Sulfur and carbon co-doped g-C₃N₄ microtubes with enhanced photocatalytic H₂ production activity

Abstract

Metal-free graphitic carbon nitride (g-C₃N₄) has captured significant attention as a low-cost and efficient hydrogen production photocatalyst through. Effectively regulating the microstructure and accelerating the separation of photogenerated carriers remain crucial strategies for promoting the photocatalytic performance of this material. Herein, a novel sulfur–carbon co-doped g-C₃N₄ (SCCN) hierarchical microtubules filled with abundant nanosheets inside by thermal polymerization is reported. Numerous nanosheets create abundant pores and cavities inside the SCCN microtubes, thereby increasing the specific surface area of g-C₃N₄ and providing sufficient reactant attachment sites. Besides, the hierarchical structure of SCCN microtubules strengthens the reflection and scattering of light, and the utilization of visible light is favorably affected. More importantly, co-doping S and C has greatly improved the photocatalytic performance of graphitic carbon nitride, optimized the band gap structure and enhanced the photogenerated carrier splitting. Consequently, the SCCN exhibits a remarkable photocatalytic H₂ evolution rate of 4868 μmol/(g·h). This work demonstrates the potential of multi-nonmetal doped g-C₃N₄ as the ideal photocatalyst for H₂ evolution.

Graphical abstract

Keywords

carbon nitride / photocatalysis / hydrogen production

1 Introduction

Considering the increasing energy crisis, numerous efforts have been made for exploiting sustainable and promising clean energy sources to supersede limited nonrenewable fossil fuels. Application of semiconductor photocatalysts to boost the photo-splitting of water is an attractive route for converting endless solar energy into convenient hydrogen fuel [1–3]. With excellent physiochemical properties, such as reasonable band gap energy, thermal/chemical stability, and low cost, non-metal polymeric graphitic carbon nitride (g-C₃N₄) has received significant interest [4–6]. However, drawbacks such as the poor surface area, low density of active sites, and the rapid recombination of the electron-hole pairs still limit the photocatalytic activity of bulk carbon nitride. There have been various efforts to boost the photocatalytic performance of carbon nitride (g-C₃N₄), including nanostructure synthesis, elemental doping, heterojunction construction, and defect engineering [7–12].

Unique nanostructures including nanosheets, nanorods, and microspheres, are capable of enhancing the photocatalytic activity of g-C₃N₄ [13–16]. 2D nanosheets, in particular, exhibit a high specific area and favorable photocatalytic efficiency [17,18]. However, the serious agglomeration of carbon nitride nanosheets, due to their high surface energy, has limited their application in most photocatalytic reactions [19,20]. The hierarchical structure has been extensively researched as an effective approach to address the challenges in photocatalysis [21]. Fu et al. [22] for instance, reported O-doped g-C₃N₄ with hierarchical porous nanotube structure and applied them as the photocatalysts in the photocatalytic reduction of CO₂. Their study demonstrated that stable hierarchical structures offer significant benefits by preventing the aggregation of carbon nitride nanostructures. The open-up nanosheets with defined edges provide high specific surface areas and an abundance of reactive sites. The numerous porous structures facilitate the transportation of reactant molecules and enhance light absorption of the catalysts through light-scattering.

Apart from designing nanoscale structures, non-metal heteroatom doping is a reliable method for further enhancing the functionality of carbon nitride. Zhang et al. [23] showed that P doping can improve visible-light absorption and electronic properties, leading to a 5-fold increase in photocurrent response compared to pure g-C₃N₄. Additionally, Dong et al. [24] revealed the fabrication of an innovative C self-doped g-C₃N₄ catalyst using absolute ethanol pretreated melamine. In this novel approach, the homogeneous replacement of C atoms in the lattice significantly enhanced the absorbance and electron transfer in visible-light. Wang et al. [25] reported a highly crystalline S-doped g-C₃N₄ with rectangular rod-like morphology, which exhibited an excellent H₂ evolution performance. Moreover, compared with mono-element doping, binary co-doping of different elements into g-C₃N₄ can further optimize the band gap structure and construct photogenerated charge transport channels, improving the photocatalytic performance as a result [26–29]. Through thermal condensation of sulfur, melamine and hexachlorotriphosphazene, Hu et al. [30] successfully synthesized phosphorus and sulfur co-doped g-C₃N₄. The presence of phosphorus and sulfur in the g-C₃N₄ allowed for their occupation of C vacancies, which led to the formation of numerous active sites within the material. As a result, the synthesized phosphorus and sulfur co-doped g-C₃N₄ displayed superior photocatalytic degradation properties toward methylene blue compared to non-doped g-C₃N₄.

Based on the above background, the multiple superiorities of hierarchical structures in combination with elemental co-doping were expected to optimize the photocatalytic properties of g-C₃N₄. Herein, the S, C co-doped g-C₃N₄ (SCCN) microtubes with hierarchical structure via the bottom-to-top strategy is reported. The as-prepared samples were synthesized by the thermal polymerization method and applied in hydrogen production reactions under visible-light irradiation. In-depth characterizations were performed to understand the crystal structure, chemical composition, morphology, and photoelectrochemical behavior of SCCN. The SCCN exhibited a remarkably improved photocatalytic activity towards H₂ production, particularly when compared to sulfur mono-doped g-C₃N₄ (SCN). Finally, a possible mechanism was presented to explicate the boosted photocatalytic performance of SCCN.

2 Materials and methods

2.1 Chemicals and materials

The chemicals and materials involved in the present work are melamine (C₃H₆N₆, Sinopharm Chemical Reagent), tri-thiocyanuric acid (C₂H₆N₃S₃, Sinopharm Chemical Reagent), 2,4,6-triaminopyrimidine (C₄H₇N₅, Aladdin), ethylene glycol (C₂H₆O₂, Sinopharm Chemical Reagent), ethanol (C₂H₅OH, Sinopharm Chemical Reagent), sodium sulfate (N₂SO₄, Sinopharm Chemical Reagent), nafion (Aladdin, 5 wt.% (mass fraction)), triethanolamine (C₆H₁₅NO₃, Sinopharm Chemical Reagent, 10 vol.% (volume fraction)), and chloroplatinic acid (H₂PtCl₆·xH₂O, Aladdin).

2.1.1 Preparation of sulfur-doped g-C₃N₄ (SCN)

The synthesis of this compound demands a modified protocol derived from literature. The preparation of the precursor was started by dissolving 2.0 g melamine into 70 mL of ethylene glycol at 100 °C. Meanwhile, 2.8 g of trithiocyanuric acid was dispersed into 70 mL of deionized water. Next, the melamine supramolecular aggregates (MSA) was formed by gradually introducing the trithiocyanuric acid solution into the aqueous solution of melamine. After 4 h of stirring, the sample was filtered and the solid obtained was dried under vacuum for 12 h. The SCN was finally obtained with a calcination process which involved calcining precisely 2.0 g of supramolecular precursor at 550 °C for 4 h.

2.1.2 Preparation of sulfur and carbon co-doped g-C₃N₄ (SCCN)

To synthesize the SCCN, 50 mL of ethylene glycol and 2.8 g of trithiocyanuric acid were dissolved at 100 °C. Concurrently, an equimolar amount of melamine was dissolved into 50 mL of deionized water at 100 °C with a specific proportion of 2,4,6-triaminopyrimidine. The solution was completely dissolved after 30 min of stirring, before the addition of the trithiocyanuric acid solution dropwise. This resulted in the formation of melamine and 2,4,6-triaminopyrimidine supramolecular aggregates (MSTA) with a primrose yellow appearance. The supramolecular precursor was filtered to obtain the solid sample, which was then vacuum-dried for 12 h at 60 °C after washing with methanol. The dried precursor was meticulously heated to 550 °C, at a heating rate of 3 °C/min, and then calcined for 4 h. The resultant powder was then collected and given the designation SCCN. The obtained samples were labeled as 0.03SCCN, 0.05SCCN, 0.10SCCN, and 0.15SCCN based on the amount of 2,4,6-triaminopyrimidine used.

2.2 Characterization

The X-ray diffraction (XRD) analysis using a Bruker D8-Discover diffractometer revealed the crystal phases of SCCN. The Fourier transform infrared (FT-IR) spectroscopy was analyzed by Thermo Nicolet 5700 spectrometer. Both transmission electron microscopy (TEM, Tecnai G2 F20) and scanning electron microscopy (SEM, FEI inspect F50) were used to analyze the morphology and microstructure of the SCCN samples. Surface properties were characterized using X-ray photoelectron spectroscopy (XPS) with Al-Kα X-rays (Thermo ESCALAB 250XI). The Brunauer-Emmett-Teller (BET) method, which characterizes the surface area of a material, was employed by Micromeritics ASAP 2460. Utilizing the Micromeritics ASAP 2020 HD88 nitrogen-adsorption equipment, measurements of Nitrogen adsorption-desorption were also performed. In addition, UV-Vis diffuse reflectance spectra (DRS, Shimadzu UV-2600 spectrophotometer, BaSO₄) were also utilized to evaluate the optical properties of SCCN. The photoluminescence (PL) spectra were measured by a Horiba Fluoromax-4 spectrofluorometer, with 365 nm as the excitation wavelength. Time-resolved PL spectra were recorded on an Edinburgh FLS-1000 spectrophotometer.

2.3 Photoelectrochemical measurements

The photoelectrochemical tests were assessed via an electrochemical workstation (CHI660E) with 0.1 mol/L of Na₂SO₄ solution as electrolyte. The Ag/AgCl and Pt foil worked as reference electrode and counter electrode. 5 mg of photocatalysts and 20 µL of Nafion ethanol solution (5 wt.%) were dispersed throughout 1 mL of ethanol to form a solution suspension which was coated on the surface of fluorine-doped tin oxide (FTO) to fabricate the work electrode. The light source being used was a 300 W Xenon lamp (CELL-HXF 300, λ > 420 nm). The electrochemical impedance spectroscopy (EIS) was performed in a frequency range from 10⁵ to 0.1 Hz. The flat-band potential was analyzed based on the Mott-Schottky (M-S) plots at a frequency of 100 Hz.

2.4 Photocatalytic H₂ production evaluation

The photocatalytic H₂ generation experiment was conducted in the vessel at nitrogen atmosphere. The co-catalyst (3 wt.% Pt) was deposited on the surface of SCN (50 mg) and SCCN (50 mg) samples. The photocatalysts obtained were mixed with 100 mL of triethanolamine (TEOA, 10 vol.%). H₂ was produced under simulated visible light irradiation (λ > 420 nm), and then studied after detection by gas chromatograph (GC-9860, 5 Å molecular sieve-packed column) in combination with a thermal conductivity detector (TCD).

3 Results and discussion

3.1 Structural characterization

Fig.1(b) shows XRD patterns of SCN and SCCN samples. Two diffraction peaks are noticed at 13.1° and 27.4°, which distinctly correspond to the characteristic graphite phase g-C₃N₄ structure [31]. The dominant peak at 27.4° (002) belongs to the interstratified stacking of g-C₃N₄ [32]. Meanwhile, the minor peak observed at 13.1° is attributed to the repeating in-plane heptazine unit along the (100) plane [33]. The SCCN sample exhibits a weaker diffraction peak intensity compared to SCN, which decreases further with increasing concentrations of 2,4,6-triaminopyrimidine [34]. These observations confirm the distortion of the SCCN structure, previously reported for carbon-doped g-C₃N₄ [34]. Moreover, the migration of the diffraction peak for the (002) plane from 27.48° to 27.34° in the SCCN demonstrates the increase in the interlayer distance, further confirming the distortion of the SCCN structure [34,35].

The FT-IR spectra of SCN and SCCN are shown in Fig.1(c). The apparent peak at 809 cm⁻¹ indicates the bending oscillation of heptazine ring. Two broad peaks ranging from 2700 to 3600 cm⁻¹ indicates the residue amine terminal groups, including −NH and −NH₂ groups [35]. For SCCN samples, the intensive peaks from 900 to 1800 cm⁻¹, which explicitly correspond to the tensile vibration of C−N and C=N bonds inside the heptazine ring, shift to higher wavenumbers. This indicates a higher vibration frequency of the C−N bond, consistent with Ref. [31]. The introduction of extra carbon atoms contributes to changing the charge density on the conjugated heterocyclic ring, ultimately leading to the increase in the energy of C−N bond [35]. The analyses suggest that SCCN samples have been successfully synthesized, while still retaining their crystal structure and bond composition.

The morphologies of as-prepared samples were investigated by SEM and TEM, as depicted in Fig.2. Noticeably, the supramolecular precursor MSA exhibits a regular rod-like structure (Fig.2(a)). Upon calcination, the appearance of SCN remains to be a microtube structure due to hydrogen-bonding interactions in the precursor (Fig.2(b)). The external morphology of 0.10SCCN exhibits a microtube structure which is similar to that of SCN (Fig.2(c)). The addition of 2,4,6-triaminopyrimidine to the precursor has seldom destroyed the final external structure of g-C₃N₄. To examine the internal structure of the samples, the cross-section of the broken 0.10SCCN microtubes are chosen for further observation (Fig.2(d)). SCN exhibits microtubule structures filled with g-C₃N₄ fragments, which are confirmed through TEM characterization (Fig.2(e)). The interior of the SCN microtubes contain numerous g-C₃N₄ fragments which are further proved to be SCN nanosheets (Fig.2(f)). TEM and SEM analyses of 0.10SCCN reveal the same microtubule structure (Fig.2(i)). Normally, the g-C₃N₄ obtained by calcining self-assembled supramolecules exhibits a hollow tubular structure due to the high-temperature sublimation phenomenon of melamine. The abundant nanosheets inside SCN and SCCN microtubes are attributed to the internal condensation of resided melamine during the thermal condensation process. It is inferred that the hierarchical microtubes are served to shorten the migration distance of photo-generated charges while preventing unnecessary recombination.

The characteristics of pores in SCN and 0.10SCCN are shown in Fig.3, which display highly similar isotherms and corresponding pore size distributions, characterized by type IV isotherms with H3 hysteresis loops [36]. Moreover, the N₂ adsorption-desorption isotherms of the samples show rapid increases at high P/P₀ values, indicating mesopores. Fig.3 shows that the N₂ adsorption–desorption isotherms and the pore-size distribution curves of both SCN and SCCN samples exhibit a close similarity, indicating that carbon doping has minimal variation on the pore structure of g-C₃N₄. Besides, the specific surface area of SCN and 0.10SCCN are 84.97 and 77.99 m²/cm³, much larger than that of bulk g-C₃N₄ (S_BET = 8.38 m²/g, Fig. S6). The primary pore size ranges for both SCN and 0.10SCCN are within 2‒110 nm, primarily dominated by mesopores [37]. The pore volume of SCN and 0.10SCCN are determined to be 0.49 and 0.50 cm³/g, respectively. Therefore, the internal nanosheets present complex pores and cavities in the microtubes (SCN and 0.10SCCN) provide larger surface areas with abundant reactive sites, accelerating the mass transfer kinetics of the catalytic process [38].

XPS is performed to analyze the chemical states of SCN and 0.10SCCN. Fig.4(a) illustrates the predominant composition of carbon, nitrogen, and oxygen with minimal sulfur observed. Fig.4(b) displays the C 1s spectra, which exhibit three peaks at 284.6, 286.4, and 287.9 eV, representing graphitic carbon, sp²-bonded carbon (C−NH_x), and sp²-bonded carbon (N−C=N), respectively. For the N 1s spectra in Fig.4(c), the crucial peaks at 398.3 and 399.9 eV belong to sp²-bonded nitrogen (C−N=C) and tertiary N (N−C3) while the other two peaks relate to C−NH_x, and π-excitation [39]. The binding energies of the nitrogen and carbon atoms in SCCN are lowered by 0.1‒0.2 eV due to carbon doping. This enhanced electron density in aromatic heterocycles has led to the increase in the relative electron density of carbon with neighboring nitrogen atoms [34]. Fig.4(d) presents the high-resolution S 2p spectra for both SCN and 0.10SCCN. In the S 2p spectra, three distinct binding energies of 163.6, 165.0, and 168.9 eV can be expected, corresponding to 2p_3/2 and 2p_1/2 of low-valent sulfur and high-valent sulfur, respectively [40]. The binding energy of the low-valent sulfur is similar to CS₂, suggesting that the sulfur atoms have replaced the nitrogen atom in the heptazine ring to form C−S=C bonds [41]. The presence of high-valent sulfur (S⁴⁺ and S⁶⁺ species) is explained by the oxidation of C−S=C bonds into C−SO_x−C in the calcination process [42,43].

Tab.1 exhibits the surface element composition from XPS survey spectra and indicates that 0.10SCCN comprises a more significant proportion of carbon and sulfur elements than SCN. These findings suggest a successful synthesis of sulfur and carbon co-doped material through supramolecular precursor calcination. Tab.2 presents the peak position of the low-valence sulfur element in SCCN. A comparative analysis of SCN shows minimal changes in peak position, yet the peak area is remarkably higher in SCCN. Interestingly, the peak position of the high valence sulfur element has shifted to the low binding energy, and its peak area is lower than that of SCN. These results offer compelling evidence that carbon doping enhances the doping amount of sulfur.

3.2 Optical and photoelectrochemical properties

Optimizing the band gap structure of semiconductor catalysts is necessary for effective utilization of visible light [44]. The UV-Vis diffuse reflection spectra of SCN and SCCN are displayed in Fig.5(a). SCN displays an obvious absorption tailing due to sulfur doping. Moreover, numerous holes and cavities in the microtube morphology formed by the nanosheets cause multiple light refractions and reflections in the microtubes, which significantly enhance the absorption and utilization of visible light [22]. Notably, the optical absorption ability of SCCN is greatly enhanced with increasing 2,4,6-triaminopyrimidine precursor content. Co-doping sulfur and carbon effectively broadens the visible light absorption range of SCCN and increases its utilization range. In Fig.5(b), the Tauc method is used to calculate the band gap widths of photocatalysts, and the band gap widths of SCN and SCCN photocatalysts gradually decrease while 2,4,6-triaminopyrimidine content increase [45]. Compared with SCN, the minimum band gap of SCCN is reduced by 0.70 eV.

The M-S plots of SCN and SCCN photocatalysts, shown in Fig.5(c), reveal flat-band potentials of −0.54, −0.44, −0.40, −0.37, and −0.33 V, indicating that the flat-band potential lowers with increased carbon and sulfur doping. Another obvious discovery is that all samples are n-type semiconductors according the positive slope in the plots. Additionally, UV photoelectron spectroscopy (UPS) were employed to determine the Fermi levels of SCN and 0.10SCCN samples. According to Fig. S1, the Fermi energies of SCN and 0.10SCCN photocatalysts are calculated to be −0.14 and −0.24 eV, classifying them as n-type semiconductors, which further confirms the results of M-S plots [46,47]. Generally, the n-type semiconductor conduction band (CB) potential is approximately 0.1‒0.3 eV less than the flat-band potential, indicating a slight variation in conduction band potential of this semiconductor photocatalysts [48]. The XPS valence band spectra of SCN and SCCN photocatalysts in Fig.5(d) show valence band potentials of 1.84, 1.75, 1.64, 1.58, and 1.49 eV, respectively. Calculated conduction band potentials are −0.79, −0.60, −0.54, −0.49, and −0.44 eV, respectively, consistent with the flat band potential variation provided by the Mott-Schottky plots. In conclusion, co-doping sulfur and carbon effectively regulate g-C₃N₄ band gaps, improving visible light utilization.

In addition to excellent visible-light absorption, effective charge carrier separation is also a requirement for good photocatalysts. The PL spectrum characterizes the radioluminescence recombination of charge carriers. As shown in Fig.6(a), the SCN photocatalyst exhibits the highest fluorescence intensity, indicating that the charge carriers are severely recombined [49]. The fluorescence intensity of sulfur and carbon co-doped SCCN photocatalyst decreases significantly with increased 2,4,6-triaminopyrimidine content, indicating that greater carbon and sulfur doping improves the separation of photo-generated charge carriers. Notably, 0.10SCCN samples exhibit the poorest fluorescence intensity, reflecting the greatest charge separation efficiency. The SCN and SCCN fluorescence peaks also display conspicuous redshifts that are consistent with the results of UV-Vis diffuse reflectance spectroscopy. To further probe the effect of sulfur-carbon co-doping on the lifetime of photoexcited carriers, the time-resolved PL spectra (Fig. S2) of SCN and 0.10SCCN samples were fitted to a second-order decay function. The fitting results are shown in Table S1. Evidently, the 0.10SCCN sample presents a prolonged average (τ_ave = 4.81 ns) emission lifetime in comparison with the SCN (τ_ave = 3.29 ns), which confirms the efficient separation of photoinduced carriers which derived from the co-doping of C and S elements [50]. Photoelectrochemical tests are another means of evaluating charge separation ability in photocatalysts. The transient photocurrent response curve (Fig.6(b)) demonstrates a considerable visible light response in the photocatalysts. By adding 2,4,6-triaminopyrimidine, SCCN photocurrent density increases, reaching its maximum of 0.10SCCN. The results of this test are consistent with PL spectra data. The minor decrease in 0.15SCCN photo-current density is ascribed to the formation of composite centers arising from excessive element doping [51]. Fig.6(c) presents the EIS plots of modified g-C₃N₄, with the charge transfer resistance of SCCN decreasing at a lower 2,4,6-triaminopyrimidine content and then increasing at a higher content. This trend is consistent with the transient photocurrent test (Fig.6(b)). In conclusion, sulfur and carbon co-doping greatly promotes charge transfer, reduces charge transfer resistance, and provides sufficient photogenerated charge for photocatalytic reactions.

3.3 Photocatalytic H₂ production

Photocatalytic H₂ production was performed under visible-light irradiation (see Fig.7(a)), which depicts the results of the H₂ production of photocatalysts prepared with SCN and SCCN. The absence of H₂ production in the reaction system without photocatalysts indicates the photoactivity of the process. The H₂ production of all photocatalysts show a linear increase depending on the stability of the photocatalyst. The co-doping of sulfur and carbon in SCCN samples led to a higher H₂ production than SCN samples, demonstrating that the co-doping process can improve the photocatalytic performance. Fig.7(b) displays the average H₂ production rate of the as-prepared samples under the same conditions. The sample of 0.10 SCCN demonstrates the highest H₂ production rate of 4868 μmol/(g·h), much better than that of g-C₃N₄ (29.6 times, Fig. S8) and SCN (2.3 times). Further, the apparent quantum efficiency (AQE) of the SCCN composite was measured as 3.93% at the wavelength of 420 nm. Consistent with the results of BET analysis, the surface areas of the 0.10SCCN and SCN samples are similar, indicating that the increase in photocatalytic H₂ production attributes to the sulfur and carbon co-doping with strengthened light absorption capacity and accelerated photogenerated charge separation rate. While increasing the amount of carbon doping can promote the visible-light absorption of SCCN photocatalysts, excessive doping creates aggregation and recombination of photogenerated charges which reduces the quantity of electrons and holes involved in the interfacial reactions [51]. Consequently, the H₂ production rate of 0.15SCCN was lower than that of 0.10SCCN. The stability of the photocatalyst is crucial for the potential application [52,53]. As shown in Fig.7(c), the hydrogen production of 0.10 SCCN remains stable in eight rounds of experiments. Furthermore, the hardly changed XRD pattern of the as-prepared sample after the cycling tests, indicates the feasible photostability and physicochemical stability (Fig. S3(a)).

3.4 Mechanism investigation

According to the comprehensive analysis of the above results, a potential mechanism for the boosted photocatalytic performance of SCCN photocatalyst for visible-light-driven H₂ evolution is proposed and described clearly in Fig.8. The hierarchical structure of the nanosheet-filled SCCN microtubes with abundant pores and cavities enables the incident light to undergo multiple refraction and reflection inside the microtubes, resulting in a higher efficiency of light absorption and utilization. Moreover, this remarkable multilayer structure enlarges the specific surface area of carbon nitride and provides accessible sites for Pt deposition. With S, C co-doping, the SCCN photocatalyst possesses the narrower band gap width and wider visible-light absorption, which makes more electrons excited from the valence band (VB) to the conduction band (CB) and then overflow to the well-dispersed Pt surface, leaving holes in the valence band. In this stage, it is crucial to prevent the recombination of electron-hole pairs and provide sufficient reactive sites. Obviously, the SCCN photocatalyst achieves effective separation of photogenerated carriers by rapid transfer of photogenerated electrons to active sites, which can be confirmed by PL spectroscopy. Therefore, under visible-light irradiation, SCCN can absorb more photon energy and form multiple photogenerated e⁻–h⁺ pairs. Subsequently, the photo-generated electrons are rapidly consumed via proton reduction to H₂ on the Pt surface, while the holes are oxidized by TEOA sacrificial agent. Consequently, the effective separation of photogenerated electron-hole pairs leads to a notable enhancement of the photocatalytic H₂ production performance.

4 Conclusions

In summary, this work successfully prepared SCCN photocatalyst with hierarchical structure and investigated its photocatalytic properties in H₂ production driven by visible-light. The introduction of 2,4,6-triaminopyrimidine during the thermal polymerization are found to be significant in forming the microtubes structure and muti-element doping carbon nitride. The SCCN samples possess microtubes structures similar to SCN, which provide them with large surface areas and rich reactive sites. In terms of photoelectrochemical behavior, SCCN photocatalysts exhibit a wider visible-light absorption, a shorter carrier migration distance, and a lower carrier complexation rate. Under visible-light irradiation, 0.10SCCN as the optimal co-doping amount demonstrates an outstanding photocatalytic H₂ production rate of 4868 μmol/(g·h). This superior rate is 2.3 times higher than that of SCN photocatalyst under identical conditions. This study presents a feasible strategy for synthesizing a high-efficiency, heteroatom binary-doped, hierarchical g-C₃N₄ photocatalyst.

References

[1] Sun L J, Dong H L, Xu J. . Unravelling the synergy between phase engineering and interface regulation in TiO₂/1T-rich MoSe₂ heterostructures for efficient photocatalytic hydrogen evolution. ACS Sustainable Chemistry & Engineering, 2023, 11(21): 8009–8019

[2] Jourshabani M, Asrami M, Lee B. Advanced functional carbon nitride by implanting semi-isolated VO₂ active sites for photocatalytic H₂ production and organic pollutant degradation. Small, 2023, 19(28): 2300147

[3] Du S W, Lin S Q, Ren K K. . Revealing the effects of transition metal doping on CoSe cocatalyst for enhancing photocatalytic H₂ production. Applied Catalysis B: Environmental, 2023, 328: 122503

[4] Wang X, Maeda K, Chen X. . Polymer semiconductors for artificial photosynthesis: Hydrogen evolution by mesoporous graphitic carbon nitride with visible light. Journal of the American Chemical Society, 2009, 131(5): 1680–1681

[5] Xiao S T, Yin R, Wu L. . Hierarchically porous few-layer carbon nitride and its high H⁺ selectivity for efficient photocatalytic seawater splitting. Nano Letters, 2023, 23(10): 4390–4398

[6] Lin F, Zhou S, Wang G H. . Electrostatic self-assembly combined with microwave hydrothermal strategy: Construction of 1D/1D carbon nanofibers/crystalline g-C₃N₄ heterojunction for boosting photocatalytic hydrogen production. Nano Energy, 2022, 99: 107432–107441

[7] Zheng Y, Lin L H, Wang B. . Graphitic carbon nitride polymers toward sustainable photoredox catalysis. Angewandte Chemie International Edition, 2015, 54(44): 12868–12884

[8] Wang X C, Chen X F, Thomas A. . Metal-containing carbon nitride compounds: A new functional organic-metal hybrid material. Advanced Materials, 2009, 21(16): 1609–1612

[9] Liu Z X, Liu Y D, Sun X B. . Construction of Z-scheme Ag/AgVO₃/carbon-rich g-C₃N₄ heterojunction for enhanced photocatalytic degradation of sulfamethiadiazole: DFT calculation and mechanism study. Chemical Engineering Journal, 2022, 433: 133604–133616

[10] Liu Y J, Tayyab M, Pei W K. . The precision defect engineering with nonmetallic element refilling strategy in g-C₃N₄ for enhanced photocatalytic hydrogen production. Small, 2023, 19(21): 2208117

[11] Shen R C, Hao L, Chen Q. . P-doped g-C₃N₄ nanosheets with highly dispersed Co_0.2Ni_1.6Fe_0.2P cocatalyst for efficient photocatalytic hydrogen evolution. Acta Physico-Chimica Sinica, 2022, 38(7): 2110014

[12] Zhang J J, Wang L X, Mitra M. . Molecular-level engineering of S-scheme heterojunction: The sitespecific role for directional charge transfer. Chinese Journal of Structural Chemistry, 2022, 41(6): 2206003–2206005

[13] Jun Y S, Lee E Z, Wang X C. . From melamine-cyanuric acid supramolecular aggregates to carbon nitride hollow spheres. Advanced Functional Materials, 2013, 23(29): 3661–3667

[14] Sun J H, Zhang J S, Zhang M W. . Bioinspired hollow semiconductor nanospheres as photosynthetic nanoparticles. Nature Communications, 2012, 3(1): 1139

[15] Cui L F, Song J L, McGuire A. . Constructing highly uniform onion-ring-like graphitic carbon nitride for efficient visible-light-driven photocatalytic hydrogen evolution. ACS Nano, 2018, 12(6): 5551–5558

[16] Shen R C, He K L, Zhang A P. . In-situ construction of metallic Ni₃C@Ni core–shell cocatalysts over g-C₃N₄ nanosheets for shell-thickness-dependent photocatalytic H₂ production. Applied Catalysis B: Environmental, 2021, 291: 120104

[17] Huang H W, Xiao K, Tian N. . Template-free precursor-surface-etching route to porous, thin g-C₃N₄ nanosheets for enhancing photocatalytic reduction and oxidation activity. Journal of Materials Chemistry. A, Materials for Energy and Sustainability, 2017, 5(33): 17452–17463

[18] Zhao Z W, Dai K, Zhang J F. . In situ preparation of Mn_0·2Cd_0·8S-diethylenetriamine/porous g-C₃N₄ S-scheme heterojunction with enhanced photocatalytic hydrogen production. Advanced Sustainable Systems, 2023, 7(1): 2100498

[19] Ou H H, Lin L H, Zheng Y. . Tri-s-triazine-based crystalline carbon nitride nanosheets for an improved hydrogen evolution. Advanced Materials, 2017, 29(22): 1700008

[20] Zhang J S, Zhang M W, Yang C. . Nanospherical carbon nitride frameworks with sharp edges accelerating charge collection and separation at a soft photocatalytic interface. Advanced Materials, 2014, 26(24): 4121–4126

[21] Sun Z Z, Dong H Z, Yuan Q. . Self-supported hierarchical crystalline carbon nitride arrays with triazine-heptazine heterojunctions for highly efficient photoredox catalysis. Chemical Engineering Journal, 2022, 435: 134865

[22] Fu J W, Zhu B C, Jiang C J. . Hierarchical porous O-doped g-C₃N₄ with enhanced photocatalytic CO₂ reduction activity. Small, 2017, 13(15): 1603938

[23] Zhang Y, Mori T, Ye J. . Phosphorus-doped carbon nitride solid: enhanced electrical conductivity and photocurrent generation. Journal of the American Chemical Society, 2010, 132(18): 6294–6295

[24] Dong G H, Zhao K, Zhang L Z. Carbon self-doping induced high electronic conductivity and photoreactivity of g-C₃N₄. Chemical Communications, 2012, 48(49): 6178–6180

[25] Wang H, Bian Y R, Hu J T. . Highly crystalline sulfur-doped carbon nitride as photocatalyst for efficient visible-light hydrogen generation. Applied Catalysis B: Environmental, 2018, 238: 592–598

[26] Chu Y C, Lin T J, Lin Y R. . Influence of P,S,O-doping on g-C₃N₄ for hydrogel formation and photocatalysis: An experimental and theoretical study. Carbon, 2020, 169: 338–348

[27] Samanta S, Yadav R, Kumar A. . Surface modified C, O co-doped polymeric g-C₃N₄ as an efficient photocatalyst for visible light assisted CO₂ reduction and H₂O₂ production. Applied Catalysis B: Environmental, 2019, 259: 118054

[28] Huang J X, Li D G, Li R B. . An efficient metal-free phosphorus and oxygen co-doped g-C₃N₄ photocatalyst with enhanced visible light photocatalytic activity for the degradation of fluoroquinolone antibiotics. Chemical Engineering Journal, 2019, 374: 242–253

[29] Zheng X S, Zhang Q X, Chen T S. . A novel synthetic carbon and oxygen doped stalactite-like g-C₃N₄ for broad-spectrum-driven indometacin degradation. Journal of Hazardous Materials, 2020, 386: 121961

[30] Hu C C, Hung W Z, Wang M S. . Phosphorus and sulfur codoped g-C₃N₄ as an efficient metal-free photocatalyst. Carbon, 2018, 127: 374–383

[31] Ye M Y, Zhao Z H, Hu Z F. . 0D/2D heterojunctions of vanadate quantum dots/graphitic carbon nitride nanosheets for enhanced visible-light-driven photocatalysis. Angewandte Chemie International Edition, 2017, 56(29): 8407–8411

[32] Wang Y G, Xia Q N, Bai X. . Carbothermal activation synthesis of 3D porous g-C₃N₄/carbon nanosheets composite with superior performance for CO₂ photoreduction. Applied Catalysis B: Environmental, 2018, 239: 196–203

[33] Liang Z Z, Bai J X, Hao L. . Photodeposition of NiS cocatalysts on g-C₃N₄ with edge grafting of 4-(1H-imidazol-2-yl) benzoic acid for highly elevated photocatalytic H₂ evolution. Advanced Sustainable Systems, 2023, 7(1): 2200143

[34] Da Silva E, Moura N, Coutinho A. . β-cyclodextrin as a precursor to holey C-doped g-C₃N₄ nanosheets for photocatalytic hydrogen generation. ChemSusChem, 2018, 11(16): 2681–2694

[35] Ho W K, Zhang Z Z, Lin W. . Copolymerization with 2,4,6-triaminopyrimidine for the rolling-up the layer structure, tunable electronic properties, and photocatalysis of g-C₃N₄. ACS Applied Materials & Interfaces, 2015, 7(9): 5497–5505

[36] Liu C, Wu K L, Meng G H. . Explore the properties and photocatalytic performance of iron-doped g-C₃N₄ nanosheets decorated with Ni₂P. Molecular Catalysis, 2017, 437: 80–88

[37] Zhu Y P, Ren T Z, Yuan Z Y. Mesoporous phosphorus-doped g-C₃N₄ nanostructured flowers with superior photocatalytic hydrogen evolution performance. ACS Applied Materials & Interfaces, 2015, 7(30): 16850–16856

[38] Shi L, Chang K, Zhang H B. . Drastic enhancement of photocatalytic activities over phosphoric acid protonated porous g-C₃N₄ nanosheets under visible light. Small, 2016, 12(32): 4431–4439

[39] Li S, Dong G, Hailili R. . Effective photocatalytic H₂O₂ production under visible light irradiation at g-C₃N₄ modulated by carbon vacancies. Applied Catalysis B: Environmental, 2016, 190: 26–35

[40] WangKLi QLiuB S, . Sulfur-doped g-C₃N₄ with enhanced photocatalytic CO₂-reduction performance. Applied Catalysis B: Environmental, 2015, 176–177: 44–52

[41] Liu T, Li Y F, Sun H J. . Asymmetric structure awakened n-π* electron transition in sulfur and selenium co-doped g-C₃N₄ with efficient photocatalytic performance. Chinese Journal of Structural Chemistry, 2022, 41(6): 2206055–2206061

[42] Ge F Y, Huang S Q, Yan J. . Sulfur promoted n-π* electron transitions in thiophene-doped g-C₃N₄ for enhanced photocatalytic activity. Chinese Journal of Catalysis, 2021, 42(3): 450–459

[43] Liu Y, Wang H, Yuan X Z. . Roles of sulfur-edge sites, metal-edge sites, terrace sites, and defects in metal sulfides for photocatalysis. Chem Catalysis, 2021, 1(1): 44–68

[44] Lu X Y, Xie J, Chen X B. . Engineering MP_X (M = Fe, Co or Ni) interface electron transfer channels for boosting photocatalytic H₂ evolution over g-C₃N₄/MoS₂ layered heterojunctions. Applied Catalysis B: Environmental, 2019, 252: 250–259

[45] Wang Y B, Zhao X, Cao D. . Peroxymonosulfate enhanced visible light photocatalytic degradation bisphenol A by single-atom dispersed Ag mesoporous g-C₃N₄ hybrid. Applied Catalysis B: Environmental, 2017, 211: 79–88

[46] Xu H T, Xiao R, Huang J R. . In situ construction of protonated g-C₃N₄/Ti₃C₂ MXene Schottky heterojunctions for efficient photocatalytic hydrogen production. Chinese Journal of Catalysis, 2021, 42(1): 107–114

[47] Zong X P, Miao X, Hua S X. . Structure defects assisted photocatalytic H₂ production for polythiophene nanofibers. Applied Catalysis B: Environmental, 2017, 211: 98–105

[48] Tian N, Huang H W, Wang S B. . Facet-charge-induced coupling dependent interfacial photocharge separation: A case of BiOI/g-C₃N₄ p-n junction. Applied Catalysis B: Environmental, 2020, 267: 118697

[49] Guo Q Y, Zhang Y H, Zhang H S. . 3D foam strutted graphene carbon nitride with highly stable optoelectronic properties. Advanced Functional Materials, 2017, 27(42): 1703711

[50] Sahoo R C, Lu H J, Garg D. . Bandgap engineered g-C₃N₄ and its graphene composites for stable photoreduction of CO₂ to methanol. Carbon, 2022, 192: 101–108

[51] Zhang J G, Zhu Q H, Ma Y F. . Photo-generated charges escape from P⁺ center through the chemical bridges between P-doped g-C₃N₄ and Ru_xP nanoparticles to enhance the photocatalytic hydrogen evolution. Catalysis Today, 2021, 380: 223–229

[52] Wang H, Wu Y, Feng M B. . Visible-light-driven removal of tetracycline antibiotics and reclamation of hydrogen energy from natural water matrices and wastewater by polymeric carbon nitride foam. Water Research, 2018, 144: 215–225

[53] Chen J Y, Qin C C, Mou Y. . Linker regulation of iron-based MOFs for highly effective Fenton-like degradation of refractory organic contaminants. Chemical Engineering Journal, 2023, 459: 141588

RIGHTS & PERMISSIONS

Higher Education Press 2023