Magnetic Co-doped 1D/2D structured γ-Fe₂O₃/MoS₂ effectively activated peroxymonosulfate for efficient abatement of bisphenol A via both radical and non-radical pathways

Abstract

● Magnetic Co- γ -Fe₂O₃/MoS₂ were prepared via facile hydrothermal methods.

● Doping γ -Fe₂O₃ with cobalt greatly increased PMS activation for BPA abatement.

● The compounding of MoS₂ significantly enhanced the stability of the catalyst.

● Hybrid radical-nonradical pathways acted for effective degradation of BPA.

● The toxicity of intermediates was lower than BPA via T.E.S.T analysis.

Iron-based catalysts have been widely used to treat refractory organic pollutants in wastewater. In this paper, magnetic Co-γ-Fe₂O₃ was synthesized by a facile tartaric acid-assisted hydrothermal method, and Co-γ-Fe₂O₃/MoS₂ nanocomposite catalyst was obtained via in situ growth of MoS₂ nanosheets on Co-γ-Fe₂O₃ nanoparticles. The nanocomposite catalysts were used to decompose bisphenol A (BPA) by activating peroxymonosulfate (PMS). It was shown that only 0.15 g/L catalyst and 0.5 mmol/L PMS degraded 10 mg/L of BPA (99.3% within 10 min) in the pH range of 3–9. PMS was activated due to redox cycling among the pairs Co(III)/Co(II), Fe(III)/Fe(II), and Mo(VI)/Mo(IV). Quenching experiments and electron paramagnetic resonance spectroscopy demonstrated that both radical and non-radical pathways were involved in BPA degradation, in which active radical sulfate radical and non-radical singlet oxygen were the main reactive oxygen species. Ten intermediates were identified by liquid chromatography-coupled mass spectrometry, and three possible BPA degradation pathways were proposed. The toxicity of several degradation intermediates was lower, and Co-γ-Fe₂O₃/MoS₂ exhibited excellent reusability and could be magnetically recovered.

Graphical abstract

Keywords

Magnetic Co- γ-Fe ₂O ₃/MoS ₂ / Hydrothermal method / Bisphenol A / Degradation pathways / Toxicity analysis

1 Introduction

In recent years, bisphenol A (BPA) has become one of the most broadly utilized and consumed chemicals for plasticizers, hardeners, and adhesives in plastic products (Wang et al., 2020). However, after entering aquatic environments and subsequent exposure to human beings and animals, BPA interacts with their endocrine receptors and impacts their reproduction, development, and neurological and immune systems (Gao et al., 2021; Liu et al., 2022). BPA has been frequently detected in aquatic ecosystems at concentrations of up to 21 μg/L (Sun et al., 2022), which has raised growing public concern. Hence, the removal of BPA has become an urgent problem.

Among the techniques for the removal of bio-recalcitrant contaminants, advanced oxidation processes based on sulfate radicals (SO₄^•−) generated by activating persulfates have attracted increasing research attention. Compared with hydroxyl radicals (•OH), SO₄^•− exhibits some advantages, including: 1) a higher oxidizing potential (E₀ = 2.5–3.1 V vs. 1.8–2.7 V for •OH) (Pan et al., 2018; Sun et al., 2022), 2) a longer half-life (30–40 μs vs. < 1 μs for •OH), and 3) applicability over a wider pH range. Regarding persulfates, peroxymonosulfate (PMS) is a better oxidant than peroxydisulfate (PDS) due to its asymmetric structure and longer superoxide O–O bond, which is more easily activated by the catalyst to produce reactive oxygen species (ROSs) (Duan et al., 2018). Notably, several methods have been used to activate PMS, mainly including heat-activated (Liu et al., 2023), ultraviolet-activated (Lu et al., 2019), ultrasound-activated (Xu et al., 2020), and transition metal-based heterogeneous activation methods (Lu et al., 2022). Among them, transition metal-based (e.g., Fe, Co, and Cu) heterogeneous activation allows catalysts to be developed that show magnetic recovery, excellent catalytic performance, and application over a broad pH range (Chen et al., 2019b; Long et al., 2023). Recently, iron-based catalysts, especially one-dimensional (1D) γ-Fe₂O₃, have a high natural abundance, eco-friendliness, stability, and ease of recovery (Niu et al., 2021; Wang et al., 2022b; Zhao et al., 2022). However, the application of iron-based catalysts is usually limited due to the low conversion of Fe³⁺ to Fe²⁺ (Mao et al., 2018). Ma et al. (2019) found that rhodamine B could barely be removed by γ-Fe₂O₃/PMS within 150 min. Compounding γ-Fe₂O₃ with other metal oxides or doping other transition metal ions on γ-Fe₂O₃ are efficient ways to enhance the catalytic performance of γ-Fe₂O₃. Notably, the integration of cobalt ions (Co²⁺) or cobalt oxide can generate one of the most effective activators for PMS. Zhu et al. (2019a) reported that atrazine degradation was more efficient in the Co-doped mesoporous FePO₄/PMS system than in the FePO₄/PMS system. However, these activation systems were susceptible to the leaching of toxic Co²⁺ after oxidation (Duan et al., 2018; Zhu et al., 2019a; Sun et al., 2020a). Additionally, earlier investigations demonstrated that the stability and recoverability of the Co-based catalysts require improvements, and the ROS generation mechanism during the reaction requires further investigation (Ma et al., 2019). Molybdenum sulfide (MoS₂), a transition metal dichalcogenide with a two-dimensional (2D) layered structure with high electron mobility, accelerated the production of ROSs (Xie et al., 2022; Yan et al., 2022). Moreover, it prevented the leaching of metal ions and thereby enhanced the stability and catalytic performance of the catalyst for recycling. He et al. (2022b) used a SrCoO₃/MnFe₂O₄/MoS₂ nanocomposite to remove levofloxacin (LVF) by activating PMS under visible light, and the removal efficiency of LVF remained at 85.2% after five cycles. It has been demonstrated that catalyst performance for the decomposition of different target pollutants is also connected to the activation mechanism of PMS. Recently, apart from radical pathway activation, several studies have demonstrated that PMS can also be activated via non-radical pathways, such as the generation of single oxygen (¹O₂) with a gentle oxidative capacity (2.2 V) and a high selectivity for attacking organic pollutants (Zhu et al., 2019b).

Considering the excellent properties of MoS₂, the integration of Co-γ-Fe₂O₃ and MoS₂ is expected to enhance the cyclic activity and stability of the material, and the novel magnetic nanomaterial Co-γ-Fe₂O₃/MoS₂ was fabricated via a two-step hydrothermal method in this work. The structure, morphology, surface composition, and magnetic properties of the compound materials were characterized by various techniques such as X-ray diffraction (XRD), transmission electron microscopy (TEM), scanning electron microscopy (SEM), and vibrating sample magnetometry (VSM). Co-γ-Fe₂O₃/MoS₂/PMS was employed to remove BPA in aqueous solution. The effects of reaction systems, initial pH, catalyst dosage, PMS concentration, coexisting anions, and natural organic matter on BPA degradation were investigated and discussed. As expected, Co-γ-Fe₂O₃/MoS₂ demonstrated excellent degradation of BPA and enhanced cycling stability. Furthermore, the main active species were investigated by quenching experiments and electron paramagnetic resonance (EPR) analysis. The potential decomposition pathways of BPA were proposed according to liquid chromatography-coupled mass spectrometry (LC-MS), and the toxicity of the substances formed during degradation was explored by a toxicity estimation software tool (T.E.S.T).

2 Materials and methods

2.1 Materials

Cobalt nitrate hexahydrate (Co(NO₃)₂·6H₂O, 99%), iron nitrate nonahydrate (Fe(NO₃)₃·9H₂O, 99.9%), ammonium molybdate tetrahydrate ((NH₄)₆Mo₇O₂₄·4H₂O, 99%), thiourea (CH₄N₂S, 99%), ethanol (C₂H₆O, 75%), p-benzoquinone (p-BQ, 97%), sodium nitrate (NaNO₃), sodium sulfate (Na₂SO₄), sodium chloride (NaCl), and sodium bicarbonate (NaHCO₃) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Polyethylene glycol 400 (PEG 400, H(OCH₂CH₂)_nOH), tartaric acid (C₄H₆O₆), bisphenol A (BPA, C₁₄H₁₈N₄O₃, 99%), methanol (MeOH, CH₄O), tert-butyl alcohol (TBA, C₄H₁₀O), L-histidine (C₆H₉N₃O₂), and acetonitrile (CH₃CN) were purchased from Aladdin Chemical Reagent Co., Ltd. (Shanghai, China). Hydrochloric acid (HCl) and sodium hydroxide (NaOH) were purchased from Tianjin Fuchen Chemical Reagent Co., Ltd. (Tianjin, China). Acetonitrile and methanol were HPLC grade, and the remaining chemicals were analytical grade purity. All chemicals were used without further purification.

2.2 Catalyst synthesis

The magnetic Co-γ-Fe₂O₃ and Co-γ-Fe₂O₃/MoS₂ composites were fabricated via hydrothermal methods. Co-γ-Fe₂O₃ was synthesized by using environment-friendly and readily available tartaric acid as an assisting agent. The detailed preparation processes are as follows.

2.2.1 Preparation of Co-γ-Fe₂O₃

The Co-γ-Fe₂O₃ complex was obtained by a simple hydrothermal method based on the synthesis method of Liu et al. (2019). The schematic diagram of the synthesis is shown in Fig. S1. Briefly, Fe(NO₃)₃·9H₂O was dissolved in a solution of PEG 400 in deionized water. Then, magnetic stirring was performed, while tartaric acid and Co(NO₃)₂·6H₂O were added to the solution. The optimal ratio of Co to Fe for BPA degradation was 3: 100 (Text S1 and Fig. S2). The mixture was moved into a Teflonlined autoclave and kept at 180 °C for 9 h. Finally, the black products were washed multiple times with deionized water and anhydrous ethanol and then dried to gain the final products. In this work, tartaric acid was chosen as an auxiliary synthesis material because of its low toxicity and low cost, as well as its ability to provide the carboxyl groups necessary for the reaction (Golestani et al., 2018).

2.2.2 Preparation of Co-γ-Fe₂O₃/MoS₂

A novel hydrothermal method was adopted for the in situ loading of MoS₂ nanosheets on the synthesized Co-γ-Fe₂O₃ nanomaterials. Specifically, (NH₄)₆Mo₇O₂₄·4H₂O and CH₄N₂S were dissolved in 30 mL of deionized water under ultrasound. Co-γ-Fe₂O₃ powder was mixed into the above solution, and the obtained black suspension was ultrasonicated intermittently. Then, the mixture was transferred into a 50-mL Teflonlined autoclave and heated at 180 °C for 10 h. After the reaction, the black precipitate was separated by centrifugation, washed several times with deionized water and anhydrous ethanol, and finally dried at 60 °C for 12 h. Pure MoS₂ was obtained via a similar method.

2.3 Catalyst characterization

XRD was performed using Cu Kα radiation to observe the crystal structure of the catalysts (Bruker AXS, Germany). X-ray photoelectron spectroscopy (XPS; ESCALAB Xi⁺, Thermo Fisher Scientific, USA) was conducted to identify the chemical components on the catalyst surface before and after the reaction. To study the surface morphology and obtain the element maps of the catalysts, field emission scanning electron microscopy (FE-SEM; GeminiSEM 300, ZEISS, Germany) and high-resolution transmission electron microscopy (HRTEM; FEI Tecnai F30, USA) were employed. The zeta potentials of the catalysts were determined with a nanoparticle potentiometer (Malvern, Nano-ZS90, UK). Magnetic properties were measured at room temperature by VSM (Quantum Design, PPMS-9, USA). The surface functional groups of the samples were identified and analyzed by Fourier-transform infrared (FTIR) spectroscopy.

2.4 Evaluation methods

Details of the used evaluation methods regarding the determination of BPA degradation efficiency, investigation of major ROSs, assessment of catalyst reusability, identification of generated intermediates, possible degradation pathways, and toxicity of intermediates are provided in Text S2.

3 Results and discussion

3.1 Characteristics of photocatalysts

3.1.1 Structural features

The XRD patterns of pure MoS₂, γ-Fe₂O₃, Co-γ-Fe₂O₃, and Co-γ-Fe₂O₃/MoS₂ are depicted in Fig.1(a). The diffraction peaks of Co-doped catalyst at 2θ = 18.44°, 29.74°, 35.44°, 42.94°, 53.50°, 57.08°, and 62.48° corresponded to the (111), (220), (311), (440), (422), (511), and (400) crystal planes of γ-Fe₂O₃, respectively, consistent with data in JCPDS No. 39-1346. No characteristic peaks of Co or Co-based compounds were observed, indicating the atomic dispersion of Co species in γ-Fe₂O₃ (Bai et al., 2020). Co-γ-Fe₂O₃/MoS₂ retained the characteristic peaks of γ-Fe₂O₃. However, the intensity of the corresponding peaks was reduced, indicating that loading MoS₂ did not affect the crystal shape and phase of γ-Fe₂O₃, but it did decrease the crystallinity of the oxide (He et al., 2022b). No obvious peaks of MoS₂ were observed, most likely due to its low crystallinity, low content, and high diffusion of MoS₂ on the Co-γ-Fe₂O₃ surface (Zheng et al., 2019; Zhou et al., 2020).

3.1.2 Morphological analysis

The microscopic features of the compounds were observed by SEM. Fig.1(b) displays the SEM image of Co-γ-Fe₂O₃, which consisted of fine nanoparticles with an average particle size of about 20.88 nm. The morphology of Co-γ-Fe₂O₃/MoS₂ showed that Co-γ-Fe₂O₃ nanoparticles were well combined with 2D MoS₂ nanosheets (Fig.1(c)), indicating that MoS₂ was successfully anchored to the surface of Co-γ-Fe₂O₃ (He et al., 2022a). A previous study showed that ultra-thin 2D nanosheets could expose sufficient active sites and reduce the charge diffusion length to accelerate charge carrier transfer, which facilitated the improvement of catalyst activity (Geng et al., 2021). Furthermore, the MoS₂ nuclei were grown in situ on the Co-γ-Fe₂O₃ nanoparticles in the hydrothermal reaction, and the MoS₂ nanosheets were also loaded with Co-γ-Fe₂O₃ nanoparticles. This generated strong interactions between them and promoted the formation of a favorable heterojunction at the interface, ensuring the stability of the nanometer-sized thin flake flower-like structure.

In addition, HRTEM and FTIR spectroscopy were performed to reveal more definite evidence for the nano-heterostructures of Co-γ-Fe₂O₃/MoS₂. As shown in Fig.1(d), the samples consisted of nearly spherical nanoparticles, which indicated the successful synthesis of γ-Fe₂O₃ (Li et al., 2015). Furthermore, it could be observed that some nanoparticles were attached to the nanosheets. The HRTEM images in Figs. S3(a)–S3(c) display the narrow interface between the two composites. The lattice spacings 0.251, 0.241, and 0.295 nm were attributed to the (311), (222), and (220) crystal planes of γ-Fe₂O₃ (Ge et al., 2021; Wu et al., 2022). The MoS₂ nanosheets exhibited lattice spacings of 0.614 and 0.227 nm, which were attributed to the (002) and (103) crystal planes, respectively (Zhang et al., 2018), compatible with the XRD results. Besides, as demonstrated in Fig. S4, the surface functional groups of γ-Fe₂O₃, Co-γ-Fe₂O₃, MoS₂, and Co-γ-Fe₂O₃/MoS₂ were studied by FTIR spectroscopy. The small peak at about 685 cm⁻¹ was related to the stretching of the Co–O bond in Co-γ-Fe₂O₃ (Sakthi Athithan et al., 2021). In addition, the peaks located at about 1089, 1629, and 2919 cm⁻¹ were assigned to C–O–C, C–O, and C–H bond vibrations of PEG, respectively (Abareshi and Salehi, 2022; Li et al., 2022b). The peak near 699 cm⁻¹ was assigned to the Mo–O bonds of MoS₂ (Ikram et al., 2020). The small peak at about 1126 cm⁻¹ was assigned to the S=O functional group, while the broad band at ~3441 cm⁻¹ was ascribed to the symmetric stretching vibration of O–H (Zhang et al., 2023). All of the above characteristic peaks were also present in the spectra of the composite catalyst Co-γ-Fe₂O₃/MoS₂, indicating the presence of both Co-γ-Fe₂O₃ and MoS₂. In Fig.1(e), the selected area electron diffraction (SAED) plot of Co-γ-Fe₂O₃/MoS₂ displays a set of concentric rings, and the rings and shiny spots matched the reflections of γ-Fe₂O₃ nanoparticles during the synthesis of Co-γ-Fe₂O₃/MoS₂ (Ge et al., 2021), indicating that the sample was composed of polycrystals (Xu et al., 2011). The elemental composition of Co-γ-Fe₂O₃/MoS₂ is displayed in Fig.1(f). The sample consisted of six elements, Fe, Co, O, Mo, S, and C, with percentages of 24.58%, 1.28%, 48.73%, 6.49%, 8.25%, and 10.67%, respectively. The presence of C was attributed to the use of Lacey carbon carriers during the test and the high C content in the precursors PEG 400 and tartaric acid (Liu et al., 2021a). These results demonstrated the successful doping of Co and compounding of MoS₂.

3.1.3 Elemental surface composition analysis

To further identify the surface composition and chemical states of the samples, the XPS spectra of the composites were recorded, as displayed in Fig.2(a)–2(f). The XPS survey spectrum of the Co-γ-Fe₂O₃/MoS₂ composite is displayed in Fig.2(a), and all constituent elements of the catalyst (i.e., Fe, Co, O, Mo, S, and C) were examined and characterized. In Fig.2(b), the binding energies at 724.5 and 710.9 eV were characteristic double peaks of Fe³⁺ 2p_1/2 and Fe³⁺ 2p_3/2, respectively (Flak et al., 2018). The peaks at 719.0 and 733.0 eV were assigned to satellite peaks and were also characteristic features of Fe³⁺ rather than Fe²⁺ (Wang et al., 2022c). The variation of binding energy between different orbitals was due to spin-orbit coupling and was equivalent to that of γ-Fe₂O₃ (Su et al., 2022). Therefore, the combination of XRD and XPS proved that the catalyst was γ-Fe₂O₃ rather than Fe₃O₄. The Co 2p XPS spectra (Fig.2(c)) displayed two peaks at binding energies of 780.5 and 795.9 eV, which corresponded to Co 2p_3/2 and Co 2p_1/2, respectively (Barik et al., 2017). The peaks at 780.3 and 795.5 eV in the Co 2p spectrum were attributed to Co²⁺ (Guo and Hu, 2022), while the peaks at 782.4 and 797.1 eV originated from Co³⁺. Furthermore, the peaks at 787.1 eV and 803.2 eV were identified as satellite peaks of Co³⁺ and Co²⁺, respectively (Li et al., 2022a). In Fig.2(d), the peaks at 228.7 and 232.1 eV were attributed to the Mo⁴⁺ 3d_5/2 and Mo⁴⁺ 3d_3/2 orbitals of MoS₂. The peak at 225.9 eV was assigned to the S 2s orbital of MoS₂, and the peak at 235.8 eV originated from Mo⁶⁺ 3d_3/2, which may indicate the presence of MoO₃ produced by partial oxidation of MoS₂ in air atmosphere (Qi et al., 2019). In the S 2p spectrum, the peaks at 161.5 and 162.7 eV (Fig.2(e)) were attributed to S 2p_3/2 and S 2p_1/2 (Yue et al., 2022). These results revealed that MoS₂ was synthesized successfully and exhibited excellent complexation with Co-γ-Fe₂O₃. The O 1s XPS spectrum of the catalyst could be deconvoluted into two components (Fig.2(f)) at 530.1 and 531.4 eV, which were ascribed to lattice oxygen (O_L) and highly oxidizing oxygen (O²⁻, O⁻, or –OH), respectively (Zeng et al., 2018). Additionally, the peak at 533.2 eV corresponded to physically adsorbed oxygen (O_ads) on the composite (Zhang et al., 2018).

3.1.4 Analyses of textural, surface charge, and magnetic properties

The specific surface areas (S_BET) and porous structures of synthesized Co-γ-Fe₂O₃ and Co-γ-Fe₂O₃/MoS₂ were studied by Brunauer-Emmett-Teller (BET) analysis. The obtained N₂ adsorption-desorption isotherms and the Barrett-Joyner-Halenda (BJH) porous structure distribution curves are displayed in Figs. S5(a) and S5(b). In the relative pressure (P/P₀) range of 0.1–0.9, all composites exhibited type IV isotherms with H3 hysteresis loops, indicating the mesoporosity of the composites (Sarkar et al., 2022a). Compared with Co-γ-Fe₂O₃ (S_BET = 33.42 m²/g), the surface area of Co-γ-Fe₂O₃/MoS₂ was larger (45.26 m²/g) (Fig. S5(a)). The relevant pore diameter distribution of composites peaked at about 10 nm (Fig. S5(b)), and the average pore size of Co-γ-Fe₂O₃/MoS₂ (11.36 nm) was slightly larger than Co-γ-Fe₂O₃ (10.09 nm). Additionally, the total pore capacity of Co-γ-Fe₂O₃/MoS₂ was higher than that of Co-γ-Fe₂O₃, which could be attributed to the uniform distribution of the Co-γ-Fe₂O₃ nanoparticles on the surface of MoS₂. The porous structure of Co-γ-Fe₂O₃/MoS₂ exhibited better adsorption performance and provided more active sites exposed to the material surface, which contributed to the degradation of the BPA molecules.

The variation in the surface zeta potential of the composite with solution pH is displayed in Fig. S5(c), indicating that the zero point charge (pH_zpc) of Co-γ-Fe₂O₃/MoS₂ was 4.25. Under acidic conditions (pH < 4.25), the material had a positive charge because of the protonation of the surface functional groups, and the negative charges at pH > 4.25 due to metal oxides being hydroxylated. Additionally, the surface zeta potential of Co-γ-Fe₂O₃/MoS₂ changed from 24.2 mV (pH 2) to −35.8 mV (pH 11) (Luciano et al., 2020).

The magnetic hysteresis loops of the samples were analyzed by using VSM, and the related curves were presented in Fig. S5(d). The magnetic moment of Co-γ-Fe₂O₃ decreased slightly from 46.03 to 41.82 emu/g after compounding the non-magnetic MoS₂. Nevertheless, the Co-γ-Fe₂O₃/MoS₂ still demonstrated excellent magnetic properties, and it could be easily separated from the mixture by applying a magnetic field to rapidly recover Co-γ-Fe₂O₃/MoS₂.

3.2 Catalytic performance of synthesized catalysts

3.2.1 Influence of molar Mo/Fe ratio

The effect of varying the molar Mo/Fe ratio on PMS activation by the catalysts for BPA removal is presented in Fig. S6(a). Compared with Co-γ-Fe₂O₃/PMS (98.2%), a slight increase in the BPA degradation efficiency was observed when the molar Mo/Fe ratio was raised from 1% to 4% in Co-γ-Fe₂O₃/MoS₂ catalysts, resulting in BPA degradation of 85.0%, 89.7%, 93.9%, and 95.9% at 30 min, respectively. When the Mo/Fe ratio was further increased to 6%, BPA removal was significantly improved (91.8% for Mo/Fe (5%) and 99.3% for Mo/Fe (6%) within 10 min). When the Mo/Fe ratio exceeded 6%, MoS₂ doping did not further influence BPA removal (99.2% for Mo/Fe (7%) and 98.1% for Mo/Fe (8%)). In Fig. S6(b), lnC₀/C exhibited a linear relationship with t, which can be expressed by Eq. (1):

(1)\ln C_0/C=K_{\mathrm{a}\mathrm{p}\mathrm{p}}t,

where C₀, C, and K_app denote the initial concentration of BPA, the concentration of BPA after t min of reaction, and the pseudo-first-order rate constant, respectively. It could be observed that K_app remarkably increased from CFM₀ to CFM₆. However, when the proportion of MoS₂ continued to increase, K_app remained nearly constant or slightly decreased. The appropriate amount of MoS₂ improved the catalytic activity, and CFM₆ expressed the highest catalytic performance with a K_app of 0.4799 min⁻¹, which was 3.83 times higher than that of CFM₀ (0.1253 min⁻¹). Overall, the degradation rate of BPA was accelerated as more active sites were available upon increasing the MoS₂ content. In addition, the total BPA degradation was enhanced by activating PMS and promoting Fe²⁺/Fe³⁺ cycling when MoS₂ was introduced, which generated active species capable of adsorbing BPA (Bai et al., 2020). The results show that only the optimal ratio of Mo/Fe demonstrated a significant promoting effect. Hence, a molar Mo/Fe ratio of 6% was identified as the optimal ratio in this work.

3.2.2 Influence of different reaction systems

The degradation efficiency of BPA was assessed for different reaction systems. As indicated in Fig.3(a), in the presence of only γ-Fe₂O₃, Co-γ-Fe₂O₃, or Co-γ-Fe₂O₃/MoS₂, 8.0%, 9.0%, and 6.4% of BPA was removed, respectively. The results implied that only a small amount of BPA could be removed by adsorption on the surface of the materials (Sheng et al., 2019; Sun et al., 2020a). In the presence of PMS alone, only a small amount of BPA was degraded due to the weak oxidation capacity of small amounts of PDS generated by PMS (Zhang et al., 2020). For the MoS₂/PMS and γ-Fe₂O₃/PMS catalyst systems, removal efficiencies of 15% and 60% BPA were achieved, respectively, because the transition metal ions activated PMS to produce ROSs, which accelerated the removal of BPA. However, the decomposition of BPA was limited due to the low catalytic performance of MoS₂ and γ-Fe₂O₃. When using Co-γ-Fe₂O₃ and Co-γ-Fe₂O₃/MoS₂ as the heterogeneous catalysts, significant degradation of BPA was observed (Bai et al., 2020). Co-γ-Fe₂O₃ activated PMS to remove 98.3% of BPA in 30 min, while 99.3% of BPA could be removed by the Co-γ-Fe₂O₃/MoS₂/PMS system in 10 min. This demonstrated that Co doping dramatically improved the catalytic performance, and the decomposition rate was accelerated via compounding MoS₂. As evident from the data in Table S1, Co-γ-Fe₂O₃/MoS₂ exhibited excellent catalytic activity for PMS to degrade BPA in a shorter reaction time compared with previously reported catalysts.

3.2.3 Influence of different reaction conditions

Variations in the initial pH, PMS, and catalyst concentration are essential factors during catalytic oxidation. The effect of PMS concentration on the degradation of BPA via the Co-γ-Fe₂O₃/MoS₂/PMS system is displayed in Fig.3(b). The efficiency of BPA removal gradually improved when the PMS concentration increased from 0.3 to 0.5 mmol/L, since higher PMS concentrations generated more ROSs (Li et al., 2020). However, the degradation efficiency slightly declined when the PMS concentration was further increased, which was attributed to several reasons: 1) PMS could not be fully activated due to the limited active sites of Co-γ-Fe₂O₃/MoS₂; 2) excess PMS might have reacted with free radicals (•OH, SO₄^•−), as shown in Eqs. (2) and (3) (Bai et al., 2020).

(2){\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}+{\mathrm{S}\mathrm{O}}_4^{\text{•}-}\to {\mathrm{S}\mathrm{O}}_5^{\text{•}-}+{\mathrm{S}\mathrm{O}}_4^{2-}+{\mathrm{H}}^{+}

(3){\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}+\text{•}\mathrm{O}\mathrm{H}\to {\mathrm{S}\mathrm{O}}_5^{\text{•}-}+{\mathrm{H}}_2\mathrm{O}

As demonstrated in Fig.3(c), there was a significant enhancement in the BPA removal efficiency from 84.0% to 99.1% when the amount of Co-γ-Fe₂O₃/MoS₂ was raised from 0.05 to 0.15 g/L. The removal efficiency remained almost constant upon further increasing it to 0.25 g/L, which implies that 0.15 g/L of catalyst provided enough active sites to activate PMS. Therefore, 0.15 g/L of Co-γ-Fe₂O₃/MoS₂ was used for subsequent experiments.

The influence of initial pH on BPA degradation was studied. In Fig.3(d), the efficiency of BPA decomposition changed with initial pH. This impact of pH on BPA removal was caused by the formation of different active species. The first pK_a of BPA is 9.2 (Zhang et al., 2013), i.e., at pH values below 9.2, the cationic and zwitterionic molecular structures of BPA dominated, while the anionic form of BPA dominated above pH 9.2. Since the pK_a of PMS is 9.4, PMS mainly exists as HSO₅⁻ below pH 9.4 and is primarily composed of SO₅²⁻ at higher pH. Thus, at pH < 5.0, the combination of positive or neutral BPA molecules, negatively charged PMS, and positively charged catalyst surfaces promoted the removal of BPA. Simultaneously, due to the similar charge Co-γ-Fe₂O₃/MoS₂ and PMS, the decomposition efficiency of BPA was slightly reduced in the pH range of 5–9 (Sarkar et al., 2022b). Under highly alkaline conditions, SO₄^•− would be eliminated by hydroxyl ions to form •OH (Niu et al., 2022), and PMS would be deprotonated and produce •SO₅²⁻ with weak oxidizing ability (Yao et al., 2022). Thus, the efficiency of BPA removal was noticeably reduced in highly alkaline environments.

3.2.4 Influence of environmental factors

Inorganic ions including Cl⁻, NO₃⁻, HCO₃⁻, and SO₄²⁻ as well as natural organics such as humic acid (HA) are commonly found in natural water and might affect the removal efficiency of BPA. Therefore, the effects of these representative substances on the efficiency of BPA removal were studied, as presented in Fig.3(e) and 3(f). The BPA removal efficiency was significantly promoted when the Cl⁻ concentration ranged from 2 to 10 mmol/L. This could be attributed to the following reasons: 1) a series of reactions occurred between Cl⁻ and SO₄^•− in the Co-γ-Fe₂O₃/MoS₂/PMS system to form highly oxidizing chlorine species, such as Cl^• and Cl₂^•− (Zhu et al., 2022) (Eqs. (4)–(7)); 2) the direct reactions of Cl⁻ with PMS generated HOCl and Cl₂, which promoted the degradation of BPA (Giannakis et al., 2021) (Eqs. (8) and (9)). In the presence of SO₄²⁻, the BPA removal efficiency was slightly promoted when increasing the concentration from 2 to 10 mmol/L. This was possibly because SO₄²⁻ targeted the π electrons of BPA and facilitated the breaking of bonds in the organic molecule, which decreased the energy barriers of SO₄^•− to attack BPA.

(4){\mathrm{C}\mathrm{l}}^{-}+{\mathrm{S}\mathrm{O}}_4^{\text{•}-}\to {\mathrm{C}\mathrm{l}}^{\text{•}}+{\mathrm{S}\mathrm{O}}_4^{2-}

(5){\mathrm{C}\mathrm{l}}^{-}+\text{•}\mathrm{O}\mathrm{H}\to {\mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}}^{\text{•}-}

(6){\mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}}^{\text{•}-}+{\mathrm{H}}^{+}\to {\mathrm{C}\mathrm{l}}^{\text{•}}+{\mathrm{H}}_2\mathrm{O}

(7){\mathrm{C}\mathrm{l}}^{-}+{\mathrm{C}\mathrm{l}}^{\text{•}}\to {\mathrm{C}\mathrm{l}}_2^{\text{•}-}

(8){\mathrm{C}\mathrm{l}}^{-}+{\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}\to \mathrm{H}\mathrm{O}\mathrm{C}\mathrm{l}+{\mathrm{S}\mathrm{O}}_4^{2-}

(9){\mathrm{C}\mathrm{l}}^{-}+{\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}+{\mathrm{H}}^{+}\to {\mathrm{S}\mathrm{O}}_4^{2-}+{\mathrm{C}\mathrm{l}}_2+{\mathrm{H}}_2\mathrm{O}

Generally, NO₃⁻ in solution reacts with •OH and SO₄^•− to form the less-oxidizing NO₃^• (Eqs. (10) and (11)) but it suppresses the BPA removal efficiency only to a small extent, which could be basically ignored even at high concentrations (Niu et al., 2022).

(10){\mathrm{S}\mathrm{O}}_4^{\text{•}-}+{\mathrm{N}\mathrm{O}}_3^{-}\to {\mathrm{S}\mathrm{O}}_4^{2-}+{\mathrm{N}\mathrm{O}}_3^{\text{•}}

(11){\mathrm{N}\mathrm{O}}_3^{-}+\text{•}\mathrm{O}\mathrm{H}\to {\mathrm{N}\mathrm{O}}_3^{\text{•}}+{\mathrm{O}\mathrm{H}}^{-}

However, in the presence of HCO₃⁻, the degradation of BPA was reduced, even at low concentrations (2 mmol/L). As an explanation for this strong effect, HCO₃⁻ reacted with H⁺ to neutralize the environment, the marginal S atoms in MoS₂ were not easily trapped, and the activation of PMS by Mo(IV) was remarkably inhibited.

Natural organic matter (NOM) in the environment often competes with ROSs to decrease the removal efficiency of advanced oxidation processes for organic pollutants. The BPA removal percentage declined gradually as the concentration of HA increased from 5 to 10 mg/L because HA competed with BPA for the generated free radical species and reduced the number of active species in the reaction system. In addition, HA was adsorbed on the catalyst surface, covering the active sites to weaken interactions of the catalyst with PMS and BPA (Chen et al., 2018; Giannakis et al., 2021).

3.3 Mechanism of BPA degradation

3.3.1 Identification of major ROSs

To analyze the dominant ROSs generated in the Co-γ-Fe₂O₃/MoS₂/PMS system, MeOH, TBA, and p-BQ were added to the degradation system to quench various ROSs. The reaction rates of these quenching agents are kSO₄^•− = 2.5 × 10⁷ M⁻¹·s⁻¹ and k_•OH = 9.7 × 10⁸ M⁻¹·s⁻¹ for MeOH, k•OH = 3.8 × 10⁸–7.6 × 10⁸ M⁻¹·s⁻¹ for TBA, and kO₂^•− = 3.5 × 10⁸–7.8 × 10⁸ M⁻¹·s⁻¹ for p-BQ (Bai et al., 2020). As shown in Fig.4(a), the BPA degradation efficiency was hardly affected by adding 25 mmol/L p-BQ (50 times the PMS concentration). When 250 mmol/L TBA (500 times the PMS concentration) was added, the removal of BPA was slightly inhibited, and its removal rate decreased to 94.1% in 30 min. Furthermore, the removal efficiency dropped sharply to 36.4% in the presence of 250 mmol/L MeOH (500 times the PMS concentration). The above results indicated that SO₄^•− played a major role in the reaction. To assess the possible involvement of ¹O₂ as part of a non-radical decomposition pathway, furfuryl alcohol (FFA) was chosen as a quenching agent with reaction rates of 1.5 × 10¹⁰ M⁻¹·s⁻¹ and 1.2 × 10⁸ M⁻¹·s⁻¹ for •OH and ¹O₂ (Ren et al., 2022), respectively. Specifically, the addition of 10 mmol/L FFA (20 times the PMS concentration) remarkably inhibited BPA degradation, which was more pronounced than MeOH, indicating that the contribution of ¹O₂ to BPA decomposition could not be overlooked.

To further determine the ROSs generated in the system, the EPR test was performed using DMPO and TEMP as spin-trapping agents. As displayed in Fig.4(b), no signal of free radicals was observed in the pure PMS system because PMS could not be activated without a catalyst. In contrast, the typical signal of DMPO-•OH with a peak ratio of 1:2:2:1 was detected in the Co-γ-Fe₂O₃/MoS₂/PMS system, while some weak signals around these four peaks could be attributed to DMPO-SO₄^•−. When TEMP was used instead of DMPO, a weak 1:1:1 triplet signal peak of ¹O₂ could be discovered in the pure PMS system (Fig.4(c)), which was the result of the autolysis of PMS, which would reasonably explain why BPA was slightly degraded in the pure PMS system (Sarkar et al., 2022b). The intensity of the TEMP-¹O₂ signal increased significantly with the addition of the catalyst, indicating that PMS produced a large amount of ¹O₂ due to activation by the catalyst. To further examine the generation of ¹O₂, we tested the singlet oxygen in D₂O by ESR because ¹O₂ had a longer lifetime in D₂O (20–32 μs) (Yun et al., 2018). As demonstrated in Fig.4(c), by comparing the typical triple ESR signal intensity of ¹O₂ in both water bodies, the signal intensity in D₂O was stronger than that in H₂O. Previous experiments have demonstrated that singlet oxygen plays an important role in the BPA degradation process. Higher reaction kinetic constants are predicted for the degradation of BPA in D₂O than in H₂O (Chen et al., 2019a; Wu et al., 2020). The above results indicated that BPA could be degraded via both radical and non-radical pathways, where ¹O₂ and SO₄^•− were the major ROSs.

3.3.2 Mechanistic analysis of Co-γ-Fe₂O₃/MoS₂/PMS system

The chemical surface compositions of fresh and used Co-γ-Fe₂O₃/MoS₂ were determined by XPS, as presented in Fig.5. In the Fe 2p XPS spectra (Fig.5(a)), the Fe 2p_3/2 peak of used Co-γ-Fe₂O₃/MoS₂ was split into two peaks at 710.6 and 712.4 eV, relating to Fe²⁺ and Fe³⁺, respectively. The Fe 2p_1/2 peak was also split into two peaks at 724.1 eV (Fe²⁺) and 726.3 eV (Fe³⁺) (Niu et al., 2021), indicating that Fe³⁺ on the catalyst surface was partially reduced to Fe²⁺. Correspondingly, the relative contents of different Co valence states varied slightly before and after the reaction (Fig.5(b)), and a decrease in the amount of Co²⁺ from 54.4% to 51.1% was accompanied by an increase in Co³⁺ from 45.6% to 48.9%. The rather small change in the Co²⁺/Co³⁺ ratio indicated that Co²⁺ was mostly regenerated after Co³⁺ reduction, allowing Co-γ-Fe₂O₃/MoS₂ to maintain an excellent activation potential (Fan et al., 2017). Furthermore, in Fig.5(c), the Mo^4+/Mo⁶⁺ ratios on the surface of Co-γ-Fe₂O₃/MoS₂ before and after the reaction were 2.704 and 3.167, respectively. An increase in the amount of Mo⁴⁺ indicated that Mo⁴⁺ could be rapidly regenerated after the reaction, thus accelerating the rate of Fe²⁺/Fe³⁺ cycling (Zeng et al., 2019a). Therefore, the content of Fe²⁺, Co^2+, and Mo⁴⁺ all increased, which was favorable for the decomposition of BPA.

Based on the results discussed above and previous literature, possible mechanisms of PMS activation including both radical and non-radical pathways are presented in Fig.6. First, Co³⁺ and/or Fe³⁺ were consumed by PMS to regenerate Co²⁺ and/or Fe²⁺, while SO₅^•− with a lower activity was formed (Eq. (12)). Then, the HSO₅⁻ that originated from PMS was activated by Co²⁺ and/or Fe²⁺ to reduce to SO₄^•− (Eq. (13)) (Liu et al., 2021b). Since the standard redox potential of Co²⁺/Co³⁺ (1.81 V) is higher than that of Fe²⁺/Fe³⁺ (0.77 V), Co³⁺ could be reduced by Fe²⁺ (Eq. (14)). Secondly, SO₅^•− could be further converted to ¹O₂ by reaction with H₂O (Eq. (15)) (Sun et al., 2020b). Furthermore, ¹O₂ could also be generated by chemisorbed oxygen (O*) from lattice oxygen (O_Lat) in the presence of PMS (Eqs. (16) and (17)), while SO₄^•− could be hydrolyzed to produce •OH (Eq. (18)) (Dong et al., 2019). Mo⁴⁺ might be oxidized to Mo⁶⁺ by PMS, accompanied by the formation of SO₄^•− and •OH (Eq. (19)). After the removal of S atoms, the exposed Mo⁴⁺ became very reactive and contributed to the Fe³⁺/Fe²⁺ cycle (Eq. (20)) (Zeng et al., 2019b). Finally, BPA would be removed and mineralized by large amounts of ¹O₂, SO₄^•−, and •OH (Eq. (21)).

(12){\mathrm{C}\mathrm{o}}^{3+}/{\mathrm{F}\mathrm{e}}^{3+}+{\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{2+}/{\mathrm{F}\mathrm{e}}^{2+}+{\mathrm{S}\mathrm{O}}_5^{\text{•}-}+{\mathrm{H}}^{+}

(13){\mathrm{C}\mathrm{o}}^{2+}/{\mathrm{F}\mathrm{e}}^{2+}+{\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}\to {\mathrm{C}\mathrm{o}}^{3+}/{\mathrm{F}\mathrm{e}}^{3+}+{\mathrm{S}\mathrm{O}}_4^{\text{•}-}+{\mathrm{O}\mathrm{H}}^{-}

(14){\mathrm{F}\mathrm{e}}^{2+}+{\mathrm{C}\mathrm{o}}^{3+}\rightleftharpoons {\mathrm{F}\mathrm{e}}^{3+}+{\mathrm{C}\mathrm{o}}^{2+}

(15)2{\mathrm{S}\mathrm{O}}_5^{\text{•}-}+{\mathrm{H}}_2\mathrm{O}\to {1.5}^1{\mathrm{O}}_2+2{\mathrm{H}\mathrm{S}\mathrm{O}}_4^{-}

(16){\mathrm{O}}_{\mathrm{L}\mathrm{a}\mathrm{t}}\to {\mathrm{O}}^{\ast }

(17){\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}+{\mathrm{O}}^{\ast }{\to }^1{\mathrm{O}}_2+{\mathrm{H}\mathrm{S}\mathrm{O}}_4^{-}

(18){\mathrm{S}\mathrm{O}}_4^{\text{•}-}+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{S}\mathrm{O}}_4^{2-}+\text{•}\mathrm{O}\mathrm{H}+{\mathrm{H}}^{+}

(19){\mathrm{M}\mathrm{o}}^{4+}+{\mathrm{H}\mathrm{S}\mathrm{O}}_5^{-}\to {\mathrm{M}\mathrm{o}}^{6+}+{\mathrm{S}\mathrm{O}}_4^{\text{•}-}+{\mathrm{O}\mathrm{H}}^{-}

(20){\mathrm{M}\mathrm{o}}^{4+}+{\mathrm{F}\mathrm{e}}^{3+}\to {\mathrm{M}\mathrm{o}}^{6+}+{\mathrm{F}\mathrm{e}}^{2+}

(21)\begin{array}{rl}& {\mathrm{S}\mathrm{O}}_4^{\text{•}-}{+}^1{\mathrm{O}}_2+\text{•}\mathrm{O}\mathrm{H}+\mathrm{B}\mathrm{P}\mathrm{A}\to \\ & \mathrm{D}\mathrm{e}\mathrm{g}\mathrm{r}\mathrm{a}\mathrm{d}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{t}\mathrm{s}\to {\mathrm{C}\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}\end{array}

3.3.3 Possible degradation pathways of BPA

LC-MS was used to identify the intermediates in the Co-γ-Fe₂O₃/MoS₂ reaction system to explore the possible degradation pathways of BPA. Ten products (P1–P10) were formed mainly via oxidation, hydroxylation, and ring cleavage. The obtained mass spectra of the intermediates and the corresponding molecular formulae and different mass-to-charge (m/z) ratios are provided in Fig. S7 and Table S2. According to the products detected by LC-MS and previous studies, three possible decomposition pathways of BPA were formulated (Fig.7). In the first possible route, the BPA molecule was hydroxylated to form P1 (Kakavandi et al., 2022) due to the attack of ROSs, after which P1 was deprotonated and dehydrated to open the ring to form P2. Then, P4 and P5 were formed after ring opening, hydroxylation, and oxidation of the benzene ring of P1 and P2 by ROS attack and finally further oxidized to P7. In pathway II, the BPA molecule was attacked by ROSs causing ring cleavage to form P3. The electron-rich C–C bond on P3 was broken due to the attack of its methyl group by the ROSs, followed by conversion to P6 (Wang et al., 2022a). Finally, the para-phenol P10 was generated due to oxidation and C–C bond fission of P6 (Yang et al., 2022). In another pathway, the dimer P8 was produced by the oxidation and polymerization of BPA, which was then further oxidized to P9. Eventually, the resulting intermediates were degraded to small-molecule products or mineralized to CO₂ and H₂O (Sun et al., 2020a; Liu et al., 2021b; Liu et al., 2022).

3.3.4 Toxicity analysis of intermediate products

The toxicity of BPA and its intermediates was predicted by quantitative structure-activity relationship (QSAR) using a toxicity estimation software tool (T.E.S.T.), as shown in Fig.8. In Fig.8(a), although most of the products were more mutagenic than BPA, all of them were negative, including P3, P4, and P5, which had positive mutagenicity, which became negative after being attacked by ROSs (Jiang et al., 2020). In addition, all the products except P2, P8, and P9 were less developmentally toxic than BPA (Fig.8(b)). Interestingly, Fig.8(c) reveals that the bioconcentration factors of the intermediate products were smaller than those of BPA except for P1, P6, P7, P8, and P9. They were further reduced due to hydroxylation reactions, diminishing their threat to the environment (Yang et al., 2021). In addition, the acute toxicity of BPA and its intermediates was evaluated by fathead minnow LC₅₀ (96 h) (i.e., the compound concentration that killed 50% of fathead minnow after 96 h), and the results are displayed in Fig.8(d), where the acute toxicity of all products was lower than that of BPA except for P1, P7, P8, and P9. Overall, the overall toxicity of BPA and its intermediates was reduced after degradation in the Co-γ-Fe₂O₃/MoS₂/PMS system.

3.4 Evaluation of catalyst applicability

To investigate the degradation ability of Co-γ-Fe₂O₃/MoS₂ toward other representative organic pollutants, levofloxacin (LFX), sulfamethoxazole (SMX), carbamazepine (CBZ), tetracycline (TC), and ciprofloxacin (CIP) were studied as contaminants. The experimental conditions were consistent with the conditions for BPA degradation, and the results are shown in Fig. S8. According to Fig. S8(a), the removal rates of the selected pollutants by the Co-γ-Fe₂O₃/MoS₂/PMS system all exceeded 80%, and their corresponding reaction rate constants well all high, with the largest value corresponding to BPA (0.4799 min⁻¹) and a relatively slow reaction rate for TC (0.0751 min⁻¹) (Fig. S8(b)). The lower TC removal efficiency and corresponding reaction rate constant may be related to its relatively large molecular weight (He et al., 2022a), which decreased the removal efficiency. Thus, TC may require a large amount of PMS to be effectively degraded.

The reusability of catalysts is indispensable for practical applications. To examine the reusability of Co-γ-Fe₂O₃ and Co-γ-Fe₂O₃/MoS₂, cycling experiments were performed to remove BPA. As demonstrated in Fig.9(a), the speed of BPA removal by Co-γ-Fe₂O₃ was significantly reduced in the fourth and fifth cycles. This might be explained by the damage of the active sites of Co-γ-Fe₂O₃ caused by the intermediate products of BPA during the removal process and the adsorption of SO₄²⁻ and organic pollutants on the surface of Co-γ-Fe₂O₃ (Zheng et al., 2019). Additionally, the XRD patterns of Co-γ-Fe₂O₃ before use and after the fifth cycle (Fig.9(b)) revealed the disappearance of the diffraction peak corresponding to the (111) crystal plane, indicating the poor stability of Co-γ-Fe₂O₃. The results of the cycling experiment after compounding MoS₂ are shown in Fig.9(c). After five cycles, the removal of BPA still exceeded 99%. In addition, the comparison of XRD patterns and the XPS and VSM results before and after the reaction indicated that the structural stability of Co-γ-Fe₂O₃ was enhanced after MoS₂ compounding (He et al., 2022a). Furthermore, in the Co-γ-Fe₂O₃/PMS system, the leached contents of Fe ions and Co ions after the reaction were 6.983 and 0.697 mg/L, respectively. However, they were only 1.904 and 0.201 mg/L in the Co-γ-Fe₂O₃/MoS₂/PMS system (Fig. S9), which could be assigned to the synergistic effect between Fe, Co, and Mo (Sun et al., 2020b).

4 Conclusions

The environmentally friendly magnetic Co-γ-Fe₂O₃/MoS₂ nanocomposite was synthesized by a two-step hydrothermal method through in situ growth of MoS₂ nanosheets on Co-γ-Fe₂O₃ nanoparticles. The Co-γ-Fe₂O₃/MoS₂/PMS system exhibited an excellent BPA removal efficiency due to the synergistic effect between Fe, Co, and Mo. In the Co-γ-Fe₂O₃/MoS₂/PMS system, the BPA removal rate remained above 99% within 30 min after five reuse cycles of the catalyst. Its ability to maintain a high efficiency indicated that Co-γ-Fe₂O₃ possessed excellent stability and low ion leaching after compounding with MoS₂. In addition, both radicals and non-radicals were involved in the degradation of BPA, and ¹O₂ and SO₄^•− were the main ROSs in the Co-γ-Fe₂O₃/MoS₂/PMS system. Ten intermediates of BPA degradation were identified by LC-MS, and three possible decomposition pathways were proposed. The toxicities of BPA and its intermediates were predicted by T.E.S.T, indicating that the toxicity of BPA was reduced after degradation by the Co-γ-Fe₂O₃/MoS₂/PMS system, which demonstrated its potential applications for degrading BPA in actual water bodies. Overall, Co-γ-Fe₂O₃/MoS₂ is an efficient and durable catalyst with wide applicability and a promising future for organic pollutant treatment.

References

[1] Abareshi A , Salehi N . (2022). The effect of Fe₃O₄ nanoparticles on structural, optical, and thermal properties MoS₂ nanoflakes. Journal of Materials Science Materials in Electronics, 33(33): 25153–25162

[2] Bai R , Yan W , Xiao Y , Wang S , Tian X , Li J , Xiao X , Lu X , Zhao F . (2020). Acceleration of peroxymonosulfate decomposition by a magnetic MoS₂/CuFe₂O₄ heterogeneous catalyst for rapid degradation of fluoxetine. Chemical Engineering Journal, 397: 125501

[3] Barik R , Jena B K , Mohapatra M . (2017). Metal doped mesoporous FeOOH nanorods for high performance supercapacitors. RSC Advances, 7(77): 49083–49090

[4] Chen C , Ma T , Shang Y , Gao B , Jin B , Dan H , Li Q , Yue Q , Li Y , Wang Y . . (2019a). In-situ pyrolysis of enteromorpha as carbocatalyst for catalytic removal of organic contaminants: considering the intrinsic N/Fe in enteromorpha and non-radical reaction. Applied Catalysis B: Environmental, 250: 382–395

[5] Chen G , Zhang X , Gao Y , Zhu G , Cheng Q , Cheng X . (2019b). Novel magnetic MnO₂/MnFe₂O₄ nanocomposite as a heterogeneous catalyst for activation of peroxymonosulfate (PMS) toward oxidation of organic pollutants. Separation and Purification Technology, 213: 456–464

[6] Chen L , Ding D , Liu C , Cai H , Qu Y , Yang S , Gao Y , Cai T . (2018). Degradation of norfloxacin by CoFe₂O₄-GO composite coupled with peroxymonosulfate: a comparative study and mechanistic consideration. Chemical Engineering Journal, 334: 273–284

[7] Dong H , Chen J , Feng L , Zhang W , Guan X , Strathmann T J . (2019). Degradation of organic contaminants through activating bisulfite by cerium(IV): a sulfate radical-predominant oxidation process. Chemical Engineering Journal, 357: 328–336

[8] Duan X , Su C , Miao J , Zhong Y , Shao Z , Wang S , Sun H . (2018). Insights into perovskite-catalyzed peroxymonosulfate activation: maneuverable cobalt sites for promoted evolution of sulfate radicals. Applied Catalysis B: Environmental, 220: 626–634

[9] Fan Y , Ji Y , Zheng G , Lu J , Kong D , Yin X , Zhou Q . (2017). Degradation of atrazine in heterogeneous Co₃O₄ activated peroxymonosulfate oxidation process: kinetics, mechanisms, and reaction pathways. Chemical Engineering Journal, 330: 831–839

[10] Flak D , Chen Q , Mun B S , Liu Z , Rękas M , Braun A . (2018). In situ ambient pressure XPS observation of surface chemistry and electronic structure of α-Fe₂O₃ and γ-Fe₂O₃ nanoparticles. Applied Surface Science, 455: 1019–1028

[11] Gao J , Song J , Ye J , Duan X , Dionysiou D D , Yadav J S , Nadagouda M N , Yang L , Luo S . (2021). Comparative toxicity reduction potential of UV/sodium percarbonate and UV/hydrogen peroxide treatments for bisphenol A in water: an integrated analysis using chemical, computational, biological, and metabolomic approaches. Water Research, 190: 116755

[12] Ge Y , Li C , Waterhouse G I N , Jiang X , Zhang Z , Yu L . (2021). Polypyrrole/γ-Fe₂O₃/g-C₃N₄ nanocomposites for high-performance electromagnetic wave absorption. Synthetic Metals, 274: 116716

[13] Geng Y , Chen D , Li N , Xu Q , Li H , He J , Lu J . (2021). Z-scheme 2D/2D α-Fe₂O₃/g-C₃N₄ heterojunction for photocatalytic oxidation of nitric oxide. Applied Catalysis B: Environmental, 280: 119409

[14] Giannakis S , Lin K A , Ghanbari F . (2021). A review of the recent advances on the treatment of industrial wastewaters by sulfate radical-based advanced oxidation processes (SR-AOPs). Chemical Engineering Journal, 406: 127083

[15] Golestani E , Javanbakht M , Ghafarian-Zahmatkesh H , Beydaghi H , Ghaemi M . (2018). Tartaric acid assisted carbonization of LiFePO₄ synthesized through in situ hydrothermal process in aqueous glycerol solution. Electrochimica Acta, 259: 903–915

[16] Guo P , Hu X . (2022). Co, Fe co-doped g-C₃N₄ composites as peroxymonosulfate activators under visible light irradiation for levofloxacin degradation: characterization, performance and synergy mechanism. Colloids and Surfaces. A, Physicochemical and Engineering Aspects, 648: 129423

[17] He Y , Qian J , Wang P , Wu J , Lu B , Tang S , Gao P . (2022a). Acceleration of levofloxacin degradation by combination of multiple free radicals via MoS₂ anchored in manganese ferrite doped perovskite activated PMS under visible light. Chemical Engineering Journal, 431: 133933

[18] He Y , Qian J , Xu B , Wang P , Lu B , Tang S , Gao P . (2022b). Encapsulate SrCoO₃ perovskite crystal within molybdenum disulfide layer as core-shell structure to enhance electron transfer for peroxymonosulfate activation. Separation and Purification Technology, 283: 120199

[19] Ikram M , Khan M I , Raza A , Imran M , Ul-Hamid A , Ali S . (2020). Outstanding performance of silver-decorated MoS₂ nanopetals used as nanocatalyst for synthetic dye degradation. Physica E, Low-Dimensional Systems and Nanostructures, 124: 114246

[20] Jiang J , Wang X , Liu Y , Ma Y , Li T , Lin Y , Xie T , Dong S . (2020). Photo-Fenton degradation of emerging pollutants over Fe-POM nanoparticle/porous and ultrathin g-C₃N₄ nanosheet with rich nitrogen defect: degradation mechanism, pathways, and products toxicity assessment. Applied Catalysis B: Environmental, 278: 119349

[21] Kakavandi B , Alavi S , Ghanbari F , Ahmadi M . (2022). Bisphenol A degradation by peroxymonosulfate photo-activation coupled with carbon-based cobalt ferrite nanocomposite: Performance, upgrading synergy and mechanistic pathway. Chemosphere, 287: 132024

[22] Li J , Li J , Chen L , Lin Y , Liu X , Gong X , Li D . (2015). Characterization and magnetism of Co-modified γ-Fe₂O₃ core-shell nanoparticles by enhancement using NaOH. Journal of Magnetism and Magnetic Materials, 374: 157–163

[23] Li M , Li Y , Yu P , Zhao H , Xiang L , Feng N , Li Q , He K , Luo X , Cai Q . . (2022a). Exploring degradation mechanism of tetracycline via high-effective peroxymonosulfate catalysts of montmorillonite hybridized CoFe composites and safety assessment. Chemical Engineering Journal, 427: 130930

[24] Li Q , Lv C , Xia X , Peng C , Yang Y , Guo F , Zhang J . (2022b). Separation/degradation behavior and mechanism for cationic/anionic dyes by Ag-functionalized Fe₃O₄-PDA core-shell adsorbents. Frontiers of Environmental Science & Engineering, 16(11): 138

[25] Li R , Huang J , Cai M , Huang J , Xie Z , Zhang Q , Liu Y , Liu H , Lv W , Liu G . (2020). Activation of peroxymonosulfate by Fe doped g-C₃N₄/graphene under visible light irradiation for Trimethoprim degradation. Journal of Hazardous Materials, 384: 121435

[26] Liu H , Luo S , Yan S , Wang Q , Hu D , Wang Y , Feng J , Yi T . (2019). High-performance α-Fe₂O₃/C composite anodes for lithium-ion batteries synthesized by hydrothermal carbonization glucose method used pickled iron oxide red as raw material. Composites. Part B, Engineering, 164: 576–582

[27] Liu L , Hu T , Dai K , Zhang J , Liang C . (2021a). A novel step-scheme BiVO₄/Ag₃VO₄ photocatalyst for enhanced photocatalytic degradation activity under visible light irradiation. Chinese Journal of Catalysis, 42(1): 46–55

[28] Liu M, Zhang L, Han L, Mei C, Xu C, Yuan R, Geng C (2023). Can heat-activated peroxymonosulfate be used as a pretreatment to mitigate fouling for membrane distillation: performance of individual organics? Water, 15(6): 1148

[29] Liu T , Wang C , Han Y , Bai C , Ren H , Liu Y , Han X . (2022). Oxidative polymerization of bisphenol A (BPA) via H-abstraction by Bi_2.15WO₆ and persulfate: importance of the surface complexes. Chemical Engineering Journal, 435: 134816

[30] Liu Z , Wan J , Ma Y , Wang Y . (2021b). In situ synthesis of FeOCl@MoS₂ on graphite felt as novel electro-fenton cathode for efficient degradation of antibiotic ciprofloxacin at mild pH. Chemosphere, 273: 129747

[31] Long X , Luo J , Zhong Z , Zhu Y , Zhang C , Wan J , Zhou H , Zhang B , Xia D . (2023). Performance and mechanism of carbamazepine removal by FeS-S₂O₈²⁻ process: experimental investigation and DFT calculations. Frontiers of Environmental Science & Engineering, 17(9): 113

[32] Lu J , Zhou Y , Ling L , Zhou Y . (2022). Enhanced activation of PMS by a novel Fenton-like composite Fe₃O_4/S-WO₃ for rapid chloroxylenol degradation. Chemical Engineering Journal, 446: 137067

[33] Lu Y , Ding Z , Zhang J , Fu C , Xia X , Fang Y . (2019). Degradation of atrazine by UV/PMS in phosphate buffer. Polish Journal of Environmental Studies, 28(4): 2735–2744

[34] Luciano A J R , de Sousa Soletti L , Ferreira M E C , Cusioli L F , de Andrade M B , Bergamasco R , Yamaguchi N U . (2020). Manganese ferrite dispersed over graphene sand composite for methylene blue photocatalytic degradation. Journal of Environmental Chemical Engineering, 8(5): 104191

[35] Ma Q , Zhang X , Guo R , Zhang H , Cheng Q , Xie M , Cheng X . (2019). Persulfate activation by magnetic γ-Fe₂O₃/Mn₃O₄ nanocomposites for degradation of organic pollutants. Separation and Purification Technology, 210: 335–342

[36] Mao J , Quan X , Wang J , Gao C , Chen S , Yu H , Zhang Y . (2018). Enhanced heterogeneous Fenton-like activity by Cu-doped BiFeO₃ perovskite for degradation of organic pollutants. Frontiers of Environmental Science & Engineering, 12(6): 10

[37] Niu B , Wang L , Li M , Yao W , Zang K , Zhou L , Hu X , Zheng Y . (2022). Lattice B-doping evolved ferromagnetic perovskite-like catalyst for enhancing persulfate-based degradation of norfloxacin. Journal of Hazardous Materials, 425: 127949

[38] Niu L , Zhang G , Xian G , Ren Z , Wei T , Li Q , Zhang Y , Zou Z . (2021). Tetracycline degradation by persulfate activated with magnetic γ-Fe₂O₃/CeO₂ catalyst: performance, activation mechanism and degradation pathway. Separation and Purification Technology, 259: 118156

[39] Pan X , Yan L , Qu R , Wang Z . (2018). Degradation of the UV-filter benzophenone-3 in aqueous solution using persulfate activated by heat, metal ions and light. Chemosphere, 196: 95–104

[40] Qi K , Yuan Z , Hou Y , Zhao R , Zhang B . (2019). Facile synthesis and improved Li-storage performance of Fe-doped MoS₂/reduced graphene oxide as anode materials. Applied Surface Science, 483: 688–695

[41] Ren W , Cheng C , Shao P , Luo X , Zhang H , Wang S , Duan X . (2022). Origins of electron-transfer regime in persulfate-based nonradical oxidation processes. Environmental Science & Technology, 56(1): 78–97

[42] Sakthi Athithan A S , Jeyasundari J , Jacob Y B A . (2021). Biological synthesis, physico-chemical characterization of undoped and Co doped α-Fe₂O₃ nanoparticles using Tribulus terrestris leaf extract and its antidiabetic, antimicrobial applications. Advances in Natural Sciences. Nanoscience and Nanotechnology, 12(4): 045003

[43] Sarkar P , De S , Neogi S . (2022a). Microwave assisted facile fabrication of dual Z-scheme g-C₃N₄ /ZnFe₂O₄/Bi₂S₃ photocatalyst for peroxymonosulphate mediated degradation of 2,4,6-trichlorophenol: the mechanistic insights. Applied Catalysis B: Environmental, 307: 121165

[44] Sarkar P , Roy D , Bera B , De S , Neogi S . (2022b). Efficient photocatalytic degradation of ciprofloxacin using novel dual Z-scheme gCN/CuFe₂O₄/MoS₂ mediated peroxymonosulphate activation. Chemical Engineering Journal, 430: 132834

[45] Sheng B , Yang F , Wang Y , Wang Z , Li Q , Guo Y , Lou X , Liu J . (2019). Pivotal roles of MoS₂ in boosting catalytic degradation of aqueous organic pollutants by Fe(II)/PMS. Chemical Engineering Journal, 375: 121989

[46] Su M , Li H , Liu Z , Peng H , Huang S , Zhou Y , Liao C , Song G , Chen D . (2022). Highly-efficient and easy separation of γ-Fe₂O₃ selectively adsorbs U(VI) in waters. Environmental Research, 210: 112917

[47] Sun Q , Xu B , Yang J , Qian T , Jiang H . (2020a). Layered oxides supported Co–Fe bimetal catalyst for carbamazepine degradation via the catalytic activation of peroxymonosulfate. Chemical Engineering Journal, 400: 125899

[48] Sun X , Xu D , Dai P , Liu X , Tan F , Guo Q . (2020b). Efficient degradation of methyl orange in water via both radical and non-radical pathways using Fe–Co bimetal-doped MCM-41 as peroxymonosulfate activator. Chemical Engineering Journal, 402: 125881

[49] Sun Z , Zhu Y , Deng Y , Liu F , Ruan W , Xie L , Beadham I . (2022). Nature of surface active centers in activation of peroxydisulfate by CuO for degradation of BPA with non-radical pathway. Colloids and Surfaces. A, Physicochemical and Engineering Aspects, 643: 128731

[50] Wang F , Lai Y , Fang Q , Li Z , Ou P , Wu P , Duan Y , Chen Z , Li S , Zhang Y . (2020). Facile fabricate of novel Co(OH)F@MXenes catalysts and their catalytic activity on bisphenol A by peroxymonosulfate activation: the reaction kinetics and mechanism. Applied Catalysis B: Environmental, 262: 118099

[51] Wang J , Xie Y , Hou J , Zhou X , Chen J , Yao C , Zhang Y , Li Y . (2022a). Biodegradation of bisphenol A by alginate immobilized phanerochaete chrysosporium beads: continuous cyclic treatment and degradation pathway analysis. Biochemical Engineering Journal, 177: 108212

[52] Wang Y , Wu L , Zhou Y , Zhang Y , Sun S , Duo Wu W , Wang X , Wu Z . (2022b). Ternary FeS/γ-Fe₂O₃@N/S-doped carbon nanohybrids dispersed in an ordered mesoporous silica for efficient peroxymonosulfate activation. Chemical Engineering Journal, 435: 135124

[53] Wang Z , Yang B , Zhao X , Chen Y , Wei D , Zhang L , Su X . (2022c). Facile synthesis of ultrathin γ-Fe₂O₃ magnetic nanosheets rich in oxygen vacancies and their photocatalytic activity for water oxidation. Applied Surface Science, 578: 151999

[54] Wu S , Liu H , Yang C , Li X , Lin Y , Yin K , Sun J , Teng Q , Du C , Zhong Y . (2020). High-performance porous carbon catalysts doped by iron and nitrogen for degradation of bisphenol F via peroxymonosulfate activation. Chemical Engineering Journal, 392: 123683

[55] Wu S , Yang Z , Wang F , Jin X , Kengara F , Xi K , Fang W , Yang W , Zhang Y . (2022). Effect of γ-Fe₂O₃ nanoparticles on the composition of montmorillonite and its sorption capacity for pyrene. Science of the Total Environment, 813: 151893

[56] Xie X , Cao J , Xiang Y , Xie R , Suo Z , Ao Z , Yang X , Huang H . (2022). Accelerated iron cycle inducing molecular oxygen activation for deep oxidation of aromatic VOCs in MoS₂ co-catalytic Fe³⁺/PMS system. Applied Catalysis B: Environmental, 309: 121235

[57] Xu B , Wang J , Yu H , Gao H . (2011). Large-scale synthesis of hierarchical flowerlike boehmite architectures. Journal of Environmental Sciences (China), 23(Suppl): S49–S52

[58] Xu L , Wang X , Sun Y , Gong H , Guo M , Zhang X , Meng L , Gan L . (2020). Mechanistic study on the combination of ultrasound and peroxymonosulfate for the decomposition of endocrine disrupting compounds. Ultrasonics Sonochemistry, 60: 104749

[59] Yan T , Yang Q , Feng R , Ren X , Zhao Y , Sun M , Yan L , Wei Q . (2022). Highly effective visible-photocatalytic hydrogen evolution and simultaneous organic pollutant degradation over an urchin-like oxygen-doped MoS₂/ZnIn₂S₄ composite. Frontiers of Environmental Science & Engineering, 16(10): 131

[60] Yang G , Liang Y , Xiong Z , Yang J , Wang K , Zeng Z . (2021). Molten salt-assisted synthesis of Ce₄O₇/Bi₄MoO₉ heterojunction photocatalysts for photo-fenton degradation of Tetracycline: enhanced mechanism, degradation pathway and products toxicity assessment. Chemical Engineering Journal, 425: 130689

[61] Yang J , Zeng D , Hassan M , Ma Z , Dong L , Xie Y , He Y . (2022). Efficient degradation of bisphenol A by dielectric barrier discharge non-thermal plasma: performance, degradation pathways and mechanistic consideration. Chemosphere, 286: 131627

[62] Yao L , He X , Lv J , Xu G , Bao Z , Cui J , Yu D , Wu Y . (2022). Efficient degradation of ciprofloxacin by Co₃O₄/Si nanoarrays heterojunction activated peroxymonosulfate under simulated sunlight: performance and mechanism. Journal of Environmental Chemical Engineering, 10(3): 107397

[63] Yue Y , Shen S , Cheng W , Han G , Wu Q , Jiang J . (2022). Construction of mechanically robust and recyclable photocatalytic hydrogel based on nanocellulose-supported CdS/MoS₂/montmorillonite hybrid for antibiotic degradation. Colloids and Surfaces. A, Physicochemical and Engineering Aspects, 636: 128035

[64] Yun E , Lee J H , Kim J , Park H , Lee J . (2018). Identifying the nonradical mechanism in the peroxymonosulfate activation process: singlet oxygenation versus mediated electron transfer. Environmental Science & Technology, 52(12): 7032–7042

[65] Zeng L , Xiao L , Shi X , Wei M , Cao J , Long Y . (2019a). Core-shell prussian blue analogues@ poly(m-phenylenediamine) as efficient peroxymonosulfate activators for degradation of rhodamine B with reduced metal leaching. Journal of Colloid and Interface Science, 534: 586–594

[66] Zeng Y , Guo N , Song Y , Zhao Y , Li H , Xu X , Qiu J , Yu H . (2018). Fabrication of Z-scheme magnetic MoS₂/CoFe₂O₄ nanocomposites with highly efficient photocatalytic activity. Journal of Colloid and Interface Science, 514: 664–674

[67] Zeng Y , Guo N , Xu X , Yu Y , Wang Q , Wang N , Han X , Yu H . (2019b). Degradation of bisphenol A using peroxymonosulfate activated by WO₃@MoS₂/Ag hollow nanotubes photocatalyst. Chemosphere, 227: 589–597

[68] Zhang D , Fan X , Yang A , Zong X . (2018). Hierarchical assembly of urchin-like alpha-iron oxide hollow microspheres and molybdenum disulphide nanosheets for ethanol gas sensing. Journal of Colloid and Interface Science, 523: 217–225

[69] Zhang H , Du L , Xing J , Wei G , Quan X . (2023). Electro-conductive crosslinked polyaniline/carbon nanotube nanofiltration membrane for electro-enhanced removal of bisphenol A. Frontiers of Environmental Science & Engineering, 17(5): 59

[70] Zhang J , Sun B , Guan X . (2013). Oxidative removal of bisphenol A by permanganate: kinetics, pathways and influences of co-existing chemicals. Separation and Purification Technology, 107: 48–53

[71] Zhang W , Zhang H , Yan X , Zhang M , Luo R , Qi J , Sun X , Shen J , Han W , Wang L . . (2020). Controlled synthesis of bimetallic Prussian blue analogues to activate peroxymonosulfate for efficient bisphenol A degradation. Journal of Hazardous Materials, 387: 121701

[72] Zhao S , Chen C , Ding J , Yang S , Zang Y , Ren N . (2022). One-pot hydrothermal fabrication of BiVO₄/Fe₃O₄/rGO composite photocatalyst for the simulated solar light-driven degradation of Rhodamine B. Frontiers of Environmental Science & Engineering, 16(3): 36

[73] Zheng K , Sun Y , Gong S , Jiang G , Zheng X , Yu Z . (2019). Degradation of sulfamethoxazole in aqueous solution by dielectric barrier discharge plasma combined with Bi₂WO₆-rMoS₂ nanocomposite: mechanism and degradation pathway. Chemosphere, 222: 872–883

[74] Zhou Y , Zhou L , Zhou Y , Xing M , Zhang J . (2020). Z-scheme photo-Fenton system for efficiency synchronous oxidation of organic contaminants and reduction of metal ions. Applied Catalysis B: Environmental, 279: 119365

[75] Zhu J , Wang J , Shan C , Zhang J , Lv L , Pan B . (2019a). Durable activation of peroxymonosulfate mediated by Co-doped mesoporous FePO₄ via charge redistribution for atrazine degradation. Chemical Engineering Journal, 375: 122009

[76] Zhu M , Yang J E , Delai Sun D , Yuan B , Fu M . (2022). Deciphering the simultaneous removal of carbamazepine and metronidazole by monolithic Co₂AlO₄@Al₂O₃ activated peroxymonosulfate. Chemical Engineering Journal, 436: 135201

[77] Zhu S , Li X , Kang J , Duan X , Wang S . (2019b). Persulfate activation on crystallographic manganese oxides: mechanism of singlet oxygen evolution for nonradical selective degradation of aqueous contaminants. Environmental Science & Technology, 53(1): 307–315

RIGHTS & PERMISSIONS

The Author(s) 2024. This article is published with open access at link.springer.com and journal.hep.com.cn