English

• Quinoline, as a typical nitrogenous heterocyclic compound, is widely used in dye, rubber, pharmaceutical, food and chemical industries, and its residue is inevitable [1-3]. However, quinoline has been found to have toxicity, carcinogenicity, mutagenicity and serious odor potential, which makes its environmental pollution nonnegligible [4]. The discharge of quinoline wastes causes great harm to human health and environmental quality. Unfortunately, its double ring fusion structure means serious steric hindrance, which makes quinoline difficult to degrade naturally [5]. The effective degradation of quinoline has attracted more and more attention. Advanced oxidation processes (AOPs) are methods of degrading and mineralizing refractory organic pollutants such as quinoline by producing strong oxidizing species. Varieties of AOPs, including electrochemical oxidation, Fenton, ozone oxidation and photocatalysis, have been widely studied and applied [6]. Among AOPs, Fenton reaction can activate H₂O₂ with Fe²⁺ to generate strong oxidizing ^•OH to decompose and mineralize a variety of organic pollutants. Fenton reaction has been already applied maturely and widely in environmental remediation. Unfortunately, the classic Fenton system is hindered by its difficult recovery of homogeneous catalyst [7], resulting in poor long-term activity. Therefore, the heterogeneous Fenton system has been extensively studied. Fe-based materials are commonly used in heterogeneous Fenton systems, including hematite (α-Fe₂O₃), goethite (α-FeOOH), magnetite (Fe₃O₄), and pyrite (FeS₂). The heterogeneous Fenton system fixes iron in lattice, which solves the problem of difficult catalyst recovery and reduces the formation of iron sludge to a certain extent. However, as the reaction progress, surface ≡Fe(Ⅱ) reacts with H₂O₂ to generate reactive oxygen species (ROSs) and transforms into ≡Fe(Ⅲ), the proportion of ≡Fe(Ⅱ) decreases continuously, resulting in inevitable deactivation of heterogeneous catalyst. Comparing to that of the reaction between Fe(Ⅱ) and H₂O₂ (40–80 L mol⁻¹ s⁻¹), the reaction rate constant (0.001–0.01 L mol⁻¹ s⁻¹) of H₂O₂ reducing Fe(Ⅲ) is relatively smaller, which limits the regeneration of Fe(Ⅱ) [8,9]. Thus, expediting the ≡Fe(Ⅱ)/≡Fe(Ⅲ) cycle is the essential question.

  To overcome the mentioned problems, large numbers of studies have been carried out. It is a common strategy to combine iron-based materials with semiconductor materials and construct a photo-Fenton system to provide exogenous electrons to reduce ≡Fe(Ⅲ). Li et al. doped Fe on g-C₃N₄, promoted ≡Fe(Ⅲ)/≡Fe(Ⅱ) cycle by photogenerated electrons provided by g-C₃N₄, maintaining the catalytic activity during the photo-Fenton reaction [10]. However, the amount of photogenerated electrons depends on the photocatalytic performance of the composite semiconductor, but the light conditions do not always meet its needs. In the aspect of providing exogenous electrons, the electro-Fenton system is more direct. In the electro-Fenton system, external electrons are directly transported from external circuits to iron-based materials [11]. Both photo-Fenton and electro-Fenton systems require additional energy consumption. Adding organic reagents to the reaction system is also studied. Li et al. discovered that antibiotics adsorbed on the surface of Schwertmannite/Bamboo biochar can provide extra supplement electrons to maintain the continuous reduction of ≡Fe(Ⅲ) and produce higher ^•OH yield [12]. This means that the direct reaction between the catalyst and organic reagent can also help the regeneration of ≡Fe(Ⅱ). However, organic reagents were hard to be reused or recovered. Treatment of excessive organic reagents would also consume the oxidation capacity of the system. Adding cocatalysts to the reaction system was also a common strategy. Zhou et al. introduced boron as electron donor into the Fenton system [13,14]. The boron can donate electrons to promote fast Fe^Ⅲ reduction to Fe^Ⅱ and activate persulfate directly in Fenton reaction. Vacancies can capture electrons to introduce additional electrons on the surface of the catalyst and form local electron-rich centers, and these defect electrons are conducive to interfacial charge transfer [15-17]. This strategy helps to construct local charge centers for the heterogeneous Fenton reaction, which is hopeful to solve the problem of catalyst deactivation.

  Here, we come up with a novel Fe-based heterogeneous Fenton catalyst named as FeS_xO_y-X (X is the ratio of ethylene glycol (EG) to N, N-dimethylformamide (DMF)) by simple one-pot solvothermal method. The resulting material demonstrated outstanding degradation activity on varieties of pollutants. Electron paramagnetic resonance (EPR), X-ray photoelectron spectroscopy (XPS) and sacrificial agent experiments indicated that the defect electrons on sulfur vacancy constructed a stable iron cycle on the surface of FeS_xO_y-1:1, resulting in a continuously stable supply of ROSs. Thus, without any organic reagents or cocatalysts, FeS_xO_y-1:1-based Fenton system realized effective long-term degradation of 560 mg/L quinoline within only 7 days. In addition, the practical application potential of the system was comprehensively investigated through quinoline degradation in different water samples and real wastewater treatment.

  Synthesis of catalyst can be summarized as the schematic diagram of Fig. 1a. After a facile solvothermal and impurity removal process, black solid with magnetism was obtained. The details were given in the supporting information. The morphology of the synthesized catalyst was observed to be nano flake by SEM and TEM (Fig. S1 in Supporting information and Fig. 1b). XRD was carried out to identify the synthesized catalyst and analyze their crystal structure (Fig. 1c and Fig. S2 in Supporting information). Compared with the standard PDF card, the synthesized nano flake cannot be identified as any single standard sample. For further identification, the samples were compared with commercial iron oxide and iron sulfide by Raman spectroscopy (Fig. 1d). The results showed that the synthesized samples did not have enough similarity with any single standard samples. Since the spike of the synthesized samples cannot be facilely attributed to any standard, elemental analysis and Solid ICP were applied to confirm its chemical composition (Fig. S3 in Supporting information). In detail, the catalyst contained 37.81% Fe, 32.45% S and 29.74% O. The result confirmed the existence of Fe, S and O on the nano flake. Then, the bonding condition of the material surface was confirmed by Raman spectroscopy and Fourier transform infrared (FT-IR) spectroscopy. Strong peaks appeared at 214.89 cm⁻¹ (A_g), 278.19 cm⁻¹ (E_g) and 387.42 cm⁻¹ (E_g) in Raman spectra can be attributed to Fe−O bond. As for FT-IR (Fig. S4 in Supporting information), the strong absorption peak at 3440 cm⁻¹ indicated the presence of −OH on the catalyst surface. S−O bond (1382 cm⁻¹ and 1085 cm⁻¹), Fe−S (620 cm⁻¹) and S−S bond or metal-S (480 cm⁻¹) were also detected [18]. This showed that Fe, S and O were effectively bonded on the surface of the sample. Then, SAED pattern was obtained from the diffracted electrons when the beam was focused on the nano flakes (Fig. 1e). The discontinuous diffraction ring composed of diffraction points on SAED image indicated that the nano flake was a poly-crystalline consisting of fine single crystals, which also explained the large envelope peak structure in XRD pattern in Fig. 1c. On this basis, the characteristic lattice fringes of FeS_xO_y-1:1 were observed by HRTEM, and the interplanar spacing was 0.353 nm (Fig. 1f). EDX mapping also demonstrated the uniform distribution of Fe, S and O on the nano flake (Fig. 1g). Therefore, we had enough evidence to prove that the synthesized sample was a new type of Fe-based compound, and the synthesized nano flakes were named as FeS_xO_y-X (X: pure EG, pure DMF, or EG/DMF ratio).

  Figure 1

  

  Figure 1.  (a) Schematic illustration showing the preparation of FeS_xO_y-X. (b) TEM image of FeS_xO_y-1:1. (c) XRD pattern of FeS_xO_y-1:1 and Powder Diffraction File card of Fe₂O₃, Fe₃O₄, FeS₂ and FeS. (d) Raman spectra of FeS_xO_y-1:1, Commercial Fe₂O₃, Commercial Fe₃O₄ and Commercial FeS₂. (e) SAED image of FeS_xO_y-1:1. (f) HRTEM image of FeS_xO_y-1:1. (g) Solid EPR spectra of FeS_xO_y-X at room temperature. (h) EDX Mapping of FeS_xO_y-1:1.

  DownLoad: Full-Size Img PowerPoint

  From Raman spectra in Fig. 1d, we also found the existence of defects. Compared to that of standard sample of commercial Fe₂O₃, the Fe−O bond peak position shifted to lower wavenumbers with larger half peak width, donating the existence of more defects in the corresponding catalyst [19]. Given the existence of defects on the surface of FeS_xO_y-1:1, solid electron paramagnetic resonance (EPR) was conducted for further identification of the defects (Fig. 1h). Surprisingly, FeS_xO_y-1:1 showed relatively stronger peak intensity compared to FeS_xO_y-EG and FeS_xO_y-DMF. The g factor was 2.0278, indicating the existence of S vacancy. Moreover, FeS_xO_y-1:1 had the highest Fe proportion (37.81%) and the lowest sulfur ratio (32.45%), indicating that its S vacancy might be the most. Therefore, along with these vacancies, there may be abundant defect electrons on the surface of FeS_xO_y-1:1, which can accelerate electron transfer as a local electron rich center. FeS_xO_y-1:1 may have the highest Fenton activity.

  More information about the activities of FeS_xO_y-X can be obtained by analyzing XPS results. As shown in Fig. 2a, the high resolution Fe 2p XPS spectrum of FeS_xO_y-X can be split into seven peaks. Satellite peaks were found at 732.23 eV and 720.42 eV for FeS_xO_y-1:1, 732.72 eV and 720.21 eV for FeS_xO_y-EG and 731.88 eV and 719.71 eV for FeS_xO_y-DMF, respectively. For FeS_xO_y-1:1, peaks at 726.54 eV and 712.92 eV can be attributed to Fe^Ⅲ 2p_1/2 and Fe^Ⅲ 2p_3/2, respectively, while peaks at 724.28 eV and 710.43 eV can be attributed to Fe^Ⅱ 2p_1/2 and Fe^Ⅱ 2p_3/2, respectively. Peak at 707.47 eV can be attributed to Fe−S bond [20], while binding energy of Fe−S bond were 707.42 eV and 707.01 eV for FeS_xO_y-EG and FeS_xO_y-DMF, respectively. That meant the order of electron cloud density on Fe−S bond was: FeS_xO_y-1:1 < FeS_xO_y-EG < FeS_xO_y-DMF, which suggested that FeS_xO_y-1:1 may have the weakest Fe−S bond among FeS_xO_y-X. High resolution S 2p XPS spectrum of FeS_xO_y-X can be split into five peaks (Fig. 2b). For FeS_xO_y-1:1, the typical peak of S−O bond appeared at 168.55 eV. Peaks at 164.22 eV, 163.51 eV, 162.18 eV and 160.97 eV can be attributed to edge S, sulfur atoms of intermediate oxidation state, lattice S₂²⁻ and surface S²⁻, respectively [18,21]. Peaks at 164.22 eV and 160.97 eV can be typically associated with vacancy or damaged material [22], which also reiterated the existence of sulfur vacancy. The O 1s XPS spectrum (Fig. 2c) can be split into three peaks, corresponding to C−O bond, OH (O_OH) and lattice O (O_L) [19,23]. Elements Proportion on the surface of FeS_xO_y-X was shown in Fig. S5 (Supporting information). In conclusion, both defects and electron bias suggested that 1:1 sample may have the highest Fenton activity.

  Figure 2

  

  Figure 2.  (a) Fe 2p, (b) S 2p and (c) O 1s XPS spectra of FeS_xO_y-X.

  DownLoad: Full-Size Img PowerPoint

  To comprehensively investigate the performance of FeS_xO_y-X in activating oxidant and degrading organic pollutants, a series of degradation experiment were carried out. It was proved in advance that the effect of absorption on FeS_xO_y-X-based Fenton system was negligible (Fig. S6 in Supporting information). The optimal conditions of FeS_xO_y-X-based Fenton system were explored by using the control variable method. First, the optimal dosage of H₂O₂ was investigated (Fig. S7 in Supporting information). When the concentration of H₂O₂ increased from 0.49 mmol/L to 1.47 mmol/L, the degradation rate of quinoline increased from 58.79% to 98.75%, while dosage of H₂O₂ exceeded 1.47 mmol/L, degradation efficiency no longer increased. Taking cost of H₂O₂ into consideration, 1.47 mmol/L was determined to be the optimal dosage of H₂O₂. Second, the optimal dosage of catalyst was discussed (Fig. S8 in Supporting information). When the dosage of FeS_xO_y-1:1 increased from 3 mg to 10 mg, the degradation rate rose from 64.17% to 98.75%. However, while the dosage of FeS_xO_y-1:1 further rose to 15 mg, the degradation efficiency showed no increase. For the sake of cost of catalyst, it was appropriate to determine 10 mg as the optimal dosage of FeS_xO_y-1:1. Finally, the initial pH of FeS_xO_y-1:1-based Fenton system was studied (Fig. S9 in Supporting information). FeS_xO_y-1:1-based Fenton system has better degradation activity under acidic environment. While pH_i increased from 4 to 8, activity of FeS_xO_y-1:1 decreased. It is worth noting that when pH_i decreased from 4 to 2 degradation activity of FeS_xO_y-1:1-based Fenton system showed no attenuation but increased. Typically in classic Fenton system, when pH is too low, especially pH is lower than 3, exorbitant H⁺ will inhibit the formation of FeOOH²⁺, which consecutively hinders the production of Fe²⁺ and hydroxyl radicals [24]. FeS_xO_y-1:1-based Fenton system showed the ability adapting to strong acidic environment, indicating that the activity of the system may be dominated by Fe(Ⅱ) on the surface rather than free Fe²⁺ in the solution. The pH_i = 4 was chose as the optimal pH for further study as pH_i = 4 was just sufficient for complete quinoline degradation in FeS_xO_y-1:1-based Fenton system. Briefly speaking, the optimum reaction conditions in FeS_xO_y-1:1-based Fenton system can be concluded as follows: for 100 mL water containing 20 mg/L quinoline, 10 mg FeS_xO_y-1:1 and 1.47 mmol/L H₂O₂ was added with initial pH ≈ 4.

  Quinoline degradation experiments by different catalysts were conducted under optimal conditions (Fig. 3a). Degradation rate of FeS_xO_y-1:1, FeS_xO_y-EG and FeS_xO_y-DMF within 25 min were 98.75%, 87.42% and 41.41%, respectively. In order to further compare of activity of FeS_xO_y-X, pseudo first order dynamic fitting and calculation of observed rate constant (k_obs) were conducted (Fig. 3b). FeS_xO_y-X-based Fenton system accorded with pseudo-first order kinetic equation. The observed rate constant was 0.1739 min⁻¹ for FeS_xO_y-1:1, 0.08769 min⁻¹ for FeS_xO_y-EG and 0.02023 min⁻¹ for FeS_xO_y-DMF. Notably, observed rate constant of FeS_xO_y-1:1 is almost twice that of FeS_xO_y-EG and eight times that of FeS_xO_y-DMF. In the later stage of the reaction, the reduction of the degradation rate gap is actually subject to the low substrate concentration.

  Figure 3

  

  Figure 3.  Degradation behavior of the FeS_xO_y-X-based Fenton system (pH_i ≈ 4.0, [Quinoline] = 20 mg/L, [Catalyst] = 10 mg, [H₂O₂] = 1.47 mmol/L): (a) Quinoline degradation by different catalysts. (b) Observed rate constant of FeS_xO_y-1:1, FeS_xO_y-EG and FeS_xO_y-DMF. (c) Hydrogen dioxide consumption of FeS_xO_y-1:1, FeS_xO_y-EG and FeS_xO_y-DMF. (d) Degradation rate of nitro compounds, phenolics, antibiotics and dyes of FeS_xO_y-1:1 ([pollutant] = 20 mg/L). (e) Inhibition effect of radical scavengers on Quinoline degradation in the FeS_xO_y-1:1-based Fenton system ([t-BA] = 50 µL, [BQ] = 2.31 mmol/L, [TEMP] = 0.462 mmol/L). EPR spectra of (f) DMPO-^•OH, (g) DMPO-^•O₂⁻ and (h) TEMP-¹O₂ from the FeS_xO_y-1:1-based Fenton system.

  DownLoad: Full-Size Img PowerPoint

  FeS_xO_y-1:1-based Fenton system has shown the highest degradation efficiency among FeS_xO_y-X-based Fenton system, but the reason for its advantage remains to be clarified. Hydrogen peroxide consumption in FeS_xO_y-X-based Fenton systems were compared in Fig. 3c. In detail, FeS_xO_y-1:1 has the highest H₂O₂ activating efficiency, while FeS_xO_y-EG and FeS_xO_y-DMF activate H₂O₂ much slower. Fig. S10 (Supporting information) displayed the working curve of determination of hydrogen peroxide concentration by iodometry. The curve had good linearity, and the specific test method was given in the supporting information. Interesting conclusion can be drawn by comparing results of Figs. 3a and c. When the reaction proceeded to the twentieth minute in FeS_xO_y-EG-based Fenton system, the hydrogen peroxide consumption was almost the same as that at the fifth minute in FeS_xO_y-1:1-based Fenton system. At the above point of time, the degradation rate of quinoline was also almost the same. Coincidentally, the same situation can be observed at the twenty-fifth minute of FeS_xO_y-EG-based Fenton system and the tenth minute of FeS_xO_y-1:1-based Fenton system. That means, quinoline degradation rate may keep the same whenever the same amount of H₂O₂ was consumed in FeS_xO_y-X-based Fenton system. Thus, the degradation mechanisms of the three systems were the same. It was the stronger ability to decompose H₂O₂ that accounted for the advantage of FeS_xO_y-1:1 over FeS_xO_y-EG and FeS_xO_y-DMF. This may be attributed to the abundant defect electrons on the surface of FeS_xO_y-1:1, which is conducive to interfacial charge transfer [15].

  Since FeS_xO_y-1:1 was a novel Fe-based Fenton catalyst containing both sulfur and oxygen, it is necessary to compare it with common Fe-based Fenton catalysts. A special FeS_xO_y-1:1 sample was prepared without sulfur source. Common iron sulfides and the special sample were also compared with FeS_xO_y-1:1 under optimum reaction condition (Fig. 3a and Fig. S11 in Supporting information). The quinoline degradation activity of commercial FeS₂ and FeS were 9.36% and 31.32%, respectively. While the FeS_xO_y-1:1 sample prepared without sulfur source exhibited a degradation rate of 12.19%. These samples, either oxygen-free or sulfur-free, did not show activity comparable to FeS_xO_y-1:1. FeS_xO_y-1:1-based Fenton system was also compared to other heterogeneous Fenton systems reported recently in Table S1 (Supporting information). In comparison, FeS_xO_y-1:1-based Fenton system obtained stronger oxidation capacity with less hydrogen peroxide demand and catalyst dosage.

  For assessment of the general applicability of the oxidation system, different environmental pollutants, such as nitro compounds (nitro benzene (NB), p-chloronitrobenzene (p-CNB) and p-nitrophenol (PNP)), phenols (phenol and tetrachlorophenol (4-CP)), antibiotics (enrofloxacin (ENR), norfloxacin (NOR) and sulfadiazine (SD)) and dyes (rhodamine B (RhB), methylene blue (MB), tartrazine (TA), malachite green (MG) and orange Ⅱ), were selected as the target pollutants to evaluate the activity of the FeS_xO_y-1:1-based Fenton system. As shown in Fig. 3d, the FeS_xO_y-1:1-based Fenton system manifested excellent degradation activity for all these organic pollutants, suggesting that the main active species of the system have broad-spectrum oxidation.

  ROS quenching experiments were carried out to verify main radical species of the oxidation system (Fig. 3e). The selection and dosage of quencher are referred to the literature [25]. Typically, quenching agents were added before any catalyst or oxidant under optimum reaction conditions. When tert-butanol (t-BA, a sacrificial agent of ^•OH, 50 µL), p-benzoquinone (p-BQ, a sacrificial agent of ^•O₂⁻, 2.31 mmol/L) and 2, 2, 6, 6-tetramethylplperldine (TEMP, a sacrificial agent of ¹O₂, 0.462 mmol/L) was added respectively, the degradation rates of quinoline in FeS_xO_y-1:1-based Fenton system were greatly reduced. All these oxidizing species might be involved in the FeS_xO_y-1:1-based Fenton system. EPR was carried out for further verification of radical species in FeS_xO_y-1:1-based Fenton system (Figs. 3f–h). As expected, signal of DMPO-^•OH of FeS_xO_y-1:1 is stronger than that of FeS_xO_y-EG and FeS_xO_y-DMF. The result is consistent with degradation experiment (Fig. 3f). However, no significant signal of DMPO-^•O₂⁻ was detected. In the HPLC chromatogram of active oxygen quenching experiment determined by BQ (Fig. S12 in Supporting information), as the reaction progress, signal peak intensity of BQ also evidently decreased. Both BQ and quinoline were degraded by FeS_xO_y-1:1-based Fenton system. Therefore, it was the competing reaction between BQ and quinoline, rather than the specific capture effect of BQ on ^•O₂⁻ that hindered the quinoline degradation. The typical triplet peak in Fig. 3h indicated the existence of TEMP-¹O₂. FeS_xO_y-1:1 had the strongest signal of TEMP-¹O₂ among FeS_xO_y-X. High-valent iron is a common oxidative species in Fenton reaction. Phenyl methyl sulfoxide (PMSO) was used as a reactant to figure out the contribution of high-valent iron in the FeS_xO_y-1:1-based Fenton system (Fig. S13 in Supporting information). As the reaction proceeded, the concentration of PMSO decreased continuously, while the concentration of PMSO₂ remained at an extremely low level. The absence of high-valent iron was abundantly clear. Accordingly, the FeS_xO_y-1:1-based Fenton system was dominated by the activation of H₂O₂ for the production of ^•OH and ¹O₂.

  Leaching of ions is an important feature of heterogeneous Fenton catalysts. Concentration of Fe²⁺ was determined by a typical chromogenic method with 1, 10-phenanthroline as chromogenic agent. Fig. S14 (Supporting information) displayed the working curve of Fe²⁺ concentration detection by 1, 10-phenanthroline colorimetric method. The curve had good linearity, and the specific test method was given in the supporting information. As shown in Fig. 4a, both total iron ion and ferrous ion decreased rapidly at the beginning of oxidation reaction and reached dynamic equilibrium. Total iron concentration was below 3.43 mg/L, while ferrous ion concentration was below 0.39 mg/L. Then, overdosed 1, 10-phenanthroline was introduced into the FeS_xO_y-1:1-based Fenton system to completely capture leached Fe²⁺ (Fig. 4b), resulting in a decreased degradation rate of 67.79%. Quinoline degradation was also carried out by applying mean value leached Fe²⁺, and the degradation rate of was only 36.21% (Fig. S15 in Supporting information). The homogeneous process contributed to the reaction to a certain extent, but it was not the dominant factor. Therefore, the excellent activity of the system was not dominated by the rapid dissolution of Fe²⁺, suggesting the significant role of heterogeneous process in the system.

  Figure 4

  

  Figure 4.  (a) Leaching of Fe²⁺ and total iron ions in the FeS_xO_y-1:1-based Fenton reaction. (b) Inhibition effect of AgNO₃, 1, 10-phenanthroline and N₂ on Quinoline degradation in the FeS_xO_y-1:1-based Fenton system ([AgNO₃] = 1 mmol/L, [o-Phenanthroline] = 0.1 mmol/L, [Quinoline] = 20 mg/L), [Catalyst] = 10 mg, [H₂O₂] = 1.47 mmol/L, pH_i ≈ 4). (c) Fe 2p spectra of fresh and used FeS_xO_y-1:1. (d) Fe 2p spectra of fresh and used FeS_xO_y-1:1-S. (e) Divalent iron loss rate on the surface of catalysts. (f) Inhibition effect of different anions on Quinoline degradation in the FeS_xO_y-1:1-based Fenton system. ([Anions] = 10 mg/L, [Quinoline] = 20 mg/L), [Catalyst] = 10 mg, [H₂O₂] = 1.47 mmol/L). (g) Long term degradation of Quinoline by FeS, SV-FeS₂ (SV: Sulfur vacancy) and FeS_xO_y-1:1, respectively. (h) Quinoline degradation in natural water in the FeS_xO_y-1:1-based Fenton system. (i) Wastewater treatment by the FeS_xO_y-1:1-based Fenton system.

  DownLoad: Full-Size Img PowerPoint

  Given the existence of sulfur vacancy, experiments and characterization were carried out to clarify its role in FeS_xO_y-1:1-based Fenton system. Under nitrogen atmosphere, the catalyst was treated with sulfur powder to remove the sulfur vacancies on its surface. As shown in Fig. S16 (Supporting information), the XRD peak position of FeS_xO_y-1:1 remained basically unchanged compared with that before the treatment. The annealing treatment did not change the crystal form of the catalyst. As shown in Fig. S17 (Supporting information), after annealing, the defect concentration on the catalyst surface decreased significantly. FeS_xO_y-1:1 with less surface defects was successfully prepared labeled as FeS_xO_y-1:1-S. Compared to that of FeS_xO_y-1:1, quinoline degradation rate of FeS_xO_y-1:1-S significantly decreased with its observed rate constant reduced from 0.17385 min⁻¹ to 0.0568 min⁻¹ (Figs. S18a and b in Supporting information). The great decay in activity implies the important role of sulfur vacancies in the system. In order to understand the change on the surface of the samples, XPS characterization was conducted on fresh and used FeS_xO_y-1:1 and FeS_xO_y-1:1-S (Figs. 4c and d). For used FeS_xO_y-1:1, Fe 2p XPS peaks at 726.48 eV and 712.81 eV can be attributed to Fe^Ⅲ 2p_1/2 and Fe^Ⅲ 2p_3/2, respectively, while peaks at 723.96 eV and 710.38 eV can be attributed to Fe^Ⅱ 2p_1/2 and Fe^Ⅱ 2p_3/2, respectively. It was worth noting that, compared with that of fresh sample, the Fe 2p peak of used FeS_xO_y-1:1 shifted toward lower binding energy, while the O 1s XPS peak and S 2p XPS peak shifted toward higher binding energy (Figs. S19 and S20 in Supporting information). Thus, Fe obtained electrons and was reduced during the Fenton reaction, which was conducive to the circulation of ≡Fe(Ⅱ) on the catalyst surface. For contrast, FeS_xO_y-1:1 after a H₂O₂ free reaction was recovered and characterized (Fig. S21 in Supporting information). ≡Fe(Ⅱ) loss rate on the surface of the catalyst was calculated by the results of XPS (Fig. 4e). The calculation formula is as follows (Eq. 1):

  (1)

  After Fenton reaction, the Fe(Ⅱ) on the surface of FeS_xO_y-1:1 declined only by 2.48%. When hydrogen peroxide was removed from the system, ≡Fe(Ⅱ) loss rate increased to 5.64%, while FeS_xO_y-1:1-S lost 14.49% ≡Fe(Ⅱ) with the absence of S vacancy. Therefore, both H₂O₂ and S vacancy were involved in the regeneration of ≡Fe(Ⅱ) on the surface of FeS_xO_y-1:1. FeS_xO_y-1:1 might have the potential of effective long term degradation. The used FeS_xO_y-1:1-S still retained the peak of Fe−S bond, which may be caused by the formation of more Fe−S bonds due to the filling of sulfur vacancies by sulfur atoms during the annealing treatment, and these Fe−S bonds are retained in the Fenton reaction. The S 2p spectrum of FeS_xO_y-1:1-S was shown in Fig. S22a (Supporting information). Peaks at 164.00 eV and 160.91 eV can be typically associated with vacancy or damaged material [22]. The peak intensity of both two peaks significantly declined after annealing, demonstrating the reduction of Sulfur vacancy on the surface of the catalyst. The S−O bond peak originally located at 185.55 eV splited into two peaks, which was caused by the reduction of sulfur vapor. The O 1s spectrum of FeS_xO_y-1:1-S was shown in Fig. S22b (Supporting information). Due to the reaction with sulfur powder during annealing treatment, the ratio of lattice O decreased. After the reaction, the lattice O was re exposed, which reaffirmed that the crystal structure of the catalyst was not damaged during the annealing treatment. The S 2p spectrum of FeS_xO_y-1:1 after a H₂O₂ free reaction was shown in Fig. S23a (Supporting information). The typical peak of S−O bond appeared at 168.51 eV. Peaks at 164.89 eV, 163.87 eV, 163.31 eV and 161.98 eV can be attributed to edge S, sulfur atoms of intermediate oxidation state, lattice S₂²⁻ and surface S²⁻, respectively. The O 1s spectrum of FeS_xO_y-1:1-S was shown in Fig. S23b (Supporting information).

  AgNO₃ was an inorganic electron trapping agent, which would not compete with pollutants for ROS and interfere with Fenton reaction. Under optimal conditions, 1 mmol/L AgNO₃ was added to capture the defect electron of FeS_xO_y-1:1, resulting in the degradation rate to be inhibited to 35.61%, indicating the significant role of defect electrons in the Fenton reaction (Fig. 4b). Therefore, S vacancy was indirectly involved in the regeneration of ≡Fe(Ⅱ) in FeS_xO_y-1:1-based Fenton system.

  For further investigation of mechanism of FeS_xO_y-1:1-based Fenton system, other sacrificial agent was added (Fig. 4b). First, N₂ was bubbled into the reaction system under optimum conditions to remove dissolved oxygen. O₂ was not involved in the reaction as the activity of FeS_xO_y-1:1 based Fenton system exhibited no obviously difference before and after bubbled N₂. This meant, the ¹O₂ in FeS_xO_y-1:1-based Fenton system came from the conversion of H₂O₂ rather than the molecular oxygen pathway [26]. Experiments were conducted to explore the influence of S vacancy on ROSs generation. As it is proved that ^•O₂⁻ was nonexistent in FeS_xO_y-1:1 based Fenton system, quantitative tests were only performed for ^•OH and ¹O₂ (Figs. S24–S26 in Supporting information). The specific calculation method was given in the Supporting Information. After removing the sulfur vacancy, the ^•OH and ¹O₂ concentration decreased to 53.25% and 35.87% of the initial level, respectively. Sulfur vacancies had a greater influence on the generation of ¹O₂, and the relationship between S vacancies and ¹O₂ may be more direct than that of ^•OH. Data of Zeta potential at different pH was carried out to clarify the surface charge of FeS_xO_y-1:1 (Fig. S27 in Supporting information). Mutation of zeta potential occurred at pH 6, with a sharp decrease in degradation rate (Fig. S9). As pH of bulk fluid rise, the positive charge decreased, less H⁺ were involved in the stable layer. Meanwhile, quinoline was a typical lewis base, mutation of zeta potential made it harder for quinoline to reach the interface of stable layer and sliding plane.

  Combining all the above investigations and analyses, the mechanism of FeS_xO_y-1:1-based Fenton system can be summarized. The degradation of organics by FeS_xO_y-1:1 was primarily conducted by activating H₂O₂ to generate strong oxidizing ^•OH and ¹O₂. H₂O₂ diffused from the bulk fluid to the stable layer and was decomposed by ≡Fe(Ⅱ) on the surface of FeS_xO_y-1:1 to produce ^•OH (Eq. 2). S vacancy seized one electron of hydrogen peroxide to generate HO₂^• (Eq. 3). Then, ≡Fe(Ⅲ) obtain one electron from HO₂^• and produce ¹O₂ (Eq. 4) [27]. Meanwhile, ≡Fe(Ⅱ) can leach to produce Fe²⁺, which diffused into the bulk fluid to decompose H₂O₂ and generate ^•OH (Eqs. 5 and 6). Abundant defect electrons participated in the ≡Fe(Ⅱ)/≡Fe(Ⅲ) cycle on the surface of FeS_xO_y-1:1: ≡Fe(Ⅲ) on the surface of FeS_xO_y-1:1 obtain an electron to regenerate ≡Fe(Ⅱ) (Eq. 7). This spontaneous iron circulation behavior maintained timely regeneration of ≡Fe(Ⅱ) on the surface of FeS_xO_y-1:1, which indicated that FeS_xO_y-1:1-based Fenton system may have the potential for promising long-term effective degradation.

  (2)

  (3)

  (4)

  (5)

  (6)

  (7)

  The intermediates and structures measured by HPLC-MS are presented in Table S2 (Supporting information). Besides quinoline, intermediates (e.g., 8-hydroxyquinolin-2(1H)-one, (Z)-3-(2-amino-3-hydroxyphenyl)acrylic acid, 2-amino-3-hydroxybenzoic acid, (+)-isopropanolamine, 2-hydroxyquinolin and 1-aminoethan-1-ol) were detected. Common characteristic of intermediates can be concluded as the preferential attack of ^•OH on nitrogen-containing ring in quinoline degradation: ^•OH might attack the nitrogen ring and the benzene ring in sequence in the process. For quinoline, the high electron density on nitrogen-containing ring set the condition for proneness of electrophilic addition reactions [28]. As a powerful electrophilic group, it was reasonable for ^•OH to attack the nitrogen-containing ring first. Singlet oxygen was mainly involved in the further mineralization of intermediate products. Probable degradation pathway of quinoline was given in Fig. S28 (Supporting information).

  Although the important role of S vacancy and defect electrons in FeS_xO_y-1:1-based Fenton system has been proved, whether the activity enhancement brought by this strategy can be applied to the activation of other oxidants remains to be discussed. Thus, degradation experiment was also carried out to investigate the ability of FeS_xO_y-1:1 to activate other oxidants (CaO₂ and peroxymonosulfate (PMS)). In detail, controlling other conditions to be the same, the same equivalent of different oxidants were added (Fig. S29 in Supporting information). FeS_xO_y-1:1 had the ability to activate all these oxidants with outstanding quinoline degradation. Considering that CaO₂ was once considered as a substitute for H₂O₂, the FeS_xO_y-1:1-CaO₂ system was explored in detail. The optimal conditions of FeS_xO_y-1:1-CaO₂ system were studied (Figs. S30–S32 in Supporting information). Briefly speaking, in FeS_xO_y-1:1-CaO₂ system, optimum reaction conditions were as follows: to 100 mL deionized water containing 20 mg/L quinoline, 10 mg catalyst and 10 mg CaO₂ was added with initial pH ≈ 3.28. In systems activating CaO₂, FeS_xO_y-1:1 still showed the highest activity (Fig. S33 in Supporting information). Compared with FeS_xO_y-1:1-based Fenton system, FeS_xO_y-1:1-CaO₂ system has higher requirements for acidic environment. This may be caused by the decomposition of calcium peroxide, which would consume a certain amount of H⁺. Moreover, high dosage of CaO₂ severely inhibited the reaction because excessive CaO₂, as a solid base, affected the pH value by releasing a large amount of OH⁻. ROSs quenching experiments were also carried out in FeS_xO_y-1:1-CaO₂ system. As expected, addition of t-BA, BQ and TEMP significantly inhibited the reaction (Fig. S34 in Supporting information). EPR was carried out for further verification of radical species in the FeS_xO_y-1:1-CaO₂ system (Figs. S35–S37 in Supporting information). Obvious signal of DMPO-^•OH was detected, while signal of TEMP-¹O₂ was relatively weak. Signal of DMPO-^•O₂⁻ was also not detected in FeS_xO_y-1:1-CaO₂ system, denoting that the inhibition of BQ may be also caused by competing reaction between BQ and quinoline. Accordingly, the dominant main active species in the FeS_xO_y-1:1-CaO₂ system were ^•OH and ¹O₂. Therefore, conclusions similar to that of FeS_xO_y-1:1-based Fenton system was obtained in the FeS_xO_y-1:1-CaO₂ system. The strategy of manufacturing defects by adjusting the solvent to obtain better Fenton activity had certain universality.

  Inorganic anions ubiquitous in environmental waters, including Cl⁻, NO₃⁻, SO₄²⁻, H₂PO₄⁻ and HCO₃⁻, were introduced to test the anti-interference ability of FeS_xO_y-1:1-based Fenton system (Fig. 4f). Among them, NO₃⁻ and SO₄²⁻ exerted a negligible effect on quinoline degradation. The addition of Cl⁻ caused slight inhibition in FeS_xO_y-1:1-based Fenton system, which can be attributed to the reaction between ^•OH and Cl⁻ to consume ^•OH and generate ^•Cl (Eq. 8). The oxidizability of ^•Cl was obviously weaker than that of ^•OH: oxidation potential of ^•OH and ^•Cl were 2.8 eV and 2.47 eV in Fenton/Cl⁻ system, respectively [28,29].

  (8)

  However, the addition of HCO₃⁻ severely inhibited quinoline degradation because the reaction solution was adjusted to a weakly alkaline environment (pH 8.31). Meanwhile, H₂PO₄⁻ also showed inhibition on quinoline degradation because of the formation of ferric phosphate complex (logK_Fe(Ⅲ)-H2PO₄− = ~3.5) [30]. The complexation may make it more difficult for trivalent iron to participate in chemical reactions including the regeneration of ≡Fe(Ⅱ).

  Given the potential for long-term degradation, FeS_xO_y-1:1 was employed to degrade quinoline for up to 7 days, and recently reported catalysts (FeS and SV-FeS₂) were used as comparison [31,32]. In detail, whenever a certain reaction time (30 min for the first two days and 60 min for the rest several days) has passed, the same amount of 20 mg/L quinoline and 1.47 mmol/L H₂O₂ were added to each system to continue the reaction. No reducing reagent was involved as electron sacrificier in any group [33,34]. Shown as Fig. 4g, within 7 days, a total of 28 times of 20 mg/L quinoline (total: 560 mg/L) concentrate was added. Among them, FeS_xO_y-1:1-based Fenton system had always maintained a significantly high reactivity with the degradation rate over 95% for almost each turn in a certain period of time, maintaining the lowest quinoline concentration. In contrast, SV-FeS₂ system always maintained certain concentration of residual quinoline. FeS_xO_y-1:1-based Fenton system exhibited obviously better long-term degradation ability than SV-FeS₂ system. As for FeS system, its degradation activity was much lower than the previous two systems, resulting in the accumulation and increase of the concentration of quinoline. Especially, FeS_xO_y-1:1-based Fenton system has a total organic carbon (TOC) removal rate of 56.72%.

  The investigation of practical application was carried out through the degradation of quinoline in natural water and the treatment of practical wastewater. Water from representative rivers in China (Pearl River and Yangtze River), local river (Huangpu River) and small river (Bailang River) was involved at optimum reaction conditions (Fig. 4h). No obvious inhibition was observed in all natural water bodies, indicating good adaptability to different water bodies. Treatment of actual wastewater (1: biochemical wastewater from industrial park, 2: printing and dyeing wastewater and 3: harbor wastewater) by FeS_xO_y-1:1-based Fenton system was carried out (Fig. 4i). In detail, 30 mg catalyst and 300 µL H₂O₂ was added into 100 mL wastewater after the initial pH was adjusted to 4.0 by 4.6 mol/L H₂SO₄. FeS_xO_y-1:1-based Fenton system achieved effective degradation on both kinds of wastewater, notably, COD removal rate of biochemical wastewater reached 79.39%.

  Heterogeneous Fenton reaction has been mature and widely studied in environmental remediation. However, the rapid deactivation of the catalyst greatly limits its potential for practical application of long-term treatment. In this work, a novel Fe-based catalyst named as FeS_xO_y-X, which can activate multiple oxidants, was prepared through a simple one pot solvothermal method. Specially, FeS_xO_y-1:1-based Fenton system exhibited relatively high degradation efficiency on various organic pollutants. Compared to traditional heterogeneous Fenton catalysts, FeS_xO_y-1:1 accelerated ≡Fe(Ⅱ) regeneration with the help of its abundant defect electrons in sulfur vacancies, maintaining a stable divalent iron ratio with continuous generation of ^•OH and ¹O₂. In long-term degradation, up to 560 mg/L quinoline was decomposed within 7 days. FeS_xO_y-1:1-based Fenton system had no dependence on water quality with effective wastewater degradation. Thus, as an efficient and universal heterogeneous Fenton catalyst, FeS_xO_y-1:1 has provided a new choice for the long-term treatment for refractory organic pollutant.

  Declaration of competing interest

  The authors declare no competing financial interest.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 22176060), Project supported by Shanghai Municipal Science and Technology Major Project (No. 2018SHZDZX03), the Program of Introducing Talents of Discipline to Universities (No. B16017), and the Science and Technology Commission of Shanghai Municipality (No. 20DZ2250400). Authors thank Research Center of Analysis and Test of East China University of Science and Technology for the help on the characterization.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.108109.

  
  
• 1. [1]

     Q. Bai, L. Yang, R. Li, et al., Environ. Sci. Technol. 49 (2015) 11536–11542. doi: 10.1021/acs.est.5b03293

  2. [2]

     A. Mermer, T. Keles, Y. Sirin, Bioorg. Chem. 114 (2021) 105076. doi: 10.1016/j.bioorg.2021.105076

  3. [3]

     D. Hu, C. Pi, W. Hu, et al., Chin. Chem. Lett. 33 (2022) 4064–4068. doi: 10.1016/j.cclet.2021.12.072

  4. [4]

     K.V. Padoley, S.N. Mudliar, R.A. Pandey, Bioresour. Technol. 99 (2008) 4029–4043. doi: 10.1016/j.biortech.2007.01.047

  5. [5]

     S. Zhu, X. Yang, W. Yang, et al., Int. J. Environ. Res. Public Health 9 (2012) 548–563. doi: 10.3390/ijerph9020548

  6. [6]

     H. Zheng, Y. Hou, S. Li, et al., Chin. Chem. Lett. 33 (2022) 5103-5022.

  7. [7]

     A.N. Soon, B.H. Hameed, Desalination 269 (2011) 1–16. doi: 10.1016/j.desal.2010.11.002

  8. [8]

     C. Lai, X. Shi, L. Li, et al., Sci. Total Environ. 775 (2021) 145850. doi: 10.1016/j.scitotenv.2021.145850

  9. [9]

     N. Thomas, D.D. Dionysiou, S.C. Pillai, J. Hazard. Mater. 404 (2021) 124082. doi: 10.1016/j.jhazmat.2020.124082

  10. [10]

      K. Li, Y. Liang, H. Yang, et al., Catal. Today 371 (2021) 58–63. doi: 10.1016/j.cattod.2020.07.026

  11. [11]

      W.P. Ting, M.C. Lu, Y.H. Huang, J. Hazard. Mater. 156 (2008) 421–427. doi: 10.1016/j.jhazmat.2007.12.031

  12. [12]

      T. Li, P. Zhu, D. Wang, Z. Zhang, L. Zhou, Environ. Res. 209 (2022) 112830. doi: 10.1016/j.envres.2022.112830

  13. [13]

      P. Zhou, W. Ren, G. Nie, et al., Angew. Chem. Int. Ed. 59 (2020) 16517–16526. doi: 10.1002/anie.202007046

  14. [14]

      W. Ren, P. Zhou, G. Nie, et al., Water Res. 186 (2020) 116361. doi: 10.1016/j.watres.2020.116361

  15. [15]

      Y. Deng, P. Gao, L. Wang, et al., J. Environ. Chem. Eng. 10 (2022) 107481. doi: 10.1016/j.jece.2022.107481

  16. [16]

      G. Zhang, X. Yuan, B. Xie, et al., Chem. Eng. J. 433 (2022) 133670. doi: 10.1016/j.cej.2021.133670

  17. [17]

      L. Cai, J. He, Q. Liu, et al., J. Am. Chem. Soc. 137 (2015) 2622–2627. doi: 10.1021/ja5120908

  18. [18]

      Q. Yan, C. Lian, K. Huang, et al., Angew. Chem. Int. Ed. 60 (2021) 17155–17163. doi: 10.1002/anie.202105736

  19. [19]

      W. Xu, W. Xue, H. Huang, et al., Appl. Catal. B: Environ. 291 (2021) 120129. doi: 10.1016/j.apcatb.2021.120129

  20. [20]

      M.V. Morales-Gallardo, A.M. Ayala, M. Pal, et al., Chem. Phys. Lett. 660 (2016) 93–98. doi: 10.1016/j.cplett.2016.07.046

  21. [21]

      F. Yu, Y. Wang, H. Ma, M. Zhou, Sep. Purif. Technol. 248 (2020) 117022. doi: 10.1016/j.seppur.2020.117022

  22. [22]

      R. Canton-Vitoria, Y. Sayed-Ahmad-Baraza, M. Pelaez-Fernandez, et al., npj 2D Mater. Appl. 1 (2017) 13. doi: 10.1038/s41699-017-0012-8

  23. [23]

      M. Xiao, R. Li, J. Yin, et al., Colloids Surf. A Physicochem. Eng. Asp. 651 (2022) 129678. doi: 10.1016/j.colsurfa.2022.129678

  24. [24]

      E. Neyens, J. Baeyens, J. Hazard. Mater. 98 (2003) 33–50. doi: 10.1016/S0304-3894(02)00282-0

  25. [25]

      Q. Yi, J. Ji, B. Shen, et al., Environ. Sci. Technol. 53 (2019) 9725–9733. doi: 10.1021/acs.est.9b01676

  26. [26]

      Y. Yang, L. Xu, J. Wang, Chem. Eng. J. 425 (2021) 131497. doi: 10.1016/j.cej.2021.131497

  27. [27]

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

  28. [28]

      Z. Jiao, X. Zhang, H. Gong, et al., J. Ind. Eng. Chem. 105 (2022) 49–57. doi: 10.1016/j.jiec.2021.10.006

  29. [29]

      X. Lai, X.A. Ning, Y. Zhang, et al., Environ. Res. 197 (2021) 110997. doi: 10.1016/j.envres.2021.110997

  30. [30]

      R.C.P. Richard, B. Wilhelmy, E. Maijevic, Inorg. Chem. 24 (1985) 3290–3297. doi: 10.1021/ic00214a039

  31. [31]

      D. Cheng, A. Neumann, S. Yuan, W. Liao, A. Qian, Environ. Sci. Technol. 54 (2020) 4091–4101. doi: 10.1021/acs.est.9b07012

  32. [32]

      C.C. Ling, X.F. Liu, M.Q. Li, et al., Appl. Catal. B: Environ. 290 (2021) 120051. doi: 10.1016/j.apcatb.2021.120051

  33. [33]

      J. Huang, A. Jones, T.D. Waite, et al., Chem. Rev. 121 (2021) 8161–8233. doi: 10.1021/acs.chemrev.0c01286

  34. [34]

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

• 

  Figure 1  (a) Schematic illustration showing the preparation of FeS_xO_y-X. (b) TEM image of FeS_xO_y-1:1. (c) XRD pattern of FeS_xO_y-1:1 and Powder Diffraction File card of Fe₂O₃, Fe₃O₄, FeS₂ and FeS. (d) Raman spectra of FeS_xO_y-1:1, Commercial Fe₂O₃, Commercial Fe₃O₄ and Commercial FeS₂. (e) SAED image of FeS_xO_y-1:1. (f) HRTEM image of FeS_xO_y-1:1. (g) Solid EPR spectra of FeS_xO_y-X at room temperature. (h) EDX Mapping of FeS_xO_y-1:1.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) Fe 2p, (b) S 2p and (c) O 1s XPS spectra of FeS_xO_y-X.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Degradation behavior of the FeS_xO_y-X-based Fenton system (pH_i ≈ 4.0, [Quinoline] = 20 mg/L, [Catalyst] = 10 mg, [H₂O₂] = 1.47 mmol/L): (a) Quinoline degradation by different catalysts. (b) Observed rate constant of FeS_xO_y-1:1, FeS_xO_y-EG and FeS_xO_y-DMF. (c) Hydrogen dioxide consumption of FeS_xO_y-1:1, FeS_xO_y-EG and FeS_xO_y-DMF. (d) Degradation rate of nitro compounds, phenolics, antibiotics and dyes of FeS_xO_y-1:1 ([pollutant] = 20 mg/L). (e) Inhibition effect of radical scavengers on Quinoline degradation in the FeS_xO_y-1:1-based Fenton system ([t-BA] = 50 µL, [BQ] = 2.31 mmol/L, [TEMP] = 0.462 mmol/L). EPR spectra of (f) DMPO-^•OH, (g) DMPO-^•O₂⁻ and (h) TEMP-¹O₂ from the FeS_xO_y-1:1-based Fenton system.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) Leaching of Fe²⁺ and total iron ions in the FeS_xO_y-1:1-based Fenton reaction. (b) Inhibition effect of AgNO₃, 1, 10-phenanthroline and N₂ on Quinoline degradation in the FeS_xO_y-1:1-based Fenton system ([AgNO₃] = 1 mmol/L, [o-Phenanthroline] = 0.1 mmol/L, [Quinoline] = 20 mg/L), [Catalyst] = 10 mg, [H₂O₂] = 1.47 mmol/L, pH_i ≈ 4). (c) Fe 2p spectra of fresh and used FeS_xO_y-1:1. (d) Fe 2p spectra of fresh and used FeS_xO_y-1:1-S. (e) Divalent iron loss rate on the surface of catalysts. (f) Inhibition effect of different anions on Quinoline degradation in the FeS_xO_y-1:1-based Fenton system. ([Anions] = 10 mg/L, [Quinoline] = 20 mg/L), [Catalyst] = 10 mg, [H₂O₂] = 1.47 mmol/L). (g) Long term degradation of Quinoline by FeS, SV-FeS₂ (SV: Sulfur vacancy) and FeS_xO_y-1:1, respectively. (h) Quinoline degradation in natural water in the FeS_xO_y-1:1-based Fenton system. (i) Wastewater treatment by the FeS_xO_y-1:1-based Fenton system.

  下载: 全尺寸图片 幻灯片
•