Self-assembly of a heterogeneous microreactor with carbon dots embedded in Ti-MOF derived ZnIn₂S₄/TiO₂ microcapsules for efficient CO₂ photoreduction

Abstract

The assembly of the heterogeneous microreactor is a promising approach for CO₂ photoreduction attributed to its abundant microchannel, intimate contact, high exposed surface area, and favorable heat-mass transfer. Herein, we developed a metal–organic framework (MOF) derived in situ transformation strategy to construct a carbon dot (CD)-decorated ZnIn₂S₄/TiO₂ (CDs/ZIS/TiO₂) microreactor. Taking advantages of this hierarchical structure, the CDs/ZnIn₂S₄/TiO₂ microreactor exhibits significantly enhanced photocatalytic CO₂ reduction activity with a CH₄ yield of 14.9 μmol g⁻¹ h⁻¹ and CH₄ selectivity of 75.6% in the absence of a sacrificial agent, where the electron consumption rate (R_electron) of 157.6 μmol g⁻¹ h⁻¹ is 1.9 and 18.3 times higher than those of ZIS(60)/TiO₂ and bare ZnIn₂S₄, respectively. The combination of transient photo-induced voltage (TPV), in situ Fourier transform infrared and electron spin resonance (ESR) spectra illustrate the photocatalytic mechanism and the effect of CDs on the electron transfer behavior. This work emphasizes a facile technique for developing a CD-based microreactor to achieve high-efficiency photocatalytic CO₂ reduction performance.

---

1. Introduction

Solar-driven CO₂ reduction is an attractive and efficient way to directly convert CO₂ into fuels and high value-added products, such as CO, CH₄, CH₃OH, and C₂H₄.^1–3 Among the C1 products, CH₄ as a popular and clean energy source is one of the most important fuels, and so far, enormous efforts have been focused on the construction of semiconductor catalysts to improve the yield and selectivity of CH₄.^4–7 Ternary chalcogenide ZnIn₂S₄ with a layered structure can convert CO₂ into useful fuels (e.g., CO, H₂, CH₄) with considerable performance due to their appropriate band gap structure and great visible-light absorption ability.⁸ Although ZnIn₂S₄-based catalysts present a high CO₂-to-(CO + H₂) conversion efficiency, the photocatalytic activity and selectivity of CH₄ in their gas products are insufficient.^9–11

On account of the enhancement of CH₄ production, the development of a ZnIn₂S₄-based heterogeneous microreactor can be considered an effective strategy due to its high surface-to-volume ratio, abundant microchannels, and favorable heat-mass transfer. This heterogeneous microreactor can strengthen the electron trapping ability, and thus requires more electrons and protons to generate CH₄. Metal–organic frameworks (MOFs) with fascinating topology, large pore volume, and chemical adjustability provide an excellent platform to fabricate the semiconductors with a microcapsule structure^12,13 including metal oxides,¹⁴ metal sulfides,¹⁵ layered double hydroxides (LDHs) by using ion-exchange or solvothermal method,¹⁶ which enables the encapsulation of numerous nanoparticles or nanosheets on the capsules. For instance, Bibi et al. found that thioacetamide (TAA) can decompose MIL-125 to form TiO₂/CdS capsules after the post-solvothermal method, which exhibited enhanced photocatalytic activity.¹⁷ MOF-derived microcapsules as a crucial part of the microreactor can provide high porosity, large inner space, the enhanced spatial density of active site as well as an unimpeded electron transport channel.

Carbon dots (CDs) have both remarkable light-harvesting and electron-transfer/reservoir abilities, which may act as an important component of the microreactor.^18–24 With CDs in a microreactor system, the transportation of photogenerated electrons in the encapsulation system will become faster and more efficient.²⁵ CDs as electron storage containers may capture more electrons from the semiconductor catalyst and regulate the local charge distribution, thus acquiring more electrons to produce CH₄.²⁶ Also, CDs may facilitate the water oxidation reaction to provide more protons for CH₄ instead of H₂.²⁷ It is predictable that rationally designing and assembling of the CD-modified ZnIn₂S₄ microreactor should be a promising approach to achieve the high activity and selectivity of CO₂-to-CH₄ conversion. While it is still a big challenge to assemble a productive heterogeneous microreactor by combining active components together through an effective and facile fabrication.

Herein, a carbon dot (CD)-modified ZnIn₂S₄/TiO₂ (CDs/ZIS/TiO₂) microreactor with hollow nanocages and a multi-shell structure was obtained by an ingenious one-step reaction strategy, in which the formed TiO₂ microcapsule benefited from the corrosion of NH₂-MIL-125 caused by thioacetamide (TAA). The as-prepared CDs/ZIS/TiO₂ microreactors exhibit excellent CO₂ photoreduction with CH₄ yield (14.9 μmol g⁻¹ h⁻¹) without a sacrificial agent, which is much higher than that of pure ZnIn₂S₄. Besides, the CDs/ZIS/TiO₂ microreactor presents a highly stable photocatalytic activity after six successive runs. The well-defined architecture with multi-shell structure, high surface area, and large inner space can improve the light absorption ability, shorten the diffusion pathway, and facilitate charge transfer. Importantly, CDs as electron “reservoirs” can effectively capture electrons and inhibit charge recombination. The proposed photocatalytic mechanism and charge transfer process were studied in detail using transient photo-induced voltage (TPV), in situ Fourier transform infrared and electron spin resonance (ESR) spectra.

2. Experimental section

2.1 Synthesis

2.1.1 Synthesis of ZnIn₂S₄/TiO₂ microcapsules. A certain amount of the as-prepared NH₂-MIL-125(Ti)²⁸ was dispersed into the deionized water (50 mL) under ultrasonication for 30 min to obtain a light-yellow solution. Then, ZnCl₂ (32.2 mg), InCl₃·4H₂O (138.5 mg), and TAA (71 mg) were added to the above mixture and stirred for 1 h at room temperature. Then, the mixture was transferred to an oil bath and kept at 110 °C for 1 h. Finally, the solid product was collected by centrifugation and washed with water and ethanol several times. Finally, the solid product was dried at 60 °C overnight. The resulting ZnIn₂S₄/TiO₂ products with different amounts of ZnIn₂S₄ were denoted as ZIS(x)/TiO₂, where x represents the mass percentage of ZnIn₂S₄ in the composite (x = 50, 60, and 80 wt%).

2.1.2 Synthesis of CDs/ZnIn₂S₄/TiO₂ microreactor. The as-prepared CDs were added to a mixture solution including NH₂-MIL-125(Ti), ZnCl₂, InCl₃·4H₂O, TAA, and deionized water, following the same procedure as that of ZnIn₂S₄/TiO₂ microcapsules. The suspension was transferred to an oil bath and kept at 110 °C for 1 h. Finally, the solid product was collected by centrifugation and washed with water and ethanol several times. The optimal weight ratio of ZnIn₂S₄/TiO₂ was selected to assemble the microreactor, where ZIS(60)/TiO₂ was used in the microreactor. The series of CDs/ZnIn₂S₄(60)/TiO₂ (y-CDs/ZIS/TiO₂) microreactors prepared were denoted as 3-CDs/ZIS/TiO₂, 5-CDs/ZIS/TiO₂, 10-CDs/ZIS/TiO₂, respectively, where y represented the mass percentage of CDs in the microreactors (y = 3, 5 and 10 wt%).

3. Results and discussion

3.1 Material characterization

The synthetic procedure of the CDs/ZIS/TiO₂ microreactor is schematically presented in Fig. 1a, in which this microreactor was prepared by a one-step in situ self-assembly method. The morphological characteristics of the ZIS/TiO₂ microcapsule and CDs/ZIS/TiO₂ microreactor were observed from SEM and TEM images. SEM images demonstrated that the pristine ZnIn₂S₄ presents flower-like microspheres including numerous nanosheets (Fig. S3a†), while NH₂-MIL-125 has a uniform pill-like morphology with an average length of 600 nm (Fig. S3b†). When ZIS/TiO₂ microcapsule was formed, its morphology was quite different from that of NH₂-MIL-125 and ZnIn₂S₄ (Fig. S3c†). The microcapsule seems to be the expansion of NH₂-MIL-125 caused by the effect of TAA and presents a round spindle shape. Besides, a few nanosheets on the surface belonged to the ZnIn₂S₄ layer. Few defects on the surface of the microcapsule can be seen from the SEM and TEM images of ZIS/TiO₂, shown in Fig. S3f,† the hollow cavity can be clearly observed, in which the inner space can be found through the sharp contrast between the rough shell and central void space. Furthermore, due to the growth of the ZnIn₂S₄ nanosheet on the TiO₂ microcapsule, the “double-shell”-like structure can be observed, and the average length of ZIS/TiO₂ seems slightly greater than NH₂-MIL-125, demonstrating that TAA affects not only the phase but also the shape of NH₂-MIL-125 during the self-assembly process, thus the coexistence of ZnIn₂S₄ and TiO₂ in the microcapsule shell. In order to further confirm the formation of TiO₂ microcapsule, we investigated the morphology of m-TiO₂, where only TAA was added to affect the morphology of NH₂-MIL-125, without the Zn/In ions. From the SEM image, as displayed in Fig. S3g,† the as-prepared m-TiO₂ still exhibits the microcapsule structure, and the HRTEM image of m-TiO₂ showed the interplanar distance of 0.35 nm (Fig. S3h†), due to the (101) crystal plane of anatase TiO₂, in agreement with the XRD results. The EDS pattern and elemental mapping (Fig. S3i and S4†) of m-TiO₂ showed that the elements of Ti and O existed without other impurities.

Fig. 1 (a) Schematic illustration for the preparation of CDs/ZIS/TiO₂ microreactor. (b and c) TEM images, (d) HRTEM image, and (e) EDS mapping of 5-CDs/ZIS/TiO₂.

From Fig. 1b, it can be seen that the CDs/ZIS/TiO₂ microreactor maintained its microcapsule structure and the morphology of CDs/ZIS/TiO₂ is similar to that of ZIS/TiO₂, indicating that the microreactor was intact, indicating that CDs coupled with the ZIS/TiO₂ microcapsule fabricated a multi-phase microreactor instead of destroying the original shape. From EDS patterns (Fig. S5†), Zn, In, S, Ti, O, and C elements can be observed in this microreactor. As shown in Fig. S2,† the pristine CDs exhibit well-dispersion with an average diameter of 3 nm, and after the addition of CDs in the in situ synthesis, the morphology of CDs/ZIS/TiO₂ was evaluated from the HRTEM image (Fig. 1d). The HRTEM image of CDs/ZIS/TiO₂ exhibits lattice fringes of 0.21, 0.32 and 0.35 nm that are ascribed to the (100), (102) and (101) crystal facets of CDs, ZnIn₂S₄ and TiO₂, respectively, revealing the successful formation of the CDs/ZIS/TiO₂ microreactor.^29–31 From the EDX mapping analysis of CDs/ZIS/TiO₂ (Fig. 1e), the uniform distribution of C, Zn, In, S, Ti, and O elements throughout the microreactor without agglomeration, further confirmed the formation of hierarchical structure.

The crystal structure information of the ZIS/TiO₂ microcapsule and CDs/ZIS/TiO₂ microreactor was investigated by XRD. As displayed in Fig. S6,† when only TAA reacted with NH₂-MIL-125 without Zn and In source, the XRD patterns of NH-MIL-125 indicated the obvious phase change with the increasing TAA content. When TAA concentration was less than 25%, both the characteristic peaks of TiO₂ and NH₂-MIL-125 were observed, indicating the coexistence of TiO₂ and NH₂-MIL-125. When the TAA concentration was more than 35%, only the peaks of TiO₂ at 25.2°, 37.9°, 48.0°, and 62.6° belonging to the (101), (004), (200) and (204) lattice planes of anatase phase (JCPDS 21-1272),³² and the peaks at 54.3° and 69.1° ascribed to (221) and (301) lattice planes of the rutile phase (JCPDS 21-1276),³³ respectively, were observed, indicating that NH-MIL-125 was completely decomposed to form TiO₂ in this case. When the Zn and In sources were added, ZnIn₂S₄ was formed on the TiO₂ microcapsule. It can be seen from Fig. 2a and S7† that the main peaks of the composite belonged to the characteristic peaks of ZnIn₂S₄ at the (006), (102), (110), (116), and (022) lattice planes, and (101) lattice planes of anatase TiO₂ were observed clearly, indicating the successful synthesis of ZIS/TiO₂ microcapsules. The broad peaks of pristine CDs emerge at 23°, corresponding to the (002) crystal plane of graphite, which illustrates the amorphous phase (Fig. S1a†).³⁴ As expected, CDs/ZIS/TiO₂ exhibited similar diffraction peaks to ZIS/TiO₂, and no characteristic peaks indexed to CDs were observed, probably caused by their low content, uniform distribution as well as small particle size, demonstrating that the microreactors were intact after the introduction of CDs through one-step self-assembly method.

Fig. 2 (a) XRD patterns of ZnIn₂S₄, m-TiO₂, ZIS(60)/TiO₂, and 5-CDs/ZIS/TiO₂. (b) Raman spectra of P25, m-TiO₂, ZIS(60)/TiO₂, and 5-CDs/ZIS/TiO₂. (c) FT-IR spectra of NH₂-MIL-125, ZnIn₂S₄, ZIS(60)/TiO₂ and 5-CDs/ZIS/TiO₂. (d) UV-vis spectra of ZnIn₂S₄, m-TiO₂, ZIS(60)/TiO₂, and 5-CDs/ZIS/TiO₂. High-resolution XPS spectra of (e) Ti 2p and (f) O 1s over NH₂-MIL-125, ZIS(60)/TiO₂, and 5-CDs/ZIS/TiO₂.

To further investigate the structure of the CDs/ZIS/TiO₂ microreactor, Raman spectra were collected and are displayed in Fig. 2b. Since ZnIn₂S₄ presents a weak Raman signal, we tested the Raman spectrum of m-TiO₂, in which the peaks of m-TiO₂ at 148, 394, 510, and 631 cm⁻¹ are ascribed to the E_g, B_1g, A_1g, and E_g modes, respectively, confirming the successful synthesis of anatase TiO₂.³⁵ Compared with commercial P25, the obvious shift of m-TiO₂ at 148 cm⁻¹ is due to the formation of the microcapsule and the size effect of TiO₂. For ZIS/TiO₂ and CDs/ZIS/TiO₂, the further shift at 151 and 155 cm⁻¹ show that the introduction of ZnIn₂S₄ and CDs can enhance the interaction of the composite catalyst.³⁶ Besides, two signals at 1335 and 1594 cm⁻¹ resulting from the D and G bands of CDs, respectively, were observed, indicating that CDs are embedded in the ZIS/TiO₂.³⁷ The FTIR spectra of the ZIS/TiO₂ microcapsule and the CDs/ZIS/TiO₂ microreactor are displayed in Fig. 2c and S8,† respectively, and the functional groups of the as-prepared samples around 3480 cm⁻¹ belonged to –OH bending vibration.³⁸ The pristine NH₂-MIL-125 exhibited the characteristic peaks around 500–800 cm⁻¹ corresponding to the bending vibrations of the Ti–O–Ti group, the band at 1257 cm⁻¹ is ascribed to the C–N group, and the bands in the range of 1350–1600 cm⁻¹ could represent the –COOH group.³⁹ Although bare ZnIn₂S₄ has no obvious characteristic peaks, after the formation of ZIS/TiO₂ and CDs/ZIS/TiO₂, the peaks in the range of 500–700 cm⁻¹ and the band at 1625 cm⁻¹ can be clearly observed, which represent the Ti–O group and –OH group, respectively, indicating that NH₂-MIL-125 was indeed converted to TiO₂ in the microreactor. Due to the small number of CDs, the characteristic peaks belonging to CDs were not observed in the microreactor. The above results verified the coexistence of the anatase TiO₂ phase and CDs in the microreactor.

The UV-vis DRS spectra of CDs/ZIS/TiO₂ microreactors were collected to investigate their optical absorption ability and are shown in Fig. 2d. It was found that NH₂-MIL-125 exhibited good visible-light absorption, attributable to the absorption of NH₂-ligand, while the pristine ZnIn₂S₄ presents an absorption edge around 520 nm. For m-TiO₂, the absorption edge was increased compared with P25, probably due to the change in morphology, as shown in Fig. S9.† When the ZIS/TiO₂ microcapsule was formed, the absorption edge was slightly decreased in comparison with bare ZnIn₂S₄, probably due to the weak light absorption of TiO₂. After the fabrication of CDs/ZIS/TiO₂, the extended absorption edges in the visible-light range can be clearly observed, indicative of the strong absorption of CDs. Since the optical bandgap (E_g) of the photocatalyst is crucial to the determination of the photocatalytic mechanism, E_g values can be obtained from a Tauc plot on the basis of UV-vis DRS spectra (Fig. S10†). E_g values of NH₂-MIL-125, ZnIn₂S₄, and m-TiO₂ were calculated to be 2.48, 2.37, and 2.94 eV, respectively.

The specific surface area of the CDs/ZIS/TiO₂ microreactor was obtained from N₂ adsorption–desorption isotherms (Fig. S11†). The pristine NH₂-MIL-125 exhibited a high BET surface area (1006 m² g⁻¹) and its N₂ sorption isotherm belonged to type I, indicating the characteristics of the microporous material.⁴⁰ In comparison, pure ZnIn₂S₄ exhibited a type-IV isotherm with an obvious hysteresis loop, and a BET surface area of 138 m² g⁻¹. After the formation of the microcapsule, some microporous structure emerges due to the TiO₂ derived from NH₂-MIL-125, and therefore, the BET surface area of ZIS(60)/TiO₂ was up to 272 m² g⁻¹, much higher than that of ZnIn₂S₄. When CDs were added and the microreactor was constructed, the CDs/ZIS/TiO₂ presented the BET surface area of 206 m² g⁻¹, slightly lower than that of ZIS/TiO₂, demonstrating that CDs exist in the microreactor and occupy part of the channel. From the corresponding BJH diagrams (inset), the microcapsule and microreactor exhibit similar pore size distribution, and possessed the coexistence of micropore and mesopore with pore diameters of <20 nm.⁴¹ The rich porosity is conducive to offering more active sites and reducing the mass transfer resistance, which contributes to robust CO₂ photoreduction.

The elemental composition and electron structure of the CDs/ZIS/TiO₂ microreactor were analyzed by XPS spectra. From the XPS survey spectra (Fig. S12a†), the ZIS/TiO₂ microcapsule and CDs/ZIS/TiO₂ microreactor exhibited the expected presence of Ti, O, C, Zn, In, and S, and the peak intensity of the C element in CDs/ZIS/TiO₂ was stronger than that in ZIS/TiO₂, indicating that CDs were successfully embedded in ZIS/TiO₂. Besides, the N element was observed in NH₂-MIL-125 instead of ZIS/TiO₂, suggesting that the N element was lost during the self-assembly process. As displayed in Fig. 2e, the Ti 2p peaks with binding energies of 458.6 and 464.4 eV were assigned to Ti 2p_3/2 and Ti 2p_1/2, respectively, indicative of the presence of Ti⁴⁺ in the pristine NH₂-MIL-125, corresponding to the previous work.⁴² For ZIS/TiO₂ and CDs/ZIS/TiO₂, the peak slightly shifted to the lower binding energy of Ti 2p_3/2, probably due to the phase change from the Ti–O cluster to TiO₂.⁴³ Notably, O 1s peak is of significance for investigating the surface unsaturated surrounding in Ti and further confirmed the formation of metal oxides (Fig. 2f). In the deconvoluted O 1s spectrum of NH₂-MIL-125, the three signals at 529.8, 531.9, and 533.4 eV correspond to the Ti–O bond, –OH bond and adsorbed oxygen (O_abs), respectively.⁴⁴ However, for ZIS/TiO₂ and CDs/ZIS/TiO₂, the area of the Ti–O band was significantly enhanced, demonstrating that a large amount of metal oxide was constructed.⁴⁵ In addition, the C 1s spectra of NH₂-MIL-125 presented four deconvoluted signals at 284.8, 285.3, 286.5, and 288.7 eV, which represent the C=C, C–N, C–NH₂, and C=O bonds, respectively, indicative of the presence of the C–N group (Fig. S12b†).⁴⁶ For ZIS/TiO₂ and CDs/ZIS/TiO₂, the peaks of the amino group disappeared, and more C–C and C=O groups emerged, indicating the presence of surface functional groups such as hydroxyl and carboxyl around catalysts.⁴⁷ From the high-resolution spectrum of N 1s (Fig. S12c†), the difference between NH₂-MIL-125 and ZIS/TiO₂ can be seen clearly. The N 1s spectrum of NH₂-MIL-125 was deconvoluted into two signals at 399.2 and 402.6 eV, which belong to the amino group (C–N) and imine group (–NH–), respectively,⁴⁸ while there were no observable N 1s signals in ZIS/TiO₂ and CDs/ZIS/TiO₂, suggesting that MOF-topological structure was completely converted into TiO₂, successfully. Besides, as for Zn (Fig. S12d†), the binding energies of 1022.3 and 1045.4 eV correspond to Zn 2p_3/2 and Zn 2p_1/2, respectively, which is typical of Zn²⁺ in ZIS/TiO₂ and CDs/ZIS/TiO₂.⁴⁹ For In 3d XPS spectra (Fig. S12e†), the peaks of ZIS/TiO₂ centered at 445.3 and 452.8 eV are due to the In 3d_5/2 and In 3d_3/2, respectively.⁴⁹ Compared with ZIS/TiO₂, the binding energies of CDs/ZIS/TiO₂ for In 3d region exhibited a slight positive shift, which resulted from the changed electron density caused by the effect of CDs. In Fig. S12f,† the binding energies of S 2p in ZIS/TiO₂ at 162.0 and 163.3 eV are indicative of the presence of S²⁻, and also CDs/ZIS/TiO₂ exhibited a larger positive shift compared with In 3d orbit, suggesting the effective electron transfer.⁵⁰

3.2 Photocatalytic CO₂ reduction

With the CDs/ZIS/TiO₂ microreactor as a photocatalyst, the photocatalytic performances for CO₂ reduction were conducted in H₂O/EA solution without any photosensitizer or sacrificial agent, in which EA could enhance the solubility of CO₂ gas. During these experiments, CO and CH₄ were detected as main products in the as-prepared catalysts under simulated solar light irradiation, as displayed in Fig. 3. Only a small amount of CO product (0.7 μmol g⁻¹ h⁻¹) over pristine NH₂-MIL-125 was detected, while the bare ZnIn₂S₄ presented low CO (3.1 μmol g⁻¹ h⁻¹) and CH₄ (0.3 μmol g⁻¹ h⁻¹) yields. After the formation of the ZIS/TiO₂ microcapsule, CO₂ photoreduction was remarkably enhanced compared with the precursors. The different weight ratios of ZnIn₂S₄ and TiO₂ were considered (Fig. 3b), and the optimal ZIS(60)/TiO₂ exhibited CO and CH₄ yields of up to 12.9 and 7.4 μmol g⁻¹ h⁻¹, respectively. Notably, the introduction of CDs on ZIS(60)/TiO₂ further boosted the photocatalytic performances, 5-CDs/ZIS/TiO₂ showed the maximum CO (19.2 μmol g⁻¹ h⁻¹) and CH₄ (14.9 μmol g⁻¹ h⁻¹) yields, in which, the concentration of products revealed a linearly enhancing trend with duration time. Such outstanding activity originated from the CDs/ZIS/TiO₂ as a productive heterogeneous microreactor. The yield of CH₄ generation over the as-prepared CDs/ZIS/TiO₂ is regarded as one of the most competitive performances for ZnIn₂S₄-based catalysts after an extensive literature search in the field of similar photocatalysts as illustrated in Table S5.†

Fig. 3 (a) Time–yield plots of 5-CDs/ZIS/TiO₂ under AM 1.5 G simulated solar irradiation. (b) Photocatalytic CO₂ reduction performances of different catalysts. (c) Photocatalytic cycling stability of 5-CDs/ZIS/TiO₂. (d) Summary of the photocatalytic performances of ZnIn₂S₄, ZIS(60)/TiO₂, and 5-CDs/ZIS/TiO₂. (e) GC-MS analysis of CO generated from the ¹³CO₂ isotope experiment. (f) Production rates of CO and CH₄ over 5-CDs/ZIS/TiO₂ under various reaction conditions.

Based on electron consumption rates (R_electron, CH₄: 8e⁻; CO: 2e⁻), CH₄ selectivity was evaluated by the following equation: CH₄ selectivity (%) = [8R(CH₄)]/[2R(CO) + 8R(CH₄)] × 100%, where R(CO) and R(CH₄) are the conversion rates of CO and CH₄, respectively. As displayed in Fig. 3d and S14,† the 5-CDs/ZIS/TiO₂ microreactor exhibited the highest CH₄ selectivity (75.6%), and R_electron was up to 157.6 μmol g⁻¹ h⁻¹, which were 1.9 and 18.3 times higher those of ZIS(60)/TiO₂ and bare ZnIn₂S₄, respectively, indicating that e more protons formed in the microreactor can facilitate the CH₄ production from CO₂ reduction. Importantly, O₂ as the main oxidation product was observed, as displayed in Fig. S16,† verifying the overall photocatalytic CO₂ reduction. The O₂ yield of 5-CDs/ZIS/TiO₂ (11.3 μmol h⁻¹ g⁻¹) was higher than that of ZIS(60)/TiO₂ (7.1 μmol h⁻¹ g⁻¹), indicating that CDs have a positive effect on H₂O oxidation reaction. Additionally, the stoichiometric ratio of the consumed electrons and holes for ZIS(60)/TiO₂ and 5-CDs/ZIS/TiO₂ was 1.1 and 1.2, respectively, which illustrated that the two values are close to 1. To further evaluate the oxidation capacity of the CDs/ZIS/TiO₂ microreactor, oxygen evolution reaction (OER) was revealed by the linear sweep voltammetry (LSV) performances (Fig. S17†). The increased current density of ZIS(60)/TiO₂ was observed compared with the pristine ZnIn₂S₄, indicating that the microcapsule presented an excellent OER activity. Obviously, after the formation of the CDs-modified microreactor, all the CDs/ZIS/TiO₂ samples possessed further enhanced current density, meaning that the microreactor with higher OER performance drove the complete CO₂ reduction reaction.

The long-term durability of CDs/ZIS/TiO₂ was investigated by cycle tests as displayed in Fig. 3c. After six successive cycles over 30 h, a negligible decrease in the photocatalytic performance of 5-CDs/ZIS/TiO₂ was detected, with only ca. 3.6% and 4.0% deactivation of CO and CH₄ yields, respectively. The ¹³C labeled isotope experiments confirmed that the generated CO and CH₄ indeed originated from catalysts instead of other carbon sources, as illustrated in Fig. 3e. ¹³CO (m/z = 29) and ¹³CH₄ (m/z = 17) were observed, revealing that the products are from the CO₂ photoreduction. Furthermore, the controlled tests were performed to determine the origin of the products (Fig. 3f). Negligible amounts of carbon-containing products were observed when the reaction system was tested without the photocatalyst, absence of light irradiation, and in Ar atmosphere instead of high-purity CO₂, demonstrating that the photo-excited process was indispensable for this reaction, and the other carbon impurities existed in the photocatalyst and the reaction system could not afford any CO₂ reduction products.

3.3 Mechanism for the CO₂ photoreduction

Steady-state and time-resolved PL spectra were used as powerful tools to investigate the electron transfer of the CDs/ZIS/TiO₂ microreactor. As displayed in Fig. 4a, the PL intensity of ZIS/TiO₂ decreased in comparison with the pure ZnIn₂S₄, which indicated that the formed hierarchical hollow morphology with a shorter charge transfer distance could reduce the charge recombination. After the introduction of CDs, the interfacial interaction between ZnIn₂S₄ and TiO₂ became stronger, which can remarkably improve the charge transfer efficiency, leading to CDs/ZIS/TiO₂ with the lowest PL intensity.⁵¹ The average PL lifetimes (τ_ave) of the representative catalysts were further investigated using TR-PL spectra with values of 1.62, 1.22, 1.06 ns for ZnIn₂S₄, ZIS(60)/TiO₂ and 5-CDs/ZIS/TiO₂, respectively, as shown in Fig. 4b. The PL lifetime of ZIS(60)/TiO₂ presents a shorter τ_ave than ZnIn₂S₄, attributable to fast electron transfer on the surface of the microcapsule, resulting from intimate contact. The faster PL decay and shorter τ_ave in 5-CDs/ZIS/TiO₂ indicate that CDs indeed act as charge mediators in a microreactor.⁵² The observations of decay in both steady-state and time-resolved PL results demonstrate that microcapsule and multi-phase microreactors can effectively suppress the charge carrier recombination.

Fig. 4 (a) PL spectra and (b) TRPL spectra of ZnIn₂S₄, ZIS(60)/TiO₂, and 5-CDs/ZIS/TiO₂. (c) Photocurrent responses and (d) EIS Nyquist plots of NH₂-MIL-125, ZnIn₂S₄, ZIS(60)/TiO₂, and 5-CDs/ZIS/TiO₂. Mott–Schottky plots of (e) ZnIn₂S₄ and (f) m-TiO₂.

The electrochemical performances were further studied to analyze the effect of the multi-phase microreactor on the photoinduced charge separation efficiency. The photocurrent of ZIS(60)/TiO₂ was much higher than that of ZnIn₂S₄ and unconverted NH₂-MIL-125 under simulated sunlight irradiation (Fig. 4c). Also, 5-CDs/ZIS/TiO₂ indicates that the further enhanced current density, which implies that CDs in the microreactor can accelerate the electron transfer rate.³⁴

Corresponding to the photocurrent results, ZIS(60)/TiO₂ and 5-CDs/ZIS/TiO₂ showed smaller radii in the EIS Nyquist diagram (Fig. 4d), which represented the decreasing charge transfer resistance, facilitating the charge transport and boosting reaction kinetics in photocatalytic performance.⁵³ Furthermore, the flat-band potentials (E_FB) of ZnIn₂S₄ and TiO₂ were obtained from Mott–Schottky (M–S) plots at frequencies of 1000, 1500, and 2000 Hz (Fig. 4e and f).⁵⁴ The plots of ZnIn₂S₄ and TiO₂ display positive slopes, indicative of typical n-type semiconductors. The E_FB values of ZnIn₂S₄ (−1.22 V vs. Ag/AgCl) and m-TiO₂ (−0.48 V vs. Ag/AgCl) were determined by extrapolating the M–S plots, and correspondingly, E_FB values are −1.02 and −0.28 V (vs. NHE) for ZnIn₂S₄ and m-TiO₂ according to the E_NHE = E_Ag/AgCl + 0.197,⁵⁵ respectively, where E_FB values are close to the conduction band (E_CB) potentials of n-type semiconductor.⁵⁶ Consequently, E_CB values of ZnIn₂S₄ and m-TiO₂ are −1.02 and −0.28 V (vs. NHE), respectively, and together with band gaps (E_g) from the Tauc plots, the corresponding valence band values (E_VB) of ZnIn₂S₄ and m-TiO₂ are 1.35 V and 2.66 V (vs. NHE), respectively.

Transient photovoltage (TPV) tests were performed to analyze the charge transfer kinetics on the interfaces of the photocatalysts. The TPV relaxation curves of m-TiO₂, ZnIn₂S₄, ZIS/TiO₂, and CDs/ZIS/TiO₂ are shown in Fig. 5a. Furthermore, as shown in Fig. 5b, the electron recombination rates existing in the photocatalysts were investigated by the use of the attenuation constants (τ). The τs of m-TiO₂ and ZnIn₂S₄ were 0.682 and 0.609 ms, respectively, while the τ of ZIS/TiO₂ was 0.544 ms. The reason for the smaller τ of ZIS/TiO₂ is that the formation of the heterojunction between m-TiO₂ and ZnIn₂S₄ causes a part of the charges to recombine before being collected by the working electrode. In addition, the formation of the heterojunction is beneficial for charge transfer. The τ of CDs/ZIS/TiO₂ is 0.421 ms, revealing that the addition of CDs further facilitates electron transport. As shown in Fig. 5c, t_max was used to estimate the rate of the charge extraction process. There is a great difference between the t_max of m-TiO₂ (t_max1 = 0.256 ms) and ZnIn₂S₄ (t_max2 = 0.101 ms). However, compared with that of m-TiO₂, t_max3 (0.114 ms) becomes much smaller, indicating that the integration of m-TiO₂ and ZnIn₂S₄ promotes the charge extraction process. After the addition of CDs, t_max4 (0.133 ms) is still much smaller than that of m-TiO₂, proving that the addition of CDs also accelerates the charge extraction process. Fig. 5d shows the area of the shadow part (A) of m-TiO₂, ZnIn₂S₄, ZIS/TiO₂, and CDs/ZIS/TiO₂, which correspond to the maximum charge extraction of the catalysts. It is worth noting that the A of m-TiO₂ (A₁ = 0.192) and ZnIn₂S₄ (A₂ = 0.0511) are larger than that of ZIS/TiO₂ (A₃ = 0.0341), which is attributed to the recombination of the charge on heterojunction interfaces after being excited by the laser, resulting in the smaller amounts of charges collected by the working electrode. Furthermore, the A of CDs/ZIS/TiO₂ (A₄ = 0.0934) is larger than that of ZIS/TiO₂, demonstrating that CDs can enhance the electron extraction ability of the photocatalyst. The surface effective charge (n_e) is used to further determine the three eigenvalues of TPV (τ, t_max, A), which can be calculated from the equation of n_e = (A × τ)/t_max. For photocatalysts, the value of n_e represents the amount of the charge that is involved in the photocatalytic redox reaction.⁵⁷ The n_e of m-TiO₂, ZnIn₂S₄, ZIS/TiO₂, and CDs/ZIS/TiO₂ are 0.510, 0.307, 0.162, and 0.295, respectively. Similarly, the n_e of ZIS/TiO₂ becomes smaller than those of m-TiO₂ and ZnIn₂S₄, which was also caused by the heterojunction formed between the two components. The n_e of CDs/ZIS/TiO₂ increases by ca. 1.82 times compared with that of ZIS/TiO₂, suggesting that the introduction of CDs is beneficial for the photocatalytic reaction. In summary, the heterojunction formed between m-TiO₂ and ZnIn₂S₄ facilitated the charge transfer process. In addition, CDs not only play the role of regulating the charge transfer process but also improve the ability of the photocatalyst to extract electrons for the photocatalytic reaction.

Fig. 5 Comparison of the TPV curves of m-TiO₂, ZnIn₂S₄, ZIS(60)/TiO₂, and 5-CDs/ZIS/TiO₂. (a) TPV relaxation curves of m-TiO₂, ZnIn₂S₄, ZIS(60)/TiO₂ and 5-CDs/ZIS/TiO₂. (b) The attenuation constants (τ) of the charge recombination process. (c) Charge extraction rate (t_max) of m-TiO₂, ZnIn₂S₄, ZIS(60)/TiO₂ and 5-CDs/ZIS/TiO₂. (d) The maximum electron extraction of m-TiO₂, ZnIn₂S₄, ZIS(60)/TiO₂ and 5-CDs/ZIS/TiO₂.

In situ TPV experiments were performed to understand the photocatalytic reaction over the catalysts. Fig. 6 displays the in situ TPV results of m-TiO₂ and ZnIn₂S₄ under an atmosphere of N₂-saturated MeCN, CO₂-saturated MeCN, and 0.5 vol% H₂O/MeCN (v/v), respectively. Compared with m-TiO₂, the TPV intensity of ZnIn₂S₄ exhibits a sharper decrease when the atmosphere changes from N₂-saturated MeCN to CO₂-saturated MeCN, indicating that ZnIn₂S₄ provides active sites for the CO₂ reduction reaction, which consumes electrons. Similarly, the H₂O oxidation reaction consumes holes, which will lead to an increase in TPV intensity. However, the increase of TPV intensity of m-TiO₂ and ZnIn₂S₄ are close when the atmosphere changes from N₂-saturated MeCN to 0.5 vol% H₂O/MeCN (v/v). Therefore, in order to further analyze the active sites of H₂O oxidation, the attenuation constants (τ) of the in situ TPV curves were calculated, as shown in Fig. 6c and d. It is worth noting that, for ZnIn₂S₄, there exist two attenuation processes both in CO₂-saturated and 0.5 vol% H₂O/MeCN (v/v). Therefore, after calculating the τ of each attenuation process (τ_5-1 = 0.055 ms, τ_5-2 = 0.412 ms, τ_6-1 = 0.410 ms, τ_6-1 = 0.561 ms), the average τs of ZnIn₂S₄ under CO₂-saturated and 0.5 vol% H₂O/MeCN (v/v) (τ_5avg = 0.190 ms, τ_6avg = 0.490 ms) were calculated using the formula (x) provided in the ESI.† The changing percentage of τ (Δτ) was calculated using the following formula (Δτ = (τ_N2 − τ_H2O/CO2)/τ_N2 × 100%) to study the influence of CO₂ or H₂O on charge recombination process. For m-TiO₂, Δτ(H₂O) (22.2%) is much higher than Δτ(CO₂) (1.56%), which proves that the introduction of H₂O makes great effect on its charge recombination process, suggesting that m-TiO₂ provides active sites for the H₂O oxidation reaction.^31,34 Similarly, for ZnIn₂S₄, Δτ(CO₂) (68.8%) is much higher than Δτ(H₂O) (19.5%), indicating that ZnIn₂S₄ provided active sites for the CO₂ reduction reaction.^34,58

Fig. 6 (a, b) Comparison of the in situ TPV curves of m-TiO₂ and ZnIn₂S₄ under N₂-saturated MeCN, CO₂-saturated MeCN, and 0.5 vol% H₂O/MeCN (v/v). (c, d) The corresponding attenuation constants (τ) of the charge recombination process.

Electron spin resonance (ESR) spectra of CDs/ZIS/TiO₂ were used to further ascertain the photocatalytic mechanism as displayed in Fig. 7a. The obvious DMPO-˙O₂⁻ signals exhibit that the generated electrons on CDs/ZIS/TiO₂ can effectively produce ˙O₂⁻ species, which means that the position of the electrons on the CB is more negative than the potential of superoxide radical (O₂/˙O₂⁻, −0.33 eV).⁵⁹ Thus, the above results indicate that the electrons are accumulated on the CB of ZnIn₂S₄ instead of m-TiO₂, which is consistent with the TPV results, both confirming the formation of the Z-scheme mechanism.⁶⁰ Besides, we conducted in situ FTIR spectroscopy to illustrate the reaction pathway in the CO₂ photoreduction over the CDs/ZIS/TiO₂ microreactor, as displayed in Fig. 7b. It is seen that some peaks of multiple intermediate products emerge, which gradually become stronger with the extension of the irradiation time. After CO₂ and H₂O gas were adsorbed on CDs/ZIS/TiO₂ in the dark for 30 min, bicarbonate species (HCO₃⁻, 1223, 1398, 1436, 1455 and 1475 cm⁻¹),⁶¹ monodentate carbonate species (m-CO₃²⁻, 1418, 1488, 1541 and 1557 cm⁻¹) and bidentate carbonate species (b-CO₃²⁻, 1387, 1524 and 1631 cm⁻¹) were observed in the reaction process,^62,63 revealing that the absorbed CO₂ and dissociative H₂O molecules exist on the surface of CDs/ZIS/TiO₂. Besides, new peaks emerge and the active CO₂⁻ peaks at 1677 cm⁻¹ can be observed. The peaks of formaldehyde (HCHO⁻, 1507 and 1788 cm⁻¹), methoxy groups (CH₃O⁻, 1688 and 1734 cm⁻¹), formic acid species (HCOO⁻, 1641 and 1658 cm⁻¹), and carboxylate species (COO⁻, 1350 cm⁻¹) are detected, indicating that they are primary intermediates during CO₂ photoreduction.^64–66 Besides, there are no CH₄ peaks, probably due to its nonpolar as well as low affinity.

Fig. 7 (a) ESR spectra of DMPO ˙O₂⁻ on 5-CDs/ZIS/TiO₂. (b) In situ DRIFTS spectra of surface adsorbed CO₂ species and photocatalytic CO₂ reduction intermediates on 5-CDs/ZIS/TiO₂. (c) Photocatalytic reaction pathways over CDs/ZIS/TiO₂.

According to the in situ FTIR results, HCHO⁻, CH₃O⁻, HCOO⁻ and COO⁻ groups are the significant intermediates, and coupled with the TPV and ESR results, the formed Z-scheme over CDs/ZIS/TiO₂ was deduced, as shown in Fig. S18.† Under sunlight irradiation, the photoinduced electrons in the CB of m-TiO₂ recombine with the holes in the VB of ZnIn₂S₄, and meanwhile, the accumulation of electrons on the CB of ZnIn₂S₄ and holes on the VB of m-TiO₂ possesses strong redox ability for CO₂ reduction and O₂ oxidation.⁶⁷ Besides, the CDs in this heterostructure act as an electron conductor and reservoir, in which electrons transported to the surface are captured by CDs, further reducing the charge recombination to promote the redox reaction. As for the formed microreactor, the numerous ultrathin nanosheets as the outer layer are beneficial to CD implantation and CO₂ adsorption. Furthermore, the multi-shell structure, high surface area, and large inner space can promote light utilization by multiple reflections, and facilitate the fast diffusion of gaseous products.

4. Conclusions

In summary, we have demonstrated a novel CDs/ZnIn₂S₄/TiO₂ microreactor prepared using a MOF-mediated strategy. During the synthetic process, the thioacetamide (TAA) can facilitate the dissolution of the inner core for NH₂-MIL-125 to form TiO₂ microcapsules, and with the precipitation of Zn²⁺/In³⁺ ions and CDs, a novel productive heterogeneous microreactor with multi-shell structure was obtained. The obtained CDs/ZIS/TiO₂ exhibited high photocatalytic performance and selectivity for CO₂ photoreduction into CH₄ without any sacrificial agent. The yield of CH₄ over optimal CDs/ZIS/TiO₂ was up to 14.9 μmol g⁻¹ h⁻¹ with CH₄ selectivity of 75.6%, and the R_electron reaches 157.6 μmol g⁻¹ h⁻¹. The in situ TPV measurements indicated that m-TiO₂ provided active sites for the H₂O oxidation reaction, and ZnIn₂S₄ provided active sites for the CO₂ reduction reaction. Coupled with ESR, photoelectrochemical and in situ Fourier transform infrared spectra, a Z-scheme mechanism for CDs/ZIS/TiO₂ is proposed. Furthermore, the CDs in this microreactor as an electron acceptor play a significant role in the improvement of charge separation efficiency. This work not only provides an effective strategy to construct MOF-derived productive heterogeneous microreactor for photocatalytic CO₂ conversion but is also an inspiration for developing the integrated catalytic system in the field of artificial photosynthesis.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work is supported by the National Key Basic Research Program of China (2020YFA0406104/2020YFA0406101), the National MCF Energy R&D Program (2018YFE0306105), the Innovative Research Group Project of the National Natural Science Foundation of China (51821002), National Natural Science Foundation of China (51725204, 51972216, 21876015, 52272043, 52271223, 52202107, 52201269), Natural Science Foundation of Jiangsu Province (BK20220028, BK20190041), Natural Science Foundation of Jiangsu Province-Excellent Youth Foundation (BK20190102), Key-Area Research and Development Program of GuangDong Province (2019B010933001), Collaborative Innovation Center of Suzhou Nano Science & Technology, Qinglan Project Foundation of Jiangsu Province, the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), and the 111 Project.

Notes and references

1. S. F. Ji, Y. Qu, T. Wang, Y. J. Chen, G. F. Wang, X. Li, J. C. Dong, Q. Y. Chen, W. Y. Zhang, Z. D. Zhang, S. Y. Liang, R. Yu, Y. Wang, D. S. Wang and Y. D. Li, Angew. Chem., Int. Ed., 2020, 59, 10651–10657.
2. C. L. Wang, Z. X. Sun, Y. Zheng and Y. H. Hu, J. Mater. Chem. A, 2019, 7, 865–887.
3. I. Shown, S. Samireddi, Y. C. Chang, R. Putikam, P. Chang, A. Sabbah, F. Fu, W. Chen, C. Wu, T. Yu, P. Chung, M. C. Lin, L. Chen and K. Chen, Nat. Commun., 2018, 9, 169.
4. J. Li, H. L. Huang, W. J. Xue, K. Sun, X. H. Song, C. R. Wu, L. Nie, Y. Li, C. Y. Liu, Y. Pan, H. L. Jiang, D. H. Mei and C. L. Zhong, Nat. Catal., 2021, 4, 719–729.
5. Y. J. Ma, X. X. Yi, S. L. Wang, T. Li, B. Tan, C. C. Chen, T. Majima, E. R. Waclawik, H. Y. Zhu and J. Y. Wang, Nat. Commun., 2022, 13, 1400.
6. H. Wu, X. Y. Kong, X. M. Wen, S. Chai, E. C. Lovell, J. W. Tang and Y. H. Ng, Angew. Chem., Int. Ed., 2021, 60, 8455–8459.
7. S. L. Wang, M. Xu, T. Y. Peng, C. X. Zhang, T. Li, I. Hussain, J. Y. Wang and B. Tan, Nat. Commun., 2019, 10, 676.
8. Y. Zhang, Y. X. Wu, L. Wan, H. J. Ding, H. X. Li, X. Y. Wang and W. H. Zhang, Appl. Catal., B, 2022, 311, 121255.
9. S. B. Wang, B. Y. Guan and X. W. Lou, J. Am. Chem. Soc., 2018, 140, 5037–5040.
10. K. Zhu, J. O. Yang, Q. Zeng, S. G. Meng, W. Teng, Y. H. Song, S. Tang and Y. J. Cui, Chin. J. Catal., 2020, 41, 454–463.
11. P. She, B. Y. Guan, J. Y. Sheng, Y. Y. Qi, G. Y. Qiao, H. B Rui, G. Y. Lu, J. S. Qin and H. Rao, Catal. Sci. Technol., 2022, 12, 1092–1099.
12. S. T. Wu, Z. Xin, S. C. Zhao and S. T. Sun, Nano Res., 2019, 12, 2736–2742.
13. T. He, X. B. Xu, B. Ni, H. F. Lin, C. Z. Li, W. P. Hu and X. Wang, Angew. Chem., Int. Ed., 2018, 57, 10148–10152.
14. Y. Liu, X. F Zhou, Z. R. Jia, H. J. Wu and G. L. Wu, Adv. Funct. Mater., 2022, 32, 2204499.
15. H. T. Fan, Y. J. Jin, K. C. Liu and W. S. Liu, Adv. Sci., 2022, 9, 2104579.
16. H. J. Xu, C. F. Shan, X. X. Wu, M. Z. Sun, B. Huang, Y. Tang and C. H. Yan, Energy Environ. Sci., 2020, 13, 2949–2956.
17. R. Bibi, H. L. Huang, M. Kalulu, Q. H. Shen, L. F. Wei, O. Oderinde, N. X. Li and J. C. Zhou, ACS Sustainable Chem. Eng., 2019, 7, 4868–4877.
18. Q. Y. Wu, J. J. Cao, X. Wang, Y. Liu, Y. J. Zhao, H. Wang, Y. Liu, H. Huang, F. Liao, M. W. Shao and Z. H. Kang, Nat. Commun., 2021, 12, 483.
19. H. Q. Song, J. K. Yu, Z. Y. Tang, B. Yang and S. Y. Lu, Adv. Energy Mater., 2022, 12, 2102573.
20. J. Wu, Y. D. Han, Y. C. Bai, X. T. Wang, Y. J. Zhou, W. X. Zhu, T. W. He, Y. M. Wang, H. Huang, Y. Liu and Z. H. Kang, Adv. Funct. Mater., 2022, 32, 2203647.
21. Y. J. Dong, Q. Han, Q. Y. Hu, C. J. Xu, C. Z. Dong, Y. Peng, Y. Ding and Y. Q. Lan, Appl. Catal., B, 2021, 293, 120214.
22. B. Li, W. Peng, J. Zhang, J. C. Lian, T. Huang, N. Cheng, Z. Y. Luo, W. Q. Huang, W. Y. Hu, A. L. Pan, L. Jiang and G. F. Huang, Adv. Funct. Mater., 2021, 31, 2100816.
23. Y. J. Zhao, L. L. Xu, X. Wang, Z. Z. Wang, Y. Liu, Y. Wang, Q. L. Wang, Z. T. Wang, H. Huang, Y. Liu, W. Y. Wong and Z. H. Kang, Nano Today, 2022, 43, 101428.
24. Y. Liu, X. Wang, Y. J. Zhao, Q. Y. Wu, H. D. Nie, H. L. Si, H. Huang, Y. Liu, M. W. Shao and Z. H. Kang, Nano Res., 2022, 15, 4000–4007.
25. S. S. Mondal, S. R. Das, L. Sahoo, S. Dutta and U. Gautam, J. Am. Chem. Soc., 2022, 144, 2580–2589.
26. Q. Li, S. C Wang, Z. X. Sun, Q. J. Tang, Y. Q. Liu, L. Z. Wang, H. Q. Wang and Z. B. Wu, Nano Res., 2019, 12, 2749–2759.
27. X. Q. Gu, Z. M. Chen, Y. Li, J. Wu, X. Wang, H. Huang, Y. Liu, B. Dong, M. W. Shao and Z. H. Kang, ACS Appl. Mater. Interfaces, 2021, 13, 24814–24823.
28. Z. Z. Su, B. X. Zhang, J. B. Shi, D. X. Tan, F. Y. Zhang, L. F. Liu, X. N. Tan, D. Shao, G. Y. Yang and J. L. Zhang, Sustain. Energy Fuels, 2019, 3, 1233–1238.
29. X. W. Shi, C. Dai, X. Wang, J. Y. Hu, J. Y. Zhang, L. X. Zheng, L. Mao, H. J. Zheng and M. S. Zhu, Nat. Commun., 2022, 13, 1287.
30. Q. Cheng, Y. J. Yuan, R. Tang, Q. Y. Liu, L. Bao, P. Wang, J. S. Zhong, Z. Y. Zhao, Z. T. Yu and Z. G. Zou, ACS Catal., 2022, 12, 2118–2125.
31. Q. Liang, X. T. Yan, Z. Y. Li, Z. Y. Wu, H. Shi, H. Huang and Z. H. Kang, J. Mater. Chem. A, 2022, 10, 4279–4287.
32. Y. Lu, W. J. Yin, K. L. Peng, K. Wang, Q. Hu, A. Selloni, F. R. Chen, L. M. Liu and M. L. Sui, Nat. Commun., 2018, 9, 2752.
33. C. Peng, T. Zhou, P. Wei, H. Q. Ai, B. P. Zhou, H. Pan, W. K. Xu, J. B. Jia, K. Zhang, H. J. Wang and H. Yu, Chem. Eng. J., 2022, 439, 135685.
34. Q. Liang, S. Zhao, Z. Y. , Li, Z. Y. Wu, H. Shi, H. Huang and Z. H. Kang, ACS Appl. Mater. Interfaces, 2021, 13, 40754–40765.
35. Y. X. Wang, Y. Y. Zhang, X. J. Zhu, Y. Liu and Z. B. Wu, Appl. Catal., B, 2022, 316, 121610.
36. X. X. Zhang, Y. G. Xiao, S. S. Cao, Z. L. Yin and Z. Q. Liu, J. Clean. Prod., 2022, 352, 131560.
37. Y. J. Bai, X. B. Yi, B. Li, S. W. Chen and Z. J. Fan, Appl. Surf. Sci., 2022, 578, 151993.
38. Z. Man, Y. Meng, X. C. Lin, X. R. Dai, L. P. Wang and D. Z. Liu, Chem. Eng. J., 2022, 431, 133952.
39. G. L. Mo, L. X. Wang and J. H. Luo, Sep. Purif. Technol., 2021, 277, 119643.
40. Z. G. Yin, T. T. Song, W. T. Zhou, Z. L. Wang and Y. Ma, J. Catal., 2021, 402, 289–299.
41. K. M. Kamal, R. Narayan, N. Chandran, S. Popović, M. A. Nazrulla, J. Kovač, N. Vrtovec, M. Bele, N. Hodnik, M. M. Kržmanc and B. Likozar, Appl. Catal., B, 2022, 307, 121181.
42. X. M. Cheng, X. Y. Dao, S. Q. Wang, J. Zhao and W. Y. Sun, ACS Catal., 2021, 11, 650–658.
43. Y. Chen, G. B. Mao, Y. W. Tang, H. Wu, G. Wang, L. Zhang and Q. Liu, Chin. J. Catal., 2021, 42, 225–234.
44. X. D. Zhang, K. Yue, R. Z. Rao, J. F. Chen, Q. Liu, Y. Yang, F. K. Bi, Y. X. Wang, J. C. Xu and N. Liu, Appl. Catal., B, 2022, 310, 121300.
45. L. B. Wang, B. Chen, L. Y. Zhang and J. G. Yu, Small, 2021, 17, 2103447.
46. S. H. Wu, X. F Xing, D. Wang, J. Z. Zhang, J. M. Chu, C. C. Yu, Z. T. Wei, M. L. Hu, X. Zhang and Z. X. Li, ACS Sustain. Chem. Eng., 2020, 8, 148–153.
47. X. J. Yang, H. W. Sun, G. Y. Li, T. C. An and W. Choi, Appl. Catal., B, 2021, 294, 120252.
48. X. J. Li, K. Y. Zhang, X. B. Huang, Z. Y. Wu, D. F. Zhao and G. Wang, Nanoscale, 2021, 13, 19671–19681.
49. S. Q. Zhang, X. Liu, C. B Liu, S. L. Luo, L. L. Wang, T. Ca, Y. X. Zeng, J. L. Yuan, W. Y. Dong, Y. Pei and Y. T. Liu, ACS Nano, 2018, 12, 751–758.
50. K. Wang, X. H. Li, N. Wang, Q. H. Shen, M. C. Liu, J. C. Zhou and N. X. Li, Ind. Eng. Chem. Res., 2021, 60, 8720–8732.
51. P. She, B. Y. Guan, J. Y. Sheng, Y. Y. Qi, G. Y. Qiao, H. B. Rui, G. Y. Lu, J.-S. Qin and H. Rao, Catal. Sci. Technol., 2022, 12, 1092–1099.
52. H. H. Li, F. Zhang, H. F. Wang, J. R. Xue, Y. M. Guo, Q. Z. Qian and G. Q. Zhang, Energy Environ. Sci., 2021, 14, 5339–5346.
53. C. H. Liu, Y. H. Yao, L. Sun, L. L. Luo, W. C. Wang and Z. D. Chen, Chem. Commun., 2021, 57, 9846–9849.
54. A. R. Amani-Ghadim, F. Khodam and M. S. Dorraji, J. Mater. Chem. A, 2019, 7, 11408–11422.
55. T. Wang, Q. Y. Men, X. Q. Liu, H. Q. Zhan and Y. Q. Wang, Sep. Purif. Technol., 2022, 294, 121215.
56. X. H. Ma, D. Y. Li, Y. H. Jiang, H. C. Jin, L. Y. Bai, J. Qi, F. F. You and F. L. Yuan, J. Colloid Interface Sci., 2022, 628, 768–776.
57. Y. Li, Y. Zhao, J. Wu, Y. D. Han, H. Huang, Y. Liu and Z. H. Kang, J. Mater. Chem. A, 2021, 9, 25453–25462.
58. Y. X. Li, Y. X. Liu, X. Liu, Y. L. Liu, Y. Y. Cheng, P. Zhang, P. J. Deng, J. J. Deng, Z. H. Kang and H. T. Li, Nano Res., 2022, 15, 6026–6035.
59. J. Z. Zhao, M. X. Ji, H. L. Chen, Y.-X. Weng, J. Zhong, Y. J. Li, S. Y. Wang, Z. R. Chen, J. X. Xia and H. M. Li, Appl. Catal., B, 2022, 307, 121162.
60. Y. Ke, Q. Liang, S. Zhao, Z. H. Zhang, X. Z. Li and Z. Y. Li, Inorg. Chem., 2022, 61, 2652–2661.
61. J. X. Xu, Y. F. Chen, M. Chen, J. Wang and L. Wang, Chem. Eng. J., 2022, 442, 136208.
62. P. Verma, F. A. Rahimi, D. Samanta, A. Kundu, J. Dasgupta and T. K. Maji, Angew. Chem., Int. Ed., 2022, 61, e202116094.
63. G. M. Ren, S. T. Liu, Z. Z. Li, H. C. Bai, X. D. Hu and X. C. Meng, Sol. RRL, 2022, 6, 2200154.
64. L. B. Wang, B. Cheng, L. Y. Zhang and J. G. Yu, Small, 2021, 17, 2103447.
65. B. X. Ni, H. Jiang, W. Y. Guo, Q. J. Xu and Y. L. Min, Appl. Catal., B, 2022, 307, 121141.
66. Z. R. Miao, Q. L. Wang, Y. F. Zhang, L. P. Meng and X. X. Wang, Appl. Catal., B, 2022, 301, 120802.
67. Y. J. Ma, Q. Tang, W. Y. Sun, Z. Y. Yao, W. H. Zhu, T. Li and J. Y. Wang, Appl. Catal., B, 2020, 270, 118856.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2ta07217c

---