Safety concerns in relation to bisphenol A (BPA) have drawn attention owing to its widespread use in producing epoxy resins and polycarbonate plastics [1]. As a representative endocrine disruptor, BPA can damage the fecundity of animals and humans [2] by disrupting endocrine effects in the reproductive systems and cause the death of some types of cells [3-5]. Thus, a rapid and efficient method of removing BPA is urgently required.

Among various technologies for removing BPA, the heterogeneous photo-Fenton process is considered to be an effective and affordable method, owing to its generation of highly active radicals and the easy separation of the catalysts from treated wastewater [6, 7]. In the past few decades, much attention has been paid to investigating heterogeneous photo-Fenton processes for organic pollutant degradation under UV irradiation. However, UV light-based heterogeneous photo-Fenton processes are limited in their practical applications because UV light accounts for only 3%–5% of solar light energy [8]. Thus, the development of heterogeneous photo-Fenton catalysts that operated based on visible light is of great importance for practical applications.

The combination of photo-Fenton catalysts and plasmonic materials might overcome this problem. Plasmonic materials based on silver/silver halide (Ag/AgX, X = Cl, Br, I) composites can strongly absorb visible light because of their surface plasmon resonance [9-11] and are extensively used as visible-light photocatalytic materials [12-15]. However, Ag/AgX composites suffer from high charge carrier recombination rates, causing losses of photocatalytic efficiency [16]. Additionally, recent evidence has suggested that iron oxides, blended with semiconductor composites, can act a selective acceptor and inhibit the recombination of electron/hole pairs [17-19]. For example, Yang et al. [20] synthesized a Fe₃O₄@rGO@TiO₂ visible light catalyst, in which the photo-induced electrons from TiO₂ could rapidly transfer to Fe³⁺, accelerating the redox transformation between Fe(Ⅲ) and Fe(Ⅱ). On the basis of the above reports, we attempted to add Ag/AgCl composites into a heterogeneous photo-Fenton system. We expected that the Ag/AgCl might promote the charge transfer between Fe(Ⅲ)/Fe(Ⅱ) by promoting photogenerated electrons and enhancing the photo-Fenton catalytic activity under visible light irradiation. The choice of catalyst support is also important for preparing efficient heterogeneous photo-Fenton catalysts. Sepiolite is a zeolite-like clay mineral that has a high specific surface area and good chemical stability [21, 22]. Thus, in this work, sepiolite was selected as the support material to enhance the synergistic effects between Ag/AgCl and hydroxy-iron, and fabricated a quaternary-composite photo-Fenton catalyst. We investigated the photocatalytic activity and stability of our synthesized Ag/AgCl/Fe-Sepiolite (Ag/AgCl/Fe-S) photocatalyst using BPA as a target contaminant. In addition, the possible photocatalytic mechanism involved in the photo-Fenton system was discussed. This work provides new insights into the preparation of visible-light responsive photo-Fenton catalysts.

The sepiolite sample used in this study was obtained from Hunan Province, China, and the clay particle size was approximately 100-mesh. All chemical reagents were of analytical grade and used without further purification. Deionized water was used throughout the experiments.

Acidified sepiolite was prepared by pouring 50 g of raw sepiolite powders into 1 L of HNO₃ solution (2 mol/L) under continuous stirring at 40 ℃ for 2 h. This mixture was then filtered and washed with deionized water repeatedly until the supernatant pH value was approximately 7, and the resulting solid was vacuum dried at 100 ℃ overnight.

The Ag/AgCl/Fe-S catalyst was prepared by ion exchange and photoreduction methods. First, Fe-sepiolite was fabricated as follows. Na₂CO₃ (0.2 mol/L) was added dropwise into 0.2 mol/L Fe(NO₃)₃ solution under stirring at 25 ℃ until the molar ratio of Na/Fe was 1:1. This solution was aged for 24 h at 25 ℃ and then added into an aqueous suspension containing 2 wt% of the above-treated sepiolite under stirring at 60 ℃ until the final Fe/clay ratio was equal to 5 mmol/g of sepiolite. After aging for 12 h at 60 ℃, the mixture was centrifuged and washed by deionized water. The resulting precipitates were dried in air overnight at 70 ℃ to obtain Fe-sepiolite. Subsequently, the Fe-sepiolite was used to fabricate the AgCl/Fe-S. A 1-g portion of Fe-sepiolite was dispersed in 100 mL of deionized water and a 20-mL AgNO₃solution (0.15 g of AgNO₃ dissolved in 20 mL of water) was added to the mixture with vigorous magnetic stirring for 12 h at 25 ℃. Then, a 20-mL KCl solution (0.15 g of KCl dissolved in 20 mL of water) was added into the mixture, which was stirred for a further 30 min. The resulting product (AgCl/Fe-S) was filtered, washed, and dried at 70 ℃. Finally, Ag/AgCl/Fe-S was prepared via a photoreduction method. A 1-g portion of the AgCl/Fe-S was dispersed in 100 mL of deionized water. An AgNO₃ solution (30 mg of AgNO₃ in 5 mL of water) was then added to the mixture, and the reaction stirred in the dark for 30 min. This mixture was irradiated with visible light (λ > 400 nm) for 30 min to partially reduce the Ag⁺ ions to Ag⁰ species. The final product was gathered by centrifugation and dried at 70 ℃. For comparison, pure Ag/AgCl was also prepared by the same method without the addition of Fe-sepiolite. All of the prepared samples were crushed and screened through a 200-mesh sieve.

The XRD patterns of the prepared materials were acquired with a diffractometer (Rigaku D/max-2550 VK/PC) equipped with Cu K_α radiation at 40 kV and 50 mA. XPS measurements were performed on a K-alpha X-ray photoelectron spectrometer (PHI Quantera Ⅱ, UIVAC) using a monochromatic Al K_α X-ray radiation source at 1486.71 eV. The morphologies of the products were observed with a scanning electron microscope (SEM, JSM-6360LV, JEOL). The nitrogen adsorption-desorption isotherms were determined with a NOVA 2200e instrument. Prior to the adsorption tests, the samples were outgassed for 12 h at 150 ℃. The Brunauer-Emmett-Teller (BET) method was used to calculate the specific surface areas of the samples. The light absorption properties were measured with a UV-vis diffuse reflectance spectrophotometer (Shimadzu, UV-2550) with a wavelength range of 200–800 nm.

All the experiments were conducted in a photoreaction apparatus (BL-GHX-V, Shanghai Depai Biotech. Co. Ltd., China), in which a 500 W xenon lamp equipped with a 420-nm cutoff filter was applied as the visible light source. The light source was positioned inside a cylindrical Pyrex vessel surrounded by a jacket with circulating water. The photocatalytic activities of the studied catalysts were evaluated from their ability to degrade BPA under various conditions. An appropriate amount of the catalyst was added into 100 mL of BPA solution (10 mg/L), and the initial pH of solution was adjusted by addition of NaOH or HNO₃ solutions. Prior to irradiation, the solution was magnetically stirred in the dark for 30 min to establish an adsorption-desorption equilibrium. The reaction was started when H₂O₂ was added to the solution and the light source was turned on. During the photocatalytic process, samples were taken from the reaction mixture at fixed intervals, and filtered immediately with a 0.45 μm membrane.

The concentration of BPA in the aqueous solution was measured by high performance liquid chromatography (HPLC, Agilent 1260), and the total organic carbon (TOC) was analyzed with a Shimadzu TOC-V CPH analyzer.

An atomic absorption spectrophotometry instrument (Shimadzu AA7000) was used to measure the quantity of iron ions leached from the catalyst into solution. The concentration of Fe²⁺ in the leachate was measured by the ο-phenanthroline spectrophotometric method (λ = 510 nm).

The active species inducing degradation of BPA during the photocatalytic reaction were detected with the use of various scavengers. Namely, isopropanol (IPA), ammonium oxalate (AO) and benzoquinone (BQ) were separately applied as scavengers for hydroxyl radicals (^•OH), holes (h⁺), and superoxide radicals (O₂^•−), respectively. The processes for the active species capture experiments were similar to those for the photocatalytic experiments.

The photoluminescence (PL) spectra were measured on a fluorescence spectrophotometer (F-4600) at room temperature with excitation at 362 nm.

The generation of ^•OH radicals was also investigated by electron spin resonance spectroscopy (ESR) with a Bruker EMX-10/12 electron paramagnetic resonance spectrometer, and 5, 5-dimethyl-1-pyrroline-N-oxide (DMPO) was chosen as the spin-trapping reagent (0.02 mol/L).

Electrochemical impedance spectroscopy (EIS) measurements were performed in three-electrode quartz cells with a 0.1 mol/L KCl electrolyte solution containing 5 mmol/L Fe(CN)₆³⁻/Fe(CN)₆⁴⁻. A glassy carbon electrode served as the working electrode, which was modified by the different catalysts. A platinum plate and saturated calomel electrode were used as the counter and reference electrodes, respectively. A CHI660 Electrochemical Workstation (Shanghai Chen Hua Instrumental Co. Ltd., China) was used to measure the EIS data over the frequency range from 100 kHz to 0.01 Hz with an AC signal amplitude of 10 mV.

Fig. 1 shows the XRD patterns of sepiolite, Fe-sepiolite, and Ag/AgCl/Fe-S. The intensities of the characteristic peaks of sepiolite at 2θ = 7.494° (110), 20.882° (131), and 26.654° (080) [23] decreased sharply after loading with Fe or Ag/AgCl, suggesting that the sepiolite phase became less crystalline. Furthermore, some new diffraction peaks appeared in the XRD pattern of Ag/AgCl/Fe-S at 27.792°, 32.216°, 46.203°, 54.786°, and 57.445°, which could be indexed to AgCl (JCPDS31-1238). These results indicate the successful loading of AgCl onto sepiolite. However, no distinct diffraction peaks from metallic Ag or Fe could be observed in the Ag/AgCl/Fe-S, which is likely because of the small-size of their crystallites and their high degree of dispersion. To further demonstrate the chemical compositions of Ag/AgCl/Fe-S, we performed XPS measurements, as shown in Figs. 2 and 3. The XPS survey spectrum of the raw sepiolite is also shown for comparison with that of Ag/AgCl/Fe-S.

The main elements found at the surface of sepiolite, were Si, Mg, C, and O. In Ag/AgCl/Fe-S clear signals from Fe and Ag also appeared. In the Fe 2p XPS spectrum of Ag/AgCl/Fe-S (Fig. 3(a)), peaks around 710.6 and 711.8 eV could be attributed to FeOOH and Fe₂O₃ [24], respectively, and indicate that the oxidation state of Fe was Fe(Ⅲ) in the Ag/AgCl/Fe-S catalyst. In the Ag 3d spectrum of the catalyst (Fig. 3(b)), the peaks observed at 367.3 and 373.3 eV belong to Ag⁺, and the peaks at 368.1 and 374.1 eV were assigned to metallic Ag [25, 26]. These results indicate that the Ag/AgCl structure was formed on the surface of the Fe-sepiolite.

The morphology and microstructure of the raw sepiolite, Fe-sepiolite, and Ag/AgCl/Fe-S were examined by SEM imaging (Fig. 4). Images of the raw sepiolite showed a smooth and dense surface (Fig. 4(a)); however, the surface of the Fe-sepiolite particles appeared to be rough and mesoporous (Fig. 4(b)), indicating that the hydroxyl-iron treatment etched channels into the sepiolite. Additionally, Ag/AgCl particles with a size of a few hundred nanometers to several micrometers were observed to be well dispersed on the surface of the Fe-sepiolite catalyst (Fig. 4(c)).

Fig. 5 shows nitrogen adsorption-desorption isotherms of sepiolite, Fe-sepiolite, and Ag/AgCl/Fe-S. All samples exhibited a type Ⅳ isotherm with a H3 type hysteresis loop, indicating that all the samples were typical mesoporous materials with a high adsorption energy [27, 28]. The specific surface areas, pore volumes, and aperture diameters of all samples are summarized in Table 1. The specific surface area and pore volume of Fe-sepiolite were much larger than those of raw sepiolite. However, the corresponding values were slightly lower in Ag/AgCl/Fe-S. These lower values were likely caused by the effects of pore blocking by the Ag/AgCl particles. Similar observations have also been reported by McEvoy et al. [29].

The light absorption ability of the prepared samples was examined by UV-vis diffuse reflectance spectroscopy (Fig. 6). The Ag/AgCl shows two strong absorption peaks in the range of 200–800 nm. The absorption peak at 350 nm is attributed to the indirect bandgap of AgCl [30], and the other intense absorption band observed in the visible light region (around 550 nm) is attributed to the surface plasmon resonance of Ag NPs [31]. By comparison, Fe-S showed much weaker absorption of visible light. However, after loading the Ag/AgCl onto the surface of the Fe-sepiolite, we found a notable enhancement of light absorption in both the UV and visible light regions. Hence, the Ag/AgCl/Fe-S catalyst is able to efficiently absorb visible light.

The degradation and mineralization of BPA in aqueous solution by various processes were investigated. As shown in Fig. 7, in the dark, only a small amount of BPA decomposition was observed after 3 h in the presence of Ag/AgCl/Fe-S and H₂O₂. This result indicated that no active radicals were released in the absence of light. In a system subjected to only visible light, the degradation and mineralization of BPA were also negligible, implying that BPA is stable in wastewater under ambient conditions. When H₂O₂ and visible light were combined, the degradation efficiency of BPA reached 96.24%; however, the mineralization of BPA was negligible. These results indicate that the hydroxyl radicals released in this system could degrade BPA into long-lived intermediates but could not further oxidize those intermediates into CO₂ and H₂O. The degradation of BPA reached more than 95% in the systems with Ag/AgCl, H₂O, visible light; Fe-sepiolite, H₂O₂, visible light; and Ag/AgCl/Fe-S, H₂O₂, visible light. However, the Ag/AgCl/Fe-S, H₂O₂, visible light system showed the highest TOC removal efficiency (60.98%) and the fastest BPA degradation rate, suggesting that Ag/AgCl/Fe-S is an efficient heterogeneous photo-Fenton catalyst for removal of BPA.

The initial H₂O₂ concentration had a notable influence in the photo-Fenton reaction owing to its influence on the production of hydroxyl radicals [32]. Thus, we examined the effects of initial H₂O₂ concentration on the degradation of BPA in our heterogeneous photo-Fenton system with the Ag/AgCl/Fe-S catalyst.

As shown in Fig. 8, when the concentration of H₂O₂ was increased to 6 mmol/L, the kinetic constant of the BPA degradation increased from 0.057 to 0.10393 min⁻¹ correspondingly. This result can be explained by the increased quantity of hydroxyl radicals produced. However, a further increase of the H₂O₂ concentration from 6 to 25 mmol/L led to a lower kinetic constant for BPA removal (0.03641 min⁻¹). This decrease may be attributed to scavenging of hydroxyl radicals by excess H₂O₂ according to Eq. (1) [33-36]:

(1)

The pH value of the reaction medium also plays an important role in photo-Fenton system; pH value not only affects the efficiency of photo-Fenton reaction but also determines the extent of Fe leaching from catalyst. To investigate the influence of initial pH value on the degradation of BPA, we performed experiments at five different pH values (pH = 2, 3, 4, 5, and 6) as shown in Fig. 9.

The kinetic constant of BPA degradation increased from 0.10882 to 0.12838 min⁻¹ as the pH was increased from 2 to 3. However, increasing the solution pH value from 3 to 6 resulted in a decrease of the kinetic constant from 0.12838 to 0.03889 min⁻¹. These results confirmed that the optimum pH for BPA degradation efficiency in the studied system was approximately pH = 3. However, as shown in Fig. 10, leaching of Fe in the system at pH = 3 was almost 3 times as high as that in the pH = 4 system by the end of reaction. This high leaching rate might cause more rapid catalyst degradation. Considering that the kinetic constant of BPA degradation in the system at pH = 4 was 0.10393 min⁻¹, and only slightly lower than that of the pH = 3 system, we used solutions with pH = 4 for the following experiments.

The concentration of Fe²⁺ in the system of pH = 4 was measured, and the results are shown in Fig. 11. The Fe²⁺ content as a proportion of total Fe was low, which may be attributed to the rapid oxidation of Fe²⁺ by H₂O₂.

The effects of light intensity on BPA removal in the heterogeneous photo-Fenton process were investigated by varying the visible light intensity from 250 to 500 W. These results are shown in Fig. 12.

The degradation of BPA was highly sensitive to the visible light intensity. When the power of the light source was increased from 250 to 500 W, the kinetic constant of BPA degradation increased from 0.00775 to 0.10393 min⁻¹over a reaction time of 60 min. This result can be explained by the higher light intensity contributing more photons for activation of the Ag/AgCl/Fe-S catalyst, which in turn promoted the degradation of BPA.

The effects of the catalyst loading on BPA degradation are shown in Fig. 13. The BPA was degraded more rapidly as the catalyst loading was increased from 0.1 to 1.5 g/L. However, for a catalyst loading higher than 1.5 g/L, the BPA removal rate was slightly decreased. Thus, higher catalyst loadings provided more active sites for H₂O₂ activation and produced more active radicals to degrade BPA; however, at catalyst loadings exceeding 1.5 g/L, the visible light penetration was decreased owing to the screening effects of excess catalyst particles in the solution [37]. Thus, the optimal catalyst loading used in this system was 1.5 g/L.

To evaluate the stability of Ag/AgCl/Fe-S in the heterogeneous photo-Fenton reaction, the catalyst was recycled three times under the conditions of 20 mg/L BPA, 6 mmol/L H₂O₂, pH = 4.0, 500-W irradiation intensity, and 1.5 g/L catalyst loading. As shown in Fig. 14, a BPA degradation of 95.41% was achieved in the first run. In the second and third runs, the BPA degradation decreased to 94.86% and 93.28%, respectively, indicating that the catalyst can be effectively reused at least three times without any major loss of its catalytic activity.

To gain insight into the reactive species involved in the heterogeneous photo-Fenton system catalyzed by Ag/AgCl/Fe-S, specific scavengers (IPA, BQ, and AO) were used in the photocatalytic process under the above optimized conditions. IPA is known to be an effective scavenger for ^•OH radicals (k_{i-PrOH, •OH} = 1.9×10⁹ L/(mol·s)) [35, 36, 38]. Fig. 15(a) shows the effects of different concentrations of IPA on the degradation of BPA. As the concentration of IPA was increased the degradation of BPA decreased. The inhibition caused the BPA degradation to approach 65.21% when the IPA concentration was increased to 15 mmol/L. This result suggests that ^•OH is the main oxidative species in the studied heterogeneous photo-Fenton system.

BQ was used to determine the contribution of O₂^•− radicals to BPA degradation [39, 40]. In this study, the addition of BQ had a slight effect on the degradation rate of BPA (Fig. 15(b)), indicating that O₂^•− has less effect on degradation of BPA than ^•OH. Holes are another possible active species in semiconductor photocatalytic systems, and AO is a typically used hole scavenger [41, 42]. As shown in Fig. 15(c), when the concentration of AO was increased from 0 to 10 mmol/L, the BPA degradation efficiency decreased from 100% to 49.23%, which indicated that holes are also an important active species in this system.

On the basis of the above experiments, we concluded that ^•OH radicals play the most important role in the studied heterogeneous photo-Fenton system. To further characterize the ^•OH radical generation ability in the different systems, we performed ESR spin-trapping measurements with DMPO as a spin trap. These results are shown in Fig. 16. No significant ESR signal was detected in the system with Ag/AgCl and visible light, indicating that no ^•OH radicals were generated in this system. A weak signal from DMPO^•−OH adducts characterized by an intensity ratio of 1:2:2:1 was observed in the Fe-S, H₂O₂, visible light system, implying that ^•OH generation was relatively slow because Fe circulation was restricted. For the Ag/AgCl/Fe-S, H₂O₂, visible light system, the intensity of the ^•OH signal was much higher than that for the Fe-S system. This result suggests that the introduction of Ag/AgCl into Fe-S induces an increase in the amount of ^•OH radicals produced in photo-Fenton systems.

We performed EIS measurements of the systems based on Fe-S, Ag/AgCl, and Ag/AgCl/Fe-S to investigate the charge transfer resistance and the separation efficiency between the photogenerated electrons and holes in the different catalysts [43]. According to previous reports [16, 44], a lower charge transfer resistance indicates more effective separation of electron-hole pairs. The results in Fig. 17 show that the charge transfer resistance of Ag/AgCl/Fe-S was lower than that of the Fe-S and Ag/AgCl individual components suggesting that the combination of Ag/AgCl and Fe-S effectively promoted separation and transfer of photogenerated electron-hole pairs.

To further understand the transfer and recombination processes of the photogenerated electron hole pairs in the studied plasmonic catalyst, we measured the photoluminescence (PL) emission spectra of Ag/AgCl and Ag/AgCl/Fe-S under excitation at 362 nm. As shown in Fig. 18, two emission peaks at approximately 475 and 550 nm were observed in the PL spectrum of pure Ag/AgCl, which indicated recombination of photo-induced electron and holes [16, 45]. The intensity of the emission peaks of Ag/AgCl/Fe-S was clearly lower than that of Ag/AgCl. This result suggests that hydroxyl-iron on the sepiolite surface acted as a trap to capture photo-induced electrons, and thus inhibited recombination of electron-hole pairs.

On the basis of the above results, a possible mechanism for BPA degradation catalyzed by Ag/AgCl/Fe-S and H₂O₂ under visible light is proposed. As shown in Fig. 19, under visible light irradiation, the metallic Ag nanoparticles absorb photons by a surface plasmon resonance effect to generate electron-hole pairs. The plasmon-excited electrons rapidly migrate to the conduction band (CB) of AgCl [46], leaving holes on the surface of the Ag nanoparticles, which can oxidize BPA. The electrons in the CB of AgCl are captured by Fe³⁺ to generate Fe²⁺, which can then react with H₂O₂ to generate ^•OH radicals through a typical Fenton reaction. Thus, the ^•OH radicals and holes are the main active species that degrade BPA in this system.

A promising heterogeneous photo-Fenton catalyst, Ag/AgCl/Fe-S, was fabricated by impregnation of Ag/AgCl onto a hydroxy-iron modified sepiolite. The photo-Fenton catalytic activity of the catalyst was tested under various reaction conditions with BPA as a target contaminant. Our results showed that the Ag/AgCl/Fe-S catalyst exhibited excellent activity and stability under visible light illumination. Active species capture experiments revealed that the major active species in the heterogeneous photo-Fenton system catalyzed by Ag/AgCl/Fe-S were ^•OH radicals and holes. These active species exhibited synergistic effects in the heterogeneous photo-Fenton system, leading to higher photocatalytic activity than that of Ag/AgCl and Fe-sepiolite.