English

• Nowadays, global energy crisis and environmental issues have become an imminent threat to human development. Solar-driven overall water splitting into H₂ and O₂ under photocatalysts is considered as the ultimate approach towards solar-to-fuel conversion [1-5], due to the carbon-free feature, simple facilities, as well as low cost [6-8]. However, this strategy has an inherent defect where the evolved gaseous H₂ and O₂ mixture needs to be separated for safe application [9-10]. Moreover, direct oxidation reaction of H₂O to O₂ was demonstrated to suffer from slow kinetics due to the complicated four-electron pathway [11]. One feasible strategy to address these problems is to avoid O₂ evolution by adding hole scavengers [12], but the resulting complicated operation and maintenance process along with the high cost of sacrificial reagents make this strategy not very progressive indeed. Another alternative solution is to split water into H₂ and H₂O₂, in which the reaction proceeds in a two-electron pathway .i.e., 2H₂O → H₂ + H₂O₂). Although the four-electron pathway for O₂ evolution is thermodynamically more favored than two-electron pathway for H₂O₂ production (1.23 eV.vs. 1.78 eV), the latter was demonstrated to be faster in kinetics [13-15]. The water-soluble H₂O₂ product avoids the gas separation process after water splitting. Additionally, H₂O₂ itself can be served as high-value oxidant with widespread application in food, medicine, industry and daily life. In industry, H₂O₂ is currently produced through an energy-intensive anthraquinone process [16]. Therefore, simultaneous H₂ and H₂O₂ production from photocatalytic water splitting provides an appealing alternative to solar-to-fuel and solar-to-oxidant conversion [17].

  The design and fabrication of efficient, stable, and low-cost photocatalysts is the core issue for photocatalytic water splitting. Since the pioneer works of Fujishima [18,19], semiconductor materials have been intensively investigated as photocatalysts for water splitting. Among them, graphitic carbon nitride (g-C₃N₄) was the first reported material for photocatalytic H₂ production from water splitting under visible light [20]. The past decade has witnessed the rapid development of g-C₃N₄-based photocatalysts for overall water splitting, due to their simple preparation method, superior chemical stability, resistance to photocorrosion, and appropriate band structure for water splitting [21-25]. g-C₃N₄ was also capable of catalyzing water splitting to produce H₂O₂ [13-15]. Unfortunately, the intrinsically poor conductivity, rapid recombination of photogenerated carriers and the inactive surface limit its practical application [26-28]. Various strategies have been adopted to address these problems, such as doping, defect engineering, and heterojunction construction [14,29-30]. Integration of functionalized cocatalyst with g-C₃N₄ primary catalyst to construct heterostructures in their interfacial region holds tremendous promise in terms of both enhanced charge separation and surface activation [31]. The intimate contact between cocatalyst and g-C₃N₄ is an essential requirement to induce vertical migration of photogenerated carriers to active sites [26,32]. For instance, Xue.et al. [30] constructed a novel [P-doped g-C₃N₄]/[red phosphorus]/[C._xP] heterojunction photocatalyst for water splitting. It was confirmed that electron transfer.via homogeneous phosphorus bridges in interfacial regions enabling boosted photocatalytic generation of H₂ and H₂O₂. Similar result was also observed in [P-doped g-C₃N₄]/[C._xN._yP] heterojunction photocatalyst, in which the chemical connection between primary and cocatalyst.via a P⁺−.^δ−−C.^δ+/N.^δ+ bridge endowed the photocatalyst with an excellent H₂ evolution rate as high as 239.3 µmol g⁻¹ h⁻¹ [31]. It is also worth noting that the size and distribution of a given cocatalyst need to be refined to provide more active sites for catalysis [31,33]. However, it usually becomes failure when researchers strive to simultaneously impart the aforementioned properties to one photocatalytic system. A readily producible, efficient, stable, and low-cost catalyst with broadband photocatalytic performance for H₂ and H₂O₂ production from water splitting is thus highly desirable.

  Herein, we report a novel g-C₃N₄/AgI._x._y heterojunction photocatalyst for the simultaneous production of H₂ and H₂O₂ from water splitting. The composite catalyst was prepared by classic thermal polycondensation method followed by simple coprecipitation process. Bimetallic sulfide was selected as the cocatalyst in view of its low cost, intensive visible-light response and widespread application in optical, optoelectronic, and photocatalytic areas [34-37]. Especially, the tubular AgI._x._y deposited on the g-C₃N₄ nanosheet is expected to construct a robust 3D-on-2D configuration for a more abundant interfacial heterostructures. Spectroscopic and photoelectrochemical studies were performed to disclose the interfacial charge transfer behavior between primary and cocatalyst. The photocatalytic performance for water splitting was evaluated under visible irradiation ( > 420 nm) to reveal the simultaneous and stoichiometric feature of H₂ and H₂O₂ production. And finally, a reasonable photocatalytic mechanism for water splitting into H₂ and H₂O₂ was proposed according to the results of active sites and band alignment analyses. g-C₃N₄ nanosheets were prepared by thermal polycondensation of melamine precursor followed by a further thermal exfoliation process (see Supporting information for detailed preparation and characterization methods). g-C₃N₄/AgI._x._y composite photocatalysts were prepared by the coprecipitation method. The pristine g-C₃N₄ nanosheets (g-C₃N₄) showed a blocky morphology with relatively flat surface (Fig. 1a) ascribed to the insufficient exfoliation of nanosheets and resultant interlayer stacking. The morphology of pristine silver indium sulfide (AIS) was considered as an aggregate of many randomly arranged nanotubes with an outer diameter of 30−50 nm and an inner diameter of 10−20 nm (Fig. 1b), respectively. AIS nanotubes were not very uniform in diameter and length. After coprecipitation reaction, many nanorods were deposited onto the flat g-C₃N₄ surface (Figs. 1c and d). The diameters of these nanorods were roughly equal to the outer diameter of pristine AIS nanotubes.

  Figure 1

  

  Figure 1.  SEM micrographs of g-C₃N₄ (a), AIS (b), and 70-GAIS (c, d). (e) TEM and (f) HRTEM micrographs of 70-GAIS. t₁, t₂, t₃ and t₄ display the inverse fast Fourier transform (IFFT) corresponding to the related regions.

  DownLoad: Full-Size Img PowerPoint

  TEM images of the composite verified the tubular structure of the deposited nanorods (Fig. 1e). The inner diameter of the deposited nanotubes was also comparable to that of pristine AIS nanotubes, indicating the g-C₃N₄ component in a coprecipitation deposition process had negligible influence on the formation and structure of AIS nanotubes. The element mapping images of the composite showed a uniform distribution of C and N elements on the surface of the material, while the distribution of Ag, In and S elements was consistent with the position of nanotubes (Fig. S1 in Supporting information). The composition of g-C₃N₄/AIS composite was further investigated by high-resolution TEM. As shown in Fig. 1f, the lattice fringes, with spacings of 0.33, 0.31 and 0.19 nm, were clearly observed in the nanotube region, corresponding to the (002), (201) lattice planes of AgInS₂ and (440) plane of AgIn₅S₈ [38], respectively. Furthermore, lattice fringe with spacing of 0.32 nm was also observed in the flat region, corresponding to the (002) plane of g-C₃N₄. All the above analyses proved that we have successfully prepared g-C₃N₄ nanosheets/AIS nanotubes composite.

  The crystalline structure of the composite was further investigated by X-ray diffraction analysis. As shown in Fig. 2a, g-C₃N₄ gave two characteristic peaks at 12.9° and 27.6°, corresponding to (100) and (002) lattice planes, respectively. The former is attributed to the in-planar repeating triazine motifs while the latter originates from the interlayer stacking of g-C₃N₄ sheets [39,40]. For AIS nanotubes, the diffraction peaks appeared at 25.4°, 26.6°, 28.4°, 28.7°, 44.5°, 52.6°, 13.2° and 47.3°, where the first six peaks were aroused by the (200), (002), (121), (201), (320) and (322) planes of AgInS₂ (PDF #25–1328) while the last two peaks were attributed to the (111) and (440) planes of AgIn₅S₈ (PDF #25–1029) [38]. The aforementioned diffraction peaks were all observed in the pattern of the g-C₃N₄/AIS composite.

  Figure 2

  

  Figure 2.  XRD patterns (a), XPS spectra (b), and C 1s high-resolution XPS spectra (c) of g-C₃N₄, AIS, and 70-GAIS. (d) N 1s high-resolution XPS spectra of g-C₃N₄ and 70-GAIS.

  DownLoad: Full-Size Img PowerPoint

  To reveal the surface chemical composition and state of the g-C₃N₄/AIS photocatalyst, X-ray photoelectron spectroscopy (XPS) characterization was then performed. All binding energies were calibrated by C 1s binding energy at 284.6 eV. As anticipated, various element signals reasonably appeared in the spectra of corresponding samples (Fig. 2b). Note that a C 1s signal was also observed in the spectrum of AIS. Since AIS is a carbon-free material, this C 1s signal could be attribute to C−C from adventitious carbon contaminant. The deconvoluted C 1s high-resolution XPS peaks of g-C₃N₄ and GAIS at 287.9 and 288.6 eV could be assigned to the sp²-hybridized carbon (N=C−(N)₂, see Fig. 2c) [41]. In N 1s XPS spectrum of g-C₃N₄, the peak at 398.3 eV was assigned to the sp²-hybridized nitrogen (C=N−C), the peak at 399.8 eV to the three-coordinate nitrogen (N−(C)₃), and the peak at 404.1 eV to the.π-excitation of C−N heterocycles (Fig. 2d) [42]. The spectrum of GAIS showed similar N species whereas the peaks were located at 399.0, 400.5, and 402.5 eV, respectively. The significant positive shift (+0.7 eV) of the N 1s binding energy for both C=N−C and N−(C)₃ groups, indicating a decreased N electron density, could be caused by the electron transfer from g-C₃N₄ to AIS at the interfacial region [42,43]. This argument can be further validated by high-resolution XPS spectra of other elements (Fig. S2 in Supporting information). As shown in Fig. S2a, AIS displayed two Ag 3d peaks centered at 373.9 and 368.0 eV, which can be assigned to Ag 3d_3/2 and Ag 3d_5/2 core levels [44,45]. For GAIS, these two peaks shifted to 373.8 and 367.8 eV respectively, suggesting slight increasing of Ag electron density in AIS component after being deposited on g-C₃N₄. The In 3d and S 2p XPS spectra gave similar results (Figs. S2b and c). The decreasing of g-C₃N₄ electron density along with the increasing of AIS electron density in g-C₃N₄/AIS composite photocatalyst verified the vital electron transfer from g-C₃N₄ to AIS.

  We next evaluated the photocatalytic performance of the as-prepared g-C₃N₄/AIS composites for water splitting under visible light. Fig. 3a showed the typical time course of H₂ evolution from water splitting over 70-GAIS within 24 h. The sample exhibited rapid H₂ evolution in the first 4 h, with a rate of about 962 µmol g⁻¹ h⁻¹, followed by a slow H₂ evolution process until 20 h, and a prolonged irradiation time over 24 h did not result in further increasing of H₂ production. The average H₂ evolution rate in 24 h was calculated to be 237 µmol g⁻¹ h⁻¹ over 70-GAIS photocatalyst. The cumulative H₂ productions of various photocatalysts under visible light irradiation in 24 h were shown in Fig. 3b. Neither pristine g-C₃N₄ nor AIS had detectable photocatalytic performance for H₂ production from water splitting. However, simple deposition of AIS onto g-C₃N₄ could significantly enhance the photocatalytic activity, and this improvement was related to the added amount of g-C₃N₄ in the reaction mixture during catalyst preparation. When feeding amout of g-C₃N₄ was less 70 mg, H₂ production increased with the increase of g-C₃N₄/AgNO₃ ratio, and the largest H₂ production was obtained in the case of 70-GAIS with the g-C₃N₄ feeding amout of 70 mg. A further increase in feeding g-C₃N₄ would suppress the hydrogen production. This result is reasonable when we consider the fact that a moderate g-C₃N₄/AgNO₃ ratio is equivalent to a moderate coverage degree of AIS nanotubes on g-C₃N₄ surface.

  Figure 3

  

  Figure 3.  (a) Typical time course of H₂ evolution from water splitting over 70-GAIS. (b) Photocatalytic water splitting performance of various photocatalysts under visible light in 24 h.

  DownLoad: Full-Size Img PowerPoint

  The gas products from water splitting are usually H₂ and O₂. However, no O₂ was detected in our photocatalytic water splitting experiment. According to the general splitting mechanism, we speculated that the water splitting reaction in our photocatalytic system proceeded in a two-electron pathway,.i.e., the oxidation product was H₂O₂. It is well-known that MnO₂ is efficient catalyst to decompose H₂O₂ into O₂ and H₂O as 2H₂O₂ → O₂ + 2H₂O. To verify our speculation, a certain amount of MnO₂ was added to the 70-GAIS reaction suspension after a 24-h photocatalytic reaction [31,46-47]. As shown in Fig. S3a (Supporting information), the amount of O₂ evolution gradually increased after the addition of MnO₂ and remained unchanged after 90 min. No additional H₂ evolution was detected throughout this process. The cumulative O₂ evolution within 120 min was 2.73 mmol/g, which was just half the H₂ production in a 24-h photocatalytic reaction. Regardless of the g-C₃N₄ content, an apparent 2:1 stoichiometric ratio of H₂ to O₂ was observed for all prepared g-C₃N₄/AIS photocatalysts (Fig. 3b), indicating that equimolar H₂ and H₂O₂ were produced in a typical photocatalytic water splitting reaction. This two-electron pathway for water splitting was further verified by rotating ring-disk electrode (RRDE) test. It is well known that a 0.9 V .vs. NHE) potential of ring electrode could ensure the generated H₂O₂ on disk electrode to be oxidized to O₂ [14,15]. The electron transfer number under this potential was 1.92 (Fig. S3b in Supporting information), indicating the crucial H₂O₂ species on disk electrode and therefore revealing a two-electron pathway of water splitting (2H₂O → H₂ + H₂O₂) over g-C₃N₄/AIS photocatalysts. The excellent photocatalytic H₂ production performance of the optimal catalyst (70-GAIS) in this work exceeded that of most of the g-C₃N₄-based photocatalysts for simultaneous H₂ and H₂O₂ production from water splitting [15,48,49] and was comparable with that of previously reported C._xN._yP clusters decorated phosphorized-g-C₃N₄ photocatalyst [31].

  Efficient charge separation and fast charge carrier transfer are essential to achieve satisfactory photocatalytic efficiency [50]. To disclose the interfacial charge transfer behavior, we next performed spectroscopic and photoelectrochemical studies. The steady-state photoluminescence (PL) spectra were depicted in Fig. 4a. It is obvious that g-C₃N₄ displayed intense emission at about 463 nm, which originated from the recombination of the excited electron-hole pairs [50]. AIS nanotubes gave a moderate emission band at 380−470 nm whereas for g-C₃N₄/AIS composite the photoluminescence nearly quenched within the determined wavelength range. This phenomenon indicated that recombination of photogenerated electron-hole pairs over g-C₃N₄ had been effectively inhibited after deposition of AIS nanotubes. Time-resolved transient fluorescence decay curves were also recorded to investigate the charge carrier lifetime. GAIS exhibited a relatively faster exponential decay than g-C₃N₄ (Fig. 4b). The average charge carrier lifetimes of GAIS and g-C₃N₄ were calculated to be 1.60 and 1.80 ns, respectively. This significantly declined lifetime signified faster transfer of photogenerated charge carriers on the surface/interface of GAIS photocatalyst. Such an enhanced charge transfer property was further validated by a larger photocurrent response of GAIS compared to pure g-C₃N₄ and AIS (Fig. 4c). Furthermore, electrochemical impedance spectroscopy (EIS) study also demonstrated that GAIS had the lowest interfacial charge transfer resistance, as evidenced by the smallest arc radius in corresponding Nyquist plots (Fig. 4d) [51,52].

  Figure 4

  

  Figure 4.  Steady-state photoluminescence (PL) spectra (a), time-resolved transient fluorescence decay curves (b), photocurrent responses (c), and Nyquist plot from an electrochemical impedance spectroscopy study (d) of g-C₃N₄, AIS, and 70-GAIS.

  DownLoad: Full-Size Img PowerPoint

  Take all the results together, we could conclude that the strong heterojunction interaction at the interfacial region of the g-C₃N₄/AIS photocatalyst facilitated the separation and transfer of photogenerated charge carriers, and therefore achieved excellent photocatalytic performance towards water splitting.

  Fig. S4 (Supporting information) displayed N₂ adsorption-desorption isotherms and the corresponding pore size distribution curves of g-C₃N₄, AIS, and GAIS. Obviously, all three materials gave typical Type Ⅳ isotherm with Type H3 hysteresis loop (Fig. S4a) according to IUPAC classification [53], indicating the mesoporous feature of these materials. The Type H3 hysteresis loop suggested that the mesoporous structures were associated with slit-shaped pores formed by aggregates of plate-like particles [53-54]. The pore size distribution curves derived from desorption isotherm branches by BJH method indicated that mesopores were deficient in AIS, whereas g-C₃N₄ and GAIS had more mesoporous structure with pore width centered at 32.5 nm (g-C₃N₄) and 3.7 nm (GAIS), respectively (Fig. S4b). g-C₃N₄ and AIS showed close BET specific surface area (19 and 21 m²/g) and pore volume (0.11 and 0.09 cm³/g). However, after deposition of AIS onto the g-C₃N₄ surface, the specific surface area and pore volume of the GAIS composite significantly elevated to 67 m²/g and 0.28 cm³/g respectively, along with a decrease of average pore size to 9.2 nm (The average pore size is 19.5 and 12.0 nm for g-C₃N₄ and AIS, respectively). This result was associated with the emerging slit-like pores in the interfacial region of g-C₃N₄/AIS. The increased specific surface area and pore volume of g-C₃N₄/AIS composite catalyst could provide more active sites for photocatalytic reaction and facilitate the escape of gaseous products.

  In the following experiments, AgNO₃ and Pb(NO₃)₂ were employed as indicators to analyze the active sites for reduction and oxidation in a photocatalytic water splitting system. As shown in Figs. 5a and b, Ag particles were deposited on the surface of g-C₃N₄ nanosheets, and clear lattice spacings of 0.23 and 0.32 nm are in good agreement with the (111) lattice planes of Ag and (002) planes of g-C₃N₄ [13,55], respectively. It turned out that the reduction active sites are located on the g-C₃N₄ surface in the process of photocatalytic water splitting. The clear lattice spacings of 0.28 and 0.31 nm in Fig. 5d correspond to (111) planes of PbO₂ and (201) planes of AgInS₂ respectively [13,38], where PbO₂ particles were deposited on the surface of AIS nanotubes (Fig. 5c). This result suggested that the oxidation sites are located on the surface of AIS nanotubes.

  Figure 5

  

  Figure 5.  HRTEM micrographs of reduction sites (a, b) and oxidation sites (c, d) over 70-GAIS.

  DownLoad: Full-Size Img PowerPoint

  The photoelectrochemical properties and the electron band structures of photocatalysts were further investigated by UV–vis diffuse reflectance spectroscopy (DRS) analysis. As can be seen from Fig. 6a, g-C₃N₄ could absorb light from ultraviolet to visible region with an absorption edge at about 465 nm. AIS nanotubes could capture photons in the whole UV–vis region with an absorption edge up to 784 nm, although its absorption intensity was relatively weaker than that of g-C₃N₄. After deposition of AIS onto g-C₃N₄, the GAIS composite exhibited excellent visible absorption property with an absorption edge up to 758 nm and enhanced absorption intensity comparable with that of g-C₃N₄. The band gap .E_g) of such photocatalysts were determined using Tauc plot analysis as described in Fig. 6b. The.E_g of g-C₃N₄ and AIS were calculated to be 2.78 and 2.49 eV, respectively. The significantly decreased band gap of GAIS contributed to its excellent visible absorption performance. In addition, the flat band potentials of g-C₃N₄ and AIS were determined by Mott-Schottky plot. As shown in Fig. 6c, the positive slopes for both g-C₃N₄ and AIS indicated their n-type semiconductor nature [56]. The flat potentials .V_FB) of g-C₃N₄ and AIS were determined to be −1.52 and −0.72 V .vs. NHE, normal hydrogen electrode), respectively. Therefore, the corresponding conduction band minima .V_CB) were calculated to be −1.32 and −0.52 V .vs. NHE) for g-C₃N₄ and AIS, respectively. The valence band maxima .V_VB) can thus be calculated as +1.46 and +1.97 eV for g-C₃N₄ and AIS, respectively.

  Figure 6

  

  Figure 6.  (a) UV–vis diffuse reflectance spectra, and (b) Tauc plots of g-C₃N₄, AIS, and 70-GAIS, where.A refers to the absorbance. (c) Mott-Schottky plots of g-C₃N₄ and AIS. (d) A simplified band structure diagram of g-C₃N₄/AIS photocatalyst and proposed photocatalytic mechanism for simultaneous production of H₂ and H₂O₂ from pure water splitting.

  DownLoad: Full-Size Img PowerPoint

  Based on all experimental results, a possible photocatalytic water splitting mechanism for simultaneous production of H₂ and H₂O₂ was proposed in Fig. 6d. Under visible irradiation, electrons located in the valance bands (VB) of g-C₃N₄ and AIS are respectively photoexcited to their conduction bands (CB), forming electron-hole pairs (e⁻−h⁺). Since low energy gap and intimate contact in interfacial region, photogenerated electrons on the CB of AIS can easily combinate with holes on the VB of g-C₃N₄, remaining high-reactive photogenerated electrons on the CB of g-C₃N₄ and holes on the VB of AIS. Subsequently, the high-reactive carriers migrate to the surface of the photocatalyst to produce active sites to catalyze the water splitting reaction. Owing to suitable redox potentials, the electrons on the CB of g-C₃N₄ can reduce the adsorbed H₂O to produce H₂. Meanwhile, the holes on the VB of AIS will oxidize H₂O into H₂O₂ through a two-electron pathway. Such a redox mediator-free S-scheme heterojunction photocatalyst and resultant efficient transfer, separation and migration of electrons and holes between g-C₃N₄ and AIS can inhibit the rapid recombination of highly-active photogenerated carriers, and thus the photocatalytic performance of the g-C₃N₄/AIS composite is remarkably improved.

  In summary, we have successfully demonstrated a S-scheme heterojunction photocatalyst for pure water splitting into stoichiometric H₂ and H₂O₂ under visible light. The catalyst is composed of AgI._x._y nanotubes deposited on g-C₃N₄ nanosheets prepared by classic thermal polycondensation method followed by simple coprecipitation process. The introduction of bimetallic sulfide endowed the composite catalyst with enhanced visible-light response. A suitable band alignment of g-C₃N₄/AgI._x._y photocatalyst well meet the thermodynamic demands for simultaneous H₂ and H₂O₂ production. Therefore, excellent photocatalytic water splitting performance under visible irradiation ( > 420 nm) can be obtained, with a 24-h H₂ evolution rate as high as 237 µmol g⁻¹ h⁻¹. H₂O₂ was the only oxidation product from water splitting and produced at a 1:1 stoichiometric ratio to H₂. As a novel 3D-on-2D composite system, the g-C₃N₄/AgI._x._y photocatalyst could provide abundant contact interface for heterojunction formation and thus efficient transfer and separation of photogenerated carriers. Water was reduced to H₂ by active e⁻ on the conduction band of g-C₃N₄ nanosheets and oxidized to H₂O₂ by active h⁺ on the valance band of AgI._x._y nanotubes. The efficient separation of photogenerated carriers inhibited their recombination and therefore significantly improved the photocatalytic performance. The novel 3D-on-2D configuration for heterojunction construction proposed in this work provided alternative research ideas toward photocatalytic reaction.

  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.

  CRediT authorship contribution statement

  Hualin Jiang: Supervision. Wenxi Ye: Investigation. Huitao Zhen: Investigation. Xubiao Luo: Resources. Vyacheslav Fominski: Resources. Long Ye: Writing – original draft. Pinghua Chen: Funding acquisition.

  Acknowledgments

  This work is financially supported by the National Natural Science Foundation of China (Nos. 52362012, 42077162, 51978323), Natural Science Foundation of Jiangxi Province (No. 2022ACB203014), Major Discipline Academic and Technical Leaders Training Program of Jiangxi Province (Nos. 20213BCJ22018, 20232BCJ22048), Natural Science Project of the Educational Department in Jiangxi Province (No. GJJ2201121), Natural Science Foundation of Nanchang Hangkong University (No. EA202202256), Educational Reform Project of Jiangxi Province (No. JXYJG-2022–135), Nanchang Hangkong University Educational Reform Project (Nos. sz2214, sz2213, JY22017, KCPY1806).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109984.

  
  
• 1. [1]

     Y. Liu, X.W. Zhang, L.S. Lu, et al., Chin. Chem. Lett. 33 (2022) 1271–1274. doi: 10.1016/j.cclet.2021.07.054

  2. [2]

     S.S. Chen, T. Takata, K. Domen, Nat. Rev. Mater. 2 (2017) 17050. doi: 10.1038/natrevmats.2017.50

  3. [3]

     C. He, F.S. Han, W.X. Zhang, Chin. Chem. Lett. 33 (2022) 404–409. doi: 10.1016/j.cclet.2021.07.010

  4. [4]

     Y.M. He, D.W. Wang, Chem 4 (2018) 405–408. doi: 10.1016/j.chempr.2018.02.013

  5. [5]

     Z. Wang, C. Li, K. Domen, Chem. Soc. Rev. 48 (2019) 2109–2125. doi: 10.1039/C8CS00542G

  6. [6]

     B.A. Pinaud, J.D. Benck, L.C. Seitz, et al., Energy Environ. Sci. 6 (2013) 1983–2002. doi: 10.1039/c3ee40831k

  7. [7]

     D.M. Fabian, S. Hu, N. Singh, et al., Energy Environ. Sci. 8 (2015) 2825–2850. doi: 10.1039/C5EE01434D

  8. [8]

     Y. Goto, T. Hisatomi, Q. Wang, et al., Joule 2 (2018) 509–520. doi: 10.1016/j.joule.2017.12.009

  9. [9]

     Z. Wang, Y. Inoue, T. Hisatomi, et al., Nat. Catal. 1 (2018) 756–763. doi: 10.1038/s41929-018-0134-1

  10. [10]

      F.A. Chowdhury, M.L. Trudeau, H. Guo, et al., Nat. Commun. 9 (2018) 1707. doi: 10.1038/s41467-018-04067-1

  11. [11]

      Y. Qi, Y. Zhao, Y.Y. Gao, et al., Joule 2 (2018) 2393–2402. doi: 10.1016/j.joule.2018.07.029

  12. [12]

      A. Kudo, Y. Miseki, Chem. Soc. Rev. 38 (2009) 253–278. doi: 10.1039/B800489G

  13. [13]

      X.Q. Wu, C. Zhu, L.P. Wang, et al., ACS Catal. 7 (2017) 1637–1645. doi: 10.1021/acscatal.6b03244

  14. [14]

      Z. Li, C. Kong, G.X. Lu, J. Phys. Chem. C 120 (2016) 56–63. doi: 10.1021/acs.jpcc.5b09469

  15. [15]

      J. Liu, Y. Liu, N.Y. Liu, et al., Science 347 (2015) 970–974. doi: 10.1126/science.aaa3145

  16. [16]

      J.Y. Tang, T.S. Zhao, D. Solanki, et al., Joule 5 (2021) 1432–1461. doi: 10.1016/j.joule.2021.04.012

  17. [17]

      Y.Y. Sun, L. Han, P. Strasser, Chem. Soc. Rev. 49 (2020) 6605–6631. doi: 10.1039/d0cs00458h

  18. [18]

      S. Nakabayashi, A. Fujishima, K. Honda, Chem. Phys. Lett. 102 (1983) 464–465. doi: 10.1016/0009-2614(83)87447-8

  19. [19]

      R. Baba, S. Nakabayashi, A. Fujishima, et al., J. Phys. Chem. C 89 (1985) 1902–1905. doi: 10.1021/j100256a018

  20. [20]

      X.C. Wang, K. Maeda, A. Thomas, et al., Nat. Mater. 8 (2009) 76–80. doi: 10.1038/nmat2317

  21. [21]

      G.G. Zhang, Z.A. Lan, X.C. Wang, Chem. Sci. 8 (2017) 5261–5274. doi: 10.1039/C7SC01747B

  22. [22]

      C. Cheng, L.H. Mao, J.W. Shi, et al., J. Mater. Chem. A 9 (2021) 12299–12306. doi: 10.1039/d1ta00241d

  23. [23]

      P. Qiu, Y. An, X. Wang, et al., Chin. Chem. Lett. 34 (2023) 108246. doi: 10.1016/j.cclet.2023.108246

  24. [24]

      B. Sun, S. Lu, Y. Qian, et al. Carbon Energy 5 (2023) e305. doi: 10.1002/cey2.305

  25. [25]

      B. Fang, Z. Xing, D. Sun, et al. Adv. Powder Technol. 1 (2022) 100021. doi: 10.1016/j.apmate.2021.11.008

  26. [26]

      W. Liu, L.L. Cao, W.R. Cheng, et al., Angew. Chem. Int. Ed. 56 (2017) 9312–9317. doi: 10.1002/anie.201704358

  27. [27]

      Z.M. Pan, Y. Zheng, F.S. Guo, et al., ChemSusChem 10 (2017) 87–90. doi: 10.1002/cssc.201600850

  28. [28]

      F. Guo, W.L. Shi, C. Zhu, et al., Appl. Catal. B l 226 (2018) 412–420. doi: 10.1016/j.apcatb.2017.12.064

  29. [29]

      C. Cheng, J.W. Shi, L.Y. Wen, et al., Carbon 181 (2021) 193–203. doi: 10.1016/j.carbon.2021.05.030

  30. [30]

      F. Xue, Y.T. Si, C. Cheng, et al., Nano Energy 103 (2022) 107799. doi: 10.1016/j.nanoen.2022.107799

  31. [31]

      F. Xue, Y.T. Si, M. Wang, et al., Nano Energy 62 (2019) 823–831. doi: 10.1016/j.nanoen.2019.05.086

  32. [32]

      P.X. Qiu, C.M. Xu, N. Zhou, et al., Appl. Catal. B 221 (2018) 27–35. doi: 10.1016/j.apcatb.2017.09.010

  33. [33]

      C.M. Li, D.Q. Zhu, S.S. Cheng, et al., Chin. Chem. Lett. 33 (2022) 1141–1153. doi: 10.1016/j.cclet.2021.07.057

  34. [34]

      K. Wu, L. Mao, X.Q. Gu, et al., Chin. Chem. Lett. 33 (2022) 926–929. doi: 10.1016/j.cclet.2021.07.011

  35. [35]

      M. Kruszynska, H. Borchert, J. Parisi, et al., J. Am. Chem. Soc. 132 (2010) 15976–15986. doi: 10.1021/ja103828f

  36. [36]

      S.T. Connor, C.M. Hsu, B.D. Weil, et al., J. Am. Chem. Soc 131 (2009) 4962–4966. doi: 10.1021/ja809901u

  37. [37]

      X.B. Hu, T. Chen, Y.Q. Xu, et al., J. Lumin. 200 (2018) 189–195. doi: 10.1016/j.jlumin.2018.04.025

  38. [38]

      W.J. Zhang, D.Z. Li, Z.X. Chen, et al., Mater. Res. Bull. 46 (2011) 975–982. doi: 10.1016/j.materresbull.2011.03.026

  39. [39]

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

  40. [40]

      G. Liu, S. Yan, L. Shi, et al., Front. Chem. 7 (2019) 639. doi: 10.3389/fchem.2019.00639

  41. [41]

      H.J. Kong, D.H. Won, J. Kim, et al., Chem. Mater. 28 (2016) 1318–1324. doi: 10.1021/acs.chemmater.5b04178

  42. [42]

      J.W. Fu, Q.L. Xu, J.X. Low, et al., Appl. Catal. B 243 (2019) 556–565. doi: 10.1016/j.apcatb.2018.11.011

  43. [43]

      Y.Z. Hong, Y.H. Jiang, C.S. Li, et al., Appl. Catal. B 180 (2016) 663–673. doi: 10.1016/j.apcatb.2015.06.057

  44. [44]

      S.N. Cui, Z.Q. Ye, C. Qian, et al., J. Mater. Sci. Mater. Electron. 29 (2018) 15138–15146. doi: 10.1007/s10854-018-9655-2

  45. [45]

      X.Q. Li, Y.M. Shi, P.H. Chen, et al., J. Chem. Technol. Biotechnol. 94 (2019) 3713–3724. doi: 10.1002/jctb.6178

  46. [46]

      Q. Tian, X.K. Zeng, C. Zhao, et al., Adv. Funct. Mater. 33 (2023) 2213173. doi: 10.1002/adfm.202213173

  47. [47]

      Q. Zhu, J. Wu, M. Du, et al., Photomat (2023) 1–20.

  48. [48]

      Y.J. Fu, C.A. Liu, M.L. Zhang, et al., Adv. Energy Mater. 8 (2018) 1802525. doi: 10.1002/aenm.201802525

  49. [49]

      J. Liu, N.Y. Liu, H. Li, et al., Nanoscale 8 (2016) 11956–11961. doi: 10.1039/C6NR02437H

  50. [50]

      B. Yan, Z.H. Chen, Y.X. Xu, Chem. Asian J. 15 (2020) 2329–2340. doi: 10.1002/asia.202000253

  51. [51]

      J. Wang, G.H. Wang, X. Wang, et al., Carbon 149 (2019) 618–626. doi: 10.1016/j.carbon.2019.04.088

  52. [52]

      X.X. Li, X.C. Liu, C. Liu, et al., Tungsten 5 (2023): 100–108. doi: 10.1007/s42864-022-00144-7

  53. [53]

      K.S.W. Sing, D.H. Everett, R.A.W. Haul, et al., Pure Appl. Chem. 57 (1985) 603–619. doi: 10.1351/pac198557040603

  54. [54]

      H.L. Jiang, W.B. Zhang, P.H. Chen, et al., J. Mater. Chem. A 4 (2016) 11897–11907. doi: 10.1039/C6TA04885D

  55. [55]

      P.H. Chen, S.Y. Li, X.Q. Tang, et al., J. Environ. Chem. Eng. 10 (2022) 107998. doi: 10.1016/j.jece.2022.107998

  56. [56]

      F.X. Dai, Y. Wang, X.R. Zhou, et al., Appl. Surf. Sci. 517 (2020) 146161. doi: 10.1016/j.apsusc.2020.146161

• 

  Figure 1  SEM micrographs of g-C₃N₄ (a), AIS (b), and 70-GAIS (c, d). (e) TEM and (f) HRTEM micrographs of 70-GAIS. t₁, t₂, t₃ and t₄ display the inverse fast Fourier transform (IFFT) corresponding to the related regions.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  XRD patterns (a), XPS spectra (b), and C 1s high-resolution XPS spectra (c) of g-C₃N₄, AIS, and 70-GAIS. (d) N 1s high-resolution XPS spectra of g-C₃N₄ and 70-GAIS.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Typical time course of H₂ evolution from water splitting over 70-GAIS. (b) Photocatalytic water splitting performance of various photocatalysts under visible light in 24 h.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Steady-state photoluminescence (PL) spectra (a), time-resolved transient fluorescence decay curves (b), photocurrent responses (c), and Nyquist plot from an electrochemical impedance spectroscopy study (d) of g-C₃N₄, AIS, and 70-GAIS.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  HRTEM micrographs of reduction sites (a, b) and oxidation sites (c, d) over 70-GAIS.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) UV–vis diffuse reflectance spectra, and (b) Tauc plots of g-C₃N₄, AIS, and 70-GAIS, where.A refers to the absorbance. (c) Mott-Schottky plots of g-C₃N₄ and AIS. (d) A simplified band structure diagram of g-C₃N₄/AIS photocatalyst and proposed photocatalytic mechanism for simultaneous production of H₂ and H₂O₂ from pure water splitting.

  下载: 全尺寸图片 幻灯片
•