisci radical coupling reaction. Black and gray ball represent hydrogen atoms.

The proposed structure of CuCl_x complex may provide insights into the underlying mechanisms responsible for the high regioselectivity of the presented C─C coupling method. First, pentadentate CuCl_x complex (Figure S20, Supporting Information; II) intermediates might be engaged in hydrogen bonding interactions with a suitable radical nucleophile as well as its quinoline counterion (Figure S20, Supporting Information; III). The subsequent step involves elimination of HCl from intermediate III to obtain intermediate IV. The hydrogen bonding would favor the alkylation at the C2 position. After radical addition, the CuCl_x complex transitions from intermediate IV to the next cycle, while simultaneously acquiring intermediate V. Finally, Cl ^· would function as the single‐electron oxidant required to rearomatize the heteroarene. This hypothesis is consistent with the result that CuCl_x complex leads to a high regioselectivity.

For the continuous production, we sought to construct a flow electrolyzer. TiO₂ was used as the anode (2 cm × 2 cm) and Ni foam as the cathode (2 cm × 2 cm), and the electrolytic cell was connected to a 500 mL glass bottle serving as a reservoir for the reaction mixture (Figure S21, Supporting Information). As illustrated in Figure S22 (Supporting Information), the TiO₂ exhibited a significantly enhanced photocurrent under 1 W cm⁻² illumination, reaching a maximum of 10 mA at 1.6 V_Ag/AgCl. The current can maintain stability under these conditions for the reaction. The synthesis of 3A could be conducted on a gram scale with production rate of 105.2 mmol cm⁻² h⁻¹ by employing this flow setup.

3. Conclusion

In summary, we developed a synergistic heterogeneous/homogeneous PEC strategy to achieve a controllable radical‐induced C─C coupling reaction mediated by Cu²⁺ on TiO₂ photoanodes, which can be performed in the absence of exogenous oxidants. The presence of Cu²⁺ at the TiO₂/electrolyte interfaces facilitates the reaction between 2‐phenylquinoline and cyclohexane, resulting in a maximum yield of 90% with a selectivity of 95% and long‐term stability over 100h. This method can also enable aliphatic compounds with a variety of inert C(sp³)─H configurations to be coupled with a wide range of aromatics. The presence of Cu²⁺ promotes the formation of Cl ^· and restrains the formation of Cl₂. The formation of the CuCl_x complex plays an important role in the high regioselectivity. We believe that the present study would provide insights for future PEC organic transformations.

Conflict of Interest

The authors declare no conflict of interest.

Author Contributions

Q.L. and K.D. contributed equally to this work. Y.Z. directed the project. Q.L. carried out most experiments. Q.L, Y.Z., and K.D. wrote the manuscript, with input from others. K.D. conducted computational studies. S.L. and L.W helped with the IMPS experiments and results simulation. All the authors analyzed the results and reviewed the paper.

Supporting information

Supporting Information

Li Q., Dang K., Wu L., Liu S., Zhang Y., Zhao J., Regulating Chlorine and Hydrogen Atom Transfer for Selective Photoelectrochemical C─C Coupling by Cu‐coordination Effect at Semiconductor/Electrolyte Interfaces. Adv. Sci. 2024, 11, 2408767. 10.1002/advs.202408767

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.