English

• Advanced oxidation technologies have the advantages of fast kinetic rates and high degrading efficiency, making them potential methods for eliminating stubborn pollutants for environmental remediation [1]. This catalytic mechanism can generate diverse reactive oxygen species (ROS), predominantly hydroxyl (^•OH) and sulfate (SO₄^•−) radicals by efficient peroxymonosulfate (PMS) activation [2], while SO₄^•− exhibits superior persistence and enhanced contaminant selectivity compared to ^•OH [3]. In general, Fe(Ⅱ) as an eco-friendly catalyst mediated Fenton-like mechanism is an effective route for activating PMS, it has also been widely selected in research reports for activating PMS and initiation of Fenton-like catalytic cycles (Eqs. 1–3) [4,5]. Nevertheless, the slow Fe(Ⅲ) reduction kinetics lead to rapid iron accumulation, which can significantly restrict sustained ROS generation.

  Consequently, various approaches have been developed to optimize the iron species cycle, thereby enhancing Fenton-like PMS activation through the incorporation of external physical stimuli and co-catalytic components [6]. While external stimuli (e.g., electrolysis, photolysis, and thermal activation) can mitigate Fe(Ⅲ) accumulation, they often suffer from energy-intensive operations, elevated costs, and complex instrumentations [7]. Iron-chelating organic ligands (e.g., pyridine carboxylic acid [8]) and electron-donating compounds (e.g., L-cysteine [9] and human acid [10]) widely serve as energy-efficient promoters for Fe(Ⅱ) regeneration [11]. Nevertheless, the application of these organic promoters may lead to secondary environmental contamination owing to their limited recyclability and inherent toxicity, while the scavenging effect for ROS induced by homogeneity will cause low removal efficiency of target pollutants. In recent years, metal sulfides (e.g., WS₂ and MoS₂ [12]) and carbonaceous materials (e.g., nitrogen-doped carbon nanotube [13] and graphene [14]) have been increasingly used to facilitate Fenton-like chemistry. However, the risk of formation of surface passivation layer and low iron cycling efficiency strongly limit their applications. Accordingly, seeking for long-lasting and environmentally friendly electron sacrificers is paramount for developing sustainable PMS-based Fenton-like techniques.

  $ \mathrm{Fe}(\mathrm{II})+\mathrm{HSO}_5{ }^{-} \rightarrow \mathrm{Fe}(\mathrm{IV})+\mathrm{SO}_4^{2-}+\mathrm{H}^{+} $

  (1)

  $ \mathrm{Fe}(\mathrm{II})+\mathrm{HSO}_5{ }^{-} \rightarrow \mathrm{Fe}(\mathrm{III})+\mathrm{SO}_4{ }^{\cdot-}+\mathrm{OH} $

  (2)

  $ \mathrm{SO}_4^{\cdot-}+\mathrm{OH}^{-} \rightarrow \mathrm{SO}_4^{2-}+{}^\cdot \mathrm{OH} $

  (3)

  Molybdenum pentaboride (Mo₂B₅) has great potential in enhancing Fenton-like reactions [15]. Boron atom has a specific electron configuration (1s²2s²2p¹) [16], the outermost layer of boron does not reach a sTable 8-electron configuration and is thus in an electron-deficient state. This electron deficient property endows boron with Lewis acidity, allowing it to accept electron pairs and form coordination adsorption with negatively charged groups, such as the O—O or S-O of PMS molecule. In addition, the electronic configuration of molybdenum is [Kr]4d⁵5s¹, with abundant D-orbital electrons that can flexibly undergo oxidation state changes from +2 to +6. The molybdenum atom of Mo₂B₅ usually exists in the +3 and may directly provide electrons for initiating Fe(Ⅲ) reduction as an electron donor [17]. Meanwhile, the boron and molybdenum in lower oxidation states on Mo₂B₅ surfaces facilitate efficient electron transfer to ferric species (e.g., Fe³⁺ and FeOH²⁺) [18]. This electron donation process may accelerate ferrous ions regeneration, thereby enhancing the kinetics of the rate-determining step in Fenton-like catalytic cycles [19]. Therefore, Mo₂B₅ also has great potential in indirectly activating PMS to produce ROS, and meanwhile, low valence molybdenum and boron form a co-catalyst with dual reduction sites for promoting Fe(Ⅲ) reduction. In addition, Mo₂B₅, a cost-effective metal boride composed of abundant natural elements, is typically prepared by straightforward solid-phase reactions involving molybdenum metal and amorphous boron [20]. Thus, Mo₂B₅ is a promising co-catalyst with high conductivity and cost-effectiveness of dual reduction sites to facilitate direct and indirect activation of PMS [7,21,22].

  In this study, Mo₂B₅ was used as a co-catalyst to enhance PMS activation via dual reaction pathways for accelerating the production of ROS and degrading organic pollutants, with sulfamethoxazole (SIZ) serving as the model contaminant. The co-catalytic performance of Mo₂B₅ was compared with common solid co-catalysts (carbon materials (CNT, Graphite, WC, Biochar, rGO), metallic species in the zero valent state (W and Mo), metal sulfides (WS₂ and MoS₂), metal borides (ZrB₂, CrB₂, CaB₆, FeB₂, TiB₂, HfB₂, NbB₂, VB₂, MoB, Co₂B, WB), phosphorus (P), and dissolved reducing agents (ascorbic acid (AA), hydroquinone (HQ), L-cysteine (L-CE), sodium sulfite (SS)). In addition, the identification and quantification of ROS were systematically investigated through an integrated approach incorporating radical scavenging experiments, chemical probe methodologies, electron paramagnetic resonance (EPR) tests as well as substrate specific reactivity analysis. The co-catalytic mechanisms of Mo₂B₅ for assisting Fe(Ⅲ)/PMS were investigated and clarified comprehensively based on the analysis of iron transformation and characterizations (X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), high resolution transmission electron microscopy (HRTEM), Raman spectroscopy, and Fourier transform infrared spectroscopy (FT-IR)) of Mo₂B₅. The details of materials and methods were shown in Text S1-S8 (Supporting information).

  The co-catalytic effects of Mo₂B₅ on both Fe(Ⅲ)/PMS and Fe(Ⅱ)/PMS were systematically investigated. Fig. 1a depicts PMS system cannot directly degrade SIZ, but the Mo₂B₅/PMS system degraded 18% of SIZ at 15 min. It is shown that Mo₂B₅ consisting of the transition metal molybdenum and the semi-metallic element boron [7,23] can effectively and directly activate PMS to generate ROS for SIZ degradation. The unique activation ability of Mo₂B₅ for PMS stems from the precise matching of the dual active sites of Mo and B on its surface with the asymmetric molecular structure of PMS. Compared to the symmetrically structured peroxydisulfate and hydrogen peroxide, the asymmetric O—O bond of PMS is more susceptible abnormal fracture on the surface of Mo₂B₅ for direct electron transfer from Mo₂B₅ to PMS [24]. Moreover, Fe(Ⅲ) could not effectively activate PMS to eliminate SIZ (< 3%), whereas Fe(Ⅱ) can effectively activate PMS to degrade SIZ (22%) via Fe(Ⅱ) induced cleavage of O—O bond of PMS (Fig. 1b). The low degradation rate of SIZ by Fe(Ⅱ)/PMS indicates that SIZ was degraded in the first stage within 2 min. The rapid accumulation of Fe(Ⅲ) coupled with inefficient Fe(Ⅱ) regeneration significantly constrains sustained SIZ degradation during secondary reaction phases [25]. When Mo₂B₅ was used as a co-catalyst in Fenton-like reaction, it significantly contributed to the oxidation capability of Fe(Ⅲ)/PMS and Fe(Ⅱ)/PMS, while SIZ was fully degraded within 4 min in both systems. Thus, Mo₂B₅ as a co-catalyst can directly activate PMS to produce ROS and simultaneously promote iron redox cycling to accelerate Fenton-like oxidation. The superior stability of Fe(Ⅲ) relative to Fe(Ⅱ) makes it more appropriate for further mechanistic study in Mo₂B₅ co-catalyzed Fenton-like systems. The performance of Fe(Ⅲ)/Mo₂B₅/PMS is pH-dependent, which exhibited optimal efficiency for degrading SIZ at pH 3.0 (Fig. S1 in Supporting information).

  Figure 1

  

  Figure 1.  Co-catalytic activity of Mo₂B₅ for enhancing Fe(Ⅲ)/PMS system. Mo₂B₅ assisted (a) Fe(Ⅲ)/PMS and (b) Fe(Ⅱ)/PMS. (c) Reducing agents assisted Fe(Ⅲ)/PMS and (d) direct activation of PMS by carbon materials. Various metal brides assisted Fe(Ⅲ)/PMS for (e) ratios and (f) k_obs of SIZ removal.

  DownLoad: Full-Size Img PowerPoint

  Then, a wide spectrum of co-catalysts were applied for assisting Fenton-like activation of PMS by comparing with Mo₂B₅. The results in Fig. 1c and Fig. S2 (Supporting information) indicate that Mo₂B₅ (100%) outperformed than other reducing agents (W (18%), WS₂ (41%), Mo (81%), MoS₂ (57%), P (13%), HQ (42%), SS (36%), L-CE (93%), and AA (13%)) for promoting Fe(Ⅲ)/PMS to remove SIZ. Previous literature reported that carbon materials mediated activation of PMS can also significantly degrade organic contaminants [26], thus, we compared the performance of Mo₂B₅/Fe(Ⅲ)/PMS with various carbon materials assisted PMS systems [27,28]. Fig. 1d and Fig. S3 (Supporting information) show that rGO, CNT, Graphite, Biochar, and WC coupled PMS systems only degraded 22%, 12%, 57%, 16%, and 23% of SIZ at 15 min, respectively, which are markedly lower than that in Mo₂B₅/Fe(Ⅲ)/PMS. Moreover, the reactivities of these carbon materials as co-catalysts of Fe(Ⅲ)/PMS were also markedly lower than that of Mo₂B₅ (Fig. S4 in Supporting information). Although the performance of Mo₂B₅ in directly activating PMS is not prominent, the performance of Mo₂B₅/Fe(Ⅲ)/PMS is superior to carbon material assisted systems. In recent years, the effective enhancement of the catalytic activity of carbon materials for PMS activation were widely investigated by defect engineering and surface modifications (e.g., surface functional groups and heteroatom doping) [29]. However, carbon materials assisted PMS systems only exhibit high activity in degrading organic compounds with electron-rich groups, while Mo₂B₅ shows higher co-catalytic efficiency and wider adaptability for diverse organic contaminants [29–32]. In addition, various metal borides (ZrB₂, CrB₂, CaB₆, FeB₂, TiB₂, HfB₂, NbB₂, VB₂, MoB, Co₂B, WB) also demonstrated capability in enhancing Fe(Ⅲ)/PMS (Fig. 1e and Fig. S5 in Supporting information), and meanwhile, Mo₂B₅ exhibited superior co-catalytic performance compared with other metal borides. Benchmarking experiments demonstrated the 1.2–10.1 times higher k_obs for SIZ removal (0.97 min^-1) in Mo₂B₅/Fe(Ⅲ)/PMS compared to alternative metal boride co-catalysts assisted systems (Fig. 1f and Text S9 in Supporting information). The exceptional co-catalytic behavior of Mo₂B₅ is mechanistically attributed to its optimized electronic structure and abundant surface active sites, which may synergistically facilitate (1) direct PMS activation and (2) stabilized Fe(Ⅲ)/Fe(Ⅱ) redox cycling for indirect Fenton-like reaction (Fig. S6 in Supporting information).

  Previous reports indicated that PMS decomposition catalyzed by iron can generate multiple ROS [33,34]. As depicted in Fig. 2a and Fig. S7 (Supporting information), comparing with extremely low PMS decomposition in Fe(Ⅲ)/PMS, the PMS depletion in Mo₂B₅/PMS is 0.15 mmol/L at 15 min, which implies the capability of Mo₂B₅ to directly decompose PMS. The Mo₂B₅/Fe(Ⅲ)/PMS system consumed 0.45 mmol/L PMS at 15 min, indicating that Mo₂B₅ can significantly promote indirect decomposition of PMS for ROS formation via Fenton-like reaction. The generated ROS were thus in-situ characterized through EPR tests using DMPO as the chemical probe (Fig. 2b and Fig. S8 in Supporting information). EPR spectrum revealed no detectable DMPO-radical adduct in Fe(Ⅲ)/PMS, confirming the limited capacity of Fe(Ⅲ) for direct PMS activation. The detection of DMPO—OH adducts (α^N = α^H = 14.9 G, with 1:2:2:1 quartet peak intensity ratio) in Mo₂B₅/PMS provides conclusive evidence that Mo₂B₅ can directly activate PMS and promote O—O bond cleavage for producing ^•OH rather than SO₄^•− [35]. Moreover, the EPR spectra revealed significantly enhanced DMPO—OH adduct signal in Mo₂B₅/Fe(Ⅲ)/PMS compared to Mo₂B₅/PMS, accompanied by characteristic DMPO-SO₄ peaks (α^H = 0.75 G, 1.42 G, 10.1 G; α^N = 13.7 G) [36]. These observations unveil the capability of Mo₂B₅ to promote Fe(Ⅲ)/Fe(Ⅱ) redox cycling, thereby facilitating indirect generation of reactive radicals (SO₄^•− and ^•OH) in Fenton-like process. Moreover, quenching tests employing tert‑butyl alcohol (TBA) and methanol (MeOH) as radical quenchers for selectively scavenging ^•OH and SO₄^•− in Mo₂B₅/Fe(Ⅲ)/PMS were carried out, resulting from the distinct rate constants of MeOH and TBA for reacting with SO₄^•− and ^•OH [37,38]. Fig. 2c clearly illustrates the differential inhibition effects of MeOH and TBA, MeOH reduced the degradation efficiency by 56%, whereas TBA caused a comparatively lower inhibition of 22%. The significant suppression by TBA confirms the non-negligible contribution of ^•OH, whereas the higher inhibition ratio by MeOH suggests concurrent involvement of SO₄^•−. The result demonstrated that both ^•OH and SO₄^•− participate in SIZ oxidation. In addition, EPR analysis of Mo₂B₅/Fe(Ⅲ)/PMS also revealed the presence of singlet oxygen (¹O₂, Fig. S9 in Supporting information), while superoxide radical was not detected (Fig. S10 in Supporting information), indicating that Mo₂B₅ can mediate energy transfer to generate ¹O₂. Nevertheless, considering the indistinctive inhibition effect of sugar alcohol on SIZ degradation (Fig. S9) and the non-negligible scavenging effect of sugar alcohol on ^•OH and SO₄^•−, ¹O₂ induced SIZ removal in Mo₂B₅/Fe(Ⅲ)/PMS is not the dominant approach.

  Figure 2

  

  Figure 2.  Identification of various ROS in the Mo₂B₅/Fe(Ⅲ)/PMS system. (a) PMS consumption, (b) EPR spectra, (c) quenching tests, (d) contributions of different ROS for PMSO oxidation.

  DownLoad: Full-Size Img PowerPoint

  Moreover, current scientific discussion propose that Fe(Ⅱ) may facilitate heterolytic dissociation of peroxide bond in PMS to produce Fe(Ⅳ) under acidic condition [39]. The limited reaction kinetics among Fe(Ⅳ) and scavengers (k_(MeOH) = 571 L mol^-1 s^-1, k_(TBA) = 60 L mol^-1 s^-1) suggest the potential Fe(Ⅳ) formation in Mo₂B₅/Fe(Ⅲ)/PMS, as evidenced by significant SIZ degradation (32%) with high dosage of MeOH for scavenging SO₄^•− and ^•OH (Fig. 2c) [22,40]. The present study investigated the role of Fe(Ⅳ) in Mo₂B₅/Fe(Ⅲ)/PMS via methyl phenyl sulfoxide (PMSO)-based probe experiments, enabling both qualitative and semi-quantitative determination of Fe(Ⅳ) generation while differentiating between radical (SO₄^•− and ^•OH) and non-radical (Fe(Ⅳ)) oxidation pathways. This result can be attributed to the unique oxygen transfer mechanism of Fe(Ⅳ), which can selectively oxidize sulfoxides to their corresponding sulfones [41]. Fig. 2d reveals that Fe(Ⅳ) contributes obviously to PMSO oxidation in Mo₂B₅/Fe(Ⅲ)/PMS, exhibiting a PMSO₂ yield of 58% relative to PMSO consumption (105 μmol/L). These experimental findings substantiate that Fe(Ⅳ) can serve as a crucial ROS for degrading organic pollutants. Moreover, by adding TBA into Mo₂B₅/Fe(Ⅲ)/PMS for selectively terminating ^•OH induced oxidation of PMSO, it was found that ^•OH and SO₄^•− contributed to 19% and 23% of PMSO degradation, respectively. Thus Mo₂B₅/Fe(Ⅲ)/PMS can synergistically degrade organic pollutants through combined radical (SO₄^•− and ^•OH) and non-radical (Fe(Ⅳ)) pathways. Due to the generation of multiple ROS for degrading organic pollutants via synergetic radical and non-radical routes, the system coupling Mo₂B₅ and Fe(Ⅲ)/PMS immunes largely from various anions (Cl⁻, SO₄^2−, NO₃⁻ and HCO₃⁻, Fig. S11 in Supporting information).

  The combined results from EPR, radical screening, and PMSO chemical probe collectively unveil that Mo₂B₅/Fe(Ⅲ)/PMS can generate multiple radicals (SO₄^•− and ^•OH) and non-radical Fe(Ⅳ). Fig. 3 depicts the Mo₂B₅ mediated direct route as an activator and indirect route as a co-catalyst for activating PMS via Fenton-like reaction. The universality of Mo₂B₅/Fe(Ⅲ)/PMS for oxidizing multifold emerging contaminants (SIZ, diethyl phthalate (DEP), dimethyl phthalate (DMP), carbamazepine (CBZ), sulfamethoxazole (SMX), ibuprofen (IBP), benzoic acid (BA), bisphenol A (BPA), and phenol (PE)) with different molecular structures was further evaluated. SO₄^•− and ^•OH typically engage with organic compounds through analogous reaction mechanisms, such as H atom abstraction, electron abstraction as well as radical addition [42]. Compared to ^•OH (2.7 V), SO₄^•− not only has comparable oxidation potential (2.5–3.1 V) but also shows more significant selective oxidation properties for electron-rich contaminants [43]. ^•OH is an unselective ROS for almost all of these contaminants (10⁹–10¹⁰ L mol^-1 s^-1) [44], while SO₄^•− exhibits comparatively lower reactivity toward DMP and DEP (6.4 × 10⁷ L mol^-1 s^-1) with reaction rates 2–3 orders of magnitude below those of ^•OH [45]. Furthermore, Fe(Ⅳ) (2.0 V) exhibits selective reactivity towards electron-donating moieties in organic molecules, attributed to its intermediate oxidation potential and distinct reaction mechanisms, including oxygen atom transfer, hydride abstraction, and electrophilic addition [46].

  Figure 3

  

  Figure 3.  The routes of the Mo₂B₅/Fe(Ⅲ)/PMS system for producing ROS.

  DownLoad: Full-Size Img PowerPoint

  Fig. 4a and Fig. S12 (Supporting information) show that Mo₂B₅/Fe(Ⅲ)/PMS can completely degrade various organic pollutants within 15 min, while all three ROS are effective in degrading SMX, BA, CBZ, PE, IBP, SIZ and BPA based on results of quenching experiments. However, both TBA and MeOH significantly inhibited this system for degrading of DEP and DMP, suggesting that ^•OH is the main ROS for oxidizing these highly refractory pollutants. Despite the strong inhibitory effects of TBA and MeOH, this system still shows high degradation efficiencies for other pollutants, which is attributed to the Fe(Ⅳ)-mediated non-radical pathway. The synergistic formation of SO₄^•−, ^•OH, and Fe(Ⅳ) thus overcomes the limitations of a single ROS-dominated oxidation system. While SO₄^•− and ^•OH excel in degrading a wide range of pollutants, Fe(Ⅳ) provided a complementary pathway for pollutants with complex structures. The substrate-dependent reactivity of Mo₂B₅/Fe(Ⅲ)/PMS is further demonstrated by the significant variation in k_obs of different pollutants, and Fig. 4b shows a significant difference in k_obs for the oxidation of different organic pollutants (0.12–0.96 min^-1). Mo₂B₅/Fe(Ⅲ)/PMS has relatively low oxidation capacity for refractory pollutants, and the k_obs values of DEP (0.12 min^-1) and DMP (0.12 min^-1) were relatively lower than those of the other seven pollutants (0.53–0.96 min^-1). These results indicate that Mo₂B₅/Fe(Ⅲ)/PMS with the composite ROS can synergistically degrade various pollutants, which ensures the robust performance of the system across a wide range of pollutants.

  Figure 4

  

  Figure 4.  (a) Efficiencies and (b) k_obs for degrading multiple pollutants in the Mo₂B₅/Fe(Ⅲ)/PMS system.

  DownLoad: Full-Size Img PowerPoint

  The recovery of Fe(Ⅱ) directly determines the efficiencies of Fenton-like systems, the conversion of iron species was thus investigated by monitoring the dynamics of various iron species (Fig. 5a). In Mo₂B₅/Fe(Ⅲ)/PMS, the content of total dissolved iron was consistently maintained at 20 μmol/L (same as the initial Fe(Ⅲ) dosage), indicating that no obvious iron ions precipitated on the surface of Mo₂B₅. Comparative experiments revealed rapid Fe(Ⅱ) oxidation in Fe(Ⅱ)/PMS, with 80% conversion to Fe(Ⅲ) occurring by 1 min and complete oxidation of Fe(Ⅱ) achieved at 3 min. Meanwhile, Fe(Ⅲ)/Mo₂B₅ exhibited markedly faster reduction kinetics, achieving complete Fe(Ⅲ) reduction to form Fe(Ⅱ) at 1 min. This stable Fe(Ⅱ) generation process suggests that Mo₂B₅ can reduce Fe(Ⅲ) on its surface by directly donating electrons. The regenerated Fe(Ⅱ) was then released into the aqueous phase for initiating Fenton-like processes to generate ROS. This confirms that Mo₂B₅ can significantly accelerate the iron cycle via electron donation. Notably, Fe(Ⅱ) was not identified in Mo₂B₅/Fe(Ⅲ)/PMS system, and it is hypothesized that newly generated Fe(Ⅱ) may be rapidly consumed by PMS. Moreover, solid-state EPR analysis revealed the presence of paramagnetic oxygen vacancy (OV) centers in the original Mo₂B₅ (g = 2.004, Fig. 5b). OVs can act as electron-rich regions to enhance the electron donation ability on the Mo₂B₅ surface. These results further confirmed that Mo₂B₅ as an excellent electron donor can provide electrons to directly activate PMS and indirectly accelerate iron-mediated Fenton-like reaction for producing ROS.

  Figure 5

  

  Figure 5.  Mechanism analysis. (a) Transformation of iron species, (b) EPR spectrum of original Mo₂B₅, (c) full XPS spectra and XPS spectrum of (d) Fe 2p, (e) Mo 3d, (f) B 1s of Mo₂B₅ before and after experiments.

  DownLoad: Full-Size Img PowerPoint

  To elucidate the conformational relationship of electron transfer routes of Mo₂B₅ in Mo₂B₅/Fe(Ⅲ)/PMS, the interfacial evolution of pristine and reacted Mo₂B₅ was systematically characterized using various surface analytical techniques. Full spectral analysis by XPS (Fig. 5c) showed that there was no dramatical change in the elemental species of the samples after reaction, with a relative enhancement of the intensity of the B 1s peaks in the post-reaction samples compared to the pre-reaction ones. In contrast, the intensity of the Mo 3d peak was reduced, which was closely related to the surface selective oxidation process. Meanwhile, Raman spectroscopic analysis revealed progressive attenuation of characteristic O-Mo-O vibrational modes at 191.2 cm^-1 (antisymmetric stretching) and 332.3 cm^-1 (symmetric stretching) with the increase in the number of reaction cycles (Fig. S13 in Supporting information) [47]. This indicates that the low-valent molybdenum in Mo₂B₅ was oxidized by PMS (for direct PMS activation) or Fe(Ⅲ) (for indirect Fenton-like reaction), and the weakening of the Mo 3d peaks in the XPS spectra may result from the generation of a thick oxide layer on the Mo₂B₅ surface or the presence of multiple oxidation states (Mo(Ⅴ) and Mo(Ⅵ)) [47]. Furthermore, Fe 2p XPS spectra in Fig. 5d shows that neither Fe(Ⅱ) nor Fe(Ⅲ) was detected on Mo₂B₅ surface both before and after experiments. This data confirms that after Fe(Ⅲ) reduction on the Mo₂B₅ surface, the generated Fe(Ⅱ) rapidly desorbs into the solution phase through the solid-liquid interfacial mass-transfer process, which in turn triggers homogeneous activation of PMS via Fenton-like reaction. This surface-body-phase synergistic mechanism avoids active site blockage and provides key experimental evidence for constructing a quasi-homogeneous Fenton-like system in dynamic equilibrium.

  Quantitative analysis of Mo 3d spectra (Fig. 5e) showed that the low-valent Mo (Mo(Ⅱ)) decreased from the initial 16.5% to 0% after one cycle, whereas Mo(Ⅳ) and Mo(Ⅴ) increased from 39.8% and 43.5% to 53.6% and 46.4%, respectively; and Mo(Ⅳ) decreased to 41.5% and Mo(Ⅴ) increased to 58.5% after five cycles. These confirm the role of Mo species as the main electron donor in Mo₂B₅/Fe(Ⅲ)/PMS. The proportion of Mo in the low valence state is always maintained during the reaction cycle, which keeps its electron-donating ability and ensures the stability of its recycling. Moreover, the analysis of the chemical state evolution of elemental boron (Fig. 5f) revealed a significant structural reconfiguration. The pristine Mo₂B₅ particles readily react with oxygen to form a surface layer of low oxidized boron and boron oxide on the Mo₂B₅ particles. This surface oxidation process explains the relatively high interfacial suboxide boron and B₂O₃ detected at the interface. The share of monomeric boron (B⁰, 187–188 eV) decreased from an initial 39.8% to 32.9% after one reaction cycle, whereas the B₂O₃ fraction (190–193 eV) correspondingly increased from 30.6% to 41.9%. This phenomenon indicates that Mo₂B₅ transfers electrons to Fe(Ⅲ) as the reaction proceeds, and the chemical valence of boron gradually increased and was accompanied by the production of intermediate oxides and B₂O₃. In addition, FT-IR spectroscopy showed that Mo₂B₅ underwent surface oxidation and hydrolysis to produce B-OH, with a slight enhancement of both B-O and Mo-O peaks during the reaction in Mo₂B₅/Fe(Ⅲ)/PMS (Fig. S14 in Supporting information).

  Cyclic stability tests (Fig. 6a) revealed that Mo₂B₅ exhibited excellent and high stable co-catalysis performance, the Mo₂B₅/Fe(Ⅲ)/PMS system consistently maintained efficient degradation kinetics for SIZ. This system could still achieve the complete degradation of SIZ within 8 min in cycle 5, which fully demonstrated the excellent stability and reusability of the co-catalytic system for the deep treatment of organic pollutants. In addition, the values of k_obs exhibited an atypical enhancement over 5 consecutive cycles (Fig. 6b), contrasting sharply with conventional Fenton-like systems that typically suffer from rapid activity decay. Characterization of XRD and HRTEM explained this phenomenon and revealed the surface reconstruction process caused by cycling tests (Figs. 6c and d). As shown in Fig. 6c, after several rounds of oxidation reaction in Mo₂B₅/Fe(Ⅲ)/PMS, the characteristic diffraction peaks of Mo₂B₅ did not show any obvious displacement or intensity attenuation, and the characteristic peaks of heterogeneous phases were not detected. This structural characterization result confirms the remarkable structural stability of Mo₂B₅ in Mo₂B₅/Fe(Ⅲ)/PMS from the crystallographic level. This excellent lattice integrity provides an important guarantee for the long-lasting operation of Mo₂B₅ and also explains the microscopic mechanism by which the system can maintain stable oxidation capability. Moreover, HRTEM images in Fig. 6d depicts that the pristine Mo₂B₅ exhibits typical crystal structure features, and the amorphous structure appears in local regions on Mo₂B₅ surface after 5 cycles. Although the formation of the surface amorphous layer was observed, there was no significant increase in the half-peak widths of the characteristic peaks in the XRD spectra (Fig. 6c), indicating that the amorphous layer was limited to the surface and the bulk structure remained intact. In addition, cyclic experiments confirmed that the amorphous layer has no significant negative impact on long-term catalytic activity. This is due to the fact that the B and Mo atoms on the Mo₂B₅ surface will be converted to ions resulting from the availability of electrons during direct and indirect routes for PMS activation, and meanwhile, the Mo₂B₅ surface can be renewed. This dynamic equilibrium phenomenon of the surface interface structure reveals a unique self-renewal mechanism of Mo₂B₅. ICP-MS analysis (Fig. S15 in Supporting information) shows the relatively low leaching concentrations of Mo (1.4 and 1.8 mg/L) and boron (1.2 and 1.4 mg/L) after 1 cycle test and 5 cycle tests, respectively, and the atom ratio of Mo to B was around 1.25 (Mo/B). The controlled decomposition of the reductive sites on the Mo₂B₅ surface not only effectively avoids the accumulation of catalytically toxic substances, but also achieves in-situ regeneration of the active sites by the exposure of fresh internal crystalline surfaces. Therefore, oxidation etching preferentially removed the passivation layer on the Mo₂B₅ surface for exposing more inside active sites, resulting in high co-catalytic performance in the cyclic tests.

  Figure 6

  

  Figure 6.  Self-cleaning effect of Mo₂B₅ surface in the Mo₂B₅/Fe(Ⅲ)/PMS system during cycling tests. (a) Curves and (b) k_obs of SIZ removal, (c) XRD spectra and (d) HRTEM images of Mo₂B₅.

  DownLoad: Full-Size Img PowerPoint

  In conclusion, we found that Mo₂B₅ as an efficient and multifunctional co-catalyst exhibited excellent enhancement in assisting Fenton-like system (Mo₂B₅/Fe(Ⅲ)/PMS) by virtue of its unique dual-active-site configuration and metal-like electron-conducting property. Mo₂B₅ can enhance the activation of PMS through dual pathways: (1) Direct activating PMS to generate ^•OH; (2) indirect activating PMS to produce SO₄^•− and Fe(Ⅳ) via Fenton-like reaction by low valent species of boron and molybdenum induced Fe(Ⅲ) reduction. Due to the production of these ROS, the Mo₂B₅/Fe(Ⅲ)/PMS system exhibits broad-spectrum oxidative capacity toward diverse organic contaminants via synergetic radical and non-radical routes. Moreover, the unique self-renewal mechanism of Mo₂B₅ ensures its long-lasting stability for co-catalyzing Fenton-like reaction during long-term operation. Therefore, this work provides a sustainable paradigm for water remediation by integrating dual co-catalytic pathways and developing a high-performance Fenton-like technology.

  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.

  CRediT authorship contribution statement

  Xinyun Zhang: Writing – review & editing, Writing – original draft, Investigation, Formal analysis, Data curation. Chenying Zhou: Investigation, Formal analysis, Data curation. Jian Zhang: Project administration, Formal analysis. Minglu Sun: Writing – review & editing, Supervision, Investigation, Formal analysis, Data curation. Yanbiao Shi: Supervision, Funding acquisition. Chuanshu He: Supervision, Formal analysis. Xiaowei Huo: Supervision, Formal analysis. Yang Liu: Supervision. Peng Zhou: Writing – review & editing, Writing – original draft, Supervision, Investigation, Formal analysis. Bo Lai: Writing – review & editing, Supervision, Project administration, Funding acquisition.

  Acknowledgments

  This research was supported financially by National Natural Science Foundation of China (Nos. U24A20561 and 22306119), Sichuan Science and Technology Program (No. 2024NSFTD0014) and Key R & D Program of Heilongjiang Province (No. 2023ZX02C01).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2025.111630.

  
  
• 1. [1]

     A. Shokri, M.S. Fard, Environ. Chall. 7 (2022) 100534. doi: 10.1016/j.envc.2022.100534

  2. [2]

     X. Li, D. Wu, T. Hua, et al., Sci. Total Environ. 804 (2022) 150096. doi: 10.1016/j.scitotenv.2021.150096

  3. [3]

     X. Zhou, X. Li, Y. Xiang, et al., Chin. Chem. Lett. 36 (2025) 110664. doi: 10.1016/j.cclet.2024.110664

  4. [4]

     Z. Li, L. Wang, Y. Liu, et al., Water Res. 168 (2020) 115093. doi: 10.1016/j.watres.2019.115093

  5. [5]

     S. Xiao, M. Cheng, H. Zhong, et al., Chem. Eng. J. 384 (2020) 123265. doi: 10.1016/j.cej.2019.123265

  6. [6]

     J. Guo, B. Gao, Q. Li, et al., Adv. Mater. 36 (2024) 2403965. doi: 10.1002/adma.202403965

  7. [7]

     S. Meng, M. Sun, P. Zhang, et al., Environ. Sci. Technol. 57 (2023) 12534–12545. doi: 10.1021/acs.est.3c03212

  8. [8]

     Z. Yang, C. Shan, B. Pan, et al., Environ. Sci. Technol. 55 (2021) 8299–8308. doi: 10.1021/acs.est.1c00230

  9. [9]

     T. Li, Z. Zhao, Q. Wang, et al., Water Res. 105 (2016) 479–486. doi: 10.1016/j.watres.2016.09.019

  10. [10]

      X. Li, Y. Liu, J. Wei, et al., Chin. Chem. Lett. 36 (2025) 110811. doi: 10.1016/j.cclet.2024.110811

  11. [11]

      P. Zhou, W. Ren, G. Nie, et al., Angew. Chem. Int. Ed. 132 (2020) 16517–16526. doi: 10.1002/anie.202007046

  12. [12]

      M. Huang, X. Wang, C. Liu, et al., J. Hazard. Mater. 404 (2021) 124175. doi: 10.1016/j.jhazmat.2020.124175

  13. [13]

      J. Guo, Y. Wang, Y.A. Shang, et al., Proc. Natl. Acad. Sci. U. S. A. 121 (2024) e2313387121. doi: 10.1073/pnas.2313387121

  14. [14]

      Y. Chen, R. Su, F. Xu, et al., J. Hazard. Mater. 470 (2024) 134162. doi: 10.1016/j.jhazmat.2024.134162

  15. [15]

      X. Liu, Y. Gong, Inorg. Chem. 60 (2021) 18075–18081. doi: 10.1021/acs.inorgchem.1c02684

  16. [16]

      H. Schaefer, F. Harris, Phys. Rev. 167 (1968) 67–73. doi: 10.1103/PhysRev.167.67

  17. [17]

      B. Albert, H. Hillebrecht, Angew. Chem. Int. Ed. 48 (2009) 8640–8668. doi: 10.1002/anie.200903246

  18. [18]

      P. Zhou, S. Meng, M. Sun, et al., Sep. Purif. Technol. 317 (2023) 123860. doi: 10.1016/j.seppur.2023.123860

  19. [19]

      M. Xing, W. Xu, C. Dong, et al., Chem 4 (2018) 1359–1372. doi: 10.1016/j.chempr.2018.03.002

  20. [20]

      X. Liu, Q. Jia, S. Zhang, et al., Int. Mater. Rev. 69 (2024) 107–141. doi: 10.1177/09506608231219864

  21. [21]

      X. Duan, W. Li, Z. Ao, et al., J. Mater. Chem. A 7 (2019) 23904–23913. doi: 10.1039/c9ta04885e

  22. [22]

      Z. Wang, W. Qiu, S. Pang, et al., Chem. Eng. J. 371 (2019) 842–847. doi: 10.1016/j.cej.2019.04.101

  23. [23]

      K. Yu, L. Li, C. Zang, et al., Colloid. Surf. A 578 (2019) 123607. doi: 10.1016/j.colsurfa.2019.123607

  24. [24]

      P. Shao, X. Duan, J. Xu, et al., J. Hazard. Mater. 322 (2017) 532–539. doi: 10.1016/j.jhazmat.2016.10.020

  25. [25]

      J. Zou, J. Ma, L. Chen, et al., Environ. Sci. Technol. 47 (2013) 11685–11691. doi: 10.1021/es4019145

  26. [26]

      Y. Peng, G. Xie, P. Shao, et al., Appl. Catal. B: Environ. 310 (2022) 121345. doi: 10.1016/j.apcatb.2022.121345

  27. [27]

      J. Espinosa, S. Navalón, M. Álvaro, et al., ChemCatChem 8 (2016) 2642–2648. doi: 10.1002/cctc.201600364

  28. [28]

      H. Cheng, H. Liu, C. Huang, et al., Sep. Purif. Technol. 330 (2024) 125311. doi: 10.1016/j.seppur.2023.125311

  29. [29]

      X. Zhang, J. Wang, Y. Wang, et al., J. Hazard. Mater. 442 (2023) 130115. doi: 10.1016/j.jhazmat.2022.130115

  30. [30]

      J. Lin, W. Tian, Z. Guan, et al., Adv. Funct. Mater. 32 (2022) 2201743. doi: 10.1002/adfm.202201743

  31. [31]

      M. Li, K. Zheng, Y. Jin, et al., J. Mater. Sci. Technol. 137 (2023) 67–78. doi: 10.1016/j.jmst.2022.08.003

  32. [32]

      R. Hirani, A. Asif, N. Rafique, et al., Sep. Purif. Technol. 319 (2023) 124110. doi: 10.1016/j.seppur.2023.124110

  33. [33]

      F. Ghanbari, M. Moradi, Chem. Eng. J. 310 (2017) 41–62. doi: 10.1016/j.cej.2016.10.064

  34. [34]

      C. Cheng, W. Ren, F. Miao, et al., Angew. Chem. Int. Ed. 62 (2023) e202218510. doi: 10.1002/anie.202218510

  35. [35]

      E. Braxton, D. Fox, B. Breeze, et al., ACS Measur. Sci. Au 3 (2023) 21–31. doi: 10.1021/acsmeasuresciau.2c00049

  36. [36]

      M. Cashman, L. Kirschenbaum, J. Holowachuk, et al., J. Hazard. Mater. 380 (2019) 120875. doi: 10.1016/j.jhazmat.2019.120875

  37. [37]

      G. Buxton, C. Greenstock, W. Helman, et al., J. Phys. Chem. Ref. Data 17 (1988) 513–886. doi: 10.1063/1.555805

  38. [38]

      P. Neta, R. Huie, A. Ross, J. Phys. Chem. Ref. Data 17 (1988) 1027–1284. doi: 10.1063/1.555808

  39. [39]

      Z. Wang, W. Qiu, S. Pang, et al., Environ. Sci. Technol. 56 (2022) 1492–1509. doi: 10.1021/acs.est.1c04530

  40. [40]

      Z. Wang, J. Jiang, S. Pang, et al., Environ. Sci. Technol. 52 (2018) 11276–11284. doi: 10.1021/acs.est.8b02266

  41. [41]

      H. Dong, Y. Li, S. Wang, et al., Environ. Sci. Technol. Lett. 7 (2020) 219–224. doi: 10.1021/acs.estlett.0c00025

  42. [42]

      U. Hübner, S. Spahr, H. Lutze, et al., Heliyon 10 (2024) e30402. doi: 10.1016/j.heliyon.2024.e30402

  43. [43]

      M. Zhang, J. Ruan, X. Wang, et al., Water Res. 246 (2023) 120697. doi: 10.1016/j.watres.2023.120697

  44. [44]

      P. Zhou, Y. Yang, W. Ren, et al., Appl. Catal. B: Environ. 319 (2022) 121916. doi: 10.1016/j.apcatb.2022.121916

  45. [45]

      R. Xiao, T. Ye, Z. Wei, et al., Environ. Sci. Technol. 49 (2015) 13394–13402. doi: 10.1021/acs.est.5b03078

  46. [46]

      C. Zhang, W. Zuo, L. Ai, et al., J. Colloid. Interf. Sci. 669 (2024) 794–803. doi: 10.1016/j.jcis.2024.05.020

  47. [47]

      J. Baltrusaitis, B. Mendoza-Sanchez, V. Fernandez, et al., Appl. Surf. Sci. 326 (2015) 151–161. doi: 10.1016/j.apsusc.2014.11.077

• 

  Figure 1  Co-catalytic activity of Mo₂B₅ for enhancing Fe(Ⅲ)/PMS system. Mo₂B₅ assisted (a) Fe(Ⅲ)/PMS and (b) Fe(Ⅱ)/PMS. (c) Reducing agents assisted Fe(Ⅲ)/PMS and (d) direct activation of PMS by carbon materials. Various metal brides assisted Fe(Ⅲ)/PMS for (e) ratios and (f) k_obs of SIZ removal.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Identification of various ROS in the Mo₂B₅/Fe(Ⅲ)/PMS system. (a) PMS consumption, (b) EPR spectra, (c) quenching tests, (d) contributions of different ROS for PMSO oxidation.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  The routes of the Mo₂B₅/Fe(Ⅲ)/PMS system for producing ROS.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Efficiencies and (b) k_obs for degrading multiple pollutants in the Mo₂B₅/Fe(Ⅲ)/PMS system.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Mechanism analysis. (a) Transformation of iron species, (b) EPR spectrum of original Mo₂B₅, (c) full XPS spectra and XPS spectrum of (d) Fe 2p, (e) Mo 3d, (f) B 1s of Mo₂B₅ before and after experiments.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Self-cleaning effect of Mo₂B₅ surface in the Mo₂B₅/Fe(Ⅲ)/PMS system during cycling tests. (a) Curves and (b) k_obs of SIZ removal, (c) XRD spectra and (d) HRTEM images of Mo₂B₅.

  下载: 全尺寸图片 幻灯片
•