Advanced oxidation processes (AOPs) have proved to be very attractive procedures to treat organic pollutants [1-7]. Since Fenton et al. [8] discovered the strong ability of an Fe²⁺/H₂O₂ system to catalyze oxidative tartaric acid, a great deal of research work has focused on the degradation of organic pollutants by the Fenton reaction [9-12]. Nevertheless, owing to the narrow pH range, pretreatment must be carried out during the Fenton reaction. Meanwhile, the iron sludge generated after the reaction demands further treatment to meet the emission requirements, thus restricting the development of the Fenton reaction to degrade organic pollutants [13, 14]. Lately, it has been confirmed that semiconductor photocatalytic materials and various iron-based and copper-based Fenton materials form photo-Fenton systems, which can solve the circulation problem of high-valence metal ions to low-valence metal ions in the Fenton reaction [15].

As another AOP, photocatalysis is being developed, having the advantages of green products, mild reaction conditions, simple operation, and good selectivity. In 1972, Fujishima et al. [16] discovered for the first time that the n-type semiconductor, TiO₂, could catalyze the decomposition of water to produce H₂ under illumination. From then on, photocatalysis has unveiled the prologue. During the past few decades, research on photocatalytic degradation of organic pollutants has developed rapidly and become a leading technology in wastewater treatment [17-19]. Because the band gap (E_g) of the metal oxide catalyst, TiO₂, is relatively large (about 3.2 eV), it can only absorb ultraviolet light with a wavelength less than or equal to 400 nm [20]. However, infrared and visible lights are the most abundant in the solar light, therefore the above limits the possibility of large-scale popularization and application of TiO₂. In order to fully utilize the potential advantages of both the Fenton oxidation technology and photocatalysis in degrading organic pollutants, it is vital to construct new catalysts that can respond to a wide solar spectrum to induce visible-light-responsive photo-Fenton catalysis.

As one of the most investigated perovskite-type oxides owing to the distinctive layered structure and outstanding performance in the photodegradation of organics, Bi₂WO₆ (BWO) has a relatively wide band gap of 2.6–2.8 eV, which can respond to a part of visible light [21-24]. However, the practical application of Bi₂WO₆ is still hindered by the high recombination rate of its photo-generated electron–hole pairs and narrow light-responsive region. To this end, many modification strategies have been developed to efficiently utilize the solar energy and promote its activity [25-29]. Among these strategies, coupling Bi₂WO₆ with suitable cocatalysts to fabricate heterojunctions is proven to be an effective strategy for promoting the separation of the photogenerated carriers and broadening the light-responsive region [30-32]. Among them, Cu₂ZnSnS₄ (CZTS) is a narrow band-gap semiconductor, which can respond to visible light and even near-infrared light. Recently, it has been widely reported that CZTS can be an effective cocatalyst that can be combined with other semiconductors to increase the photocatalytic activity [33, 34].

In this study, two-dimensional/two-dimensional (2D/2D) type-Ⅱ p-n CZTS/BWO heterojunctions were fabricated by a second solvothermal method. The photocatalytic and photo-Fenton catalytic activities were evaluated using azo dye rhodamine B (RhB) and colorless tetracycline hydrochloride (TC-HCl) as the simulated organic pollutants under visible-light irradiation. On the basis of the energy band analysis and electron spin resonance (ESR), the possible photocatalytic and photo-Fenton catalytic mechanisms were proposed. The combination of the photocatalytic and Fenton reactions could enable full utilization of their respective complementary advantages and rapidly degrade pollutants under mild conditions.

Sodium hydroxide (NaOH, 96.0%), zinc acetate dihydrate (Zn(CH₃COO)₂∙2H₂O, 99.0%), and cetyltrimethylammonium bromide (C₁₉H₄₂BrN, 99.0%) were purchased from Tianjin Kemiou Chemical Reagent Co., Ltd (Tianjin, China). Sodium tungstate dihydrate (Na₂WO₄·2H₂O, 99.5%), bismuth nitrate pentahydrate (Bi(NO₃)₃∙5H₂O, 99%), and anhydrous ethanol (C₂H₅OH, 99.7%) were purchased from Tianjin Zhiyuan Chemical Reagent Co., Ltd (Tianjin, China). Hydrochloric acid (HCl, 36%), nitric acid (HNO₃, 63%), and ammonia solution (NH₃·H₂O, 28%) were purchased form Sichuan Xilong Chemical Co., Ltd (Sichuan, China). Thiourea (CH₄N₂S, 99.0%) was purchased from Sinopharm Chemical Reagent Co., Ltd (Shanghai, China). Hydrated stannic chloride (SnCl₄∙5H₂O, 99.0%) was purchased from Shanghai Aladdin Biochemical Technology Co., Ltd (Shanghai, China). Copper(Ⅱ) acetate monohydrate (Cu(CH₃COO)₂∙H₂O, 99.0%) and TC-HCl (C₂₂H₂₄N₂O₈·HCl, biotechnology level) were purchased from Shanghai Macklin Biochemical Co., Ltd (Shanghai, China). Rhodamine B (C₂₈H₃₁C_lN₂O₃, Analytical grade) was purchased from Tianjin Tianxin Fine Chemical Development Center (Tianjin, China). Deionized water was used as a solvent. All the reagents were of analytical grade, except TC-HCl, and were used without any further purification.

BWO nanosheets were synthesized by an improved hydrothermal method according to a previous report [35]. Firstly, 1.94 g of Bi(NO₃)₃&∙5H₂O was completely dissolved in 5 mL of 4 mol/L HNO₃ under stirring for 30 min. Afterward, 0.20 g of cetyltriethylammonium bromide (CTAB) was dissolved in 300 mL of deionized water and 6.0 g of Na₂WO₄&·2H₂O was dissolved in 200 mL of deionized water to obtain homogeneous solutions, respectively. Then, 30 mL of the CTAB and 20 mL of the Na₂WO₄&·2H₂O aqueous solutions were added to the above mixture in turn. After the addition, the pH was adjusted by adding a diluent of neutralized ammonia water and deionized water in a ratio of 1:1 drop by drop until the pH was neutral. Following continuous stirring for 1 h, it was transferred to a 100-mL Teflon-lined stainless-steel autoclave and put it into an oven, reacting for 14 h at 180 ℃. Finally, the resultant was taken out to wash and dry and was put it into a muffle furnace, calcining for 3 h at 300 ℃. The obtained sample was denoted as BWO.

The 2D/2D CZTS/BWO heterojunctions were prepared by an in situ assembling strategy [34]. In a typical process, 0.0128 g of CH₄N₂S, 0.0134 g of Cu(CH₃COO)₂∙H₂O, 0.0118 g of SnCl₄∙5H₂O, and 0.0074 g of Zn(CH₃COO)₂∙2H₂O were fully dissolved in 70 mL of anhydrous ethanol. Subsequently, 1.5 g of Bi₂WO₆ was added, and the mixture was stirred for 1 h. The mixture was transferred to a 100-mL Teflon-lined stainless-steel autoclave and placed in an oven to react at 200 ℃ for 24 h. Once taken out, the obtained products were washed with ethanol and deionized water and finally dried at 60 ℃ for 8 h. This resulted in the 1.0% CZTS/BWO heterojunction. According to this method, we could sequentially prepare CZTS/BWO heterojunctions with different percentages of CZTS based on the 1.0% CZTS/BWO. The contents of CH₄N₂S, Cu(CH₃COO)₂∙H₂O, SnCl₄∙5H₂O, and Zn(CH₃COO)₂∙2H₂O were changed according to the required percentages, and the obtained samples were denoted as x% CZTS/BWO (x = 2 and 5, respectively). Moreover, pure CZTS was prepared under the same conditions, except for the addition of BWO, for comparison.

An X-ray powder diffractometer (Shimadzu XRD-7000) was used to analyze the crystallographic properties of the catalysts. X-ray photoelectron spectroscopy (XPS) was performed with a PHI-5400 (America PE) 250 xi system. The morphology of the samples was analyzed via scanning electron microscopy (SEM, JSM-6700F) and transmission electron microscopy (TEM, JEM-2100) (Japan electronics). Energy dispersive X-ray (EDX) was performed via field-emission scanning electron microscopy (FESEM, JSM-7610F) to analyze the elemental features of the samples. The UV-Vis diffuse reflectance spectra (UV-Vis-DRS) were measured on a UV-2550 UV-Vis spectrophotometer. The ESR spectra were examined on a Bruker ELEXSYS-Ⅱ E500 CW-EPR model. Time-resolved photoluminescence (TR-PL) spectra were obtained by a FLS920 fluorescence spectrometer (Edinburgh Analytical Instruments, UK).

Typically, 20 mg/L azo dye RhB and 40 mg/L colorless TC-HCl were degraded in the presence or absence of visible light and H₂O₂ at room temperature. The experimental processes were as follows: 0.02 g catalyst was added to a 50 mL-quartz tube, and then 20 mL of simulated contaminant aqueous solution was pipetted into the tube. At the same time, a circulating cooling water was introduced to avoid the influence of temperature. Before the photocatalytic reaction, the mixture was stirred in the dark to reach the adsorption/desorption equilibrium, and then a certain amount of H₂O₂ was added before turning on a 300-W metal halide lamp with a 420-nm cut-off filter as the visible light source. Irradiation was introduced, and 3 mL of the reaction liquid was taken out at a certain time interval and centrifuged to obtain the supernatants for UV analysis at 554 and 357 nm of the maximum absorption peaks of RhB and TC-HCl, respectively.

XRD was employed to investigate the phase structures of the prepared samples. Fig. 1 shows the XRD patterns of CZTS, BWO, and the 2% CZTS/BWO heterojunction, respectively. The diffraction peaks of BWO at 28.30° (131), 32.67° (060), 32.79° (200), 32.91° (002), 47.14° (202), 55.66° (191), 58.53° (262), 76.07° (193), and 78.53° (204) can be indexed to an orthorhombic system (JCPDS no. 39-0256) [36]. However, the crystallinity of CZTS is relatively lower than that of BWO. The two characteristic peaks corresponding to (112) and (220) observed at 28.5° and 47.3° can be considered as kesterite CZTS (JCPDS no. 26-0575) [34]. The prepared 2% CZTS/BWO heterojunction shows the homoplastic crystal structure of BWO, and only the peak intensity is attenuated. The latter may be due to the low loading or the main two diffraction peak positions of CZTS being basically coincident with those of BWO, and its low crystallinity.

The surface composition and elemental valence of the samples were demonstrated by XPS. As can be seen from Fig. 2a, the 2% CZTS/BWO heterojunction contains Bi, W, O, C, Cu, Zn, Sn, and S elements; Bi, W, C, and O elements appear in BWO; and Cu, Zn, Sn, C, and S elements are observed in CZTS, where C 1s belongs to the pollution of the instrument itself. Fig. 2b–g show the high-resolution XPS spectra of Bi 4f, W 4f, Cu 2p, Zn 2p, Sn 3d, and S 2p. The characteristic peaks of the 2% CZTS/BWO heterojunction at 164.45 and 159.17 eV showing a shift to a higher binding energy compared to those in BWO corresponding to Bi 4f_5/2 and Bi 4f_7/2 signify that the Bi element is in a +3 state (Fig. 2b) [37]. In Fig. 2c, the characteristic peaks at 37.6 and 35.5 eV of the 2% CZTS/BWO heterojunction show a shift to a higher binding energy compared to those of BWO corresponding to W 4f_5/2 and W4f_7/2, suggesting that the W element exists in a +6 state [38]. As can be seen from Fig. 2d, the characteristic peaks of the 2% CZTS/BWO heterojunction and CZTS at 951.57 and 931.7 eV correspond to Cu 2p_1/2 and Cu 2p_3/2, revealing the presence of a Cu⁺ state [34]. Fig. 2e shows that in the 2% CZTS/BWO heterojunction and CZTS XPS spectra, the Zn²⁺ state is proved by two distinct peaks at 1021.8 and 1044.8 eV with a separation of 23 eV, showing a shift to a higher binding energy compared to those in CZTS. The peaks are attributed to Zn 2p_3/2 and Zn 2p_1/2, respectively [39]. The two peaks in Fig. 2f centered at 486.6 and 495.02 eV correspond to Sn 3d_5/2 and Sn 3d_3/2, show a shift to a higher binding energy compared to those in CZTS and confirm rationally that the Sn element is in an Sn⁴⁺ state [40]. As can be seen from Fig. 2g, the corresponding characteristic peaks at 164.47 and 159.18 eV are Bi 4f_5/2 and Bi 4f_7/2, and no remarkable characteristic peaks of S 2p are observed. The latter may be owing to the low content of CZTS in the 2% CZTS/BWO heterojunction and close binding energy positions of S 2p and Bi 4f [41]. From Fig. 2a–g, it can be seen that CZTS nanosheets are successfully coupled with BWO nanosheets.

The morphologies and microstructures of CZTS, BWO, and the 2% CZTS/BWO heterojunction were studied by SEM, EDX mapping, TEM, and HRTEM techniques. As shown in Fig. 3a–b, BWO is remarkably observed to have a 2D nanosheet-like structure. However, Fig. 3c‒d showed that CZTS exhibited a three-dimensional (3D) spherical structure. The SEM images observed in Fig. 3e‒f showed that the 2D and 2D nanosheets are united together. The element distributions of CZTS and BWO in the CZTS/BWO heterojunction was further determined by EDX element mapping, and the results are shown in Fig. 3g–n. Seven elements: Bi, W, O, Cu, Zn, Sn, and S are uniformly distributed on the 2% CZTS/BWO heterojunction.

Fig. 4a–d show the TEM images of BWO, CZTS, and the 2% CZTS/BWO heterojunction. Fig. 4a–b further confirm that BWO exhibits a sheet-like structure, whereas CZTS exhibits a 3D spherical structure. Fig. 4c–d display the TEM images of the 2% CZTS/BWO heterojunction, further confirming that BWO and CZTS form a 2D/2D structure. The d values of three different stripes with lattice spacings are 0.273, 0.315, and 0.313 nm, respectively, which can be assigned to the (200) and (131) planes of BWO and (112) plane of CZTS separately. The results suggest that the CZTS nanosheets are uniformly assembled onto the surface of the BWO nanosheets.

As can be seen from Fig. 5, CZTS has a strong absorption in the UV-Vis light region, and BWO also exhibits a strong absorption in the visible range. It is worth noting that its response range to visible light will be red shifted to a certain extent after the CZTS/BWO heterojunction is formed, thus improving the utilization rate of visible light.

Fig. 6 shows the time-resolved fluorescence lifetime spectra of BWO and the 2% CZTS/BWO heterojunction. We mainly observe the lifetimes of the carriers. The photoelectron lifetimes conform to the two-exponential decay model [42]. The decay times (τ₁ and τ₂) and PL aptitudes (A₁ and A₂) are given in Table 1. The 2% CZTS/BWO heterojunction (2.01 ns) exhibits an enhanced lifetime compared to BWO (0.929 ns), which indicates that the coupling of CZTS can enhance the separation of the photogenerated electron–hole pairs and ultimately improve the carrier lifetime. The enhancement of the photoelectron lifetime is very helpful to improve the catalytic performances of the photocatalytic and photo-Fenton reactions.

Azo dye RhB was firstly used as a simulated pollutant to evaluate the photocatalytic and photo-Fenton catalytic performances of the catalyst. Fig. 7a displays that when no photocatalyst is added, the dye concentration hardly changes after the photocatalytic reaction, which indicates that the dye molecules have a high photostability. When BWO and CZTS are used as the catalysts alone, the degradation efficiency of RhB by BWO can reach 62%, whereas CZTS has a relatively weak photocatalytic activity, reaching only 20%. The formed CZTS/BWO heterojunctions exhibits a markedly enhanced photocatalytic activity. Among them, the photocatalytic efficiency is the best when the loading of CZTS is 2%, which can reach about 94% of the degradation efficiency. Further, the increase in the CZTS content leads to a sharp decline in the photocatalytic activity. This may be because excessive CZTS loading reduces the absorption of light, which indicates that the configuration of a heterojunction is very beneficial to enhance the photocatalytic activity. The efficiency of the photocatalytic degradation of RhB was furtherly simulated by kinetic analysis using the first-order kinetic equation: ln(C₀/C) = k_app· t, as shown in Fig. 7b. Based on the comparison chart of the rate, as shown in Fig. 7c, the photocatalytic reaction rate constants (k_app/min^‒1) of the 2% CZTS/BWO heterojunction are 3.04 and 12.5 folds compared to those of BWO and CZTS, respectively. Fig. 7d presents that the degradation efficiency of RhB by H₂O₂ alone can reach 13% after 50 min irradiation. However, when H₂O₂ is introduced into the photocatalytic system with photocatalyst, the degradation efficiency of 2% CZTS/BWO heterojunction and BWO can reach 76% and 62%, respectively, after 10 min of visible light illumination. The enhanced catalytic activity may be attributed to the synergistic effect of photocatalysis and Fenton catalysis.

The 2% CZTS/BWO heterojunction was selected for further photo-Fenton degradation of colorless pollutants. The results are presented in Fig. 8. It is notable from Fig. 8a that the 2% CZTS/BWO heterojunction also has a high activity performance for TC-HCl under the photo-Fenton condition compared to under BWO. At the same time, the effects of the catalyst dosage, H₂O₂ volume, and pH values on the photocatalytic activity were studied. The effect of the H₂O₂ volume on TC-HCl degradation is shown in Fig. 8b. The degradation efficiency of TC-HCl increases from 76.4% to 92.5% with the increase in H₂O₂ from 0.05 to 1.6 mL because more ^•OH can be generated by increasing the amount of H₂O₂. Fig. 8c displays the effect of the catalyst dosage on the degradation of TC-HCl. The degradation efficiency gradually increases as the catalyst dosage increases from 0.01 g to 0.03 g. However, as the catalyst dosage continues to increase to 0.04 g, the degradation efficiency is only slightly increased. The effect of the pH is shown in Fig. 8d. It can be seen that when the initial conditions of the pH reach neutrality, the degradation efficiency is maximized and there is a decline with the increase or decrease in the pH value. This is because neutral pH conditions may be conducive to causing the production of active radicals with a strong oxidation ability.

Fig. 9a shows the degradation efficiency of TC-HCl after the photo-Fenton catalytic cycle with the 2% CZTS/BWO heterojunction as the catalyst. It can be seen that the photo-Fenton activity of the heterojunction does not significantly decrease. Moreover, after repeated experiments, no significant new peaks are observed, as shown in Fig. 9b; after the XRD tests, only the intensity is slightly reduced, demonstrating the high stability of the catalyst.

To detect the main active species in the photocatalytic and photo-Fenton catalytic reactions, the ESR test was carried out, and the results are shown in Fig. 10. We mainly discuss the important roles of ^•OH and ^•O₂^- in the catalytic process. As can be seen from Fig. 10a, the characteristic for the strong ESR signals of the four characteristic peaks can be considered as ^•OH, and the 2% CZTS/BWO heterojunction can produce more ^•OH under the double action of visible light and H₂O₂ compared to BWO. Similarly, it can be deduced that in Fig. 10b, the ESR signal of the six characteristic peaks can be considered as ^•O₂^‒. The role of ^•O₂^‒ in the catalytic process is explored, and the ^•O₂^‒ signals of the 2% CZTS/BWO heterojunction are also significantly enhanced under the action of visible light and H₂O₂ compared to those of BWO. Therefore, the additional ^•OH and ^•O₂^‒ generated in the photo-Fenton oxidation process can carry out the oxidation reaction together, thereby remarkably enhancing the catalytic degradation capability of the catalyst.

Through a series of investigations, we can infer that the enhanced mechanisms of the photocatalytic and photo-Fenton reactions for the degradation of organic pollutants over the 2% CZTS/BWO heterojunction are as follows: according to the results of the UV-Vis-DRS and valence band XPS spectra, the band gap of BWO and CZTS were estimated as 2.64 and 1.52 eV, respectively (Fig. 11a–b), and the valence band maximum (VBM) was estimated as 2.22 and 1.00 eV, respectively (Fig. 11c–d). Moreover, BWO and CZTS have been widely recognized as n-type and p-type semiconductors, respectively [43, 34]. The Fermi levels (E_F) of BWO and CZTS are located near to the conduction band (CB) and valence band (VB), respectively. Accordingly, the work function (W_Ⅰ) of BWO is more negative than that of CZTS (W_Ⅱ), as depicted in Fig. 12. Upon contact, free electrons migrate from the BWO side to the CZTS side until the two semiconductors reach equal E_F values. Thus, the BWO side is positively charged, whereas the CZTS side is negatively charged at the interface. A built-in electric field (BIEF) from the BWO to CZTS forms at the interface, and the band edge bending occurs. Under light irradiation, the photogenerated electrons of CZTS (E_CBM = –0.52 eV vs. NHE) can transfer to BWO (E_CBM = –0.42 eV vs. NHE), whereas the photogenerated holes can transfer from BWO (E_VBM = 2.22 eV vs. NHE) to CZTS (E_VBM = 1.00 eV vs. NHE) under BIEF. Accordingly, a type-Ⅱ p–n junction charge carrier transfer mode is formed because of the existing BIEF.

At this time, electrons can reduce O₂ to ^•O₂^–, promoting the degradation of pollutants into small molecular substances. The photogenerated holes left on the VB are positively charged and have strong oxidation properties, which can oxidize and degrade pollutants to form harmless small molecular inorganic substances, such as CO₂ and H₂O. Moreover, these holes can oxidize OH⁻ to form ^•OH with a strong activity to promote the oxidative degradation of pollutants. The above charge carrier transfer mode actually causes an enhancement in the separation ability of the photogenerated electron–hole pairs of BWO. In the traditional Fenton reaction, Cu(Ⅰ) mainly reacts with H₂O₂ to generate Cu(Ⅱ) and ^•OH, but the slow conversion between Cu(Ⅰ) and Cu(Ⅱ) results in an unsatisfactory actual processing effect. However, the introduction of photogenerated electrons can accelerate the reduction of Cu(Ⅱ) to Cu(Ⅰ) and improve the utilization rate of H₂O₂ in the system to generate ^•OH, thereby enhancing the ultra-strong ability to degrade pollutants. It is presumed that the mechanism of the photo-Fenton oxidation degradation of pollutants in the presence of the 2D/2D type-Ⅱ p-n CZTS/BWO heterojunctions is as follows:

(1)

(2)

(3)

(4)

(5)

(6)

We designed a new 2D/2D CZTS/BWO type-Ⅱ p-n heterojunction for the first time to achieve the photo-Fenton reaction. The method of hydrothermal synthesis could easily produce 2D Bi₂WO₆ nanosheets, then CZTS nanosheets were grown on the surface of the BWO nanosheets by a second solvothermal method, and finally, the heterojunction was formed. Owing to the enhanced visible light absorption capability, large charge carrier separation, and mobility exhibited by the heterojunction material, its catalytic activity was greatly enhanced. The synergistic effect between the photocatalysis and Fenton catalysis was utilized to the greatest extent to improve the photo-Fenton activity. At the same time, the ESR technique was used to study the mechanism of the photo-Fenton catalysis. It was found that ^•O₂^- and ^•OH were the main active transient species in the degradation process. At the same time, the catalyst still maintained high photo-Fenton catalytic activity after 4 cycles, indicating a relatively high photostability. The synthesized CZTS/BWO heterojunction could provide a concept to researchers to modify traditional photocatalytic materials with Cu-based semiconductor materials to realize photo-Fenton catalysis.