Porous graphitic carbon nitride with high concentration of oxygen promotes photocatalytic H₂ evolution

Abstract

Porous structure design and the content regulation of heteroelements have been proved to be effective strategies to boost photocatalytic H₂ generation activity of graphitic carbon nitride (g-C₃N₄) based photocatalyst. Herein, a series of porous graphitic carbon nitride with high concentration of oxygen (g-C₃N₄–O) photocatalysts were synthesized via in situ polymerization process using colloidal SiO₂ as oxygen source. The content of oxygen within the g-C₃N₄–O photocatalysts could be tuned by adjusting the amount of added colloidal SiO₂ during the preparation procedure. The introduced oxygen replaced two-coordinated nitrogen atom, influencing band edge position and localized electron distribution, thereby enhancing visible light harvesting and photoelectric conversion. As a result, the g-C₃N₄–O photocatalyst with an optimal oxygen content (8.39 wt%) showed 10.5 fold enhancement in H₂ evolution rate compared to that of bulk g-C₃N₄, attributed to the porous structure and high concentration of incorporated oxygen.

---

Introduction

Limited fossil energy and serious environmental issues impel people to develop green and sustainable energy.¹ Hydrogen energy, as one of the most promising types of clean energy, could be directly prepared via water splitting driven by light or electricity and the combustion products are non-polluting to the environment.^2,3 Realizing the industrial production and application of hydrogen energy is a major demand for sustainable social development.⁴ Water splitting under solar light irradiation by semiconductor photocatalysis technology appears to be an ideal strategy to generate H₂ energy.^5,6 The yield and industrialization prospects are determined by the visible light utilization ability and photoelectric conversion efficiency of the employed semiconductor photocatalysts.^7,8 Considering the problem of metal percolation of metal-based catalysts, graphitic carbon nitride (g-C₃N₄) is recognized as an appealing photocatalyst attributed to its metal-free properties, visible light response and sufficient reduction capacity.^9,10 Meanwhile, its excellent photo-stability and physicochemical stability are able to preserve long-term photocatalytic activity to satisfy industrial application requirements.¹¹ However, the inherent π-deficient conjugated characteristic of bulk g-C₃N₄ implies low charge density and carrier mobility, which are not favorable for the charge conduction.¹² The localization of π conjugated system leads to difficult spatial separation of photoexcited electrons and holes for g-C₃N₄.¹³ Meanwhile, bulk g-C₃N₄ suffers from low surface areas and limited exposed active sites.¹⁴ Hence, without modification, bulk g-C₃N₄ presentes poor photoelectric conversion efficiency and low photocatalytic H₂ production rate.

Incorporation of heteroelements to modulate the composition of g-C₃N₄ is effective to optimize its electronic properties and photocatalytic activity.^15,16 The introduced heteroelements could disturb local electron distribution of π conjugated system and expand delocalization to promote spatial separation of photoexcited charge carriers.^17,18 Generally, the selected heteroelements possess more valence electrons in comparison with C/N atoms, such as P, Cl, Br and I etc.^19–22 The introduced extra electrons would increase electron density of π conjugated system and thus elevate the conductivity, which was beneficial to improve transportation efficiency of photoinduced carriers.²³ For instance, after incorporation with iodine, the photocurrent and photocatalytic H₂ generation rate of g-C₃N₄–I significantly increased.²⁴ Another factor to consider is electronegativity, which could give much promotional effect on the electron distribution perturbation of π conjugated system.²⁵ O and F with high electronegativity are preferred candidates to optimize the electronic properties of g-C₃N₄.^26,27 Fluorination has been used to introduce F into g-C₃N₄ substrate by direct adding of HF/NH₄F into the typical condensation procedure. However, the process does not meet the requirements of environmental protection due to the employment of harmful etchants.²⁸ Besides, the overlarge atomic radius of incorporated heteroelements is thermodynamically and geometrically difficult to substitute the C/N atom in the heptazine unites to form stable structure, which restrain the long-term stability and photocatalytic performance enhancement. Thereby, O with high electronegativity and comparable atom radius compared to C/N atoms is a recommended choice to modify g-C₃N₄.

In addition to heteroelement incorporation into g-C₃N₄ matrix, increase of surface area is desirable to elevate the overall photocatalytic performance. Researchers adopted surfactants, block polymers and ionic liquids as soft templates to synthesize microporous g-C₃N₄.^29–31 Mesoporous silica, silica nanoparticles and CaCO₃ nanosphere were selected as hard templates to prepared mesoporous g-C₃N₄.^32–34 The obtained porous architecture increased surface area, exposed active sites and promoted mass transfer, thus enhancing the photocatalytic performance of g-C₃N₄.³³ Therefore, simultaneously achievement of oxygen incorporation and textural pores within g-C₃N₄ is desirable to further improve the H₂ evolution performance.

In this work, we developed a facile approach to obtain porous g-C₃N₄ with high content of oxygen using colloidal SiO₂ as oxygen source. The concentration of introduced oxygen could be tuned by amount adjustment of added colloidal SiO₂ during the synthesis procedure. By substituting two-coordinated N atoms, the introduced O induced narrowed band gap and electron redistribution of HOMO/LUMO, thus contributing to extended visible light adsorption and promoted separation and transportation of photoexcited charge carriers. With respect to bulk g-C₃N₄, the obtained g-C₃N₄–O with an optimal oxygen content (8.39 wt%) exhibited 10.5 times enhancement in H₂ generation rate originating from mesoporous architecture formation and high oxygen content.

Experimental

Materials

Colloidal SiO₂ and Nafion were purchased from Sigma-Aldrich Company. Dicyandiamide (DCDA), ammonium bifluoride (NH₄·HF₂), ethanol and triethanolamine (TEOA) were purchased from Beijing Chemical Works. Chloroplatinic acid (H₂PtCl₆·6H₂O) was purchased from Aladdin Company. Sodium sulfate (Na₂SO₄) was purchased from Shanghai Macklin Biochemical Technology Co., Ltd. All the reagents used in this work were analytical grade without further purification. Indium tin oxide (ITO) glass was purchased from Kaisheng Technology Co., Ltd. Deionized water (Millipore, 18.2 MΩ cm) was used throughout all the experiments.

Synthetic procedures

Bulk g-C₃N₄ was prepared via thermal polymerization of dicyandiamide (DCDA). The porous g-C₃N₄ with high concentration of oxygen (g-C₃N₄–O) were prepared via in situ thermal copolymerization using DCDA as g-C₃N₄ precursor and colloidal SiO₂ as oxygen source. Typically, a certain amount of colloidal SiO₂ and 1 g of DCDA were added into 2 mL of water. After stirring and sufficiently mixing, the mixtures were placed in a vacuum oven at 50 °C until the samples were completely dried. Then the white powders were calcined in a muffle furnace at 520 °C for 4 h with a heating rate of 10 °C min⁻¹. The resulting yellow powders were thoroughly ground in an agate mortar. Subsequently, the powders were washed with saturated NH₄·HF₂ solutions for 48 h. The obtained precipitates were washed by water for several times until the pH of the supernatant was neutral. The resulting brown products were then dried in a vacuum oven for further use. The amount of added colloidal SiO₂ during the preparation procedure were 500, 1000 and 2000 mg, respectively. The obtained samples were donated as xSiO₂–g-C₃N₄, where the x presented the amount of added colloidal SiO₂.

Materials characterizations

The morphology of the as-prepared samples was observed by a Regulus 8100 scanning electron microscope (SEM) and a FEI Tecnai F20 transmission electron microscopy (TEM). The element mapping were conducted on an OXFORD XPLORE30 Energy Dispersive Spectrometer operating at an acceleration voltage of 15 kV and using a secondary electron detector. The contents of C, N and O elements were evaluated by an Elementar Vario EL cube elemental analyzer. The amount of residual Si within the g-C₃N₄–O matrix was measured by an Agilent 725 inductively coupled plasma-mass spectrometry. Surface areas and pore size distributions of the photocatalysts were acquired by nitrogen physisorption at 77 K on an ASAP2460 Surface Area and Porosity Analyzer. Fourier transformed infrared (FTIR) spectra were collected using an IRAffinity-1 FTIR spectrometer. Powder X-ray diffraction (XRD) patterns were measured with a D/MAX-2200/PC X-ray diffractometer with Cu Kα radiation at a scanning rate of 5° min⁻¹. The valence states of elements were obtained using a Thermo Scientific K-Alpha X-ray photoelectron spectroscopy (XPS). The wavelength range of visible light adsorption was tested by a on a Shimadzu UV-2550 spectrophotometer in the spectral region of 300–800 nm. The recombination efficiency of photoexcited charge carrier was monitored by a HORIBA Scientific DeltaflexFlex fluorescence lifetime system and a Shimadzu RF-5301PC Luminescence spectrometer.

Photoelectrochemical measurements

Photoelectrochemical measurements were conducted on a CHI760E electrochemical workstation with a standard three-electrode system employing gauze platinum and Ag/AgCl (saturated KCl) as the counter and reference electrode. A slurry was prepared by mixing of 10 mg of as-prepared photocatalyst, 100 μL of Nafion (5 wt%) and 900 μL of ethanol, which was spin coated onto the surface of FTO glass to obtain the working electrode. After air-drying, the working electrode was immersed into a Na₂SO₄ solution (0.5 M) saturated with N₂ and irradiated by a 300 W Xe lamp (PLS-SXE300, Perfect Light). Mott–Schottky plots were recorded at potential range of 0–0.6 V under voltage magnitude of 10 mV and frequency of 1 kHz. The electrochemical impedance spectra (EIS) was measured at frequency range from 0.1 Hz to 100 kHz under a voltage magnitude of 10 mV.

Computational method and models

All calculations were performed by the Gaussian 16 package based on the density functional theory (DFT). The B3LYP with the 6-311G(d,p) basic set was used for geometries optimization and the Grimme's DFT-D3 method was applied to correct the van der Waals of all molecules. The full width at half maximum was set as 0.2 eV. A minimum g-C₃N₄ trimer with one nitrogen pot was built as the computational model, and its structure was optimized without any symmetry constraints. To investigate the location of the incorporated oxygen, one O atom was introduced into the g-C₃N₄ trimer by replacing one C/N atom. The Multiwfn was employed to analyze the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) of g-C₃N₄ and g-C₃N₄–O.

Photocatalytic performance measurement

Photocatalytic water splitting towards H₂ generation reactions of the as-prepared photocatalysts were performed in a 500 mL closed quartz reactor under a 300 W Xe lamp irradiation (PLS-SXE300, Perfect Light). The 100 mg of photocatalyst was dispersed in 300 mL aqueous solution containing 10 vol% triethanolamine (TEOA). Before the reaction, the reactor was evacuated to remove dissolved O₂ and bubbled with N₂ until saturation. The light source with a 420 nm cutoff filter was turned on to initiate the photocatalytic reaction and the gaseous products were quantified by a Shimadzu GC-2014 gas chromatograph.

Results and discussion

Powder X-ray diffraction (XRD) patterns of the colloidal SiO₂, bulk g-C₃N₄, and the modified g-C₃N₄ samples synthesized by using varied amount of colloidal SiO₂ were displayed in Fig. 1a. A wide XRD diffraction peak at 22.11° was observed for colloidal SiO₂, which was typical feature of amorphous silica.³⁵ Compared to the bulk g-C₃N₄, all the modified g-C₃N₄ samples prepared with different amount of colloidal SiO₂ exhibited identical characteristic diffraction peaks at 13.10° and 27.37°, suggesting the complete removal of SiO₂ and the basic heptazine-based structure of as-synthesized samples.³⁶ The decreased crystallinity of modified g-C₃N₄ sample prepared with 2000 mg colloidal SiO₂ was due to the inhibited condensation in the presence of excess SiO₂.³⁷ The molecular structures of the colloidal SiO₂, bulk g-C₃N₄ and the modified g-C₃N₄ samples were confirmed by FTIR spectroscopy.

Fig. 1 (a) XRD patterns and (b) FTIR spectra of colloidal SiO₂, bulk g-C₃N₄ and modified g-C₃N₄ samples prepared with different amount of colloidal SiO₂.

The colloidal SiO₂ presented a strong asymmetric stretching vibration of Si–O–Si band at 1070 cm⁻¹ and the characteristic bands of SiO₄ tetrahedra at 796 and 472 cm⁻¹ (Fig. 1b).³⁸ The FTIR spectra of the bulk g-C₃N₄ and modified g-C₃N₄ samples exhibited analogous stretching vibrations of aromatic CN heterocycles at 1200–1700 cm⁻¹ and the breathing vibration of heptazine units at 810 cm⁻¹ associated with a broad band of terminal –NH₂ in the range of 2800–3500 cm⁻¹ (Fig. 1b).³⁶ The FTIR results demonstrated that the modified g-C₃N₄ samples after the removal of colloidal SiO₂ maintained the intrinsic molecular skeleton of g-C₃N₄.

To verify the composition variation between the bulk g-C₃N₄ and the modified g-C₃N₄ samples, elemental levels of C, N and O for all the samples were analyzed. The samples prepared in three times were measured to obtain the average elemental content within the substrates, as shown in Table 1. In comparison with the relative low oxygen concentration of bulk g-C₃N₄ (1.93 wt%), the amounts of oxygen within the g-C₃N₄ substrate could be tuned from 7.88 to 8.95 wt% by increasing the amount of colloidal SiO₂ added from 500 to 2000 mg. The gradually increased oxygen content of the modified g-C₃N₄ samples illustrated that the selection of colloidal SiO₂ as oxygen source was valid to tune the oxygen concentration of g-C₃N₄. According to the previous work reported, the increased oxygen concentration was contributed by the presence of abundant hydroxyls on the surface of colloidal SiO₂ during the polymerization procedure.³⁹ Then the modified g-C₃N₄ photocatalysts were concentrated HNO₃ and maintained named as g-C₃N₄–O_x, where x presented the content of oxygen within the matrix. To confirm the content of residual Si element within the g-C₃N₄–O matrix, the photocatalyst was transferred into in polytetrafluoroethylene-lined stainless-steel autoclave at 140 °C for 2 h. No data of Si content was detected in the sample solution, demonstrating the complete removal of colloidal SiO₂ within the g-C₃N₄–O substrate.

Table 1 Elemental analysis of bulk g-C₃N₄ and modified g-C₃N₄ samples prepared with different amount of colloidal SiO₂

Sample	C/wt%	N/wt%	O/wt%
g-C3N4	34.34	61.69	1.93
500SiO2–g-C3N4	22.86	45.77	7.88
1000SiO2–g-C3N4	19.97	39.30	8.39
2000SiO2–g-C3N4	26.41	48.36	8.95

---

The X-ray photoelectron spectroscopy (XPS) and theoretical simulation were conducted to elucidate the elemental state and the location of incorporated oxygen within the g-C₃N₄–O photocatalyst. The survey XPS spectra of bulk g-C₃N₄ and g-C₃N₄–O photocatalysts both confirmed co-existence of C, N and O elements (Fig. 2a). The absence of Si element further confirmed the complete removal of colloidal SiO₂. The C 1s peaks at 284.8, 286.2 and 288.3 eV were assigned to the C–C, C–OH and C–N=C groups, respectively (Fig. 2b). New peaks ascribed to C–O bond emerged in the high-resolution C and O 1s XPS spectra of g-C₃N₄–O (Fig. 2b and d), proving the introduced oxygen existed in the form of C–O bond.⁴⁰ Combining with the increased C/N mol ratio of all the g-C₃N₄–O samples (0.687–0.709) compared to the bulk g-C₃N₄ (0.649), the introduced oxygen replaced N atoms of the g-C₃N₄ (Table 1). The N 1s peaks centered at 398.8 and 400.2 eV were assigned to C–N=C and N–(C)₃ groups (namely, N_2C and N_3C). The weak N 1s peak at 401.3 eV was ascribed to the terminal –NH_x groups.³⁴ The bulk g-C₃N₄ and g-C₃N₄–O photocatalysts exhibited identical characteristic peak positions in the high-resolution N 1s XPS spectra (Fig. 2c), implying the elevated concentration of oxygen exerted negligible impact on N species.

Fig. 2 (a) Survey XPS spectra, (b) high resolution C 1s XPS spectra, (c) high resolution N 1s XPS spectra and (d) high resolution O 1s XPS spectra of bulk g-C₃N₄ and g-C₃N₄–O_8.39 photocatalysts.

While the decreased peak area ratios of N_2C to N_3C for g-C₃N₄–O (3.48) in comparison with the bulk g-C₃N₄ (5.96) indicated that the substitute of sp²-hybrided N atoms by incorporated oxygen (Table 2).

Table 2 Relative ratios of C–N=C, N–(C)₃ and –NH_x determined by high resolution N 1s XPS spectra

Sample	C–N=C		N–(C)3		–NHx		N2C/N3C
	Area	(%)	Area	(%)	Area	(%)
g-C3N4	42 172	77.23	7072.12	12.95	5362.36	9.82	5.96
g-C3N4–O	31 427.28	69.52	9040.26	19.99	4739.98	10.48	3.48

---

Theoretical simulation further confirmed the location of incorporated oxygen atom in the heptazine framework. The model containing a trimer was constructed to represent the repeated heptazine units of g-C₃N₄ (Fig. 3a). One O atom was introduced into the model by substitution of one N atom. The more energy-favorable nitrogen site for O replacement was sp²-hybrided N site after configuration optimization (Fig. 3b). Results from elemental analysis, XPS measurements combined with theoretical calculations suggested the incorporated oxygen existed in the form of C–O bond by replacing sp²-hybrided N atom of the heptazine unit.

Fig. 3 Optimized structures of (a) g-C₃N₄ trimer and (b) g-C₃N₄–O trimer. The grey, brown, white and red balls represented C, N, H and O atoms, respectively.

The photocatalytic H₂ evolution rate of the bulk g-C₃N₄ and g-C₃N₄–O photocatalysts with different amount of oxygen were evaluated. As shown in Fig. 4a, all the g-C₃N₄–O photocatalysts showed obviously enhanced H₂ evolution rate compared to the bulk g-C₃N₄ (417.2 μmol g⁻¹ h⁻¹), indicating that the high oxygen concentration realized by the employment of colloidal SiO₂ could boost the photocatalytic activity of g-C₃N₄ based photocatalyst. When the content of oxygen increased from 7.88 to 8.39 wt%, the H₂ production rate increased from 3475.4 to 4396.8, and then decreased to 2559.2 μmol g⁻¹ h⁻¹. This revealed that the photocatalytic H₂ production performance could be tuned through amount regulation of oxygen within the g-C₃N₄ matrix. The recycling stability of g-C₃N₄–O photocatalyst with an optimal content of oxygen (8.39 wt%) was exhibited in Fig. 4b. Its photocatalytic activity maintained after four cycles of H₂ evolution tests, confirming the superior recycling stability of the as-synthesized g-C₃N₄–O photocatalyst.

Fig. 4 (a) Photocatalytic H₂ evolution rate of bulk g-C₃N₄ and g-C₃N₄–O photocatalysts with different content of oxygen, (b) cyclic H₂ evolution test of g-C₃N₄–O photocatalyst with an optimal amount of oxygen (8.39 wt%).

To understand the mechanism of improved H₂ production performance induced by increased oxygen content, the morphology, porosity, optical and optoelectronic properties of the bulk g-C₃N₄ and g-C₃N₄–O photocatalysts were characterized. The SEM image of colloidal SiO₂ displayed spherical shape with average size of ∼50 nm (Fig. 5a). SEM observations showed the typical sheet-like structure of bulk g-C₃N₄ with smooth surface (Fig. 5b). The morphologies of intermediate products were characterized for understanding the preparation process of the g-C₃N₄–O photocatalyst. After initial drying, phase separation between the silica nanoparticles and DCDA did not occur. It could be seen that the colloidal SiO₂ agglomerated around the DCDA (Fig. S1†). Then the products obtained from the calcination procedure exhibited agglomerated silica nanoparticles dispersed on/around the sheet-like g-C₃N₄ (Fig. S2†). The surface of obtained g-C₃N₄–O after removal of SiO₂ became rough and porous (Fig. 5c and d). The corresponding energy dispersive spectroscopy (EDS) elemental mapping (Fig. S3†) images indicated the successful uniform distribution of C, N and O elements within the g-C₃N₄–O matrix. BET results indicated the mesoporous structure and increased surface area of g-C₃N₄–O photocatalyst (Fig. 5e and f), which was one reason responsible for the improved photocatalytic H₂ production performance.

Fig. 5 (a–c) SEM images of (a) colloidal SiO₂, (b) bulk g-C₃N₄ and (c and d) g-C₃N₄–O_8.39 photocatalysts. (e) Nitrogen adsorption–desorption isotherms and (f) corresponding pore size distribution of bulk g-C₃N₄ and g-C₃N₄–O_8.39 photocatalysts.

The optical and optoelectronic properties of g-C₃N₄ with high concentration of oxygen was evaluated by UV-vis diffuse reflectance spectra. The bulk g-C₃N₄ photocatalyst exhibited an absorption edge around 460 nm attributed to semiconductor band–band transition (Fig. 6a). The deeper colour and tail adsorption in the visible light region g-C₃N₄–O photocatalyst showed red-shifted adsorption edge and tail adsorption in the visible light range of 500–800 nm, indicating that the increased oxygen content tail adsorption in the visible light region of g-C₃N₄–O photocatalysts implied enhanced visible light utilization ability, which was favorable for photocatalysis. The PL emission spectra of the samples at an excitation wavelength of 365 nm was shown in Fig. 6b. The decreased PL emission peak intensity indicated the decreased recombination rate of photoexcited electron–hole pairs for g-C₃N₄–O photocatalyst. The faster PL decay spectra confirmed the facilitated separation kinetics realized by the increased concentration of oxygen (Fig. 6c).^41,42 This synergistically demonstrated the increased oxygen content promoted photoelectric conversion for H₂O reduction. The periodic photocurrent tests were performed with a light on/off interval of 20 s under open circuit voltage. The enhanced photocurrent density of g-C₃N₄–O photocatalyst implied the migration promotion of photoinduced charge carriers realized by high concentration of oxygen introduction (Fig. 6d).

Fig. 6 (a) UV-vis diffusion reflectance spectra (inner pictures: optical pictures of bulk g-C₃N₄ and g-C₃N₄–O samples), (b) PL emission spectra, (c) time-resolved PL decay spectra and (d) periodic photocurrent response of bulk g-C₃N₄ and g-C₃N₄–O photocatalysts.

The optical property of semiconductor photocatalyst was determined by its electronic energy band structure.⁴³ Thereby, the tauc plots and Mott–Schottky plots were measured to obtain the electronic energy band structure of bulk g-C₃N₄ and g-C₃N₄–O photocatalysts. The band gap of g-C₃N₄ and g-C₃N₄–O photocatalysts were determined to be 2.70 and 2.48 eV from the Tauc plots (Fig. 7a). The high concentration of oxygen induced narrowed band gap and thus red-shifted adsorption edge of g-C₃N₄–O photocatalyst, elevating utilization efficiency of visible light adsorption with longer wavelength. The flat band potentials (E_fb vs. Ag/AgCl) of the g-C₃N₄ and g-C₃N₄–O photocatalysts were −1.05 and −0.8 V according to the X intercept at the tangents of the Mott–Schottky plots (Fig. 7b). The corresponding E_fb (vs. NHE, pH = 0) potentials of g-C₃N₄ and g-C₃N₄–O photocatalysts were derived to be −0.45 and −0.20 V, respectively.⁴⁴ Generally, the E_fb potential of the n-type semiconductor was 0.3 V positive than conduction band minimum (CBM) potential.⁴⁵ Hence, the CBM potentials of g-C₃N₄ and g-C₃N₄–O photocatalysts were identified to be −0.75 and −0.50 V, respectively. Combining with the E_g values, the valence band minimum (VBM) potentials of the g-C₃N₄ and g-C₃N₄–O photocatalysts were calculated to be 1.95 and 1.98 V, respectively. On the basis of above results, the electronic energy band structures of g-C₃N₄ and g-C₃N₄–O photocatalysts were depicted in Fig. 7c. It could be known that the high concentration of oxygen influenced the energy band potential and narrowed the band gap, accounting for its optimized optical property.

Fig. 7 (a) Tauc plots, (b) Mott–Schottky plots and (c) energy band structure illustration for bulk g-C₃N₄ and g-C₃N₄–O photocatalysts.

The highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) were calculated to verify the effect of introduced oxygen on the local electron distribution of the π conjugated system (Fig. 8). For the bulk g-C₃N₄, the HOMO is primarily derived from the p_z orbitals of N atoms and the LUMO mainly localized in C–N bond orbitals, which usually resulted in easy recombination of photoexcited electrons and holes.^46,47 The incorporated O atom induced evidently local electron distribution perturbation of HOMO, manifesting in enhanced electron localization located in the oxygen-free subunit. The difference in HOMO and LUMO led to spatial separation of photoexcited electron and holes, accounting for the inhibited recombination of charge carriers for the g-C₃N₄–O photocatalyst. Besides, the O atom within the heptazine unit introduced extra electrons to increase charge density of the π conjugated system. The improved electronic conductivity was responsible for the promoted transportation of charge carriers.

Fig. 8 Calculated HOMO and LUMO for (a and c) bulk g-C₃N₄ and (b and d) g-C₃N₄–O trimer.

Conclusions

In summary, porous g-C₃N₄ photocatalyst with tunable high concentration of oxygen was synthesized via copolymerization of DCDA and colloidal SiO₂. The removal of colloidal SiO₂ contributed to porous structure of g-C₃N₄–O and increased surface area, favoring reactants adsorption and mass transfer. The introduced oxygen by replacement of two-coordinated N atoms narrowed the band gap and disturbed the local electron distribution, thus extended the visible light harvesting and promoted the separation and transportation of photogenerated charge carriers. The obtained g-C₃N₄–O photocatalyst with an optimal oxygen content (8.39 wt%) showed 10.5 times enhancement in H₂ production rate in comparison with bulk g-C₃N₄, attributed to porous structure formation and high concentration of oxygen incorporation. Such work revealed the effect of colloidal SiO₂ on the composition regulation of g-C₃N₄ based photocatalysts.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the Science and Technology Development Project of Jilin Province (20220101248JC), the 13th Five-Year Science and Technology Project of the Education Department of Jilin Province (JJKH20200124KJ) and the Doctoral Research Start-up Fund of Jilin Medical University (JYBS2021032LK).

References

1. Y. Ping, X. Liu, J. Iocozzia, Y. Yuan, L. Gu, G. Xu and Z. Lin, Highly stable non-noble metal Ni₂P co-catalyst for increased H₂ generation by g-C₃N₄ under visible light irradiation, J. Mater. Chem. A, 2017, 5, 8493–8498.
2. Z. Hui, Y. Dong, P. Jiang, H. Miao and J. Zhang, In situ light-assisted preparation of MoS₂ on graphitic C₃N₄ nanosheets for enhanced photocatalytic H₂ production from water, J. Mater. Chem. A, 2015, 3, 7375–7381.
3. A. Fujishima and K. Honda, Electrochemical photolysis of water at a semiconductor electrode, Nature, 1972, 238, 37–38.
4. L. Ge, C. C. Han, X. L. Xiao, L. L. Guo and Y. J. Li, Enhanced visible light photocatalytic hydrogen evolution of sulfur-doped polymeric g-C₃N₄ photocatalysts, Mater. Res. Bull., 2013, 48, 3919–3925.
5. Z. Lin, S. J. Wang, Y. Q. Miao, J. P. Yuan, Y. L. Liu, S. G. Xu and S. K. Cao, One-step preparation of halogenated aminobenzonitrile modified g-C₃N₄ via copolymerization and in situ halogen doping for highly enhanced visible light hydrogen evolution, Int. J. Hydrogen Energy, 2020, 45, 6341–6351.
6. J. K. Lin, W. J. Tian, H. Y. Zhang, X. G. Duan, H. Q. Sun and S. B. Wang, Graphitic carbon nitride-based Z-scheme structure for photocatalytic CO₂ reduction, Energy Fuels, 2020, 35, 7–24.
7. K. Z. Qi, Y. B. Xie, R. D. Wang, S. Y. Liu and Z. Zhao, Electroless plating Ni-P cocatalyst decorated g-C₃N₄ with enhanced photocatalytic water splitting for H₂ generation, Appl. Surf. Sci., 2019, 466, 847–853.
8. P. Karthika, T. R. N. Kumarab and B. Neppolian, Redox couple mediated charge carrier separation in g-C₃N₄/CuO photocatalyst for enhanced photocatalytic H₂, Int. J. Hydrogen Energy, 2020, 45, 7541–7551.
9. J. Feng, M. M. Gao, Z. Q. Zhang, M. Z. Gu, J. X. Wang, W. J. Zeng and Y. M. Ren, Comparing the photocatalytic properties of g-C₃N₄ treated by thermal decomposition, solvothermal and protonation, Results Phys., 2018, 11, 331–334.
10. B. R. Bhagat and A. Dashora, Understanding the synergistic effect of co-loading and B-doping in g-C₃N₄ for enhanced photocatalytic activity for overall solar water splitting, Carbon, 2021, 178, 666–677.
11. H. Z. Liu, X. L. Wan, H. Xu and C. X. Luo, Facile synthesis of F-doped g-C₃N₄/Bi₂Fe₄O₉ heterostructure with Z-scheme for enhanced photocatalytic performance in NO oxidation, J. Phys. Chem. Solids, 2020, 146, 109500–109506.
12. L. B. Jiang, X. Z. Yuan, Y. Pan, J. Liang, G. M. Zeng, Z. B. Wu and H. Wang, Doping of graphitic carbon nitride for photocatalysis: A review, Appl. Catal., B, 2018, 220, 222–233.
13. H. Starukh and P. Praus, Doping of graphitic carbon nitride with non-metal elements and its applications in photocatalysis, Catal, 2020, 10, 1119–1157.
14. S. Patnaik, D. P. Sahoo and K. Parida, Recent advances in anion doped g-C₃N₄ photocatalysts: A review, Carbon, 2021, 172, 682–711.
15. H. L. Dou, S. H. Zheng and Y. P. Zhang, Graphitic carbon nitride with S and Fe(III) codoping for improved photodegradation performance, Catal. Lett., 2017, 148, 601–611.
16. C. Ye, J. X. Li, Z. J. Li, X. B. Li, X. B. Fan, L. P. Zhang, B. Chen, C. H. Tung and L. Z. Wu, Enhanced driving force and charge separation efficiency of protonated g-C₃N₄ for photocatalytic O₂ evolution, ACS Catal., 2015, 5, 6973–6979.
17. Y. O. Wang, H. Suzuki, J. J. Xie, O. Tomita, D. J. Martin, M. Higashi, R. Abe and J. W. Tang, Mimicking natural photosynthesis:solar to renewable H₂ fuel synthesis by Z-scheme water splitting systems, Chem. Rev., 2018, 118, 5201–5241.
18. K. Pandi, S. K. Lakhera and N. Bernaurdshaw, Efficient promotion and transfer of excited charge carriers in phosphorus doped and Ni complex modified g-C₃N₄, Catal. Today, 2021, 370, 161–172.
19. L. F. Cui, J. L. Song, A. F. McGuire, S. F. Kang, X. Y. Fang, J. J. Wang, C. C. Yin, Y. G. Wang and B. X. Cui, Constructing highly uniform onion-ring-like graphitic carbon nitride for efficient visible-light-driven photocatalytic hydrogen evolution, ACS Nano, 2018, 12, 5551–5558.
20. M. Faisal, A. A. Ismail, F. A. Harraz, S. A. Al-Sayari, A. M. El-Toni, A. E. Al-Salami and M. S. Al-Assiri, Fabrication of highly efficient TiO₂/C₃N₄ visible light driven photocatalysts with enhanced photocatalytic activity, J. Mol. Struct., 2018, 1173, 428–438.
21. N. I. M. Rosli, S. M. Lam, J. C. Sin, I. Satoshi and A. R. Mohamed, Photocatalytic performance of ZnO/g-C₃N₄ for removal of phenol under simulated sunlight irradiation, J. Environ. Eng., 2018, 144, 04017091–04017103.
22. J. H. Ma, Z. Q. Wei, L. Li, L. Ma, C. Li and S. P. Huang, Synthesis and photoelectrochemical properties of visible-light response g-C₃N₄@CdS heterojunctions photocatalyst, Desalin. Water Treat., 2021, 231, 287–296.
23. M. I. Chebanenko, A. A. Lobinsky, V. N. Nevedomskiy and V. I. Popkov, NiO-decorated graphitic carbon nitride toward electrocatalytic hydrogen production from ethanol, Dalton Trans., 2020, 49, 12088–12097.
24. Z. Feng, L. Zeng, Y. J. Chen, Y. Y. Ma, C. R. Zhao, R. S. Jin, Y. Lu, Y. Wu and Y. M. He, In situ preparation of Z-scheme MoO₃/g-C₃N₄ composite with high performance in photocatalytic CO₂ reduction and RhB degradation, J. Mater. Res., 2017, 32, 3660–3668.
25. B. Lin, C. Xue, X. Q. Yan, G. D. Yang, G. Yang and B. L. Yang, Facile fabrication of novel SiO₂/g-C₃N₄ core shell nanosphere photocatalysts with enhanced visible light activity, Appl. Surf. Sci., 2015, 357, 346–355.
26. L. Peng, Z. W. Li, R. R. Zheng, H. Yu and X. T. Dong, Preparation and characterization of mesoporous g-Facile fabrication of novel SiO₂ g-C₃N₄ core shell nanosphere photocatalysts with enhanced visible light activity/SiO₂ material with enhanced photocatalytic activity, J. Mater. Res., 2019, 34, 1785–1794.
27. S. B. Zhang, M. Li, W. J. Qiu, Y. Wei, G. F. Zhang, J. Y. Han, H. Wang and X. Liu, Super small polymeric carbon nitride nanospheres with core-shell structure for photocatalysis, ChemistrySelect, 2017, 2, 10580–10585.
28. X. Wang, X. Wang X, J. F. Huang, S. X. Li, A. Meng and Z. J. Li, Interfacial chemical bond and internal electric field modulated Z-scheme Sv-ZnIn₂S₄/MoSe₂ photocatalyst for efficient hydrogen evolution, Nat. Commun., 2021, 12, 1–11.
29. X. H. Dang, M. S. Xie, F. F. Dai, J. N. Guo, J. Liu and X. Q. Lu, Ultrathin 2D/2D ZnIn₂S₄/g-C3N4 nanosheet heterojunction with atomic-level intimate interface for photocatalytic hydrogen evolution under visible light, Adv. Mater. Interfaces, 2021, 8, 2100151–2100164.
30. X. H. Yang, C. Cao, Z. L. Guo, X. Y. Zhang, Y. X. Wang and W. S. Yang, Promoting hydrogen evolution of a g-C₃N₄-based photocatalyst by indium and phosphorus co-doping, New J. Chem., 2021, 45, 7231–7238.
31. M. Shalom, S. Inal, D. Neher and M. Antoniettia, SiO₂/carbon nitride composite materials: the role of surfaces for enhanced photocatalysis, Catal. Today, 2014, 225, 185–190.
32. Q. Hao, X. X. Niu, C. S. Nie, S. M. Hao, W. Zou, J. M. Ge, D. M. Chen and W. Q. Yao, A highly efficient g-C₃N₄/SiO₂ heterojunction: the role of SiO₂ in the enhancement of visible light photocatalytic activity, Phys. Chem. Chem. Phys., 2016, 18, 31410–31418.
33. M. T. Xue, G. Q. Tan, T. Liu, L. Lv, B. Li, D. Zhang, M. Y. Dang, H. J. Ren and A. Xia, Insights into the improved photocatalytic performance of fluorine surface modified mpg-C₃N₄ at room temperature under aqueous conditions, Appl. Catal., A, 2019, 578, 89–97.
34. M. S. Akple, T. Ishigaki and P. Madhusudan, Bio-inspired honeycomb-like graphitic carbon nitride for enhanced visible light photocatalytic CO₂ reduction activity, Environ. Sci. Pollut. Res., 2020, 27, 22604–22618.
35. R. K. Jia, Y. L. Zhang and X. H. Yang, High efficiency photocatalytic CO₂ reduction realized by Ca²⁺ and HDMP group co-modified graphitic carbon nitride, Int. J. Hydrogen Energy, 2021, 46, 32893–32903.
36. B. Lin, S. Chen, F. Dong and G. D. Yang, A ball-in-ball g-C₃N₄@SiO₂ nano-photoreactor for highly efficient H₂ generation and NO removal, Nanoscale, 2017, 9, 5273–5279.
37. X. X. Wang, S. S. Wang, W. D. Hu, J. Cai, L. H. Zhang, L. Z. Dong, L. H. Zhao and Y. M. He, Synthesis and photocatalytic activity of SiO₂/g-C₃N₄ composite photocatalyst, Mater. Lett., 2014, 115, 53–56.
38. S. Narzary, K. Alamelu, V. Raja and B. M. J. Ali, Visible light active, magnetically retrievable Fe₃O₄@SiO₂@g-C₃N₄/TiO₂ nanocomposite as efficient photocatalyst for removal of dye pollutants, J. Environ. Chem. Eng., 2020, 8, 104373–104382.
39. M. Q. Xu, B. Chai, J. T. Yan, H. B. Wang, Z. D. Ren and K. W. Paik, Facile synthesis of fluorine doped graphitic carbon nitride with enhanced visible light photocatalytic activity, Nano, 2016, 11, 68–78.
40. F. He, G. Chen, Y. S. Zhou, Y. G. Yu, Y. Zheng and S. Hao, The facile synthesis of mesoporous g-C₃N₄ with highly enhanced photocatalytic H₂ evolution performance, Chem. Commun., 2015, 51, 16244–16246.
41. F. K. Ma, C. L. Sun, Y. L. Shao, Y. Z. Wu, B. B. Huang and X. P. Hao, One-step exfoliation and fluorination of g-C₃N₄ nanosheets with enhanced photocatalytic activities, New J. Chem., 2017, 41, 3061–3067.
42. A. Allahresani, B. Taheri and M. A. Nasseri, A green synthesis of spirooxindole derivatives catalyzed by SiO₂@g-C₃N₄ nanocomposite, New J. Chem., 2017, 41, 3061–3067.
43. G. L. Xu, J. C. Shen, S. M. Chen, Y. J. Gao, H. B. Zhang and J. Zhang, Double defects modified carbon nitride nanosheets with enhanced photocatalytic hydrogen evolution, Phys. Chem. Chem. Phys., 2018, 20, 17471–17476.
44. W. Wang and J. J. Fang, Mesoporous SiO₂-derived g-C₃N₄@CdS core-shell heteronanostructure for efficient and stable photocatalytic H₂ production, Ceram. Int., 2020, 46, 2384–2391.
45. X. Yang, Z. Guo, X. Zhang, Y. Han, Z. Xue and T. Xie, et al., The effect of indium doping on the hydrogen evolution performance of g-C₃N₄ based photocatalysts, New J. Chem., 2021, 45, 544–550.
46. X. Wang, X. Chen, A. Thomas, X. Fu and M. Antonietti, Metal-containing carbon nitride compounds: a new functional organic-metal hybrid material, Adv. Mater., 2009, 21, 1609–1612.
47. Z. A. Lan, G. Zhang and X. Wang, A facile synthesis of Br-modified g-C₃N₄ semiconductors for photoredox water splitting, Appl. Catal., B, 2016, 192, 116–125.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2ra05662c

---