Construction of Curly-like CN@CdS Z-Scheme Heterojunction to Boost Visible-Light-Driven H₂O₂ Evolution

Abstract

Photocatalytic H₂O₂ production with H₂O and O₂ as resources is a promising technique. Herein, a g-C₃N₄@CdS (CN@CdS) Z-scheme heterojunction is prepared for photocatalytic H₂O₂ production with a rate of 167.5 μmol/h, which is 5.43 and 5.15 times higher than that of pure g-C₃N₄ and pure CdS. Photoelectronic characterization results reveal the existence of a strong built-in electric field between CdS and g-C₃N₄, which significantly facilitates the separation of photogenerated carriers and preserves the strong redox capacity. Density functional theory (DFT) calculations show that CN@CdS displays lower O₂ adsorption energy and the H₂O₂ is more readily formed. This work provides a novel strategy for the design of photocatalysts with excellent H₂O₂ evolution efficiency.

1. Introduction

As one of the most important chemical substances in the world, H₂O₂ is widely applied in the paper industry, chemical synthesis, antimicrobial bleaching and wastewater treatment [1]. Currently, H₂O₂ is produced through an oxidation of anthraquinone process, which requires the direct use of hydrogen and oxygen gases and generates wastewater, exhaust gas and solid waste [2,3]. Photocatalytic H₂O₂ production, using H₂O and O₂ as raw materials and solar energy as the only driving force, has attracted attention for its green, economic and safety advantages [4].

Graphite-phase carbon nitride (g-C₃N₄), a graphite-phase nonmetallic polymer with tris-s-triazine as a repeating unit, is considered to be a common and effective photocatalyst. Since Wang et al. [5] first reported photocatalytic overall water splitting by g-C₃N₄ under visible light irradiation, g-C₃N₄ has been used for various fields. However, pristine g-C₃N₄ leads to inferior photocatalytic H₂O₂ production activity owing to disadvantages such as the high complexation efficiency of photogenerated carriers, small specific surface area and weak redox capacity. Meanwhile, single-component photocatalysts usually cannot simultaneously satisfy requirements for strong light absorption capacity, strong redox properties and efficient photogenerated carrier separation. Constructing a Z-scheme heterojunction has been an effective approach to improve the photocatalytic performance of g-C₃N₄-based photocatalysts [6]. Unlike conventional type-II heterojunctions, Z-scheme heterojunctions retain strongly reducing conduction-band electrons and strongly oxidizing valence-band holes in addition to improved charge separation efficiency, thus providing both highly efficient charge separation and strong redox capability [7].

Metal sulfides (CdS, MoS₂, ZnS, etc.) have attracted much attention in the photocatalytic field due to their unique physicochemical properties [8]. In particular, CdS is a promising semiconductor material owing to its excellent band gap (≈2.4 eV) and superior response to visible light [9]. The d orbitals of Cd in CdS contribute to a significant role in the formation of the valence band (VB), which facilitates electron transfer in photocatalysis [10]. The suitable energy band position of CdS makes it an excellent choice for coupling with g-C₃N₄ to construct a Z-scheme heterojunction. Zhang et al. [11] prepared a novel Z-scheme O-CN/CdS heterojunction by the hydrothermal method. It is found that constructing a Z-scheme heterojunction promotes the separation of photogenerated carriers, which greatly enhances photocatalytic performance. Close interfacial contact between heterojunctions improves charge transport pathways, thereby facilitating interfacial charge separation and transfer. Lin et al. [12] designed and synthesized a novel defective C₃N₄/CdS direct Z-scheme heterojunction with significantly enhanced light absorption, redox properties, charge separation efficiency and carrier lifetime. However, the above Z-scheme heterojunction application process only utilizes the conduction band electrons for the target product synthesis, and the holes on the valence band of CdS are not utilized for production. Therefore, it is necessary to develop a photocatalyst that utilizes both a conduction band and valence band for the synthesis of target products.

It has been shown that co-modification can effectively improve the photocatalytic efficiency. A curly-like photocatalyst enables multiple reflections to maximize the utilization of light [13]. In this study, we constructed curly-like CN@CdS Z-scheme heterojunction photocatalysts for visible-light photocatalytic H₂O₂ production. The mechanism of photocatalytic H₂O₂ production from CN@CdS photocatalysts was revealed by DFT calculations, active group capture experiments and ESR characterization. The convoluted structure of the photocatalyst supplies more adsorption and active sites for the reaction, and the Z-scheme charge transfer pathway retains the strong redox ability of the photocatalyst. Therefore, the CN@CdS heterojunction exhibited significantly improved yields for photocatalytic H₂O₂ production. This study provides a rational design pathway for the construction of the Z-scheme heterojunction of g-C₃N₄-based photocatalysts aimed at enhancing charge separation efficiency and boosting photocatalytic H₂O₂ production.

2. Results

2.1. Structural Characterizations

The crystal structures of g-C₃N₄, CdS, CN@CdS-1, CN@CdS-2 and CN@CdS-3 were analyzed by XRD. As shown in Figure 1a, two characteristic peaks of g-C₃N₄ at 2θ = 12.9° and 27.7° can be observed, which correspond to the (100) and (002) crystal planes of g-C₃N₄ (JCPDS No. 87-1526), assigned to the in-plane structural stacking motifs of tri-s-triazine in g-C₃N₄ and the interplanar spacing between C₃N₄ layers, respectively [14]. The XRD spectrum of CdS shows diffraction peaks at 2θ = 25.1°, 26.2°, 28.3°, 43.9°, 48.0°and 51.9°, which correspond to the (100), (002), (101), (110), (103) and (112) crystal planes of the hexagonal wurtzite structure of CdS (JCPDS No. 41-1049) [15]. In addition, a series of CN@CdS photocatalysts were prepared with good crystalline phase matching with CdS and g-C₃N₄. For CN@CdS composites, the peak intensity of the CdS phase increases with its content. No impurity diffraction peaks are detected, indicating that no other chemical reactions occurred during the electrostatic self-assembly of g-C₃N₄ with CdS nanoparticles. To further understand the chemical state of CN@CdS composites, XPS was employed. The XPS survey spectrum (Figure 1b) reveals that C, N, Cd, S and O elements co-exist in CN@CdS-2. The observed presence of a small amount of elemental O can be attributed to the adsorption of H₂O on the surface of CN@CdS-2. As depicted in the high-resolution spectrum of C 1s (Figure 1c), the characteristic peaks at 284.7 eV and 287.9 eV belong to graphitic carbon (C=C) and sp²-bonded carbon (N-C=N), respectively [16]. The three characteristic peaks of N 1s spectra (Figure 1d) are attributed to the sp² hybridized nitrogen (C-N=C), N-(C)₃ group and π-excited state [17]. In Figure 1e, the Cd 3d high-resolution spectra have two characteristic peaks corresponding with Cd 3d_5/2 and Cd 3d_3/2 of Cd²⁺ at 405.2 and 412.0 eV [18,19]. Similarly, the binding energies of S 2p (Figure 1f) bimodal peaks are located at 163.4 and 161.8 eV [20]. The XPS results further demonstrate the favorable loading of CdS nanoparticles on CN to form a CN@CdS heterojunction.

SEM, TEM and HRTEM were employed to observe the microscopic morphology of samples. Figure 2a shows SEM image of CdS, and it can be clearly seen that a large number of CdS nanospheres with diameters of about 100 nm were aggregated together to form CdS nanospheres with diameters of about 500 nm. Figure 2b shows the SEM image of g-C₃N₄, in which it is obvious that g-C₃N₄ has a curly structure. This structure can expose a higher surface area, which can reflect the light multiple times in the photocatalyst to maximize the utilization of the light source. In addition, it also can shorten the transmission distance of the photogenerated electrons and inhibit the complexation of the photogenerated carriers, which promotes the improvement of the photocatalytic performance. Figure 2c shows an SEM image of CN@CdS-2. After self-assembly by electrostatic force, close interfacial contacts are formed between CdS nanospheres and g-C₃N₄. Compared with pure CdS, the size of CdS nanospheres in CN@CdS-2 is significantly reduced, which reflects the positive effect of g-C₃N₄ in enhancing the dispersion of CdS and effectively suppressing the agglomeration of CdS nanospheres. Figure 2d,e show TEM images of g-C₃N₄ and CN@CdS-2, respectively, which are consistent with SEM results, exhibiting the curly structure of g-C₃N₄ and nano-microsphere structure of CdS. The HRTEM image of CN@CdS-2 is illustrated in Figure 2f, where two crystalline planes with lattice spacings of 0.34 and 0.31 nm can be observed, corresponding to the (002) and (101) crystalline planes of CdS [15,21]. The amorphous region is attributed to g-C₃N₄. The above results provide strong proof of the successful anchoring of CdS nanospheres on the g-C₃N₄ surface.

2.2. Optical and Photoelectrochemical Characterization

The light absorption properties of CdS, g-C₃N₄ and CN@CdS were investigated by UV-Vis DRS. The band edges of g-C₃N₄ and CdS are 452 and 583 nm, respectively, as seen in Figure 3a. As expected, the band edges of the CN@CdS heterojunction move to longer wavelengths with the increase in CdS loading. This indicates that the construction of the CN@CdS heterojunction can enhance the light absorption ability, which is more favorable to enhance the photocatalytic activity. The band gap (E_g) of each catalyst can be obtained by converting the UV-Vis DRS (Figure 3a) into Trac spectra (Figure 3b). The E_g can be derived from the Kubellka–Munk equation ($\alpha \mathrm{h}\nu =\mathrm{A}{(\mathrm{h}\nu -{\mathrm{E}}_{\mathrm{g}})}^{{\displaystyle \frac{\mathrm{n}}{2}}}$, where α, A, h, ν and E_g are the absorption coefficient, scaling factor, Planck coefficient, optical frequency and band gap. Both g-C₃N₄ and CdS are direct bandgap semiconductors, so n = 1. From the above equation, it can be seen that by plots (αhν)^1/2 versus hν, the slope of the curves is E_g. For g-C₃N₄ and CdS, the E_g are 2.75 and 2.30 eV (Figure 3b). For the CN@CdS heterojunction, the E_g of three catalysts are in between g-C₃N₄ and CdS. The E_g of CN@CdS-1, CN@CdS-2 and CN@CdS-3 are 2.74, 2.73 and 2.71 eV, respectively. The flat potential (E_fb) of g-C₃N₄ and CdS are measured based on Mott Schottky (Figure 3c). The slopes of g-C₃N₄ and CdS are both positive, consistent with an n-type semiconductor. The E_fb of g-C₃N₄ and CdS are −1.17 and −0.40 eV (vs. NHE), respectively. Usually the E_CB of an n-type semiconductor approximates the E_fb. The valence band positions (E_VB) of g-C₃N₄ and CdS can be calculated from E_g = E_VB − E_CB to be 1.58 eV and 1.90 eV (vs. NHE), respectively.

DFT was used to investigate the interfacial charge transfer in the heterojunction. Combining the results of HRTEM and XRD, we selected the (101) crystal plane of CdS and the (002) crystal plane of g-C₃N₄ for DFT calculations. As can be observed from Figure 4a,b, the work functions (Φ) of the g-C₃N₄ (002) surface and CdS (101) surface are 5.657 and 6.936 eV, and the Fermi energy (E_f) level of CdS is higher than that of g-C₃N₄ (Figure 4c). Upon contact between the two, the CB of g-C₃N₄ bends upward at the interface while the CB of CdS bends downward, induced by an internal electric field (IEF). Upon photoexcitation, an e⁻ is spontaneously transferred from g-C₃N₄ to CdS under the effect of IEF, coulomb attraction and energy band bending until the two Fermi energy levels reach equilibrium. As a result, an e⁻ on the CB of CdS binds to h⁺ on the VB of g-C₃N₄ due to the presence of the IEF. The formation of a Z-scheme heterojunction between CdS and g-C₃N₄ improves the photogenerated carrier separation rate and maintains a high redox activity.

The effect of the Z-scheme heterojunction on photogenerated carrier separation efficiency has been investigated by PL and transient photocurrent response. Compared with g-C₃N₄, CN@CdS-2 exhibits remarkably low intensity, indicating the construction of the heterojunction effectively suppresses the complexation rate of the photogenerated carrier (Figure 5a). The PL spectrum of g-C₃N₄ shows a small peak at ~ 450 nm in addition to a main peak at 430 nm. The reason for the double peak may be due to the defect states derived from the structure, which form intermediate energy levels in the forbidden band, resulting in low-energy emission peaks [22]. The PL peak position also appears red-shifted, which is consistent with the results of the UV-Vis DRS. In addition, the prepared samples were also analyzed for photocurrent to compare their photogenerated carrier separation ability (Figure 5b). The highest photocurrent density is observed for CN@CdS-2, indicating the highest separation efficiency of its photogenerated carriers. All of the above results indicate that construction of the Z-scheme heterojunction can significantly enhance the separation efficiency of the photogenerated carriers and thus possibly boost the photocatalytic activity of the photocatalyst.

2.3. Photocatalytic Performance of H₂O₂ Production

The photocatalytic performance of prepared samples was evaluated by photocatalytic H₂O₂ production. The photocatalytic reaction was carried out under air conditions without adding any sacrificial agent and the specific experimental steps are illustrated in the Materials and Methods section. Figure 6a exhibits curves of H₂O₂ produced by g-C₃N₄, CdS, CN@CdS-1, CN@CdS-2 and CN@CdS-3 with time. The concentrations of H₂O₂ increase with the longer irradiation time for all samples, indicating that all of them have the ability to produce H₂O₂ in the photocatalytic reaction system. After irradiation for 2 h, CN@CdS series samples all show superior photocatalytic H₂O₂ production performance to that of g-C₃N₄ and CdS. CN@CdS-2 exhibits the highest H₂O₂ yield, reaching up to 167.5 μM/h, which is 5.43 and 5.15 times higher than that of g-C₃N₄ and CdS, respectively. The excellent photocatalytic performance of CN@CdS may be attributed to the construction of the Z-scheme heterojunction, which enhances the light-absorption capacity and separation efficiency of photogenerated carriers. In addition, its curled structure multiplies adsorption and active sites for O₂ and H₂O, which promotes its visible-light photocatalytic H₂O₂ production performance. The yield of CN@CdS-3 is inferior to that of CN@CdS-2, which is probably due to the fact that the excessive loading of CdS will cover the active sites of the photocatalyst, leading to an inhibition of photocatalytic H₂O₂ production performance. To investigate whether the prepared samples could maintain high photocatalytic H₂O₂ production activity in real water samples, photocatalytic H₂O₂ production experiments were carried out in seawater and pure water. As illustrated in Figure 6b, g-C₃N₄ and CN@CdS-2 both have greater H₂O₂ yields in seawater than in pure water. CN@CdS-2 has an improved H₂O₂ yield from 167.5 to 287.5 μM/h. This result can be attributed to the ability of cations in seawater to play a facilitating role in photocatalytic reactions [23].

The stability of CN@CdS-2 was investigated. As demonstrated in Figure 6c, the photocatalytic H₂O₂ production of CN@CdS-2 can still maintain relative stability after three cycling experiments. The decrease in production during cycling may be attributed to unavoidable loss during recycling. The photocatalytic H₂O₂ production is normally associated with a decomposition of H₂O₂; hence, a kinetic model is developed to describe it. The production rate constant (K_f) and decomposition rate constant (K_d) of H₂O₂ are in accordance with the kinetic equations of zero and first order, respectively [24].

$$\left[{\mathrm{H}}_2{\mathrm{O}}_2\right]={\displaystyle \frac{{\mathrm{K}}_{\mathrm{f}}}{{\mathrm{K}}_{\mathrm{d}}}}\left(1-\exp \left(-{\mathrm{K}}_{\mathrm{d}}\times \mathrm{t}\right)\right)$$

(1)

$${\mathrm{K}}_{\mathrm{d}}=-\ln \left({\mathrm{C}}_{\mathrm{t}}/{\mathrm{C}}_0\right)/\mathrm{t}$$

(2)

where K_f and K_d are the formation and decomposition rate constants, respectively. As shown in Figure 6d, CN@CdS-2 possesses the highest K_f value (2.804 μM/min) and the lowest K_d value (0.745 × 10⁻⁴ min⁻¹). This indicates that the construction of a Z-scheme heterojunction is not only favorable to promote H₂O₂ generation but also to inhibit the decomposition of H₂O₂.

Table 1. H₂O₂ production and AQY (%) of g-C₃N₄, CdS and CN@CdS-2 at 450 nm.

Photocatalyst	g-C3N4	CdS	CN@CdS-2
H2O2 production (μM)	38.97	13.97	86.7
AQY (%)	1.01	0.36	2.24

demonstrates yields and AQY values of photocatalytic H₂O₂ production from CdS, g-C₃N₄ and CN@CdS-2 under 450 nm monochromatic light irradiation. As can be seen from Table 1, the AQY value of CN@CdS-2 is significantly higher than that of pristine CdS and g-C₃N₄, which is attributed to a Z-scheme heterojunction constructed to promote visible light absorption. In addition, we compared the performance of photocatalytic H₂O₂ synthesis on CN@CdS-2 with other photocatalysts reported in the literature, as shown in

Table 2. Comparison of photocatalytic H₂O₂ synthesis performance of CN@CdS-2 with that of reported photocatalysts.

Photocatalyst	H2O2 Yield(μmol g−1 h−1)	AQY (%)	Refs
CN@CdS-2	287.5	2.24 (450 nm)	This work
ZnO/CuInS2	91	1.2 (365 nm)	[25]
CdS/K2Ta2O6	160.89	-	[26]
Bi4O5Br2/g-C3N4	124	11.8 (420 nm)	[27]
Resorcinol-formaldehyde Resin/g-C3N4	140	-	[28]
rGO decorated W18O49@g-C3N4 (r-CNW-2)	49.4	1.2 (420 nm)	[29]

, demonstrating its excellent photocatalytic performance.

2.4. Possible Photocatalysis Mechanism

The pathways for photocatalytic H₂O₂ production include the ORR pathway and WOR pathway. Active substance capture experiments were performed to elucidate the mechanism of H₂O₂ generation. The experiments were carried out by bubbling Ar to eliminate O₂. The results are illustrated in Figure 7a. When Ar was continuously passed into the reaction solution, the H₂O₂ generation decreased significantly to only 16.89% of that under saturated air. It is suggested that O₂ is the main reactor for photocatalytic H₂O₂ production. In addition, P-benzoquinone (BQ), isopropanol (IPA), EDTA-2Na and AgNO₃ were used as ·O₂⁻, ·OH, h⁺ and e⁻ trapping agents, respectively. The ·OH scavenger (IPA) has little inhibitory effect against H₂O₂ production on CN@CdS-2. The e⁻ and ·O₂⁻ capture agent significantly decreased H₂O₂ production, confirming that e⁻ and ·O₂⁻ are the main active species and the photocatalysts perform a two-electron reduction from O₂ to produce H₂O₂. The H₂O₂ production increases by 10.18% after adding EDTA-2Na, which is attributed to the fact that EDTA-2Na consumes h⁺ and inhibits the complexation of photogenerated electron-hole pairs, thus allowing more e⁻ to participate in the surface reduction reaction to generate H₂O₂. To reveal the main contributing radical for photocatalytic H₂O₂ production, ESR experiments were performed using 5,5-dimethyl-1-pyrrolidine N-oxide (DMPO) as a radical trapping agent, as depicted in Figure 7b. The distinct peaks of DMPO-·O₂⁻ can be observed, confirming that H₂O₂ can be generated from the ORR pathway.

In addition, the Gibbs free energy diagram is calculated for the conversion of O₂ to H₂O₂ to obtain further insight into the mechanism of highly reactive H₂O₂ production on CN@CdS (Figure 8). The adsorption of O₂ (O₂ → *O₂) is the rate determining step of the 2e⁻ ORR process. During the conversion of O₂ to H₂O₂, CN@CdS exhibits a smoother path compared to g-C₃N₄, suggesting a kinetic advantage [16]. In the O₂ absorption process, a lower ΔG value is observed on CN@CdS (−1.77 eV) than g-C₃N₄ (−2.01 eV). In addition, OOH*, the key intermediate of 2e⁻ ORR, is more readily formed on CN@CdS than on g-C₃N₄. The relatively low ΔG value suggests that CN@CdS prefers to produce H₂O₂ via a two-step one-electron ORR pathway in comparison with g-C₃N₄.

Based on the above results, the mechanism of photocatalytic H₂O₂ production by the CN@CdS Z-scheme heterojunction is proposed, as shown in Figure 9. The photocatalytic H₂O₂ production process consists of the WOR pathway and ORR pathway, where the ORR pathway is the main source of H₂O₂. The high CB potential (−1.17 eV vs. NHE) on g-C₃N₄ is sufficient to drive a single-step two-electron ORR for the direct reduction of O₂ to form H₂O₂ in the presence of protons (H⁺) (Equations (4–6)). In addition, the accumulation of holes (h⁺) on CdS accelerates the kinetics of the WOR pathway. h⁺ in VB on CdS oxidizes H₂O in situ to produce H₂O₂ (Equation (7)). This process is accompanied by the 4e⁻ WOR process (Equation (8)). Based on the above discussion, it can be concluded that enhanced H₂O₂ production can be attributed to the high specific surface area of the convoluted structure and the construction of the Z-scheme heterojunction, which facilitates the spatial separation of redox centers and the exposure of active sites to promote WOR and ORR reactions.

$$\mathrm{CN}@\mathrm{CdS}+\mathrm{h}\nu \to {\mathrm{e}}^{-}+{\mathrm{h}}^{+}$$

(3)

$${\mathrm{O}}_2+{\mathrm{e}}^{-}\to ·{\mathrm{O}}_2^{-}$$

(4)

$${·\mathrm{O}}_2^{-}+{\mathrm{H}}^{+}\to ·\mathrm{OOH}$$

(5)

$$·\mathrm{OOH}+{\mathrm{e}}^{-}+{\mathrm{H}}^{+}\to {\mathrm{H}}_2{\mathrm{O}}_2$$

(6)

$$2{\mathrm{H}}_2\mathrm{O}+2{\mathrm{h}}^{+}\to {\mathrm{H}}_2{\mathrm{O}}_2+2{\mathrm{H}}^{+}$$

(7)

$$2{\mathrm{H}}_2\mathrm{O}+4{\mathrm{h}}^{+}\to {\mathrm{O}}_2+4{\mathrm{H}}^{+}$$

(8)

3. Materials and Methods

3.1. Materials

All of the reagents were purchased from Sinopharm Chemical Reagent Co. (Shanghai, China), and were not further treated. Urea (CO(NH₂)₂), melamine (C₃H₆N₆), potassium iodide (KI), cadmium acetate (C₄H₆CdO₄·2H₂O), thiourea (CH₄N₂S), hydrochloric acid (HCl), ethylene glycol (C₂H₄O₂) and ammonium molybdate (H₈MoN₂O₄) were analytical reagents.

3.2. Synthesis of Photocatalysts

3.2.1. Synthesis of g-C₃N₄

g-C₃N₄ was obtained by a thermal polymerization of precursors. First, 10 g CO(NH₂)₂ and 2 g C₃H₆N₆ were taken into a mortar and ground to a uniform mixture. The mixed powder was placed in a porcelain boat and placed in a tube furnace. The temperature was gradually increased to 550 °C in Ar atmosphere at a rate of 5 °C/min and held for 4 h. The product was allowed to cool naturally to room temperature. The product was ground and washed three times with deionized water to remove residual impurities. Finally, g-C₃N₄ was obtained by drying at 60 °C.

3.2.2. Synthesis of CdS

A total of 0.533 g C₄H₆CdO₄·2H₂O and 0.4567 g CH₄N₂S were dissolved in 60 mL of deionized water. After continuous stirring for 1 h, the mixed solution was transferred to a 100 mL autoclave and reacted at 180 °C for 12 h. The product was obtained by centrifugation and was washed three times alternately using deionized water and ethanol, respectively. The final CdS was obtained by drying at 60 °C overnight.

3.2.3. Synthesis of CN@CdS Z-Scheme Heterojunction

The g-C₃N₄@CdS (CN@CdS) Z-scheme heterojunction was obtained by an electrostatic self-assembly method as shown in Figure 10. A total of 0.1 g g-C₃N₄ was added into 50 mL deionized (DI) water and ultrasonically dispersed for 1 h. Subsequently, a certain amount of CdS was added proportionally into the g-C₃N₄ dispersion, and stirring continued for 24 h. The products were collected by centrifugation, washed with DI water and dried overnight at 60 °C, and different mass ratios of CN@CdS composites were obtained. According to the mass ratio of CdS to g-C₃N₄, the prepared composites were named g-C₃N₄@CdS-1(CN@CdS-1), g-C₃N₄@CdS-2(CN@CdS-2) and g-C₃N₄@CdS-3(CN@CdS-3), in which the percentages of the mass ratio of CdS to g-C₃N₄ were 10%, 15% and 20%, respectively.

3.3. Characterizations

The crystal structure of the samples was measured using a D8 Advance X-ray diffractometer from Bruker, Munich, Germany. X-ray photoelectron spectroscopy (XPS) was performed with an Esca Lab 250XI X-ray photoelectron spectrometer from Thermo Fisher Scientific, Waltham, MA, USA. The optical absorption properties of samples were analyzed by ultraviolet visible diffuse reflectance spectroscopy (UV-vis DRS) on a UV-3600 ultraviolet visible spectrometer manufactured by Shimadzu, Kyoto, Japan, with BaSO₄ powder as a substrate blank, and the detection range was 200~800 nm. The S-4800 scanning electron microscope from Hitachi, Chiyoda City, Japan, was used to observe the surface morphology of the samples. TEM images were obtained using a JEM-2100 transmission electron microscope. The radical components were monitored by an electron paramagnetic resonance spectrometer (A300 ESR, Bruker, Munich, Germany). The photoluminescence (PL) spectra of the synthesized samples were analyzed by a fluorescence spectrometer, F-7000, from Hitachi, Chiyoda City, Japan, with an excitation wavelength of 366 nm.

The electrochemical properties of the samples were determined using an electrochemical workstation model CS350M (Wuhan COEETEST Instruments Co., Ltd., Wuhan, China). The tests were performed using three electrodes, a Pt electrode as the counter electrode, Ag/AgCl electrode as the reference electrode and FTO conductive glass coated with photocatalytic material as the working electrode (samples are coated with an area of 1 × 1 cm²), and 0.1 M NaSO₄ solution as the electrolyte. When conducting transient photocurrent tests, the light source is a 300 W Xe lamp (CEL-HXF300, λ > 420 nm) with a switching cycle of 60 s. Mott–Schottky measurements were performed at a frequency of 1 kHz.

3.4. Photocatalytic Synthesis and Degradation of H₂O₂ and Analytical Methods

3.4.1. Photocatalytic Synthesis of H₂O₂

In the experiment, 50 mg of photocatalyst was dispersed in 50 mL of DI water, sonicated for 10 min and then placed in a 100 mL photochemical reactor. Before dark treatment for 30 min, the mixture was adjusted to pH = 3 with hydrochloric acid solution. The dark treatment process involves continuous stirring and the injection of enough air into the suspension to achieve an adsorption–dissolution equilibrium between the photocatalyst and dissolved oxygen in the water. A 300 W Xe lamp (CEL-HXF300, λ > 420 nm) was used as the light source, and the reaction temperature was controlled by a thermostatic water circulation system. After activating the light source, samples for measurement are taken every 20 min and the reaction solution is collected through a 0.22 μM PES filter head. To evaluate the stability of the photocatalyst, the photocatalyst was collected through centrifugation, washed with deionized water and subsequently dried at 60 °C for 6 h after each photocatalytic reaction experiment. Following the drying process, the samples were reintroduced into the photochemical reactor and subjected to a repeated photocatalytic H₂O₂ production experiment under the aforementioned experimental conditions.

The experimental procedure for apparent quantum yield (AQY) is as follows: 3 mg of photocatalyst was dispersed into 3 mL of DI water, and the pH of the suspension was adjusted to 3 with hydrochloric acid solution. Then, it was added to a test tube with a 1 × 1 cm² area receiving light controlled using masking tape. A single-wavelength LED lamp (λ = 450 nm) was applied as the light source with a light intensity of 9.55 mW/cm². Other experimental procedures were performed as the photocatalytic H₂O₂ production experiments. The AQY was calculated by the formula below:

$$\mathrm{AQY}={\displaystyle \frac{{\mathrm{N}}_{\mathrm{e}}}{{\mathrm{N}}_{\mathrm{p}}}}\times 100\%={\displaystyle \frac{2\times \mathrm{n}\times {\mathrm{N}}_{\mathrm{A}}\times \mathrm{h}\times \mathrm{c}}{\mathrm{S}\times \mathrm{P}\times \mathrm{t}\times \lambda }}\times 100\%$$

where AQY—apparent quantum yield;

N_e—electrons involved in the photocatalytic reaction;

N_p—incident photons;

n—the amount of H₂O₂ molecule (mol);

N_A—Avogadro’s constant (6.022 × 10²³/mol);

h—Planck’s constant (6.626 × 10⁻³⁴ J·s);

c—the speed of light (3 × 10⁸ m/s).

S—the area of irradiation (cm²);

P—the energy density of incident light (W/cm²);

t—the time of photocatalytic reaction (s);

λ—the wavelength of monochromatic light (nm).

3.4.2. Photocatalytic Decomposition of H₂O₂

The experimental conditions for the photocatalytic decomposition of H₂O₂ are identical to photocatalytic H₂O₂ production. 50 mg of photocatalyst was dispersed in 50 mL of H₂O₂ (0.5 mM) solution followed by ultrasonication for 10 min and then placement in a photochemical reactor. After being stirred in the dark for 30 min, the light source was activated and a 1 mL suspension was extracted every 20 min to measure the concentration of H₂O₂.

3.4.3. Analytical Methods

The concentration of H₂O₂ was determined according to the iodometric method [13]. The working curve can be established by configuring H₂O₂ solutions with different standard concentrations, as illustrated in Figure 11. Experimental details are as follows: 1 mL of reacted solution was added to 2 mL of 0.1 M KI solution and 50 μL of 0.01 M (NH₄)₂MoO₄ solution, and the absorbance of the solution was measured by a UV-visible spectrophotometer at the wavelength of 350 nm. The resulting H₂O₂ concentration was calculated using a standard curve.

3.5. Computational Detail

In this study, the Perdew–Burke–Ernzerhof (PBE) exchange-correlation generalization and generalized gradient approximation generalization (GGA) in the Materials Studio software (2018) were used and a plane-wave truncation energy of 540 eV was set to perform the DFT calculation [30]. For each photocatalyst single-cell model, a 3 × 3 × 1 k-point grid was used for sampling and geometry optimization in the self-consistent field (SCF) calculations, while a 4 × 4 × 1 k-point grid was chosen for the density of states (DOS) calculations. The convergence thresholds for energy and force were set to 10⁻⁴ eV and 0.02 eV Å⁻¹, respectively. To improve the accuracy and reliability of the calculations, we also used the DFT + U method to correct for van der Waals forces, and applied dipole corrections, the cancellation of symmetry settings and spin polarization. The calculation model for photocatalysts is shown in Figure 12.

4. Conclusions

In conclusion, CN@CdS Z-scheme heterojunctions with high photocatalytic activity and stability have been constructed by in situ anchoring CdS nanoparticles onto curly-like g-C₃N₄. The optimized CN@CdS shows significantly enhanced visible-light-driven photocatalytic H₂O₂ production, which is superior to that of pure CdS and g-C₃N₄. The convoluted structure of g-C₃N₄ provides abundant surface-anchored sites, which restrict the agglomeration of CdS nanoparticles. More importantly, the well-matched energy band structure and close contact between CdS and g-C₃N₄ induced the formation of IEF, which accelerated the photogenerated electron migration. The Z-scheme heterojunction has been constructed, which not only achieves a spatial separation of photogenerated carriers, but also maintains strong oxidation and reduction capabilities. The results of DFT, the ESR test and trapping experiments show that CN@CdS generates H₂O₂ through the dual channels of the ORR and WOR pathways, dominated by the ORR pathway. This study provides promising insights for the rational design of photocatalysts for efficient photocatalytic H₂O₂ production.