English

• Nowadays, the energy demands of human society continue to be primarily dependent on fossil fuels, including coal, oil, and natural gas. However, the overuse of these resources and the environmental pollution associated with modern industrial society have given rise to a range of environmental and energy issues, including clean water shortages, a global energy crisis and the greenhouse effects [1–3]. In light of these challenges, the development of new energy and water environment purification technologies has become a pressing necessity in the future. Photocatalysis has been identified as a promising technology due to its numerous advantages, including environmental friendliness, sustainability, low cost, and the potential for solar energy utilisation [4,5]. In particular, photocatalytic H₂ production from dyestuffs wastewater not only converts solar energy into hydrogen, but accompany the purification of polluted water, presenting a promising prospect for the integration of energy and environmental solutions [6].

  Carbon nitride (g-C₃N₄), composed of earth-abundant nitrogen and carbon elements, has emerged as a promising polymer photocatalyst [7,8]. Nevertheless, due to the high loss rate of precursors in polycondensation reactions and the difficulty in precisely controlling the micro-polycondensation reactions of precursors, regulating the microstructure of g-C₃N₄ has become exceptionally challenging. Herein, a novel 3D cross-linked g-C₃N₄ network (SCN) assembling with 1D nanowires was proposed. This cross-linked structure creates a sponge-like porous structure and introduces abundant defects, which exhibits compelling advantages in photocatalytic H₂ production coupled with the photodegradation reaction [9–11], with the following aspects [12–14] 3D cross-linked network structure increases the number of light-trafficking pathways and promotes multiple reflective absorption, thereby enhancing the absorption efficiency within the porous structures. Furthermore, 3D cross-linked network structure improves molecular transport by constructing mass transfer channels to facilitate migration of reactant molecules to reactive sites, and the secondary nanostructure within 3D cross-linked network structure efficiently shortens the diffusion distance of photoinduced electrons from interior of SCN to the surface-active sites. Additionally, 3D cross-linked porous structure in photocatalysts can efficiently prevents particle agglomeration and achieves a high surface area, providing numerous active adsorption and photocatalytic reaction sites, thereby enhancing photocatalytic performance. The distinctive attributes of 3D cross-linked network structure position them as promising candidates for the development of novel coupled hydrogen production reactions.

  The preparation process of SCN is depicted in Fig. 1a. UCN, synthesized through the direct urea pyrolysis, and semicarbazide hydrochloride (SH) was uniformly mixed by a vacuum freezing-assisted heat treatment. As a material susceptible to pyrolysis, SH decomposes and escapes from within the UCN at elevated temperatures, ultimately giving rise to a 3D cross-linked network with a sponge-like porous architecture in SCN. A distinct 3D cross-linked network with sponge-like pores is evident in the SCN framework as illustrated in Figs. 1b and c. Moreover, the high-magnification SEM image reveals that this 3D cross-linked network is assembled from nanowire-like secondary structure with diameters of –10 nm, which are intricately interwoven to form the sponge-like porous configurations as presented in Fig. 1d. Transmission electron microscopy (TEM) images also highlight this 3D cross-linked network in SCN, aligning with the SEM results as shown in Figs. 1e and f. Additionally, the elemental mapping results indicate that SCN primarily comprises C and N elements, with no evidence of elemental segregation, as demonstrated in Figs. 1g-j.

  Figure 1

  

  Figure 1.  (a) Illustration of the synthetic process of SCN sample. (b–d) Field emission scanning electron microscopy (FESEM), (e, f) TEM images and (g–j) elemental mapping images of SCN.

  DownLoad: Full-Size Img PowerPoint

  It is noteworthy that the XRD peak intensities of SCN are markedly diminished compared to those of UCN, as displayed in Fig. 2a. This implies that the long-range ordered structure of the crystal lattice in SCN undergoes some disruption due to the incorporation of the SH pore-forming agent relative to UCN [15,16]. To further elucidate framework of SCN, the ¹⁵N NMR spectra of SCN and UCN were acquired and presented in Fig. 2b. There are three typical chemical shifts: N₁ (114 ppm), N₂ (135 ppm) and N₃ (187 ppm), corresponding to the typical N atom chemical shifts of –NH₂ and C=N—C, NC₃ in the g-C₃N₄ framework, respectively [17,18]. Notably, the SCN displays a minor peak at 104 ppm but there is absent in the UCN spectrum, which can be indexed as the chemical shift of the N element in the dangling bond of C=N—H, most probably originating from carbon defects in the tri-s-triazine rings as depicted in insert of Fig. 2b [19].

  Figure 2

  

  Figure 2.  (a) XRD, (b) ¹⁵N cross‐polarization solid-state NMR, (c) FT-IR, (d) HR-XPS, and (e) EPR spectra of SCN and UCN. The optimized structures together with corresponding spin-polarized total density of states (TDOS) and partial density of states (PDOS) plots of (f) UCN and (g) SCN, respectively.

  DownLoad: Full-Size Img PowerPoint

  FTIR spectra also reveal the vibration modes of C=N—H at 1535 cm^–1, which is proximate to the typical vibration of C=N—C in the SCN spectrum [20]. However, no vibration peak is observed at this position in UCN, as shown in Fig. 2c [21,22]. Other obvious vibrations in both SCN and UCN at 810, 1200–1700 cm^–1 and 3100–3400 cm^–1 represent the breathing vibration mode, various vibration modes of the C–N heterocycle, and vibration mode of amino groups in carbon nitride, respectively [23]. Moreover, the =NH species caused by the carbon defects can also be examined by the high-resolution N 1s spectrum in SCN as presented in Fig. 2d [24,25]. Additionally, the molar ratios of C/N in UCN and SCN can also be determined from their high-resolution XPS spectra in Fig. 2d and Fig. S2 (Supporting information). The results indicate that the molar ratios of C/N in UCN and SCN are 0.74 and 0.69, respectively, suggesting that SCN is a nitrogen-rich framework [26]. EPR spectrum of SCN manifests a significantly stronger signal than that of UCN as illustrated in Fig. 2e, indicating that the SCN framework contains abundant defect states [10,20], which can be ascribed to be the introduction of carbon defects based on aforementioned experimental results. To delve deeper into their defective structures, spin-polarized total density of states (TDOS) and partial density of states (PDOS) PDOS were employed, as shown in Figs. 2f and g. In comparison to the PDOS of UCN, a small energy level emerges in the band gap region of SCN after the introduction of C defects [27]. This level is situated in close proximity to the conduction band and is postulated to be the defect energy level.

  Furthermore, the DRS spectra of the SCN with the 3D cross-linked network in Fig. 3a reveal the enhanced light harvesting relative to UCN, which is primarily attributed to the defect structure of SCN [28]. The band gaps of UCN and SCN are calculated to be 2.73 and 3.04 eV, respectively, signifying that a blue shift occurs at the absorption band edge of SCN compared to UCN, which can be ascribed to the quantum size confinement effect induced by the secondary nanowire structure of SCN [23,29,30]. Additionally, the Mott–Schottky plot and XPS valence band spectra were employed to elucidate their band structures, as illustrated in Figs. 3b–d. The valence band (VB) position of SCN and UCN samples are determined to be 1.28 and 1.32 V versus NHE (pH 7), respectively. Meanwhile, the conduct band (CB) positions of SCN and UCN samples are calculated to be −1.41 and −1.76 V versus NHE (pH 7), respectively.

  Figure 3

  

  Figure 3.  (a) UV–vis absorption spectra (inset: bandgap plot), (b) Mott-Schottky plots, (c) XPS valence band spectra, (d) band structure, (e) fluorescence spectra, (f) TRPL spectra, (g) transient photocurrent response curves, (h) electrochemical impedance spectra and (i) N₂ adsorption-desorption isothermal curves (inset: pore size distribution) of SCN and UCN.

  DownLoad: Full-Size Img PowerPoint

  Photoluminescence (PL) spectra of SCN and UCN samples demonstrate that the peak intensity of SCN is significantly lower than that of UCN depicted in Fig. 3e. This result indicates that SCN exhibits a relatively low photogenerated carrier recombination rate compared with UCN [9,31]. Moreover, the photogenerated carrier dynamics of SCN and UCN were investigated using time-resolved fluorescence spectroscopy (TRPL), as presented in Fig. 3f. The fitting parameters obtained from their fluorescence decay curves were detailed in Table S2 (Supporting information). For UCN sample, the carrier lifetime τ₁, representing the radiative recombination process of photogenerated electrons and holes, is 1.6 and constitutes 73.3% of the total. Meanwhile, the carrier lifetime τ₂, associated with the non-radiative energy transfer process, is 8.5 ns and accounts for 26.7%. This implies that the majority of photogenerated carriers in UCN undergo a radiative recombination process, leading to the limitation of their intrinsic photocatalytic activity [32]. Contrastively, SCN exhibits different dynamic behaviour with τ₁ and τ₂ values of 1.4 ns and 15.9 ns, accounting for 56.1% and 43.9%, respectively. This suggests that SCN with a 3D network structure and carbon defects can effectively suppress the recombination of photogenerated carriers, thereby an increased number of photocarriers partake in the non-radiative energy transfer process [33,34]. This not only amplifies the number of photogenerated carriers but also prolongs their lifetime, which is conducive to increasing the possibility of photogenerated carriers to participate in photocatalytic reactions and improving photocatalytic performance [9,35].

  Particularly, a notably higher photocurrent intensity for SCN compared with that of UCN reveals that separation of photogenerated charge carriers is more efficient over SCN than UCN, as shown in Fig. 3g, thus generating more effective photoelectrons to participate in the photocatalytic reaction [7,16]. Typically, a low impedance value in EIS signifies that the photoelectrons encounter fewer obstacles during migration [9,24]. The impedance values of SCN and UCN samples are determined to be 175 and 250 Ω in Fig. 3h, respectively. This result implies that the SCN facilitates a more efficient interfacial charge transfer compared to UCN [3,20].

  Their pore size distribution (PSD) curves (inset in Fig. 3i) reveal that SCN sample exhibits two typical pore structures, small mesopores (–3 nm) and macropores (–80 nm) [12]. The former typically forms within or on the surface of the photocatalyst particles, providing additional catalytic active sites [36]; the latter originates from the spaces between stacked carbon nitride nanowires, facilitating the mass transport process of reactant molecules [37,38]. In comparison with UCN, the SCN sample contains a higher proportion of both small mesopores and macropores, which can be further reflected by the total pore volume (V_pore) and specific surface area (S_BET) as summarized in Table S1 (Supporting information). This suggests that the augmented mesopore and macropore volume of SCN contributes to the enhancement of the mass transfer process in photocatalytic reactions [12,39]. Meanwhile, the high specific surface area of SCN furnishes a multitude of surface-active sites and photocatalytic reaction sites, which are highly advantageous for augmenting photocatalytic H₂ evolution in SCN [40].

  Their photocatalytic performance was evaluated through photocatalytic H₂ production coupled with RhB photodegradation, as described in Fig. 4. The photocatalytic hydrogen production rate of SCN reaches 283 µmol h⁻¹ g⁻¹, markedly surpassing the 31 µmol L⁻¹ h⁻¹ achieved by UCN in Fig. 4a. Meanwhile, in the accompanying photocatalytic degradation reactions, attains a 97% degradation efficiency of RhB within 90 min, in contrast to UCN’s 64% degradation rate under analogous conditions in Figs. 4b and c. Moreover, the photocatalytic degradation rate constant for SCN towards RhB is 0.035, approximately four-fold that of UCN (K_ap = 0.009) as presented in Fig. 4d. These results indicate that SCN exhibits enhanced photocatalytic H₂ production and photodegradation performance compared to UCN in RhB-laden wastewater. Notably, in SCN photocatalytic reaction, the hydrogen production efficiency scarcely enhances beyond 90 min, whereas UCN’s hydrogen yield continues to ascend. This discrepancy primarily arises because the RhB in the SCN photocatalytic reaction is nearly entirely degraded by 90 min, rendering it unable to furnish sufficient RhB agents. Conversely, in the UCN system, an undegraded fraction of RhB persists, sustaining the supply of sacrificial agents and thereby bolstering hydrogen production. Furthermore, the apparent quantum efficiencies (AQE) of the SCN sample were an ascertained under diverse monochromatic light wavelengths (380 nm, 400 nm, and 450 nm). The AQE of SCN could reach up to 23.7% under 380 nm illumination, while it was 12.9% and 2.1% under 400 nm and 450 nm illumination, respectively. As elucidated in Fig. 4e, the AQE variation trend for the SCN sample across wavelength is positively correlated with its absorption spectrum.

  Figure 4

  

  Figure 4.  (a) Photocatalytic H₂ production rates coupled with degradation of RhB over SCN and UCN. (b) Photocatalytic degradation spectra of RhB in SCN system. (c) Photocatalytic degradation rates of RhB and (d) rate constants after quasi-primary kinetic simulations. (e) Apparent quantum efficiencies of photocatalytic H₂ production coupled degradation of RhB over SCN under monochromatic illumination at 380 nm, 400 nm and 450 nm, respectively. (f) Photocatalytic hydrogen production rates coupled degradation on SCN + Pt (3% Pt) and UCN + Pt (3% Pt) samples in RhB solution. (g) Photocatalytic degradation of RhB spectra in SCN + Pt system, and corresponding (h) photocatalytic degradation rate of RhB and (i) rate constants after quasi-primary kinetic simulations.

  DownLoad: Full-Size Img PowerPoint

  Interestingly, if Pt nanoparticles were deposited on the surface of SCN and UCN photocatalysts as co-catalysts, they exhibited significantly enhanced performance in the simultaneous photocatalytic H₂ production and RhB photodegradation, as depicted in Fig. 4f. After surface loading of Pt, the photocatalytic H₂ production rates of SCN and UCN photocatalysts increase to 1180–266 µmol L⁻¹ h⁻¹, respectively, which is approximately 4.2-fold and 8.5-fold higher than that of the scenarios without Pt addition. Meanwhile, SCN achieve a 95% degradation rate for RhB within 60 min, while UCN almost completely degraded RhB after 120 min, as shown in Figs. 4g and h. Based on the fitting of the approximate first-order kinetic equation [41], the photocatalytic degradation rate constants for RhB by SCN + Pt and UCN + Pt are determined to be 0.056–0.02 in Fig. 4i, respectively. These experimental results reveal that the 3D cross-linked network structure with carbon defects of SCN not only provide abundant active sites for the photocatalytic degradation and enhance the mass-transfer process relative to UCN, but also significantly inhibit the recombination of photogenerated carriers, improving the separation efficiency. Moreover, the surface-deposited Pt co-catalyst generates a localized surface electric field, which further promotes the separation and migration of photogenerated carriers, resulting in more effective photogenerated carriers involved in the photocatalytic redox reactions [42]. As a result, SCN with Pt cocatalysts demonstrates superior hydrogen production performance in simulated wastewater containing RhB. Furthermore, the supplementary experiments were conducted to assess the photocatalytic stability of SCN, as depicted in Fig. S3 (Supporting information). Upon completion of four cycles of photocatalytic hydrogen evolution coupled with RhB degradation, the hydrogen production capability of SCN exhibited a negligible decline. Concurrently, RhB photocatalytic degradation retained a high efficiency throughout these four cycles. Moreover, X-ray diffraction (XRD) patterns of SCN were acquired both prior to and following the photocatalytic reaction, revealing a high degree of similarity between the pre- and post-reaction patterns, thereby indicating minimal structural alterations in SCN subsequent to the photocatalytic process. Collectively, these findings underscore the robust stability and durability of SCN in the photocatalytic reaction for hydrogen production coupled with RhB degradation.

  Photoelectrons can be captured by oxygen molecules to form the superoxide radicals (^•O₂⁻) because the CB positions of SCN and UCN are more negative than the O₂/^•O₂⁻ potential (−0.33 V vs. NHE, pH 7). Fig. 5a demonstrates that SCN exhibits a higher ESR signal than that of UCN in atmospheric environments, suggesting that SCN could produce more ^•O₂^- radicals relative to UCN. This implies that SCN has the higher concentration photoelectron than that of UCN under the identical illumination conditions in atmospheric environments [43]. This is significant for achieving high-efficiency photocatalytic H₂ evolution activity as these photoelectrons ultimately participates in the photoreduced reactions from H⁺ or H₂O to H₂ molecules [44,45]. It is noteworthy that during the photocatalytic H₂ production coupled with RhB photodegradation reaction, air was completely excluded from the reaction system and no oxygen was contained, thus the signals of superoxide radicals could not be detected, as shown in Fig. 5b. To further explore the active species involved in the photocatalytic H₂ production coupled with RhB photodegradation, isopropanol, benzoquinone, and oxalic acid were added to the reaction systems of SCN and UCN to remove hydroxyl radicals, superoxide radicals, and photoinduced holes as shown in Figs. 5c and d [46]. Compared with the photocatalytic H₂ performance without adding any sacrificial agents, the photocatalytic H₂ production activities of UCN and SCN decrease only slightly after the addition of these reagents. This result indicates that consuming hydroxyl radicals, superoxide radicals and photoinduced holes has a negligible effect on photocatalytic H₂ production in a closed oxygen-free environment. Meanwhile, adding isopropanol or benzoquinone has almost no effect on the efficiency of SCN in photodegrading RhB. However, when oxalic acid was introduced into the SCN reaction, it significantly inhibited the degradation of RhB, with the degradation rates of SCN and UCN for RhB being only 18% and 10%, respectively. This was primarily because SCN and UCN are unable to generate superoxide radicals and hydroxyl radicals in the closed anaerobic environment. Therefore, isopropanol and benzoquinone have no significant effect on photocatalytic degradation activity. In contrast, when oxalic acid was added to the abovementioned anaerobic reaction, it can consume photoholes and form a competitive reaction with RhB, leading to a significant decrease in the degradation efficiency of SCN and UCN towards RhB. These results demonstrate that RhB itself is photodegraded by photogenerated holes and acts as a sacrificial agent, while photogenerated electrons involved in the process of reducing H⁺ to H₂ in the photocatalytic H₂ production reaction coupled with RhB photodegradation under anaerobic conditions.

  Figure 5

  

  Figure 5.  EPR signals of DMPO-^•O₂⁻ adducts in SCN and UCN (a) in atmospheric environments and (b) in photocatalytic H₂ production coupled with RhB degradation system. (c) Photocatalytic hydrogen production rate and (d) RhB degradation efficiency of SCN and UCN after adding isopropanol, benzoquinone and oxalic acid, respectively. Charge density difference (the yellow and sky-blue color loops represent charge accumulation and depletion region) of (e) UCN and (f) SCN. (g) The optimized structures of the H atom adsorbed on C-site and N-site of UCN, and on N1-site and N2-site of SCN. (h) The HER reaction pathways at zero voltage on above optimized structures.

  DownLoad: Full-Size Img PowerPoint

  To understand the mechanism of the C defect in g-C₃N₄ in photocatalytic hydrogen production from RhB-containing wastewater, the charge density distribution was computed to examine the charge transfer, as shown in Figs. 5e and f. The charge transfers from H atoms to SCN is 0.59 e whereas it is 0.46 e for UCN. This indicate that H adsorption on defective site is comparatively stronger than pristine one and therefore shows good HER performance [47]. Furthermore, the thermodynamic process of H₂ evolution reaction (HER) is determined by the hydrogen adsorption energy onto the surface of the photocatalyst, which can be reflected by hydrogen adsorption free energy ∆G(H*) [48]. The results are summarized in Table S3 (Supporting information). A value of ∆G(H*) close to zero suggests that both adsorption and desorption steps have low reaction barriers, favoring HER and indicating a promising HER photocatalyst. The optimized structures of the H atom adsorbed on different active sites are displayed in Fig. 5g, which depicts the H adsorption on different active sites for theoretical calculations of free-energy reaction pathways at 0 V vs. reversible hydrogen electrode (RHE). The transition states for photocatalytic H₂ evolution were calculated to probe the HER reaction pathways, as shown in Fig. 5h. The C-site of UCN has large positive ∆G(H*) (0.42 eV), suggesting that these sites are not catalytically active for HER [49]. Whereas, the hydrogen binding energies become moderate at N-site of SCN. The values of ∆G(H*) for N1-site and N2-site of SCN are −0.20 and −0.16 eV, respectively. These values are much closer to zero, indicating that the adsorption–desorption barrier was significantly reduced, which renders SCN more favorable for H* adsorption and demonstrates outstanding HER performance of N1-site of SCN.

  In summary, a novel 3D cross-linked carbon nitride network (SCN) assembling with nanowires was successfully synthesized via a vacuum freezing-assisted heat treatment strategy. This 3D cross-linked network consists of one-dimensional g-C₃N₄ nanowires interspersed with each other, creating a sponge-like porous structure. Moreover, SCN contains abundant C defects and has newly formed a defect energy level at its forbidden band position, which promotes the efficient separation of photogenerated carriers, and extends the lifetime of photogenerated carriers from 8.5 ns in UCN to 15.9 ns in SCN. In the photocatalytic H₂ production coupled with degradation of RhB reaction, the photocatalytic H₂ production rate of SCN is 283 µmol h^-1 g^-1, with a degradation rate of up to 97% for RhB, corresponding to a degradation rate constant of 0.035. Furthermore, under illumination at 380 nm, 400 nm and 450 nm, the AQY of SCN is 23.7%, 12.9% and 2.1%, respectively. Theoretical calculations show that the values of ∆G(H*) for the N1-site and N2-site of SCN are −0.20 and −0.16 eV, respectively. The N1 site in SCN facilitates H* adsorption and desorption, thereby generating hydrogen on the surface of SCN.

  Declaration of competing interest

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

  CRediT authorship contribution statement

  Linyu Zhu: Writing – original draft, Investigation, Funding acquisition, Conceptualization. Xu Tian: Investigation, Conceptualization. Guang Shi: Methodology, Investigation. Wenchi Zhang: Formal analysis, Conceptualization. Peisong Tang: Methodology, Conceptualization. Mohamed Bououdina: Software, Investigation. Sajjad Ali: Software, Resources, Methodology. Pengfei Xia: Writing – review & editing, Project administration, Funding acquisition, Conceptualization.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 22202033), National College Students Innovation and Entrepreneurship Training Program (No. 202310347048), and Zhejiang Provincial Training Programs of Innovation and Entrepreneurship for Undergraduates (No. S202310347032). The author would like to thank Prince Sultan University for computational support.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111088.

  
  
• 1. [1]

     J. Dong, J. Zhao, X. Yan, et al., Appl. Catal. B: Environ. Energy351 (2024) 123993. doi: 10.1016/j.apcatb.2024.123993

  2. [2]

     C. Feng, T. Bo, P. Maity, et al., Adv. Funct. Mater. 34 (2024) 2309761. doi: 10.1002/adfm.202309761

  3. [3]

     Y. Gao, X. Zhai, Y. Zhang, et al., Nano Mater. Sci. 5 (2023) 177–188. doi: 10.1016/j.nanoms.2022.01.004

  4. [4]

     J. Luo, S. Zhang, M. Sun, L. Yang, J.C. Crittenden, ACS. Nano13 (2019) 9811–9840. doi: 10.1021/acsnano.9b03649

  5. [5]

     M. Xiao, Z. Wang, M. Lyu, et al., Adv. Mater. 31 (2019) 1801369. doi: 10.1002/adma.201801369

  6. [6]

     S. Meng, C. Chen, X. Gu, et al., Appl. Catal. B: Environ. 285 (2021) 119789. doi: 10.1016/j.apcatb.2020.119789

  7. [7]

     P.F. Xia, S.W. Cao, B.C. Zhu, et al., Angew. Chem. Int. Ed. 59 (2020) 5218–5225. doi: 10.1002/anie.201916012

  8. [8]

     A. Genoux, M. Pauly, C.L. Rooney, et al., J. Am. Chem. Soc. 145 (2023) 20739–20744. doi: 10.1021/jacs.3c05417

  9. [9]

     P.F. Xia, M. Antonietti, B.C. Zhu, et al., Adv. Funct. Mater. 29 (2019) 1900093. doi: 10.1002/adfm.201900093

  10. [10]

      G.H. Dong, D.L. Jacobs, L. Zang, C.Y. Wang, Appl. Catal. B: Environ. 218 (2017) 515–524. doi: 10.1016/j.apcatb.2017.07.010

  11. [11]

      M. Jourshabani, M.R. Asrami, B.K. Lee, Appl. Catal. B: Environ. 302 (2022) 120839. doi: 10.1016/j.apcatb.2021.120839

  12. [12]

      X. Li, J. Yu, M. Jaroniec, Chem. Soc. Rev. 45 (2016) 2603–2636. doi: 10.1039/C5CS00838G

  13. [13]

      P.Y. Kuang, M. Sayed, J.J. Fan, B. Cheng, J.G. Yu, Adv. Energy Mater. 10 (2020) 1903802. doi: 10.1002/aenm.201903802

  14. [14]

      S. An, G. Zhang, K. Li, et al., Adv. Mater. 33 (2021) 2104361. doi: 10.1002/adma.202104361

  15. [15]

      J. Fu, J. Yu, C. Jiang, B. Cheng, Adv. Energy Mater. 8 (2017) 1701503.

  16. [16]

      X.H. Wu, H.Q. Ma, W. Zhong, J.J. Fan, H.G. Yu, Appl. Catal. B: Environ. 271 (2020) 118899. doi: 10.1016/j.apcatb.2020.118899

  17. [17]

      L. Lin, Z. Yu, X. Wang, Angew. Chem. Int. Edit. 58 (2019) 6164–6175. doi: 10.1002/anie.201809897

  18. [18]

      Y.P. Zhu, Y.J. Lei, F.W. Ming, et al., Nano Energy65 (2019) 104030. doi: 10.1016/j.nanoen.2019.104030

  19. [19]

      Y.C. Hu, Y. Shim, J. Oh, et al., Chem. Mater. 29 (2017) 5080–5089. doi: 10.1021/acs.chemmater.7b00069

  20. [20]

      Z. Yang, M. Yuan, Z. Cheng, et al., Angew. Chem. Int. Ed. 63 (2024) e202401758. doi: 10.1002/anie.202401758

  21. [21]

      G. Zhao, Y.L. Cheng, Y.Z. Wu, X.J. Xu, X.P. Hao, Small. 14 (2018) 1704138. doi: 10.1002/smll.201704138

  22. [22]

      T.T. Bartels, The Sadtler Handbook of Infrared Spectra, Sadtler, Philadelphia, 1978.

  23. [23]

      P. Xia, B. Zhu, J. Yu, S. Cao, M. Jaroniec, J. Mater. Chem. A5 (2017) 3230–3238. doi: 10.1039/C6TA08310B

  24. [24]

      Y.B. Jiang, Z.Z. Sun, C. Tang, et al., Appl. Catal. B: Environ. 240 (2019) 30–38. doi: 10.1145/3373509.3373568

  25. [25]

      M.H. Vu, M. Sakar, C.C. Nguyen, T.O. Do, A.C.S. Sustain. Chem. Eng. 6 (2018) 4194–4203. doi: 10.1021/acssuschemeng.7b04598

  26. [26]

      M. Groenewolt, M. Antonietti, Adv. Mater. 17 (2005) 1789–1792. doi: 10.1002/adma.200401756

  27. [27]

      X. Yu, S.F. Ng, L.K. Putri, et al., Small. 17 (2021) 2006851. doi: 10.1002/smll.202006851

  28. [28]

      X. Chen, R. Shi, Q. Chen, et al., Nano Energy59 (2019) 644–650. doi: 10.1016/j.nanoen.2019.03.010

  29. [29]

      S. Mondal, L. Sahoo, Y. Vaishnav, et al., J. Mater. Chem. A8 (2020) 20581–20592. doi: 10.1039/d0ta08001b

  30. [30]

      L. Liang, L. Shi, F. Wang, et al., Appl. Catal. A599 (2020) 117618. doi: 10.1016/j.apcata.2020.117618

  31. [31]

      Q. Han, B. Wang, Y. Zhao, C.G. Hu, L.T. Qu, Angew. Chem. Int. Ed. 54 (2015) 11433–11437. doi: 10.1002/anie.201504985

  32. [32]

      Y. Liang, X. Wu, X. Liu, C. Li, S. Liu, Appl. Catal. B: Environ. 304 (2022) 120978. doi: 10.1016/j.apcatb.2021.120978

  33. [33]

      S. Gao, X. Wang, C. Song, et al., Appl. Catal. B: Environ. 295 (2021) 120272. doi: 10.1016/j.apcatb.2021.120272

  34. [34]

      B. Yang, J. Han, Q. Zhang, et al., Carbon. N. Y. 202 (2023) 348–357. doi: 10.1016/j.carbon.2022.10.074

  35. [35]

      Y. Li, M. Zhou, B. Cheng, Y. Shao, J. Mater. Sci. Technol. 56 (2020) 1–17. doi: 10.1016/j.jmst.2020.04.028

  36. [36]

      Y. Mun, M.J. Kim, S.A. Park, et al., Appl. Catal. B: Environ. 222 (2018) 191–199. doi: 10.1016/j.apcatb.2017.10.015

  37. [37]

      D. Gu, C.J. Jia, C. Weidenthaler, et al., J. Am. Chem. Soc. 137 (2015) 11407–11418. doi: 10.1021/jacs.5b06336

  38. [38]

      M.A. Isaacs, N. Robinson, B. Barbero, et al., J. Mater. Chem. A7 (2019) 11814–11825. doi: 10.1039/c9ta01867k

  39. [39]

      Y. Wei, N. Yang, K. Huang, et al., Adv. Mater. 32 (2020) 2002556. doi: 10.1002/adma.202002556

  40. [40]

      Y. Chen, M. Cheng, C. Lai, et al., Small. 19 (2023) 2205902. doi: 10.1002/smll.202205902

  41. [41]

      P. Shoghi, M. Hamzehloo, J. Colloid Interf. Sci. 616 (2022) 453–464. doi: 10.1016/j.jcis.2022.02.028

  42. [42]

      T.K.A. Nguyen, T. Trần-Phú, X.M.C. Ta, et al., Small Methods8 (2024) 2300427. doi: 10.1002/smtd.202300427

  43. [43]

      L. Zhu, H. Li, Q. Xu, D. Xiong, P. Xia, J. Colloid Interf. Sci. 564 (2020) 303–312. doi: 10.1016/j.jcis.2019.12.088

  44. [44]

      C. Han, P. Su, B. Tan, et al., J. Colloid Interf. Sci. 581 (2021) 159–166. doi: 10.1016/j.jcis.2020.07.119

  45. [45]

      Q. Xu, B. Zhu, B. Cheng, et al., Appl. Catal. B: Environ. 255 (2019) 117770. doi: 10.1016/j.apcatb.2019.117770

  46. [46]

      S. Zhou, L. Jiang, H. Wang, et al., Adv. Funct. Mater. 33 (2023) 2307702. doi: 10.1002/adfm.202307702

  47. [47]

      T. Tang, Z. Wang, J. Guan, Chin. J. Catal. 43 (2022) 636–678. doi: 10.1016/S1872-2067(21)63945-1

  48. [48]

      I.T. McCrum, M.T. Koper, Nat. Energy5 (2020) 891–899. doi: 10.1038/s41560-020-00710-8

  49. [49]

      W. Zhong, B. Xiao, Z. Lin, et al., Adv. Mater. 33 (2021) 2007894. doi: 10.1002/adma.202007894

• 

  Figure 1  (a) Illustration of the synthetic process of SCN sample. (b–d) Field emission scanning electron microscopy (FESEM), (e, f) TEM images and (g–j) elemental mapping images of SCN.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) XRD, (b) ¹⁵N cross‐polarization solid-state NMR, (c) FT-IR, (d) HR-XPS, and (e) EPR spectra of SCN and UCN. The optimized structures together with corresponding spin-polarized total density of states (TDOS) and partial density of states (PDOS) plots of (f) UCN and (g) SCN, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) UV–vis absorption spectra (inset: bandgap plot), (b) Mott-Schottky plots, (c) XPS valence band spectra, (d) band structure, (e) fluorescence spectra, (f) TRPL spectra, (g) transient photocurrent response curves, (h) electrochemical impedance spectra and (i) N₂ adsorption-desorption isothermal curves (inset: pore size distribution) of SCN and UCN.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Photocatalytic H₂ production rates coupled with degradation of RhB over SCN and UCN. (b) Photocatalytic degradation spectra of RhB in SCN system. (c) Photocatalytic degradation rates of RhB and (d) rate constants after quasi-primary kinetic simulations. (e) Apparent quantum efficiencies of photocatalytic H₂ production coupled degradation of RhB over SCN under monochromatic illumination at 380 nm, 400 nm and 450 nm, respectively. (f) Photocatalytic hydrogen production rates coupled degradation on SCN + Pt (3% Pt) and UCN + Pt (3% Pt) samples in RhB solution. (g) Photocatalytic degradation of RhB spectra in SCN + Pt system, and corresponding (h) photocatalytic degradation rate of RhB and (i) rate constants after quasi-primary kinetic simulations.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  EPR signals of DMPO-^•O₂⁻ adducts in SCN and UCN (a) in atmospheric environments and (b) in photocatalytic H₂ production coupled with RhB degradation system. (c) Photocatalytic hydrogen production rate and (d) RhB degradation efficiency of SCN and UCN after adding isopropanol, benzoquinone and oxalic acid, respectively. Charge density difference (the yellow and sky-blue color loops represent charge accumulation and depletion region) of (e) UCN and (f) SCN. (g) The optimized structures of the H atom adsorbed on C-site and N-site of UCN, and on N1-site and N2-site of SCN. (h) The HER reaction pathways at zero voltage on above optimized structures.

  下载: 全尺寸图片 幻灯片
•