Charge separation and molecule activation promoted by Pd/MIL-125-NH₂ hybrid structures for selective oxidation reactions

Abstract

While heterogeneous photocatalysis is regarded as a sustainable strategy to achieve organic reactions for the production of high-value-added organics, photocatalytic efficiency is generally limited by poor charge separation, dull surface catalytic activity and the consequently caused insufficient generation of active species. Herein, the hybrid structures of metal–organic frameworks (MOFs) incorporated with Pd ultrasmall nanocrystals are synthesized, and their photocatalytic performance is evaluated by the oxidative coupling of benzylamine as a model reaction. The photocatalytic activity of the Pd/MIL-125-NH₂ hybrid material with an optimal Pd loading is greatly superior to that of the pristine MIL-125-NH₂, achieving a high benzylamine conversion rate of 3136 μmol h⁻¹ g_cat⁻¹ at 94.08% conversion. The enhanced charge transfer and separation in Pd/MIL-125-NH₂ are verified by photoluminescence spectroscopy and photoelectrochemical experiments. Photogenerated reactive oxygen species are probed by 3,3′,5,5′-tetramethylbenzidine (TMB) as an indicator and the reactive intermediates in the photocatalytic systems are identified by in situ ESR detection. The results confirm that the optimized Pd/MIL-125-NH₂ hybrid catalyst possesses higher ability for molecule activation and that both superoxide radical (˙O₂⁻) and singlet oxygen (¹O₂) are the dominant oxidative species in the present photocatalytic system. Therefore, the dual advantages of Pd/MIL-125-NH₂ hybrid structures for promoting the photogenerated charge transfer/separation and providing highly catalytic sites for molecule activation contribute to higher photocatalytic activity.

---

Introduction

In the past decades, in order to improve the conversion rate and product selectivity of organic transformations, high temperature and/or high pressure were usually required in thermally-driven catalytic reactions.^1,2 A great deal of non-renewable energy sources are consumed and some toxic or corrosive chemicals are involved in these processes. Considering the increasing interest in green and sustainable chemistry, it is highly desirable that organic reactions with a high conversion rate and product selectivity can be achieved through readily available and environmentally benign reactants (such as O₂ in air) as well as at the low cost of traditional energy consumption.^3–7 The development of photocatalysts has offered great possibilities for the realization of this goal.^8–11 In recent years, more and more efforts are devoted to improving the photocatalytic transformations of organics, such as the selective oxidation of amines into imines and their derivatives, which are high-value-added organic pharmaceutical intermediates.^12–14 However, the photocatalytic selective oxidation of organics still suffers from low conversion rate or/and poor selectivity, because of the sluggish photocatalytic activity or the presence of the reactive oxygen species (ROSs) having strong oxidizing power (such as ˙OH).^15,16 Among various photogenerated ROSs, superoxide radicals (˙O₂⁻) and singlet oxygen (¹O₂) are regarded as the most appropriate candidates for the selective oxidation of organics due to their moderate oxidation ability.^17–20 Therefore, developing photocatalysts being adequate in moderate ROS production and organic molecule activation is the key point to boost the selective oxidation of organics with molecular oxygen under ambient conditions.

Recently, the rapid development of metal–organic frameworks (MOFs) has captured a great deal of research interest in exploiting MOF-based materials as types of promising photocatalysts.^21–23 The high porosity and large specific surface area of MOFs endow them with efficient adsorption and gathering of reactant molecules in pores as well as the relatively rapid charge migration to the pore surface for reactions.²⁴ Nonetheless, the limited active sites in pristine MOFs actually hinder their performance in photocatalytic reactions. Various modification methods were adopted to improve the photocatalytic efficiency, such as incorporating metal particles,^25,26 coupling with inorganic semiconductors,^27,28 grafting photoactive metal complexes,²⁹ and constructing MOF–MOF (bi-MOF) heterostructures.³⁰

Pd nanocrystals have been demonstrated to be effective for the activation of molecular oxygen (O₂) to form ¹O₂ through electron transfer from Pd to the antibonding π* orbital of O₂.³¹ Moreover, the Pd–semiconductor hybrid configuration enabled the improvement of O₂ activation under appropriate illumination due to the increased electron density on the Pd surface in the hybrid structures, which is significant for enhancing the efficiency and selectivity of organic oxidation reactions.³² In a similar way, the integration of Pd nanocrystals into semiconductor-like MOFs is expected to cause enhanced photocatalytic performance due to the transfer and separation of the photoinduced carriers in MOFs as well as the adsorption and activation of O₂ and organic molecules. Moreover, the Pd nanocrystals with ultrasmall sizes (<2 nm) often present prominent catalytic activity because their dramatically high surface-to-volume ratios offer abundant available active/reactive sites.^33,34 The rich ordered pores in MOFs make them suitable solids for confining the ultrasmall Pd nanocrystals, ensuring the uniform dispersions of the Pd nanocrystals during preparation and catalytic reactions.³⁵ The small-sized Pd nanoclusters/nanocrystals encapsulated inside MOFs have showed superior catalytic activity for thermally-driven oxidation reactions,³⁶ hydrogenation,³⁷ hydrogen absorption,³⁸ Suzuki coupling and Heck arylation reactions,^39,40 whereas the synergistic effect between Pd nanocrystals and MOFs on their photocatalysis for the selective oxidation of organics needs to be further studied. In this work, Pd nanocrystals with ultrasmall sizes (<2 nm) are embedded in porous MIL-125-NH₂ (Ti₈O₈(OH)₄(BDC-NH₂)₆, BDC-NH₂ = 2-aminobenzene-1,4-dicarboxylate) to form hybrid nanostructures. The Pd(1.0)/MIL-125-NH₂ hybrid catalyst with an optimal Pd loading content exhibits much higher photocatalytic activity toward the selective oxidation of organic molecules than the pristine MIL-125-NH₂. The enhanced photocatalysis of Pd(1.0)/MIL-125-NH₂ results from the superior charge separation efficiency and the enhanced active species generation due to the charge transfer from MIL-125-NH₂ to Pd ultrasmall nanocrystals upon photoexcitation.

Experimental section

Materials

2-Aminobenzene-1,4-dicarboxylate (BDC-NH₂), titanium(IV) isopropoxide, polyvinylpyrrolidone (PVP, K-88, M_w = 1 300 000), ammonium tetrachloropalladate ((NH₄)₂PdCl₄), ascorbic acid (AA), 4-fluorobenzylamine, 4-methoxybenzylamine, 4-chlorobenzylamine, 4-methylbenzylamine, benzylamine, β-carotene, catalase (≥200 000 unit per g), mannitol, superoxide dismutase (SOD) (≥1400 units per mg dry weight), 3,3′,5,5′-tetramethylbenzidine (TMB), acetate (HAc), sodium acetate anhydrous (NaAc), 5,5-dimethyl-pyrroline-N-oxide (DMPO), 2,2,6,6-tetramethylpiperidine (TEMP), potassium hydrogen phthalate (C₈H₅KO₄), potassium iodide (KI), methanol and acetonitrile. All of the chemicals were of analytical grade and purchased from Aladdin Industrial Corporation.

Synthesis of MIL-125-NH₂

MIL-125-NH₂ was synthesized by a previously reported procedure with minor modification.⁴¹ Typically, 0.628 g BDC-NH₂ and 0.531 mL titanium(IV) isopropoxide were dissolved in the mixed solvent consisting of 15 mL DMF and 15 mL MeOH by magnetic stirring for 30 min. Then, the solution was transferred into a 50 mL Teflon-lined autoclave, which was sealed and heated at 150 °C for 15 h. Finally, the yellow powder of MIL-125-NH₂ was collected, washed with DMF and MeOH, and dried at 60 °C overnight under vacuum.

Synthesis of Pd/MIL-125-NH₂ hybrid structures

The ultrasmall Pd nanocrystals were embedded in MIL-125-NH₂ through mild reduction with ascorbic acid as a reducing agent. 100 mg of the as-prepared MIL-125-NH₂ powder was ultrasonically dispersed in 20 mL of PVP aqueous solution (0.50 g L⁻¹). After 1 h stirring, a certain volume of (NH₄)₂PdCl₄ solution (0.10 M) was injected into the above suspension, followed by another 8 h stirring. The solid particles were collected by centrifugation, and then re-dispersed in 20 mL deionized water. 0.24 g of ascorbic acid was added in this suspension, and the mixture was then stirred at 60 °C for 2 h. The precipitate was collected by centrifugation and washed with deionized water several times. The final product was dried at 60 °C overnight in a vacuum. The Pd(x)/MIL-125-NH₂ hybrid structures with different Pd loading contents (x wt%) were obtained by varying the injected volume of (NH₄)₂PdCl₄ solution, and denoted as Pd(0.2)/MIL-125-NH₂, Pd(0.5)/MIL-125-NH₂, Pd(1.0)/MIL-125-NH₂ and Pd(1.5)/MIL-125-NH₂, respectively.

It should be noted that the Pd loading amounts (x wt%) herein are the theoretical values calculated based on the assumption that all the injected Pd²⁺ ions were transformed into Pd nanocrystals in the Pd(x)/MIL-125-NH₂ hybrid structures. The actual Pd loading amounts should be much lower than the corresponding theoretical values.

Characterization

Powder X-ray diffraction (XRD) was conducted on an X-ray diffractometer (Japan Shimadzu XRD-6000) with Cu Kα radiation. The morphology and microstructures were observed by field emission scanning electronic microscopy (FESEM, Zeiss, Ultra 55) and high resolution transmission electron microscopy (HRTEM, Hitachi S4800). The specific surface area was calculated by the Brunauer–Emmett–Teller (BET) method on the basis of N₂ adsorption/desorption measurement on a Micromeritics ASAP 2020 analyzer, and the pore size distribution was obtained by the Barrett–Joyner–Halenda (BJH) method. X-ray photoelectron spectroscopy (XPS) analysis was conducted using a Thermo Scientific ESCALAB 250Xi X-ray photoelectron spectrometer using Al Kα radiation as the excitation source. UV-vis diffuse reflection spectra (DRS) were recorded on a spectrophotometer (UV-2700, Shimadzu). Photoluminescence (PL) spectroscopy was analyzed a HORIBA HuoroMax-4P spectrometer. The photoelectrochemical experiments were carried out on an electrochemical workstation (CHI660E, Shanghai Chenhua, China) with a standard three electrode system and under irradiation (360–780 nm) using a 300 W xenon lamp (CEL-HXF300). The electron spin resonance (ESR) signals were obtained on an ESR JES-FA200 spectrometer, according to the photocatalytic conditions, with DMPO for the detection of ˙O₂⁻ and TEMP for the detection of ¹O₂, respectively.

Evaluation of photocatalytic properties

The photocatalytic properties of the as-prepared catalysts were evaluated by the oxidative coupling of benzylamine under a 300 W Xe lamp equipped with a reflector that could afford irradiation mainly within the wavelength range of 360–780 nm, as shown in Fig. S1.† The optical power density that irradiated on the reaction solution is 100 mW cm⁻². In a typical reaction, 5 mg catalyst was suspended in 2 mL CH₃CN containing 0.05 mmol benzylamine in a quartz tube with a magnetic stirrer, using molecular O₂ from atmospheric air as the oxidizing agent. After being stirred in the dark for 30 min, the reaction suspension was exposed to irradiation for a certain time. Finally, the reaction solution was filtered through a syringe filter with a 0.22 μm polymer membrane to remove the catalyst particles, and the concentrations of reactants and products were analyzed by gas chromatography (GC, Shimadzu 2014C).

Molecular oxygen activation tests

O₂ activation was tested by 3,3′,5,5′-tetramethylbenzidine (TMB) oxidation according to the evolution in the spectral absorption of TMB solution along the irradiation time.³¹ In a typical test, 15 mg TMB was suspended in 8 mL deionized water, and 24 mL HAc/NaAc buffer solution (pH = 4.7) was added in the suspension. Subsequently, 100 μL of the photocatalyst aqueous suspension (5 g L⁻¹) was injected into the TMB suspension. After the adsorption equilibrium in the dark, the mixture was exposed to the Xe lamp irradiation of 360–780 nm. The reaction was probed by recording the absorbance around 370 nm on a UV-vis spectrophotometer.

To verify the generation of different reactive oxygen species, the contrast experiments with adding the related scavenger were conducted similarly according to the above experimental procedure. The added amounts of scavengers are listed as follows: catalase (3500 unit per mL, 200 μL), superoxide dismutase (2000 unit per mL, 2 mL), mannite (10 mg), and carotene (5 mg).

Results and discussion

Characterization of MIL-125-NH₂ and Pd(x)/MIL-125-NH₂ materials

The Pd(x)/MIL-125-NH₂ hybrid structures are synthesized through a two-step process (Fig. 1a), where MIL-125-NH₂ is synthesized through a solvothermal method and Pd nanocrystals are then embedded in MIL-125-NH₂ by a mild reaction with ascorbic acid as the reducing agent. The XRD pattern of the as-prepared MIL-125-NH₂ sample (Fig. 1b) is in good agreement with the simulated pattern of MIL-125-NH₂ which has a quasi-cubic tetragonal structure.⁴² After loading Pd, the crystal structure of MIL-125-NH₂ remains unchanged, but the diffraction peaks of Pd nanocrystals are unobservable in the XRD patterns of Pd(x)/MIL-125-NH₂ hybrid structures (Fig. 1b and S2†), suggesting the ultrasmall size and high dispersion of Pd nanocrystals in MIL-125-NH₂. The MIL-125-NH₂ sample consists of well-dispersed hexahedron-shaped nanocrystals with a length/width of around 600 nm and thickness of about 150 nm (Fig. S3a and b†). Fig. 1c–f show the SEM and TEM images of the Pd(1.0)/MIL-125-NH₂ sample which was optimized to display the best photocatalytic performance. The Pd(1.0)/MIL-125-NH₂ hybrid structures maintain well the morphology of MIL-125-NH₂. The HRTEM image (Fig. 1e) reveals that the ultrasmall Pd nanocrystals with an average size of ∼1.6 nm are highly dispersed in MIL-125-NH₂. The distribution of Pd element is further demonstrated by HAADF-STEM imaging and EDX element mapping (Fig. 1f and g). During the synthesis process, Pd²⁺ ions were readily adsorbed onto the pores of MIL-125-NH₂ after the long-time (8 h) stirring at room temperature due to the capillary force and the interaction between Pd²⁺ ions and amino groups in MIL-125-NH₂. The subsequent reduction of Pd²⁺ ions to Pd nanocrystals occurred in the MOF pores, which prevented the migration and aggregation of Pd nanocrystals. As a result, the Pd nanocrystals with ultrasmall sizes were produced and stabilized in MIL-125-NH₂. The Pd nanocrystals of the Pd(0.5)/MIL-125-NH₂ sample also presents a similar ultrasmall particle size of around 1.5 nm (Fig. S3c†). However, superfluous Pd²⁺ ions in the solution would lead to the formation of larger Pd nanocrystals attached on the surface of MIL-125-NH₂. The Pd particle sizes of Pd(1.5)/MIL-125-NH₂ are estimated to be around 7.0 nm (Fig. S3d†). In addition, a contrast sample of Pd/MIL-125-NH₂ (Fig. S4†) prepared by the direct reduction-deposition of Pd nanocrystals onto MIL-125-NH₂, without the long-time adsorption before ascorbic-acid reduction, presents the larger particle sizes (∼3.9 nm) and inferior distribution uniformity of Pd nanocrystals, which indirectly proves that the ultrasmall Pd nanocrystals of the Pd(x)/MIL-125-NH₂ samples synthesized by our absorption–reduction method are highly dispersed and embedded inside the pores of MIL-125-NH₂.

Fig. 1 (a) Schematic illustration for the synthesis of Pd/MIL-125-NH₂ hybrid structures. (b) Powder XRD patterns of the simulated MIL-125-NH₂, the as-synthesized MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂. (c) SEM image (inset: lateral view), (d) TEM image, (e) HRTEM image (inset: size distribution of Pd nanocrystals), (f) HAADF-STEM image and (g) the corresponding elemental mapping images of Pd(1.0)/MIL-125-NH₂.

The specific surface areas and porous structures of MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ are determined using the N₂ adsorption–desorption isotherms (Fig. 2). Both samples exhibit type IV isotherms with obvious hysteresis loops, suggesting their mesoporous structures. The specific surface areas of MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ are calculated by the BET method to be 857.1 and 789.7 m² g⁻¹, as listed in Table S1,† and their average pore sizes are 2.3 and 2.8 nm, respectively. The slight decrease in the specific surface area of Pd(1.0)/MIL-125-NH₂ possibly results from the presence of Pd nanocrystals.

Fig. 2 N₂ adsorption–desorption isotherms and pore size distributions (inset) of (a) MIL-125-NH₂ and (b) Pd(1.0)/MIL-125-NH₂.

XPS characterization techniques are applied to analyse the chemical states of surface elements as well as their interactions in MIL-125-NH₂ before and after loading Pd nanocrystals. The survey spectra (Fig. S5†) indicate that the additional Pd signal emerges in Pd(1.0)/MIL-125-NH₂ by comparison with that of the pristine MIL-125-NH₂ sample. The XPS spectra (Fig. 3a–c) of Ti 2p, O 1s, and N 1s of Pd(1.0)/MIL-125-NH₂ present negative shifts with respect to those of pristine MIL-125-NH₂. The Ti 2p spectra (Fig. 3a) of Pd(1.0)/MIL-125-NH₂ involve two pairs of Ti 2p_3/2 and Ti 2p_1/2 peaks, corresponding to Ti(IV) 2p (458.78 and 464.61 eV) and Ti(III) 2p (457.67 and 462.88 eV), respectively, which indicates that the introduction of Pd ultrasmall nanocrystals engenders Ti(III) in MIL-125-NH₂ due to the oxygen vacancies and linker defects.⁴³ The O 1s spectra (Fig. 3b) of both MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ samples are divided into three peaks centered at 529.95 (529.55), 531.55 (531.17), and 533.40 (533.05) eV, respectively, corresponding to O–Ti bonds in the titanium-oxo cluster, O–C and O=C groups in the framework.⁴⁴ In the N 1s spectra of both samples (Fig. 3c),the fitted two peaks at 399.36 (399.20) and 402.28 (400.84) eV can be ascribed to the –N=⁺ and –NH−⁺ groups, respectively.^45,46 Moreover, the N 1s binding energy for –NH−⁺ of Pd(1.0)/MIL-125-NH₂ presents a 1.44 eV shift towards lower values, which can be attributed to the interaction between Pd and the amidogen of MIL-125-NH₂. The Pd 3d spectrum (Fig. 3d) of Pd(1.0)/MIL-125-NH₂ shows a pair of spin–orbit doublet of Pd 3d_5/2 and Pd 3d_3/2 peaks at 338.07 and 343.25 eV, respectively, corresponding to Pd(II).⁴⁷ The positive shift in the binding energies of Pd 3d confirms the strong interaction between Pd and MIL-125-NH₂ in the hybrid structures, which is consistent with the negative shifts in the Ti 2p, O 1s, and N 1s peaks of Pd(1.0)/MIL-125-NH₂ in comparison to the pristine MIL-125-NH₂.³⁶ The electron transfer from Pd ultrasmall nanocrystals to MIL-125-NH₂ reduces more Ti(III) sites and oxygen vacancies, which are beneficial to catalysis by promoting charge transfer and providing more active sites.³⁹

Fig. 3 High-resolution XPS spectra of (a) Ti 2p, (b) O 1s, (c) N 1s and (d) Pd 3d for MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂, respectively.

Photoabsorption and photogenerated carrier separation

The photoabsorption properties of the MIL-125-NH₂ sample and different Pd(x)/MIL-125-NH₂ hybrid materials are determined by DRS analysis (Fig. 4a). The pristine MIL-125-NH₂ presents strong absorption in the range of wavelengths less than ca. 475 nm. The absorption bands of Pd(x)/MIL-125-NH₂ hybrid structures show a little shift towards longer wavelengths due to interband transition and light scattering of Pd nanocrystals.^48,49 The HOMO/LUMO gap of the MIL-125-NH₂ sample is calculated to be 2.71 eV by a Tauc plot (Fig. S6†). To elucidate the semiconductor-like character of MIL-125-NH₂, its Mott–Schottky plot was measured (Fig. 4b), and the positive slope of the plot demonstrates the n-type semiconductor feature of MIL-125-NH₂. The flat band potential (V_fb) of MIL-125-NH₂ was determined from the intersection value to be −1.09 V vs. Ag/AgCl. Accordingly, the LUMO potential of MIL-125-NH₂ was determined to be −0.89 V vs. NHE, and the HOMO potential was calculated to be 1.81 V vs. NHE, on the basis of the HOMO/LUMO gap of 2.71 eV.^50,51

Fig. 4 (a) UV-vis DSR of MIL-125-NH₂ and different Pd(x)/MIL-125-NH₂ samples, (b) Mott–Schottky plots and HOMO/LUMO levels (inset) of the pristine MIL-125-NH₂.

The PL spectra of MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ are measured to reflect the photogenerated carrier separation of Pd/MIL-125-NH₂ hybrid structures. Under the excitation wavelength of 300 nm, the Pd(1.0)/MIL-125-NH₂ hybrid structures show a quenched fluorescence emission intensity compared with pure MIL-125-NH₂ (Fig. 5a), suggesting the suppressed recombination of photogenerated electron–hole pairs due to the efficient electron transfer from MIL-125-NH₂ to Pd NPs. The time-resolved photoluminescence (TRPL) decay profiles of the two samples are shown in Fig. 5b, to probe their photogenerated carrier dynamics. The TRPL decay profiles can be fitted by the biexponential decay function:

I_(t) = A₁ exp(−t/τ₁) + A₂ exp(−t/τ₂) (1)

where τ_i and A_i represent the corresponding lifetime and weighing factor, respectively.⁵² The average PL lifetimes (τ_m) of MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ are calculated to be 7.2 ns and 5.9 ns (Table S2†), respectively. In the present case, the shorter PL lifetime of the Pd(1.0)/MIL-125-NH₂ hybrid structures should be attributed to the faster interfacial electron transfer from MOFs to Pd nanocrystals. Both the quenched PL intensity and the shorter PL lifetime consistently reflect the faster interfacial electron transfer and more efficient charge separation in the Pd(1.0)/MIL-125-NH₂ hybrid structures.^51,53 Photoelectrochemical tests are also performed on the MIL-125-NH₂ and Pd(x)/MIL-125-NH₂ samples under the Xe lamp irradiation, to further reveal their charge separation properties. All the Pd(x)/MIL-125-NH₂ electrodes exhibit notably increased photocurrent intensity than the pristine MIL-125-NH₂ one (Fig. 5c), confirming more efficient charge separation in Pd(x)/MIL-125-NH₂ hybrid structures. Moreover, the comparison in Nyquist plots (Fig. 5d) of Pd(x)/MIL-125-NH₂ and MIL-125-NH₂ indicates that the incorporation of Pd nanocrystals in MIL-125-NH₂ can remarkably reduce the interfacial charge transfer resistance and accelerate the photogenerated charge transfer. The high efficiency of charge transfer and separation in Pd(x)/MIL-125-NH₂ hybrid structures is undoubtedly advantageous for enhancing photocatalytic efficiency. Although both the photocurrent intensity and the charge transfer efficiency increase slightly along with the increased Pd loading amounts, the size-determined surface catalysis of Pd nanocrystals should also be considered simultaneously. The smaller Pd nanocrystals would show higher catalytic activity toward the adsorption and activation of O₂ as well as organic molecules.

Fig. 5 (a) Steady PL spectra under the excitation wavelength of 300 nm and (b) time-resolved photoluminescence (TRPL) decay profiles of MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂. (c) Photocurrents and (d) EIS Nyquist plots of MIL-125-NH₂ and Pd(x)/MIL-125-NH₂ with different Pd loading amounts.

Photocatalytic selective oxidation of benzylamines

The above analyses predicted the great potential of Pd(x)/MIL-125-NH₂ hybrid materials as photocatalysts. The photocatalytic performance of Pd(x)/MIL-125-NH₂ with varied Pd loading amounts is assessed by the selectively oxidative coupling of benzylamine to imine (Fig. 6a). Photocatalytic reactions were carried out under ambient conditions and with an irradiation of 360–780 nm for 3 h. As illustrated in Fig. 6b and Table S3,† the pristine MIL-125-NH₂ shows a moderate activity to achieve a benzylamine conversion of 68.82% and a dissatisfactory selectivity of 86.01% for producing the corresponding imine. With introducing ultrasmall Pd nanocrystals, Pd(0.2)/MIL-125-NH₂, Pd(0.5)/MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ hybrid photocatalysts exhibit both the incremental activity and the excellent selectivity (>99%) for the oxidative coupling of benzylamine. The highest benzylamine conversion of 94.08% was achieved in the presence of the Pd(1.0)/MIL-125-NH₂ catalyst after 3 h reaction, corresponding to a conversion rate of 3136 μmol h⁻¹ g_cat⁻¹, which is about 1.37 times that of pristine MIL-125-NH₂. However, with further increasing Pd-loading amount, the Pd(1.5)/MIL-125-NH₂ sample exhibited a much decreased benzylamine conversion (62.46%) in spite of maintaining the excellent selectivity. The variations in the photocatalysis of Pd(x)/MIL-125-NH₂ hybrid structures along with their varied Pd loading are possibly due to that the properly increased Pd content provides more active sites in the Pd(1.0)/MIL-125-NH₂ catalyst, but excess Pd loading leads to the formation of larger Pd nanocrystals which blocks the structural pores of MIL-125-NH₂ and reduces the efficient active sites. As a result, the Pd(1.0)/MIL-125-NH₂ sample exhibits the best photocatalytic activity for selective oxidation. Then the photocatalytic performance of Pd(1.0)/MIL-125-NH₂ is determined under different conditions (Fig. 6c and Table S4†). The oxidation of benzylamine cannot proceed in the absence of either catalyst or light irradiation. Under visible light irradiation (λ > 420 nm), the Pd(1.0)/MIL-125-NH₂ hybrid catalyst also shows higher benzylamine conversion (77.41% after 3 h) and product selectivity (99.70%) than pristine MIL-125-NH₂ (52.61% for conversion and 82.61% for selectivity), confirming the significant role of Pd ultrasmall nanocrystals in improving the photocatalytic activity and selectivity. In addition, the result for photocatalytic selective oxidation of benzylamine in air is almost identical to that obtained under an O₂ atmosphere, indicating the high oxygen activation ability of the Pd(1.0)/MIL-125-NH₂ hybrid catalyst. Therefore, selective oxidation reactions are carried out in air in view of the sustainable chemistry. When the photocatalytic reaction is conducted in a N₂ atmosphere, the benzylamine conversion decreases to 42.74% after the same time, indicating that the molecular O₂ is crucial for the oxidative coupling of benzylamine to imine. Furthermore, the photocatalytic benzylamine conversion over Pd(1.0)/MIL-125-NH₂ is investigated under different monochromatic lights from the Xe lamp with optical filters. As shown in Fig. 6d and Table S5,† the variation tendency of benzylamine conversion under various monochromatic wavelengths from 365 to 700 nm coincides with the photoabsorption intensity of the Pd(1.0)/MIL-125-NH₂ catalyst at the corresponding wavelengths. The photocatalytic efficiency of Pd(1.0)/MIL-125-NH₂ is dominated in the wavelength range of 365–450 nm corresponding to its intensive photo-absorption in this region. Moreover, the selective oxidation of benzylamine can proceed slowly in the visible light region from 500 to 700 nm, in agreement with the visible light harvesting capability of this photocatalyst. The photoabsorption of Pd(1.0)/MIL-125-NH₂ hybrid structures covers almost the whole UV-visible region of the solar spectrum, which contributes to achieving high utilization of solar energy for photocatalytic reactions.

Fig. 6 (a) Photocatalytic oxidative coupling of benzylamines to corresponding imines, (b) conversion and selectivity of benzamine oxidation after 3 h over the Pd(x)/MIL-125-NH₂ photocatalyst with different Pd loading amounts, (c) the results of selective benzylamine oxidation under different conditions, (d) the relationship between conversion of benzylamine and light harvesting of Pd(1.0)/MIL-125-NH₂, and (e) photocatalytic selective oxidation of benzylamines with different substituent groups catalyzed by Pd(1.0)/MIL-125-NH₂.

On the basis of the optimal photocatalytic performance of the Pd(1.0)/MIL-125-NH₂ hybrid catalyst, the selective oxidation of benzylamines with various substituents is carried out under the same conditions for 3 h (Fig. 6e and Table S6†). Among various substrates, unsubstituted benzylamine holds the highest conversion rate (94.08%), compared with benzylamines with 4-methoxy (91.76%), 4-methyl (77.22%), 4-F (84.82%) and 4-Cl (80.22%) for producing the corresponding imines with high selectivity (>99%), indicating that the universality of the Pd(1.0)/MIL-125-NH₂ photocatalyst for the oxidative coupling of amines. The differences in the conversion rates of various substrates are determined by both steric and electronic effects of the reactants. Finally, the recyclability of the Pd(1.0)/MIL-125-NH₂ photocatalyst is tested by the selective oxidation of benzylamine under the same irradiation (Fig. S7a and Table S7†). The cyclic experiment demonstrates that the activity and selectivity of the Pd(1.0)/MIL-125-NH₂ catalyst are well maintained after 3 cycles except a slight decrease in the activity due to the loss of the photocatalyst during the separation using a centrifuge. Moreover, the XRD patterns (Fig. S7b†) of Pd(1.0)/MIL-125-NH₂ before and after the three cycles remained unchanged, suggesting that the crystal structure of the catalyst are well preserved and it has good stability to undergo a long-term reaction. In addition, the Pd(x)/MIL-125-NH₂ catalysts present the same variation tendency in their photocatalysis toward selective oxidation of benzyl alcohol (Table S8†) along with the varied Pd loading contents, and the selectivity for producing benzaldehyde is 100%. Meanwhile, benzyl alcohol oxidation needs a much longer reaction time due to the moderate oxidizing power of the present photocatalytic system, which can avoid the overoxidation of organic molecules.

Insight into the photocatalytic mechanism

The oxygen activation process plays an important role in the selective oxidation reactions with molecular oxygen as an oxidant. To account for the enhanced activity and selectivity of Pd(1.0)/MIL-125-NH₂ hybrid structures toward the selective oxidation of benzylamine, molecular oxygen activation was probed by TMB oxidation through monitoring the evolution in the spectral absorption of TMB solution along the irradiation time. As shown in Fig. S8,† upon one-electron oxidation by the reactive oxygen species (ROS) generated from the Pd(1.0)/MIL-125-NH₂ photocatalyst, the colorless TMB (λ_max = 285 nm) gradually turns into a blue charge-transfer complex (λ_max = 370, 650 nm) of the parent diamine and the diimine oxidation product.^17,54 The charge-transfer complex can both exist in rapid equilibrium with the radical cation. The time-dependent absorption spectra of TMB solution (Fig. 7a) reflect the higher O₂ activation efficiency of Pd(1.0)/MIL-125-NH₂ than that of pristine MIL-125-NH₂. Subsequently, in order to identify the type of ROS generated in the photocatalytic system, different scavengers including superoxide dismutase (SOD), carotene, mannitol, and catalase were introduced into the TMB solution to consume superoxide (˙O₂⁻), singlet oxygen (¹O₂), hydroxyl radicals (˙OH), and hydrogen peroxide (H₂O₂), respectively (Fig. 7b, S9 and S10†). For MIL-125-NH₂, TMB oxidation is significantly suppressed by the addition of SOD but little influenced by other scavengers, suggesting that ˙O₂⁻ is the primary ROS generated during photocatalytic O₂ activation over MIL-125-NH₂. In the case of the Pd(1.0)/MIL-125-NH₂ hybrid catalyst, TMB oxidation can be remarkably restrained by both SOD and carotene, indicating that ˙O₂⁻ and ¹O₂ are dominated together in this reaction system. In consequence, the incorporation of Pd ultrasmall nanocrystals into MIL-125-NH₂ can enhance the O₂ activation efficiency as well as facilitate the generation of ¹O₂ species, thereby greatly improving the photocatalytic activity and selectivity toward selective oxidation of benzylamines.

Fig. 7 (a) Time-dependent absorption spectra of TMB oxidation with MIL-125-NH₂ and Pd 1.0/MIL 125-NH₂ in air, and (b) the absorbance evolutions monitored for TMB oxidation at 370 nm over MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ in the presence of different scavengers.

Furthermore, ESR detections provide more specific information about the photogenerated ROS and other radical intermediates during the photocatalytic oxidative coupling of benzylamine. With TEMP as a trapping agent, the 1 : 1 : 1 triplet signals are the characteristic of 2,2,6,6-tetramethylpiperidine-N-oxyl (TEMPO), corresponding to the generation of ¹O₂.²⁰ With MIL-125-NH₂ as a photocatalyst, the weak triple peaks remained the same whether in the dark or under irradiation (Fig. 8a), which indicates that the pristine MIL-125-NH₂ cannot effectively produce ¹O₂ by photoexcitation. The emergence of weak triple signals in the dark is possibly due to the inevitable oxidation of TEMP upon exposure to air.⁵⁵ However, in the case of the Pd(1.0)/MIL-125-NH₂ hybrid photocatalyst, the triple-signal intensity is significantly intensified upon illumination (Fig. 8b), confirming the photogeneration of ¹O₂ on Pd(1.0)/MIL-125-NH₂. After introducing benzylamine into the reaction system, the triple signal is weakened because of the consumption of ¹O₂ by the selective oxidation of benzylamine. In addition, both MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ can create the sextet signal (Fig. 8c) under illumination, which proves the formation of DMPO–OOH, a spin adduct derived from DMPO–˙O₂⁻.⁵⁶ Moreover, Pd(1.0)/MIL-125-NH₂ exhibits a stronger sextet signal than that generated by MIL-125-NH₂ under the same conditions, demonstrating the higher ability of Pd(1.0)/MIL-125-NH₂ for producing ˙O₂⁻ radical. The addition of benzylamine into the system also remarkably suppresses the DMPO–˙O₂⁻ signal (●) from Pd(1.0)/MIL-125-NH₂ (Fig. 8d), and a set of strong ESR signals (♦) appear simultaneously, which can be assigned to the benzyl-type nitrogen-centered or carbon-centered radicals.⁵⁷ In contrast, pristine MIL-125-NH₂ shows a very low activity to generate the benzyl-type nitrogen-centered or carbon-centered radicals. Therefore, the Pd(1.0)/MIL-125-NH₂ hybrid photocatalyst possesses higher abilities for activating both O₂ and benzylamine molecules, which jointly boosted the photocatalytic efficiency for benzylamine oxidation.

Fig. 8 ESR detection of ¹O₂ generated by (a) MIL-125-NH₂ and (b) Pd(1.0)/MIL-125-NH₂ with TEMP as the ¹O₂ trapping agent. (c) ESR detection of ˙O₂⁻ generated by MIL-125-NH₂ and Pd(1.0)/MIL-125-NH₂ with DMPO as the ˙O₂⁻ trapping agent. (d) ESR signals of DMPO–˙O₂⁻ (●) and DMPO–nitrogen (or carbon)-centered radicals (♦).

Based on the above results, the photocatalytic mechanism for selective oxidation of benzylamine to imine is illustrated in Fig. 9. Upon excitation under light irradiation, the electron–hole pairs are generated in MIL-125-NH₂ by ligand-to-metal charge transfer (LMCT) where the electron transfer occurs from an organic ligand to Ti⁴⁺, similar to the generation of electrons and holes in semiconductors.⁴¹ Ti⁴⁺ is reduced to Ti³⁺ after accepting electrons, leaving holes in the organic ligand. Ti³⁺ is an active site for the absorption and reduction of O₂ to form ˙O₂⁻ radicals.⁵⁸ Simultaneously, the holes left in ligands oxidize benzylamine to PhCH₂NH₂˙⁺. Proton transfer takes place from PhCH₂NH₂˙⁺ to ˙O₂⁻, resulting in the appearance of PhCH₂NH˙ or PhCH˙NH₂ and HO₂˙, and the subsequent transformation into PhCH=NH and H₂O₂. The final product of imine is produced by the condensation reaction of a PhCH=NH intermediate with a benzylamine molecule. In the case of anchoring Pd ultrasmall nanocrystals into MIL-125-NH₂, the electrons of Ti³⁺ sites in MIL-125-NH₂ readily flow to the Fermi level of Pd, increasing the electron intensity on the surface of Pd nanocrystals, which promotes the activation of O₂ to ¹O₂ on Pd nanocrystals.³² The generated ¹O₂ having high electrophilicity can oxidize amine to a PhCH=NH intermediate, along with the release of H₂O₂.^55,59 As a consequence, both ˙O₂⁻ and ¹O₂ are the dominating reactive oxygen species contributing the selective oxidation of benzylamine in the presence of Pd(1.0)/MIL-125-NH₂ as a photocatalyst. H₂O₂ production has been detected in the photocatalytic reaction solution of benzylamine oxidation (Fig. S11†). The evolution trends of both benzylamine conversion and H₂O₂ concentration are coincident, suggesting that the benzylamine oxidation process is accompanied by H₂O₂ production. Moreover, the H₂O₂ amount produced by the Pd(1.0)/MIL-125-NH₂ hybrid catalyst is much higher than that by pristine MIL-125-NH₂, confirming the high efficiency of O₂ activation on Pd(1.0)/MIL-125-NH₂ catalyst. Therefore, the Pd/MIL-125-NH₂ hybrid nanostructures have been verified to possess remarkable superiority in the spatial separation of photoinduced electron–hole pairs as well as the activation of O₂ and benzylamine molecules. These abilities endow the Pd/MIL-125-NH₂ hybrid material with excellent photocatalytic activity toward the selective oxidation of organics.

Fig. 9 (a) Schematic illustration for the photocatalytic mechanism of Pd/MIL-125-NH₂ hybrid structures, and (b) the possible pathway of oxidative coupling of benzylamine.

Conclusions

In summary, Pd/MIL-125-NH₂ hybrid structures are synthesized by controllably anchoring Pd ultrasmall nanocrystals (<2 nm) in the structural pores of MIL-125-NH₂. The optimal photocatalyst of Pd(1.0)/MIL-125-NH₂ exhibits excellent activity, selectivity and recyclability in the oxidative coupling of aromatic amines under ambient conditions, far surpassing the pristine MIL-125-NH₂. The introduction of ultrasmall Pd nanocrystals plays a significant role in promoting the generation and separation of photo-induced electron–hole pairs as well as the catalytic activation of substrate molecules. The enhancement in charge-carrier kinetics and ROS generation in the system of the Pd/MIL-125-NH₂ hybrid catalyst are demonstrated by the joint observations from PL spectroscopy, photoelectrochemical tests, and TMB oxidation experiments. The simultaneous generation of ˙O₂⁻ and ¹O₂ is dominated in this system. The reactive intermediates generated during the photocatalytic oxidative coupling of benzylamine are further verified by the in situ ESR technique and H₂O₂ detection, providing an in-depth understanding on the photocatalytic processes. Benefiting from the dual advantages of promoted charge separation and efficient O₂ activation, the Pd/MIL-125-NH₂ hybrid material with an appropriate Pd content is identified as a promising photocatalyst in solar-driven selective oxidation of organics. This work not only demonstrates the potential of Pd/MOF hybrid structures as photocatalysts for selective organic transformations but also provides an in-depth understanding on charge separation and molecule activation in such hybrid photocatalysts.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 21471004), the Natural Science Foundation of Zhejiang Province of China (No. LY19B010005), and the Science Foundation of Zhejiang Sci-Tech University (No. 17062002-Y).

Notes and references

1. F. A. Westerhaus, R. V. Jagadeesh, G. Wienhofer, M. M. Pohl, J. Radnik, A. E. Surkus, J. Rabeah, K. Junge, H. Junge, M. Nielsen, A. Bruckner and M. Beller, Nat. Chem., 2013, 5, 537–543.
2. I. Sorribes, L. Liu and A. Corma, ACS Catal., 2017, 7, 2698–2708.
3. J. Long, L. Wang, X. Gao, C. Bai, H. Jiang and Y. Li, Chem. Commun., 2012, 48, 12109–12111.
4. A. N. Campbell and S. S. Stahl, Acc. Chem. Res., 2012, 45, 851–863.
5. Z. Guo, B. Liu, Q. Zhang, W. Deng, Y. Wang and Y. Yang, Chem. Soc. Rev., 2014, 43, 3480–3524.
6. B. Zhao, Y. Huang, D. Liu, Y. Yu and B. Zhang, Sci. China: Chem., 2019 DOI:10.1007/s11426-019-9620-1.
7. X. Chong, C. Liu, Y. Huang, C. Huang and B. Zhang, Natl. Sci. Rev., 2019 DOI:10.1093/nsr/nwz146.
8. A. Dhakshinamoorthy, Z. Li and H. Garcia, Chem. Soc. Rev., 2018, 47, 8134–8172.
9. A. Hu, J.-J. Guo, H. Pan and Z. Zuo, Science, 2018, 361, 668–672.
10. J. Kou, C. Lu, J. Wang, Y. Chen, Z. Xu and R. S. Varma, Chem. Rev., 2017, 117, 1445–1514.
11. Y. Xin, Y. Huang, K. Lin, Y. Yu and B. Zhang, Sci. Bull., 2018, 63, 601–608.
12. B. Yuan, R. Chong, B. Zhang, J. Li, Y. Liu and C. Li, Chem. Commun., 2014, 50, 15593–15596.
13. H. Chen, C. Liu, M. Wang, C. Zhang, N. Luo, Y. Wang, H. Abroshan, G. Li and F. Wang, ACS Catal., 2017, 7, 3632–3638.
14. H. Wei, Z. Guo, X. Liang, P. Chen, H. Liu and H. Xing, ACS Appl. Mater. Interfaces, 2019, 11, 3016–3023.
15. Y. Nosaka and A. Y. Nosaka, Chem. Rev., 2017, 117, 11302–11336.
16. H. Hao, L. Zhang, W. Wang and S. Zeng, Catal. Sci. Technol., 2018, 8, 1229–1250.
17. Y.-Z. Chen, Z. U. Wang, H. Wang, J. Lu, S.-H. Yu and H.-L. Jiang, J. Am. Chem. Soc., 2017, 139, 2035–2044.
18. H. Li, F. Qin, Z. Yang, X. Cui, J. Wang and L. Zhang, J. Am. Chem. Soc., 2017, 139, 3513–3521.
19. H. Wang, S. Chen, D. Yong, X. Zhang, S. Li, W. Shao, X. Sun, B. Pan and Y. Xie, J. Am. Chem. Soc., 2017, 139, 4737–4742.
20. X. Sun, X. Luo, X. Zhang, J. Xie, S. Jin, H. Wang, X. Zheng, X. Wu and Y. Xie, J. Am. Chem. Soc., 2019, 141, 3797–3801.
21. J. D. Xiao and H. L. Jiang, Acc. Chem. Res., 2019, 52, 356–366.
22. G. Li, S. Zhao, Y. Zhang and Z. Tang, Adv. Mater., 2018, 30, 1800702.
23. J. A. Johnson, J. Luo, X. Zhang, Y.-S. Chen, M. D. Morton, E. Echeverría, F. E. Torres and J. Zhang, ACS Catal., 2015, 5, 5283–5291.
24. L. Jiao, Y. Wang, H. L. Jiang and Q. Xu, Adv. Mater., 2017, 1703663.
25. Q. Yang, Q. Xu and H. L. Jiang, Chem. Soc. Rev., 2017, 46, 4774–4808.
26. J. D. Xiao, L. Han, J. Luo, S. H. Yu and H. L. Jiang, Angew. Chem., Int. Ed., 2018, 57, 1103–1107.
27. H.-Q. Xu, S. Yang, X. Ma, J. Huang and H.-L. Jiang, ACS Catal., 2018, 8, 11615–11621.
28. J. W. Yoon, D. H. Kim, J.-H. Kim, H. W. Jang and J.-H. Lee, Appl. Catal., A, 2019, 244, 511–518.
29. Y. Isaka, Y. Kondo, Y. Kuwahara, K. Mori and H. Yamashita, Catal. Sci. Technol., 2019, 9, 1511–1517.
30. L. Liu, L. Zhang, F. Wang, K. Qi, H. Zhang, X. Cui and W. Zheng, Nanoscale, 2019, 11, 7554–7559.
31. R. Long, K. Mao, X. Ye, W. Yan, Y. Huang, J. Wang, Y. Fu, X. Wang, X. Wu, Y. Xie and Y. Xiong, J. Am. Chem. Soc., 2013, 135, 3200–3207.
32. R. Long, K. Mao, M. Gong, S. Zhou, J. Hu, M. Zhi, Y. You, S. Bai, J. Jiang, Q. Zhang, X. Wu and Y. Xiong, Angew. Chem., Int. Ed., 2014, 53, 3205–3209.
33. D. Gao, H. Zhou, J. Wang, S. Miao, F. Yang, G. Wang, J. Wang and X. Bao, J. Am. Chem. Soc., 2015, 137, 4288–4291.
34. X. Pei, Y. Deng, Y. Li, Y. Huang, K. Yuan, J.-F. Lee, T.-S. Chan, J. Zhou, A. Lei and L. Zhang, Nanoscale, 2018, 10, 14719–14725.
35. L. Chen, W. Huang, X. Wang, Z. Chen, X. Yang, R. Luque and Y. Li, Chem. Commun., 2017, 53, 1184–1187.
36. P. Zhang, C. Chen, X. Kang, L. Zhang, C. Wu, J. Zhang and B. Han, Chem. Sci., 2018, 9, 1339–1343.
37. X. Qiu, J. Chen, X. Zou, R. Fang, L. Chen, Z. Chen, K. Shen and Y. Li, Chem. Sci., 2018, 9, 8962–8968.
38. A. Malouche, G. Blanita, D. Lupu, J. Bourgon, J. Nelayah and C. Zlotea, J. Mater. Chem. A, 2017, 5, 23043–23052.
39. D. Sun and Z. Li, J. Phys. Chem. C, 2016, 120, 19744–19750.
40. K. Titov, D. B. Eremin, A. S. Kashin, R. Boada, B. E. Souza, C. S. Kelley, M. D. Frogley, G. Cinque, D. Gianolio, G. Cibin, S. Rudić, V. P. Ananikov and J.-C. Tan, ACS Sustainable Chem. Eng., 2019, 7, 5875–5885.
41. Y. Fu, D. Sun, Y. Chen, R. Huang, Z. Ding, X. Fu and Z. Li, Angew. Chem., Int. Ed., 2012, 51, 3364–3367.
42. A. P. Smalley, D. G. Reid, J. C. Tan and G. O. Lloyd, CrystEngComm, 2013, 15, 9368–9371.
43. B. Zhang, J. Zhang, X. Tan, D. Shao, J. Shi, L. Zheng, J. Zhang, G. Yang and B. Han, ACS Appl. Mater. Interfaces, 2018, 10, 16418–16423.
44. S. Venkateswarlu, A. Panda, E. Kim and M. Yoon, ACS Appl. Nano Mater., 2018, 1, 4198–4210.
45. H. Wang, X. Yuan, Y. Wu, G. Zeng, X. Chen, L. Leng, Z. Wu, L. Jiang and H. Li, J. Hazard. Mater., 2015, 286, 187–194.
46. R. Bibi, H. Huang, M. Kalulu, Q. Shen, L. Wei, O. Oderinde, N. Li and J. Zhou, ACS Sustainable Chem. Eng., 2018, 7, 4868–4877.
47. F. R. Fortea-Pérez, M. Mon, J. Ferrando-Soria, M. Boronat, A. Leyva-Pérez, A. Corma, J. M. Herrera, D. Osadchii, J. Gascon, D. Armentano and E. Pardo, Nat. Mater., 2017, 16, 760–766.
48. S. Sarina, H.-Y. Zhu, Q. Xiao, E. Jaatinen, J. Jia, Y. Huang, Z. Zheng and H. Wu, Angew. Chem., Int. Ed., 2014, 53, 2935–2940.
49. R. Wang, B. Li, Y. Xiao, X. Tao, X. Su and X. Dong, J. Catal., 2018, 364, 154–165.
50. C. H. Hendon, D. Tiana, M. Fontecave, C. Sanchez, L. D'arras, C. Sassoye, L. Rozes, C. Mellot-Draznieks and A. Walsh, J. Am. Chem. Soc., 2013, 135, 10942–10945.
51. S. Gao, W. Cen, Q. Li, J. Li, Y. Lu, H. Wang and Z. Wu, Appl. Catal., A, 2018, 227, 190–197.
52. X. Wang, C. Liow, A. Bisht, X. Liu, T. C. Sum, X. Chen and S. Li, Adv. Mater., 2015, 27, 2207–2214.
53. M. V. Dozzi, A. Candeo, G. Marra, C. D'Andrea, G. Valentini and E. Selli, J. Phys. Chem. C, 2018, 122, 14326–14335.
54. P. Josephy, T. Eling and R. Mason, J. Biol. Chem., 1982, 257, 3669–3675.
55. C. Xu, H. Liu, D. Li, J. H. Su and H. L. Jiang, Chem. Sci., 2018, 9, 3152–3158.
56. H. Wang, D. Yong, S. Chen, S. Jiang, X. Zhang, W. Shao, Q. Zhang, W. Yan, B. Pan and Y. Xie, J. Am. Chem. Soc., 2018, 140, 1760–1766.
57. R. Wang, G. Qiu, Y. Xiao, X. Tao, W. Peng and B. Li, J. Catal., 2019, 374, 378–390.
58. D. Sun, L. Ye and Z. Li, Appl. Catal., A, 2015, 164, 428–432.
59. E. Baciocchi, T. Del Giacco, O. Lanzalunga and A. Lapi, J. Org. Chem., 2007, 72, 9582–9589.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9cy01690b

---