English

• Optimizing carbon fuel consumption and mitigating CO₂ greenhouse gas emissions are imperative for fostering the sustainable evolution of human civilization, pivotal in averting the hastening of energy shortages and the exacerbation of global warming [1-5]. The photocatalytic CO₂ reduction reaction (CRR) has been considered an environmentally friendly method to "kill two birds with one stone" in the conversion of CO₂ for carbon fuel technologies [6-9]. Among the vast array of photocatalysts, bismuth tungstate (Bi₂WO₆) stands out as a promising member of the Aurivillius family, characterized by its layered perovskite-like structure and appealing band diagram. The energy level of the W 5d orbital presents potential advantages for CO₂ binding on the surface of catalysts, while the interaction between the Bi 6s and O 2p orbitals at the top of the valence band leads to a relatively narrow band gap, facilitating light absorption. Typically, photocatalytic efficiency is mainly dependent on light absorption of catalyst, charge transfer, and surface reaction. Tremendous investigations have been devoted to enhancing the charge transfer and optimizing the surface-active sites of Bi₂WO₆, such as heterojunction construction of Bi₂WO₆ and BiOX (X=Cl, Br, I) [10-13], downsizing to ultrathin layer or single unit cell [14-18], and doping single atom sites or defects [19-25]. The separation and transfer of electron-hole pairs are recognized to play major roles in all these investigations [26-27]. For instance, Li et al. created oxygen vacancies in plasma-treated Bi₂WO₆ nanosheets to boost charge transfer, thus enhancing photocatalytic CO₂ reduction [28]. Recently, Di and Chen fabricated distorted BiOCl induced by a bilayer thickness of tube geometric structure, and accelerated the photocarrier migration and photocatalytic performance [29]. Moreover, Li et al. reported an electric field would generate via the polarization of the non-uniform charge distribution between different constituent layers of Bi₂WO₆-BiOCl, and the interlayer carbon impurities can boost separation of electrons and holes immediately after their generation [30]. Nevertheless, single atoms, defects, and interfaces might induce atomic-scale lattice mismatch, which can have a crucial impact on charge transfer dynamics. Especially in two-phase heterojunction, the barrier for charge transfer is inevitable due to the inherent interface. Therefore, the establishment of lattice-matched interfaces, known as coherent interfaces, a term in the field of materials science, is crucial for enhancing interfacial charge transfer in composite materials.

  The charge migration in Bi₂WO₆ semiconductor is another issue of bismuth photocatalysts. The distorted or defective Bi₂WO₆ can induce a higher charge carrier concentration within catalysts. For instance, Chung found higher concentration of tungsten incorporation into Bi₂WO₆ can self-doping and lead to a higher bulk carrier density and improved conductivity [31]. Furthermore, the band structure is the thermodynamics prerequisite for the charge transfer. To improve the activity of Bi₂WO₆, Zhao et al. made a hybrid Ti₃C₂-Bi₂WO₆ to increase the efficiency of photocatalytic CO₂ reduction, assuring that the interfacial charge transfer ability is dominant [32]. Thus, inspired by the semi-metal phenomena in Bi₂WO₆ and the charge transfer optimization by hybridization, distorted Bi₂WO₆ with reduced metal sites can be developed [33]. Although many combinations of Bi₂WO₆, Bi oxide, BiOX have been widely studied, especially in photocatalytic CO₂ reduction, most research focus on the band alignment, regardless of interface contact modification, which is crucial for the inter-phase charge carries transport. in-situ grown heterostructure has promising aspects providing the ideal compatible components for the layered Bi-based structures due to the close packing of atoms at the interface.

  In this study, we have proposed a facile method to fabricate a coherent interface aligned Bi₂WO₆-BiOCl heterostructure, facilitated by the presence of external WO₄^2‒ during hydrothermal synthesis. The resultant in-situ grown BiOCl crystals facilitate the reduction of metal sites of W and Bi, thereby significantly enhancing charge transfer efficiency and consequently improving the photocatalytic CO₂ reduction efficiency. Our approach based on atomic lattice matched Bi₂WO₆-BiOCl, facilitating dynamic charge carrier movement through meticulous interface engineering. This leads to a remarkable photocatalytic efficiency for CO₂ reduction to CO, reaching 68.03 µmol g^‒1 h^‒1. This enhancement can be attributed to optimized surface bonding via strain effects and a rational band alignment that facilitates carrier movement onto the surface.

  As depicted in Fig. 1, the structure models of single-phase catalyst, ball-milling sample and Bi₂WO₆-BiOCl are presented. Single phases in Fig. 1a consist of Bi-O layer, Cl atoms as well as WO₄^2‒ units featuring layered structures. Combined compositions can align surfaces and tune the bandgap to surpass the limit of single catalyst. In Fig. 1b, with the help of mechanically mix process, perovskite-like Bi₂WO₆ can contact Bi-O layer with BiOCl with physically-built interfaces. The exogenous assembly form rough contact within layers and pristine interfaces were reserved in bulk materials. Beyond hetero-interface between existing structures, the heterostructure of Bi₂WO₆-BiOCl was synthesized via homology-atom doping strategy in Fig. 1c. Based on the nanosheet reported before [33], we introduce 10% tungstate in acid solution as exogenous part to intercalate in the Bi-O layer. OTAC serve as trapping agent to slow the z-axis growth and chlorine will redistribute owing to the overdosed WO₄^2‒. Through content growth in water for 21 h in 140 ℃, coherent interface was built in halide and WO₄^2‒ layers with co-sharing Bi-O layer acting as transition area. Assemble from solution can build precise alignment to form tandem junctions. According to the previous method [31,33], the interlayer of WO₄^2‒ can be routinely and flexibly coordinated with bismuth-oxygen bond, when strain or exceeding elements was induced in Bi₂WO₆. Cl atoms was dispersed within octyltrimethyl ammonium chloride (OTAC), while an excess of W elements could trigger the crystallization of Cl component, originating from amorphous state with Bi-O bonding in the crystal growth. Exquisite phase contact between the BiOCl and Bi₂WO₆ ensure excellent atomic transition across the interface to promote the charge dynamics and band gaps. On the basis of coherent structure, superb CO₂ reduction reaction (CRR) performance could be realized.

  Figure 1

  

  Figure 1.  Schematic illustration of catalysts with interlayer configuration. Structure models of (a) single phase Bi₂WO₆ and BiOCl. (b) Ball-milling sample via mechanical forces. (c) Tandem Bi₂WO₆-BiOCl with coherent interfaces.

  DownLoad: Full-Size Img PowerPoint

  The components of the tandem Bi₂WO₆-BiOCl, were validated by X-ray diffraction (XRD) pattern and FTIR spectrum as shown in Fig. S1 (Supporting information). In addition, Fig. S1a demonstrates the XRD patterns of four samples. The characteristic peaks can be identified as BiOCl (JCPDS No. 006–0249) and Bi₂WO₆ (JCPDS No. 01–073–2020), while Ball-milling sample shows minor peak owing to the poor crystallinity from BiOCl component. Fourier transform infrared spectroscopy (FTIR) of the tandem Bi₂WO₆-BiOCl catalyst confirm bonds of W-O, Bi-O and Bi-Cl (Fig. S1b), Bi-Cl bond at 1155 and 1458 cm^‒1 confirm the incorporation of crystal BiOCl.

  Brunauer-Emmett-Teller (BET) analysis shows that specific surface area of Bi₂WO₆-BiOCl heterostructure was 31.82 m²/g (Fig. S3a in Supporting information), larger than that of pristine Bi₂WO₆ with 23.42 m²/g (Fig. S3c in Supporting information), the ball-milling sample possesses a surface area of 33.58 m²/g as depicted in Fig. S3b (Supporting information). The morphology of Bi₂WO₆-BiOCl was further investigated using aberration-corrected transmission electron microscopy (AC-TEM), as presented in Fig. 2a. The nanosheet shows an uneven and sphere-like surface. Energy dispersive X-ray spectroscopy (EDS) results from Fig. 2 and Fig. S7 (Supporting information) manifest the element dispersion. Bi₂WO₆-BiOCl showed bright region on the surface, where the location of chlorine and oxygen element follow the morphology. Tungsten was dispersing below while Bismuth was dominant on the surface, fitting the interlayer configuration of WO₄^2‒ layer in the Bi₂WO₆ substrate. The nanosheet shows an uneven and sphere-like surface. The element dispersion spectra (EDS) of W, Cl, O and Bi demonstrate a mixed distribution of W and Cl within the composite. In Fig. S4 (Supporting information), detailed aggregation of Cl element can be observed in Bi₂WO₆-BiOCl before and after the reduction reaction process, arising from the BiOCl formation on the Bi₂WO₆ surface, while Bi₂WO₆ itself as flat nanosheet. Fig. 2b was obtained via high-angle annular dark field (HAADF) in scanning transmission electron microscopy (STEM). The lattice spacing around 2.80 Å is assigned to the (110) plane of BiOCl as well as (200) planes of Bi₂WO₆. Corresponding diffraction patterns further elucidate two phases as well for 〈110〉 of BiOCl and 〈101〉 of Bi₂WO₆. The fast Fourier transform (FFT) filtered image can visualize the phase transitions as false colored image (Fig. 2c). Bi₂WO₆ is depicted as blue area, while BiOCl occupies the right-up area with orange background. The red rectangle area confirms the consistent transition between Bi₂WO₆ and BiOCl. The boundary is continuous to reach almost coherent condition. The lattice distortions occur within the contacting area, and the deviation is less than both lattice constants. To unravel the interfacial structure, atomic arrangement and lattice spacing was analyzed. Fig. 2d evidently show the reconstructed boundary area with atomic-scale lattice-match. Lattice spacing was surveyed with four parallel line intensity profiles across the interface. Lines (2) and (3) locates within the boundary, line (1) fits the BiOCl phase and line (4) is close to Bi₂WO₆. According to the location of two bold line position, the periodic atom arrangement shows little tensile in line (2) and moderate strain in line (3) with the same length, in comparison with line (1). All the mismatched range is within 100 pm. On the basis of lattice arrangements, structure of Bi₂WO₆-BiOCl can be deduced. The formation of the coherent interface was surveyed, for research on the influence of the interface modulation constructions that exert on the lattice strain engineering. Atomic lattice adjustment in the heterostructure features the strain effects [15], reflecting on the HAADF and geometric phase analysis (GPA)-images. Based on the atomic HAADF image, Ball-milling sample was chosen to contrast with the in-situ grown procedures. In Fig. 2e, smooth atomic contact was labeled as typical balls, while the interface in Fig. 2g was distorted extensively owing to numerous line defects as dislocations. Dislocations in ball-milling sample build disconnected borders between BiOCl and Bi₂WO₆, displaying as obscure blue pits in the middle of Fig. 2g. As shown in Fig. S11b (Supporting information), dislocations were dominant and distorted the crystal lattice. In terms of GPA strain maps of Fig. 2f, the coherent interface exhibits intensive color contrast, implying the BiOCl undergoes tensile strain as red zone. Blue contrast in the strain field appeared around the border of BiOCl crystal, which means that Bi₂WO₆ substrate around the -interface suffers compressive interface suffers compressive strain to maintain the coherent structure, although minor lattice distortion appeared. The stress distribution in Fig. 2h shows relatively homogeneous stresses in the ball milling sample, indicating moderate phase connection with shallow color and uniform stress distribution. The strain values that the BiOCl exhibited are more average and smaller. In contrast to Fig. 2f, solid interactions were achieved via the tandem structure with the segregation of BiOCl onto surface. According to results above, we infer that Bi₂WO₆-BiOCl was atomic-precise formed through the intense strain effects, while the ball-milling sample was filled with dislocations and loose bonded.

  Figure 2

  

  Figure 2.  STEM characterization of Bi₂WO₆-BiOCl. (a) HAADF-STEM image and elemental dispersive scanning of composite with Bi, O, W, Cl. (b) Atomic resolution HAADF-STEM image for (ⅰ) BiOCl and (ⅱ) Bi₂WO₆ phase, the (ⅲ) area was selected to manifest the interface structure. Diffraction patterns of (ⅰ) BiOCl and (ⅲ) Bi₂WO₆ were shown right of the panel. (c) FFT-filtered atomic resolution image using two sets of diffraction spots in red area. The lattice distortions can be visualized by comparing the array disorder. (d) FFT-filtered atomic image of left-up area in Fig. 2b. Four parallel lines were selected in BiOCl (1), interface (2), interface (3) and Bi₂WO₆ (4) to check atomic deviation profile around the interface. (e, f) FFT-filtered atomic resolution image visualization of Fig. 2c and corresponding GPA image of Bi₂WO₆-BiOCl. (g, h) False color image around the interface and GPA image for the ball-milling sample.

  DownLoad: Full-Size Img PowerPoint

  To uncover the formation of coherent-matched heterostructure induced by WO₄^2‒, electronic states of elements and coordination environment were surveyed by X-ray photoelectron spectroscopy (XPS) and X-ray absorption spectroscopy (XAS). Fig. 3a distinguish the electron states of W_4f. Bi₂WO₆ consisting of W⁶⁺ was set as the reference. More reduced W was detected in the ball-milling sample and Bi₂WO₆-BiOCl as lower binding energy, indicating the W⁵⁺ states arising from the electron injection. Fig. S12a (Supporting information) displayed that reduced species of W⁵⁺ appeared in the heterostructure with a ratio of 57.1%, while ball-milling sample occupied less percentage as 42.8%, elucidating better incorporation of tungsten with a majority of reduced tungsten elements. As a result, the coherent interface would be accompanied by the tuning of electronic structure with coordination and valence state changes of tungsten element [18]. The valence states of bismuth and oxygen are depicted in Fig. 3b. In Bi₂WO₆-BiOCl, the Bi element shifted to lower binding energy. Ball-milling samples also shifted with relatively higher states. Single phases exhibited more oxidized states of Bismuth. Fig. 3c proves the oxide exists as lower binding in Bi₂WO₆-BiOCl and ball-milling sample. Single BiOCl and Bi₂WO₆ were more oxidized. Cl elements stay relatively stable with minor changes.

  Figure 3

  

  Figure 3.  Electronic states of elements and local environmental coordination of W. (a) The XPS peak of W 4f for Bi₂WO₆-BiOCl. (b, c) was XPS of Bi 4f and O 1s. (d) The W L_Ⅲ-edge XANES spectra. (e) The corresponding k³-weighted FT spectra of Bi₂WO₆-BiOCl, Bi₂WO₆ was used as reference sample.

  DownLoad: Full-Size Img PowerPoint

  Since the structure of Bi₂WO₆-BiOCl was induced by WO₄^2‒, whether external tungsten altered the coordination environment was surveyed. As displayed in W LIII-edge X-ray absorption near-edge structure (XANES) spectra (Fig. 3d), peaks of Bi₂WO₆ and Bi₂WO₆-BiOCl located between the W foil and WO₃. The absorption threshold of Bi₂WO₆-BiOCl shifted to lower energy than Bi₂WO₆, as well as lower white line intensity, indicating an increase of reduced state of tungsten as W⁵⁺. The XPS results presented before fits well with XANES result. The Fourier-transformed extended X-ray absorption fine structure (FT-EXAFS) spectra (Fig. 3e) displays that the main peak of W-O in Bi₂WO₆-BiOCl is shorter than that of pristine Bi₂WO₆ by 0.06 Å, implying the dense coordination of tungsten and oxygen. The fitted curves confirmed the octahedral model of WO₄^2‒. Detailed results are summarized in Table S1 (Supporting information). Compared with Bi₂WO₆, the Bi₂WO₆-BiOCl sample has one more coordination number of planar W-O bond as 4.8, and simultaneously has one less coordination number of vertical W-O bonding. The distorted configuration may result from z-axis coordination with less oxygen atoms while the planar coordination number increases. Shorter bond length justifies spatially change within the strained Bi₂WO₆, whose confined structure benefits the crystallization of Cl into BiOCl [34]. The compressed geometry of WO₄^2‒ and shorter W-O bonding was consistent with HAADF-STEM parameters, aside from chemical states analysis, the coordination of tungsten was examined by XAS. Since the distorted W-O and Bi-O both have lower binding energy than that in pristine Bi₂WO₆, the external WO₄^2‒ is likely to trigger the reconstruction of Bi₂WO₆ nanosheet with distortion interlayer during the synthesis. Meanwhile, in-situ grown BiOCl crystal could compete Bi-O bonding with neighbored WO₄^2‒ unit to occupy the z-axis room to form heterostructure, consequently coherent interface stabilizes the Bi-O layer with more reduced metal-oxygen bonding.

  To further investigate photocatalytic CO₂ reduction performance of Bi₂WO₆-BiOCl, a gas-solid interfacial CO₂ photo-reduction was performed to evaluate the photocatalytic efficiency and selectivity. No external sacrifice agent was induced in the reaction process, and all the tests were conducted with AM 1.5 light. Fig. 4a shows the Bi₂WO₆-BiOCl has a CO yield of 68.03 µmol g^‒1 h^‒1, which is 5.62 times and 5.76 times larger than that of pristine Bi₂WO₆ and BiOCl, respectively. The efficiency reaches superior performance comparing with the literature reported bismuth tungstate structures as list in Table S3 (Supporting information). Moreover, the Bi₂WO₆-BiOCl heterostructure with atomic-scale lattice-match exhibits a greater performance than that of ball-milling sample with numerous dislocations, suggesting the lattice-matched interface contributes enhanced activity.

  Figure 4

  

  Figure 4.  Photocatalytic CO₂ reduction (a) efficiency, (b) selectivity and stability of Bi₂WO₆-BiOCl composite, the reference samples involve ball-milling sample, pristine Bi₂WO₆ and pristine BiOCl. (c) The performance comparison of CO yield with pure Bi-based materials.

  DownLoad: Full-Size Img PowerPoint

  Performance of Ball-milling sample were inferior to single phase, which may arise from loosen contact around interface. Furthermore, Fig. 4b shows the selectivity of the photocatalytic CO₂ reduction reactions, the products were mainly CO with a minor amount of CH₄, the electron selectivity of Bi₂WO₆-BiOCl is almost 98%. The reactivity of Bi₂WO₆-BiOCl for CRR was superb and highly selective to CO Additionally, the stability test (Fig. 4b) illustrates that the Bi₂WO₆-BiOCl is relatively stable [36], while the catalytic performance can maintain 90% above. In Fig. 4c, compared with current works of pure Bi-based materials, the reactivity ranks among the top catalysts.

  To disclose the mechanism for the enhanced efficiency of Bi₂WO₆-BiOCl, the photoelectric properties and kinetic pathways of photo carriers were investigated. We surveyed photocurrent tests and electrochemical impedance spectroscopy (EIS) to deduce the charge transfer resistance. Fig. 5a showed a significant higher photocurrent by almost 4–5 times larger than those of pristine Bi₂WO₆ and BiOCl. The ball-milling sample was insufficient for carrier transportation as almost zero current density. As exhibited in Fig. S11b, dislocations serving as trapping sites can accumulate charge carriers to combine before they contribute to the photocurrent, leading to the drastic decrease in the overall photocurrent efficiency. In ball-milling sample, mismatched interface inhibited sufficiently utilization of electrons, resulting poor photocurrent performance. In Fig. 5b, the Nyquist plots of Bi₂WO₆-BiOCl shows lower arc-resistance compared with those of BiOCl and Bi₂WO₆, indicating the charge transfer can be improved between the surfaces. Bi₂WO₆ was less conductive than BiOCl with bigger arc-resistance.

  Figure 5

  

  Figure 5.  Photocarrier separation and migration. (a) Photo-current performance and (b) Nyquist plots. (c) PL spectra. (d) Transient time-resolved PL decay spectra. (e) UV–vis spectra of four samples. (f) Band diagram of single phase and Bi₂WO₆-BiOCl.

  DownLoad: Full-Size Img PowerPoint

  The reason why ball-milling sample exhibited lowest arc resistance may result from the poor crystallinity owing to the dominant dosage of OTAC surfactant, for the amorphous contributes to higher conductivity in EIS and broad peak in XRD [25]. Wholly covered surface of BiOCl in the ball-milling sample determine the performance. Bi₂WO₆-BiOCl originates from substrate of Bi₂WO₆, consistent Bi₂WO₆ determine the inferior electrochemical impedance.

  Photoluminescence (PL) was conducted to reveal the separation of electron-hole pairs. As shown in Fig. 5c. Peaks around 464 nm relate to isolated vacancy bi defects, while the 393 and 400 nm peaks of Bi₂WO₆-BiOCl and ball-milling sample were red shift than that of BiOCl at 375 nm, indicating electronic coupling between nanocrystals when they are packed into a composite. All the intensities were elevated in contrast with Bi₂WO₆ sample, arising from the BiOCl crystal upon the surface [36,38]. Furthermore, a kinetic investigation of transient photoelectron's behavior on Bi₂WO₆-BiOCl was also accomplished. Analysis of the curves with re-convolution fitting manifests that the decays of each sample follow a bi-exponential model with τ₁ and τ₂ (Fig. 5d and Table S4 in Supporting information). In pristine Bi₂WO₆, the photo generated electrons from WO₄^2- layer shall move to the surface of the nanosheet to participate in the reaction; such a long migration path may increase the probability of electron-hole recombination and lower the photocatalytic activity [33]. Interestingly, Bi₂WO₆-BiOCl and ball-milling samples demonstrates less τ₁ and τ₂ than those of single phases. Reduced τ values indicate that new electron channels were built, resulting in no obvious recombination suppression of electron-hole separation. τ₂ was greatly reduced in Bi₂WO₆-BiOCl than other samples, which confirms that the interface pronounced to boost the separation [37]. In matched-lattice epitaxy in Bi₂WO₆-BiOCl, and fine electron channel was in-built across the coherent interface. As for ball-milling sample, photocurrent transportation was severely hampered, although better EIS and PL performance [38]. In ball-milling sample, numerous dislocations around interface trap the carriers, while Bi₂WO₆-BiOCl only build smooth interface. We presume that interface and defects both inhibit the recombination of electron and holes. In summary, the mechanism of charge transfer enhancement in Bi₂WO₆-BiOCl can be ascribed to the reduced tungsten with more W⁵⁺ states and reduced Bi atoms, which is also consistent with previous researches that this conductivity improves with the presence of the semi-metal of W⁵⁺ sites and distorted W-O unit cell [31,33,39]. Fig. 5e shows UV–vis spectra of the samples, in terms of the band gap in Fig. S14 (Supporting information), the 2.91 eV of composites is between 2.70 eV and 3.17 eV, which corresponds to BiOCl and Bi₂WO₆ [40-42]. The energy level results are shown in Fig. 5f and Table S5 (Supporting information). In the heterostructure, the induced BiOCl phase alters the Bi₂WO₆ nanosheet and the electron flows to the BiOCl phase. N type BiOCl and Bi₂WO₆ will bend upward in both CB and VB band to transport charge carriers. Energy level in the heterostructure is examined to confirm the thermodynamics.

  To verify the dynamic reaction process and the reaction route of CO₂ to CO, in-situ FTIR spectra were conducted to elucidate the intermediates. The reaction parameters were recorded with the interval of 6 min. As depicted in Figs. 6a and b, peaks concerning with the adsorbed species of monodentate carbonate (m-CO₃^2‒) located at 1560 cm^‒1and 1398 cm^‒1, while bicarbonate (HCO₃^‒) peaks located at 1472 cm^‒1 and 1523 cm^‒1. Peaks at 1697 cm^‒1 and 1620 cm^‒1 resulted from the vibration of bidentate carbonates (b-CO₃^2‒) [43]. All peaks associated with CO₃^2‒ species were enhanced in intensity after light. Particularly, during the reaction period, the concentration of COOH* (1550 cm^‒1) are significantly increased, indicating which were the key intermediates during the process of CO₂ to CO [44]. Besides, the peaks at 1360 cm^‒1 were assigned to the specie formate (HCO^2‒) [45,46]. Therefore, based on the data from in situ FT-IR spectra, the possible reaction path for the catalytic process was proposed as follows (Eqs. 1-4).

  $ \mathrm{CO}_2(\mathrm{~g})+{ }^* \rightarrow \mathrm{CO}_2{ }^* $

  (1)

  $ \mathrm{CO}_2+\mathrm{H}^{+}+\mathrm{e}^{-} \rightarrow \mathrm{COOH}^* $

  (2)

  $ \mathrm{COOH}^*+\mathrm{H}^{+}+\mathrm{e}^{-} \rightarrow \mathrm{CO}^*+\mathrm{H}_2 \mathrm{O} $

  (3)

  $ \mathrm{CO}^* \rightarrow \mathrm{CO}(\mathrm{g}) $

  (4)

  Figure 6

  

  Figure 6.  Dynamic characterization of CO₂ reduction. In situ FTIR spectrum of Bi₂WO₆-BiOCl heterostructure under light (a) and dark (b) conditions. The time interval was 15 min. (c) In situ Raman spectrum Bi₂WO₆-BiOCl with step-wise voltage in the electro CO₂ reduction process. Electron spin resonance (ESR) test of Bi₂WO₆-BiOCl with (d) hydroxyl radicals and (e) superoxide radicals. (f) Charge distribution of CO₂ on BiOCl and Bi₂WO₆-BiOCl. (g) The free energy diagram for CO₂ reduction to CO over Bi₂WO₆-BiOCl and pure BiOCl phase.

  DownLoad: Full-Size Img PowerPoint

  As shown in Fig. 6c, peaks range from 830 to 900 cm^‒1 can be assigned to the peroxo-species. Signals at 139 cm^‒1 and 192 cm^‒1 arises from A_1g internal stretching and E_g vibration of Bi-Cl bond in the crystal BiOCl. The peroxo bond with metals around 817 cm^‒1 appeared with Bi-Cl bond during the reaction, along with O—O stretching bond located at 901 cm^‒1 [47-49]. All the four peaks were not observed in the pristine sample, indicating that the reaction proceeds with Bi-Cl and surface water to form the metal-peroxo complex. The bands at 784 and 817 cm^‒1 are associated with the antisymmetric and symmetric Ag modes of the terminal O-W-O bond, respectively. O-W-O bond exhibited reduced intensity at 817 cm^‒1 and shorter length around 784 cm^‒1, reflecting more distorted bridge bonds in the WO₆ octahedral layers [33,50]. Hence, the peroxo species produced with Bi-Cl bond simultaneously in the in-situ reaction process, while W-O bonds showed disorder and less active in the reaction process. We can infer that Cl take part in the reaction process to form O—O and peroxo species, surface anchored BiOCl provide the active sites for the intermediates. Electron spin resonance spectrum in Figs. 6d and e showed the reaction intermediate was OH and O^2‒ radicals [35,48]. Fig. 6f displayed the charge distribution, and BiOCl was content with electrons rather than Bi₂WO₆ in the heterostructure. Combined with the efficient charge dynamics and band alignment, density functional theory (DFT) calculations were further conducted to verify the reactivity of Bi₂WO₆-BiOCl. As shown in Fig. 6g, CO₂ underwent three steps to form CO, where COOH* was the intermediate. Both the active sites are on the surface of BiOCl for two samples, and the process of CO₂ to COOH was the most endothermic step. In Bi₂WO₆-BiOCl, the electron flowed from Bi₂WO₆ to BiOCl, which enabled more electrons to participate into the surface reaction. The energy difference was as high as 1.895 eV for BiOCl from CO₂ to COOH. In comparison, the most endothermic step on Bi₂WO₆-BiOCl was the formation of COOH* intermediate with an energy barrier of 1.085 eV The reduced energy value even reached 0.810 eV, which further uncovered the mechanisms of excellent photocatalytic CO₂ reduction on Bi₂WO₆-BiOCl. The charge distribution in Fig. S18 (Supporting information) showed CO₂ adsorbed on the Bi₂WO₆-BiOCl was more active on the surface [51,52].

  In summary, the construction of in-situ grown Bi₂WO₆-BiOCl heterostructure with coherent interface was achieved via tungstate doping, whose lattice enriched with abundant reduced metal sites with atomic precision. The catalyst exhibits an enhanced photocatalytic efficiency of 68.03 µmol g⁻¹ h⁻¹ with 98% selectivity to CO production, which is higher than those of pristine BiOCl and Bi₂WO₆, propelling itself as the state-of-the-art Bi-based photocatalyst. The promoted performance can be ascribed to the enhanced electron transfer between the atomic level coherent Bi₂WO₆-BiOCl interface, which was confirmed by local coordination environment and valence analysis. The DFT and in-situ FT-IR tests reveal the lower energy barriers of rate-determining step the conversion of adsorbed CO₂ to COOH* over the Bi₂WO₆-BiOCl with coherent interface. This coherent interface engineering route can optimize interfacial charge transfer dynamics beyond energy alignment engineering.

  CRediT authorship contribution statement

  Hui Li: Writing – original draft, Investigation, Formal analysis, Conceptualization. Chunlang Gao: Visualization, Resources, Data curation. Guo Yang: Visualization, Validation, Software. Lu Xia: Software, Methodology, Formal analysis, Data curation. Wulyu Jiang: Writing – review & editing, Visualization, Supervision, Software, Formal analysis, Conceptualization. Cheng Wu: Visualization, Formal analysis. Kaiwen Wang: Writing – review & editing, Resources, Data curation. Yingtang Zhou: Writing – review & editing, Writing – original draft, Validation, Supervision, Resources, Data curation. Xiaodong Han: Writing – review & editing, Validation, Supervision, Resources, Funding acquisition, Data curation.

  Acknowledgments

  This work was supported by the National Key R&D Program of China (No. 2021YFA1200201), the Beijing Outstanding Young Scientists Projects (No. BJJWZYJH01201910005018), The Basic Science Center Program for Multiphase Evolution in Hypergravity of the National Natural Science Foundation of China (No. 51988101), the National Natural Science Foundation of China (Nos. 52071003 and 91860202).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110547.

  
  
• 1. [1]

     C. Vogt. C, M. Monai, G.J. Kramer, et al., Nat. Catal. 2 (2019) 188–197. doi: 10.1038/s41929-019-0244-4

  2. [2]

     S. Zhang, Q. Fan, R. Xia, et al., Acc. Chem. Res. 53 (2020) 255–264. doi: 10.1021/acs.accounts.9b00496

  3. [3]

     D.H. Nam, P. De. Luna, A. Rosas-Hernández, et al., Nat. Mater. 19 (2020) 266–276. doi: 10.1038/s41563-020-0610-2

  4. [4]

     S. Xu, E.A. Carter, Chem. Rev. 119 (2018) 6631–6669.

  5. [5]

     X. Li, J. Yu, M. Jaroniec, et al., Chem. Rev. 119 (2019) 3962–4179. doi: 10.1021/acs.chemrev.8b00400

  6. [6]

     S. Ali, M.C. Flores A. Razzaq, et al., Catalysts. 9 (2019) 727. doi: 10.3390/catal9090727

  7. [7]

     H. Yu, J. Li, Y. Zhang, et al., Angew. Chem. Int. Ed. 58 (2019) 3880–3884. doi: 10.1002/anie.201813967

  8. [8]

     S. Li, L. Bai, N. Ji et al., J. Mater. Chem. A. 8 (2020) 9268–9277. doi: 10.1039/d0ta02102d

  9. [9]

     Y. Hu, R. Abazari, S. Sanati, et al., ACS Appl. Mater. Interfaces 5 (2023) 37300–37311. doi: 10.1021/acsami.3c04506

  10. [10]

      M. Li, S.X. Yu, H. Huang, et al., Angew. Chem. Int. Ed. 58 (2019) 9517–9521. doi: 10.1002/anie.201904921

  11. [11]

      Z. He, Y. Shi, C. Gao, et al., J. Phys. Chem. C. 118 (2014) 389–398. doi: 10.1021/jp409598s

  12. [12]

      S. Shenawi-Khalil, V. Uvarov, S. Fronton, et al., J. Phys. Chem. C 116 (2012) 11004–11012. doi: 10.1021/jp3009964

  13. [13]

      C. Chang, L. Zhu, S. Wang, et al., ACS Appl. Mater. Interfaces 6 (2014) 5083–5093. doi: 10.1021/am5002597

  14. [14]

      M. Sun, M. Chen, X. Cui, et al., Surf. Interfaces 42 (2023) 103313. doi: 10.1016/j.surfin.2023.103313

  15. [15]

      Z. Hou, C. Cui, Y. Li, et al., Adv. Mater. 35 (2023) 2209876. doi: 10.1002/adma.202209876

  16. [16]

      Y. Zhou, Y. Zhang, M. Lin, et al., Nat. Commun. 6 (2015) 8340. doi: 10.1038/ncomms9340

  17. [17]

      F. Chen, Y. Zhang, H. Huang, Chin. Chem. Lett. 34 (2023) 107523. doi: 10.1016/j.cclet.2022.05.037

  18. [18]

      J. Di, C. Chen, S.Z. Yang, et al., Nat. Commun. 10 (2019) 2840. doi: 10.1038/s41467-019-10392-w

  19. [19]

      C. Hu, H. Huang, Acta Phys. Chim. Sin. 39 (2023) 2212048. doi: 10.3866/pku.whxb202212048

  20. [20]

      X. Jin, Y. Xu, X. Zhou, et al., ACS Mater. Lett. 3 (2021) 364–371. doi: 10.1021/acsmaterialslett.1c00091

  21. [21]

      X. Luo, R. Abazari, M. Tahir, et al., Coordin. Chem. Rev. 461 (2022) 214505. doi: 10.1016/j.ccr.2022.214505

  22. [22]

      L.J. McGilly, A. Kerelsky, N.R. Finney, et al., Nat. Nanotechnol. 15 (2020) 580–584. doi: 10.1038/s41565-020-0708-3

  23. [23]

      F. Rao, Y. An, X. Huang, et al., ACS Catal. 13 (2023) 2523–2533. doi: 10.1021/acscatal.2c04954

  24. [24]

      Q. Bie, H. Yin, Y. Wang, Chin. J. Catal. 57 (2024)123–132. doi: 10.1016/S1872-2067(23)64591-7

  25. [25]

      B. Aziz S, S. Marf A, E.M.A. Dannoun, et al., Polymers 12 (2020) 2184. doi: 10.3390/polym12102184

  26. [26]

      Z. Li, G. Zhu, W. Zhang, et al., Chem. Eng. J. 452 (2023) 139378. doi: 10.1016/j.cej.2022.139378

  27. [27]

      J. Cao, J. Wang, Z. Wang, et al., Surf. Interfaces. 45 (2024) 103875. doi: 10.1016/j.surfin.2024.103875

  28. [28]

      Q. Li, X. Zhu, J. Yang, et al., Inorg. Chem. Front. 7 (2020) 597–602. doi: 10.1039/c9qi01370a

  29. [29]

      J. Di, C. Zhu, M. Ji, et al., Angew. Chem. Int. Ed. 57 (2018) 14847–14851. doi: 10.1002/anie.201809492

  30. [30]

      J. Li, L. Cai, J. Shang, et al., Adv. Mater. 28 (2016) 4059–4064. doi: 10.1002/adma.201600301

  31. [31]

      H.Y. Chung, C.Y. Toe, W. Chen, et al., Small 17 (2021) 2170183. doi: 10.1002/smll.202170183

  32. [32]

      D. Zhao, C. Cai. Mater. Chem. Front. 3 (2019) 2521–2528. doi: 10.1039/c9qm00570f

  33. [33]

      X. Cao, Z. Chen, R. Lin, et al., Nat. Catal. 1 (2018) 704–710. doi: 10.1038/s41929-018-0128-z

  34. [34]

      J. He, X. Wang, S. Lan, et al., Appl. Catal. B: Environ. 317 (2022) 121747. doi: 10.1016/j.apcatb.2022.121747

  35. [35]

      C. Yang, Y. Wang, J. Yu, et al., ACS Appl. Energy Mater. 4 (2021) 8734–8738. doi: 10.1021/acsaem.1c02122

  36. [36]

      Y. Zhang, Z. Xu, Q. Wang, et al., Appl. Catal B: Environ. 299 (2021) 120679. doi: 10.1016/j.apcatb.2021.120679

  37. [37]

      Y. Zhou, Z. Tian, Z. Zhao, et al., ACS Appl. Mater. Interfaces 3 (2011) 3594–3601. doi: 10.1021/am2008147

  38. [38]

      J. Huang, G. Tan, H. Ren, et al., ACS Appl. Mater. Interfaces 6 (2014) 21041–21050. doi: 10.1021/am505817h

  39. [39]

      J. Zhang, Z. Li, J. He, et al., ACS Catal. 13 (2023) 785–795. doi: 10.1021/acscatal.2c05545

  40. [40]

      X. YingáKong, Y. YeeáChoo, A. KaháSoh, et al., Chem. Commun. 52 (2016) 14242–14245. doi: 10.1039/C6CC07750A

  41. [41]

      Y. Liu, D. Shen, Q. Zhang, et al., Appl. Catal B: Environ. 283 (2021) 119630. doi: 10.1016/j.apcatb.2020.119630

  42. [42]

      Y. Wang, T. Chen, F. Chen, et al., Sci. China Mater. 65 (2022) 3497–3503. doi: 10.1007/s40843-022-2093-x

  43. [43]

      J. Wu, K. Li, S. Yang, et al., Chem. Eng. J. 452 (2023) 139493. doi: 10.1016/j.cej.2022.139493

  44. [44]

      N. Li, B. Wang, Y. Si, et al., ACS Catal. 9 (2019) 5590–5602. doi: 10.1021/acscatal.9b00223

  45. [45]

      S.S. Bhosale, A.K. Kharade, E. Jokar, et al., J. Am. Chem. Soc. 141 (2019) 20434–20442. doi: 10.1021/jacs.9b11089

  46. [46]

      J. Meng, Y. Duan, S. Jing, et al., Nano Energy 92 (2022) 106671. doi: 10.1016/j.nanoen.2021.106671

  47. [47]

      Y. Bai, L. Ye, T. Chen, et al., Appl. Catal B: Environ. 203 (2017) 633–640. doi: 10.1016/j.apcatb.2016.10.066

  48. [48]

      J. Li, B. Huang, Q. Guo, et al., Appl. Catal B: Environ. 284 (2021) 119733. doi: 10.1016/j.apcatb.2020.119733

  49. [49]

      X. Zhu, G. Zhou, J. Yi, et al., ACS Appl. Mater. Interfaces 13 (2021) 39523–39532. doi: 10.1021/acsami.1c12692

  50. [50]

      X. Zhang, H. Su, P. Cui, et al., Nat Commun. 14 (2023) 7115. doi: 10.1038/s41467-023-42887-y

  51. [51]

      N.V. Maksimchuk, G.M. Maksimov, V.Y. Evtushok, et al., ACS Catal. 8 (2018) 9722–9737. doi: 10.1021/acscatal.8b02761

  52. [52]

      L. Wang, X. Zhao, D. Lv, et al., Adv. Mater. 32 (2020) 2004311. doi: 10.1002/adma.202004311

• 

  Figure 1  Schematic illustration of catalysts with interlayer configuration. Structure models of (a) single phase Bi₂WO₆ and BiOCl. (b) Ball-milling sample via mechanical forces. (c) Tandem Bi₂WO₆-BiOCl with coherent interfaces.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  STEM characterization of Bi₂WO₆-BiOCl. (a) HAADF-STEM image and elemental dispersive scanning of composite with Bi, O, W, Cl. (b) Atomic resolution HAADF-STEM image for (ⅰ) BiOCl and (ⅱ) Bi₂WO₆ phase, the (ⅲ) area was selected to manifest the interface structure. Diffraction patterns of (ⅰ) BiOCl and (ⅲ) Bi₂WO₆ were shown right of the panel. (c) FFT-filtered atomic resolution image using two sets of diffraction spots in red area. The lattice distortions can be visualized by comparing the array disorder. (d) FFT-filtered atomic image of left-up area in Fig. 2b. Four parallel lines were selected in BiOCl (1), interface (2), interface (3) and Bi₂WO₆ (4) to check atomic deviation profile around the interface. (e, f) FFT-filtered atomic resolution image visualization of Fig. 2c and corresponding GPA image of Bi₂WO₆-BiOCl. (g, h) False color image around the interface and GPA image for the ball-milling sample.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Electronic states of elements and local environmental coordination of W. (a) The XPS peak of W 4f for Bi₂WO₆-BiOCl. (b, c) was XPS of Bi 4f and O 1s. (d) The W L_Ⅲ-edge XANES spectra. (e) The corresponding k³-weighted FT spectra of Bi₂WO₆-BiOCl, Bi₂WO₆ was used as reference sample.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Photocatalytic CO₂ reduction (a) efficiency, (b) selectivity and stability of Bi₂WO₆-BiOCl composite, the reference samples involve ball-milling sample, pristine Bi₂WO₆ and pristine BiOCl. (c) The performance comparison of CO yield with pure Bi-based materials.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Photocarrier separation and migration. (a) Photo-current performance and (b) Nyquist plots. (c) PL spectra. (d) Transient time-resolved PL decay spectra. (e) UV–vis spectra of four samples. (f) Band diagram of single phase and Bi₂WO₆-BiOCl.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Dynamic characterization of CO₂ reduction. In situ FTIR spectrum of Bi₂WO₆-BiOCl heterostructure under light (a) and dark (b) conditions. The time interval was 15 min. (c) In situ Raman spectrum Bi₂WO₆-BiOCl with step-wise voltage in the electro CO₂ reduction process. Electron spin resonance (ESR) test of Bi₂WO₆-BiOCl with (d) hydroxyl radicals and (e) superoxide radicals. (f) Charge distribution of CO₂ on BiOCl and Bi₂WO₆-BiOCl. (g) The free energy diagram for CO₂ reduction to CO over Bi₂WO₆-BiOCl and pure BiOCl phase.

  下载: 全尺寸图片 幻灯片
•