English

• With the acceleration of industrialization, antibiotic pollution has intensified and new antibiotics continue to emerge, posing a serious threat to the global water environment [1,2]. The electro-Fenton (EF) process generates hydrogen peroxide (H₂O₂) in situ by reducing oxygen (O₂) [3], followed by the production of various reactive oxygen species (ROS) [4,5], such as hydroxyl radicals (^•OH) [6], enabling efficient degradation of antibiotic pollutants [7]. The EF technology, as a type of advanced oxidation process (AOPs), has attracted widespread attention in the field of wastewater treatment in recent years due to its advantages such as high oxidation efficiency, no secondary pollution, and low generation of toxic intermediate byproducts [8]. These characteristics make it more in line with the current concept of "green development". However, despite its promising application potential, the EF process still faces several key challenges, such as the low efficiency of H₂O₂ generation and activation, poor selectivity of ROS, and relatively high overall energy consumption [9]. These limitations significantly restrict the practical implementation and widespread application of the EF technology in the treatment of antibiotic-containing wastewater. Therefore, how to achieve efficient generation and activation of H₂O₂, and thereby optimize the overall performance and cost-effectiveness of the EF system, has become an important research direction in this field [10].

  To promote the efficient in situ generation of H₂O₂, researchers have explored various materials as supports for electrocatalysts. Among them, biochar, as a novel carbon-rich material, has the advantages of large specific surface area, low cost, abundant mesoporous structures, self-doped heteroatoms, and fewer conjugated π structures, which contributes to the production of H₂O₂ [11-13]. It demonstrates the potential for efficient in situ generation of H₂O₂ via the 2e^- pathway for O₂ reduction [14,15]. Nevertheless, carbon-based materials still face certain limitations in activation of H₂O₂.

  Metallic nanoparticles (PCs) possess good stability, high activation capability, and abundant metal-metal active sites, which can effectively promote the generation of Fe²⁺/Fe³⁺ redox cycling through the modulation of metal-metal interactions, thereby enhancing H₂O₂ activation [16-18]. However, current PCs catalysts still suffer from insufficient control over metal site exposure and low atomic utilization, which can lead to high costs and poor catalytic efficiency, hindering the effective activation of H₂O₂. In contrast, single-atom catalysts (SACs) offer well-defined active sites, higher atomic utilization, and tunable coordination environments, making them promising candidates for efficient H₂O₂ activation [19-21]. Nevertheless, SACs face challenges such as inhomogeneous active site distribution and susceptibility of the metal centers to being occupied, which limit their catalytic performance [22,23]. Therefore, it has been proposed that when SACs and PCs coexist, the synergistic effect between metal entities of different scales has the potential to overcome the individual limitations of SACs and PCs, significantly enhancing the reaction activity, and thereby effectively improving the catalytic efficiency [24,25]. It is thus clear that the construction of cooperative active sites between SACs and PCs holds great promise for the precise regulation of the electronic structure and coordination environment of SACs, accelerating the Fe²⁺/Fe³⁺ redox cycle, and ultimately achieving efficient in situ activation of H₂O₂.

  In this study, based on the synergistic interaction between Fe SACs and Fe⁰/Fe₇C₃ PCs, we successfully constructed a Fe single-atom-nanoparticle composite catalyst (Fe_SAC-PC@NC) by employing a simple impregnation-calcination method. And it was applied to the EF system, achieving highly efficient removal of antibiotic pollutants. The incorporation of Fe⁰/Fe₇C₃ PCs can precisely regulate the electronic structure of Fe SAC sites, thereby significantly improving their adsorption capacity for H₂O₂ and enhancing electron transfer efficiency, which further promotes the efficient activation of H₂O₂. Moreover, in practical degradation tests, the catalyst achieved a removal efficiency of over 95% for sulfamethoxazole (SMX) within 60 min, and maintained an efficiency above 90% after four cycles, showcasing excellent catalytic activity and structural stability, and indicating great potential for real-world applications.

  All chemical reagents used in the experiment are shown in Text S1 (Supporting information). All experimental methods are shown in Texts S2-S9 (Supporting information).

  The Fe single-atom catalyst (Fe_SAC@NC) and its Fe_SAC-PC@NC counterpart were successfully synthesized by using a simple impregnation-pyrolysis method, with N-doped biochar serving as the support (Fig. 1a). Scanning electron microscopy (SEM) analysis reveals that Fe_SAC-PC@NC has a porous 3D structure, indicating a large specific surface area that is beneficial for enhancing catalytic activity (Fig. S1 of Supporting information). High resolution transmission electron microscope (HR-TEM) reveals that no obvious lattice structure is observed on the surface of Fe_SAC@NC particles, and weak carbon shell lattice fringes can be seen at the particle edges, which preliminarily confirms that Fe species in Fe_SAC@NC do not form PCs (Fig. S2a in Supporting information). In contrast, some obvious lattice structures can be observed on the surface of Fe_SAC-PC@NC (Fig. 1b), with lattice spacings of 1.97 Å and 2.45 Å, corresponding to the (110) plane of Fe and the (122) plane of Fe₇C₃, respectively, suggesting that Fe species in Fe_SAC-PC@NC exist in the form of nanoparticles. Further high-angle annular dark-field-scanning TEM (HAADF-STEM) images show that Fe SACs are uniformly dispersed on the carbon surface of Fe_SAC@NC at the material edges, and distinct lattice fringes with a spacing of 3.36 Å can be observed corresponding to the (111) plane of carbon, accurately confirming the presence of Fe in the form of SACs in Fe_SAC@NC (Fig. S2b in Supporting information). In Fig. 1c, Fe SACs and distinct lattice fringes are clearly visible on the surface of Fe_SAC-PC@NC with lattice spacings of 2.62, 2.98, 1.67, and 1.72 Å, corresponding to the (121) and (103) planes of Fe₇C₃, the (222) plane of C, and the (200) plane of Fe, respectively. This indicates that Fe in Fe_SAC-PC@NC exists both as Fe SACs and as Fe⁰/Fe₇C₃ PCs, further confirming the successful preparation of both Fe_SAC@NC and Fe_SAC-PC@NC materials. The energy-dispersive X-ray spectroscopy (EDS) mapping (Fig. 1d) demonstrates that C and N elements are uniformly distributed in the Fe_SAC-PC@NC material, while Fe is embedded in the carbon-nitrogen matrix in the form of particulates and atoms, which is consistent with the TEM and HAADF-STEM results. X-ray diffraction (XRD) analysis further confirms the successful synthesis of the Fe_SAC-PC@NC material with a synergistic combination of Fe SACs and Fe⁰/Fe₇C₃ PCs. As shown in Fig. 1e, no characteristic diffraction peaks of Fe are observed in Fe_SAC@NC. In contrast, Fe_SAC-PC@NC exhibits diffraction peaks at 29.88°, 33.48°, and 44.67°, corresponding to the (103) and (121) planes of Fe₇C₃ (PDF #75–1499) and the (110) plane of Fe (PDF #99–0064), respectively, indicating the formation of Fe⁰/Fe₇C₃ PCs in Fe_SAC-PC@NC. Additionally, a peak at 26.61° is observed in both materials, which corresponds to the (111) plane of graphite carbon (PDF #75–2078). This peak has a sharp shape, and the intensity is higher in Fe_SAC-PC@NC than that in Fe_SAC@NC, suggesting that the former contains a higher degree of graphitized carbon, which is beneficial for enhancing electron transport efficiency [26]. As exhibited in Fig. S3 and Table S1 (Supporting information), the isotherms of both catalysts exhibited type IV isotherms [27], which are consistent with mesoporous structures [28]. With the introduction of Fe⁰/Fe₇C₃ PCs, the specific surface area increases by approximately 1.5 times. The optimized catalyst exposes more active sites, thereby significantly enhancing its catalytic activity.

  Figure 1

  

  Figure 1.  (a) Schematic diagram of the preparation of Fe_SAC-PC@NC. (b) HR-TEM image, (c) HAADF-STEM image and (d) EDS images of Fe_SAC-PC@NC. (e) The XRD pattern of samples. High resolution XPS spectra of (f) Fe 2p and (g) N 1s.

  DownLoad: Full-Size Img PowerPoint

  X-ray photoelectron spectrometry (XPS) was conducted to investigate the chemical states of the elements on the catalyst surface. The survey spectra reveal the presence of C, N, and Fe elements (Fig. S4a in Supporting information). In the Fe 2p spectrum, characteristic peaks corresponding to Fe⁰ (707.2 eV), Fe-N (710.5 eV), Fe²⁺ 2p_3/2 (711.9 eV), and Fe³⁺ 2p_3/2 (713.9 eV) are observed, further confirming the coexistence of Fe SACs and Fe⁰/Fe₇C₃ PCs. The incorporation of Fe⁰ facilitates the redox cycle of Fe²⁺/Fe³⁺ (Eq. 1), thereby promoting the efficient progression of the EF reaction.

  $ \mathrm{Fe}^0+\mathrm{Fe}^{3+} \rightarrow \mathrm{Fe}^{2+} $

  (1)

  Additionally, the characteristic peaks in Fe_SAC-PC@NC are shifted toward lower binding energies, indicating an increase in electron cloud density around the Fe active sites (Fig. 1f). Fig. 1g displays the high-resolution XPS of N 1s, which can be divided into five peaks corresponding to pyridinic N (398.3 eV), pyrrolic N (400.1 eV), graphitic N (401.4 eV), N-O (404.5 eV) [29]. After the introduction of Fe⁰/Fe₇C₃ PCs, the content of graphitic N increases from 10.38% to 52.20%, which is beneficial for the highly selective generation of singlet oxygen (¹O₂) (Fig. S4b in Supporting information) [30].

  Based on the above characterizations, the catalyst with synergistic coordination of Fe SACs and Fe⁰/Fe₇C₃ PCs has been successfully fabricated. The introduction of nanoparticles precisely regulates the electronic structure of Fe active sites, optimizing their coordination environment and thereby enhancing catalytic activity. To further evaluate the electrochemical degradation performance of the Fe_SAC-PC@NC catalyst, we conducted degradation experiments using SMX as the target pollutant. As shown in Fig. 2a, the degradation efficiency of SMX exceeds 95% within 60 min in the Fe_SAC-PC@NC system, whereas the removal rate in the Fe_SAC@NC system is only 45%. The first-order kinetic fitting result indicates that the rate constant for Fe_SAC-PC@NC is 0.049 min^-1, which is 3.8 times higher than that of Fe_SAC@NC (0.013 min^-1), demonstrating the superior degradation activity of Fe_SAC-PC@NC toward pollutants. The catalyst itself showed no adsorption removal of the pollutant (Fig. S5 in Supporting information), indicating that the activation of H₂O₂ to ROS plays a crucial role in the rapid removal of SMX.

  Figure 2

  

  Figure 2.  (a) SMX removal in different systems. (b) Electrochemical impedance curves of samples. (c) LSV curves at 1600 rpm. (d) Electron transfer number (n) and (e) H₂O₂ selectivity of samples. (f) EPR spectrum of NC. (g) CV curves of Fe_SAC-PC@NC at different scan rates, and ECSA measurements of samples. (h) H₂O₂ concentrations of samples.

  DownLoad: Full-Size Img PowerPoint

  In order to gain a deeper understanding of the reaction mechanism, we conducted a series of electrochemical tests. The electrochemical impedance spectroscopy (EIS) (Fig. 2b) demonstrate that Fe_SAC-PC@NC exhibits smaller charge transfer resistance and the faster charge transfer rate, and thus exhibits the faster catalytic activity [31]. In addition, rotating ring-disk electrode (RRDE) is widely recognized as having a significant advantage in evaluating the oxygen reduction reaction (ORR) performance of materials [32]. The ORR polarization curves of the Fe_SAC-PC@NC were obtained using the RRDE in an aqueous solution containing 50.0 mmol/L Na₂SO₄ at pH 3.0, as depicted in Fig. 2c. After the introduction of Fe⁰/Fe₇C₃ PCs, the ring current of the sample significantly increases. The analysis of the electron transfer number (n, Fig. 2d) and H₂O₂ selectivity (Fig. 2e) reveals that Fe_SAC-PC@NC exhibits a n value closer to 2 and higher H₂O₂ selectivity compared to Fe_SAC@NC, indicating a significant improvement in the efficiency of H₂O₂ production [33]. The LSV curves of Fe_SAC-PC@NC at different rotation rates (Fig. S6 in Supporting information) show that the disk current increases steadily with the increase of rotation rate, indicating enhanced mass transport and electrochemical activity. The n value (Fig. S7a in Supporting information) and H₂O₂ selectivity (Fig. S7b in Supporting information) exhibit minimal variation, demonstrating the stable ORR performance of Fe_SAC-PC@NC under diverse rotational conditions. The EPR analysis (Fig. 2f) of biochar (NC) reveals the presence of abundant carbon defects on the catalyst surface. It has been reported that intrinsic carbon defects in carbon materials exhibit excellent selectivity for the electrochemical synthesis of H₂O₂ [34,35]. Therefore, it is inferred that the carbon defects on the catalyst surface in this system serve as effective active sites for H₂O₂ generation. In addition, as shown in Fig. 2g and Fig. S8 (Supporting information), the CV curves of the sample measured at various scan rates (20, 40, 60, 80, and 100 mV/s) exhibit a rectangular shape, indicating that the electrochemical behavior originated from the double-layer capacitance (C_dl) [36]. The calculated C_dl value for Fe_SAC-PC@NC is 0.0541 mF/cm², which is higher than those of Fe_SAC@NC and NC. This indicates that Fe_SAC-PC@NC possesses a larger effective electrochemical surface area (ECSA) [37], which provides more active sites and promotes the activation of H₂O₂. Furthermore, the concentration of H₂O₂ (Fig. 2h) confirmed the strong capability of carbon defects in H₂O₂ generation, while also demonstrating that the incorporation of PCs significantly enhances the efficient activation of H₂O₂. These results suggest that the synergistic interaction between Fe⁰/Fe₇C₃ PCs and Fe SACs effectively enhances electron transfer efficiency, enabling the efficient generation and activation of H₂O₂.

  In order to identify the dominant ROS in the system, we used t‑butyl alcohol (TBA), p-benzoquinone (p-BQ), l-histidine (l-his), and catalase (Cat.) to quench ^•OH, superoxide free radical (O₂^•-), ¹O₂, and H₂O₂ in the Fe_SAC-PC@NC system [38], respectively. As shown in Figs. 3a and b, after quenching ¹O₂ and ^•OH, SMX removal efficiency drops to 19.1% and 25.6%, respectively, with kinetic rate constants of 0.0033 and 0.0050 min^-1, suggesting that ¹O₂ and ^•OH are the main ROS driving SMX degradation. The electron paramagnetic resonance (EPR) measurements (Figs. 3c-e) confirm the presence of ^•OH, O₂^•-, and ¹O₂ in both Fe_SAC@NC and Fe_SAC-PC@NC systems. Fe_SAC@NC exhibits a stronger ^•OH signal peak at 10 min, while Fe_SAC-PC@NC shows a stronger ¹O₂ signal peak [39]. Based on these findings, it is inferred that the introduction of Fe⁰/Fe₇C₃ PCs enables the construction of a dual-site synergistic structure in SACs and PCs, which effectively regulates the activation pathway of H₂O₂, making it more favorable for the generation of ¹O₂. Fluorescence tests further support these trends by using terephthalic acid (TA) and singlet oxygen green fluorescent probe (SOSG). As shown in Fig. 3f, the fluorescence response of ^•OH in Fe_SAC@NC system first increases and then decreases, while in Fe_SAC-PC@NC system, it rapidly drops to zero after applying the voltage. This indicates that ^•OH is continuously generated in Fe_SAC@NC system, but it is likely rapidly consumed or almost no longer produced in Fe_SAC-PC@NC system, which is consistent with the 3e^- ORR process observed for Fe_SAC@NC in Fig. 2d. In contrast, the fluorescence response of ¹O₂ shows the opposite trend (Fig. 3g). This suggests that ¹O₂ is constantly consumed in Fe_SAC@NC system but continuously produced in Fe_SAC-PC@NC system. This indicates that the synergistic effect between SACs and PCs enables the regulation of H₂O₂ activation pathway, making it more favorable for the selective generation of ¹O₂, which is consistent with the EPR test results. From above tests, it can be observed that ^•OH in Fe_SAC-PC@NC system rapidly disappears after the voltage is applied. However, the quenching experiments (Fig. 3a) shows that when ^•OH is quenched, the removal efficiency of SMX significantly decreases, suggesting that ^•OH may act as an important intermediate in the generation of ¹O₂. To verify this hypothesis, we quench H₂O₂, O₂^•-, and ^•OH in Fe_SAC-PC@NC system, then detect the concentration of ¹O₂ using EPR and SOSG. As shown in Fig. 3h and Fig. S9 (Supporting information), the suppression of the EPR signal of ¹O₂ is most pronounced in the presence of p-BQ and TBA, confirming that O₂^•- and ^•OH serve as key intermediates in the H₂O₂-mediated generation of ¹O₂ and play a crucial role in its formation.

  Figure 3

  

  Figure 3.  (a) Degradation and (b) first-order reaction kinetics of SMX removal in Fe_SAC-PC@NC system. EPR spectra of (c) DMPO-^•OH, (d) DMPO-O₂^•- and (e) TEMP-¹O₂ in Fe_SAC@NC and Fe_SAC-PC@NC system. Fluorescence response of (f) TA upon treatment and (g) SOSG upon treatment in Fe_SAC@NC and Fe_SAC-PC@NC system. (h) EPR spectra of TEMP-¹O₂ in Fe_SAC-PC@NC system with quencher. (i) The electrochemical in-situ Raman spectra of Fe_SAC-PC@NC in Na₂SO₄ solution with pH 3 saturated with O₂.

  DownLoad: Full-Size Img PowerPoint

  In the H₂O₂-mediated electrocatalytic process, *OOH is a key intermediate in the generation of ^•OH from H₂O₂ [40,41]. Therefore, to verify the main source of ^•OH in Fe_SAC-PC@NC system, we monitor the reaction process using in situ Raman spectroscopy. In Fig. 3i, the stretching vibration signals of the *OOH intermediate at 858 cm^-1 and the O-O stretching vibration peak of *OOH at 985 cm^-1 are clearly observed [42]. As the applied potential gradually decreases, the Raman signals of these characteristic peaks gradually increase, reaching the highest intensity at 0.4 V. This stage may correspond to the transformation process of *OOH, indicating that *OOH acts as a reactive intermediate in the formation of ^•OH in Fe_SAC-PC@NC system.

  Therefore, it is inferred that the electrocatalytic removal of SMX in Fe_SAC-PC@NC system occurs via 3e^- ORR pathway mediated by O₂ and H₂O₂. Specifically, O₂ first adsorbs onto surface active sites (*), and is then reduced to form the *O₂^•- intermediate (Eq. 2). In the presence of H₂O or other proton sources, O₂^•- is protonated to generate *OOH (Eq. 3) [43], which is subsequently reduced and protonated to yield H₂O₂ (Eq. 4). Subsequently, H₂O₂ may be transformed into ^•OH through the *OOH intermediate (Eq. 5) [44,45]. The electron-donating effect of Fe nanoparticles facilitates the redox cycle of Fe (Eq. 1), resulting in the abundant generation of ^•OH. The reaction between ^•OOH and ^•OH then leads to the formation of ¹O₂ (Eq. 6) [46,47]. Afterwards, ¹O₂, as a key ROS involved in SMX degradation, oxidizes SMX into intermediates and ultimately mineralizes into CO₂ and H₂O (Eq. 7).

  $ \begin{aligned} { }^*+\mathrm{O}_2+\mathrm{e}^{-} \rightarrow{ }^* \mathrm{O}_2{ }^{•-} \end{aligned} $

  (2)

  $ { }^* \mathrm{O}_2{ }^{\cdot-}+\mathrm{H}^{+} \rightarrow{ }^*+^{\cdot} \mathrm{OOH} $

  (3)

  $ { }^*+^{\cdot} \mathrm{OOH}+\mathrm{H}^{+}+e^{-} \rightarrow{ }^* \mathrm{H}_2 \mathrm{O}_2 $

  (4)

  $\begin{aligned} \mathrm{H}^{+}+\mathrm{H}_2 \mathrm{O}_2+\mathrm{Fe}^{2+} \rightarrow \mathrm{Fe}^{3+}+^{\cdot} \mathrm{OH}+\mathrm{H}_2 \mathrm{O} \end{aligned} $

  (5)

  $ ^{\cdot} \mathrm{OH}+^{\cdot} \mathrm{OOH} \rightarrow \mathrm{H}_2 \mathrm{O}+{ }^1 \mathrm{O}_2 $

  (6)

  $ { }^1 \mathrm{O}_2+\mathrm{SMX} \rightarrow \text { intermediates }+\mathrm{CO}_2+\mathrm{H}_2 \mathrm{O} $

  (7)

  DFT calculations elucidate the synergistic interaction between SACs and PCs in activation of H₂O₂ at the atomic level. Fig. 4a presents the molecular structure model of Fe SACs serving as H₂O₂ adsorption sites in Fe_SAC@NC system. Structural optimizations are performed using both Fe SACs and Fe PCs as active sites in Fe_SAC-PC@NC system (Fig. 4b), and the corresponding total energies are calculated. The results show that the total energy of Fe SACs (−980.40 eV) is lower than that of Fe PCs (−978.25 eV), indicating that Fe SACs possess a more stable structure. Therefore, Fe SACs are selected as the adsorption sites in both systems for further calculations.

  Figure 4

  

  Figure 4.  Adsorption of H₂O₂ on the Fe site of (a) Fe_SAC@NC and (b) Fe_SAC-PC@NC surface. (c) The adsorption energy of Fe_SAC@NC and Fe_SAC-PC@NC system and (d) the charge density difference. The yellow and blue isosurfaces depict charge accumulation and depletion in the space, respectively.

  DownLoad: Full-Size Img PowerPoint

  As shown in Fig. 4c, the adsorption energy of H₂O₂ on Fe_SAC-PC@NC (−2.47 eV) is significantly lower than that on Fe_SAC@NC (−1.15 eV), suggesting that the introduction of Fe nanoparticles strengthens the adsorption between Fe SACs and H₂O₂, thus promoting its activation. Differential charge density analysis (Fig. 4d) further demonstrates that the synergistic interaction between Fe PC and Fe SACs induces an asymmetric redistribution of electrons around the Fe SACs, resulting in an increased electron density at the Fe sites. This electron enrichment subsequently enhances the adsorption capacity of H₂O₂ at the active sites, which is in agreement with the XPS results. In summary, the synergy between Fe SACs and Fe PCs not only improves the adsorption capacity of H₂O₂, but also promotes its efficient activation, enabling large-scale generation of ^•OH radicals and facilitating the highly selective production of ¹O₂.

  Moreover, to assess the ecological risks associated with this system, the intermediate products generated during the degradation of SMX are identified using high-performance liquid chromatography coupled with mass spectrometry (HPLC-MS). Based on these findings (Fig. 5a and Table S2 in Supporting information), four possible degradation pathways are proposed (Text S10 in Supporting information), all consistent with the main oxidation route of SMX [48,49]. As the reaction progresses, SMX and its intermediates are ultimately mineralized into harmless products such as H₂O, CO₂, and inorganic salts. Subsequently, the ecological structure-activity relationships (ECOSAR) program is employed to predict both acute and chronic toxicity based on indicator species, including fish, daphnids, and green algae (Figs. 5b and c) [50-52]. The results show that in the EF system, most of the intermediates exhibit lower toxicity than SMX.

  Figure 5

  

  Figure 5.  (a) Possible degradation pathways of SMX. (b) Acute toxicity and (c) chronic toxicity of SMX and the degradation intermediates to three aquatic organisms by ECOSAR program.

  DownLoad: Full-Size Img PowerPoint

  In addition to electrocatalytic activity, the universality and cycling stability are also important indicators for evaluating catalyst performance. As shown in Fig. 6a, Fe_SAC-PC@NC achieves a removal efficiency of over 90% for SMX solutions with varying concentrations (30–90 mg/L). Furthermore, its removal efficiencies for sulfadimidine (SDM), sulfisoxazole (SIZ) and tetracycline hydrochloride (TC) all exceed 93% (Fig. 6b), demonstrating its excellent universality. In terms of cycle stability, the removal efficiency of SMX remains above 90% after four consecutive cycles (Fig. 6c). Moreover, as shown in Fig. S10 (Supporting information), no obvious structural degradation is observed in the Fe_SAC-PC@NC catalyst after the reaction, indicating that it possesses good reusability.

  Figure 6

  

  Figure 6.  (a) Effect of different catalyst dosage, SMX concentration and initial pH on removal efficiency of SMX. (b) Removal efficiencies of SDM, SIZ and TC. (c) The stability test after 5 cycles. (d) The electrocatalytic continuous reaction device and (e) the degradation of Methyl Orange. (f) Scaled-up experimental setup and (g) SMX removal efficiency. (h) Comparison of E_EO of several catalytic materials for pollutant removal.

  DownLoad: Full-Size Img PowerPoint

  To further explore the potential application value of Fe_SAC-PC@NC, we immobilized Fe_SAC-PC@NC on polyvinylidene difluoride (PVDF) membranes by vacuum filtration and built a continuous-flow experiment for wastewater treatment (Fig. 6d). This approach successfully solved the problem that powder catalysts are difficult to recycle in practical applications. As shown in Fig. 6e, Fe_SAC-PC@NC membrane reaction unit maintained above 92% removal of Methyl Orange after 640 min of continuous operation. This demonstrates that it can be used consistently over a long period of time. Meanwhile, we also evaluated the degradation effect of Fe_SAC-PC@NC system in scaled-up experiment (Fig. 6f). As shown in Fig. 6g, the scaled-up experiment achieves an SMX removal rate of 88% within 300 min, demonstrating the feasibility of Fe_SAC-PC@NC for large-scale applications. Moreover, operational parameters such as applied voltage and current intensity play a critical role in determing energy consumption and cost efficiency. The energy efficiency of oxidation (E_EO) for SMX degradation in this efficient filtration system is 3.7 kWh/m³, much lower than that of other reported catalytic oxidation technologies for wastewater treatment as illustrated in Fig. 6h and Table 3 (Supporting information), highlighting the practical application of Fe_SAC-PC@NC system.

  In summary, Fe SACs and Fe⁰/Fe₇C₃ PCs were successfully constructed on a lignin-derived nitrogen-doped biochar support and applied in an EF system for efficient degradation of organic pollutants. The synergistic effect between these two distinct metallic entities: Fe SACs and Fe⁰/Fe₇C₃ PCs, not only enhances electron transfer and promotes the Fe²⁺/Fe³⁺ redox cycle, but also collaboratively modulates the electronic structure of the active sites, leading to an increased electron density. This electron enrichment subsequently enhances the adsorption capacity of H₂O₂ at the active sites and thereby improving its activation efficiency. Radical scavenging experiments further demonstrate that the synergistic interaction between Fe SACs and Fe⁰/Fe₇C₃ PCs leads to an optimized H₂O₂ activation pathway, enabling highly selective generation of ¹O₂ in the EF system. Moreover, under optimal conditions (catalyst dosage: 0.17 g/L, pollution concentration: 50 mg/L, pH 3, voltage 4 V), the removal efficiency of various pollutions reach over 93% within 60 min. Toxicity evaluation and extended experiments indicated that this EF system exhibits low ecotoxicity and excellent continuous degradation performance, making it a promising candidate for continuous and scalable wastewater treatment with an energy efficiency of E_EO as low as 3.7 kWh/m³ and controllable operational costs. This study demonstrates that the synergistic strategy between Fe SACs and Fe⁰/Fe₇C₃ PCs can precisely regulate the electronic structure of active sites, providing a strategic direction and technical pathway for improving H₂O₂ activation efficiency in EF systems. The findings hold significant theoretical value and application potential.

  CRediT authorship contribution statement

  Chunhe Cao: Writing – original draft, Investigation, Formal analysis, Conceptualization. Wenwen Zheng: Writing – original draft, Methodology, Investigation. Suhang Meng: Methodology, Data curation. Qiuyu Rong: Methodology. Yi Li: Writing – review & editing, Project administration, Methodology, Funding acquisition, Data curation, Conceptualization.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by the Natural Science Foundation of China as general projects (Nos. 52570088 and 22376157), the National Key Research and Development Program of China (No. 2024YFC3908501) and Tianjin Commission of Science and Technology as key technologies R&D projects (Nos. 23YFZCSN00130 and 25ZYCGCG00370).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2026.112570.

  
  
• 1. [1]

     M.R. Wang, B.L. Bodirsky, R. Rijneveld, et al., Nat. Commun. 15 (2024) 880. doi: 10.1038/s41467-024-44947-3

  2. [2]

     Z.Y. Zhao, Y.N. Mi, S.L. Wang, X.D. Du, Q.R. Zhang, Water Res. 280 (2025) 123510. doi: 10.1016/j.watres.2025.123510

  3. [3]

     M. Hassan, H. Olvera-Vargas, X.P. Zhu, B. Zhang, Y.L. He, J. Power Sources 424 (2019) 220–244. doi: 10.1016/j.jpowsour.2019.03.112

  4. [4]

     S.L. Wang, M.R. Razanajatovo, X.D. Du, et al., Chin. Chem. Lett. 35 (2024) 109140. doi: 10.1016/j.cclet.2023.109140

  5. [5]

     S. Wang, S.M. Guo, X.D. Du, et al., Chem. Eng. J. 495 (2024) 153604. doi: 10.1016/j.cej.2024.153604

  6. [6]

     J.Y. Yu, Y. Cheng, A.H. Cai, et al., Chin. Chem. Lett. 36 (2025) 110549. doi: 10.1016/j.cclet.2024.110549

  7. [7]

     Y. Jiang, J.B. Ran, K. Mao, et al., Ecotoxicol. Environ. Saf. 236 (2022) 113464. doi: 10.1016/j.ecoenv.2022.113464

  8. [8]

     H.Y. Sheng, A.N. Janes, R.D. Ross, et al., Energy Environ. Sci. 13 (2020) 4189–4203. doi: 10.1039/D0EE01925A

  9. [9]

     F. Miao, M.M. Gao, X. Yu, et al., Electrochem. Commun. 113 (2020) 106687. doi: 10.1016/j.elecom.2020.106687

  10. [10]

      X.L. Zhang, C.Y. Wu, Z.H. Wang, et al., Environ. Res. 263 (2024) 120020. doi: 10.1016/j.envres.2024.120020

  11. [11]

      Z.H. Chen, D. Yan, X.S. Wang, et al., ACS Catal. 15 (2025) 13568–13580. doi: 10.1021/acscatal.5c02889

  12. [12]

      Z.P. Lou, M.T. Liu, D.Y. Tang, et al., Appl. Catal. B: Environ. 315 (2022) 121523. doi: 10.1016/j.apcatb.2022.121523

  13. [13]

      H.F. Chan, Q. Du, C.L. Yue, et al., Appl. Catal. B: Environ. 382 (2026) 125953. doi: 10.1016/j.apcatb.2025.125953

  14. [14]

      M. Tian, Y.Z. Zhu, D.L. Zhang, et al., Nano Energ. 64 (2019) 103940. doi: 10.1016/j.nanoen.2019.103940

  15. [15]

      S. Kabir, K. Artyushkova, A. Serov, P. Atanassov, ACS Appl. Mater. Interfaces 10 (2018) 11623–11632. doi: 10.1021/acsami.7b18651

  16. [16]

      J.L. Li, W.W. Zheng, S.H. Meng, et al., Adv. Funct. Mater. 35 (2025) 2417538. doi: 10.1002/adfm.202417538

  17. [17]

      H.P. Rong, S.F. Ji, J.T. Zhang, D.S. Wang, Y.D. Li, Nat. Commun. 11 (2020) 5884. doi: 10.1038/s41467-020-19571-6

  18. [18]

      Y. Wang, X.B. Zheng, D.S. Wang, Nano Res. 15 (2022) 1730–1752. doi: 10.1007/s12274-021-3794-0

  19. [19]

      S.F. Ji, B. Jiang, H.G. Hao, et al., Nat. Catal. 4 (2021) 407–417. doi: 10.1038/s41929-021-00609-x

  20. [20]

      J.R. Yang, W.H. Li, D.S. Wang, Chem. Catal. 1 (2021) 24–26.

  21. [21]

      Z.C. Zhuang, Y. Li, Y.H. Li, et al., Energy Environ. Sci. 14 (2021) 1016–1028. doi: 10.1039/D0EE03701J

  22. [22]

      H.Y. Jing, Z.Y. Zhao, C.Y. Zhang, et al., Nano Res. 14 (2021) 4025–4032. doi: 10.1007/s12274-021-3331-1

  23. [23]

      D. Zhou, L.L. Zhang, X.Y. Liu, et al., Nano Res. 15 (2022) 519–527. doi: 10.1007/s12274-021-3511-z

  24. [24]

      L.C. Liu, D.M. Meira, R. Arenal, et al., ACS Catal. 9 (2019) 10626–10639. doi: 10.1021/acscatal.9b04214

  25. [25]

      C. Mu, Z.H. Liu, Q.F. Yao, Q. He, J.P. Xie, SmartMat 6 (2025) e1317. doi: 10.1002/smm2.1317

  26. [26]

      X. Guo, H. Zhang, Y.L. Wang, et al., Angew. Chem. Int. Ed. 64 (2025) e202511266. doi: 10.1002/anie.202511266

  27. [27]

      Z.R. Zhou, Y.N. Li, M.M. Li, Y. Li, S.H. Zhan, Chin. Chem. Lett. 31 (2020) 2698–2704. doi: 10.1016/j.cclet.2020.07.003

  28. [28]

      J.L. Li, Y.P. Liu, S.H. Meng, et al., ACS ES&T Water 4 (2024) 5868–5879.

  29. [29]

      H.Y. Gang, H.Y. Deng, L. Yan, et al., J. Colloid Interface Sci. 638 (2023) 252–262. doi: 10.1016/j.jcis.2023.01.093

  30. [30]

      Y.J. Ji, F.Y. Xu, K. Fang, et al., Environ. Technol. Innov. 38 (2025) 104079. doi: 10.1016/j.eti.2025.104079

  31. [31]

      S.H. Meng, J.L. Li, S.Y. Wang, et al., Adv. Mater. 37 (2025) 2420080. doi: 10.1002/adma.202420080

  32. [32]

      C. Xia, S. Back, S. Ringe, et al., Nat. Catal. 3 (2020) 125–134. doi: 10.1038/s41929-019-0402-8

  33. [33]

      Y.P. Liu, P.F. Wang, L.B. Xie, et al., Angew. Chem. Int. Ed. 63 (2024) e202319470. doi: 10.1002/anie.202319470

  34. [34]

      X.Y. Lu, D. Wang, K.H. Wu, X.L. Guo, W. Qi, J. Colloid Interface Sci. 573 (2020) 376–383. doi: 10.1016/j.jcis.2020.04.030

  35. [35]

      W. Wang, Y.X. Zheng, Y.C. Hu, Y.C. Liu, S.L. Chen, J. Phys. Chem. Lett. 13 (2022) 8914–8920. doi: 10.1021/acs.jpclett.2c02684

  36. [36]

      C.C.L. McCrory, S. Jung, J.C. Peters, T.F. Jaramillo, J. Am. Chem. Soc. 135 (2013) 16977–16987. doi: 10.1021/ja407115p

  37. [37]

      Y. Tong, J.C. Wu, P.Z. Chen, et al., J. Am. Chem. Soc. 140 (2018) 11165–11169. doi: 10.1021/jacs.8b06108

  38. [38]

      F. Ye, Y. Liu, Q.R. Lv, et al., Chin. Chem. Lett. 37 (2026) 111136. doi: 10.1016/j.cclet.2025.111136

  39. [39]

      X.M. Chen, Y.Y. Wu, Y. Tang, et al., Chin. Chem. Lett. 35 (2024) 109245. doi: 10.1016/j.cclet.2023.109245

  40. [40]

      X. Huang, M. Song, J. Zhang, et al., Nano Res. 15 (2022) 3927–3932. doi: 10.1007/s12274-021-4057-9

  41. [41]

      L.L. Wang, X. Lan, W.Y. Peng, Z.H. Wang, J. Hazard. Mater. 408 (2021) 124436. doi: 10.1016/j.jhazmat.2020.124436

  42. [42]

      P. Chen, E.I. Solomon, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 13105–13110. doi: 10.1073/pnas.0402114101

  43. [43]

      N. Mahne, B. Schafzahl, C. Leypold, et al., Nat. Energy 2 (2017) 17036. doi: 10.1038/nenergy.2017.36

  44. [44]

      F. Xiao, Z.N. Wang, J.Q. Fan, et al., Angew. Chem. Int. Ed. 60 (2021) 10375–10383. doi: 10.1002/anie.202101804

  45. [45]

      X.M. Xu, Y.M. Zhang, Y. Chen, et al., Proc. Natl. Acad. Sci. U. S. A. 119 (2022) e2205562119. doi: 10.1073/pnas.2205562119

  46. [46]

      Z.C. Yang, J.S. Qian, A.Q. Yu, B.C. Pan, Proc. Natl. Acad. Sci. U. S. A. 116 (2019) 6659–6664. doi: 10.1073/pnas.1819382116

  47. [47]

      A.U. Khan, M. Kasha, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 12365–12367. doi: 10.1073/pnas.91.26.12365

  48. [48]

      X.F. Feng, R.X. Long, C.C. Liu, Y. Lu, X.B. Liu, Chem. Eng. J. 471 (2023) 144447. doi: 10.1016/j.cej.2023.144447

  49. [49]

      J.L. Zheng, Q.T. Lin, Y.X. Liu, et al., Chem. Eng. J. 452 (2023) 139233. doi: 10.1016/j.cej.2022.139233

  50. [50]

      A.V. Mohod, M. Momotko, N.S. Shah, et al., Water Resour. Ind. 30 (2023) 100220. doi: 10.1016/j.wri.2023.100220

  51. [51]

      S. Li, X.R. Liu, Y.J. Zheng, et al., Chin. Chem. Lett. 35 (2024) 108971. doi: 10.1016/j.cclet.2023.108971

  52. [52]

      S.Y. Hou, H.L. Hu, Q. Fu, et al., Sep. Purif. Technol. 333 (2024) 125980. doi: 10.1016/j.seppur.2023.125980

• 

  Figure 1  (a) Schematic diagram of the preparation of Fe_SAC-PC@NC. (b) HR-TEM image, (c) HAADF-STEM image and (d) EDS images of Fe_SAC-PC@NC. (e) The XRD pattern of samples. High resolution XPS spectra of (f) Fe 2p and (g) N 1s.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) SMX removal in different systems. (b) Electrochemical impedance curves of samples. (c) LSV curves at 1600 rpm. (d) Electron transfer number (n) and (e) H₂O₂ selectivity of samples. (f) EPR spectrum of NC. (g) CV curves of Fe_SAC-PC@NC at different scan rates, and ECSA measurements of samples. (h) H₂O₂ concentrations of samples.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Degradation and (b) first-order reaction kinetics of SMX removal in Fe_SAC-PC@NC system. EPR spectra of (c) DMPO-^•OH, (d) DMPO-O₂^•- and (e) TEMP-¹O₂ in Fe_SAC@NC and Fe_SAC-PC@NC system. Fluorescence response of (f) TA upon treatment and (g) SOSG upon treatment in Fe_SAC@NC and Fe_SAC-PC@NC system. (h) EPR spectra of TEMP-¹O₂ in Fe_SAC-PC@NC system with quencher. (i) The electrochemical in-situ Raman spectra of Fe_SAC-PC@NC in Na₂SO₄ solution with pH 3 saturated with O₂.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Adsorption of H₂O₂ on the Fe site of (a) Fe_SAC@NC and (b) Fe_SAC-PC@NC surface. (c) The adsorption energy of Fe_SAC@NC and Fe_SAC-PC@NC system and (d) the charge density difference. The yellow and blue isosurfaces depict charge accumulation and depletion in the space, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) Possible degradation pathways of SMX. (b) Acute toxicity and (c) chronic toxicity of SMX and the degradation intermediates to three aquatic organisms by ECOSAR program.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  (a) Effect of different catalyst dosage, SMX concentration and initial pH on removal efficiency of SMX. (b) Removal efficiencies of SDM, SIZ and TC. (c) The stability test after 5 cycles. (d) The electrocatalytic continuous reaction device and (e) the degradation of Methyl Orange. (f) Scaled-up experimental setup and (g) SMX removal efficiency. (h) Comparison of E_EO of several catalytic materials for pollutant removal.

  下载: 全尺寸图片 幻灯片
•