Activating atomically dispersed Co–N/C sites on g-C₃N₄ nanosheets via incorporating sulfur enables efficient visible light H₂ evolution

Abstract

Embedding atomically dispersed metal (M)–N/C sites within two-dimensional (2D) g-C₃N₄ nanosheets (CNs) is able to greatly improve charge separation efficiency; however, these sites are typically inert for catalyzing the photochemical H₂ evolution reaction (HER). In this work, a simple room-temperature sulfurization strategy is proposed to activate the inert Co–N/C sites on CNs for efficiently catalyzing the dye-sensitized photocatalytic HER. By simply immersing cobalt-doped CNs (Co-CNs) into aqueous solution with high-concentration S²⁻, the coordinated O atoms at Co–N/C sites are partially replaced by S atoms at room temperature, generating ample highly active S-coordinated Co–N/C sites with excellent dispersion preservation. The as-prepared S-Co-CNs catalysts exhibit enhanced activity in catalyzing the HER in an Erythrosin B–triethanolamine (ErB–TEOA) system under visible light irradiation, while both the pristine CNs and Co-CNs show negligible activity under the same reaction conditions. The most efficient S-Co-CNs catalyst achieves a H₂ evolution rate of 6.38 mmol h⁻¹ g⁻¹ and a quantum efficiency (QE) of 13.02% at 520 nm, and this S-Co-CNs catalyst shows excellent HER stability when sensitized with a more stable fluorescein (FL) dye. The enhanced catalytic performance originates from the favorable electronic structure of the active sites adjusted by S modulation, leading to reduced H₂ evolution overpotential while maintaining the favorable electron transfer kinetics. This work provides an effective strategy to activate inert sites embedded in CNs for high-performance photocatalytic water splitting, organic transformation, and CO₂ reduction.

---

Introduction

The solar photocatalytic H₂ evolution from water splitting has been recognized as a promising route to solve the ever-increasing environmental issues and energy crisis.^1–8 Among the photocatalysts developed so far, metal-free graphitic carbon nitride (g-C₃N₄) is a highly promising candidate for photocatalytic H₂ evolution due to its suitable band gap of ca. 2.7 eV, good chemical and thermal stability, and low cost.^9–15 However, pristine g-C₃N₄ shows low photocatalytic activity, attributed to the insufficient visible light absorption (λ < 470 nm), the fast recombination of charge carriers, and the lack of active sites for the H₂ evolution reaction (HER).^12–15 Several strategies, including dye-sensitization,^16–21 metal and/or non-metal doping,^22–32 nanostructuring,^33–35 coupling with other inorganic semiconductors and conductive materials,^36–41 and cocatalyst loading,^{9–12,42–46} have been proposed to enhance the photocatalytic activity of g-C₃N₄. Among these methods, dye-sensitization and loading cocatalysts have been proved to be facile yet effective methods for improving the photocatalytic HER performance of g-C₃N₄. Dye-sensitization of g-C₃N₄ could greatly extend the visible light absorption for maximizing the utilization efficiency of solar light,¹⁷ while loading a suitable cocatalyst on g-C₃N₄ can accelerate carrier transportation and reduce the barriers for the HER.⁹ Although noble metals, such as Pt-group metals, are the best cocatalysts for g-C₃N₄, their scarcity and high cost limit their practical applications. Therefore, it is still desirable to develop an alternative cocatalyst based on earth-abundant elements.^47–55

On the other hand, it has been found that introducing M–N/C sites (M = Fe, Co, Ni, Zn, etc.) at the atomic level within the framework of g-C₃N₄ could effectively tune the electronic structure of g-C₃N₄, and the resulting M-doped g-C₃N₄ (M-g-C₃N₄) materials typically possess reduced bandgap energy and enhanced charge separation efficiency,^{21–26,56,57} and thus improved photocatalytic performances. However, the confined M–N/C sites are hardly active for catalyzing the photocatalytic HER over M-g-C₃N₄, and an additional cocatalyst, such as Pt, still needs to be loaded on M-g-C₃N₄ to provide active sites for enhancing the photocatalytic HER activity. Fortunately, it has been found that the M-N/C sites could be activated to achieve reactivity for catalyzing the photocatalytic HER by rationally regulating the electronic structures of coordinatively unsaturated M–N/C sites via introducing an additional active ligand. In this context, Liu et al.⁵⁸ reported that phosphidation of a single Co₁-oxo site confined on g-C₃N₄ nanosheets led to a single Co₁–P₄ site that can act as an efficient reaction center for the overall water splitting reaction. Chen et al.⁵⁹ discovered that single Ni active sites coordinated to a S-containing ligand chemically bonded to g-C₃N₄ can efficiently catalyze visible light H₂ evolution with an activity comparable to that of 3 wt% Pt loaded g-C₃N₄ under the identical conditions. While these studies provided a new approach to construct active single transition metal-based active sites on the g-C₃N₄ matrix by deliberately choosing secondary active ligands, the synthesis of these materials typically involved multistep reactions using toxic reagents under harsh reaction conditions. Therefore, it is still highly desirable to develop a facile yet effective approach to activate inert M–N/C sites on g-C₃N₄ to obtain H₂ evolution catalysts with high intrinsic activity.

Herein, we present a facile yet effective approach to activate inert Co–N/C sites on 2D g-C₃N₄ nanosheets (CNs) to highly active S-coordinated Co–N/C (S-Co–N/C) sites for catalyzing the H₂ evolution reaction in a dye-sensitized system under visible light. The S-Co–N/C sites are generated in situ on CNs via a room-temperature sulfurization strategy by simply immersing Co-doped CNs (Co-CNs) precursor in a solution with a high concentration of S²⁻, and the O atoms (O²⁻) that are coordinated to Co–N/C sites are partially replaced by S²⁻ according to the solubility equilibrium. The resulting S-Co–N/C sites can precisely retain the atomic-level dispersion of the Co–N/C sites and avoid the formation of aggregated nanoparticles of Co sulfides. Benefitting from the high dispersion and favorable electronic structure, the S-Co–N/C sites on CNs can not only lead to the efficient transportation of electrons but also greatly reduce the reaction barriers for the H₂ evolution reaction. As a result, the best S-Co-CNs catalyst sensitized by Erythrosin B (ErB) exhibits a much higher H₂ evolution rate of 6.38 mmol h⁻¹ g⁻¹ under visible light irradiation and an AQY of 13.02% at 520 nm, and greatly outperforms both pristine CNs and Co-CNs. In addition, when sensitized with a stable dye, the S-Co-CNs catalyst also shows an excellent long-term stability.

Experimental

Chemicals and reagents

All reagents including (NH₄)₂S solution (21 wt%), Co(NO₃)₂·6H₂O, Ni(NO₃)₂·6H₂O, CuSO₄·5H₂O, Zn(NO₃)₂·6H₂O, and urea were of analytical grade and used without further purification. Triethanolamine (TEOA) was purchased from Xilong Scientific (>99.8%). Erythrosin B (ErB) and fluorescein (FL) were obtained from Tianjin Guangfu Fine Chemical Research Institute. All solutions in the experiments were prepared with deionized water (>15 MΩ) that was obtained from a YL-100B-D water-purification system.

Synthesis of S-Co-CN catalyst

Co-doped g-C₃N₄ nanosheets (Co-CNs) precursor was first synthesized by a well-established thermal polymerization method.¹⁷ Typically, 20 g of urea and 40 mg of Co(NO₃)₂·6H₂O were dissolved into 100 mL of water under stirring to form a homogeneous solution. Then, the water was evaporated to generate a solid mixture. After being ground into powders, the solid mixture was placed in a crucible with a cover and calcined in a muffle furnace at 550 °C for 2 h with a heating rate of 5 °C min⁻¹. The obtained products were washed with water and ethanol several times, and dried in a vacuum oven at 60 °C overnight. The final product was labeled as Co-CNs. The pristine CNs was prepared by identical processes to Co-CNs without the addition of Co salt.

S-Co-CNs catalyst was then synthesized via the room-temperature sulfurization of Co-CNs. Typically, 150 mg of Co-CNs was dispersed into 25 mL of water and ultrasonicated for 10 min to form a homogeneous suspension. After that, 500 μL of (NH₄)₂S solution was added into the above suspension and the resulting mixture was stirred for 6 h at room temperature. The obtained product, denoted as S-Co-CNs, was collected by centrifugation and then rinsed with water and ethanol several times, and finally dried in a vacuum oven at 60 °C overnight. A series of metal-doped (Ni, Cu, Zn, and Fe) CNs catalysts were also synthesized using the corresponding metal salts and sulfurized with the exact same procedures as for the synthesis of S-Co-CNs catalyst.

Characterization

Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) images of the samples were recorded using a FEI Tecnai-G²-F30 transmission electron microscope. X-ray diffraction (XRD) patterns of the samples were recorded with a Rigaku smartlab powder X-ray diffractometer with nickel filtered Cu Kα radiation from 5 to 80°. Fourier transform infrared spectroscopy (FTIR) spectra of the samples were collected by using a Thermo Nicolet Avatar 380 FT-IR spectrometer. X-ray photoelectron spectroscopy (XPS) measurements of the samples were performed on a Thermo Scientific Escalab-250Xi electron spectrometer and the C 1s peak at 284.8 eV was a signal-calibrating standard for all the binding energies. UV-vis absorption spectra of the reaction solutions were recorded by using a Thermo Scientific-Evolution 220 spectrophotometer. UV-vis diffuse reflectance spectra of the solid samples were recorded on a PerkinElmer Lambda-750 UV-vis-near-IR spectrometer. The textile structures of the samples were determined by using a Micromeritics ASAP 2020 HD88 adsorption apparatus. Photoluminescence (PL) spectra were collected on a Horiba Scientific FluoroMax-4 spectrofluorometer spectrometer. The PL decay profiles were measured using the Horiba Jobin Yvon Data Station HUB operating in time correlated single photon counting mode (TCSPC) (time resolution: 200 ps) with a nano LED diode emitting pulses at 369 nm with 1 MHz repetition rate as an excitation source. Light-scattering Ludox solution was used to obtain the instrument response function (prompt). Horiba Jobin Yvon DAS6 fluorescence decay analysis software was used to fit the model functions to the experimental data.

Photocatalytic H₂ evolution experiments

Photocatalytic H₂ evolution experiments were performed in a sealed Pyrex reactor (340 mL) with a top flat quartz window (38.47 cm²) for light irradiation and a silicone rubber septum was fixed on its side for sampling the produced H₂. In a typical procedure, the S-Co-CNs catalyst (50 mg) and ErB (0.2 mM) were added to a reaction cell containing 100 mL of 10 vol% TEOA (pH 8) aqueous solution under vigorous stirring. The reaction system was then thoroughly degassed for 10 min and backfilled with N₂. After that, the reaction solution was irradiated by using a 30 W LED lamp equipped with a cut-off filter (λ ≥ 450 nm). The amount of H₂ evolved was measured by using a gas chromatograph (Tech comp; GC-7900) with a thermal conductivity detector, a 5 A molecular sieve column (4 mm × 5 m), and with N₂ as carrier gas. The apparent quantum yield (AQY) for the H₂ evolution reaction was measured under light irradiation of different wavelengths. The photon flux of the incident monochromatic light was determined using a ray virtual radiation actinometer (Apogee MQ-500), and the AQY was calculated based on the following equation:

Electrochemical measurements

The electrochemical measurements were carried out in a standard three-electrode system with an electrochemical workstation (CHI760E, Shanghai, China). Saturated Ag/AgCl and a graphite rod were used as the reference and counter electrodes, respectively. The supporting electrolyte was degassed 10% TEOA (pH 8) solution containing 0.5 M Na₂SO₄. The working electrodes were prepared by drop-casting 0.12 mL of the catalyst (CNs, Co-CNs, or S-Co-CNs) suspension in ethanol (5 mg mL⁻¹) onto both sides of carbon paper (HESEN, HCP030P, thickness: 0.3 mm, 1.5 cm × 0.5 cm) and dried at room temperature. The electrocatalytic H₂ evolution activity was examined by measuring polarization curves using the linear sweep voltammetry (LSV) technique at a scan rate of 5 mV s⁻¹. Electrochemical impedance spectroscopy (EIS) was carried out with an AC amplitude of 5 mV and a frequency range of 10 mHz to 100 kHz. The Mott–Schottky analysis was conducted at a frequency of 1 kHz and a scan rate of 50 mV s⁻¹.

Results and discussion

The S-Co-CNs catalyst was prepared by a simple liquid-phase anion-exchange reaction with Co-CNs and high-concentration S²⁻ at room temperature. As schematically illustrated in Fig. 1, Co-CNs precursor was first prepared by pyrolyzing a mixture of urea and Co(NO₃)₂ in air, which produces numerous O-coordinated Co–N/C sites that are highly dispersed and firmly embedded in the frameworks of CNs. Then, the Co-CNs precursor was directly reacted with (NH₄)₂S in aqueous solution at room temperature, during which the coordinated O atoms at Co–N/C sites were partially replaced by S atoms, generating ample highly active S-coordinated Co–N/C sites. The successful anion-exchange along with S incorporation can be visually confirmed by the color change from light yellow for Co-CNs to grey for S-Co-CNs. Unlike conventional vapor-phase and hydrothermal sulfurization processes that are typically performed at high temperature involving the use of toxic chemicals (such as thiourea and thioacetamide), and in which the active sites tend to agglomerate to form larger isolated nanoparticles,⁵² the present developed synthetic route based on the solution-based sulfurization reaction allows the efficient activation of Co–N/C sites while maintaining the high dispersion, which would be beneficial to increase the catalytic activity for photocatalytic H₂ evolution.

Fig. 1 Schematic illustration of the synthesis of S-Co-CNs catalyst via the room-temperature sulfurization strategy and their applications in the visible light H₂ evolution.

X-ray diffraction (XRD) patterns of pristine CNs, Co-CNs, and S-Co-CNs are shown in Fig. 2A. For pristine CNs, two characteristic peaks at around 13.0° and 27.5° are observed, which can be indexed to the (100) plane of the in-plane ordering of tri-s-triazine repeating units and the (002) plane of the graphitic stacking of C₃N₄ layers, respectively.⁹ The XRD pattern of Co-CNs does not differ significantly from that of pristine CNs, except for a decrease in the intensity of the (002) peak due to the interruption of polymeric condensation by Co incorporation.^23,26 No diffraction peaks related to Co species, such as metallic Co, Co oxides, Co nitrides, and Co carbides, could be detectable. These results suggest that the Co species were homogeneously dispersed in the CN matrix, most likely via forming atomically dispersed Co–N/C species.²⁶ After room-temperature sulfurization, the diffraction pattern of S-Co-CNs catalyst shows no significant difference compared to those of pristine CNs and Co-CNs, and no peaks originating from Co-based sulfides could be found, indicating that the coordination of S atoms to the Co atoms has no significant effect on the phase structure of CNs and the resulting Co species still preserve their high dispersion in the frameworks of CNs.

Fig. 2 (A) XRD patterns, (B) UV-Vis DRS, (C) FT-IR spectra, and (D) N₂ adsorption–desorption isotherms of pristine CNs, Co-CNs, and S-Co-CNs. Insets in panel (A) and (D) show the digital images and pore size distribution curves of pristine CNs, Co-CNs, and S-Co-CNs, respectively.

Fig. 2B shows the UV-visible diffuse reflectance spectra (UV-vis-DRS) of pristine CNs, Co-CNs, and S-Co-CNs. The pristine CNs exhibit a strong absorption from 200 to 380 nm with a sharp absorption edge at around 438 nm, corresponding to a band gap (E_g) of 2.83 eV (Table 1). Compared to the pristine CNs, it is clearly observed that the absorption band edge of Co-CNs is shifted to the red wavelength region (∼447 nm, E_g = 2.77 eV). In addition, obvious enhanced optical absorption was found in the visible light spectral range of 450–650 nm. This indicates that a ligand-to-metal charge transition (LMCT) similar to molecular systems occurs in solid transition metal-doped CNs,^23–26 demonstrating the success of Co incorporation in CNs at the atomic level and the strong electronic interaction between Co and CNs. After sulfurization, the obtained S-Co-CNs shows a similar band gap to Co-CNs, but the absorption in the range of 450–800 nm is slightly enhanced along with a color change from light yellow to grey, indicating the formation of S-coordinated Co–N/C species on CNs. Fourier transform infrared (FTIR) spectroscopy characterization was performed to confirm the chemical structure of the samples. As shown in Fig. 2C, the FTIR spectrum of pristine CNs shows the vibrations of various C/N-containing groups, e.g., the characteristic mode of the triazine units (≈807 cm⁻¹) and CN heterocycles (1200–1700 cm⁻¹), as well as a broad peak in the range of 3000–3300 cm⁻¹ corresponding to the stretching modes of the –OH/–NH₂ groups.¹⁷ All these characteristic bands are well-preserved after the Co incorporation and subsequent sulfurization, and no new peaks attributed to the C–S/C=S bands typically at 900–1300 cm⁻¹ can be observed in the spectrum of S-Co-CNs. These results suggest that the S atoms are coordinated to the unsaturated Co atoms in Co–N/C sites instead of incorporating into the frameworks of CNs.

Table 1 Physical and textural properties of the samples

Sample	S BET (m^2 g^−1)	Pore size^b (nm)	V pore (cm^3 g^−1)	E g (eV)
a BET surface area is calculated from the linear part of the BET plot. b Adsorption average pore width (4 V A^−1 by BET). c Single point total pore volume of the pores at P/P0 = 0.99.
CNs	48.1	21.0	0.252	2.83
Co-CNs	92.5	21.6	0.499	2.77
S-Co-CNs	94.7	22.8	0.541	2.77

---

The N₂ adsorption–desorption isotherms measured at 77 K and pore size distribution curves of the samples are given in Fig. 2D, and the calculated textural data are summarized in Table 1. The isotherms for all samples are identified as typical type IV with a H3-type hysteresis loop, which are characteristics of mesoporous materials. The pore size distribution curves indicate the presence of ample mesopores with an average size of ∼20 nm for all the samples.¹⁷ The calculated Brunauer–Emmett–Teller (BET) specific surface area (S_BET) for pristine CNs is 48.2 m² g⁻¹. Incorporating Co into CNs nearly doubles the S_BET (92.5 m² g⁻¹) due to the generation of more micro- and mesopores, as evidenced by the increase of the total pore volume from 0.252 to 0.499 cm³ g⁻¹. This is probably due to the disturbance of the CN network structure by the incorporated Co species or the stabilization of micropores via introducing Co species.²⁶ After sulfurization, the S-Co-CNs catalyst shows slightly higher S_BET, pore size, and pore volume than Co-CNs, indicating that introducing larger S atoms to replace smaller O atoms at Co–N/C sites will further expand the pore size and thus enhance the S_BET but maintain the higher dispersion of in situ formed S-Co–C/N sites. The larger surface area and highly dispersed S-Co–C/N sites would be expected to provide a better H₂ evolution activity.

The microstructures of Co-CNs and S-Co-CNs were further investigated by transmission electron microscopy (TEM). The pristine CNs have a thin sheet structure with abundant pores (20–40 nm) on its surface (Fig. S1, ESI†), consistent with the result of N₂ adsorption–desorption analysis, and the thin sheets are folded at their edges due to the high surface energy. As shown in Fig. 3A, the Co-CNs exhibits a similar sheet structure to pristine CNs, indicating that the basic morphology does not change after Co incorporation. Additionally, no aggregated nanoparticles or clusters can be found on the entire surface of CNs from its high-resolution TEM images (Fig. 3B and S2A, ESI†), revealing that the Co dopants are distributed in the conjugated frameworks of CNs at the atomic level. This is further confirmed by the high angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and the corresponding energy-dispersive X-ray spectroscopy (EDX) mapping analysis (Fig. 3C), revealing that C, N, Co, and O are homogeneously dispersed in the CN matrix. According to the EDX analysis (Table S1, ESI†), the loading amount of Co in the Co-CNs catalyst is determined to be 0.18 at%. Fig. 3D shows the TEM image of S-Co-CNs, revealing that the thin layer sheet structure is well preserved after room-temperature sulfurization. As shown in the HRTEM images in Fig. 3E and S2B, ESI,† it can also be seen that no particles are formed, which indicates that the high dispersion of Co-N/C sites remains unchanged after sulfurization. Furthermore, the HAADF-STEM image and the relevant EDX mapping images in Fig. 3F further confirm a homogeneous distribution of the S element in the S-Co-CNs, while other elements maintain their original distributions without any aggregation. In addition, the EDX analysis indicates that the S content increased from zero to 0.19 at% for the resulting S-Co-CNs catalyst. These results show that the S atoms have been successfully coordinated to the Co-N/C sites and that the in situ formed S-Co–N/C maintains the high dispersion of original Co–N/C sites.

Fig. 3 (A) TEM and (B) HRTEM images of Co-CNs. (C) HAADF-STEM image of Co-CNs and the corresponding elemental maps (C, N, Co, and O). (D) TEM and (E) HRTEM images of S-Co-CNs. (F) HAADF-STEM images of S-Co-CNs and the corresponding elemental maps (C, N, Co, O, and S).

The electronic structures of Co-CNs and S-Co-CNs were investigated by using X-ray photoelectron spectroscopy (XPS). As shown in Fig. 4A, compared to Co-CNs that are composed of C, N, O, and Co elements, an additional S 2p signal can be found in the spectrum of the S-Co-CNs. This result is consistent with the observation of EDX analysis, further confirming the successful S incorporation via facile room-temperature sulfurization. For the Co 2p XPS spectrum of Co-CNs (Fig. 4B), the two Co 2p_3/2 peaks centered at 779.8 and 781.5 eV can be assigned to the Co–O and Co–N species, respectively. The appearance of Co–O species indicates the coordination of O atoms to Co–N/C sites as the Co-CNs was prepared in air.²⁵ The other two peaks at 784.5 and 788.1 eV are the satellite peaks of both Co–O and Co–N species.⁶⁰ In contrast to Co-CNs, besides the Co–O and Co–N species, the Co 2p XPS spectrum of S-Co-CNs shows a new Co 2p_3/2 peak at 778.9 eV, which can be identified as the peak of Co–S species.^61,62 In addition, the binding energies of Co–O (780.2 eV) and Co–N (781.8 eV) species for S-Co-CNs reveal a shift to higher binding energies due to S incorporation. The high-resolution S 2p XPS spectrum of S-Co-CNs further confirms the formation of S-coordinated Co species. As shown in Fig. 4C, the two peaks observed at 163.2 and 161.9 eV are assigned to 2p_1/2 and 2p_3/2 of Co–S species, and the two peaks at 167.4 and 168.7 eV correspond to 2p_1/2 and 2p_3/2 of surface-adsorbed S oxides.⁶³ These firmly confirm that O²⁻ coordinated to Co–N/C sites was partly replaced by S²⁻. In the N 1s spectra (Fig. 4D), the Co-CNs shows three main peaks at 397.7, 399.0 and 400.1 eV that correspond to the C–NC=C, N–C₃, and C–NH_x functional groups, respectively.¹⁷ All these peaks can also be observed for S-Co-CNs but slightly shift to lower binding energies. This is attributed to the fact that the S atoms are coordinated to Co and the electron transfer from CNs to in situ generated S-Co–N/C sites.

Fig. 4 (A) XPS survey spectra of Co-CNs and S-Co-CNs. (B) Co 2p XPS spectra of Co-CNs and S-Co-CNs. (C) S 2p XPS spectrum of S-Co-CNs. (D) N 1s XPS spectra of Co-CNs and S-Co-CNs.

The activities of the samples towards photocatalytic H₂ evolution were evaluated in a dye-sensitized system composed of xanthene dye Erythrosin B (ErB) as the photosensitizer and triethanolamine (TEOA) as the sacrificial reagent under visible light irradiation (λ ≥ 450 nm). As shown in Fig. 5A, the ErB– TEOA system shows no activity towards H₂ evolution in the absence of a catalyst. When CNs were added as the catalyst into the above system, only trace amounts of H₂ were detected. The low activity of CNs is attributed to the lack of active sites. Similarly, the Co-CNs also exhibits a very low H₂ evolution activity, demonstrating that the Co–N/C sites are inert for the H₂ evolution reaction though they have been reported to effectively facilitate the charge separation.^25,57 This result is consistent with previously reported results that an additional cocatalyst is still needed to be loaded on metal-doped CNs to promote the H₂ evolution reaction.^25,57 In strong contrast, the S-Co-CNs catalyst exhibits superior H₂ evolution activity to CNs and Co-CNs under the same reaction conditions. In a 5 h reaction, 616.5 μmol of H₂ is produced over S-Co-CNs catalyst in the ErB–TEOA system, corresponding to a turnover number (TON) of 62 based on ErB. Since there is no significant change in the textural properties (surface area, pore size, and pore volume) for the Co-CNs and S-Co-CNs, the significantly enhanced catalytic H₂ evolution activity of S-Co-CNs over Co-CNs can be attributed to the activation of Co–N/C sites by S incorporation, and the in situ formed S-Co–N/C sites are highly active for H₂ evolution. The time courses of H₂ evolution over ErB-sensitized S-Co-CNs catalysts prepared with different amounts of Co salt are given in Fig. S3, ESI.† It can be noted that H₂ evolves rapidly in the first 1 h of reaction over all of the S-Co-CNs catalysts. After 1 h of reaction, H₂ evolution becomes slow probably due to the degradation of ErB. Fig. 5B shows the initial H₂ evolution rates on S-Co-CNs catalysts sensitized by ErB. It is clear that the H₂ evolution rate increases linearly with increasing Co content from 5 to 40 mg. As shown in the XRD patterns of Co-CNs and S-Co-CNs catalysts (Fig. S4, ESI†), it is clear that no peaks of crystalline Co species for Co-CNs and Co sulfides for S-Co-CNs can be observed even at higher Co contents, indicating that the enhanced H₂ evolution activity with Co content is due to the formation of more S-Co–N/C active sites on CNs. Further increasing the Co content to 50 mg and higher will in turn deteriorate the activity most likely due to the aggregation of S-Co–N/C sites. In addition, the light-shielding caused by deep-colored S-Co-CNs catalysts at higher Co contents (inset in Fig. 5B) would also limit the light utilization of the reaction system and thus reduce the H₂ evolution activity. This facile room-temperature sulfurization strategy was also extended to prepare various transition metal-doped CNs (S-M-CNs, M = Cu, Zn, Ni) catalysts aiming to develop more active catalysts. The XRD patterns in Fig. S5, ESI† indicate that Cu and Zn can be homogeneously distributed on CNs, but the Ni dopant will aggregate to form Ni₇S₆ particles in the final S-Ni-CNs catalyst. Their activities towards H₂ evolution in the ErB–TEOA system were also tested. As shown in Fig. S6, ESI,† the as-prepared S-Cu-CNs and S-Zn-CNs show no H₂ evolution activity in the ErB–TEOA system, while the S-Ni-CNs exhibit a H₂ evolution rate of 1.81 mmol h⁻¹ g⁻¹, which is much lower than that obtained on S-Co-CNs (6.38 mmol h⁻¹ g⁻¹). These results further emphasize the importance of constructing highly active sites on CNs via simultaneously regulating the central metal and coordination structure.

Fig. 5 (A) Time course of H₂ evolution over different catalysts in the ErB–TEOA system. (B) Average H₂ evolution rates over S-Co-CN catalysts prepared with different masses of Co(NO₃)₂·6H₂O (5 to 60 mg) in the ErB–TEOA system. The inset shows the pictures of the S-Co-CNs catalysts. (C) Dependence of the H₂ evolution rate and of AQY on an ErB-sensitized S-Co-CNs catalyst as a function of wavelength of incident light. (D) Time course of cyclic H₂ evolution from the ErB-sensitized S-Co-CNs system. The inset shows the H₂ evolution from the Fl-sensitized S-Co-CNs system. Reaction conditions: catalyst, 50 mg; 100 mL of TEOA solution, 10%, pH 8; dye, 0.2 mM; light source, 30 W LED lamp; λ ≥ 450 nm.

The high catalytic activity of the S-Co-CNs catalyst was further demonstrated by measuring the apparent quantum yield (AQY) of H₂ evolution under visible light irradiation of different wavelengths. As displayed in Fig. 5C, the H₂ evolution rate and the corresponding AQY first increase and then decrease with the increase of wavelength of incident light from 450 to 650 nm, which is fully consistent with the absorption spectrum of ErB. This indicates that ErB is a photosensitizer for capturing light and the S-Co-CNs act as a H₂ evolution catalyst. At 520 nm, both the H₂ evolution rate and AQY reach the highest values of 129.1 μmol h⁻¹ and 13.02%, respectively. These values are among the best obtained on dye-sensitized g-C₃N₄-based catalysts under visible light irradiation (Table S2, ESI†), further demonstrating that our S-Co-CNs catalyst, albeit with very low Co content, is a highly efficient cocatalyst for photocatalytic H₂ evolution in a dye-sensitized system.

The catalytic stability of S-Co-CNs catalyst towards H₂ evolution in the ErB–TEOA system was also evaluated by performing a continuous repeated reaction over 24 h, as shown in Fig. 5D. As can be seen the system exhibits a steady H₂ evolution activity in the first 6 h, after which the activity decreases with time, which is due to the gradual degradation of dye ErB (Fig. S7, ESI†). After the first run reaction, the S-Co-CNs catalyst was collected, washed, and redispersed into fresh ErB–TEOA solution for the next run reaction. Although the activity of the system decreases more significantly in the following runs compared to the first run, the S-Co-CNs catalyst is still active. The results of XRD (Fig. 6A) and FTIR (Fig. 6B) analyses reveal that the chemical structures of CNs and the high dispersion nature of S-Co–N/C sites are well retained in the used S-Co-CNs catalyst. XPS analysis (Fig. 6C and D) shows that the chemical states of the S-Co–N/C catalyst are also well maintained after the stability test. Specifically, the Co–S species can still be observed in the Co 2p and S 2p XPS spectra of the used S-Co-CNs catalyst. In addition, TEM analysis reveals that the microstructure of the used S-Co-CNs catalyst is similar to that of the fresh one, and no particle aggregation can be observed (Fig. S8, ESI†). These results reveal that the S-Co-CNs catalyst is robust with well-retained highly dispersed S-Co–N/C active sites during the long-term H₂ evolution in the ErB–TEOA system. To improve the stability of the dye-sensitized S-Co-CNs system for H₂ evolution, a more stable dye fluorescein (FL) was used to replace ErB. As shown in the inset of Fig. 5D, a steady H₂ evolution activity is obtained during a continuous reaction for a period of 24 h, further demonstrating the good catalytic stability of the S-Co-CNs catalyst.

Fig. 6 (A) XRD pattern, (B) FTIR spectrum, and high-resolution XPS spectra of (C) Co 2p and (D) S 2p of the S-Co-CNs catalyst after the stability test.

To elucidate the origin of the high activity of the S-Co-CNs catalyst, the electrocatalytic H₂ evolution experiments were first carried out by linear sweep voltammetry (LSV). As shown in the LSV plots in Fig. 7A, the overpotential to achieve a current density of 10 mA cm⁻² on S-Co-CNs catalyst is 0.56 V, much lower than those of Co-CNs (0.63 V) and pristine CNs (0.86 V). This result indicates that the S incorporation is effective to activate the Co–N/C sites for enhancing the H₂ evolution activity, which is consistent with previous observations that the Co sulfides are typically highly active catalysts for electrocatalytic and photocatalytic H₂ evolution reactions. Electrochemical impedance spectroscopy (EIS) measurements were then performed to further disclose the origin of the activity enhancement on the S-Co-CNs catalyst (Fig. 7B). It can be seen that the Co-CNs catalyst shows a much smaller semicircle diameter and a much lower interfacial charge-transfer resistance (R_ct = 38.0 kΩ) than those of pristine CNs (R_ct = 113.3 kΩ), indicating that doping of Co into the CNs will significantly reduce the electron transfer resistance.²⁵ After sulfurization treatment, the charge transfer resistance of S-Co-CNs is further reduced (R_ct = 21.3 kΩ), which is beneficial for improving electron transfer from the excited dye to S-Co–N/C active sites in the ErB–TEOA system and thus leading to a better H₂ evolution activity. The efficient charge separation promoted by Co–N/C and S-Co–N/C active sites was further revealed by PL analysis. As demonstrated in Fig. 7C, the strong PL emission of CNs in the wavelength range of 400–650 nm can be significantly quenched after the Co–N/C sites were embedded in the CNs, indicating that the charge recombination rate on CNs is greatly restricted due to the presence of Co–N/C sites.⁵⁸ Time-resolved PL (TRPL) spectra (Fig. 7D) recorded at the corresponding steady-state emission peaks further reveal the average lifetimes (τ_ave) of approximately 12.5 and 5.8 ns for pristine CNs and Co-CNs, respectively. This greatly decreased lifetime for charge carriers firmly reveals significantly enhanced charge separation in the Co-CNs due to the presence of Co–N/C sites.⁵⁸ These results further confirm that embedding of Co–N/C sites is an effective strategy to promote the photoexcitation dissociation. After the S incorporation, the resulting S-Co-CNs catalyst, however, shows a slightly higher PL emission intensity compared to Co-CNs. Moreover, the PL emission lifetime of the S-Co-CNs catalyst (6.0 ns) is nearly the same as that of the Co-CNs catalyst. These results reveal that the efficient charge separation nature of Co-N/C sites is well-retained after the S incorporation, and the enhanced H₂ evolution activity achieved on S-Co-CNs with respect to Co-CNs can be attributed to the enhanced intrinsic catalytic activity of S-Co–N/C towards the H₂ evolution reaction, as evidenced by the LSV results. Furthermore, the Mott–Schottky (MS) measurements were performed to determine the possible change of the CB (E_CB) and VB (E_VB) potentials of CNs after Co doping and subsequent room-temperature sulfurization (Fig. S9, ESI†). The E_CB of CNs, Co-CNs, and S-Co-CNs is determined to be −0.91, −0.86, and −0.83 V vs. Ag/AgCl, respectively. Based on the above results and the bandgap energy obtained from the UV-vis-DRS in Table 1, the VB potentials (E_VB) were calculated to be +1.92, +1.91, and +1.94 V vs. Ag/AgCl for CNs, Co-CNs, and S-Co-CNs, respectively. These results clearly confirm that the embedding of Co-N/C sites and subsequent sulfurization will slightly reduce the E_CB of CNs due to the strong interactions between the CN matrix and embedded S-Co–N/C active sites, which would be beneficial for the fast electron transfer from excited ErB to S-Co-CNs for further enhancing the photocatalytic H₂ evolution activity. Based on the above results, the efficient photocatalytic H₂ evolution catalyzed by S-Co-CNs in ErB–TEOA can be illustrated in Fig. 8. Upon visible light irradiation, ErB is excited to be ErB*. Owing to the estimated reduction potential of ErB of −0.91 V vs. NHE (1.15 V vs. SCE),⁶⁴ and the E_CB of S-Co-CNs of −0.63 V vs. NHE (−0.83 V vs. Ag/AgCl), the electrons will efficiently transfer from ErB* to the CB of S-Co-CNs catalyst. Owing to the efficient electron separation nature of S-Co-CNs, the accepted electrons can be rapidly transferred to the embedded S-Co–N/C sites and effectively catalyze the H₂O/H⁺ reduction to H₂, while the formed ErB⁺ can be regenerated to ErB by TEOA to complete the catalytic cycle.

Fig. 7 (A) H₂ evolution LSV plots and (B) Nyquist plots of pristine CNs, Co-CNs, and S-Co-CNs recorded in 10% TEOA (pH 8) solution containing 0.5 M Na₂SO₄. The inset shows the equivalent circuit used for data fitting. (C) Steady-state PL spectra (excitation wavelength: 380 nm) and (D) time-resolved PL decay profiles of pristine CNs, Co-CNs, and S-Co-CNs.

Fig. 8 Schematic illustration of the band alignment and the electron transfer from ErB to CNs and the H₂ evolution on the S-Co–N/C sites under visible light irradiation.

Conclusions

In summary, we have proposed a facile and effective strategy to activate inert Co–N/C sites on CNs to highly active S-Co–N/C sites for catalyzing the photochemical H₂ evolution reaction via a facile and effective room-temperature sulfurization method. Attributed to the mild sulfurization conditions, the high dispersion of active sites can be well preserved, and the obtained S-Co-CNs catalysts render excellent catalytic activity for H₂ evolution in an ErB-sensitized system under visible light irradiation (λ ≥ 450 nm). The most active S-Co-CNs catalyst shows a high H₂ evolution rate of 6.38 mmol h⁻¹ g⁻¹ and a highest AQY of 13.02% at 520 nm, while both pristine CNs and Co-CNs are totally inactive. The S incorporation is found to effectively enhance the intrinsic activity of Co–N/C sites while maintaining the efficient charge separation nature of the active sites. This work affords a simple but effective method for the fabrication of highly dispersed and highly efficient active sites for the H₂ evolution reaction on the 2D g-C₃N₄ matrix and is expected to provide new insights into the materials design of transition metal-based catalysts and their applications in various energy-conversion-related photo/electrocatalysis fields.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (22162001), the Natural Science Foundation of Ningxia Province (2021AAC03201), the Foundation of Academic Top-notch Talent Support Program of North Minzu University (2019BGBZ08), the Leading Talents Program of Science and Technology Innovation in Ningxia Province (2020GKLRLX14), the West Light Foundation of the Chinese Academy of Sciences (XAB2018AW13), and the Innovation and Entrepreneurship Projects for Returnees of Ningxia Province.

References

1. X. B. Chen, S. H. Shen, L. J. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570.
2. Y. Ma, X. L. Wang, Y. S. Jia, X. B Chen, H. X Han and C. Li, Chem. Rev., 2014, 19, 9987–10043.
3. L. Lei, W. J. Wang, Z. F. Xie, X. B. Wu, A. K. Yadav, P. M. Buschbaum and H. Q. Fan, Sustainable Energy Fuels, 2021, 5, 5227–5235.
4. Y. Q. Xing, Z. R. Tan, J. Z. Cheng, Z. Q. Shen, Y. J. Zhang and S. Y. Liu, Sustainable Energy Fuels, 2021, 5, 5166–5174.
5. S. Wageh, A. A. Al-Ghamdi, R. Jafer, X. Li and P. Zhang, Chin. J. Catal., 2021, 42, 667–669.
6. R. C. Shen, D. D Ren, Y. N. Ding, Y. T. Guan, Y. H. Ng, P. Zhang and X. Li, Sci. China Mater., 2020, 63, 2153–2188.
7. Z. Z. Liang, R. C. Shen, Y. H. Ng, P. Zhang, Q. J. Xiang and X. Li, J. Mater. Sci. Technol., 2020, 56, 89–121.
8. R. C. Shen, J. Xie, Q. J. Xiang, X. B. Chen, J. Z. Jiang and X. Li, Chin. J. Catal., 2019, 40, 240–288.
9. X. C. Wang, K. Maeda, A. Thomas, K. Takanabe, G. Xin, J. M. Carlsson, K. Domen and M. Antonietti, Nat. Mater., 2008, 8, 76–80.
10. X. C. Wang, K. Maeda, X. F. Chen, K. Takanabe, K. Domen, Y. D. Hou, X. Z. Fu and M. Antonietti, J. Am. Chem. Soc., 2009, 131, 1680–1681.
11. Y. Zheng, J. Liu, J. Liang, M. Jaroniec and S. Z. Qiao, Energy Environ. Sci., 2012, 5, 6717–6731.
12. X. C. Wang, S. Blechert and M. Antonietti, ACS Catal., 2012, 2, 1596–1606.
13. Y. Wang, X. C. Wang and M. Antonietti, Angew. Chem., Int. Ed., 2012, 51, 68–69.
14. S. W. Cao and J. G. Yu, J. Phys. Chem. Lett., 2014, 5, 2101–2107.
15. Z. Chen, S. Zhang, J. Yang, C. Chen, Y. C. Song, C. L. Xu, M. Q. Wu and J. X. Liao, Sustainable Energy Fuels, 2021, 5, 4973–4980.
16. K. Takanabe, K. Kamata, X. C. Wang, M. Antonietti, J. Kubota and K. Domen, Phys. Chem. Chem. Phys., 2010, 12, 13020–13025.
17. S. X. Min and G. X. Lu, J. Phys. Chem. C, 2012, 116, 19644–19652.
18. J. Y. Xu, Y. X. Li, S. Q. Peng, G. X. Lu and S. B. Li, Phys. Chem. Chem. Phys., 2013, 15, 7657–7665.
19. W. J. Wang, H. Xu, G. X. Yu, D. M. Chen, S. Q. Sun, M. Zhou and J. Chen, Sustainable Energy Fuels, 2021, 5, 5247–5256.
20. Y. N. Liu, C. C. Shen, N. Jiang, Z. W. Zhao, X. Zhou, S. J. Zhao and A. W. Xu, ACS Catal., 2017, 7, 8228–8234.
21. Y. Xue, Y. G. Lei, X. Y. Liu, Y. N. Li, W. A. Deng, F. Wang and S. X. Min, New J. Chem., 2018, 42, 14083–14086.
22. Y. F. Li, M. Zhang, L. Zhou, S. J. Yang, Z. S. Wu and Y. H. Ma, Acta Phys. -Chim. Sin., 2021, 37, 2009030.
23. X. C. Wang, X. F. Chen, A. Thomas, X. Z. Fu and M. Antonietti, Adv. Mater., 2009, 21, 1609–1612.
24. X. F. Chen, J. S. Zhang, X. Z. Fu, M. Antonietti and X. C. Wang, J. Am. Chem. Soc., 2009, 131, 11658–11659.
25. G. G. Zhang, C. J. Huang and X. C. Wang, Small, 2015, 11, 1215–1221.
26. G. G. Zhang, S. H. Zang, L. H. Lin, Z. A. Lan, G. S. Li and X. C. Wang, ACS Appl. Mater. Interfaces, 2016, 8, 2287–2296.
27. G. Liu, P. Niu, C. H. Sun, S. C. Smith, Z. G. Chen, G. Q. Lu and H. M. Cheng, J. Am. Chem. Soc., 2010, 132, 11642–11648.
28. J. H. Zhang, J. H. Sun, K. Maeda, K. Domen, P. Liu, M. Antonietti, X. Z. Fu and X. C. Wang, Energy Environ. Sci., 2014, 39, 15373–15379.
29. C. H. Ng, S. H. Teo, N. Mansir, A. Islam, C. G. Joseph, S. Hayased and Y. H. Taufiq-Yap, Sustainable Energy Fuels, 2021, 5, 4457–4511.
30. Y. P. Zhu, T. Z. Ren and Z. Y. Yuan, ACS Appl. Mater. Interfaces, 2015, 7, 1685–1686.
31. X. J. Shen, L. Liu, H. Y. Ji, Z. Mo, Y. P. Li, L. Y. Huang, D. L. Du, H. Xu and H. M. Li, Appl. Catal., B, 2016, 187, 144–153.
32. Z. Z. Lin and X. C. Wang, Angew. Chem., Int. Ed., 2013, 52, 1735–1738.
33. D. J. Martin, K. P. Qiu, S. A. Shevin, A. D. Handoko, X. W. Chen, Z. X. Guo and J. W. Tang, Angew. Chem., Int. Ed., 2014, 53, 9240–9245.
34. Y. Yu, W. Yan, X. F. Wang, P. Li, W. Y. Gao, H. H. Zou, S. M. Wu and K. J. Ding, Adv. Mater., 2018, 30, 1705060.
35. J. H. Sun, J. S. Zhang, M. W. Zhang, M. Antonietti, X. Z. Fu and X. C. Wang, Nat. Commun., 2012, 3, 1139–1146.
36. C. Han, Y. D. Wang, Y. P. Lei, B. Wang, N. Wu, Q. Shi and Q. Li, Nano Res., 2015, 8, 1199–1209.
37. R. Q. Ye, H. B. Fang, Y. Z. Zheng, N. Li, Y. Wang and X. Tao, ACS Appl. Mater. Interfaces, 2016, 8, 13879–13889.
38. H. J. Yan and Y. Huang, Chem. Commun., 2011, 47, 4168–4170.
39. J. J. Duan, S. Chen, M. Jaroniec and S. Z. Qiao, ACS Nano, 2015, 9, 931–940.
40. L. Xu, W. Q. Huang, L. L. Wang, Z. A. Tian, W. Y. Hu, Y. M. Ma, X. Wang, A. L. Pan and G. F. Huang, Chem. Mater., 2015, 27, 1612–1621.
41. T. Ishaq, M. Yousaf, L. A. Bhattia, M. Ahmadc, M. Ikramd, M. UsmanKhanef and A. Qayyuma, Int. J. Hydrogen Energy, 2020, 45, 31574–31584.
42. D. D. Ren, W. N. Zhang, Y. N. Ding, R. C. Shen, Z. M. Jiang, X. Y. Lu and X. Li, Solar PRL, 2020, 4, 1900423.
43. C. Liu, Y. Feng, Z. T. Han, Y. Sun, X. Q. Wang, Q. F. Zhang and Z. G. Zou, Chin. J. Catal., 2021, 42, 164–174.
44. H. T. Xu, R. Xiao, J. R. Huang, Y. Jiang, C. X. Zhao and X. Yang, Chin. J. Catal., 2021, 42, 107–114.
45. F. H. Li, R. R. Zhao, B. Y. Yang, W. Wang, Y. Liu, J. P. Gao and Y. L. Gong, Int. J. Hydrogen Energy, 2019, 44, 30185–30195.
46. L. Ge, C. C. Han, J. Liu and Y. F. Li, Appl. Catal., A, 2011, 409, 215–222.
47. R. C. Shen, K. L. He, A. P. Zhang, N. Li, Y. H. Ng, P. Zhang, J. Hu and X. Li, Appl. Catal., B, 2021, 291, 120104.
48. M. Wang, J. J. Cheng, X. F. Wang, X. K. Hong, J. J. Fan and H. G. Yu, Chin. J. Catal., 2021, 42, 37–45.
49. D. D. Zheng, G. G. Zhang, Y. D. Hou and X. C. Wang, Appl. Catal., A, 2016, 521, 2–8.
50. D. Q. Zeng, T. Zhou, W. J. Ong, M. D. Wu, X. G. Duan, W. J. Xu, Y. Z. Chen, Y. A. Zhu and D. L. Peng, ACS Appl. Mater. Interfaces, 2019, 11, 5651–5660.
51. M. Ghaemmaghami and R. Mohammadi, Sustainable Energy Fuels, 2019, 3, 2176–2204.
52. H. T. Li, M. Wang, Y. P. Wei and F. Long, J. Colloid Interface Sci., 2019, 534, 343–349.
53. R. C. Zhen, J. Xie, X. Y. Lu, X. B. Chen and X. Li, ACS Sustainable Chem. Eng., 2018, 6, 4026–4036.
54. R. C. Zhen, J. Xie, H. D. Zhang, A. P. Zhang, X. B. Chen and X. Li, ACS Sustainable Chem. Eng., 2018, 6, 816–826.
55. N. Zhao, L. G. Kong, Y. M. Dong, G. L. Wang, X. M. Wu and P. P. Jiang, ACS Appl. Mater. Interfaces, 2018, 10, 9522–9531.
56. L. F. Gao, T. Wen, J. Y. Xu, X. P. Zhai, M. Zhao, G. W. Hu, P. Chen, Q. Wang and H. L. Zhang, Appl. Surf. Sci., 2017, 392, 608–615.
57. P. W. Chen, K. Li, Y. X. Yu and W. D. Zhang, Appl. Surf. Sci., 2017, 392, 608–615.
58. W. Lin, L. L. Gao, W. R. Cheng, Y. J. Cao, X. K. Liu, W. Zhang, X. L. Mou, L. L. Jin, X. S. Zheng, W. C. Prof, T. Yao and S. Q. Wei, Angew. Chem., Int. Ed., 2017, 56, 9312–9317.
59. Y. Chen, B. Lin, W. L. Yu, Y. Yang, H. Huang, K. Z. Takanabe and J. M. Basset, Chem.–Eur. J., 2015, 21, 10290–10295.
60. X. R. Wang, J. Y. Liu, Z. W. Liu, W. C. Wang, J. Luo, X. P. Han, X. W. Du, X. Z. Qiao and J. Yang, Adv. Mater., 2018, 30, 1800005.
61. P. W. Cai, J. H. Huang, J. X. Chen and Z. H. Wen, Angew. Chem., Int. Ed., 2017, 56, 4858–4861.
62. S. H. Chang, M. D. Lu, Y. L. Tung and H. Y. Tuan, ACS Nano, 2013, 7, 9443–9451.
63. C. B. Ouyang, X. Wang and S. Y. Wang, Chem. Commun., 2015, 51, 14160–14163.
64. J. Y. Xu, Y. X. Li and S. Q. Peng, Int. J. Hydrogen Energy, 2015, 40, 353–362.

---

Footnote

† Electronic supplementary information (ESI) available: Additional figures. See DOI: 10.1039/d1se01702k

---