Construction of a ternary staggered heterojunction of ZnO/g-C3N4/AgCl with reduced charge recombination for enhanced photocatalysis

In this study, a series of ternary composite photocatalysts ZnO/g-C3N4/AgCl(x) (x is the amount of AgCl added, x=0.05, 0.1, 0.15 g) were synthesized, and various characterization methods were used to analyze the morphology, structural, and photochemical properties of the above samples. The photocatalytic activity of the obtained samples was evaluated by degrading rhodamine B (RhB) and acid orange (AOII) under xenon lamp irradiation. The results show that the degradation rate of ZnO/g-C3N4/AgCl (0.1 g) is 99% within 60 min, which is much higher than the 38% degradation rate of g-C3N4. After five cycles, the degradation efficiency of RhB and AOII by ZnO/g-C3N4/AgCl (0.1 g) were still 85% and 94%, respectively. In addition to colored dyes, the photocatalytic degradation of colorless tetracycline hydrochloride (TCH) compounds was also investigated to understand the effect of photosensitization on the degradation process. Based on the electronic structure analysis of the DFT calculations, a staggered heterojunction structure was found between g-C3N4, ZnO, and AgCl. The enhanced photocatalytic activity of the ternary composite is mainly attributed to the efficient separation of charge carriers through the synergistic removal of photogenerated electrons in g-C3N4 by ZnO and AgCl. Radical trapping experiments confirmed that •O2− and h+ were the main active species in the reaction system. As a visible light-responsive catalyst, ZnO/g-C3N4/AgCl can be effectively applied to the degradation of organic dye pollutants and has broad application prospects.


Introduction
Since the twenty-first century, the problem of global environmental pollution has become increasingly serious. There is an urgent need to find a green, environmentally friendly, and sustainable method to solve the above problem. Among them, photocatalytic technology has received extensive attention due to its utilization of inexhaustible solar energy to achieve pollutant degradation (Kais et al. 2019;Wu et al. 2021;Rosaline et al. 2020;Xue et al. 2019;Peng et al. 2019). A suitable photocatalyst plays a very important role in the photocatalytic degradation of dyes. ZnO, which has excellent properties such as low toxicity, strong oxidizing ability, and low cost, is one of the most widely used photocatalysts. Unfortunately, the application of ZnO in photocatalysis is limited due to its wide band gap (3.37 eV) and low e − -h + separation efficiency.
It is well known that materials with narrow bandgap below 3 eV can absorb most of the visible spectrum. Therefore, constructing heterojunctions with other semiconductors that have narrower band gaps is an effective method to overcome the defects of ZnO Wang and Shi 2019;Zeng et al. 2019;Selvaraj et al. 2019). According to recent reports, graphitic carbonitride (g-C 3 N 4 ) with two-dimensional (2D) graphene-like nanosheet structure is a promising photocatalyst, owing to its desirable properties of narrow bandgap (2.7 eV), tunable electronic structure, easy preparation, and excellent physicochemical stability. Zhong et al. have reported that the construction of g-C 3 N 4 /ZnO heterojunction can improve the transfer of photogenerated charges from g-C 3 N 4 to ZnO, thereby enhancing the photocatalytic performance of ZnO under visible light (Zhong et al. 2020).
Based on the above research reports, this study envisages introducing a semiconductor material into the binary g-C 3 N 4 /ZnO composite to construct a ternary composite system by forming a staggered heterojunction structure with g-C 3 N 4 and ZnO. Compared with the binary g-C 3 N 4 /ZnO composite, the ternary composite system ZnO/g-C 3 N 4 /AgCl may have more charge transfer channels. Due to its unique optical properties and strong response to visible light, the introduction of AgCl into binary g-C 3 N 4 /ZnO composites may be an effective strategy Ravichandran and Sindhuja 2019). It can be expected to further increase the separation of photogenerated e − -h + pairs, thereby enhancing the photocatalytic activity of g-C 3 N 4 /ZnO.
In this study, a series of ternary composite photocatalysts ZnO/g-C 3 N 4 /AgCl(x) (x is the AgCl additions, x=0.05, 0.1, 0.15 g) with a potentially enhanced photocatalysis ability were prepared via hydrothermal reaction, calcination, and in situ deposition. The structure, morphology, and optical properties of the samples were characterized by XRD, SEM, TEM, BET, UV-Vis DRS, FT-IR, PL, ESR, etc. The photocatalytic activity of ZnO/g-C 3 N 4 /AgCl under sunlight irradiation was studied by using xenon lamp to simulate sunlight and rhodamine B (RhB) and acid orange (AOII) to simulate dye pollutants. The reasons for the enhanced photocatalytic performance were discussed in depth by analyzing the charge separation efficiency, generation of active species, and energy band structure. The photocatalytic mechanism of ZnO/g-C 3 N 4 /AgCl was speculated by radical trapping experiments and theoretical calculations based on DFT. The reusability of ZnO/g-C 3 N 4 /AgCl was evaluated through five cycle tests.

Preparation of ZnO/g-C 3 N 4 /AgCl catalyst
The preparation of ZnO was designed as follows: first, 10 g of zinc acetate dihydrate was put into 30 mL of absolute ethanol and stirred at room temperature for 1 h. Subsequently, put the above mixture into an oven and dry it at 60 °C for 8 h. The obtained material was ground to the powder and then placed to a box muffle furnace calcined at 300 °C for 2 h at a heat rate of 3 °C/min. After cooling to room temperature, an off-white powder was obtained, namely ZnO.
The g-C 3 N 4 was obtained by calcining 5 g of melamine in a box muffle furnace at 500 °C for 2 h at a heat rate of 5 °C/min.
AgCl was synthesized in the following way: 0.1 g AgNO 3 was mixed with 30 mL of deionized water and sonicated for 30 min to ensure complete dispersion. A certain concentration of NaCl solution was added dropwise and then the mixture was stirred for 5 h in the dark. After repeated washing with ethanol and deionized water, the obtained product was dried in an oven at 60 °C for 6 h. Finally, pack it in a bag and store it away from light.
The process for the preparation of ZnO/g-C 3 N 4 /AgCl was designed as follows: first, 0.3 g/g-C 3 N 4 , 0.25 g ZnO, and a certain mass of AgCl (0.05, 0.1, 0.15 g) were added into 80 mL of isopropanol, respectively, and sonicated for 2 h. The mixture was placed in a water bath at 65 °C and refluxed for 5 h. Finally, after repeated washing with ethanol and deionized water, the obtained product was placed in an oven at 80 °C to dry for 6 h, and the obtained product was the ternary composite catalyst ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g). The same preparation method was used, except that AgCl was not added in the preparation process; that is, g-C 3 N 4 /ZnO was synthesized. Similarly, g-C 3 N 4 /AgCl was synthesized without adding ZnO in the preparation process.

Catalyst characterization
The crystal structures of materials were analyzed by XRD (Bruker D8 Advance, Bruker, Germany) via Cu Kα line of 0.15418 nm at 40 kV and 40 mA. The surface molecular functional group structures of materials were analyzed by Nicolet,USA). The chemical bonds formed between different elements in the ZnO/g-C 3 N 4 /AgCl (0.1 g) catalyst were analyzed by a multifunctional imaging electron spectrometer (XPS, ESCALAB 250XI, Thermo Fisher, USA) under Al Kα (hv = 1486.6 eV), power 150 W, 500 μm beam spot. The peak fitting analysis was carried out by XPS INSTALLATION Avantage software. The morphologies of g-C 3 N 4 and ZnO/g-C 3 N 4 /AgCl (0.1 g) catalysts were observed by SEM (QUANTAF250, FEI, USA). The internal structures of g-C 3 N 4 and ZnO/g-C 3 N 4 /AgCl (0.1 g) were observed by TEM (JEM-2100F, JEOL, Japan) at an accelerating voltage of 200 kV. The N 2 adsorption-desorption isotherms (−196 °C) of materials were measured by a physical adsorption instrument (ASAP2020, Mack Instruments, USA). The UV-Vis absorption spectrum and visible diffuse reflectance spectrum of materials were measured by a UV-Vis spectrophotometer (UV-2600, Shimadzu Corporation, Japan). The photoluminescence (PL) spectra of materials were analyzed by a fluorescence spectrometer (F-4500, Hitachi Instruments, Japan). The photoelectrochemical (PEC) properties of materials were analyzed by an electrochemical workstation (CHI660D, Shanghai Chenhua Instrument Co., Ltd., China). The electron spin resonance (ESR) signals of superoxide radical were analyzed by an electron paramagnetic resonance spectrometer (JES FA200, Japan Electronics Co., Ltd., Japan).

DFT calculation
Using the Perdew-Burke-Ernzerhof (PBE) plane wave supersoft pseudopotential method in the Generalized Gradient Approximation (GGA) of the CASTEP module of Materials studio software and the interaction between the inner electrons and their valence electrons was described by the projector augmented wave (PAW) method. In the optimized calculation process of the electronic structure, the valence electron configurations were selected as 5s 1 4d 10 (Ag), 4s 2 3d 10 (Zn), 2s 2 2p 2 (C), 2s 2 2p 3 (N), 2s 2 2p 4 (O), and 3s 2 3p 5 (Cl), and the cutoff energy was selected as 500 eV. The convergence criteria for structure optimization were as follows: the single-atom energy is 10 −6 eV/atom and the interatomic interaction force is 0.01 eV/Å.

Photocatalytic activity test
Xenon lamp (500 w) was used to simulate sunlight, and rhodamine B (RhB) and acid orange (AOII) were used as simulated dye pollutants for degradation. In a typical experiment, 50 mg of catalyst was added to 100 mL of RhB (10 mg/L) or AOII (10 mg/L) solution, respectively. Turning on the cooling water circulation system to maintain temperature at 25 °C, the solution was stirred in the dark for 30 min to reach adsorption-desorption equilibrium. Switching on the xenon lamp and stirring continuously, 3 mL samples were taken from the suspension every ten min for the next 1 h. The samples were centrifuged for 10 min to settle the photocatalyst powder. The absorbance of the contaminants was measured with a UV-Vis spectrophotometer, and the degradation rate was calculated by Formula (1).
In the formula, D is the degradation rate, C 0 is the initial absorbance of the solution, and C is the absorbance of the sample after the photocatalytic experiment. Furthermore, since the colored pollutants were affected by the photosensitizing effect, to examine the intrinsic photocatalytic ability of our synthesized photocatalysts, we also performed the degradation of 10 mg/L of tetracycline hydrochloride (TCH) under similar conditions using the same procedure. (1)

Morphology and structural analysis of catalysts
The crystal phase structures of g-C 3 N 4 , g-C 3 N 4 /ZnO, g-C 3 N 4 /AgCl, and ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g) are analyzed by XRD, and the results are shown in Fig. 1. In the XRD pattern of g-C 3 N 4 , two distinct characteristic diffraction peaks appear at 2θ of 12.9° and 27.5°. The diffraction peak at 2θ of 12.9° corresponds to the inplane structural packing motif of the conjugated aromatic system, namely the (100) crystal plane of g-C 3 N 4 ; while the diffraction peak at 27.2° corresponds to the crystal plane packing of the conjugated aromatic system, that is, the (002) crystal plane of g-C 3 N 4 (Li et al. 2017;Wang et al. 2009). They agree with the standard diffraction pattern of g-C 3 N 4 (JCPDS NO. 87-1526). In the XRD patterns of ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g), in addition to the diffraction peaks corresponding to g-C 3 N 4 , there also appeared other diffraction peaks corresponding to hexagonal wurtzite ZnO (JCPDS NO.00-036-1451) and cubic AgCl (JCPDS NO.00-031-1238) (Qin et al. 2019;Zhang et al. 2014). It can be observed that the peak shapes of the diffraction peaks corresponding to AgCl and ZnO are sharp and clear, indicating that AgCl and ZnO are well crystallized. At the same time, it can be found that the diffraction peaks corresponding to g-C 3 N 4 become very weak, presumably because g-C 3 N 4 is covered by AgCl and ZnO, or the composite samples are arranged in decreasing long-range order. In addition, in the XRD patterns of the ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g), it can be observed that the diffraction peaks of ZnO at 2θ of 47.6° Fig. 1 XRD patterns of g-C 3 N 4 , g-C 3 N 4 /ZnO, g-C 3 N 4 /AgCl, and ZnO/g-C 3 N 4 /AgClAgCl(x) (x=0.05, 0.1, 0.15 g) and 69.1° become weaker, indicating a chemical interaction between ZnO and g-C 3 N 4 (Imran et al. 2021).
In order to verify whether the combination of g-C 3 N 4 with ZnO and AgCl will affect its pore structure, the specific surface area, pore volume and pore size distribution of g-C 3 N 4 , g-C 3 N 4 /ZnO, g-C 3 N 4 /AgCl and ZnO/g-C 3 N 4 / AgCl(x) (x=0.05, 0.1, 0.15 g) were determined by N 2 adsorption-isotherm curve (77 K). Figure 2 shows the adsorption and desorption isotherms of the tested samples. According to the Brunauer-Deming-Deming-Teller (BDDT) classification rule, the samples exhibit type IV adsorption isotherms, indicating that the catalysts belong to mesoporous solid substances (2-50 nm) (Yu et al. 2010). At the same time, the relative pressure of all samples is P/P 0 ≈1, and an obvious H3-type hysteresis loop appears, which proves that the product has a slit-like pore structure and a wide pore size distribution (Fu et al. 2021). The specific surface area, pore volume and pore size of the samples are shown in Table 1. The specific surface area of the ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g) is larger than that of g-C 3 N 4 , g-C 3 N 4 / ZnO, and g-C 3 N 4 /AgCl, and the largest is ZnO/g-C 3 N 4 /AgCl (0.1 g). It can be seen from Table 1 that the specific surface area of g-C 3 N 4 is 7.19 m 2 /g, while the specific surface area of the ZnO/g-C 3 N 4 /AgCl (0.1 g) is 17.28 m 2 /g, which is 2.4 times larger. In addition, the pore volume increased from 0.041 cm 3 /g to 0.075 cm 3 /g. Since the large specific surface area can provide more reaction sites for the organic dyes to react, it is favorable for the photocatalytic reaction to occur.
Among the prepared catalyst samples, the ZnO/g-C 3 N 4 / AgCl (0.1 g) has the largest specific surface area and pore volume. Therefore, its photocatalytic effect is probably the best. In order to understand its structural characteristics further, the morphology of g-C 3 N 4 and ZnO/g-C 3 N 4 /AgCl (0.1 g) were characterized by SEM. Figure 3a, b are SEM images of the g-C 3 N 4 and the ZnO/g-C 3 N 4 /AgCl (0.1 g).
As can be seen from Fig. 3, the surface of the g-C 3 N 4 is relatively smooth, showing a block-like structure. However, the structure of the ZnO/g-C 3 N 4 /AgCl (0.1 g) changed significantly, and its particle size was smaller, showing an irregular rod-like structure with uneven surface. The uneven structure was conducive to increase surface area, and which is consistent with the BET results. With the increase of the specific surface area of the sample, the number of reactive sites and reactive molecule transport channels per unit area will increase, thereby improving the photocatalytic activity.
The morphology and pore structure of the g-C 3 N 4 and ZnO/g-C 3 N 4 /AgCl (0.1 g) sample were further analyzed by TEM, and results were shown in Fig. 4a, b. Figure 4a shows that g-C 3 N 4 is an irregular shape composed of two-dimensional (2D) curved nanosheets. It can be seen from Fig. 4b that when ZnO and AgCl are combined with g-C 3 N 4 , it can be observed that AgCl and ZnO are closely adhered to the surface of g-C 3 N 4 , indicating the formation of a ternary heterostructure. The heterojunctions between AgCl/ZnO particles and g-C 3 N 4 sheets are expected to promote the rearrangement of energy band levels and form a built-in electric field, thereby enhancing the separation and transfer of charge carriers. In addition, the presence and distribution of elements in the ZnO/g-C 3 N 4 /AgCl (0.1 g) were further confirmed by TEM elemental mapping, as shown in Fig. 4c. The results show that Zn, O, C, N, Ag, and Cl exist simultaneously in the ZnO/g-C 3 N 4 /AgCl composite.
The FT-IR spectrum can be used to further study the functional group structure of the sample, and the scanning range is 400-4000 cm −1 . Results were shown in Fig. 5.
In the pristine g-C 3 N 4 , the bands in the range of 1240-1650 cm −1 can be attributed to the stretching vibrations of C-N and C=N in heterocycles ; Fig. 2 N 2 adsorption-desorption isotherms of g-C 3 N 4 , g-C 3 N 4 /ZnO, g-C 3 N 4 /AgCl, and ZnO/g-C 3 N 4 /AgClAgCl(x) (x=0.05, 0.1, 0.15 g) Table 1 Specific surface area, pore volume and pore size of g-C 3 N 4 , g-C 3 N 4 /ZnO, g-C 3 N 4 /AgCl, and ZnO/g-C 3 N 4 /AgCl (x) Wan et al. 2016). Furthermore, the band at 806 cm −1 corresponds to the breathing mode of the heptazine arrangement (Yu et al. 2015). For the samples containing g-C 3 N 4 , the broad absorption band appearing at 3000-3300 cm −1 can be attributed to the terminal NH 2 or NH groups at the defect sites of the g-C 3 N 4 aromatic rings (Chung et al. 2016;Ma et al. 2017). For the samples containing ZnO, the band at 494 cm −1 is related to the stretching vibration of the Zn-O bond (Kumar et al. 2014). For the samples containing AgCl, there are not any peaks corresponding to the AgCl in the range of 400-4000 cm −1 (Raizada et al. 2020;Zhou et al. 2014).
The chemical composition, elemental chemical state and electronic state of g-C 3 N 4 /AgCl, g-C 3 N 4 /ZnO, ZnO/g-C 3 N 4 / AgCl (0.1 g) were further investigated by XPS. From Fig. 6a, the XPS full scan spectrum of ZnO/g-C 3 N 4 /AgCl (0.1 g) can clearly observe Cl 2p, Ag 3d, Zn 2p, C 1s, N 1s and O 1s. It can be observed from Fig. 6b that the Ag 3d peak of g-C 3 N 4 / AgCl is deconvoluted into two peaks, which belong to Ag 3d 5/2 (367.89 eV) and Ag 3d 3/2 (373.88 eV) respectively, indicating that Ag exists in the form of Ag + ). In Fig. 6c, the binding energies of 199.79 eV and 198.05 eV can correspond to Cl 2p 1/2 and Cl2p 3/2 of Cl in g-C 3 N 4 /AgCl (Asadzadeh-Khaneghah et al. 2018). In the C1s spectrum of ZnO/g-C 3 N 4 /AgCl (0.1 g) (Fig. 6d), the peaks at 284.8, 285.95 and 288.28 eV are assigned to C=C, C-N, N=C-N bonds, respectively (Wang et al. 2016a, b;Sun et al. 2019). It can be observed from Fig. 6e that the N 1s peak of ZnO/g-C 3 N 4 /AgCl (0.1 g) is deconvoluted into two peaks, which belong to the sp 2 hybrid aromatic N atom bound to the carbon atom in the form of C=N-C (398.56 eV) (Lv et al. 2018) and tertiary N atoms are bound to carbon atoms in the form of N-(C) 3 (400.17 eV) (Hu et al. 2015). In the Zn 2p spectrum of g-C 3 N 4 /ZnO (Fig. 6f), peaks at 1022.45 and 1045.44 eV are assigned to Zn2p 3/2 and Zn 2p 1/2 , respectively (Liu et al. 2022). The O 1s spectrum of g-C 3 N 4 /ZnO (Fig. 6g) is deconvoluted into two peaks, belonging to Zn-O (530.09 eV) (Chen et al. 2016) and oxygen in the hydroxyl (-OH) of surface adsorbed water (531.39 eV) (Liu et al. 2013). It is worth noting that compared with g-C 3 N 4 /ZnO and g-C 3 N 4 /AgCl, the binding energies of Zn 2p, O 1s, C 1s, N 1s, Ag 3d, and Cl 2p spectra of ZnO/g-C 3 N 4 /AgCl (0.1 g) have slight chemical changes, which may be due to the charge transfer between g-C 3 N 4 and ZnO and AgCl during the formation of heterojunction. This result confirms the successful synthesis of these composites and the electron transfer through g-C 3 N 4 and ZnO, AgCl interfaces.

Band structure and photogenerated charge properties of catalyst
The optical properties of all samples were analyzed by UV-Vis spectrometer, and the results are shown in Fig. 7a.
It can be seen from Fig. 7 that the wavelength of the absorption edge for g-C 3 N 4 is 460 nm, indicating that the photoresponse can be achieved in both visible and ultraviolet regions, while the absorbance of the ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g) reaches about 480 nm. Moreover, the absorption intensity is stronger in the visible light and ultraviolet regions, which means that the ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g) has a stronger photoresponse to the ultraviolet and visible light regions. The value of the forbidden band width (Eg) of a semiconductor can be approximated as the energy required for an electron to transit from the valence band to the conduction band. The smaller the value is, the less transition energy is required, so the accurate measurement of the Eg of a semiconductor can further predict its photophysical and chemical properties. Based on Tauc's law, the Eg of the semiconductor is calculated (Makuła et al. 2018), as shown in Eq. (2).
In the formula, B is the absorption coefficient and hv is the energy of incident light. The γ factor depends on the nature of the electronic transition, with a direct band gap of 1/2 and an indirect band gap of 2.
Using the curve obtained by the Tauc formula (see Fig. 7b), the intersection of the linear fit part and the X-axis in this figure can be regarded as the E g of the sample. The results are shown in Table 2.
It can be seen from Table 2 that the E g of g-C 3 N 4 is 2.79 eV, while the E g of ZnO/g-C 3 N 4 /AgCl (0.1 g) is 2.75 eV, and Eg value gradually decreases, indicating the energy required for the electronic transition even less. The reduction of E g indicates that the yield and lifetime of electrons and holes generated in the conduction band (CB) and valence band (VB) of the ternary composite will be fundamentally improved under visible light irradiation.
Photoluminescence (PL) is commonly used to study the lifetime and effective mobility of e − -h + pairs of photocatalysts (Bao and Chen 2018). Figure 8 is the emission spectrum obtained after testing all samples by UV light with excitation wavelength of 370 nm at room temperature. The emission spectrum show that the maximum peak intensity of all samples is around 460 nm. Compared with g-C 3 N 4 , g-C 3 N 4 /ZnO and AgCl/g-C 3 N 4 , the maximum peak intensity of ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g) is slightly shifted to the direction of short wavelength, and the peak intensity is significantly reduced, which means that the photogenerated e − -h + recombination rate is reduced, and the separation efficiency is improved (Wang et al. 2016a, b). In the ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g) sample, with the increase of AgCl content, the PL peak intensity first decreased and then increased, and the peak intensity reached the lowest when the AgCl content was 0.1 g. Combined with the results of UV-Vis DRS, it can be seen that the light absorption area of it is wider, and the addition of AgCl can well inhibit the recombination of e − -h + of g-C 3 N 4 , enhance the lifetime of active molecules and improve the  The transport properties of photogenerated carriers in ZnO/ g-C 3 N 4 /AgCl (0.1 g) and g-C 3 N 4 , ZnO/g-C 3 N 4 , AgCl/g-C 3 N 4 catalysts were investigated by transient photocurrent spectroscopy (TPC). Results are shown in Fig. 9a. It can be seen from Fig. 9a that ZnO/g-C 3 N 4 /AgCl (0.1 g) has stronger photocurrent density than g-C 3 N 4 , g-C 3 N 4 /ZnO, and AgCl/ g-C 3 N 4 . This may be attributed to the forming of a staggered heterojunction structure among g-C 3 N 4 , ZnO and AgCl, which enables more efficient separation of photogenerated electron holes at the interface of the photocatalytic material, thereby improving the photocatalytic activity.
The transport and photoelectrochemical properties of photogenerated carriers in ZnO/g-C 3 N 4 /AgCl (0.1 g) and g-C 3 N 4 , g-C 3 N 4 /ZnO, g-C 3 N 4 /AgCl catalysts samples were investigated by electrochemical impedance spectroscopy (EIS). The electrochemical properties were measured by placing the sample in a 0.1 M Na 2 SO 4 solution to obtain the impedance spectrum (Nyquist) of the sample. The size of the arc radius on the Nyquist diagram corresponds to the charge transfer resistance and the photogenerated electron holes separation efficiency, and the results are shown in Fig. 9b. In addition, the equivalent circuit diagram is obtained by fitting the ZsimpWin software. The selected circuit model is R s1 (Q(R s2 W))(CR ct ), where R s1 and R s2 are solution resistances, R ct is the charge transfer resistance, and Q, W, and C are respectively constant phase angle element, diffusion impedance and double-layer capacitance, as shown in Fig. 9c. It can be seen from Fig. 9b that ZnO/g-C 3 N 4 / AgCl (0.1 g) has a smaller arc radius than g-C 3 N 4 , g-C 3 N 4 / ZnO, and g-C 3 N 4 /AgCl, which indicates that the combination of ZnO and AgCl with g-C 3 N 4 can effectively reduce its interface charge transfer resistance (Zhang et al. 2012). It is beneficial to improve the conductivity of electrons in the ground state rather than the excited state, which is very beneficial for the separation of photo-excited charges in the photocatalytic process. The test results of EIS are consistent with the test results of TPC.

Photocatalytic degradation performance analysis
In this study, a 500 W xenon lamp was used to simulate the solar light source, the reactor was connected to circulating water to keep the reaction temperature at 25 °C, RhB and AOII were used as simulated pollutants to test the degradation performance. Since the catalyst itself may have a certain adsorption effect on organic dyes, it is necessary to carry Fig. 7 a UV-Vis DRS spectra of the samples; b Tauc plots of g-C 3 N 4 , g-C 3 N 4 /ZnO, g-C 3 N 4 / AgCl, and ZnO/g-C 3 N 4 /AgCl (0.1 g)  out a blank control test. In the absence of light, a group of adsorption dark reaction experiments were carried out.
The reaction results are shown in Fig. 10. It can be seen from Fig. 10 that the catalyst samples all showed a certain adsorption to the dye. Turn on the light source to test, and the degradation rate of ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g) is significantly higher than that of g-C 3 N 4 , g-C 3 N 4 / ZnO, g-C 3 N 4 /AgCl. And when the content of AgCl was 0.1 g, the photocatalytic performance reached the highest, and the degradation rate of RhB and AOII were 99% within 60 min. The excellent catalytic efficiency of ZnO/g-C 3 N 4 /AgCl (0.1 g) may be attributed to the synergistic effect of AgCl, g-C 3 N 4 and ZnO, which can rapidly transfer electrons and inhibit the recombination of electrons and holes, and the catalytic activity is greatly improved. Furthermore, colored dye solutions are known to influence the photocatalytic performance of catalysts to a large extent through the photosensitization effect. Therefore, to examine the intrinsic photocatalytic efficiency of all synthesized samples for wider applications, we also tested their photocatalytic performance for degrading 10 mg/L TCH solution under simulated sunlight conditions. The degradation process and experimental conditions of TCH remained the same as RhB. Figure 10c shows the degradation rate of TCH as a function of time. The results show that the degradation rate of the ternary sample is much higher than that of the binary and unit samples, and the highest efficiency reaches 84% when the AgCl content is 0.1 g. This study provides conclusive evidence that the catalytic effect is only due to the three photocatalytic ability of meta-nanocomposites. The stability and reproducibility of a photocatalyst are very important for its practical application, so the stability cycling experiments were carried out on g-C 3 N 4 and ZnO/g-C 3 N 4 /AgCl (0.1g) catalysts, and the samples after each test were centrifuged, collected, dried and reused. Results were shown in Fig. 11a, b. It is clear that after Fig. 11 Cycling tests of ZnO/g-C 3 N 4 /AgCl (0.1 g) photocatalyst: cyclic degradation graph of a rhodamine b and b acid orange; c XRD patterns ZnO/g-C 3 N 4 /AgCl (0.1 g) before and after photocatalytic degradation; d-i XPS spectrum of ZnO/g-C 3 N 4 /AgCl (0.1 g) after photocatalytic degradation  using the photocatalyst for five successive runs, there is a small decrease in its activity, and the degradation efficiency of RhB and AOII could reach 85% and 94%.indicating ZnO/g-C 3 N 4 /AgCl (0.1g) has good stability and reproducibility. The reduction of photocatalytic activity may be related to the photoetching of the AgCl part of the nanocomposite, which was confirmed by XRD and XPS after five runs with the nanocomposite (see Fig. 11c-i).
The XRD spectrum of ZnO/g-C 3 N 4 /AgCl (0.1 g) after the reaction shows a diffraction peak of silver element compared with that before the reaction. The XPS spectrum of ZnO/g-C 3 N 4 /AgCl (0.1 g) after the reaction shows two more groups of peaks in the Ag 3d spectrum compared with that before the reaction, corresponding to the Ag 3d 3/2 (374.9 eV) and Ag 3d 5/2 (368.9 eV) of metallic silver (Thang et al. 2021), while the binding energy of other substances does not change much compared with that before the reaction. It is obvious that metallic silver is produced during the photocatalytic reaction. Therefore, during the degradation reaction, the composition of the photocatalyst changes gradually. Therefore, after continuous use of photocatalyst, the degradation efficiency is slightly reduced.

Density functional theory (DFT) calculations
Through DFT calculations, the electronic coupling and charge transfer directions at the interface of the heterojunction are intuitively reflected. It can be seen from Fig. 12 that there is obvious surface charge redistribution at the g-C 3 N 4 / ZnO and g-C 3 N 4 /AgCl interfaces, and electron transfer from g-C 3 N 4 to ZnO and AgCl occurs during the hybridization process, which can effectively reduce the recombination rate of photogenerated e − -h + . In addition, we calculated the work function, as shown in Fig. 13, the work function of g-C 3 N 4 (4.28 eV) is smaller than that of ZnO (4.96eV) and AgCl (5.17 eV). When the semiconductor contacts, due to the Fermi level (E f ), an interfacial electron transfer process occurs between g-C 3 N 4 and ZnO, AgCl until their E f are aligned at the same level (Yang et al. 2021). Specifically, the photogenerated electrons will flow from g-C 3 N 4 to ZnO and AgCl through the heterojunction. This is consistent with the results of the previous charge transfer calculations. Combined with the results of PL and PEC, it can be concluded that the addition of ZnO and AgCl can well inhibit the recombination of photogenerated electron holes of g-C 3 N 4 , improve the lifetime of active molecules and promote charge separation, which are all beneficial to the improvement of photocatalytic performance.

Radical trapping experiment
The catalyst will generate many active species after being illuminated, such as h + , •O 2 − , 1O 2 , and •OH, and they will then participate in the photocatalytic process. Exploring the role of these active species in the catalytic process plays a very important role for the studying of the reaction process and catalytic mechanism. After adding the trapping agent disodium ethylenediaminetetraacetate (EDTA-2Na), p-benzoquinone (BQ), L-histidine, and tert-butanol (TBA), the degradation of RhB was carried out with the using of ZnO/g-C 3 N 4 /AgCl (0.1 g) as catalyst under light conditions, in which EDTA-2Na, BQ, L-Histidine and TBA were used to capture h + , •O 2 − , 1O 2 , and •OH, respectively. The experimental results are shown in Fig. 14a. After adding EDTA-2Na and BQ, the degradation rate was significantly reduced, which were 45% and 14%, respectively. After adding TBA and L-Histidine, the degradation rate still reach more than 85%. Furthermore, the generation of •O 2 − was detected by ESR technology and further quantitatively detected •O 2 − to explore the catalytically active species. It can be seen from Fig. 14b that no obvious signal of •O 2 − was observed under dark conditions. In contrast, the signal of •O 2 − can be clearly observed under illumination conditions, which means that light is an indispensable factor for the generation of •O 2 − . It can be seen from Fig. 14c that with the increase of illumination time, the content of •O 2 − has been significantly improved. The above results and analysis further prove that the degradation of pollutants is closely related to •O 2 − , which is consistent with the capture experimental results. Therefore, it is inferred that the main activity involved in the photocatalytic degradation is •O 2 − and h + .

Photocatalytic mechanism analysis
To determine its photocatalytic ability, the energy level positions of the conduction band and valence band of g-C 3 N 4 , ZnO, and AgCl were determined from the Pearson electronegativity principle to. The calculation formulas were shown in Formulas (3) and (4): In the formula, E VB represents the valence band energy, E CB represents the conduction band energy, E g is the band gap energy of the catalyst, X is geometric mean of the Pearson electronegativity value of each element that constitutes the semiconductor, and E e is the energy of free electrons (usually taken as 4.5 eV).
The Pearson electronegativity values of the elements C, N, Zn, O, Ag, and Cl are 6.27 eV, 7.30 eV, 4.45 eV, 7.54 eV, 4.44 eV, and 8.30 eV, respectively. Substituting the above   (3) and (4), the Pearson absolute electronegativity values of g-C 3 N 4 , ZnO, and AgCl were calculated, and the results were shown in Table 3. Combining Formulas (3) and (4), the E VB of g-C 3 N 4 is calculated to be +1.57 eV, and the E CB of it is −1.13 eV; the E VB and the E CB of ZnO are +2.89 eV and −0.31 eV; the E VB and the E CB of AgCl are +3.19 eV and −0.06 eV. In general, photogenerated electrons and holes are more inclined to transfer to lower redox potentials. When irradiated by simulated sunlight, since the CB potential (−1.13 eV) of g-C 3 N 4 is more negative than that of ZnO and AgCl, so the photogenerated electrons will transit from the CB of g-C 3 N 4 to the CB surface of ZnO and AgCl, which is consistent with the DFT calculation results. The VB potential (+1.57 eV) of g-C 3 N 4 is lower than that of ZnO (+2.89 eV) and AgCl (+3.19 eV), and the photogenerated h + are transferred from the VB of ZnO and AgCl to the VB of g-C 3 N 4 . Thus, the effective separation of photogenerated e − -h + pairs is achieved, which greatly improves the photocatalytic activity of the ZnO/g-C 3 N 4 /AgCl. Based on the above discussion, the synergetic effect of g-C 3 N 4 , ZnO and AgCl led to enhanced visible light photocatalytic activity.
Based on the analysis of the above experimental and characterization results, the possible catalytic mechanism of the ternary system ZnO/g-C 3 N 4 /AgCl is proposed (as shown in Fig. 15). Since the CB potentials of ZnO and AgCl are more negative than those of O 2 /•O 2 − (−0.046 eV), photogenerated electrons will transition from the CB of g-C 3 N 4 to the CB of ZnO and AgCl. At this time, e − on the CB of ZnO and AgCl reacts with the oxygen adsorbed on the surface to form •O 2 − , and the generated •O 2 − further reacts with RhB and AOII to complete the degradation. Furthermore, the photogenerated h + will be transferred from the VB of ZnO and AgCl to the VB of g-C 3 N 4 . Since the VB potential of g-C 3 N 4 is lower than that of H 2 O/•OH (2.4 eV) , so the h + on the VB of g-C 3 N 4 cannot react with H 2 O to generate hydroxyl radicals, and h + directly participates in the degradation reaction. In summary, the main active species for degradation are h + and •O 2 − , which is consistent with the experimental results of free radical trapping.

Comparison of photocatalytic degradation performance
The results of the decomposition of RhB under visible light irradiation by ZnO/g-C 3 N 4 /AgCl (0.1 g) and g-C 3 N 4 -based heterojunction photocatalysts reported in the literature were compared, as shown in Table 4. The photocatalysts exhibited better or comparable visible light photocatalytic Table 3 Pearson absolute electronegativity X and Eg of g-C 3 N 4 , ZnO, and AgCl

Sample
Absolute electronegativity X Eg g-C 3 N 4 4.72 2.7 eV ZnO 5.79 3.2 eV AgCl 6.07 3.25 eV Fig. 15 Photocatalytic mechanism diagram of ZnO/g-C 3 N 4 /AgCl performance. Therefore, ZnO/g-C 3 N 4 /AgCl (0.1 g) can be used as a promising and competitive visible-light-responsive catalyst for the degradation of organic dye pollutants with practical application prospects.

Conclusions
In this study, a series of ternary composite photocatalysts ZnO/g-C 3 N 4 /AgCl(x) (x=0.05, 0.1, 0.15 g) were synthesized, and the photocatalytic activity of the obtained samples were evaluated by the degradation of RhB and AOII through irradiation with a xenon light source (500 w). The experimental results show that ZnO/g-C 3 N 4 /AgCl (0.1 g) has the highest photocatalytic activity, and the degradation rate can reach 99% within 60 min. After three cycles of ZnO/g-C 3 N 4 /AgCl (0.1 g) catalyst, the degradation efficiencies of RhB and AOII can still reach 85% and 94%, respectively. Based on DFT calculations, it was found that a staggered heterojunction structure was formed between g-C 3 N 4 , ZnO, and AgCl, which would help ZnO and AgCl synergistically remove the photogenerated electrons in g-C 3 N 4 , thus achieving efficient separation of charge carriers. Radical trapping experiments confirmed that h + and •O 2 − were the main active species in the reaction system. ZnO/g-C 3 N 4 / AgCl can be used as a visible light-responsive catalyst to degrade organic dye pollutants, which has broad application prospects.