Boosted charge separation in direct Z-scheme heterojunctions of CsPbBr₃/Ultrathin carbon nitride for improved photocatalytic CO₂ reduction

Abstract

Realizing CO₂ reduction with H₂O oxidation on g-C₃N₄-based photocatalysts is quite promising yet challenging. Herein, an efficient direct Z-scheme heterostructure of ultrathin CsPbBr₃/g-C₃N₄ nanosheets is fabricated via a simple electrostatic self-assembly process, resulting in an effective photocatalytic reduction of CO₂ to CH₄ (184.0 μmol g⁻¹) and CO (105.2 μmol g⁻¹) with oxidation of H₂O under AM 1.5G irradiation. Notably, in situ X-ray photoelectron spectroscopy (XPS) analysis demonstrated that a direct Z-scheme charge transfer mechanism at the CsPbBr₃ and g-C₃N₄ interface is enabled, thereby achieving efficient charge separation and high redox potentials to drive photocatalytic CO₂ reduction and H₂O oxidation synchronously. Isotopic labelling experiments revealed that the adsorbed water is activated to release H atoms and eventually assists the CO₂ reduction. Theoretical calculations further revealed that the p orbitals of Pb atoms hybridize with the 2π* orbitals of CO₂ molecules, strengthening the affinity of CO₂ and weakening the binding strength. Moreover, overlapping between Pb p orbitals and 5σ orbitals of CO molecules promote the protonation of CO* intermediates. This work provides an in-depth understanding and guidance for constructing high-efficiency catalysts for CO₂ reduction with H₂O oxidation.

---

1. Introduction

Excessive consumption of fossil fuels and continuous emission of carbon dioxide (CO₂) have caused serious environmental problems.^1–5 Inspired by plant photosynthesis, artificial photosynthesis provides promising strategies for achieving simultaneous CO₂ reduction and H₂O oxidation into high-valued hydrocarbon fuels under mild conduction, alleviating energy crisis and environmental issues such as severe global warming and associated climate change issues.^6–9 However, this remains a huge challenge owing to the high chemical inertness of CO₂ molecules, slow kinetics of H₂O oxidation and multi-electron transfer pathways with high reaction barriers.^10–13 Therefore, it is increasingly urgent for exploring higher efficiency catalysts for the photo-reduction of CO₂ and H₂O oxidation.

Graphitic carbon nitride (g-C₃N₄) as a two-dimensional organic semiconductor exhibited excellent chemical stability and photoelectric catalytic performance, which has been investigated for photocatalytic CO₂ reduction.^14–19 However, bulk g-C₃N₄ exhibits unsatisfactory efficiency owing to the fast recombination and poor migration of photogenerated carriers. In this regard, many efforts have been explored to improve the activity of g-C₃N₄-based photocatalysts, such as heteroatom doping, defect modifying and nanostructure designing.^20,21 Among these, exfoliating bulk g-C₃N₄ into ultrathin nanosheets has attracted much attention benefiting from unique two-dimensional structural features, such as large specific surface area, promoting electron excitation and short charge diffusion distance, and the resultant ultrathin g-C₃N₄ demonstrated a significant increase in photocatalytic activity.^22–26 However, developing a single-component layered g-C₃N₄ photocatalysts for CO₂ reduction coupled with H₂O oxidation has proven difficult because of the thermodynamic conflict between optical absorption and redox potentials. Recently, constructing layered g-C₃N₄-based 2D/2D heterostructures has been regarded as a quite promising strategy for enabling spatial separation of carriers with high redox potentials. Ye et al. fabricated 2D/2D van der Waals heterojunctions via assembling UiO-66 into g-C₃N₄ nanosheets, which showed intimate electronic coupling at the interface and improved photocatalytic CO₂ conversion efficiency.²⁷ Li et al. prepared 2D/2D Co₂P@BP/g-C₃N₄ heterojunctions via a simple self-assembly method, which exhibited dramatically enhanced photocatalytic activity and selectivity for CO₂ conversion.²⁸

Besides, the properly aligned band structures and effective active sites are crucial for exploring high-efficiency g-C₃N₄ composite catalysts in photocatalytic CO₂ reduction.^12,13,29 Lead halide perovskites (LHPs) have emerged as promising materials, owing to easy facile processing and excellent optoelectronic properties.^30–33 In addition, its sufficiently negative conduction band edge highlights the excellent CO₂ reduction ability.^34–36 In particular, CsPbBr₃ is a typical halide perovskite material, which has gained significant interest in solar CO₂ conversion, due to strong light harvesting ability, high carrier mobility and long carrier diffusion length.^37–39 Considering the distinctive photocatalytic properties of CsPbBr_3, the increased photocatalytic CO₂ reduction performance can be realized by integrating it with other semiconductors.^40–42 Kuang et al. fabricated a α-Fe₂O₃ nanorod/CsPbBr₃ Z-scheme heterojunction using amine functional graphene oxide interlayers, which can integrate the strong reduction ability of CsPbBr₃ and the strong oxidation ability of α-Fe₂O₃ to efficiently convert CO₂ and H₂O into CH₄ and O₂.⁴³ Yu et al. designed a TiO₂/CsPbBr₃ heterojunction via a simple electrostatic assembly, where electron-rich CsPbBr₃ nanocrystals can serve as active sites by donating electrons to activated CO₂ molecules for producing H₂ and CO.⁴⁴ Inspired by these advantages, the 2D/2D Z-type heterostructure established by CsPbBr₃ and ultrathin g-C₃N₄ will enhance the carrier separation and transfer efficiency as well as the CO₂ adsorption and activation capacity, thereby enhancing the photocatalytic CO₂ reduction performance.

In this work, we assembled the pre-fabricated ultrathin CsPbBr₃ nanosheets (CPBNs) and g-C₃N₄ nanosheets (CNNs) to form a 2D/2D facet-contact Z-scheme heterostructure via strong electrostatic interactions. In addition, close face-contact CsPbBr₃/C₃N₄ realized a heterojunction with a large-area heterointerface by finely controlling the surface charge, thereby ensuring efficient Z-scheme charge transfer and promoting charge separation. In particular, in situ X-ray photoelectron spectroscopy (XPS) and band structure analysis showed that a direct Z-scheme charge transfer pathway was established between CsPbBr₃ and g-C₃N₄, thus effectively promoting carrier separation and generating high redox potentials. Moreover, density functional theory (DFT) calculations demonstrated that the presence of an internal electric field at CPBN/CNN interfaces increases the efficiency of photoinduced charge carrier separation. It is exciting that by optimizing the amount of CPBNs, the resulting Z-scheme CPBN/CNN photocatalysts deliver more excellent photocatalytic activity in CO₂ reduction with H₂O oxidation (184.0 μmol g⁻¹ for CH₄, 105.2 μmol g⁻¹ for CO, and 398.3 μmol g⁻¹ for O₂) compared with CNNs. This study provides a new direction for the construction of g-C₃N₄-based photocatalysts using a halide perovskite as a semiconductor in the field of photocatalytic overall reaction for CO₂ and H₂O conversion.

2. Experimental section

2.1 Materials and reagents

Lead acetate (Pb(AC)₂), caesium bromide (CsBr), oleylamine (OA), oleic acid (OLA), melamine and phosphorous acid were purchased from Aladdin Reagents. Ethanol (C₂H₅OH, 99.8%), ethylene glycol (C₂H₆O₂, 99.8%) and isopropyl alcohol (C₃H₈O, 99%) were bought from Tianjin Fuyu Fine Chemical Industry Co. Ltd. All chemical reagents were commercially available and purchased from reagent companies.

2.2 Materials synthesis

2.2.1 Preparation of ultrathin g-C₃N₄ nanosheets (CNNs). The CNNs were fabricated via intercalation of hexagonal rod-like supramolecular precursors followed by pyrolysis process. First, the hexagonal rod-like supramolecular precursors were obtained according to the reported melamine–cyanuric acid assembly process using phosphorous. Then, the precursors were intercalated using ethanol and glycol. In the typical procedure, 0.6 g precursors were suspended in an ethanol–glycerol solution (25 vol% glycerol in the mixture) at 180 °C for 3 h. The obtained powders were centrifuged with ethanol followed by drying at 60 °C. The resulting solid was further heated in a muffle furnace at a rate of 2 °C min⁻¹ for 2 h to obtain CNNs.

2.2.2 Preparation of CsPbBr₃ nanosheets (CPBNs). CPBNs were prepared by reacting Pb(AC)₂ and CsBr in an isopropanol solution. Typically, 3.75 mg of Pb(AC)₂ was dissolved in HBr (0.25 mL). Isopropanol (5 mL), OA (0.5 mL) and OLA (0.25 mL) were separately mixed with the solution followed by stirring for a few minutes. Then, 1 mL of CsBr solution (5 mg in methanol) was quickly added to the above-mentioned solution with vigorous string. The resulting products were collected by centrifugation at 6000 rpm for 2 min and then washed 2 times with isopropanol and once with n-hexane to remove organic residues. Finally, the prepared samples were redispersed in an isopropyl alcohol mixture to store for further use.

2.2.3 Preparation of CPBN/CNN. The CPBN/CNN hybrid was prepared by assembling CPBNs and CNNs via a simple electrostatic interaction. For a typical synthesis, the CPBNs were suspended in an isopropanol solution at concentrations of 0.1, 0.2, 0.3, and 0.4 mg mL⁻¹. Then, 10 mg of CNNs were immersed in 10 mL of CPBN solution and kept under stirring for 2 hours to yield a homogeneous suspension. Finally, the product was obtained by centrifugation and washed with isopropanol n-hexane to remove organic residues. The weight percentage of CPBNs with respect with CNNs was varied in the range of 10–40 wt%, and the corresponding products were denoted as CPBN/CNN-1, CPBN/CNN-2, CPBN/CNN-3, and CPBN/CNN-4 respectively.

2.3 Characterization

The zeta potential of the sample were measured using a Zeta sizer Nano ZS90 (Malvern S4 Instruments) in isopropyl alcohol. The morphology and microstructure of the samples were investigated using a Zeiss Sigma-500 field emission scanning electron microscope (SEM) operating at an accelerating voltage of 5 kV, and a JEOL JEM-F200 transmission electron microscope (TEM) operating at an accelerating voltage of 200 kV. The crystal structures were obtained by X-ray powder diffraction (XRD, Rigaku D/max-IIIB, Cu Kα, λ = 1.5406 Å radiation). X-ray photoelectron spectroscopy (XPS) analysis was performed using a VG ESCALAB MK II equipped with a Mg Kα (1253.6 eV) achromatic X-ray source. The UV-vis diffuse reflectance spectra (UV-Vis DRS) of the samples were recorded using a Shimadzu UV-2550 UV-visible spectrophotometer with BaSO₄ as the reference. The steady-state photoluminescence (PL) was recorded using an Edinburgh FSL1000 fluorescence spectrophotometer at room temperature. The time-resolved photoluminescence (TRPL) spectra were recorded using an Edinburgh instruments F980.

2.4 Photocatalytic CO₂ reduction

2.4.1 Photocatalytic carbon dioxide reduction performance test. The photocatalytic carbon dioxide reduction test was carried out in a 140 mL sealed heat-resistant glass bottle filled with 10 mL of ethyl acetate and 30 μL of water. About 10 mg of photocatalyst was added to the bottle. Before testing, the system was degassed to remove O₂ and other gases, and refilled with carbon dioxide. A 300 W xenon lamp with an AM1.5G filter was equipped to simulate solar lighting, calibrated by an NREL-calibrated Si solar cell, and the light intensity was adjusted to 150 mW cm⁻². After 4 h of irradiation, analysis was performed by gas chromatography (GC9790+, R&F Analyzer Co., Ltd). Under the same conditions, isotope labeling experiments were performed with ¹³CO₂ molecules and analyzed by gas chromatography-mass spectrometry (GC-MS).

2.5 Photoelectrochemical characterization

All electrochemical and photoelectrochemical measurements were performed on an electrochemical workstation using a three-electrode system (Princeton Versa STAT). Typically, a piece of FTO glass with a prepared photocatalyst coating served as the working electrode, a platinum foil as the counter electrode, and a saturated Ag/AgCl electrode as the reference electrode. The 0.05 M tetrabutylammonium hexafluorophosphate (TBAPF₆)/dichloromethane solution was filled in the quartz cell as the electrolyte. The photocurrents were recorded using a Chi electrochemical workstation under irradiation of a 150 W Xe lamp at a potential of −0.4 V.

2.6 DFT sections

All theoretical calculations were performed using the Vienna Ab initio Simulation Package. Projected augment wave (PAW) with generalised gradient approximation (GGA) and Perdew–Burke–Ernzerhof (PBE) functional described the interactions between ions and electrons. The energy cutoff was set at 400 eV, and the K point mesh was set as 2 × 2 × 1. In this work, a g-C₃N₄ model for melon type and a CsPbBr₃ slab with the (110) surface was constructed. All atoms were fully relaxed until the total energy difference reached below 1 × 10⁻⁶ eV per atom and force was less than 0.05 eV Å⁻¹.

3. Results and discussion

3.1 Structural characterization

Fig. 1a illustrates the fabrication route of 2D/2D CPBN/CNN hybrids. First, the CPBN was prepared via reacting Pb(AC)₂ with CsBr in isopropyl using oleic acid and oleylamine, and the ultrathin CNN was prepared via a thermal-induce exfoliation strategy. From the transmission electron microscopic (TEM) and atomic force microscopic (AFM) observation, the obtained CPBNs have a lateral size of approximately 300 nm with a vertical thickness of about 3 nm, and CNNs represent the laminar structure with a thickness of approximately 2 nm (Fig. S1†). Undoubtedly, these ultrathin nanosheets present high surface areas, abundant photocatalytic active sites and greatly short transport distance of photogenerated charge carriers.⁴⁵

Fig. 1 (a) Schematic diagram of the 2D/2D CPBN/CNN heterojunction prepared via an electrostatic self-assembly process. (b) Zeta potentials of CPBNs and CNNs in isopropanol. (c) XRD patterns of the CNN, CPBN and CPBN/CNN. (d) TEM image of CPBN/CNN and (e) HRTEM image of the selected area in (d). (f) Element mapping images of single elements.

Furthermore, it is advantageous for assembling these ultrathin nanosheets via a mutual Coulomb electrostatic interaction. The zeta potentials of the as-prepared CNNs and CPBNs in the isopropanol solvent are shown in Fig. 1b. It is observed that the surface of CPBNs is negative with a zeta potential of −6.42 mV, while that of CNNs is positive with a zeta potential of +4.06 mV. The opposite surface charges would cause strong electrostatic attraction between the negatively charged CPBNs and positively charged CNNs for constructing the CPBN/CNN 2D/2D heterostructure with close interface contacts.^46,47 X-ray diffraction (XRD) patterns of CPBN/CNN (Fig. 1c) obviously exhibit typical peaks of both CsPbBr₃ and g-C₃N₄. The diffraction peak at 27.3° matches well with the (002) plane of g-C₃N₄. Meanwhile, the diffraction patterns located at 15.0°, 21.3°, 30.4°, 34.1°, 37.4° and 43.5° corresponded to the (100), (110), (200), (210), (211) and (220) planes of cubic phase CsPbBr₃ (JCPDS No. 73-2020), which increased with the CPBN content (Fig. S2†). By comparing with the Raman spectrum of CNNs, there are many new characteristic peaks, indicating that the intimate contact and electrostatic interaction formed between CPBNs and CNNs (Fig. S3†). The TEM image of the CPBN/CNN heterojunction (Fig. 1d) clearly demonstrates a tightly packed structure with small sheets closely arranged on the big sheets. The high-resolution TEM (HRTEM) image affirmed that these small nanosheets are CsPbBr₃ with a lattice spacing of 0.419 Å, corresponding to the (110) facet (Fig. 1e). The energy-dispersive X-ray spectroscopy (EDX) and corresponding elemental mappings analysis further identified that all the characteristic elements including Cs, Pb, Br, C and N can be clearly observed in the hybrid (Fig. 1f and S4†). C and N elements are uniformly distributed in the whole sheets, and Cs, Pb and Br elements are concentrated in small range, indicating that CsPbBr₃ nanosheets are dispersedly stacked on the surface of g-C₃N₄, which matches well with the TEM image.

Furthermore, the surface chemical state and heterointerface interaction of CPBN/CNN was investigated by X-ray photoelectron spectroscopy (XPS). The XPS full survey spectrum identifies the presence of C, N, Cs, Pb and Br in CPBN/CNN hybrids (Fig. S5†). As observed in Fig. 2a, the C 1s spectrum of CPBN/CNN is resolved into three characteristic peaks centered at 284.60, 285.99 and 287.94 eV, which are assigned to C–C, C–NH₂ and N–C=N, respectively.⁴⁸ Compared with pristine CNNs, these characteristic peaks negatively shift by 0.13 eV, and the same shift can be observed in the N 1s spectrum. The binding energies of N 1s appear at 398.35, 399.36 and 400.85 eV, which are assigned to the N atoms in C–N=C, N–(C)₃ and C–NH₂, while these peaks are of high value in CNNs (Fig. 2b). In contrast, the binding energies of the Cs 3d and Pb 4f spectra show the opposite trend. The binding energies of Cs 3d_5/2 and Cs 3d_3/2 appear at 724.20 and 738.15 eV, which are higher than those of CPBNs (Fig. 2c), while the Pb 4f_7/2 (138.10 eV) and Pb 4f_5/2 (142.96 eV) peaks are 0.20 eV higher than those of CPBNs (Fig. 2d). The results of XPS spectra verified the electron migration from CPBNs to CNNs upon hybridization, indicating the intimate contact and strong chemical interaction at interfaces.

Fig. 2 XPS of the CNN, CPBN and CPBN/CNN-3: C 1s (a), N 1s (b), Cs 3d (c), and Pb 4f (d).

3.2 Influencing factors of photocatalytic activity

The band structures of CPBNs and CNNs were determined by ultraviolet visible diffusion reflectance spectroscopy (DRS) and ultraviolet photoelectron spectroscopy (UPS). As shown in Fig. 3a, the absorption edge of CNNs and CPBNs are located at 473 and 546 nm, corresponding to the intrinsic bandgaps of 2.62 and 2.27 eV, respectively. The adsorption edge of CPBN/CNN hybrids is composed of CPBNs and CNNs except a little shift, resulting from the strong interfacial interaction between CPBNs and CNNs. The position of VB edge (E_VB) in CNNs and CPBNs were further measured via UPS. As shown in Fig. 3b and c, the upper onset and secondary onset energies in the UPS spectra for CNNs were estimated to be 16.05 and 1.25 eV, and those for CPBNs are 16.99 and 1.74 eV. Thus, the E_VB values of CNNs and CPBNs were estimated to be 1.96 and 1.53 V (vs. normal hydrogen electrode (NHE) potential).^49,50 Combined with the optical bandgap, the CB edge (E_CB) of CNNs and CPBNs were estimated to be −0.66 and −0.74 V (vs. NHE), respectively. Meanwhile, the work functions of CPBNs and CNNs were estimated to be 4.23 and 5.15 eV (vs. vacuum), indicating that the Fermi level of CPBNs is higher than that of CNNs, which observe the same results in Kelvin probe spectrum (Fig. S6†).

Fig. 3 (a) UV-vis DRS spectra of photocatalysts and the calculated band gaps of CNNs and CPBNs. (b and c) UPS valence spectrum of CNNs and CPBNs. In situ irradiation XPS spectra of N 1s (d), Cs 3d (e) and Pb 4f (f). (g) Energy band diagram before and after contact, showing the formation of energy band bending and BIEF, which facilitate the Z-scheme charge transfer and photocatalytic reaction. (h) Time-resolved transient PL spectrum.

When CPBNs contact CNNs, electrons will transfer from CPBNs to CNNs to equilibrate the Fermi level and generate an IEF at interfaces. Furthermore, the charge transfer between CPBNs and CNNs was explored by in situ irradiated XPS. For CPBN/CNN in the darkness, the N 1s peaks shift positively by 0.24 eV, while the Cs 3d and Pb 4f peaks shift negatively by 0.23 and 0.22 eV under light irradiation (Fig. 3d–f).

The binding energy shift demonstrated that photogenerated electrons in CNNs migrate to CPBNs under light irradiation. The above-mentioned analysis demonstrate a direct Z-scheme charge transfer mechanism between CPBNs and CNNs (Fig. 3g).^51,52 The CPBNs have a positive Fermi level and band position than CNNs, which induce the electrons in CPBNs to transfer to CNNs until their Fermi level attains equilibrium, causing the band bending and an IEF with the direction of CPBNs to CNNs at the interface. Upon irradiation, the VB electrons in CPBNs and CNNs are excited to the CB. The IEF and band bend drive the photogenerated electrons in the CNN CB to recombine with holes in the CPBN VB. Afterward, the electron-rich CPBNs can serve as reductive sites, donating electrons to active CO₂ to producing CO and CH₄, and the hole-rich CNNs act as active sites to oxidize H₂O molecules to producing O_2.

The efficient charge separation between CPBNs and CNNs were further investigated by photoluminescence (PL) and time-resolved PL (TRPL) measurements. The PL spectrum of CNNs represents the strongest emission peak center at 450 nm, which is attributed to radiative recombination within heptazine units (Fig. S7†). As compared with CNNs, CPBN/CNN exhibits a weak emission peak, implying that the presence of CPBNs efficiently inhibits the electron–hole recombination in CNNs, and the blue-shift can be associated with the strong interfacial coupling.^53,54 Furthermore, the TRPL decay spectroscopy is analysed for exploring the lifetime of carriers transfer dynamics. The average fluorescence lifetimes of CNNs and CPBNs were calculated to be 14.63 ns and 12.57 ns, by fitting the decay plot with three-exponential equations (Fig. 3h and Table S1†). After CPBNs contact CNNs, the average lifetime of CPBN/CNN significantly prolongs to 23.72 ns, suggesting that the interfacial charge transfer is efficiently improved.⁵⁵ Photoelectric-chemical tests were performed to investigate the charge carrier separation efficiency. The CPBN/CNN-3 exhibits significantly higher anodic photocurrent in comparison with CPBNs and CNNs, indicating that the transport and separation efficiency are largely improved in the CPBN/CNN-3 heterojunction (Fig. S8†). The diameter of the Nyquist plots reflects the charge-transfer resistance at the electrode/electrolyte interface. The CPBN/CNN-3 electrode possesses a smaller radius in comparison to CPBN and CNN electrodes, demonstrating the lowest resistance for charge carrier transfer (Fig. S9†). The above-mentioned experimental results fully certify that the advantages of combination of CPBNs and CNNs in promoting charge transfer and separation, which is beneficial for photocatalytic CO₂ reduction.

3.3 Photocatalytic performance

Subsequently, the photocatalytic CO₂ conversion performance of CPBN/CNN-x, CPBNs and CNNs was investigated. The catalytic reaction was carried out in a CO₂–saturated mixture solution of ethyl acetate and water under AM 1.5G irradiation. As shown in Fig. 4a–c, all samples show that the products consist of CH₄, CO and O₂, which increase linearly with the prolonged reaction time. CPBNs and CNNs exhibit lower production of CH₄ (17.6 and 15.3 μmol g⁻¹), CO (16.8 and 12.9 μmol g⁻¹) and O₂ (35.4 and 30.6 μmol g⁻¹). Notably, the photocatalytic activity is significantly enhanced with the increase in the loading amount of CPBNs, and CPBN/CNN-3 (the mass ratio of CPBNs to CNNs is 3/10) reaches the best CO₂ reduction performance with yields of 184.0 μmol g⁻¹ and 105.2 μmol g⁻¹ for CH₄ and CO respectively after 4 h of continuous illumination. O₂ is also produced in the reaction system with a yield of 398.3 μmol g⁻¹, which could be attributed to H₂O oxidation as the number of photoelectrons participating in the reduction reaction is nearly equal to that in the oxidation reaction (Fig. S10†). The electron consumption rate (R_electron) of CPBN/CNN-3 was calculated to be 420.6 μmol g⁻¹ and the corresponding electron consumption yield (Yield_electron) was calculated to be 1682.4 μmol g⁻¹, which is 11.3-fold of CNNs and 9.6-fold of CPBNs and obviously better than most known catalysts. Further increasing the CPBN amount is detrimental to the photocatalytic activity because overloading CPBNs will decrease the effective interface area and shield light adsorption. The CPBN/CNN provides CO₂ reduction activities under different H₂O contents, in which the appropriate H₂O content is favorable for CO₂ reduction (Fig. S11†). The cycling test indicates that no morphological changes and decomposition of CPBNs are observed in the TEM image and XRD spectrum, and the photocatalytic activity remains at the origin performance (Fig. 4d and S12†). As illustrated in Fig. 4e and f, it is observed that the yield of reduction products of CH₄ and CO almost increased linearly over 20 h, which achieves a high yield of 917.2 μmol g⁻¹ for CH₄, 524.8 μmol g⁻¹ for CO and 2032.6 μmol g⁻¹ for O₂ with electron consumption of 8387.2 μmol g⁻¹, implying the excellent stability of CPBN/CNN-3 during the photocatalytic CO₂ reduction process. The amount of oxygen evolution is always close to one-fourth of electron consumption for CO₂ reduction. Furthermore, the apparent quantum efficiency (AQE) of the prepared CPBN/CNN was calculated to be 0.26% at 420 nm (Fig. S13 and Table S5†).

Fig. 4 . CO₂ photoreduction performance. (a) CH₄, (b) CO and (c) O₂ yields of the CNN, CPBN and CPBN/CNN-3 by recording every 1 h upon illumination for 4 h. (d and e) Long-term photocatalytic tests of CPBN/CNN-3. (f) Amounts of electron consumption (e_amount) for the generation of CH₄ and CO (left vertical axis) and the yield of O₂ (right vertical axis) of CPBN/CNN-3. (g) Evolution of CH₄ and CO under different reaction conditions. (h) GC-MS spectrum of the gas-phase products driven for CPBN/CNN-3 in the photocatalytic reduction of ¹³CO₂. (i) Schematic diagram of the interfacial electron transfer of electrons and holes on g-C₃N₄ and CsPbBr₃ nanosheets.

In order to confirm the C and O source in the photocatalytic reaction, control photocatalytic experiments were carried out (Fig. 4g). CH₄ and CO are not detected in the absence of light or catalysts, indicating that products only originated from photocatalytic process. By replacing CO₂ with N₂, only a small amount of product was detected, which should originate from the partial oxidation of ethyl acetate or organic ligands in CPBNs. The 13CO₂ isotope trace experim-ent observed ¹³CH₄ (m/z = 17), ¹³CO (m/z = 29) and ¹⁶O₂ (m/z = 32) in the gas chromatography mass spectrometer spectrum (Fig. 4h). When H₂O is replaced with H₂¹⁸O, ¹⁸O₂ (m/z = 36) and ¹⁶O₂ (m/z = 32) are observed, which originate from H₂¹⁸O or H₂¹⁶O (Fig. S14†). Additional experiments by replacing H₂O with D₂O shows that H₂O is the proton source for CH₄ as CD₄ is not detected, which is attributed to the heavier atomic mass of D₂O that weakened its reactivity (Fig. S15†). These results indicated that the CH₄ and CO production is derived from CO₂ reduction along with H₂O oxidation into O₂, in which the photogenerated electrons in the CPBNs participate in the reduction of CO₂ molecules with the assistance of protons, while H₂O is oxidized to O₂ and protons by the holes in CNNs (Fig. 4i).

To deeply investigate the charge transfer mechanism and internal electric field, theoretical calculations based on density functional theory were performed. The work function of g-C₃N₄ (001) slab and CsPbBr₃ (110) slab was estimated from the electrostatic potential of the material surface, which is an important parameter to investigate the electron transfer of heterojunctions. As shown in Fig. 5a and b, the work function of CsPbBr₃ (110) slab was calculated as 2.91 eV similar to that of g-C₃N₄ (001) slab (5.68 eV). Therefore, when the CsPbBr₃ (110) slab contacts the g-C₃N₄ (001) slab, the free electrons in CsPbBr₃ with a high Fermi level will spontaneously transfer to g-C₃N₄ with a low Fermi level via the interface until achieving equilibrium. The charge difference based on the CsPbBr₃/g-C₃N₄ heterojunctions verifies that the diffusion of electrons induces the charge redistribution at the interface, in which the interface near the CsPbBr₃ side generates an electron depletion layer with a positive charge and the electron accumulation layer with a negative charge appears near the g-C₃N₄ side. As a result, the charge redistribution at the interface contributes to the IEF point from CsPbBr₃ to g-C₃N₄, providing a strong drive force for promoting the separation and migration of photogenerated charges, which is consistent with XPS and UPS results (Fig. 5c). Subsequently, the electronic location function was employed to investigate the interaction between CO₂ and the surface of photocatalysts. As depicted in Fig. 5d, the Pb and Br atoms in the CsPbBr₃ (110) surface exhibit a strong electrostatic attraction toward C and O atoms in CO₂, resulting in more variation in O=C=O bending than that in the g-C₃N₄ (100) surface, and further indicating that adsorbed CO₂ is activated more easily in the CsPbBr₃ (110) surface. The project density of states (PDOS) and crystal orbital Hamilton population (COHP) were calculated to evaluate the CO₂ internal band interaction in the difference surface model. After interaction with CO₂, CsPbBr₃ and g-C₃N₄ show the fast adsorption of CO₂, as the 1π and 7σ orbitals of CO₂ shift downwards the Fermi level. Notably, for the CsPbBr₃ (110) slab, the Pb p orbitals overlap the 2π* orbitals, indicating the favorable activation of CO₂ molecules.⁵⁶ Besides, a lower 2π* energy level of CO₂ in the CsPbBr₃ (110) surface demonstrates stronger CO₂ adsorption and activation in the COHP spectrum (Fig. 5f). For CO adsorption, the 1π and 5σ orbitals of CsPbBr₃ shift downwards the Fermi level as compared to g-C₃N₄. Meanwhile, the Pb p orbitals overlap the 5σ and 2π* orbitals, which is favorable for the interaction with CO* intermediates.⁵⁷ Combined with the COHP changes of the C–O bond, it can be verified that the CsPbBr₃ (110) surface is favorable for protonation of CO* intermediates to CH₄.

Fig. 5 Electrostatic potentials of (a) g-C₃N₄ and (b) CsPbBr₃. (c) Charge difference between distribution for the CsPbBr₃/C₃N₄ heterojunction interface; the blue and yellow iso-surfaces stand for electron depletion and accumulation respectively. (d) Electron local function of CO₂ adsorption on CsPbBr₃/g-C₃N₄, DOS and COHP analysis in terms of adsorption for g-C₃N₄ and CsPbBr₃ with CO₂ (e and f) and CO (g and h) interaction.

4. Conclusions

In summary, a direct Z-scheme 2D/2D heterojunction has been fabricated via assembling CsPbBr₃ nanosheets (CPBNs) onto ultrathin g-C₃N₄ nanosheets (CNNs), realizing the enhanced photocatalytic performance toward synchronous CO₂ reduction and H₂O oxidation. Benefiting from appropriate band alignment and close interfacial contact, an effective Z-scheme charge transfer pathway was established between CPBNs and CNNs, promoting charge separation and ensuring high redox ability, which was confirmed by in situ XPS, UPS and TRPL characterizations. The theoretical calculations further verified that the p–p hybridization between Pb atoms and CO₂ and CO intermediates enables the efficient activation of CO₂ and weakens the binding strength of C–O bonds. Therefore, the CPBN/CNN heterojunctions exhibit a remarkable photocatalytic performance for reducing CO₂ to CH₄ (184.0 μmol g⁻¹) and CO (105.2 μmol g⁻¹) with H₂O oxidation. We believe that this 2D/2D heterostructure may provide a new perspective for improving the photocatalytic performance of C₃N₄ and supply an important guidance for the design of photocatalytic overall reaction systems.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (52273264), Outstanding Youth Fund of Heilongjiang Province (JQ 2020B002). Guangxi Science and Technology Base and Talent Special Project NO. AD21075001.

Notes and references

1. X. X. Chang, T. Wang and J. L. Gong, Energy Environ. Sci., 2016, 9, 2177–2196.
2. L. Ran, Z. W. Li, B. Ran, J. Q. Cao, Y. Zhao, T. Shao, Y. R. Song, M. K. H. Leung, L. C. Sun and J. G. Hou, J. Am. Chem. Soc., 2022, 144, 17097–17109.
3. J. J. Wang, S. Lin, N. Tian, T. Y. Ma, Y. H. Zhang and H. W. Huang, Adv. Funct. Mater., 2021, 31, 2008008.
4. J. W. Wang, L. Z. Qiao, H. D. Nie, H. H. Huang, Y. Li, S. Yao, M. Liu, Z. M. Zhang, Z. H. Kang and T. B. Lu, Nat. Commun., 2021, 12, 813.
5. Y. Shi, J. Li, C. Mao, S. Liu, X. Wang, X. Liu, S. Zhao, X. Liu, Y. Huang and L. Zhang, Nat. Commun., 2021, 12, 5923.
6. C. j. Wang, Y. L. Zhao, H. Xu, Y. F. Li, Y. C. Wei, J. Liu and Z. Zhao, Appl. Catal., B, 2020, 263, 118314.
7. H. Ge, Y. Kuwahara, K. Kusu and H. Yamashita, J. Mater. Chem. A, 2021, 9, 13898–13907.
8. L. Romani, A. Speltini, C. N. Dibenedetto, A. Listorti, F. Ambrosio, E. Mosconi, A. Simbula, M. Saba, A. Profumo, P. Quadrelli, F. De Angelis and L. Malavasi, Adv. Funct. Mater., 2021, 31, 2104428.
9. 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.
10. J. Wu, Y. Huang, W. Ye and Y. Li, Adv. Sci., 2017, 4, 1700194.
11. J. Li, W. F. Pan, Q. Y. Liu, Z. Q. Chen, Z. J. Chen, X. Z. Feng and H. Chen, J. Am. Chem. Soc., 2021, 143, 6551–6559.
12. Y. N. Bo, C. Gao and Y. J. Xiong, Nanoscale, 2020, 12, 12196.
13. H. Huang, B. Pradhan, J. Hofkens, M. B. J. Roeffaers and J. A. Steele, ACS Energy Lett., 2020, 5, 1107–1123.
14. Q. Liu, H. Cheng, T. Chen, T. W. B. Lo, Z. Xiang and F. Wang, Energy Environ. Sci., 2022, 15, 225–233.
15. X. Jiao, K. Zheng, L. Liang, X. Li, Y. Sun and Y. Xie, Chem. Soc. Rev., 2020, 49, 6592–6604.
16. S. Guo, Z. Deng, M. Li, B. Jiang, C. Tian, Q. Pan and H. Fu, Angew. Chem., Int. Ed., 2016, 55, 1830–1834.
17. C. Zhao, Q. Li, Y. Xie, L. Zhang, X. Xiao, D. Wang, Y. Jiao, C. A. Hurd Price, B. Jiang and J. Liu, J. Mater. Chem. A, 2020, 8, 305–312.
18. B. Wu, L. Zhang, B. Jiang, Q. Li, C. Tian, Y. Xie, W. Li and H. Fu, Angew. Chem., Int. Ed., 2021, 60, 4815–4822.
19. X. Xiao, Z. Ruan, Q. Li, L. Zhang, H. Meng, Q. Zhang, H. Bao, B. Jiang, J. Zhou, C. Guo, X. Wang and H. Fu, Adv. Mater., 2022, 34, 2200612.
20. M. Zhu, S. Kim, M. Liang, M. Fujitsuka, J. Zhang, X. Wang and T. Majima, J. Am. Chem. Soc., 2017, 139, 13234–13242.
21. J. Liu, H. Q. Wang and M. Antonietti, Chem. Soc. Rev., 2016, 45, 2308–2326.
22. X. Zhang, X. Xie, H. Wang, J. Zhang, B. Pan and Y. Xie, J. Am. Chem. Soc., 2013, 135, 18–21.
23. Q. Han, B. Wang, J. Gao, Z. Cheng, Y. Zhao, Z. Zhang and L. Qu, ACS Nano, 2016, 10, 2745–2751.
24. X. Zhang, P. J. Ma, C. Wang, L. Y. Gan, X. J. Chen, P. Zhang, Y. Wang, H. Li, L. H. Wang, X. Y. Zhou and K. Zheng, Energy Environ. Sci., 2022, 15, 830–842.
25. Q. Li, L. M. Zhang, J. N. Liu, J. Zhou, Y. Q. Jiao, X. D. Xiao, C. Zhao, Y. Zhou, S. Ye and B. J. Jiang, Small, 2021, 17, 2006622.
26. X. Xiao, Y. Gao, L. Zhang, J. Zhang, Q. Zhang, Q. Li, H. Bao, J. Zhou, S. Miao, N. Chen, J. Wang, B. Jiang, C. Tian and H. Fu, Adv. Mater., 2020, 32, 2003082.
27. L. Shi, T. Wang, H. Zhang, K. Chang and J. Ye, Adv. Funct. Mater., 2015, 25, 5360–5367.
28. J. D. Hu, T. Y. Yang, X. G. Yang, J. F. Qu, Y. H. Cai and C. M. Li, Small, 2022, 18, 2105376.
29. Z. Z. Zhang, M. Y. Wang, Z. X. Chi, W. J. Li, H. Yu, N. Yang and H. B. Yu, Appl. Catal., B, 2022, 313, 121426.
30. S. Kumar, I. Hassan and M. Regue, et al. , J. Mater. Chem. A, 2021, 9, 12179–12187.
31. J. Liang, D. Chen, X. Yao, K. Zhang, F. Qu, L. Qin, Y. Huang and J. Li, Small, 2020, 16, 1903398–1903417.
32. V. Malgras, S. Tominaka, J. W. Ryan, J. Henzie, T. Takei, K. Ohara and Y. Yamauchi, J. Am. Chem. Soc., 2016, 138, 13874–13881.
33. J. Zhang, G. Hodes, Z. Jin and S. Liu, Angew. Chem., Int. Ed., 2019, 58, 15596–15618.
34. L. Zhou, Y. F. Xu and B. X. Chen, et al. , Small, 2018, 14, 1703762.
35. J. Sheng, Y. He, J. Li, C. Yuan, H. Huang, S. Wang, Y. Sun, Z. Wang and F. Dong, ACS Nano, 2020, 14, 13103–13114.
36. X. Wang, Y. Huang, J. Liao, Y. Jiang, L. Zhou, X. Zhang, H. Chen and D. Kuang, J. Am. Chem. Soc., 2019, 141, 13434–13441.
37. J. Hou, S. Cao, Y. Wu, Z. Gao, F. Liang, Y. Sun, Z. Lin and L. Sun, Chemistry, 2017, 23, 9481–9485.
38. H. J. Li, W. G. Tu, Y. Zhou and Z. G. Zou, Adv. Sci., 2016, 3, 1500389.
39. L. Wang, X. Zheng, L. Chen, Y. Xiong and H. Xu, Angew. Chem., Int. Ed., 2018, 57, 3454–3458.
40. J. Wang, S. Lin, N. Tian, T. Ma, Y. Zhang and H. Huang, Adv. Funct. Mater., 2021, 31, 2008008.
41. N. Li, X. Chen, J. Wang, X. Liang, L. Ma, X. Jing, D. L. Chen and Z. Li, ACS Nano, 2022, 16, 3332–3340.
42. Q. M. Sun, J. J. Xu, F. F. Tao, W. Ye, C. Zhou, J. H. He and J. M. Lu, Angew. Chem., Int. Ed., 2022, 61, e202200872.
43. Y. Jiang, H. Chen, J. Li, J. Liao, H. Zhang, X. Wang and D. Kuang, Adv. Funct. Mater., 2020, 30, 2004293.
44. F. Xu, K. Meng, B. Cheng, S. Wang, J. Xu and J. Yu, Nat. Commun., 2020, 11, 4613.
45. Y. Xiao, G. Tian, W. Li, Y. Xie, B. Jiang, C. Tian, D. Zhao and H. Fu, J. Am. Chem. Soc., 2019, 141, 2508–2515.
46. C. W. Ahn, P. H. Borse, J. H. Kim, J. Y. Kim, J. S. Jang, C.-R. Cho, J.-H. Yoon, B.-s. Lee, J.-S. Bae, H. G. Kim and J. S. Lee, Appl. Catal., B, 2018, 224, 804–809.
47. D. M. Zhao, Y. Q. Wang, C. L. Dong, Y. C. Huang, J. Chen, F. Xue, S. H. Shen and L. J. Guo, Nat. Energy, 2021, 6, 388–397.
48. M. Zhu, S. Kim, M. Liang, M. Fujitsuka, J. Zhang, X. Wang and T. Majima, J. Am. Chem. Soc., 2017, 139, 13234–13242.
49. J. D. Hu, D. Y. Chen, Z. Mo, N. J. Li, Q. F. Xu, H. Li, J. H. He, H. Xu and J. M. Lu, Angew. Chem., Int. Ed., 2019, 58, 2073–2077.
50. L. Yuan, K.-Q. Lu, F. Zhang, X. Fu and Y.-J. Xu, Appl. Catal., B, 2018, 237, 424–431.
51. Q. Xu, L. Zhang, B. Cheng, J. Fan and J. Yu, Chem, 2020, 6, 1543–1559.
52. Q. Li, W. J. Feng, Y. Q. Liu, D. Z. Chen, Z. B. Wu and H. Q. Wang, J. Mater. Chem. A, 2022, 10, 15752.
53. N. Li, X. Chen, J. Wang, X. Liang, L. Ma, X. Jing, D.-L. Chen and Z. Li, ACS Nano, 2022, 16, 3332–3340.
54. F. Chen, Z. Y. Ma, L. Q. Ye, T. Y. Ma, T. R. Zhang, Y. H. Zhang and H. W. Huang, Adv. Mater., 2020, 1908350.
55. Z. Z. Zhang, Y. H. Zhao, J. T. Shen, Z. W. Pan, Y. F. Guo, P. K. Wong and H. B. Yu, Appl. Catal., B, 2020, 269, 118774.
56. X. Y. Yue, L. Cheng, F. Li, J. J. Fan and Q. J. Xiang, Angew. Chem., Int. Ed., 2022, e202208414.
57. G. Qian, W. Y. Lyu, X. Zhan, J. Y. Zhou, R. Q. Fang, F. L. Wang and Y. W. Li, Angew. Chem., Int. Ed., 2022, e202210576.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2ta07965h

---