Boosting PMS Activation Through Fe₃S₄/WO₃: The Essential Impact of W_X and S_X on Catalyst Activity and Regeneration Fe Active Sites for Efficient Pollutant Removal

Abstract

Fe-based heterogeneous catalytic advanced oxidation processes show great potential for treating wastewater. However, catalyst instability often hinders their practical use, mainly due to the slow regeneration of Fe²⁺ sites. Herein, we developed a Fe₃S₄/WO₃ catalyst, where the electron-rich Wx and Sx sites promoted efficient electron transfer, enabling continuous regeneration of Fe²⁺ active sites on the catalyst surface. The Fe₃S₄/WO₃ catalyst exhibited outstanding degradation efficiency for tetracycline (TC) in the peroxymonosulfate (PMS) system, achieving a 92.5% removal efficiency, significantly higher than its individual components of Fe₃S₄ (52.8%), WO₃ (43.1%), and WS₂ (53.2%). Moreover, the Fe₃S₄/WO₃/PMS system demonstrated a broad operational pH range (3.0–9.0), excellent degradation efficiency for various emerging pollutants, minimal interference from background electrolytes and organic matter, and strong stability in real water treatment. Chemical scavenger tests and electron paramagnetic resonance (EPR) analysis confirmed that the oxidative degradation of TC was driven by multiple reactive species, including SO₄^•−, ^•OH, ^•O₂⁻, and ¹O₂. This study provides a novel strategy for regulating active sites in Fe-based catalysts to ensure sustained performance, offering a pathway for the rational design of next-generation Fenton-like catalysts for efficient and sustainable micropollutant removal from wastewater.

1. Introduction

Antibiotics are among the most extensively utilized pharmaceutical products globally, finding widespread application in human healthcare, agriculture, livestock farming, and aquaculture [1,2,3]. TC, as one of the most frequently utilized antibiotics globally, is of particular interest due to its inherent strong hydrophilic nature and low volatility, which contribute to its notable persistence in aquatic environments [4,5,6]. Currently, TC has been detected in various environmental matrices, including soil, surface water, groundwater, and even drinking water [7]. This widespread presence highlights the urgent need for effective methods to remove TC. Among the commonly employed techniques for TC removal are adsorption [8], biofilm method [9], ultrasonication [10], and advanced oxidation processes [11]. In these mentioned treatment methods, catalytic advanced oxidation processes utilizing PMS (peroxymonosulfate) or PDS (peroxydisulfate) have garnered significant attention. This is due to several advantages, such as the production of SO₄^•− radicals, which possess a very high oxidation potential (2.5–3.1 V) and a longer half-life, making them highly effective for the removal of pollutants [12,13]. The primary methods for activating persulfate involve several techniques, including UV activation [14], thermal activation [15], ultrasonic activation [16], base activation [17], and transition metal activation (Au, Fe, Co, Mn, and Cu) [18,19,20,21,22,23].

Researchers especially show greater interest in persulfate activation processes involving transition metals as homogeneous or heterogeneous catalysts, owing to their enhanced reactivity, cost efficiency, and user-friendly operation [24]. The homogeneous activation of persulfate is less appealing for real application due to heavy-metal pollution, the low pH dependency, and the challenges associated with recycling. In contrast, heterogeneous catalysts can be more easily separated from treated wastewater for reuse and exhibit a wider effective pH range during pollutant removal. In heterogeneous catalysts, Fe-based catalysts stand out as the best choice due to their abundance, high efficiency, non-toxicity, environmental friendliness, and affordability. Over the past few years, various Fe-based catalysts, including iron oxide (Fe₂O₃ and Fe₃O₄) and bimetallic Fe-catalysts (Co-Fe, Cu-Fe, and Mn-Fe), have been predominantly used in persulfate systems [25,26,27,28,29,30,31]. Notably, iron oxides tend to be less reactive, mainly because a significant portion exists as Fe³⁺, which reduces their effectiveness in AOPs. This limitation is further illustrated by the study of Ruan et al. [32] which showed that the complete removal of trichloroethylene (TCE) required 63 mM of persulfate and 5 g/L of catalyst, and the process took five hours. Similarly, bimetallic Fe-catalysts such as CoFe₂O₄, CuFe₂O₄, MnFe₂O₄, etc., were considered very effective due to their vigorous catalytic activity during pollutant treatment [33]. Their good catalytic activity is related to two redox pair reactions (Equation (1)). Due to the differences in redox potential, electrons commonly transfer from the lower valence metal Mⁿ⁺ (with a lower redox potential) to the higher valence metal M⁽ⁿ⁺¹⁾⁺ (with a higher redox potential) [33]. However, bimetallic Fe-catalysts are more reactive, as they can regenerate the Fe²⁺ active site at the expense of another active site (Equation (1)). However, such catalysts face stability and leaching issues due to frequent changes in lattice defects. These changes weaken the coordination structure, resulting in poor stability during recycling studies and posing environmental concerns.

Fe³⁺ + Mⁿ⁺ → Fe²⁺ + M⁽ⁿ⁺¹⁾⁺

(1)

To tackle the issue of instability of Fe-based catalysts, considerable focus was placed on employing solid reductants on the catalyst’s surface. For example, Lu et al. [34,35] transferred Fe₃O₄ into Fe₃O₄@MoS₂ by coating the Fe₃O₄ surface with MoS₂. Similarly, Li et al. [36] and Yi et al. [37] successfully prepared FeOOH@WS₂ and FeOOH@MoS₂ by loading WS₂ or MoS₂ on the surface of FeOOH. Building on this approach, our team has developed innovative catalysts by incorporating MoS₄²⁻ or reduced graphene oxide (rGO) along with Fe ions into the structure of FeMgAl-layered double hydroxides [38,39,40]. The superior performance and enhanced stability of these catalyst systems were attributed to the minimum agglomeration and rapid Fe²⁺/Fe³⁺ redox cycle. Greigite (Fe₃S₄) is a naturally occurring metallic mineral known for its high electronic conductivity and strong magnetic properties. The study of Lin et al. [41] demonstrated that Fe₃S₄ can effectively remove bisphenol A in a PMS-based system. The reductive properties of sulfide in Fe₃S₄ enhance its catalytic activity by promoting the rapid regeneration of Fe²⁺ active sites. However, the practical application of Fe₃S₄ faces challenges due to the poor stability and low recyclability, which are caused by significant particle agglomeration and active site leaching. Tungsten, with its lower charge transfer resistance and greater reductive capacity, offers enhanced regeneration of the Fe²⁺ oxidation state in Fe-based catalysts [36]. Nguyen et al. [42] leveraged the reductive nature of WO₃ to modify the catalytic performance of Fe₂O₃, resulting in a Fe₂O₃/WO₃ composite that exhibited superior performance compared to individual Fe₂O₃. Therefore, incorporating WO₃ into the structure of Fe₃S₄ is expected to improve the catalyst performance and stability by enhancing active site regeneration and minimizing agglomeration.

In this study, we synthesized a bimetallic Fe₃S₄/WO₃ catalyst using a one-step hydrothermal method and utilized it to activate PMS for the degradation of TC. The catalyst’s morphology, structure, porosity, and chemical properties were systematically characterized using XRD, SEM, BET, XPS, and electrochemical techniques. The performance of the Fe₃S₄/WO₃ catalyst was comprehensively evaluated through its catalytic activity in both laboratory and real water samples. Additionally, we conducted an in-depth investigation of the Fe₃S₄/WO₃/PMS system, examining factors such as the catalyst dosage, PMS concentration, TC concentration, pH influence, efficiency against various organic pollutants, and catalyst stability. Ultimately, we establish a plausible PMS activation mechanism and confirm the regeneration of the Fe²⁺ oxidation state through extensive use of chemical scavengers, EPR, electrochemical tests, and XPS analysis.

2. Results and Discussion

2.1. Characterization

The Fe₃S₄/WO₃ catalyst was characterized using XRD, SEM, BET, and XPS analysis. First, XRD analysis was performed to investigate the crystal structure of Fe₃S₄/WO₃. As shown in Figure 1A, the XRD pattern displayed distinct peaks at 2θ values of 15.63°, 30.50°, 36.25°, 53.17°, and 61.53°, which correspond to the (111), (311), (400), (440), and (533) crystal planes of Fe₃S₄ (PDF#16-0713) [43]. Additionally, the weak peaks observed at 2θ values of 23.64°, 24.36°, 41.80° and 53.85° were attributed to the (002), (200), (222), and (042) crystal planes of WO₃ (PDF#20-1323) [44]. These observations confirmed the presence of characteristic peaks for both Fe₃S₄ and WO₃ phases. However, the crystallinity of the crystal planes of WO₃, such as the (002), (022), and (222) planes, was significantly lower than that of Fe₃S₄. Specifically, the crystallinity values for WO₃ were 6.72%, 13.2%, and 6.89% for the (002), (022), and (222) planes, respectively, compared to the higher crystallinity of Fe₃S₄ (Table S1). This suggests that Fe₃S₄ is more crystalline and better ordered than WO₃ in the composite. Further analysis of the Fe₃S₄/WO₃ composite revealed significant lattice distortion, as indicated by the lattice constant and strain data (Table S1). This distortion is likely due to the incorporation of WO₃ into the Fe₃S₄ structure. The low dislocation density suggests that the crystal structure of the catalyst is relatively complete and has fewer defects, which could enhance its catalytic performance. The average crystallite size (D) was determined to be 20.29 nm using the Scherrer equation (Equation (2)) and 47.4 nm using the Williamson–Hall method (Equations (2)–(4)). The larger size obtained from the Williamson–Hall method could be attributed to the presence of impurity phases, which increase the average crystallite size in the W-H plots. The positive slope of the W-H plots (Supplementary Figure S1A) further indicated the presence of microstrains in the crystal lattice of Fe₃S₄/WO₃. Consistent with these findings, the Rietveld refinement data (Supplementary Table S2) and the refined XRD patterns (Supplementary Figure S1B) showed that the calculated crystallinity and content of WO₃ were lower than those of Fe₃S₄. This supports the overall conclusion that Fe₃S₄ dominates the crystalline structure of the composite, while WO₃ contributes less to the overall crystallinity.

$$D={\displaystyle \frac{K\lambda }{{\beta }_{D}cos\theta }}$$

(2)

$${\beta }_{\epsilon }=4\epsilon tan\theta$$

(3)

$${\beta }_T={\beta }_D+{\beta }_{\epsilon }$$

(4)

where ${\beta }_T$ is the total broadening, ${\beta }_D$ is the FWHM (i.e., broadening of the peak) in radians, $K$ = 0.9 is the slape factor, $\lambda$ = 0.15406 nm is the wavelength of X-ray source, D is the crystallites size, $\theta$ is the angle corresponding to the examined peak, and $\epsilon$ is the strain.

Furthermore, the SEM analysis was used to observe the surface morphology of the prepared Fe₃S₄/WO₃ catalyst. As shown in Figure 1B, it was evident that Fe₃S₄/WO₃ revealed two distinct morphologies: nanoflakes and granular agglomerated spots. The nanoflakes, as observed in the SEM images, were characterized by a dense arrangement of small granular spots. These spots are uniformly distributed across the surface of the flakes, suggesting a high degree of dispersion. It is hypothesized that these granular spots could be WO₃ particles that have adhered to the nanoflake structure of Fe₃S₄. The coexistence of these two morphologies within a single catalyst could provide a synergistic effect, combining the stability and reactivity of Fe₃S₄. Additionally, the EDX analysis in Figure 1C shows the elemental composition of the catalyst, revealing the presence of Fe (21.33%), W (60.42%), and S (18.25%). The N₂ adsorption–desorption isotherm depicted in Figure 1D exhibits a typical type IV curve, marked by a significant rise in adsorption capacity at elevated pressures. This pattern suggests the existence of mesoporous structures, with an average pore size of 25.15 nm and a measured BET-specific surface area of 28.68 m²/g. In addition, the porosity was 5%, as obtained by Equation (5), suggesting that the adsorption of the catalyst was low. These characteristics create numerous active sites that enhance ion diffusion, thereby promoting efficient catalytic reactions.

$$P={\displaystyle \frac{V_0-V}{V_0}}\times 100\%$$

(5)

where $P$ is the material porosity, $V_0$ is the volume in natural state, and $V$ is the absolute dense volume of the material.

The surface elemental composition and electronic charge properties of the Fe₃S₄/WO₃ catalyst were investigated using XPS analysis. The full survey scan in Figure 2A reveals that the main elements present in the Fe₃S₄/WO₃ catalyst are Fe, W, O, and S, aligning with the results from energy-dispersive spectroscopy (EDS). Furthermore, the deconvoluted Fe 2p spectrum shown in Figure 2B indicates that iron exists predominantly in the Fe(II) and Fe(III) oxidation states, accounting for 69.52% and 30.48%, respectively. The peaks at 710.57 eV, 716.14 eV, 723.30 eV, and 726.08 eV correspond to Fe(II) 2p_3/2, Fe(III) 2p_3/2, Fe(II) 2p_1/2, and Fe(III) 2p_1/2, respectively [45]. The deconvoluted W 4f spectrum in Figure 2C displays two peaks at binding energies of 35.54 eV and 37.71 eV, indicating the presence of tungsten in both W(IV) and W(VI) oxidation states, accounting for 55.75% and 44.25%, respectively [46]. Additionally, the S 2p spectrum in Figure 2D shows peaks at 162.70 eV, 163.90 eV, and 168.70 eV, corresponding to sulfur in the S²⁻ and SO₄²⁻ states, with relative abundances of 85.5% and 14.5%, respectively [47]. Furthermore, the three peaks, as shown in Supplementary Figure S2A, at binding energies of 284.8 eV, 286.24 eV and 288.98 eV, were assigned to various functional moieties, including C-C, C-O, and O–C=O, respectively [48]. Similarly, the peaks at 530.31, 531.62, and 533.24 eV, as shown in Figure S2, account for lattice oxygen (O²⁻), O-H bonds, and adsorbed H₂O [36].

2.2. Catalytic Performance of Fe₃S₄/WO₃ in PMS System

The catalytic performance of Fe₃S₄/WO₃ was evaluated by assessing its ability to degrade TC as the target pollutant through the activation of PMS. As shown in Figure 3A, the adsorption of TC by Fe₃S₄/WO₃ was approximately 8.7%, indicating that the catalyst had limited adsorption capacity. In comparison, the degradation of TC using PMS alone was 42.3%, likely due to the direct redox reaction between TC and PMS. Different controlled catalytic systems such as WO₃/PMS, Fe₃S₄/PMS, and WS₂/PMS exhibited 43.6%, 47.9%, and 53.2%, respectively, indicating poor performance. In contrast, the Fe₃S₄/WO₃/PMS system achieved a remarkable TC removal rate of 92.5%, demonstrating the exceptional catalytic activity of the Fe₃S₄/WO₃ catalyst. Additionally, the kinetic analysis in Figure 3B revealed the reactivity order of the catalysts, following the sequence Fe₃S₄/WO₃ > WS₂ > Fe₃S₄ > WO₃. Furthermore, the results in Supplementary Table S3 provide a comparison between the Fe₃S₄/WO₃/PMS system and other well-known catalysts in the PMS system. Notably, the Fe₃S₄/WO₃/PMS system displayed a more rapid degradation of TC at significantly lower concentrations of PMS and catalyst. However, a direct comparison between these catalysts may not be entirely accurate due to the differences in reaction conditions. Based on the data presented in Figure 2B,C and Supplementary Table S3, the outstanding catalytic performance of Fe₃S₄/WO₃ is likely attributed to the synergistic effect between the two components of the catalyst. This synergy is attributed to the strong bonding interactions between Fe₃S₄ and WO₃ that facilitate efficient electron transfer and thereby promote the regeneration of Fe active sites through the reduction of Fe³⁺ ions to Fe²⁺. To understand the enhanced reactivity of the Fe₃S₄/WO₃ catalyst in the PMS system, the decomposition of PMS was monitored across various catalytic systems. As shown in Figure 3C, only 33% of PMS was decomposed in the WO₃/PMS system, while the decomposition reached 35% and 38% in the Fe₃S₄/PMS and WS₂/PMS systems, respectively. In contrast, the Fe₃S₄/WO₃/PMS system demonstrated superior performance, achieving a PMS decomposition of 51%. To further analyze the relationship between the PMS decomposition and the catalytic activity, the PMS stoichiometric efficiency was calculated using Equation (6) [49]. As illustrated in Figure 3D, the Fe₃S₄/WO₃/PMS system exhibited the highest PMS stoichiometric efficiency of 0.016, outperforming the WO₃/PMS (0.006), WS₂/PMS (0.011), and Fe₃S₄/PMS (0.013) systems. The exceptional performance of the Fe₃S₄/WO₃ catalyst can be attributed to the synergistic interaction between the Fe₃S₄ and WO₃ phases. These phases form robust bonding interactions, which facilitate efficient electron transfer and enhance the Fe³⁺/Fe²⁺ redox reactions. This synergy is crucial for reducing the Fe³⁺ to Fe²⁺ at the Fe₃S₄/WO₃ interface, thereby significantly boosting the overall catalytic efficiency of the system.

$$\mathrm{Stoichiometric} \mathrm{efficiency}={\displaystyle \frac{\mathsf{∆}\left[\mathrm{TC}\right]}{\mathsf{∆}\left[\mathrm{PMS}\right]}}$$

(6)

In Equation (6), Δ[TC] and Δ[PMS] represent the concentration of degraded pollutants (mmol) and decomposed PMS (mmol).

2.3. Factors Affecting the Performance of Fe₃S₄/WO₃/PMS System

Previous studies emphasize the significant impact of reaction pH on the interactions between catalysts, oxidants, and target pollutants that in turn greatly influences the performance of AOP [50]. As shown in Figure 4A, the Fe₃S₄/WO₃/PMS system demonstrated over 86% degradation of TC across a wide pH range (3–11) within 30 min, indicating that the catalyst maintains high catalytic activity across a broad range. However, some variations in the performance were observed under different pH conditions. For example, as the initial pH of the solution increased, the degradation rate of TC also increased, reaching its maximum at pH 7, suggesting that the initial pH plays a crucial role in the system’s performance. Under acidic conditions, the inhibition of TC degradation can be attributed to the leaching of active sites, thus reducing the number of available active sites on the catalyst surface. Additionally, at low pH, the O-O⁻ bond of PMS may interact with H⁺ ions through hydrogen bonding, hindering the interaction between PMS and the Fe₃S₄/WO₃ catalyst surface [51]. On the other hand, when the pH was increased to 11.0, the removal efficiency of the Fe₃S₄/WO₃/PMS system decreased to 86%. At a high pH, PMS could react with excess of OH⁻ ions, leading to the formation of SO₄²⁻ and reducing the efficiency of PMS to generate SO₄^•− radicals [52]. Additionally, the aggregation of OH⁻ ions on the catalyst surface enhances the electrostatic repulsive force between the catalyst and HSO₅⁻, which interferes with PMS activation. To investigate the impact of the electrostatic repulsion, the point of zero-charge pH (pHzpc) for Fe₃S₄/WO₃ was calculated to be 7.21 (Supplementary Figure S3). This explains that a change in solution pH > pHzpc would increase the catalyst’s negative surface charge, resulting in decreased PMS/TC–catalyst contact due to electrostatic repulsion. The broader pH range of the Fe₃S₄/WO₃/PMS system from pH 5–9 may be related to the pHzpc of Fe₃S₄/WO₃ (7.21), offering the strong adsorption for PMS (pKa1 < 0, pKa2 = 9.88) and TC (pKa1 = 3.3, pKa2 = 7.69, pKa3 = 9.69) through strong electrostatic interactions.

The generation of maximum free radicals is closely linked to the optimal ratio of catalyst dosage and PMS concentration. To achieve the highest performance of the Fe₃S₄/WO₃/PMS system, the optimal amounts of catalyst, PMS, and target pollutant were optimized through systematic study. The effect of Fe₃S₄/WO₃ dosage on the degradation of TC is shown in Figure 4B. Increasing the Fe₃S₄/WO₃ dosage from 0.06 to 0.1 g/L led to an increase in TC degradation efficiency from 85.5% to 92.5% within 30 min. Correspondingly, the observed kinetic constant (k) values also rose from 0.019 min⁻¹ to 0.024 min⁻¹. This indicates that a higher catalyst dosage provides more surface area and available active sites, facilitating the PMS activation and consequently improving the degradation rate of TC [36]. As depicted in Figure 4C, the removal rate of TC increased from 85.2% to 92.5% when the PMS dosage was raised from 1 mM to 2 mM. This enhancement is primarily due to the increased generation of free radicals, thus accelerating the TC degradation. However, the degradation rate did not continue to increase proportionally as PMS dosage was further increased to 3 mM and 4 mM. This subsequent decline in degradation efficiency could be attributed to the self-scavenging effect of free radicals, leading to the formation of non-reactive species such as S₂O₈²⁻ and weakly reactive radicals like SO₅^−• (Equations (7) and (8)) [51,53,54].

$${{\mathrm{HSO}}_5}^{-}+{{\mathrm{SO}}_4}^{-}•\to {{\mathrm{SO}}_4}^{2-}+{{\mathrm{SO}}_5}^{-}•+{\mathrm{H}}^{+}$$

(7)

$${{\mathrm{HSO}}_5}^{-}+•\mathrm{HO}\to {{\mathrm{SO}}_5}^{-}•+{\mathrm{H}}_2\mathrm{O}$$

(8)

Figure 4D illustrates the effect of varying the initial TC concentrations, ranging from 10.0 to 40.0 mg/L, on the performance of the Fe₃S₄/WO₃/PMS system. As shown, the degradation efficiency of the system decreased with higher initial TC concentrations. Specifically, the degradation rate constant for TC decreased from 0.024 min⁻¹ to 0.017 min⁻¹ as the initial TC concentration increased from 10.0 mg/L to 40.0 mg/L. This reduction in the degradation rate constant can be attributed to the saturation of active sites on the catalyst surface due to the excessive adsorption of TC. Additionally, the limited availability of PMS at higher TC concentrations may result in insufficient free radical production, with the radicals being quickly consumed by the increased amount of TC, thereby reducing the overall degradation efficiency.

To further explore the practical application potential of the Fe₃S₄/WO₃ catalyst, we conducted the degradation tests on a variety of compounds, including Rhodamine B (RhB), Methylene Blue (MB), Ciprofloxacin (CIP), Sulfamethoxazole (SMX), and Phenol. As shown in Figure 5A, the degradation efficiency for these compounds followed the order of MB > RhB > TC > CIP > Phenol > SMX. Notably, the Fe₃S₄/WO₃ catalyst demonstrated high reactivity, achieving removal efficiencies of over 80% for all tested pollutants. This indicates that the Fe₃S₄/WO₃ catalyst possesses broad applicability and underscores its potential for use in various industrial and environmental remediation scenarios. Additionally, the practical application potential of the Fe₃S₄/WO₃ catalyst was explored by testing its ability to degrade TC in real water samples sourced from a chemical industry, lake, and tap water. As shown in Figure 5B, although some inhibition was observed, the removal of TC was still significant, reaching 86.3% in tap water, 75.7% in lake water, and 60.2% in water from the chemical industry. The lower efficiency observed in the chemical-industry water sample can be attributed to its complex composition, which may interfere with the catalyst’s performance. Furthermore, the recyclability of catalysts is a critical factor in their practical application. To assess this, the stability of the Fe₃S₄/WO₃ catalyst was examined through five consecutive recycling experiments. As depicted in Figure 5C, the degradation of TC decreased from 92.5% to 84.1% after five cycles. This reduction in activity is likely due to the gradual loss of active sites on the catalyst’s surface [36]. Despite this modest decline, the Fe₃S₄/WO₃ catalyst maintained a removal efficiency above 80%, indicating its good recyclability. Additionally, as shown in Figure 5D, the unchanged structure of the Fe₃S₄/WO₃ catalyst underscores its excellent stability. This stability, combined with its consistent performance over multiple cycles, highlights the promising potential of the Fe₃S₄/WO₃ catalyst for practical wastewater treatment applications.

2.4. The Identification of Active Species in Fe₃S₄/WO₃/PMS System

In the persulfate system, different reactive species, including free radicals (OH, SO₄^•−), non-free radicals (¹O₂, surface complexation), or multiple reactive species, were reported previously [55,56]. To identify the potential radicals involved in the Fe₃S₄/WO₃/PMS system, a series of scavenging experiments were conducted using various scavengers, including methanol (MA), tert-butanol (TBA), p-benzoquinone (p-BQ), furfural alcohol (FFA), and L-histidine (L-his). In this study, MA was employed as a scavenger for SO₄^•− and ^•OH radicals, while TBA was used to trap ^•OH radicals [57,58,59]. As shown in Figure 6A, the addition of excess MA and TBA to the Fe₃S₄/WO₃/PMS system caused a decrease in TC removal, from 92.7% to 76.4% and 82.2%, respectively. This inhibition upon adding MA and TBA indicates the involvement of both SO₄^•− and ^•OH radicals in the Fe₃S₄/WO₃/PMS system. Notably, while MA and TBA both caused some inhibition, neither completely suppressed the TC removal, suggesting that multiple reactive species are involved in the Fe₃S₄/WO₃/PMS system. FFA, a scavenger for singlet oxygen (k = 1.2 × 10⁸ M⁻¹ s⁻¹), was used to assess the contribution of ¹O₂. As depicted in Figure 6A, the introduction of FFA resulted in a marked enhancement of the inhibitory effect, reaching 69% inhibition of TC degradation. This suggests that ¹O₂ plays a crucial role in the degradation process. To further confirm the involvement of ¹O₂, L-histidine (k= 3.2 × 10⁷ M⁻¹ s⁻¹), a more selective ¹O₂ scavenger, was introduced into the system [60]. As shown in Figure 6A, the inhibitory effect increased further to 78%, confirming the substantial contribution of ¹O₂ to TC degradation in the Fe₃S₄/WO₃/PMS system. Additionally, the presence of superoxide anion (^•O₂⁻) was also evaluated using p-benzoquinone (p-BQ, k = 0.9 × 10⁹ M⁻¹ s⁻¹) [60]. As shown in Figure 6A, the addition of p-BQ caused a 34.2% inhibition in TC degradation, indicating that ^•O₂⁻ also plays a role in the degradation process. These findings collectively highlight that a range of reactive species, including SO₄^•−, ^•OH, ¹O₂, and ^•O₂⁻, contribute to the efficient degradation of TC in the Fe₃S₄/WO₃/PMS system.

The findings from chemical scavenging experiments were further verified by electron paramagnetic resonance (EPR) spectroscopy. This technique employs 5,5-dimethyl-1-pyrroline N-oxide (DMPO) and 2,2,6,6-tetramethylpiperidine (TEMP) as spin-trapping agents to identify SO₄^•−, ^•OH, ¹O₂, and ^•O₂⁻. As depicted in Figure 6B, the absence of distinct free radical signals was noted when PMS was used in conjunction with DMPO alone. Conversely, the co-application of Fe₃S₄/WO₃ with PMS yielded strong DMPO-^•OH signals, characterized by a four-line pattern with an intensity ratio of 1:2:2:1, alongside a six-line DMPO-SO₄^•− spectrum. These observations confirm the presence of ^•OH and SO₄^•− radicals within the Fe₃S₄/WO₃/PMS system. Additionally, the presence of ^•O₂⁻ and ¹O₂ was verified through supplementary EPR assessments. Figure 6C shows that no peaks were observed for DMPO in an ethanol solution containing only the catalyst or PMS. However, in the Fe₃S₄/WO₃/PMS system, significant characteristic signals for the DMPO-^•O₂⁻ adduct were evident, confirming the presence of ^•O₂⁻ in the system. Moreover, as shown in Figure 6D, no peaks were observed for TEMP in solutions containing only the catalyst. Yet, a distinct triplet peak associated with TEMP-¹O₂ was detected in the presence of PMS alone, suggesting the self-activation of PMS via the interaction between SO₅²⁻ and HSO₅⁻. The intensity of the TEMP-¹O₂ signal was markedly increased in the Fe₃S₄/WO₃/PMS system. Collectively, these results, in conjunction with the chemical scavenging data, affirm the generation of various reactive species, including SO₄^•−, ^•OH, ¹O₂, and ^•O₂⁻, which are instrumental in the degradation of TC in the Fe₃S₄/WO₃/PMS system.

2.5. Active Site Regeneration and Proposed PMS Activation Mechanism

Many previous investigations emphasize the significance of catalyst active site regeneration in the PMS-based AOP for wastewater treatment [61,62,63]. In the Fe₃S₄/WO₃ catalyst, the W and S sites are rich in electrons and have low redox potential (E° S²⁻/S = −0.407 V and E° W⁶⁺/W = −1.063 V), and they play a crucial role in enhancing the catalyst’s performance. The electron transfer mechanism is initially validated through XPS analysis, which involves comparing the Fe 2p, W 4f, and S 2p spectra of the Fe₃S₄/WO₃ catalyst before and after the reaction. As illustrated in Figure 2B,C and Figure 7A,B, the proportion of Fe³⁺ and W⁶⁺ declined from 30.48% and 44.25% to 22.28% and 35.72%, respectively, while the content of Fe²⁺ and W⁴⁺ increased from 69.52% and 55.75% to 77.12% and 64.28%, respectively. These results clearly demonstrate that Fe³⁺ and W⁶⁺ on the catalyst surface underwent partial reduction to Fe²⁺ and W⁴⁺ during the reaction. This finding supports the hypothesis that shifts in the oxidation states of iron and tungsten during the reaction can provide significant insights into the processes of electron transfer and catalyst regeneration. Furthermore, the alterations in the oxidation states of sulfur during the reaction provide additional support for this mechanism [64,65]. As shown in Figure 2D and Figure 7C, the percentage of S²⁻ dropped from 85.5% to 67.23%, while SO₄²⁻ increased from 14.5% to 32.77% after the catalytic reaction. Considering the reducing power of S²⁻ (E° S²⁻/S = −0.407 V compared to E° Fe³⁺/Fe²⁺ = 0.76 V), these changes may be associated with the electron transfer mechanism. This mechanism promotes the reduction of Fe³⁺ to Fe²⁺ and enhances the regeneration of active sites in the Fe₃S₄/WO₃ catalyst. A deeper understanding of the electron transfer mechanism was obtained through various electrochemical studies, including linear sweep voltammetry (LSV), electrochemical impedance spectroscopy (EIS), and chronoamperometry (CA) tests. As depicted in Figure 7D,E, the LSV and EIS results strongly indicate that the Fe₃S₄/WO₃ catalyst possesses superior conductivity and enhanced electron transfer capabilities compared to other control catalysts. To further elucidate the direction of electron transfer, CA experiments were conducted. As shown in Figure 7F, the abrupt changes in current upon the addition of PMS or TC provided direct evidence of electron transfer from the catalyst to PMS or to TC. The combined analyses from XPS and electrochemical tests clearly demonstrate that the tungsten (W) and sulfur (S) sites in the Fe₃S₄/WO₃ catalyst play a crucial role in accelerating the rate of electron transfer and the regeneration of Fe²⁺ active sites. This regeneration is vital to the exceptional performance of the Fe₃S₄/WO₃ catalyst in PMS-based systems, as it ensures a continuous supply of active sites for the catalytic reactions, thereby maintaining high catalytic efficiency.

Expanding on the above discussion, we have developed a plausible mechanism (Figure 8) for the activation of PMS using the Fe₃S₄/WO₃ catalyst. Initially, the active site ≡Fe²⁺/W⁴⁺ on the Fe₃S₄/WO₃ surface can directly activate PMS, resulting in the formation of reactive species (Equation (9)). Additional insights from XPS and electrochemical studies suggest that both W and S are involved in electron transfer processes throughout the reaction sequence (Equations (10)–(14)). This involvement facilitates the repeated regeneration of Fe²⁺, leading to the production of multiple reactive species. This highlights the essential role of W and S in sustaining the catalytic activity of the Fe₃S₄/WO₃ catalyst.

≡Fe²⁺/W⁴⁺ + HSO₅⁻ → ≡Fe³⁺/W⁶⁺ + SO₄•⁻ + OH⁻

(9)

2Fe³⁺ + W⁴⁺ → 2Fe²⁺ + W⁶⁺

(10)

2S²⁻ + ≡Fe³⁺ → ≡Fe²⁺ + S₂²⁻

(11)

2S₂²⁻+ ≡Fe³⁺ → ≡Fe²⁺ + Sn²⁻

(12)

Sn²⁻ + ≡Fe³⁺ → ≡Fe²⁺ + S⁰

(13)

S⁰ + ≡Fe³⁺ → ≡Fe²⁺ + SO₄²⁻

(14)

3. Materials and Methods

3.1. Chemicals and Reagents

Sodium tungstate dihydrate (≥99.5% Na₂WO₄·2H₂O), ferrous sulfate heptahydrate (≥99% FeSO₄·7H₂O), acetone (≥98.0% C₃H₆O₃), sodium nitrate (≥99.0% NaNO₃), sodium chloride (99.0% NaCl), sodium sulfate (≥99.0% Na₂SO₄), sodium carbonate (≥99.8% Na₂CO₃), sodium bicarbonate (≥99.5% NaHCO₃), sodium dihydrogen phosphate (≥99.0%, NaH₂PO₄·2H₂O), sodium acetate (≥99.0% CH₃COONa·3H₂O), ethanol (≥99.7% C₂H₆O), hydrogen chloride (≥99.0% HCl), sodium hydroxide (≥96.0% NaOH), potassium iodide (≥99.8%KI), silver nitrate (≥99.8% AgNO₃), and Rhodamine B (≥99.8% C₂₈H₃₁CIN₂O₃) were all purchased from Sinopharm Chemical Reagent Co, Ltd. (Beijing, China). Thiourea (≥99.8% CH₄N₂S), L-histidine (≥98.0% L-his), tungsten disulfide (≥99.7% WS₂), and tetracycline (≥99.0% C₂₂H₂₄N₂O₈) were purchased from Shanghai Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Ammonium sulfide (≥99.0% H₈N₂S), furfuryl alcohol (98.0% FFA), methanol (≥99.5% MA), tert-butanol (≥99.0% TBA), and potassium persulfate (≥98.0% 2KHSO₅·KHSO₄·K₂SO₄) were purchased from Shanghai Maclean’s Biochemical Technology Co. Ltd. (Shanghai, China).

3.2. Preparation of Fe₃S₄/WO₃

In a typical synthesis procedure, 1.48 g of ferrous sulfate, 3.75 g of sodium tungstate, and 1.027 g of thiourea were dissolved in 100 mL of deionized water. After stirring magnetic for 30 min, a uniform mixed solution was obtained. This solution was then poured into a 150 mL stainless-steel autoclave for a hydrothermal treatment at 180 °C for 24 h. After the reaction was complete, the resulting solid was allowed to cool down to ambient temperature. The solid was subsequently washed repeatedly with deionized water and acetone to remove any impurities, and the washing process continued until the filtrate was clear. The residue after filtration was dried in a vacuum oven at 60 °C for 6 h. Finally, the dried powder was subjected to calcination at 400 °C for 3 h in a muffle furnace and named the Fe₃S₄/WO₃ catalyst. This whole preparation process of catalyst synthesis is illustrated in Figure 9.

3.3. Catalysts Characterization

The morphology of the catalyst was characterized by scanning electron microscopy (SEM, ZEISS 300, Jena, Germany). The crystal structure of the catalyst was characterized by X-ray powder diffractometer (XRD, Shimadzu, Tokyo, Japan, XRD-700, copper source), and the test range was 10~90°. The specific surface area, pore size, pore volume, and other characteristics of the catalyst are determined using nitrogen adsorption–desorption (BET) isotherms, measured with an ASAP 2460 instrument (Agilen, Santa Clara, CA, USA). X-ray photoelectron spectroscopy (XPS, Thermo Fisher Scientific Escalab 250 Xi, Waltham, MA, USA) was used to analyze the surface elements of the catalyst and the oxidation states of the elements before and after the reaction. All the electrochemical tests and electron transfer capabilities were performed on a CHI660E electrochemical workstation (Shanghai Chenhua Instrument Co., Ltd., Shanghai, China). A three-electrode system was used. The glassy carbon electrode coated with the catalysts was used as the working electrode (WE), the platinum sheet was used as the counter electrode (CE), and the Ag/AgCl was used as the reference electrode (RE). In the electrochemical test process, the electrolyte was 0.05 M Na₂SO₄ solution, the working voltage was open-circuit voltage, and the range was 0.2~1.2 V. For LSV measurements, the scan rate was set to 5 mV/s, and the range was 0~1.2 V. For EIS, the test was performed on the different catalysts at a potential of 1.5 V. A sinusoidal voltage with an amplitude of 5 mV and a scanning frequency ranging from 100 kHz to 1 Hz were applied during the measurements. The test time of the chronoamperometry current is 500 s, and the measurement voltage is 1.2 V. The EIS spectrum was fitted using Zview (4.5.0.56400) and the detailed test parameters are shown in Supplementary Table S4.

3.4. Analytical Methods

The degradation experiments were carried out in a 100 mL beaker. First, a certain quantity of catalyst was introduced into a 10 mg/L solution of TC, and the solution was allowed to achieve adsorption equilibrium by resting for 30 min. Subsequently, 2 mM of PMS was added to initiate the reaction. During the reaction, 3 mL samples were taken at a certain time interval, filtered with 0.22 μm membrane, and 0.5 mL of MA was added to terminate the reaction. The concentrations of TC and other pollutants were detected by high-performance liquid chromatography (HPLC, Agilen, Santa Clara, CA, USA). All the degradation experiments were conducted at uncontrolled pH in triplicate, and the data presented in the figures represent the average values, accompanied by standard deviations. In the quenching experiment, quenching agents such as MA, TBA, FFA, p-BQ, and L-his were used to identify the active species in the system. The degradation degree of TC was analyzed by pseudo-first-order kinetics equation. The residual PMS in the solution was measured by a method reported previously [39,66,67,68,69].

4. Conclusions

In this study, a novel Fe₃S₄/WO₃ catalyst was synthesized using the hydrothermal method. Various characterization techniques, including XRD, SEM, EDX, BET, and XPS analysis, were employed to confirm the successful synthesis of the Fe₃S₄/WO₃ structure. The Fe₃S₄/WO₃ catalyst demonstrated exceptional performance in the PMS system, achieving over 92.5% degradation of tetracycline (TC) in less than 30 min. This efficiency surpasses that of various control catalysts, including Fe₃S₄ (52.8%), WO₃ (43.1%), and WS₂ (53.2%). The enhanced catalytic activity of the Fe₃S₄/WO₃ catalyst is attributed to the presence of electron-rich W_X and S_X sites within its structure that facilitate synergistic interactions between Fe, W, and S. These interactions promote the continuous reduction of Fe³⁺ sites, ensuring a steady supply of active sites through constant electron transfer. Furthermore, the Fe₃S₄/WO₃/PMS system demonstrated remarkable efficiency across a broad pH range from 3.0 to 9.0 and effectively degraded a wide spectrum of organic pollutants. The system exhibited good practical application potential, evidenced by its excellent stability over multiple cycles and sustained performance in real water samples collected from lakes, rivers, and chemical industries. The chemical scavengers, EPR analysis, electrochemical tests, and XPS analysis confirmed the generation of multiple reactive species (SO₄•⁻, •OH, •O₂⁻, and ¹O₂) and the effective regeneration of the Fe²⁺ state during PMS activation. Overall, this study underscores the significant advantages of employing the Fe₃S₄/WO₃ as a novel Fenton-like catalyst in wastewater treatment applications.