Hollow core–shell heterojunction TAPB-COF@ZnIn₂S₄ as highly efficient photocatalysts for carbon dioxide reduction

Abstract

The conversion of carbon dioxide (CO₂) into carbon-neutral fuels using solar energy is crucial for achieving energy sustainability. However, the high carrier charge recombination and low CO₂ adsorption capacity of the photocatalysts present significant challenges. In this paper, a TAPB-COF@ZnIn₂S₄-30 (TAPB-COFZ-30) heterojunction photocatalyst was constructed by in situ growth of ZnIn₂S₄ (ZIS) on a hollow covalent organic framework (HCOF) with a hollow core–shell structure for CO₂ to CO conversion. Both experimental studies and theoretical calculations indicate that the construction of heterojunctions improves the efficiency of carrier separation and utilisation in photocatalysis. The yield of photoreduction of CO₂ to CO by the TAPB-COFZ-30 heterojunction photocatalyst reached 2895.94 μmol g⁻¹ with high selectivity (95.75%). This study provides a feasible strategy for constructing highly active core–shell composite photocatalysts to optimize CO₂ reduction.

---

Introduction

Dependence on the combustion of fossil fuels to meet energy needs has resulted in the accumulation of large quantities of greenhouse gases in the atmosphere, which has serious environmental consequences; thus, reducing the use of fossil fuels and improving the use of clean energy are the energy challenges we face today.^1–5 Among the many clean energy sources, the utilisation of solar energy has been widely recognized. The use of semiconductors as catalysts for the photoreduction of CO₂ into valuable chemical substances is an innovative green technology, which is considered one of the most promising solutions for reducing CO₂ emissions, both in terms of solving the environmental challenges created by the greenhouse effect and alleviating the problem of energy scarcity.^6–11 Semiconductors were reported for CO₂ photocatalytic reduction at the end of the twentieth century.¹² Although most of the photocatalysts have achieved relatively good results, there are still some problems, such as limited visible light collection, poor selectivity for the desired products, poor utilization of photon-generated carriers, and other problems.^4,13–16

In recent years, among various semiconductor materials, the transition metal sulfide photocatalyst ZIS has attracted great interest by virtue of its suitable electronic energy band structure, good chemical stability, low toxicity and facile preparation.^16–19 However, the photocatalytic activity of ZIS still fails to meet the requirements for a wider range of applications due to the fast recombination and low mobility of light-induced electron–hole pairs.^20,21 Its excessive self-assembly ability also reduces its specific surface area and active sites.²² Therefore, it is necessary to find stable and reliable substrate materials to inhibit the self-loading of ZIS, and optimize the charge separation/transfer and CO₂ adsorption capabilities of ZIS photocatalysts to achieve more efficient CO₂ conversion. In recent years, an emerging class of visible light-responsive porous crystalline materials, covalent organic frameworks (COFs), has attracted much attention due to their abundant porosity, structural diversity, high thermal stability, and low density.

The benefits of COFs are as follows: (1) compared with traditional inorganic or organic semiconductors, their high surface area and plentiful nitrogen atom framework provide higher adsorption capacity and CO₂ selectivity; (2) improved thermal and chemical stability; (3) regular crystal structure and uniform channel distribution, reduced charge recombination, enhanced mass transfer ability, etc.^9,23–26 The adsorption performance of COFs on CO₂ has been greatly improved with the rapid development of research since their first report in 2009.²⁷ Considering these advantages, COF materials can be used as ideal components coupled with ZIS to construct composite photocatalysts. Among the reported COF materials, nitrogen-rich COFs such as TAPB are expected to be excellent candidates for CO₂ capture owing to the increased affinity of nitrogen atoms for interacting with CO₂.²⁸

Therefore, in this work, the nitrogen-rich TAPB-COF was synthesized from 1,3,5-tris(4-aminophenyl)benzene (TAPB) and 2,5-diethenyl-1,4-benzenedicarboxaldehyde (DVA), and the hollow core–shell photocatalyst TAPB-COF@ZIS was constructed by a one-step hydrothermal method for the photocatalytic reduction of CO₂. The CO₂ reduction mechanism was revealed by in situ Fourier transform infrared spectroscopy (in situ FTIR) and Gibbs free energy calculations. X-ray photoelectron spectroscopy (XPS) and theoretical calculations show that the carrier migration and separation processes in TAPB-COF@ZIS follow a type-II transfer mechanism. Photoelectrochemical tests and N₂, CO₂ adsorption results showed that the incorporation of TAPB-COF promoted the separation and transport of photogenerated carriers, increased the specific surface area of ZIS and CO₂ adsorption. This unique photocatalyst, TAPB-COFZ-30, has the optimal TAPB-COF content, with higher photocatalytic activity than that of the original ZIS (about 6 folds) and TAPB-COF (about 33 folds). This work highlights the potential application of ZIS-based photocatalysts for CO₂ photoreduction. Furthermore, it provides a new strategy for the design of COF@ZIS composite materials, and further lays a foundation for the development of CO₂ reduction photocatalysts.

Experimental section

Materials

Hydrogen peroxide (H₂O₂) was purchased from Tianjin Fengchuan Chemical Reagent Co., Ltd. N′N-Dimethylformamide (DMF), acetonitrile (MeCN), ethanol, and acetic acid (HAc), were purchased from Tianjin Comieo Chemical Reagent Co., Ltd. 1,3,5-Tris(4-aminophenyl)benzene (TAPB), 2,5-diethenyl-1,4-benzenedicarboxaldehyde (DVA), hydrochloric acid (HCl), thioacetamide (TAA), indium(III) chloride tetrahydrate (InCl₃·4H₂O), zinc chloride (ZnCl₂), glycerol, triethanolamine (TEOA), and 2,2′-bipyridine (2,2′-bpy) were obtained from Aladdin Reagent Co., Ltd. All reagents used were of analytical grade and were used without further purification.

Synthesis of TAPB-COF

The TAPB-COF was synthesized with some minor modifications on the basis of the original method.²⁹ In a typical synthesis, 56 mg of TAPB (0.16 mmol), 44 mg of DVA (0.24 mmol), and 80 mL of MeCN were added to a 400 mL beaker. The mixture was sonicated until the two monomers were dissolved completely. Afterwards, 8 mL of HAc was added to the mixture, and stirred in a fume hood at room temperature until the solvent was almost completely volatilized. The product was collected by centrifugation and washed with ethanol three times. The obtained TAPB-DVA-COF was dispersed in 80 mL DMF. Then, 2 mL of HAc, 1.6 mL of FeCl₃/MeCN (50 mg mL⁻¹), and 16 mL of 3% H₂O₂ were added successively, and left for 20 min in an oven at 60 °C. The product was collected by centrifugation and washed with ethanol three times. The product was dried overnight at 60 °C. The product was labeled as TAPB-COF.

Synthesis of TAPB-COFZ-X

36.26 mg of TAPB-COF was added to 8 mL of H₂O and stirred for 5 min, then the pH was adjusted to 2.5 with hydrochloric acid, followed by the addition of 3 mL of glycerol and stirred for 30 min. After that, 54.5 mg of ZnCl₂, 117.3 mg of InCl₃·4H₂O, and 60.1 mg of TAA were added to the above solution and stirred for 10 min, and then reacted in an oil bath at 80 °C for 2 h. After the reaction, the product was washed three times with water and anhydrous ethanol, centrifuged, and dried at 60 °C for 12 h. The resulting product was named TAPB-COFZ-30 (the suffix indicated the calculated weight percentage of TAPB-COF). By adjusting the amount of TAPB-COF added, with no change in the amount of other raw materials added, two other samples were added with 21.2 mg of TAPB-COF and 56.4 mg of TAPB-COF, and the resulting products were designated as TAPB-COFZ-20 and TAPB-COFZ-40, respectively.

Synthesis of ZIS

ZIS was synthesized with some minor modifications according to the previous method.^30,31 In a typical synthesis, 8 mL of H₂O was added to a 25 mL beaker and the pH was adjusted to 2.5 with hydrochloric acid; then 3 mL of glycerol was added and stirred for 3 min. After that, 54.5 mg of ZnCl₂, 117.3 mg of InCl₃·4H₂O, and 60.1 mg of TAA were added to the above solution and stirred for 10 min, and then reacted in an oil bath at 80 °C for 2 h. After the reaction, the product was washed three times with water and anhydrous ethanol, centrifuged, and dried at 60 °C for 12 h.

Characterization

The morphology and structure of the samples were examined by field emission scanning electron microscopy (FESEM; Zeiss Sigma-500) and transmission electron microscopy (TEM; JEOL, JEM-2100). In addition, the elemental mapping images was also recorded by TEM. The crystal phases of the samples were analyzed by powder X-ray diffraction (PXRD) under Cu-Kα radiation on a Bruker D8 advanced X-ray diffractometer in the scanning ranges of 0.5°–10° and 10°–80°, respectively. X-ray photoelectron spectroscopy (XPS, EscaLab Xi⁺) was used to observe the state of the surface elements using standard C 1s (284.6 eV) as a reference. UV-visible diffuse reflectance spectra (UV-vis DRS) were recorded with a Hitachi UH4150 UV-vis spectrophotometer. FTIR spectra were recorded using a Tracer-100 FT-IR spectrometer (Shimadzu Corporation). Photoluminescence (PL) spectra was obtained using a FLS980 from Edinburgh Analytical Instruments. Moreover, N₂ adsorption–desorption isotherms and CO₂ adsorption capacity were determined using a Quantachrome surface area analyzer. The specific surface area was obtained by the BET method and the pore size distribution was determined by isothermal adsorption branching. Light intensity was measured using a PL-MW 2000 photoradiometer (Perfect Light, Table S1†).

Photocatalytic CO₂ reduction

The catalyst activity was tested by photocatalytic CO₂ reduction experiments. A 60 mL reactor was charged with 5 mg of photocatalyst, 15 mg of 2,2′-bipyridine, 2 μmol of CoCl₂, 1 mL of TEOA, 2 mL of H₂O, and 3 mL of MeCN, and purified with CO₂. A 300 W Xe lamp was used as the irradiation source. In each experiment, the reaction temperature was maintained at 30 °C under magnetic stirring. The concentration of the products was determined using a Foley GC970II gas chromatograph.

Photochemical testing

Photoelectrochemical experiments were carried out using a CHI660E electrochemical workstation. The test system consisted of an Ag/AgCl electrode (reference electrode), a Pt sheet (counter electrode), a photocatalyst (working electrode). The working electrode was a glassy carbon electrode (GCE) with a photocatalyst film: 15 mg of the sample was ground homogeneously, mixed with 950 μL of ethanol and 50 μL of naphthol, sonicated for 60 min, and then 10 μL of the mixture was applied to the GCE, which was dried naturally at room temperature.

Results and discussion

Photocatalyst synthesis and characterization

The synthesis route of the photocatalyst is shown in Fig. 1a. First, the TAPB-DVA-COF was synthesized by condensation reaction using TAPB and DVA as synthetic monomers; then the TAPB-COF was derived by partially breaking the imine bond, and finally, using a straightforward hydrothermal technique, the TAPB-COFZ photocatalyst was successfully created. FESEM and TEM were used to examine the appearance and morphological information of the samples. Fig. S1 and S2† show that both the TAPB-DVA-COF and TAPB-COF are roughly spherical, except that the TAPB-COF has a hollow structure. From Fig. 1, it can be seen that the TAPB-COFZ-30 photocatalyst retained the hollow spherical shape of the TAPB-COF, and the ZIS nanosheets were uniformly grown on the TAPB-COF without excessive self-assembly (Fig. 1b and c), which was favorable for exposing more active sites. In the high-resolution transmission electron microscopy (HRTEM) image (Fig. 1d), the lattice stripes with an interfacial spacing of 0.32 nm correspond to the (102) face of ZIS, which is consistent with previous reports.^16,32 The contact interface between the TAPB-COF and ZIS is indicated by the yellow dashed line, and it can be seen that the ZIS crystal face is perfectly adhered to the TAPB-COF crystal face, which indicates the successful formation of the TAPB-COFZ-30 composite (Fig. 1d). An energy dispersive X-ray detector (EDX) was used to obtain elemental mapping images (Fig. 1e). The images showed that S, In, Zn and C, N, O were well distributed in TAPB-COFZ-30, which further proved the successful formation of the TAPB-COFZ-30 composite.

Fig. 1 (a) Synthesis route of the photocatalyst TAPB-COFZ-X. (b) FESEM, (c) TEM, (d) HRTEM, (e) EDX elemental mapping images of TAPB-COFZ-30.

The PXRD spectra of the prepared samples are shown in Fig. S3.† Each PXRD peak in the spectra demonstrates the high crystallinity of the TAPB-COF and the successful complexation of the photocatalyst. As shown in Fig. S3a,† there is a strong diffraction peak at 2.73°, and several weak diffraction peaks at 4.80°, 5.54°, 7.37° and 9.70°, which belong to the (100), (110), (200), (210) and (220) crystal planes of TAPB-COF, respectively. The result is consistent with previous reports,²⁹ indicating the successful synthesis of the TAPB-COF. With the increase in TAPB-COF content, the diffraction peaks of TAPB-COFZ-x (x = 20, 30, 40) at 2.73° were gradually enhanced, indicating the successful composite formation of the photocatalyst. As shown in Fig. S3b,† TAPB-COFZ-x (x = 20, 30, 40) perfectly inherited the diffraction peaks of ZIS and matched well with the pure hexagonal structure of ZIS (JCPDS No. 65-2023), which further indicated the successful composite formation of the photocatalyst. It is worth noting that although the diffraction peaks of the TAPB-COFZ-X heterostructure are very similar to those of COF and ZIS, there is a significant displacement at the peaks of COF (001) and ZIS (006). This may be due to the growth of ZIS along the COF (001) facet, and the strong interaction between the COF (001) facet and the ZIS (006) facet caused the ZIS layered structure to extend the d-spacing along the c-axis.³³

XPS tests of the surface chemistry of pristine ZIS, TAPB-COF and TAPB-COFZ-30 were performed. In Fig. S4,† the XPS survey spectra show that the TAPB-COFZ-30 composite contains the elements C, N, O, Zn, In, and S, which is consistent with the EDX elemental mapping. As shown in Fig. 2a, the fitted peaks located at 285.95 eV and 287.7 eV are attributed to the C=N and C=O bonds in TAPB-COF. The O 1s peak binding energies located at 532.80 eV, 531.76 eV and 531 eV are attributed to the C=O, C–O bonds and absorbed oxygen in the TAPB-COF (Fig. 2b). The N 1s peak binding energies located at 400.41 eV and 399.08 eV are attributed to the C–NH₂ and C=N–C bonds in the TAPB-COF (Fig. 2c). For Zn 2p, peaks located at 1021.74 eV and 1044.77 eV are attributed to the Zn 2p_3/2 and Zn 2p_1/2 orbitals,³⁴ respectively (Fig. 2d). The two soliton rotation peaks in Fig. 2e are divided into 3d_3/2 (452.45 eV) and 3d_5/2 (444.9 eV) for In 3d in the +3 valence state.³⁵ In Fig. 2f, the peaks centered at 161.74 and 162.95 eV, assigned to the 2p_2/3 and 2p_1/2 states of S, respectively, are characteristic of S²⁻ and are in agreement with the literature.³⁶

Fig. 2 High-resolution XPS spectra of TAPB-COF and TAPB-COFZ-30: (a) C 1s, (b) O 1s, (c) N 1s, (d) Zn 2p, (e) In 3d, and (f) S 2p.

In general, the density of the atomic electron cloud is closely related to variations in the elemental binding energy. For example, the electron cloud density decreases with increasing binding energy, and the electron cloud density increases with decreasing binding energy.^37,38 This is due to the fact that the atomic electron binding energy is determined by the strong coulombic force between the outer electrons and the nucleus.³⁸ Hence, this shift in binding energy can be used in heterojunction catalyst systems to determine the direction of electron transfer.³⁹ The binding energies of C 1s, N 1s and O 1s in the TAPB-COFZ-30 composites tend to move to higher energy levels compared to those in TAPB-COF, suggesting that the incorporation of ZIS leads to a decrease in the electron cloud density of the TAPB-COF. In contrast, the binding energies of Zn 2p, In 3d, and S 2p of TAPB-COFZ-30 composites tend to move to lower energy levels compared to pure ZIS, indicating that the incorporation of TAPB-COF leads to an increase in the electron cloud density of ZIS, proving that the electrons tend to migrate from TAPB-COF to ZIS.

The characteristic functional groups of the photocatalysts were identified by FTIR. The absorption peak around 1608 cm⁻¹ for the TAPB-COF is attributed to the C=N bond and is relatively weakened; besides, the absorption peaks for C=O (1662 cm⁻¹) and –OH (3355 cm⁻¹) are enhanced, suggesting that the C=N bond breaks to form –COOH (Fig. S5†).²⁹ In Fig. S6†, the absorption peaks about 1650 cm⁻¹ and 3440 cm⁻¹ in ZIS are attributed to surface-absorbed water molecules and hydroxyl groups.^40,41 Similarly, TAPB-COFZ has the same peaks, which further indicates the successful preparation of photocatalysts.

The optical absorption capabilities and electronic energy band structures of the photocatalysts were investigated using UV-vis diffuse reflectance spectroscopy (UV-vis DRS) and Mott–Schottky (M–S) measurements. As shown in Fig. S7,† ZIS shows visible absorption edges at ∼500 nm, and TAPB-COF and TAPB-COFZ-x (x = 20, 30, and 40) show visible absorption fringes at ∼750 nm, and the incorporation of TAPB-COF leads to a significant red shift of the absorption edges. This indicates that the composite with the TAPB-COF broadens the light absorption range of ZIS. The corresponding energy band gaps of TAPB-COF and ZIS were obtained from the Tauc plots of (αhν)ⁿ (n = 2) versus hν. The band gap values of TAPB-COF and ZIS were 2.11 eV and 2.76 eV, respectively (Fig. S8a and b†). Generally, large absorption edges typically correspond to a narrow bandgap and potentially strong photovoltaic conversion capabilities.⁴²

Positive slopes on the M–S plots of ZIS and TAPB-COF suggest that they have n-type semiconductor characteristics.⁴³ The CB potential of n-type semiconductors is commonly considered to be near the flat band potential.³⁷ Hence, from the M–S tangents (Fig. S8c and d†), the flat-band potentials of TAPB-COF and ZIS are −1.68 and −1.5 V (vs. Ag/AgCl, pH ∼ 7), which correspond to the CB values of TAPB-COF (−1.48 V) and ZIS (−1.3 V), respectively, with respect to the normal hydrogen electrode (NHE). The CB of the photocatalyst is higher than the reduction potential required for CO₂ reduction to CO (−0.53 V versus NHE), which confirmed the thermodynamic feasibility of CO₂ reduction by TAPB-COFZ-30. The VB value of the samples can be defined using the equation E_CB = E_VB − E_g,⁴⁴ which results in 0.63 V and 1.46 V for TAPB-COF and ZIS, respectively.

Photocatalytic properties of CO₂ reduction

The CO₂ photoreduction activity of the materials was investigated in acetonitrile/water mixtures under visible light using TEOA as the electron donor and Co(bpy)₃²⁺ (bpy = 2,2′-bipyridine) as the co-catalyst. As shown in Fig. 3a, the CO production rate of TAPB-COFZ-30 was measured to be 2895.94 μmol g⁻¹, apparently higher than that of ZIS (466.63 μmol g⁻¹) and TAPB-COF (87.42 μmol g⁻¹). Except for small amounts of CH₄ and H₂, no other reduction by-products were detected after the 4 h testing period (Fig. 3b), and the CO selectivity of TAPB-COFZ-30 was as high as 95.75% (Table S2†). To verify the origin of the product CO, control experiments were performed under dark conditions and without a catalyst, and no photocatalytic reduction reaction occurred, indicating that CO originated from the photocatalytic CO₂ reduction reaction driven by light irradiation. N₂ was used to replace CO₂ in the photoreduction experiment, and only a small amount of H₂ was generated, and no CO and CH₄ were generated, indicating that CO produced in the reaction originated from CO₂. In the absence of TEOA, no obvious light attenuation process was observed. When Co(bpy)₃²⁺ is not added, there are fewer photoreduction products, which indicates that Co(bpy)₃²⁺ is crucial for promoting the conversion of CO₂ to CO, indirectly indicating that the main active center of photoreduction is Co(bpy)₃²⁺, which is consistent with previous studies on such CO₂ photoreduction systems (Fig. S9†). Moreover, ¹³C isotope labeling confirmed that the CO product predominantly originated from CO₂, as shown in Fig. 3c (m/z = 13, 16, 29).⁴⁵ The catalytic performance of photocatalysts for the CO₂ photoreduction reaction at different wavelengths was investigated. It can be seen from Fig. S10† that the gas evolution trend matches well with the light absorption intensity of the catalyst material, indicating that the photo-excitation of the photocatalyst triggers the CO₂ reduction reaction and realizes the generation of electric charge. Four cycles of CO₂ photoreduction experiments were carried out on TAPB-COFZ-30 under the same conditions, with each cycle lasting for 4 h, to evaluate the optical stability of the TAPB-COFZ-30 photocatalyst (Fig. 3d). The efficiency of CO generation is comparable or even higher than that of most photocatalysts reported in recent years, as shown in Fig. 3e and Table S3.†

Fig. 3 (a) CO₂ photoreduction performance of different samples. (b) Time-dependent CO/CH₄/H₂ yields. (c) GC-MS spectrum of CO generated from the ¹³CO₂ isotope experiment. (d) CO₂ photoreduction stability of TAPB-COFZ-30. (e) Comparison of photocatalytic CO₂ reduction performance with recently reported photocatalysts.

Photochemical performance analysis

Photoelectrochemical properties are widely considered as a strong indicator of the efficiency of photogenerated charge carrier separation and migration in photocatalysts.⁴⁶ The efficacy of photogenerated carrier separation was evaluated using photoluminescence spectroscopy (PL). As shown in Fig. 4a, the PL peak intensity of TAPB-COFZ-30 is significantly lower than that of the original ZIS, implying that more electrons can participate in the photocatalytic reaction due to the inhibition of ZIS light-induced electron–hole pair recombination caused by heterojunction formation.³⁴ Time-resolved photoluminescence spectroscopy (TRPL) was used to further investigate the carrier dynamics of TAPB-COFZ-30 and a double-exponential kinetic function was used to fit the emission decay curve of the sample (Table S4†). As shown in Fig. 4b, the average lifetime of the pure ZIS was 1.68 ns, while the average lifetime of the TAPB-COFZ-30 composite sample increased to 2.17 ns. The significant increase in the lifetime of TAPB-COFZ-30 suggests that it effectively promotes the photo-induced separation of electrons and holes.³² From electrochemical impedance spectroscopy (EIS), it can be seen that the semicircle of the TAPB-COFZ-30 catalyst is smaller than that of pure ZIS (Fig. 4c), which suggests that the charge transfer resistance at the electrolyte interface of TAPB-COFZ-30 is small, and interfacial charge transfer is faster.^47,48 As shown in Fig. 4d, the photocurrent of TAPB-COFZ-30 is much higher than that of ZIS, indicating that TAPB-COFZ-30 enhances the charge transfer at the interface while suppressing the compounding rate of photogenerated electrons and holes.⁴⁹

Fig. 4 (a) Steady-state PL spectra, (b) TRPL spectra of the prepared photocatalysts, (c) EIS spectra, and (d) transient photocurrent responses of ZIS and TAPB-COFZ-30.

The specific surface area of the photocatalysts as well as their CO₂ adsorption capacity are vital factors influencing photoreduction performance.⁶ N₂ and CO₂ adsorption characterization of TAPB-COF, ZIS and TAPB-COFZ-30 composites was performed, and the results are shown in Fig. S11.† The specific surface area of TAPB-COF was found to be 1042.98 m² g^—1, that of ZIS was 38.69 m² g⁻¹, and that of TAPB-COFZ-30 was 196.93 m² g⁻¹ (Fig. S11a–c†). This indicates that the incorporation of TAPB-COF effectively suppressed the agglomeration of ZIS nanosheets, increased the specific surface area of ZIS, and provided more catalytic active sites for CO₂ adsorption. Fig. S11d† shows the adsorption capacities of TAPB-COF, ZIS, and TAPB-COFZ-30 for CO₂, and it is evident that the incorporation of TAPB-COF increased the adsorption capacity of ZIS for CO₂, which enhanced the photocatalytic overall performance of the photocatalyst.

To further investigate the photocatalytic mechanism, the theoretical structures of the COF and ZIS were optimized and calculated. As shown in Fig. 5a and b, the density of states (DOS) indicates that C atoms dominate the COF valence and conduction band electrons, S atoms dominate the ZIS valence band electrons, and S and In atoms dominate the electron orbitals in the conduction band, respectively. Moreover, the estimated bandgaps of TAPB-COF and ZIS were found to be 1.98 eV and 2.52 eV, respectively, that are lower than our experimental results (2.11 eV and 2.76 eV). Since the accuracy of the exchange-correlation potential calculated by using density-functional theory (DFT) is limited, it is common for the calculated values to be smaller than the experimental values.⁵⁰ DFT calculations were used to investigate the electronic structure and charge transfer direction of the COF@ZIS composites, and the work function (Φ), which determines the charge transport at the interface, was obtained. As shown in Fig. 5c and d, the Φ values for COF and ZIS surfaces were 4.32 eV and 6.04 eV. The charge density distribution of the COF@ZIS heterojunction is shown in Fig. 5e, where the yellow colour represents the electron accumulation region and the cyan colour represents the electron depletion region. It can be observed that the surface of the TAPB-COF is dominated by the cyan colour, indicating charge depletion, while the surface of ZIS is dominated by a yellow colour, indicating charge accumulation. These results theoretically confirm the direction of electron transfer from the COF to ZIS.

Fig. 5 (a) TAPB-COF, (b) ZIS density of states (DOS), (c) TAPB-COF, (d) ZIS surfaces for the work function, and (e) average charge density difference for TAPB-COFZ-30.

Meanwhile, to investigate the role of the Co(bpy)₃²⁺ co-catalyst in the whole photoreduction process, a steady-state fluorescence test was performed on the reaction system. As shown in Fig. S12,† the PL intensity of the mixture decreased significantly after adding Co(bpy)₃²⁺ to the system, which indicated that Co(bpy)₃²⁺ can inhibit the recombination of photogenerated carriers in the system.⁵¹ At the same time, the Co(bpy)₃²⁺ co-catalyst is an indispensable active site in the photoreduction process, which can be seen from previous comparative test and reports.^52,53 Moreover, DFT calculations also showed that Co(bpy)₃²⁺ can spontaneously adsorb on the surface of ZIS (Fig. S13†).

In order to deeply explore the reaction intermediates in the photocatalytic carbon dioxide reduction process, we carried out in situ FTIR spectroscopic studies. As shown in Fig. 6a, there were no obvious characteristic peaks before light exposure, while a series of intermediate peaks appeared after light exposure. With increasing irradiation time, the peaks included HCO₃⁻ (1435, 1688, 1783 cm⁻¹),^40,54,55 m-CO₃²⁻ (1283, 1513 cm⁻¹),^56,57 b-CO₃²⁻ (1356, 1585, 1660 cm⁻¹),^40,58,59 and CO₂⁻ (1708 cm⁻¹),⁶⁰ indicating absorption and activation of CO₂ and H₂O onto the photocatalyst. Characteristic peaks of *COOH can also be found, which is usually considered the key intermediate for CO₂ reaction to CO or methane (1530, 1550, 1570, and 1633 cm⁻¹).^{54–56,61–63} The peak at 2078 cm⁻¹ belongs to *CO and marks the formation of the final product.⁶⁴ In addition, the photoreduction mechanisms of ZIS and TAPB-COFZ-30 were further investigated by Gibbs free energy calculations. Based on the in situ FTIR results of TAPB-COFZ-30, it is presumed that the photoreduction reaction proceeds through CO₂(g), *+CO₂, *COOH, *CO, and *+CO(g) (Fig. 6b and S14–S18†). The DFT results show that the formation of *COOH has the highest energy barrier, suggesting that this step is the rate-limiting step for the photoreduction of CO₂ to CO.⁶⁵ Among them, TAPB-COFZ-30 generates *COOH with a ΔG of 1.02 eV, which is much lower than that of ZIS (ΔG = 2.58 eV). It demonstrated that the creation of the heterojunction was useful for optimizing the rate-limiting step, thus accelerating the CO₂ reduction reaction kinetics. In addition, TAPB-COFZ-30 exhibited a lower energy barrier than pure ZIS in the step from *CO to *+CO(g), suggesting that CO is more easily desorbed from TAPB-COFZ-30. From the results of the ΔG calculations above, it can be shown that the formation of the heterojunction can decrease the reaction energy barrier during the CO₂ photoreduction process, which is one of the important reasons for TAPB-COFZ-30's improved photocatalytic performance for CO₂ reduction to CO. Moreover, the calculations show that the Gibbs free energy for generating *HCO from *CO is greater than that for generating CO(g), which explains the high selectivity for the reduction of CO₂ to CO.

Fig. 6 (a) In situ FTIR spectra of the photoreduction process of CO₂ on TAPB-COFZ-30 under different irradiation times. (b) Gibbs free energy diagrams of CO₂ reduction to CO for ZIS and TAPB-COFZ-30 photocatalysts. (“*” represents adsorption on the substrate; the figure shows the calculation model of TAPB-COFZ-30 adsorption energy).

Therefore, a type-II transfer mechanism (Fig. 7) at the TAPB-COFZ heterojunction interface is proposed after combining our previously described experimental results and theoretical calculations on the charge transfer dynamics between the two energy levels of COF@ZIS. The potential difference resulting from the different work functions of COF and ZIS drives the electrons to diffuse from the COF to ZIS, resulting in ZIS acquiring electrons, whereas the COF becomes electron-deficient. Meanwhile, the trapping of holes by the sacrificial agent TEOA on the VB of the COF can effectively avoid the occurrence of reversible reactions, which makes it easier to separate the photogenerated electrons and holes; CO₂ is reduced to CO in the presence of ZIS and co-catalysts.

Fig. 7 Proposed reaction mechanism for photocatalytic CO₂ reduction on the TAPB-COFZ-30 hollow core–shell heterojunction.

Conclusions

By in situ growth of ZIS on the surface of hollow TAPB-COF nanospheres, TAPB-COFZ heterojunctions with hollow core–shell structures were constructed. For the photocatalyst, the nitrogen-rich TAPB-COF can effectively adsorb and activate CO₂, while the heterojunction interface effectively promotes carrier separation and transfer. The TAPB-COFZ photocatalysts exhibited an efficient CO₂ reduction rate and excellent stability in the water/acetonitrile (H₂O/MeCN) tandem system. The key *COOH intermediate in the conversion of CO₂ to CO was confirmed by in situ FTIR and density-functional theory calculations, further supporting the proposed reaction mechanism. This work bridges the research on COF@ZIS heterojunctions for CO₂ reduction and may encourage the development of photocatalysts with more desirable COF heterojunction-based photocatalysts, as well as the coupling of molecular co-catalysts for efficient CO₂ photo-fixation.

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Hui-Tao Fan: conceptualization, lead; data curation, equal; supervision, lead; visualization, lead; writing – original draft, lead; writing – review & editing, lead. Minglin Hu: data curation, supporting; formal analysis, lead; investigation, equal; methodology, equal; writing – original draft, equal. Yabing Duan: data curation, equal; investigation, equal; methodology, equal. Luyang Zuo: investigation, equal; software, lead. Ronggui Yu: data curation, supporting; investigation, equal. Zhuwei Li: investigation, supporting; methodology, supporting. Qi Liu: data curation, equal; formal analysis, equal; investigation, equal. Bo Li: funding acquisition, lead; resources, lead; supervision, equal. Liya Wang: supervision, equal; writing – review & editing, equal.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the Program for Science and Technology Innovation Talents in Universities of Henan Province (No. 22HASTIT007), and the Special Foundation of Nanyang Normal University (2024ZX004).

References

1. S. Chu and A. Majumdar, Nature, 2012, 488, 294–303.
2. P. Prabhu, V. Jose and J. M. Lee, Adv. Funct. Mater., 2020, 30, 1910768.
3. A. R. Woldu, P. Talebi, A. G. Yohannes, J. Xu, X. D. Wu, S. Siahrostami, L. Hu and X. C. Huang, Angew. Chem., Int. Ed., 2023, 62, e202301621.
4. Q. Wang and K. Domen, Chem. Rev., 2019, 120, 919–985.
5. X. Lin, S. Xia, L. Zhang, Y. Zhang, S. Sun, Y. Chen, S. Chen, B. Ding, J. Yu and J. Yan, Adv. Mater., 2022, 34, 2200756.
6. G. Zhang, X. Li, D. Chen, N. Li, Q. Xu, H. Li and J. Lu, Adv. Funct. Mater., 2023, 33, 2308553.
7. Y. Feng, C. Wang, P. Cui, C. Li, B. Zhang, L. Gan, S. Zhang, X. Zhang, X. Zhou, Z. Sun, K. Wang, Y. Duan, H. Li, K. Zhou, H. Huang, A. Li, C. Zhuang, L. Wang, Z. Zhang and X. Han, Adv. Mater., 2022, 34, 2109074.
8. F. Chen, Z. Ma, L. Ye, T. Ma, T. Zhang, Y. Zhang and H. Huang, Adv. Mater., 2020, 32, 1908350.
9. K. Guo, X. Zhu, L. Peng, Y. Fu, R. Ma, X. Lu, F. Zhang, W. Zhu and M. Fan, Chem. Eng. J., 2021, 405, 127011.
10. A. Nakada, H. Kumagai, M. Robert, O. Ishitani and K. Maeda, Acc. Mater. Res., 2021, 2, 458–470.
11. R. T. Guo, X. Hu, X. Chen, Z. X. Bi, J. Wang and W. G. Pan, Small, 2023, 19, 2207767.
12. T. Inoue, A. Fujishima, S. Konishi and K. Honda, Nature, 1979, 277, 637–638.
13. C. Dong, C. Lian, S. Hu, Z. Deng, J. Gong, M. Li, H. Liu, M. Xing and J. Zhang, Nat. Commun., 2018, 9, 1252.
14. C. Gao, T. Wei, Y. Zhang, X. Song, Y. Huan, H. Liu, M. Zhao, J. Yu and X. Chen, Adv. Mater., 2019, 31, 1806596.
15. H. Ou, L. Lin, Y. Zheng, P. Yang, Y. Fang and X. Wang, Adv. Mater., 2017, 29, 1700008.
16. H. Fan, Y. Jin, K. Liu and W. Liu, Adv. Sci., 2022, 9, 2104579.
17. X. Zheng, Y. Song, Y. Liu, Y. Yang, D. Wu, Y. Yang, S. Feng, J. Li, W. Liu, Y. Shen and X. Tian, Coord. Chem. Rev., 2023, 475, 214898.
18. W. Yang, L. Zhang, J. Xie, X. Zhang, Q. Liu, T. Yao, S. Wei, Q. Zhang and Y. Xie, Angew. Chem., Int. Ed., 2016, 55, 6716–6720.
19. Y. Pan, X. Yuan, L. Jiang, H. Yu, J. Zhang, H. Wang, R. Guan and G. Zeng, Chem. Eng. J., 2018, 354, 407–431.
20. G. Zuo, Y. Wang, W. L. Teo, A. Xie, Y. Guo, Y. Dai, W. Zhou, D. Jana, Q. Xian, W. Dong and Y. Zhao, Angew. Chem., Int. Ed., 2020, 59, 11287–11292.
21. G. Zuo, Y. Wang, W. L. Teo, Q. Xian and Y. Zhao, Appl. Catal., B, 2021, 291, 120126.
22. S. Wang, Y. Wang, S. L. Zhang, S. Q. Zang and X. W. Lou, Adv. Mater., 2019, 31, 1903404.
23. S. Chandra, T. Kundu, S. Kandambeth, R. BabaRao, Y. Marathe, S. M. Kunjir and R. Banerjee, J. Am. Chem. Soc., 2014, 136, 6570–6573.
24. H. Ma, B. Liu, B. Li, L. Zhang, Y.-G. Li, H.-Q. Tan, H.-Y. Zang and G. Zhu, J. Am. Chem. Soc., 2016, 138, 5897–5903.
25. A. P. Côté, A. I. Benin, N. W. Ockwig, M. O'Keeffe, A. J. Matzger and O. M. Yaghi, Science, 2005, 310, 1166–1170.
26. H. M. El-Kaderi, J. R. Hunt, J. L. Mendoza-Cortés, A. P. Côté, R. E. Taylor, M. O'Keeffe and O. M. Yaghi, Science, 2007, 316, 268–272.
27. H. Furukawa and O. M. Yaghi, J. Am. Chem. Soc., 2009, 131(25), 8875–8883.
28. Y. Zeng, R. Zou and Y. Zhao, Adv. Mater., 2016, 28, 2855–2873.
29. S. Liu, K. Dou, B. Liu, M. Pang, P. a. Ma and J. Lin, Angew. Chem., Int. Ed., 2023, 62, e202301831.
30. S. Wang, B. Y. Guan and X. W. D. Lou, J. Am. Chem. Soc., 2018, 140, 5037–5040.
31. M. Zhou, S. Li, S. Wang, Z. Jiang, C. Yang, F. Guo, X. Wang and W.-K. Ho, Appl. Surf. Sci., 2022, 599, 153985.
32. S. Wang, B. Y. Guan, X. Wang and X. W. D. Lou, J. Am. Chem. Soc., 2018, 140, 15145–15148.
33. C. Cui, X. Xu, X. Zhao, N. Xi, M. Li, X. Wang, Y. Sang, X. Yu, H. Liu and J. Wang, Nano Energy, 2024, 126, 109632.
34. C. Q. Li, X. Du, S. Jiang, Y. Liu, Z. L. Niu, Z. Y. Liu, S. S. Yi and X. Z. Yue, Adv. Sci., 2022, 9, 2201773.
35. S. Lu, S. Zhang, L. Li, C. Liu, Z. Li and D. Luo, Chem. Eng. J., 2024, 483, 149058.
36. X. Jia, Y. Lu, K. Du, H. Zheng, L. Mao, H. Li, Z. Ma, R. Wang and J. Zhang, Adv. Funct. Mater., 2023, 33, 2304072.
37. S. Bao, Q. Tan, S. Wang, J. Guo, K. Lv, S. A. C. Carabineiro and L. Wen, Appl. Catal., B, 2023, 330, 122624.
38. P. Zhang, Y. Li, Y. Zhang, R. Hou, X. Zhang, C. Xue, S. Wang, B. Zhu, N. Li and G. Shao, Small Methods, 2020, 4, 2000214.
39. J. Low, B. Dai, T. Tong, C. Jiang and J. Yu, Adv. Mater., 2019, 31, 1802981.
40. M. Song, X. Song, X. Liu, W. Zhou and P. Huo, Chin. J. Catal., 2023, 51, 180–192.
41. M. Zhou, S. Wang, P. Yang, Z. Luo, R. Yuan, A. M. Asiri, M. Wakeel and X. Wang, Chem.–Eur. J., 2018, 24, 18529–18534.
42. M. Liu, Y. Ning, M. Ren, X. Fu, X. Cui, D. Hou, Z. Wang, J. Cui and A. Lin, Small, 2023, 19, 2303876.
43. P. Xia, S. Cao, B. Zhu, M. Liu, M. Shi, J. Yu and Y. Zhang, Angew. Chem., Int. Ed., 2020, 59, 5218–5225.
44. T. Zhang, F. Meng, M. Gao, W. L. Ong, K. G. Haw, T. Ding, G. W. Ho and S. Kawi, EcoMat, 2021, 3, e12152.
45. S. Wang, X. Hai, X. Ding, S. Jin, Y. Xiang, P. Wang, B. Jiang, F. Ichihara, M. Oshikiri, X. Meng, Y. Li, W. Matsuda, J. Ma, S. Seki, X. Wang, H. Huang, Y. Wada, H. Chen and J. Ye, Nat. Commun., 2020, 11, 1149.
46. K. Dai, L. Lu, C. Liang, Q. Liu and G. Zhu, Appl. Catal., B, 2014, 156–157, 331–340.
47. Y. Zheng, J. Dong, C. Huang, L. Xia, Q. Wu, Q. Xu and W. Yao, Appl. Catal., B, 2020, 260, 118220.
48. Z. Lin, C. Shao, S. Jiang, C. Sun and S. Song, Appl. Catal., B, 2020, 268, 118742.
49. Z. Wang, Y. Chen, L. Zhang, B. Cheng, J. Yu and J. Fan, J. Mater. Sci. Technol., 2020, 56, 143–150.
50. M. Dong, P. Zhou, C. Jiang, B. Cheng and J. Yu, Chem. Phys. Lett., 2017, 668, 1–6.
51. R. Wang, P. Yang, S. Wang and X. Wang, J. Catal., 2021, 402, 166–176.
52. B. Su, H. Huang, Z. Ding, M. B. J. Roeffaers, S. Wang and J. Long, J. Mater. Sci. Technol., 2022, 124, 164–170.
53. L. Huang, B. Li, B. Su, Z. Xiong, C. Zhang, Y. Hou, Z. Ding and S. Wang, J. Mater. Chem. A, 2020, 8, 7177–7183.
54. Y. Zhang, J. Li, W. Zhou, X. Liu, X. Song, S. Chen, H. Wang and P. Huo, Appl. Catal., B, 2024, 342, 123449.
55. Y. Zhang, M. Xu, W. Zhou, X. Song, X. Liu, J. Zhang, S. Chen and P. Huo, J. Colloid Interface Sci., 2023, 650, 1762–1772.
56. S. Si, H. Shou, Y. Mao, X. Bao, G. Zhai, K. Song, Z. Wang, P. Wang, Y. Liu, Z. Zheng, Y. Dai, L. Song, B. Huang and H. Cheng, Angew. Chem., Int. Ed., 2022, 61, e202209446.
57. J. Wang, L. Huang, B. Sun, H. Zhang, D. Hou, X.-q. Qiao, H. Ma and D.-S. Li, Chem. Eng. J., 2024, 479, 147719.
58. Y. He, C. Chen, Y. Liu, Y. Yang, C. Li, Z. Shi, Y. Han and S. Feng, Nano Lett., 2022, 22, 4970–4978.
59. H. Wang, J. Li, Y. Wan, A. Nazir, X. Song, P. Huo and H. Wang, Appl. Surf. Sci., 2023, 613, 155989.
60. C.-L. Tan, M.-Y. Qi, Z.-R. Tang and Y.-J. Xu, ACS Catal., 2023, 13, 8317–8329.
61. F. Wang, T. Hou, X. Zhao, W. Yao, R. Fang, K. Shen and Y. Li, Adv. Mater., 2021, 33, 2102690.
62. X. Li, Y. Sun, J. Xu, Y. Shao, J. Wu, X. Xu, Y. Pan, H. Ju, J. Zhu and Y. Xie, Nat. Energy, 2019, 4, 690–699.
63. S. Karmakar, S. Barman, F. A. Rahimi and T. K. Maji, Energy Environ. Sci., 2021, 14, 2429–2440.
64. X. Yi, S. Zhang, H. Shen, B. Li, L. Yang, W. Dai, R. Song, J. Zou and S. Luo, Appl. Catal., B, 2023, 338, 123003.
65. B. Su, M. Zheng, W. Lin, X. F. Lu, D. Luan, S. Wang and X. W. Lou, Adv. Energy Mater., 2023, 13, 2203290.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4sc07077a

---