High-Efficient Elimination of Spiramycin by Fe₃O₄/ZSM-5/Sch via Heterogeneous Photo-Fenton Oxidation at Neutral pH

Abstract

Spiramycin (SPM), a widely employed antibiotic in both clinical therapy and the livestock industry, poses significant challenges in terms of safe and efficacious management. A heterogeneous photo-Fenton system, devised using Schwertmannite (Sch), can effectively degrade contaminants. However, it is accompanied by a relatively low conversion efficiency of ≡Fe³⁺/≡Fe²⁺ and a significant iron loss. In this study, a catalyst featuring Fe₃O₄ and ZSM-5 molecular sieve-modified Sch (Fe₃O₄/ZSM-5/Sch) was devised to enhance the catalytic activity and stability. The findings revealed that Fe₃O₄/ZSM-5/Sch exhibited exceptional catalytic activity, with the reaction first-order kinetic exceeding that of pure Sch. The active species including ·OH, h⁺, e⁻, ·O₂⁻ and SO₄^·− were identified in the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system. The enhanced catalytic activity of Fe₃O₄/ZSM-5/Sch could be ascribed to the effective conversion of ≡Fe³⁺/≡Fe²⁺. The photogenerated electrons within Fe₃O₄ were transported to Sch via ZSM-5, which effectually reduced ≡Fe³⁺/≡Fe². Moreover, Fe₃O₄/ZSM-5/Sch demonstrated outstanding stability; even after six cycles, the degradation efficiency of SPM remained above 86.50%, and the leaching quantity of Fe remained below 0.24 mg/L. This research not only develops an excellent catalyst for the safe treatment of SPM but also proffers innovative perspectives for the future design of efficient iron-based catalysts.

1. Introduction

Spiramycin (SPM), a broad-spectrum macrolide antibiotic, is extensively employed in the livestock industry and clinical hospital treatments [1,2]. Owing to the incomplete absorption of SPM within the animal’s intestinal tract, a substantial quantity of unmetabolized SPM and its metabolites are excreted externally via urine and feces, subsequently entering urban wastewater treatment plants [3]. However, conventional wastewater treatment facilities cannot entirely remove antibiotics, leading to their eventual release into aquatic and terrestrial environments through treated wastewater and sewage sludge [4]. The continuous exposure to residual antibiotics in the environment can enhance bacterial antibiotic resistance, giving rise to multi-drug-resistant strains and even the emergence of superbugs [5]. This poses significant implications for ecosystems and human health. Consequently, the pollution and treatment of SPM in the environment have garnered considerable attention.

Advanced oxidation processes (AOPs) are widely employed in the treatment of organic pollutants in wastewater, utilizing highly oxidative reactive species (such as ·OH, ·OOH and O₂^·−) generated during the reaction process to effectively oxidize contaminants in water without selectivity, ultimately decomposing them into CO₂ and H₂O [6]. Compared to traditional homogeneous Fenton processes (Fe²⁺/H₂O₂), heterogeneous Fenton using iron-based solid catalysts offers advantages such as an expanded pH range of applicability and the prevention of generating copious iron-containing sludge [7,8,9]. Within the heterogeneous Fenton process, the type of iron-based catalyst proves to be a crucial determinant of degradation efficiency.

Schwertmannite (Sch) is a sulfate mineral rich in iron and hydroxide, with a chemical formula of Fe₈O₈(OH)_8-2x(SO₄)_x · nH₂O (1 ≤ x ≤ 1.75) [10], primarily found in acidic mine waters and acid sulfate soils. As a heterogeneous Fenton catalyst, Sch possesses the ability to activate hydrogen peroxide (H₂O₂) to generate hydroxyl radicals (·OH, E₀ = 2.83 eV) for pollutant degradation. Due to its ease of availability and low cost, it has been studied and applied in the degradation of water contaminants [11,12,13]. The conversion of Fe³⁺ to Fe²⁺ is a prerequisite for triggering Fenton reactions. However, as Fe³⁺ predominantly occupies the Sch surface, Sch-driven heterogeneous Fenton reactions typically experience extended induction periods [14], resulting in reduced catalytic efficiency. Therefore, enhancing catalytic activity by accelerating the conversion of Fe³⁺/Fe²⁺ is of great significance. It has been reported that UV can rapidly initiate the redox reactions between Fe³⁺ and Fe²⁺, decomposing H₂O₂ to generate ·OH for pollutant degradation. In addition to constructing photo-driven heterogeneous Fenton processes, another direct approach is to modify Sch with the aid of other catalysts [15].

Magnetite (Fe₃O₄), readily obtainable from the natural environment, boasts a unique inverse spinel structure that facilitates electron transfer through the Fe₃O₄ body itself, accelerating the redox reaction between ≡Fe³⁺ and ≡Fe²⁺. Fe₃O₄, as a suitable co-composite catalyst, can enhance the catalytic activity of other iron oxides. However, following the composite formation of Fe₃O₄/Sch, agglomeration was still noticeable, and the surface area was not particularly expansive, merely 59.7 m²/g [16]. Another drawback of Fe₃O₄/Sch is that the iron leaching from the active components is high (2.7 mg/L), exceeding the discharge standard stipulated by the European Union (2 mg/L) [17].

Previous studies have reported that modifying catalysts with carriers can provide a structural framework for the active components of the catalyst, enabling their full distribution on the surface of the composite catalyst [7,18]. This exposes more active sites and simultaneously enhances the mechanical strength of the catalyst, which is beneficial for its stability. Compared to other carriers (clay, bentonite, etc.), synthetic zeolites, with their larger surface area, are more conducive to the adsorption of smaller organic molecules. This adsorption function is believed to be advantageous for the degradation of pollutants, thus offering extensive application prospects. 10-Membered oxygen ring zeolites possess the advantages of high hydrothermal stability, high coke resistance and high activity. Among the family of 10-Membered oxygen ring zeolites, MFI type (ZSM-5) zeolite is a microporous molecular sieve with a shape-selective high silica three-dimensional pore structure. Its unit cell composition is Na_nAl_nSi_96-nO₁₉₂·16H₂O (n ≤ 27), where n is the number of aluminum atoms in the unit cell. The ZSM-5 molecular sieve has a large surface area, which can effectively enhance the catalytic activity of the catalyst and boost its stability. For instance, Kasiri et al., using ZSM-5 as a carrier and loading it with Fe³⁺, synthesized the Fe-ZSM5 composite catalyst [19]. In the UV/Fe-ZSM5/H₂O₂ system, under optimal conditions (21.4 mmol/L of H₂O₂, 0.5 g/L of catalyst and pH 5), 51.28% of Acid Blue 74 could be degraded in 120 min, and the iron ion leaching was less than 0.3 mg/L. However, studies on simultaneously modifying Sch with Fe₃O₄ and ZSM-5 have not yet been reported.

Thus, the objectives of this study are: (1) to fabricate a highly active catalyst using Fe₃O₄ and ZSM-5-modified Sch, (2) to evaluate the catalytic activity of the Fe₃O₄/ZSM-5/Sch composite material through the degradation of SPM and explore the mechanism of catalytic activity enhancement and (3) to assess the recyclability and stability of Fe₃O₄/ZSM-5/Sch during repeated usage. The findings of this study can provide a novel eco-friendly material for the removal of SPM in water.

2. Materials and Methods

2.1. Synthesis of the Catalysts

According to the method described in previous studies, Sch was synthesized in the laboratory using chemical methods [20]. Specifically, 11.12 g of FeSO₄‧7H₂O was weighed and dissolved in 494 mL of deionized water under stirring until fully dissolved. Then, 6 mL of 30% (v/v) H₂O₂ was added dropwise to the solution within 10 min under magnetic stirring. The resulting mixture was incubated at 180 rpm and 28 °C for 24 h, and the precipitate formed in the system was collected using a 0.45 μm microporous membrane filter. To remove soluble impurities from the precipitate, it was washed three times with 500 mL of acidic water (pH = 2.0) and 500 mL of deionized water, respectively. The collected precipitate was then freeze-dried for 24 h and stored for future use. The sample prepared in this manner was labeled as Sch.

The Fe₃O₄/ZSM-5/Sch catalyst was prepared by adding 1 g of Fe₃O₄ and 6 g of ZSM-5 powder during the synthesis of Sch. In brief, 1.0 g of nano-Fe₃O₄ was weighed and added to a conical flask containing 494 mL of acidified water (pH = 2.0). The mixture was sonicated for 10 min to obtain a suspension. Then, 11.12 g of FeSO₄‧7H₂O was added to the suspension and stirred until fully dissolved. Subsequently, 6 g of an H-type ZSM-5 molecular sieve with a silicon-to-aluminum ratio of 110 was added, and the mixture was sonicated for an additional 3–5 min. The subsequent steps were the same as those of the method described earlier for synthesizing Sch. The resulting sample was labeled as Fe₃O₄/ZSM-5/Sch.

2.2. Photocatalytic Activity Experiments

The catalyst (30 mg) was added into a 50 mL SPM solution with a concentration of 10 mg/L. After sonication for 2–3 min, the mixture was placed into a multi-channel photoreactor (PCX 50C Discover, Beijing Perfect light Technology Co., Ltd., Beijing, China) and irradiated with UV light (365 nm, 3 W). A given amount of H₂O₂ (0.5 mmol/L) was added to start the reaction, and the temperature was maintained at 25 ± 3 °C. At 0, 30, 60, 90, 150 and 180 min, 2 mL samples were taken and quenched with 2 mL of methanol. The resulting mixture was filtered through a 0.45 μm filter, and the absorbance was measured at a UV wavelength of 232 nm. Each experiment was performed with three parallel replicates.

The effects of the initial pH values (3, 5, 6.52, 7, 9, 11), H₂O₂ concentrations (0, 0.5, 1, 2, 4, 8 mmol/L) and catalyst loading amounts (0, 0.1, 0.6, 1.2, 2.4 and 4.8 g/L) on the degradation efficiency of SPM were investigated. The impact of free radical scavengers on the reaction system was also studied. Specifically, for the photocatalytic degradation of SPM, 20 vol.% tertbutanol (TBA, to scavenge ·OH), 20 vol.% methanol (to scavenge ·OH and SO₄^·−), 10 mmol/L potassium iodide (KI, to scavenge h⁺), 0.5 mmol/L AgNO₃ (to scavenge e⁻) and 0.5 mmol/L p-benzoquinone (BQ, to scavenge ·O₂⁻) were added. The recyclability and stability of the catalyst were evaluated through six continuous degradation experiments. After each experiment, the catalyst was separated by centrifugation and directly reused in the next experiment.

2.3. Analytical Methods

The concentration of SPM was measured using a UV spectrophotometer (752N, Shanghai Jingke, Shanghai, China) at a wavelength of 232 nm. The TOC was measured using a TOC analyzer (TOC-L, Shimadzu, Beijing, China). The active radicals were analyzed using electron paramagnetic resonance (EPR, Bruker EMX-10/12, Karlsruhe, Baden-Württemberg, Germany). The concentrations of Fe²⁺, Fe³⁺ and total iron were determined using the 1,10-phenanthroline colorimetric assay at 510 nm on a UV–vis spectrophotometer (752N, Shanghai Jingke, Shanghai, China). The degradation intermediates of SPM were analyzed using liquid chromatography–mass spectrometry (LC–MS, TSQ QUANTUM ACCESS MAX, Thermo Scientific, Waltham, MA, USA).

The morphology and characteristics of the catalyst were analyzed using scanning electron microscopy (SEM, FEI Quanta 400 FEG, Hillsboro, OR, USA), transmission electron microscopy (TEM, FEI Talos F200x, Hillsboro, OR, USA), energy dispersive spectroscopy (EDS, FEI Talos F200x, Hillsboro, OR, USA), Brunauer–Emmett–Teller (BET, Micromeritics ASAP2460, Norcross, GA, USA), X-ray diffraction (XRD, Rigaku Rotaflex D/max, Akishima, Tokyo, Japan), Fourier transform infrared (FTIR, NEXUS870, Madison, WI, USA) spectra, zeta potential measurements (Malvern Nano-ZS90, Malvern, Worcestershire, UK), UV–vis diffuse reflectance spectra (DRS, PE-Lambda-750, Waltham, MA, USA), X-ray photoelectron spectroscopy (XPS, ESCALAB 25, Waltham, MA, USA), cyclic voltammetry (CV, CHI 660E, Shanghai Chenhua, Shanghai, China), electrochemical impedance spectroscopy (EIS, CHI 660E, Shanghai Chenhua, Shanghai, China), fluorescence spectroscopy analysis (PL, FLS980, Edinburgh, UK) and transient photocurrent response (CHI 660E, Shanghai Chenhua, Shanghai, China). The detailed test method can be found in the Supplementary Materials.

2.4. Calculation

The conduction band (CB) and valence band (VB) energies were calculated as described in reference [21]. The specific calculation method can be found in the Supplementary Materials.

3. Results and Discussion

3.1. Characterization of Catalysts

The Sch particles appeared as spherical aggregates with an average particle size of around 600–700 nm (Figure 1a). When Fe₃O₄ and ZSM-5 were added to the Sch precursor solution, the synthesized Fe₃O₄/ZSM-5/Sch presented an irregular, filamentous spherical appearance with an average particle size of 400–600 nm (Figure 1b). Compared to Sch alone, the Fe₃O₄/ZSM-5/Sch particles showed a reduction in agglomeration and an increase in dispersion, with more pronounced surface edges and roughness. The specific surface areas of the Sch and Fe₃O₄/ZSM-5/Sch catalysts were 20.04 m²/g and 205.59 m²/g, respectively (Table S1). The surface of the Fe₃O₄/ZSM-5/Sch catalyst prepared showed an irregular massive structure, indicating that a small number of ZSM-5 particles were embedded in the edge of the Sch structure. Additionally, TEM images of Fe₃O₄/ZSM-5/Sch catalysts revealed that Fe₃O₄ and ZSM-5 promoted Sch growth on the surface as heterogeneous nuclei (Figure 1c). This finding was supported by high-resolution transmission electron microscopy (HRTEM) results (Figure 1d). The lattice fringe spacings of 0.21 nm and 0.26 nm were characteristic of Fe₃O₄ and Sch, respectively, corresponding to the (400) surface of Fe₃O₄ and the (212) surface of Sch. Meanwhile, some creases and dislocations could also be observed in the HRTEM diagram of the Fe₃O₄/ZSM-5/Sch catalyst, indicating that Fe₃O₄/ZSM-5/Sch possessed structural defects and poor crystallinity. This unique defect structure was beneficial for the activation of H₂O₂ [22]. In addition, the EDS-mapping elemental profiles of Fe, O, S, Si and Al elements recorded on the Fe₃O₄/ZSM-5/Sch catalyst (Figure 1e–f) further indicated that Fe, O, S, Si and Al elements were uniformly distributed in all regions of the Fe₃O₄/ZSM-5/Sch composite catalyst.

XRD patterns revealed that the characteristic diffraction peaks of Fe₃O₄, Sch and ZSM-5 all appear in the Fe₃O₄/ZSM-5/Sch structure (Figure 2a). The diffraction angles 2θ were 18.45°, 30.36°, 35.76°, 43.47° and 63.16°, which were in good agreement with the standard Fe₃O₄ diffraction data (JCPDS 75-0449), but the diffraction peak intensity decreased. This suggested that Fe₃O₄ mainly existed within the composite catalyst rather than on the surface [16]. A diffraction peak with more burrs was observed at 2θ, which was 35.16°, consistent with the comparison with the standard Sch card (JCPDS 47-1775). However, the intensity of the diffraction peak was stronger than that of pure Sch, and the peak shape also changed, which may be due to the slight alteration of the Sch structure caused by the addition of Fe₃O₄. The diffraction peaks in the 2θ range of 23°~24° were consistent with those of the ZSM-5 standard (JCPDS 37-0359), whereas the intensity of the ZSM-5 main diffraction peaks decreased, indicating that Fe₃O₄ and Sch were dispersed on the surface of ZSM-5 molecular sieves, forming a certain degree of crystallization. The infrared spectrum of Fe₃O₄/ZSM-5/Sch also revealed the coexistence of Fe₃O₄, ZSM-5 and Sch (Figure 2b). The absorption peaks observed at 1250 cm⁻¹, 450–790 cm⁻¹ and 550 cm⁻¹ were characteristic vibrational peaks of the MFI topology structure of ZSM-5, corresponding to Si-O-Si(Al), Si(Al)-O and the anti-symmetric stretching vibration absorption peak of the double five-membered rings in the ZSM-5 framework [23,24]. These peaks indicated that the framework structure of ZSM-5 remained intact as a carrier for Fe₃O₄/ZSM-5/Sch. The absorption peaks at 981–1121 cm⁻¹ and 3385 cm⁻¹ corresponded to the stretching vibration absorption peaks of SO₄²⁻ and O-H, respectively [20]. This indicated that the Fe₃O₄/ZSM-5/Sch composite contained numerous -OH and SO₄²⁻ functional groups, which could capture holes (h⁺) and electrons (e⁻) [25,26].

The light absorption properties of a catalyst are a crucial factor influencing its photocatalytic performance [27]. Sch exhibited a distinct absorption band in the 200–580 nm range, Fe₃O₄ demonstrated a strong absorption intensity within the 200–800 nm range, and ZSM-5 possessed a relatively low absorption intensity between 200 and 800 nm (Figure 2c). Compared to pure Sch, Fe₃O₄/ZSM-5/Sch displayed a noticeable redshift, presenting an absorption band within the 200–800 nm range, which was advantageous for enhancing photocatalytic activity. The respective bandgap widths (E_g) of Fe₃O₄, Sch and Fe₃O₄/ZSM-5/Sch were 0.93, 2.28 and 1.39 eV (Figure 2d). Calculations reveal that the conduction band (CB) of Fe₃O₄ (0.82 eV) was lower than that of Sch (1.14 eV); thus, upon generating photoelectrons in the composite catalyst, Fe₃O₄ could rapidly transfer them to the Sch surface, accelerating the reduction in ≡Fe³⁺.

3.2. Catalytic Activity for Heterogeneous Fenton

The degradation effect of SPM was evaluated in the UV/Fe₃O₄-H₂O₂, UV/ZSM-5-H₂O₂, UV/Sch-H₂O₂, UV/Fe₃O₄/Sch-H₂O₂, UV/ZSM-5/Sch-H₂O₂, UV/Fe₃O₄/ZSM-5/Sch, UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ and Fe₃O₄/ZSM-5/Sch-H₂O₂ systems (Figure 3a). The degradation of SPM was fitted to a pseudo first-order kinetic model (R² > 0.96). In the UV/Fe₃O₄-H₂O₂ system, the pseudo first-order rate constant (k) was extremely low (k < 0.0033 min⁻¹), indicating that ·OH could not be rapidly and effectively produced in the system. This is primarily due to the agglomeration of Fe₃O₄, which reduces the active sites for H₂O₂ activation [28]. Within the UV/ZSM-5-H₂O₂ system, the k was 0.0047 min⁻¹. Given the large surface area of ZSM-5 (Table S1), its electrostatic adsorption of SPM could be the primary cause for SPM removal [29]. In the UV/Sch-H₂O₂ system, the k amounted to 0.0128 min⁻¹, indicating that Sch can effectively activate H₂O₂ to generate ·OH. Previous studies demonstrated that ≡Fe³⁺ within Sch was the predominant iron species, which must first be reduced to ≡Fe²⁺ and then utilized to activate H₂O₂, generating ·OH to degrade pollutants, leading to an induction period and rapid degradation stage [30]. However, no induction period appeared in the UV/Sch-H₂O₂ system, primarily owing to UV’s acceleration of the reduction in ≡Fe³⁺ within Sch. Upon combining ZSM-5 and Sch, within the UV/ZSM-5/Sch-H₂O₂ system, the k ascended to 0.0139 min⁻¹, principally attributed to a greater dispersion of Sch upon loading onto ZSM-5, exposing a larger number of active sites and generating a greater amount of ·OH.

When Fe₃O₄ and Sch were combined, within the UV/Fe₃O₄/Sch-H₂O₂ system, the k increased to 0.0158 min-1. Li et al. reported that after Fe₃O₄ was combined with Sch, the electrons in Fe₃O₄ readily transfer to Sch, promoting the reduction in Fe³⁺ in Sch to Fe²⁺, generating ·OH and improving the degradation rate of pollutants [28]. When Fe₃O₄ and ZSM-5 were simultaneously combined with Sch, within the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system, the k further increased to 0.0203 min⁻¹, indicating that Fe₃O₄/ZSM-5/Sch had excellent catalytic performance. Without UV irradiation, the k in the Fe₃O₄/ZSM-5/Sch-H₂O₂ system was only 0.0063 min⁻¹, suggesting that UV irradiation significantly enhanced the catalytic activity of Fe₃O₄/ZSM-5/Sch. Moreover, in the UV/Fe₃O₄/ZSM-5/Sch system, the k was the lowest, only 0.0008 min⁻¹, indicating that adsorption was not the main mechanism for the removal of SPM in the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system.

The Fe₃O₄/ZSM-5/Sch catalyst exhibited satisfactory performance within a pH range of 3–9 (Figure 3b), suggesting that Fe₃O₄/ZSM-5/Sch broadened the applicable pH range for degrading SPM in heterogeneous Fenton systems. Surprisingly, unlike other iron-based catalysts, the Fe₃O₄/ZSM-5/Sch catalyst exhibited the highest catalytic activity when the initial solution was unadjusted (pH = 6.52). The main reasons were as follows: (1) SPM was a weakly alkaline substance with different acid dissociation constant values (pK_a of 7.1 and 8.4), and the amine and hydroxyl groups in its molecular structure were protonated and deprotonated, respectively, depending on pH changes [4]. When the pH was adjusted to 3, SPM assumed a cationic state due to protonation, resulting in a stable structure. The Zeta potential results indicated that the point of zero charge (pH_pzc) value of Fe₃O₄/ZSM-5/Sch was 3.1 (Figure S1). At a pH of 3, the surface of Fe₃O₄/ZSM-5/Sch also carried a positive charge (pH < pH_pzc), and the electrostatic force inhibited the adsorption of SPM on the Fe₃O₄/ZSM-5/Sch surface, ultimately reducing SPM degradation. (2) As the pH increased from 5 to 6.52, SPM became positively charged, while the negative charge on the Fe₃O₄/ZSM-5/Sch surface increased. This results in enhanced catalytic activity due to electrostatic attraction. (3) When the pH increased from 6.52 to 7, the oxidation potential of ·OH decreased, and iron hydroxide complexes were more likely to form on the Fe₃O₄/ZSM-5/Sch surface, covering active sites [31] and reducing SPM degradation. (4) When the pH increased from 7 to 9 (pH_pzc < pK_a < pH), SPM became deprotonated, carrying a negative charge, and the negative charge on the Fe₃O₄/ZSM-5/Sch surface continued to increase. The enhanced electrostatic repulsion resulted in a slight decrease in SPM degradation. Considering that domestic wastewater or natural water typically had a neutral pH, the subsequent experiments were conducted under natural pH conditions.

This study also explored the influence of the H₂O₂ concentration (Figure 3c). When the H₂O₂ concentration increased from 0 to 0.5 mmol/L, the k (R² > 0.98) rose from 0.0008 min⁻¹ to 0.0167 min⁻¹. This result was similar to the previous study of Luo et al., who found that trimethoprim degradation increased when the persulfate concentration increased [32]. However, a further increase in H₂O₂ led to a decrease in the k value. Excess H₂O₂ reacts with ·OH to produce ·HO₂, which is less reactive (Equation (1)). This finding was consistent with the results of Su et al. [33]. Meanwhile, the effect of the Fe₃O₄/ZSM-5/Sch catalyst dosage was investigated (Figure 3d). When the catalyst dosage increased from 0 g/L to 0.6 g/L, the k (R² > 0.98) increased from 0.0037 min⁻¹ to 0.0197 min⁻¹. A further increase in the catalyst dosage to 4.8 g/L, however, led to a decrease in the k value to 0.0126 min⁻¹. There are mainly two reasons: (1) An excess of the Fe₃O₄/ZSM-5/Sch catalyst can hinder the propagation of UV light in the system, prevent efficient electron transfer in the catalyst, produce insufficient Fe²⁺ and result in an insufficient amount of ·OH generated [34]; (2) A large quantity of the catalyst can easily lead to agglomeration [35], preventing the full exposure of active sites and reducing the removal efficiency of SPM.

·OH + H₂O₂ → H₂O + ·HO₂

(1)

3.3. Mechanism Consideration of SPM Degradation

3.3.1. Identification of Active Species

It is common knowledge that ·OH represents the active free radical in Fenton reactions. Upon introducing TBA (a potent scavenger of ·OH) to the reaction system (Figure 4a), the degradation efficiency of SPM declined from 95.21% to 18.17%, indicating that ·OH was the primary reactive free radical. When an abundance of ·OH formed, it could activate trace-dissolved SO₄²⁻ in Sch through the reaction SO₄²⁻ + ·OH → SO₄^·− + OH⁻, generating SO₄^·− (a free radical with a longer lifespan and high redox potential). Methanol was considered a more robust scavenger of both ·OH and SO₄^·−, further reducing the degradation efficiency to 7.13%. This suggested that the reaction was not entirely inhibited, and other reactive species were involved in the degradation of SPM. In addition to producing ·OH radicals, the photoexcited heterogeneous Fenton system also generates a hole (h⁺) and electron (e⁻), as well as ·O₂⁻. When KI was introduced to the system, the degradation efficiency of SPM decreased by 40.37%, primarily through the generation of ·OH via the reaction H₂O/OH⁻ + h⁺ → ·OH + H⁺ [36]. Upon adding AgNO₃ to the system, the SPM degradation efficiency dropped to 69.97%, indicating that e⁻ played a significant role in accelerating the reduction in ≡Fe³⁺. BQ, a scavenger of ·O₂⁻, only slightly inhibited the degradation efficiency of SPM. By quantitatively calculating inhibition levels (Figure 4b, Figure S2 and Table S2), the contribution of the five generated reactive species to SPM degradation followed the order: ·OH (93.90%) > h⁺ (77.80%) > e⁻ (62.11%) > ·O₂⁻ (16.73%) > SO₄^·− (3.57%). This indicated that ·OH played the dominant role in the degradation of SPM and was the main active species in the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system. Furthermore, EPR analysis delved deeper into the reactive species involved in the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ reaction. In the Fe₃O₄/ZSM-5/Sch catalyzed photo-Fenton system, a four-line signal peak of DMPO-·OH was observed, with a characteristic peak ratio of 1:2:2:1, indicating the production of ·OH (Figure 4c). In addition to the strong DMPO-·OH signal, a faint six-line signal peak with a 1:1:1:1:1:1 ratio was observed, attributable to DMPO-SO₄^·−. When the reaction solvent was methanol, a weak four-line signal peak was observed (with an intensity ratio of 1:1:1:1), indicating the generation of ·O₂⁻ (Figure 4d). The EPR analysis was consistent with the quenching results in Figure 4a.

3.3.2. Enhanced Mechanism for Catalytic Activity

During the heterogeneous Fenton process, some of the active iron species may leach from the catalyst surface and be released into the solution, forming a homogeneous Fenton system with H₂O₂ in the system, thereby contributing to the degradation of pollutants [16]. However, the leached Fe concentration in the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system remained consistently within a low range (Figure S3a), indicating the exceptional stability of the Fe₃O₄/ZSM-5/Sch structure. An analysis of the degradation efficiency of the UV/Fe³⁺-H₂O₂ system, constructed by the leached Fe³⁺ ions, indicated that it contributed approximately 26.77% to the degradation of SPM (Figure S3b), suggesting that the degradation process was mainly dominated by the heterogeneous Fenton reactions.

XPS analysis was employed to compare the changes in the valence states of Fe 2p_3/2 on the surface of Fe₃O₄/ZSM-5/Sch before and after the reaction (Figure 5a). The Fe 2p_3/2 peaks located at 710.89 eV and 712.87 eV correspond to ≡Fe²⁺ and ≡Fe³⁺, respectively [37,38]. Prior to the reaction, the molar ratio of ≡Fe²⁺/≡Fe³⁺ in Fe₃O₄/ZSM-5/Sch was 0.50 (Table S3). Upon the reaction completion, a shift in the binding energy of the two individual Fe 2p_3/2 peaks was observed, moving towards a higher binding energy direction. This suggested that redox reactions occurred between Fe²⁺ and Fe³⁺ on the catalyst surface during the reaction. After the reaction, the ratio of ≡Fe²⁺ to ≡Fe³⁺ in Fe₃O₄/Sch/ZSM-5 increased to 1.25, a 2.5-fold increase from its pre-reaction state. This was greater than the increase factor for both the ratio of ≡Fe²⁺ to ≡Fe³⁺ in Fe₃O₄/Sch (1.25) [28] and the ratio of ≡Fe²⁺ to ≡Fe³⁺ in Sch (1.67) [17].

It is well known that the reduction of ≡Fe³⁺ to ≡Fe²⁺ is the rate-limiting step in iron-based catalyst-mediated heterogeneous Fenton reactions, and the efficiency of the conventional H₂O₂-induced reduction in ≡Fe³⁺ is low. Under ultraviolet light irradiation, Fe₃O₄ and Sch in Fe₃O₄/ZSM-5/Sch were excited to generate a photogenerated electron (e⁻)–hole (h⁺) pair, and ZSM-5 acted as an electron transfer carrier to accelerate electron transfer. A photogenerated electron could rapidly reduce ≡Fe³⁺ to ≡Fe²⁺. Interestingly, the conduction band edge of Fe₃O₄ (+0.82 eV) and Sch (+1.14 eV) was more positive than the redox potential of ≡Fe³⁺/≡Fe²⁺ (+0.77 eV), making them unable to reduce ≡Fe³⁺. However, under UV irradiation at an energy of 3.40 eV (λ = 365 nm), e⁻ in the valence bands of Fe₃O₄ and Sch could be excited to higher potential edges (−1.62 eV and 0.02 eV, respectively). In the Fe₃O₄/ZSM-5/Sch composite catalyst, the modified e⁻ reduced ≡Fe³⁺ to ≡Fe²⁺, and the e⁻ in Fe₃O₄ could also reduce O₂ to ·O₂⁻ (−0.33 eV). The CV curve further demonstrated that Fe₃O₄/ZSM-5/Sch had the highest reduction current (Figure 5b). Additionally, EIS measurements (Figure 5c) revealed that Fe₃O₄/ZSM-5/Sch had the smallest radius, indicating the lowest charge transfer resistance at the electrolyte interface and a faster electron transfer rate.

Photogenerated e⁻–h⁺ pairs are prone to recombination, leading to a decrease in the oxidation and reduction abilities of photocatalysts. However, h⁺ was the main active species involved in the reaction (Figure 4b), indicating the effective separation of e⁻–h⁺ in the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system. The PL spectrum (Figure 5d) showed that Fe₃O₄/ZSM-5/Sch had the smallest fluorescence emission peak, indicating a high separation efficiency of photogenerated e⁻–h⁺ pairs in Fe₃O₄/ZSM-5/Sch [39]. Furthermore, it was found that the photocurrent density generated by the Fe₃O₄/ZSM-5/Sch composite catalyst was significantly higher than that of Fe₃O₄, Sch and ZSM-5 (Figure 5e). This suggested that rapid charge transfer occurred within Fe₃O₄/ZSM-5/Sch, which may be due to the electron donor or acceptor characteristics of the ZSM-5 molecular sieve, promoting the separation of charge carriers and hindering e⁻–h⁺ recombination [40,41].

Therefore, Fe₃O₄/ZSM-5/Sch could easily be excited by UV irradiation, resulting in the separation of e⁻–h⁺ pairs (Equation (2)). The oxidative h⁺ reacted with H₂O₂ to generate ·OH, while e⁻ promoted the production of ≡Fe²⁺ and ·O₂⁻ (Equations (3)–(5)). Subsequently, the more efficient conversion of ≡Fe³⁺/≡Fe²⁺ led to the generation of a large amount of ·OH in the system, enhancing the catalytic activity of Fe₃O₄/ZSM-5/Sch (Equations (6)–(9)). In addition, when ·OH was produced in the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system, SO₄^·− was mainly formed through the redox reaction between SO₄²⁻ and ·OH (Equation (10)). Finally, SPM adsorbed on the surface of Fe₃O₄/ZSM-5/Sch reacted with ·OH, ·O₂⁻ and SO₄^·−, degrading into small molecular substances, which were eventually mineralized into H₂O and CO₂ (Equation (11)). The enhancement mechanism of the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system is shown in Figure 5f.

Fe₃O₄/ZSM-5/Sch → h⁺ + e⁻

(2)

H₂O/OH⁻ + h⁺ → ·OH + H⁺

(3)

O₂ + e⁻ → ·O₂⁻

(4)

≡Fe(III) + e⁻ → ≡Fe(II)

(5)

≡Fe(II) + H₂O₂ → ≡Fe(III) + ·OH + OH⁻

(6)

≡Fe(III) + H₂O₂ → ≡Fe(II) + ·HO₂ + H⁺

(7)

≡Fe(II) + ·HO₂ + H⁺ → ≡Fe(III) + H₂O₂

(8)

≡Fe(III) + ·HO₂ → ≡Fe(II) + H⁺ + O₂

(9)

SO₄²⁻ + ·OH → SO₄^·− + OH⁻

(10)

SPM + ·OH, SO₄^·−, ·O₂⁻ → intermediate products → H₂O + CO₂

(11)

3.4. SPM Degradation Pathways

The main intermediates of SPM during the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ reaction were determined by LC–MS analysis and summarized in Figure S4 and Table S4. Based on the identified intermediates, we proposed the transformation pathway of SPM in this system (Figure 6).

(1) Aldoxy oxidation (Pathway 1): The two dimethylamine (DMA) groups at the C5 and C9 positions of SPM were first attacked by ·OH, generating TP859-1 and TP859-2. Upon ·OH attack, TP859-1 or TP859-2 underwent oxidation to form TP875. Subsequently, the aldehyde group at the C6 position of TP875 was oxidized by ·OH to form a carboxylic acid group, resulting in the formation of TP891. (2) C-O bond cleavage (Pathway 2 and Pathway 3): Under the attack of active radicals (·OH, ·O₂⁻, SO₄^·−, e⁻ and h⁺), the disaccharide (mycaminose and mycarose) linked at the C5 position lost the outermost mycarose, resulting in the formation of TP699. Subsequently, the C-O bond at the C9 position was cleaved, generating TP159 and TP685. Subsequently, the C-O bond at the C5 position of TP685 was cleaved, resulting in the formation of TP352 and TP335. Moreover, because TP159, TP699 and TP335 contain DMA functional groups, these three TPs can generate DMA through the cleavage of N–C bonds. (3) Demethylation (Pathway 4): The N–C bond of forosamine at the C9 position of SPM and the N–C bond at the mycaminose position connected to the C5 position were attacked by active radicals to undergo a demethylation reaction, forming TP814. TP814 was further oxidized to form DMA. Finally, DMA, TP699, TP352, TP335 and TP814 generated in Pathways 2–4 could be further mineralized into CO₂ and H₂O. In this process, approximately 23.45% of TOC was mineralized to CO₂ (Figure S5). The proposed SPM degradation pathway in the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system was consistent with the reaction pathways in UV/nano-zerovalent iron/peroxyacetic acid oxidation [42] and visible light/sulfur-doped g-C₃N₄/persulfate oxidation [43].

3.5. Reusability and Stability of the Fe₃O₄/ZSM-5/Sch Catalyst

The reusability and stability of catalysts play an important role in practical applications. When the Fe₃O₄/ZSM-5/Sch composite catalyst was reused for six cycles, the degradation efficiency of SPM remained above 86.50% (Figure 7a), clearly indicating that Fe₃O₄/ZSM-5/Sch could be repeatedly used as a heterogeneous Fenton-like catalyst. As seen in Figure S3a, before the recovery of Fe₃O₄/ZSM-5/Sch, the leached Fe concentration was 0.24 mg/L, far lower than the iron leaching amount (2.7 mg) in Fe₃O₄/Sch [28]. After Fe₃O₄/ZSM-5/Sch underwent six cycles, the leaching amount of Fe remained in the range of 0.18–0.23 mg/L (Figure 7b), which is also lower than the dissolution amount of Fe (0.38–0.58 mg/L) after multiple cycles of Sch [17]. According to the European Union emission standards (<2 mg/L) [17], this was acceptable.

As shown in Figure 2a,b, the XRD and FT-IR spectra of the reused Fe₃O₄/ZSM-5/Sch still showed little difference compared to the XRD and FTIR spectra of the newly synthesized Fe₃O₄/ZSM-5/Sch. These results indicated that the photocatalytic Fenton process did not destroy the structure and functional groups of Fe₃O₄/ZSM-5/Sch. Additionally, the photocurrent curves of the Fe₃O₄/ZSM-5/Sch catalyst did not show a decay trend after multiple cycles of light interruption (Figure 5e). These results suggested that Fe₃O₄/ZSM-5/Sch had good stability.

4. Conclusions

The UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ system is capable of efficiently removing SPM from water. The composition of Fe₃O₄ and ZSM-5 reduced the aggregation phenomenon of chemically synthesized Sch, significantly increased the edge-ridged structure on the surface and greatly enlarged the specific surface area. At the same time, compared to the Sch-mediated Fenton reaction, the degradation efficiency of SPM was substantially improved. Moreover, Fe₃O₄/ZSM-5/Sch exhibited broad applicability, with an initial pH range of 3–9, and demonstrated excellent heterogeneous Fenton-like activity for SPM degradation under near-neutral pH conditions. The effective decomposition of H₂O₂ driven by the Fe₃O₄/ZSM-5/Sch heterogeneous photocatalytic Fenton catalyst was mainly related to the efficient conversion of ≡Fe³⁺/≡Fe²⁺. Before the reaction, the molar ratio of ≡Fe²⁺/≡Fe³⁺ in Fe₃O₄/Sch/ZSM-5 was 0.50. After the reaction, the molar ratio of ≡Fe²⁺/≡Fe³⁺ increased to 1.25. Under UV irradiation, Fe₃O₄ and Sch in Fe₃O₄/ZSM-5/Sch generated hole–electron pairs. The photogenerated electrons in Fe₃O₄ were transferred to Sch via ZSM-5, which effectively reduced the ≡Fe³⁺ in Sch to ≡Fe²⁺, which then reacted with H₂O₂ to generate ·OH. The photogenerated electrons in Fe₃O₄ also reacted with O₂ to generate a small amount of ·O₂⁻. Due to the effective transfer of electrons, the holes were allowed to react with surface H₂O and OH⁻. The abundant ·OH within the system effectively triggers the production of SO₄^·− from SO₄²⁻ in Sch. When ·OH, ·O₂⁻ and SO₄^·− coexist in the reaction system, the degradation efficiency of organic compounds was significantly enhanced. In the UV/Fe₃O₄/ZSM-5/Sch-H₂O₂ reaction, SPM can ultimately be converted into CO₂ and H₂O. After six cycles of reuse, the Fe₃O₄/ZSM-5/Sch as a catalyst in the heterogeneous photo-Fenton process maintained a degradation efficiency for SPM above 86.50%, with iron leaching maintained between 0.18 and 0.24 mg/L, indicating the good stability and reusability of Fe₃O₄/ZSM-5/Sch. This research has significant implications for the future design of efficient iron-based catalysts for the degradation of recalcitrant organic pollutants.