English

• In recent years, the excessive exploitation and use of fossil fuels have caused severe energy and environmental problems, with the intensified greenhouse effect driven by massive CO₂ emissions being particularly urgent [1–3]. Among various carbon-mitigation technologies, photocatalytic CO₂ reduction to high-value carbon products is especially promising [4,5]. Conventional photocatalytic CO₂ reduction typically employs triethanolamine as a sacrificial agent to capture photogenerated carriers, which shortens catalyst lifetime and introduces secondary pollution [6]. Moreover, single-component photocatalysts face an irreconcilable trade-off between light-harvesting efficiency and redox capability, leading to low CO₂-reduction efficiencies [7–9]. To overcome these limitations, researchers have advanced photocatalytic CO₂-reduction systems and constructed heterojunctions to address the band-gap constraints of single photocatalysts. Such heterojunctions generally deliver higher carrier-separation efficiencies and superior activities than their single-component counterparts [10–13]. Among these heterojunction architectures, the conventional Type-Ⅱ heterojunction enables spatial charge carrier separation but inevitably compromises the overall redox capability of the photocatalytic system. In sharp contrast, the S-scheme heterojunction, fabricated by integrating oxidation-type and reduction-type semiconductors, not only preserves robust redox capability but also establishes a prominent internal electric field (IEF) that facilitates the selective recombination of useless carriers [14–17]. Among numerous semiconductors, titanium dioxide (TiO₂) has garnered extensive research attention owing to its excellent photocatalytic activity and superior chemical stability [18]. However, as a metal oxide with a 3.0–3.4 eV band gap, TiO₂ suffers from limited visible-light absorption and rapid carrier recombination [19–21]. Enhancing carrier transport and visible-light harvesting-thereby unlocking its CO₂-reduction potential-requires coupling TiO₂ with a narrow-band-gap oxidation-type semiconductor to construct an optimal S-scheme heterojunction. Copper tungstate (CuWO₄) is an n-type semiconductor with a suitable band gap, low toxicity, and strong oxidation capability [22,23]. Thus, assembling an efficient S-scheme photocatalyst from CuWO₄ and TiO₂ is a promising strategy, whereas TiO₂ nanoparticles are uniformly loaded onto CuWO₄. The strong electrostatic assembly yields intimately bonded crystalline interfaces, which reduces interfacial transport resistance and strengthens the IEF-driven charge-transfer force. Here, we couple TiO₂ with CuWO₄ via electrostatic self-assembly to construct TiO₂/CuWO₄ heterojunctions (abbreviated TC) with excellent photocatalytic performance. Photoelectrochemical tests confirm that TC sample exhibits an excellent IEF and strong redox capability. The CO yield reaches 9.53 μmol g⁻¹ h⁻¹. Kinetic analyses show that the TC heterojunction possesses higher carrier density, longer carrier lifetimes, and faster charge-transfer rates than single-component photocatalysts, explaining its superior performance. Density functional theory (DFT) calculations further indicate that TC heterojunction lowers the reaction barrier of the rate-determining step *CO₂ to *COOH thereby enabling efficient CO₂ reduction.

  Herein, TC composite photocatalysts were prepared via electrostatic self-assembly (Fig. 1a), where the zeta potentials of TiO₂ and CuWO₄ were +13.0 mV and −12.31 mV, respectively (Fig. S1 in Supporting information). Transmission electron microscopy (TEM) was employed to examine the microstructure, revealing that TiO₂ nanoparticles are uniformly coupled to the CuWO₄ surface, forming a tight heterojunction interface (Figs. 1b and c). HRTEM provided lattice information for TC heterojunction [24]. As shown in Fig. 1d, the measured interplanar spacings are 0.24 nm (TiO₂) and 0.29 nm (CuWO₄), corresponding to the characteristic crystal planes of the respective phases [25,26]. The lattice fringes of the two phases overlap distinctly at the interface, further confirming the intimate coupling of the heterojunction. In addition, as shown in Figs. 1e-i, energy-dispersive X-ray spectroscopy (EDS) elemental mapping clearly shows a uniform distribution of Ti, W, Cu, and O elements indicates that the TC heterojunction was successfully synthesized [21]. As shown in Fig. S2 (Supporting information), the water contact angles of TiO₂, CuWO₄, and TC-3 are 30.1°, 40.6°, and 30.7°, respectively, indicating that the prepared samples all exhibit excellent hydrophilicity [27,28]. The diffraction peaks of TiO₂ match those of the standard TiO₂ card (JCPDS No. 21–1272), indicating the successful synthesis of high-purity TiO₂ (Fig. S3 in Supporting information) [29]. In Fig. S3, CuWO₄ exhibits strong diffraction peaks at 15.4°, 18.8°, 30.2°, and 36.9°, which can be attributed to the (010), (001), (111), and (002) crystal planes of CuWO₄ (JCPDS No. 70–1732) [26]. The XRD pattern of the TC-3 composite also shows diffraction peaks at the same angles, corresponding to the crystal planes of both TiO₂ and CuWO₄, further confirming the successful construction of the TC-3 composite photocatalyst.

  Figure 1

  

  Figure 1.  (a) Diagram of TC heterojunction sample preparation. (b, c) TEM and (d) HRTEM image of TC sample. (e-i) EDS elemental mapping of TC heterojunction.

  DownLoad: Full-Size Img PowerPoint

  The TC-3 composite photocatalyst contains Ti, W, Cu, and O elements (Fig. S4 in Supporting information) [30]. In Fig. 2a, the X-ray photoelectron spectroscopy (XPS) of Ti 2p in TiO₂ sample can be deconvoluted into two peaks at 458.4 eV (Ti 2p_3/2) and 464.3 eV (Ti 2p_1/2) [31]. In the TC-3 composite, these two characteristic peaks shift slightly to 458.8 and 464.7 eV, respectively. The Cu 2p XPS spectrum of CuWO₄ shows two characteristic peaks at 953.8 and 933.8 eV, corresponding to the 2p_3/2 and 2p_1/2 orbitals of Cu (Fig. 2b) [32]. After constructing the heterojunction, the Cu 2p of TC-3 shift to lower binding energies (0.6 eV). In Fig. 2c, the W 4f XPS spectrum shows two peaks at 37.1 and 34.9 eV, corresponding to the 4f_5/2 and 4f_7/2 orbitals of W [20,32]. In the TC-3 heterojunction, the W 4f XPS spectrum of CuWO₄ also shift slightly to lower binding energies, reaching 36.7 and 34.5 eV. Additionally, the Ti 2p peak in the TC-3 heterojunction shifts to higher binding energies under dark conditions, indicating that TiO₂ acts as an electron (e^-) donor, while the Cu 2p and W 4f peaks shift to lower binding energies, suggesting that CuWO₄ acts as an e^- acceptor [33]. As shown in Fig. S5 (Supporting information), the light absorption edge of TiO₂ is 385 nm. Compared with pure TiO₂, the TC-3 heterojunction composite photocatalyst exhibits a broader light absorption range, laying a foundation for efficient CO₂ photocatalytic reduction. Fig. 2d clearly shows the charge migration from CuWO₄ to TiO₂, which is consistent with the XPS characterization results, confirming the strong interaction between TiO₂ and CuWO₄. As shown in Fig. S6 (Supporting information), the work functions of TiO₂ and CuWO₄ are 4.14 and 6.29 eV, respectively. When TiO₂ and CuWO₄ are in close contact, the mismatch in their Fermi energy levels causes charge migration from TiO₂ to CuWO₄, resulting in the formation of a IEF at the interface of heterojunction, which aligns with the XPS results. Under light exposure, the characteristic peaks of Cu and W elements in the TC-3 composite shift toward higher binding energies, while the Ti 2p peaks show the opposite trend (Figs. 2e and f, Fig. S7 in Supporting information) [13]. This indicates that the photogenerated charges migrate from CuWO₄ to TiO₂, driven by the IEF, forming an S-scheme charge transfer pathway (Figs. 2g and h) [34,35]. Therefore, under light illumination, photogenerated charges in CuWO₄ combine with holes in TiO₂, accelerating the recombination of nonproductive carriers and leaving behind the strongly redox-active carriers to participate in the reaction (Fig. 2i). The band structures of TiO₂ and CuWO₄ were investigated using the Kubelka-Munk function and XPS valence band spectra [13]. As shown in Figs. S8 and S9 (Supporting information), the band gaps of TiO₂ and CuWO₄ were calculated to be 3.26 and 1.94 eV, respectively, based on the Kubelka-Munk function. As shown in Figs. S10 and S11 (Supporting information), the valence band (VB) levels of TiO₂ and CuWO₄ were determined using XPS valence band spectra, which were found to be 1.85 and 2.46 V, respectively [12]. Additionally, the conduction band (CB) values of TiO₂ and CuWO₄ were calculated using the formula (E_g = E_VB - E_CB), which yielded values of −1.41 and 0.52 V, respectively. These values are consistent with the results obtained from the Mott-Schottky test (Fig. S12 in Supporting information) [36,37].

  Figure 2

  

  Figure 2.  The XPS spectra of Ti 2p (a), Cu 2p (b), and W 4f (c). (d) The charge-density-difference images of TC-3 heterojunction interfacial structure (the green area represents the consumption of e^-, and the yellow regions represent e^- aggregation regions). In situ XPS spectra of Ti 2p (e) and Cu 2p (f). (g-i) Sketches of TiO₂ and CuWO₄ before/after contact and light irradiation.

  DownLoad: Full-Size Img PowerPoint

  In Figs. 3a and b, pure TiO₂ and CuWO₄ exhibits low photocatalytic activity due to the rapid recombination of photogenerated carriers. The CO yield of TiO₂ and CuWO₄ is 2.13 and 0.71 μmol g^-1 h^-1, respectively. The highest CO yield of 9.53 μmol g^-1 h^-1 with 96.7% carbon products selectivity is observed for the TC-3, which is 4.47 times that of pure TiO₂ (Fig. S13 in Supporting information). Compared with recently reported photocatalysts (Tables S1 in Supporting information), the prepared TC-3 S-scheme heterojunction shows excellent photocatalytic activity. In a sealed reaction vessel, photogenerated holes oxidize H₂O to produce oxygen (O₂). However, due to the thermodynamically favorable O₂ reduction reaction, the produced O₂ is consumed, which is consistent with previous report [38]. As displayed in Figs. S14 and S15 (Supporting information), the hydrogen peroxide (H₂O₂) is the only oxidation production. The H₂O₂ production rate of TC-3 is 9.07 μmol g^-1 h^-1. The apparent quantum yield of the TC-3 S-scheme heterojunction at 420 nm reaches 0.46%. In Fig. 3c, no CO production was detected under the conditions of argon (Ar), darkness, and no catalyst. A peak at m/z = 29 (¹³CO) was observed, further confirming that the detected CO indeed originates from the photoreduction of CO₂ (Fig. 3d) [38]. After continuous photocatalytic CO₂ reduction experiments, the CO yield of the TC-3 S-scheme heterojunction only slightly decreased (Fig. 3e). After continuous photocatalytic CO₂ reduction reactions, the XPS spectra of the TC-3 sample (including the characteristic peaks of Ti 2p, W 4f, Cu 2p, and O 1s) exhibited no discernible shifts, indicating the preservation of the elemental valence states and surface chemical composition (Figs. S16 in Supporting information). Meanwhile, the X-ray diffraction (XRD) patterns revealed that the crystallinity of TC-3 remained intact (Fig. 3f). These results collectively demonstrate the excellent structural and chemical stability of this catalyst during long-term photocatalytic CO₂ conversion. As shown in Fig. S17a (Supporting information), the surface charge densities of TiO₂ and TC-3 composites are 2.3 and 0.1 μC/cm², respectively. The steady-state surface photovoltage (SPV) of the TC-3 composite material reaches 170.2 μV, which is higher than that of TiO₂ (52.5 μV) (Fig. S17b in Supporting information). Using formula F_s = (−2V_sρ/εε₀)^1/2, the IEF strength of the prepared photocatalysts was calculated [13]. As shown in Fig. S17c (Supporting information), the IEF strength of the TC-3 composite is 8.6 times of TiO₂. The fluorescence (PL) intensity of TC-3 is much weaker than that of TiO₂ and CuWO₄, indicating that the construction of the S-scheme heterojunction facilitates the rapid separation of charge carriers (Fig. S18a in Supporting information). TiO₂ and CuWO₄ samples display a larger high-frequency semicircle in the Nyquist plot compared to the TC-3 S-scheme heterojunction, indicating that TC-3 has the lowest charge transfer resistance (Fig. S18b in Supporting information). The photocurrent response of the TC-3 composite photocatalyst is significantly higher than that of pure TiO₂ and CuWO₄, suggesting that the construction of an excellent IEF promotes the rapid migration of charge carriers (Fig. S18c in Supporting information). As shown in Fig. S19 (Supporting information), TC-3 exhibits a larger electrochemical active surface area, indicating that its surface has more active sites, which is favorable for the photocatalytic reduction of CO₂ to CO [39,40].

  Figure 3

  

  Figure 3.  (a, b) Photocatalytic activity of the TiO₂, TC-1, TC-2, TC-3, and TC-4. (c) CO evolution under different reaction conditions. Mass spectra of ¹³CO result (d), stability test (e), and XRD patterns of TC-3 before and after photoreduction CO₂ reaction cycles (f).

  DownLoad: Full-Size Img PowerPoint

  After pump pulse excitation, the TiO₂ sample exhibits significant stimulated emission (SE) and ground state bleaching (GSB) signals (Fig. 4a and Fig. S20a in Supporting information) [41,42]. In contrast, CuWO₄ shows no distinct SE signals (Fig. 4b and Fig. S20b in Supporting information), but a negative peak at 528 nm, which corresponds to a typical GSB signal. The TC-3 composite photocatalyst exhibits clear GSB and SE signals (Fig. 4c and Fig. S20c in Supporting information). The signal intensity of femtosecond transient absorption spectroscopy (fs-TAS) for TiO₂ and CuWO₄ is significantly lower than that of TC-3, indicating a relatively fast recombination rate of their photogenerated charge carriers [43]. Additionally, an extra negative peak (573 nm) is observed in the fs-TA spectrum of the TC-3 sample, which is not present in the fs-TA spectra of TiO₂ and CuWO₄ (Fig. S20c). Therefore, this extra negative peak is attributed to charge transfer at the heterojunction interface of the TC-3 composite material, providing direct evidence for the S-scheme charge transfer pathway. The decay curve of the TiO₂ sample at 570 nm, fitted with a biexponential model, reveals two relaxation pathways for the photogenerated charges: Charge diffusion (τ₁ = 5.3 ps, shorter) and carrier recombination (τ₂ = 306.1 ps) (Figs. 4d and g) [44]. For the CuWO₄ photocatalyst, the decay curve, also fitted using the biexponential model, shows that τ₁ = 3.3 ps is associated with charge diffusion, and τ₂= 226.8 ps corresponds to the recombination of photogenerated charges and holes (Figs. 4e and h). As shown in Figs. 4f and i, the TC-3 composite photocatalyst exhibits three distinct relaxation pathways. Specifically, τ₁ = 4.8 ps is assigned to charge diffusion, and τ₃ = 268.8 ps is attributed to the recombination of photogenerated e^- and holes. However, τ₂ = 18.9 ps is ascribed to the recombination of photogenerated holes in TiO₂ with photogenerated charges in CuWO₄, promoting the consumption of nonproductive carriers [42].

  Figure 4

  

  Figure 4.  2D pseudo-color fs-TA spectra of TiO₂ (a), CuWO₄ (b), and TC-3 (c) under 365 nm pump. The extracted decay kinetics of TiO₂ (e), CuWO₄ (e), and TC-3 (f). The schematic illustration of e^- dynamics in TiO₂ (g), CuWO₄ (h), and TC-3 (i).

  DownLoad: Full-Size Img PowerPoint

  These characteristic peaks correspond to monodentate carbonate (m-CO₃^2-), bicarbonate (HCO₃^-), carboxyl (COOH), and bidentate carbonate (b-CO₃^2-) groups (Fig. 5a) [45]. The peaks at 1582 and 1558 cm^-1 are attributed to m-CO₃^2-, while the peaks at 1330 cm^-1 correspond to adsorbed b-CO₃^2-. The peaks at 1470 cm^-1 correspond to adsorbed HCO₃^-. The characteristic peak at 1652 and 1130 cm^-1 are assigned to the key intermediate *COOH [46]. The following photocatalytic CO₂ reduction process is proposed: First, some CO₂ molecules are adsorbed and activated on the TiO₂ surface (*CO₂), while other CO₂ molecules react with water to form H₂CO₃. Then, H₂CO₃ dissociates into HCO₃^-, H⁺, and CO₃^2-. Next, *CO₂ reacts with H⁺ and photogenerated charges on the TiO₂ material surface to generate the key intermediate *COOH (*CO₂ + H⁺+ e^- → *COOH). The key intermediate *COOH is reduced by H⁺ and photogenerated charges to *CO (*COOH + H⁺+ e^- → *CO + H₂O). Finally, the CO molecule desorbs from the TiO₂ surface (Figs. S21 and S22 in Supporting information). First, CO₂ is adsorbed and activated at the active sites on TiO₂ and TC-3. Next, proton-coupled e^- transfer occurs, leading to the formation of the key intermediate *COOH (Fig. 5b) [47]. The reaction free energy barriers for the formation of *COOH on TiO₂ and TC-3 are 2.41 and 1.73 eV, respectively [48]. The DFT results confirm that the construction of the S-scheme heterojunction promotes proton-coupled e^- transfer to generate the key intermediate *COOH, which is further converted into *CO, thereby enabling efficient photocatalytic CO₂ reduction. We calculated the charge density differences before and after CO₂ and *COOH adsorption on the TiO₂ and TC-3 S-scheme heterojunction, showing that the charge redistribution at the heterojunction interface facilitates CO₂ activation and *COOH stabilization (Figs. S23 and S24 in Supporting information). The mechanism of photocatalytic CO₂ reduction under UV–vis-IR light for the TC-3 composite sample is shown in Fig. 5c. The TC-3 composite sample follows an S-scheme charge transfer mechanism. Under UV–vis-IR illumination, TiO₂ and CuWO₄ in the TC-3 composite photocatalyst are rapidly excited, and a large number of holes and charges accumulate in the VB and CB of the TiO₂ and CuWO₄ semiconductor photocatalysts, respectively. Due to the IEF between TiO₂ and CuWO₄, which provides a strong driving force for the S-scheme charge transfer pathway, useless carriers are quickly recombined, increasing the utilization of carriers with strong redox capabilities. Finally, the photogenerated e^- with strong reducing power in the CB of TiO₂ participate in the CO₂ reduction reaction. Eventually, efficient conversion of CO₂ into CO via photocatalysis is accomplished.

  Figure 5

  

  Figure 5.  In situ DRIFTS spectra (a) and free energy diagrams (b) of the TiO₂ and TC-3, schematic diagram photoreduction CO₂ (c) of TC-3 S-scheme heterojunction.

  DownLoad: Full-Size Img PowerPoint

  In summary, we designed an S-scheme heterojunction based on nanoscale TiO₂ using hydrothermal and electrostatic self-assembly techniques. The fs-TA spectroscopy analysis confirmed that the IEF accelerates useless carrier recombination at the heterojunction interface, promoting the participation of strongly reducing carriers in the photocatalytic reaction. The IEF strength of TC-3 is 7.6 times higher than that of TiO₂. The optimal TC-3 S-scheme heterojunction achieved a CO yield of 9.53 μmol g^-1 h^-1, while the CO yield of pure TiO₂ was only 2.13 μmol g^-1 h^-1. This work provides new insights for constructing efficient and stable S-scheme heterojunction photocatalysts, addressing the inherent limitations of single-component photocatalysis.

  CRediT authorship contribution statement

  Heng Zhang: Writing – original draft, Software, Resources, Investigation. Jiayi Zhang: Project administration, Investigation. Wenke Wang: Supervision, Formal analysis. Shuang Ma: Writing – original draft, Validation. Peiying Yang: Supervision, Funding acquisition. Changdong Chen: Software, Funding acquisition. Haopeng Jiang: Writing – review & editing, Visualization, Validation. Hao Xu: Writing – review & editing, Supervision, Investigation, Data curation.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. 22571067, U1904184 and U1904191) and the Natural Science Foundation of Henan Province (No. 252300423125).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2026.112725.

  
  
• 1. [1]

     Z.K. Xie, W.J. Cheng, H.Y. Luo, et al., Adv. Energy Mater. 16 (2026) 2501840. doi: 10.1002/aenm.202501840

  2. [2]

     Q. Cheng, J.M. Li, Y.X. Huang, et al., J. Mater. Sci. Technol. 261 (2026) 83–93. doi: 10.1016/j.jmst.2025.10.030

  3. [3]

     T.W. Zhao, W.X. Zhang, J.Y. Xiong, et al., Sep. Purif. Technol. 360 (2025) 130849. doi: 10.1016/j.seppur.2024.130849

  4. [4]

     K. Wang, Y.J. Hu, X.F. Liu, et al., Nat. Commun. 16 (2025) 2094. doi: 10.1038/s41467-025-57140-x

  5. [5]

     G. Yang, J.Y. Xiong, M.J. Lu, et al., J. Colloid. Interface Sci. 624 (2022) 348–361. doi: 10.1016/j.jcis.2022.05.092

  6. [6]

     K.N. Song, S.J. Liang, X.H. Zhong, et al., Appl. Catal. B: Environ. 309 (2022) 121232. doi: 10.1016/j.apcatb.2022.121232

  7. [7]

     K. Wang, C. Liu, J.P. Li, et al., Appl. Catal. B: Environ. 361 (2025) 124560. doi: 10.1016/j.apcatb.2024.124560

  8. [8]

     G. Yang, P. Qiu, J.Y. Xiong, et al., Chin. Chem. Lett. 33 (2022) 3709–3712. doi: 10.1016/j.cclet.2021.10.047

  9. [9]

     K. Wang, Q. Cheng, W.D. Hou, et al., Adv. Funct. Mater. 34 (2024) 2309603. doi: 10.1002/adfm.202309603

  10. [10]

      L.H. Tan, X.H. Wu, J.C. Xu, et al., Chin. J. Catal. 79 (2025) 174–185. doi: 10.1016/S1872-2067(25)64849-2

  11. [11]

      F.Y. Xu, F.F. Zhao, X.Y. Deng, et al., Nat. Commun. 16 (2025) 6882. doi: 10.1038/s41467-025-60961-5

  12. [12]

      G. Santos, L. Tian, R. Gonçalves, et al., Adv. Funct. Mater. 35 (2025) 242205.

  13. [13]

      H. Jiang, L. Wang, X. Yu, et al., Chem. Eng. J. 414 (2023) 3129.

  14. [14]

      X.H. Wu, M. Sayed, G.H. Wang, et al., Adv. Mater. 38 (2026) e11322. doi: 10.1002/adma.202511322

  15. [15]

      L.J. Sun, J.S. Gao, J. Shen, et al., Ceram. Int. 50 (2024) 53801–53811. doi: 10.1016/j.ceramint.2024.10.234

  16. [16]

      J.C. Wang, H.R. Ma, W.N. Shi, et al., ACS Sustain. Chem. Eng. 13 (2025) 11537–11546. doi: 10.1021/acssuschemeng.5c04025

  17. [17]

      X.H. Wu, L.H. Tan, G.Q. Chen, et al., Sci. China Mater. 67 (2024) 444–472. doi: 10.1007/s40843-023-2755-2

  18. [18]

      L.J. Sun, X.H. Yu, L.Y. Tang, et al., Chin. J. Catal. 52 (2023) 164–175. doi: 10.1016/S1872-2067(23)64507-3

  19. [19]

      A. Zhou, Y. Dou, C. Zhao, Appl. Catal. B: Environ. Energy 264 (2020) 118519. doi: 10.1016/j.apcatb.2019.118519

  20. [20]

      X. Bai, Z.Y. Fu, X.Y. Ma, et al., J. Cleaner Prod. 369 (2022) 133099. doi: 10.1016/j.jclepro.2022.133099

  21. [21]

      R.T. Han, K. Wang, Q.C. Jiang, et al., J. Colloid Interface Sci. 671 (2024) 680–691. doi: 10.1016/j.jcis.2024.05.206

  22. [22]

      Y. Ding, B. Ding, H. Kanda, et al., Nat. Nanotechnol. 17 (2022) 598–605. doi: 10.1038/s41565-022-01108-1

  23. [23]

      H.W. Su, H.B. Yin, W. Orbell, et al., Angew. Chem. Int. Ed. 64 (2025) e202425446. doi: 10.1002/anie.202425446

  24. [24]

      S. Ma, P.Y. Yang, H. Sun, et al., Sep. Purif. Technol. 354 (2025) 128976. doi: 10.1016/j.seppur.2024.128976

  25. [25]

      G. Wang, C. Wang, J. Zheng, J. Mater. Sci. 55 (2020) 3940–3950. doi: 10.1007/s10853-019-04283-0

  26. [26]

      C. Wang, Y.Q. Zhang, L.J. Zhao, et al., J. Colloid Interface Sci. 592 (2021) 186–194. doi: 10.1016/j.jcis.2021.02.047

  27. [27]

      L. Liu, W.K. Zhao, T. Gao, et al., Mater. Today Chem. 48 (2025) 103000. doi: 10.1016/j.mtchem.2025.103000

  28. [28]

      F.L. Zhao, M.L. Zhang, D. Yan, et al., ACS Appl. Nano Mater. 6 (2023) 14478–14487. doi: 10.1021/acsanm.3c02534

  29. [29]

      N. Sondezi, Z. Njengele-Tetyana, K.P. Matabola, ACS Omega 9 (2024) 15959–15970. doi: 10.1021/acsomega.3c09308

  30. [30]

      S. Ma, X.X. Wang, X. Zhang, et al., Inorg. Chem. Commun. 180 (2025) 114981. doi: 10.1016/j.inoche.2025.114981

  31. [31]

      M. Sun, Y.R. Ma, Y.W. Tan, et al., SusMat 5 (2025) e70011. doi: 10.1002/sus2.70011

  32. [32]

      D.F. Zhang, D. Zhang, F.P. Zhao, et al., J. Mater. Chem. A 12 (2024) 33546–33558. doi: 10.1039/d4ta06654e

  33. [33]

      X.W. Shi, Z.Q. Su, C. Deng, et al., Adv. Mater. 37 (2025) 2508492. doi: 10.1002/adma.202508492

  34. [34]

      Z.K. Xie, L.H. Li, S.H. Gong, Angew. Chem. Int. Ed. 63 (2024) e202410250. doi: 10.1002/anie.202410250

  35. [35]

      Z.Q. Yang, S.Y. Yang, H.J. Yang, et al., Small 21 (2025) e08960. doi: 10.1002/smll.202508960

  36. [36]

      D. Yan, H. Miao, J. Fan, et al., Langmuir 39 (2023) 16648–16656. doi: 10.1021/acs.langmuir.3c02719

  37. [37]

      T. Yang, X.Y. Hu, J. Fan, et al., Surf. Interfaces 42 (2023) 103352. doi: 10.1016/j.surfin.2023.103352

  38. [38]

      W. Song, C. Wang, Y. Liu, et al., J. Am. Chem. Soc. 146 (2024) 29028–29039. doi: 10.1021/jacs.4c10023

  39. [39]

      M.L. Zhang, F.L. Zhao, K.T. Feng, et al., J. Alloys Compd. 963 (2023) 171302. doi: 10.1016/j.jallcom.2023.171302

  40. [40]

      T. Sun, C.X. Li, Y.P. Bao, et al., Acta Phys. Chim. Sin. 39 (2023) 2212009. doi: 10.3866/pku.whxb202212009

  41. [41]

      F.Y. Xu, W.T. Mei, P.Y. Hu, et al., Angew. Chem. Int. Ed. 64 (2025) e202513364. doi: 10.1002/anie.202513364

  42. [42]

      L.J. Sun, W.C. Yan, W.K. Wang, et al., Angew. Chem. Int. Ed. 64 (2025) e202511080. doi: 10.1002/anie.202511080

  43. [43]

      W.K. Wang, W. Shan, Y. Hu, et al., Chem. Eng. J. 493 (2024) 152516. doi: 10.1016/j.cej.2024.152516

  44. [44]

      H.Y. Li, J.J. Zhang, B.C. Zhu, et al., Adv. Funct. Mater. 36 (2026) e21318. doi: 10.1002/adfm.202521318

  45. [45]

      L. Hao, L. Kang, H.W. Huang, et al., Adv. Mater. 31 (2019) 1900546. doi: 10.1002/adma.201900546

  46. [46]

      D.D. Liu, L.P. Jiang, D.Q. Chen, et al., ACS Catal. 14 (2024) 5326–5343. doi: 10.1021/acscatal.4c00409

  47. [47]

      S.H. Yang, W.J. Byun, F.M. Zhao, et al., Adv. Mater. 36 (2024) 2312616. doi: 10.1002/adma.202312616

  48. [48]

      J.Y. Wang, H.Y. Zhang, Y. Nian, et al., Adv. Funct. Mater. 34 (2024) 2406549. doi: 10.1002/adfm.202406549

• 

  Figure 1  (a) Diagram of TC heterojunction sample preparation. (b, c) TEM and (d) HRTEM image of TC sample. (e-i) EDS elemental mapping of TC heterojunction.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  The XPS spectra of Ti 2p (a), Cu 2p (b), and W 4f (c). (d) The charge-density-difference images of TC-3 heterojunction interfacial structure (the green area represents the consumption of e^-, and the yellow regions represent e^- aggregation regions). In situ XPS spectra of Ti 2p (e) and Cu 2p (f). (g-i) Sketches of TiO₂ and CuWO₄ before/after contact and light irradiation.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a, b) Photocatalytic activity of the TiO₂, TC-1, TC-2, TC-3, and TC-4. (c) CO evolution under different reaction conditions. Mass spectra of ¹³CO result (d), stability test (e), and XRD patterns of TC-3 before and after photoreduction CO₂ reaction cycles (f).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  2D pseudo-color fs-TA spectra of TiO₂ (a), CuWO₄ (b), and TC-3 (c) under 365 nm pump. The extracted decay kinetics of TiO₂ (e), CuWO₄ (e), and TC-3 (f). The schematic illustration of e^- dynamics in TiO₂ (g), CuWO₄ (h), and TC-3 (i).

  下载: 全尺寸图片 幻灯片
  

  Figure 5  In situ DRIFTS spectra (a) and free energy diagrams (b) of the TiO₂ and TC-3, schematic diagram photoreduction CO₂ (c) of TC-3 S-scheme heterojunction.

  下载: 全尺寸图片 幻灯片
•