The Coupling Use of Weak Magnetic Field and Fe 0 /H 2 O 2 Process for Bisphenol a Abatement: Inﬂuence of Reaction Conditions and Mechanisms

: The coupling use of the heterogeneous Fenton-like process (zero-valent iron (Fe 0 )/H 2 O 2 ) and weak magnetic ﬁeld (MWF) for bisphenol A (BPA) abatement was systematically investigated in this study. Though both the Fe 0 /H 2 O 2 and WMF-Fe 0 /H 2 O 2 processes are sensitive to pH, WMF remarkably enhanced BPA removal under the pH range of 3.0–6.0 by 0.5–9.5 times. The characterization of Fe 0 conﬁrmed the role of WMF in promoting the corrosion of Fe 0 . Radicals, rather than Fe intermediates, were responsible for BPA degradation. Due to the presence of Cl – as the background ions and its reactivity towards HO • , reactive chlorine species (RCS, i.e., Cl • and Cl 2 •− ) were produced and considerably contributed to BPA degradation. In addition, ~37% and 54% of degraded BPA was ascribed to RCS in the presence of 2 and 100 mM of Cl − , respectively. However, 1.9 mg/L of ClO 3 − was detected in the presence of 2 mM of Cl − in the WMF- Fe 0 /H 2 O 2 process. HCO 3 − could diminish ClO 3 − generation signiﬁcantly through transforming RCS. The concentration of ClO 3 − decreased by 74% and 82% with dosing 1 and 10 mM HCO 3 − , respectively. The results of this study suggest that the WMF-Fe 0 /H 2 O 2 process is a promising approach for BPA removal.


Introduction
Bisphenol A (BPA) is one of the most commonly used chemicals, especially in the industry of polycarbonate plastics and epoxy resins, with annual global consumption of 4.6 million tons [1][2][3]. Consequently, human exposure to BPA becomes inevitable. As an endocrine disruptor, BPA can mimic the body's hormones and disrupt normal cell function [4][5][6][7]. More than 100 studies have demonstrated the negative effects of BPA on structural and neurochemical changes throughout the brain associated with behavioral changes. Worse, these effects may exist throughout the fetal period. Due to incomplete removal at the outlet of sewage treatment plants, BPA was widely detected in surface water, posing a huge challenge to drinking water treatment [8]. Drinking water has become one of the major routes of human exposure to BPA [8]. Liu et al. [3] demonstrated that the distribution of BPA in surface water is related to the development characteristics of industrial areas. The reported maximum concentration of BPA reached up to 12 µg/L in stream water [9], which seriously threatened human health. In drinking water in Italy Fe 2+ + H 2 O 2 → Fe 3+ + OH − + HO • k 1 = 63.0 M −1 s −1 [19] (1) Fe 0 + 2H + → 2Fe 2+ + H 2 As an alternative of Fe 2+ , zero-valent iron (Fe 0 ) received a lot of attention for use in heterogeneous Fenton processes recently [21]. In the Fe 0 /H 2 O 2 process, Fe 0 is firstly oxidized to Fe 2+ through oxygen absorption corrosion and hydrogen evolutional corrosion (Equations (2) and (3)). Similar to the traditional Fenton process, the generated Fe 2+ is converted to Fe 3+ by H 2 O 2 with the generation of HO • . In the meantime, the Fe 3+ reacts with Fe 0 and is transformed to Fe 2+ . Thus, the use of Fe 0 , instead of Fe 2+ salts, in the Fenton process achieved a continuous supply of dissolved Fe 2+ . Moreover, the increase of salinity induced by the adding Fe 2+ salts can be avoided with Fe 0 as the source of Fe 2+ [22].
However, the commercial iron particles are generally covered by a passive layer of iron oxides which forms during the manufacturing process under high temperature (700-1200 • C), resulting in the low reactivity of Fe 0 towards H 2 O 2 [23,24]. In addition, the oxidation of Fe 2+ is faster than the reduction of Fe 3+ in the Fe 0 /H 2 O 2 system [19]. As the reaction progressed, a thicker layer of iron oxides formed on the surface of iron particles which impedes further erosion of Fe 0 [23]. Consequently, the performance of the Fe 0 /H 2 O 2 process for pollutant abatement drops over time. In 2014, Guan et al. [23] serendipitously found the depassivation effect of weak magnetic field (WMF) during Fe 0 corrosion. This could be ascribed to the magnetic forces which give rise to convection in the solution and alleviate the accumulation of iron oxides on the Fe 0 surface. Consequently, the corrosion of Fe 0 was enhanced and the release of Fe 2+ was promoted. On this account, the efficiency of Fe 0 /H 2 O 2 process in removing pollutants from wastewater is anticipated to be improved by a superimposed WMF.
In this study, the effects of WMF on BPA removal by the Fe 0 /H 2 O 2 process under different reaction conditions were systematically investigated. The optimum reaction condition was pointed out with the 3D-response surface methodology and the degradation mechanism of BPA in the WMF-Fe 0 /H 2 O 2 process was illustrated. Previous studies have shown that the reactivity of Fe 0 highly depends on the solution chemistry, especially the coexisting anions [25]. In addition, some widespread anions, such as Cl − and HCO 3 − , are reactive to radicals (e.g., HO • ) and shift the distribution of reactive species in the AOPs. Thus, multiple roles of anions exist in the WMF-Fe 0 /H 2 O 2 process [26] and might be responsible for the performance of the WMF-Fe 0 /H 2 O 2 process for pollutant abatement in real water. Another aim of this study was to disclose the influence of common anions on BPA degradation by the WMF-Fe 0 /H 2 O 2 process and the mechanism behind it.

Experimental Procedure
A schematic diagram of the experimental setup is shown in Figure S1. Two pieces of neodymium-iron-boron permanent magnets under the reactor offered the magnetic field. A Teslameter (TD8620, Shanghai Hengtong Magnetic & Electric Technology Co., Ltd. (Shanghai, China)) was employed to detect the magnetic field intensity in the reactor and the maximum was determined to be 20 mT at the bottom.
A glass bottle filled with the 0.5 L solution containing BPA, buffer, background ions, and the constituent(s) of interest was placed in the thermostatic water bath. The temperature was stabilized at 25 • C unless otherwise noted. Fe 0 particles were uniformly dispersed in a solution by continuous mechanical stirring at 400 rpm. Sodium acetate, acetic acid, and 2-(N-morpholino) ethanesulfonic acid (MES) were used to maintain the pH constant at 3.0-4.0, pH = 5.0, and pH = 6.0, respectively. Then, predetermined amounts of H 2 O 2 and Fe 0 samples were added simultaneously while the solution was stirred. Samples were periodically withdrawn and filtered with a 0.45 µm membrane, and 10 mM of sodium thiosulfate were employed to quench the reaction. Most of the experiments were performed in duplicates and the data were averaged.

Chemical Analysis
Concentration of BPA was determined by high performance liquid chromatograph (HPLC, LC-20A, Shimadzu, Kyoto City, Japan). The mobile phase of CH 3 OH/H 2 O mixture (70/30 v/v) was used with a flow rate of 1.0 mL/min. Fe 0 samples were collected by membrane filtration of 0.45 µm after the reaction, then washed with distilled water and freeze-dried. The morphology of the corrosion products of Fe 0 was analyzed by X-ray diffraction (XRD) and scanning electron microscope (SEM). Oxidation products analysis was performed with gas chromatograph-mass spectrometer (GC-MS, Agilent) technique. The column used was a HP-17 ms quartz capillary column (30 m × 0.25 mm, 0.25 mm). The operational conditions of the GC-MS for product detection can be found in the literature [27]. The concentration of chlorate was measured by an ion chromatography system (Dionex Aquion Rfic, Thermo Scientific) equipped with a conductivity detector. The separation was achieved by a high-capacity hydroxide-selective analytical column and 30 mM KOH eluent at a flow rate of 1.2 mL min -1 .

Results and Discussion
3.1. Influence of WMF on the Removal of BPA by the Fe 0 /H 2 O 2 Process at Different pH Figure S2 shows the time course of BPA oxidation by the Fe 0 /H 2 O 2 or WMF-Fe 0 /H 2 O 2 process under the pH range of 3.0-6.0. As can be seen, the degradation rate of BPA highly depends on pH. The decrease of the concentration of BPA as the function of time was fitted with pseudo-first-order kinetics. At pH 5.0, the degradation of BPA terminated within the reaction time of this study and we only fitted the data in the period of degradation. The obtained rate constants were shown in Figure 1A. Increase of pH from 3.0 to 6.0 leads to the monotonic decrease of the degradation rate of BPA from 0.33 to 0.0012 min −1 in the Fe 0 /H 2 O 2 process and from 0.83 to 0.0021 min −1 in the WMF-Fe 0 /H 2 O 2 process, respectively. This can be mainly ascribed to two reasons: On the one hand, low pH increases the solubility of Fe 3+ and promotes the transformation of Fe 3+ to Fe 2+ [19]. On the other hand, decreasing pH accelerates the corrosion of Fe 0 [23].
formed with gas chromatograph-mass spectrometer (GC-MS, Agilent) technique. The c umn used was a HP-17 ms quartz capillary column (30 m × 0.25 mm, 0.25 mm). The op ational conditions of the GC-MS for product detection can be found in the literature [2 The concentration of chlorate was measured by an ion chromatography system (Dion Aquion Rfic, Thermo Scientific) equipped with a conductivity detector. The separat was achieved by a high-capacity hydroxide-selective analytical column and 30 mM KO eluent at a flow rate of 1.2 mL min -1 .

Results and Discussion
3.1. Influence of WMF on the Removal of BPA by the Fe 0 /H2O2 Process at Different pH Figure S2 shows the time course of BPA oxidation by the Fe 0 /H2O2 or WMF-Fe 0 /H process under the pH range of 3.0-6.0. As can be seen, the degradation rate of BPA hig depends on pH. The decrease of the concentration of BPA as the function of time w fitted with pseudo-first-order kinetics. At pH 5.0, the degradation of BPA termina within the reaction time of this study and we only fitted the data in the period of deg dation. The obtained rate constants were shown in Figure 1A. Increase of pH from 3.0 6.0 leads to the monotonic decrease of the degradation rate of BPA from 0.33 to 0.0 min −1 in the Fe 0 /H2O2 process and from 0.83 to 0.0021 min −1 in the WMF-Fe 0 /H2O2 proce respectively. This can be mainly ascribed to two reasons: On the one hand, low pH creases the solubility of Fe 3+ and promotes the transformation of Fe 3+ to Fe 2+ [19]. On other hand, decreasing pH accelerates the corrosion of Fe 0 [23]. Under the pH range applied in this study, WMF enhanced BPA removal by Fe 0 /H2O2 process ( Figure 1A). As reported by Guan et al. [23], WMF introduces the m netic field gradient force which leads the transmission of the newly generated Fe 2+ to poles of magnetized Fe 0 . In addition, Lorentz force is involved during the movemen the charged species in the presence of magnetic field. These two forces result in the unev distribution of Fe 2+ and its oxidation products. Consequently, more reactivity sites on surface were exposed and the corrosion of Fe 0 was accelerated. Thus, BPA removal w enhanced. In addition to the degradation rate, WMF also improved the removal capac of BPA at pH = 5.0 ( Figure S2). The incomplete removal of BPA at pH 5.0 suggests termination of Fe 0 corrosion and/or H2O2 exhaustion.
Negligible BPA could be removed by H2O2 alone and Fe 0 alone (not shown). Th the candidates of reactive species responsible for BPA removal can be attributed to ra cals and reactive Fe intermediates (e.g., Fe(IV), Fe(V), and Fe(VI)). Methyl phenyl sulf ide (PMSO) could be transformed to methyl phenyl sulfone (PMSO2) by Fe intermedia and is commonly used to determine the involvement of Fe intermediates in the oxidat Under the pH range applied in this study, WMF enhanced BPA removal by the Fe 0 /H 2 O 2 process ( Figure 1A). As reported by Guan et al. [23], WMF introduces the magnetic field gradient force which leads the transmission of the newly generated Fe 2+ to the poles of magnetized Fe 0 . In addition, Lorentz force is involved during the movement of the charged species in the presence of magnetic field. These two forces result in the uneven distribution of Fe 2+ and its oxidation products. Consequently, more reactivity sites on Fe 0 surface were exposed and the corrosion of Fe 0 was accelerated. Thus, BPA removal was enhanced. In addition to the degradation rate, WMF also improved the removal capacity of BPA at pH = 5.0 ( Figure S2). The incomplete removal of BPA at pH 5.0 suggests the termination of Fe 0 corrosion and/or H 2 O 2 exhaustion.
Negligible BPA could be removed by H 2 O 2 alone and Fe 0 alone (not shown). Thus, the candidates of reactive species responsible for BPA removal can be attributed to radicals and reactive Fe intermediates (e.g., Fe(IV), Fe(V), and Fe(VI)). Methyl phenyl sulfoxide (PMSO) could be transformed to methyl phenyl sulfone (PMSO 2 ) by Fe intermediates and is commonly used to determine the involvement of Fe intermediates in the oxidation processes [28]. Comparatively, the oxidation products of PMSO by radicals are hydroxylated and/or polymeric compounds but not PMSO 2 . Therefore, PMSO was adopted to differentiate the reactive species in the WMF-Fe 0 /H 2 O 2 process. As shown in Figure S3, the concentration of PMSO dropped intermediately without formation of PMSO 2 , suggesting that radicals rather than reactive Fe intermediates are responsible for BPA abatement in the WMF-Fe 0 /H 2 O 2 process.
It is well known that HO • is produced in the Fenton or Fenton-like system. Due to the addition of Cl − as the background ions in the reaction solution, reactive chlorine species (RCS, e.g., Cl • and Cl 2 •− ) might be involved in the WMF-Fe 0 /H 2 O 2 process (Table  S1). Phenol, benzoic acid, and BPA were used as the probe compounds to quantify HO • , Cl • , and Cl 2 •− , due to their relatively high reactivity towards these radicals (Table S2). Substituting the apparent rate constants and the corresponding second-order-rate constants into Equations (4)-(6), the concentrations of the three radicals could be determined and the results are presented in Table S3.
where k obs,phenol , k obs,benzoic acid , and k obs,BPA are the apparent rate constants of pollutant oxidation, and k radical-pollutant is the second-order-rate constant of pollutant oxidation by corresponding radical. The percentage contributions of HO • , Cl • , and Cl 2 •− for BPA removal were calculated and are shown in Figure 1B. In total, 3% and 34% of the removed BPA are attributed to the oxidation of Cl • and Cl 2 •− , suggesting the role of RCS cannot be neglected in the AOPs in the presence of Cl -. However, previous studies mainly focus on the influence of anions on Fe 0 corrosion in Fe 0 -based AOPs and the role of RCS in the presence of Cl − has often been neglected. To fill the research gap of this issue, the last part of this study was constructed to clarify the impacts of Cl − in the WMF-Fe 0 /H 2 O 2 process.

Influence of the Dosages of Fe 0 and H 2 O 2 on BPA Removal in the WMF-Fe 0 /H 2 O 2 Process
The influence of reaction conditions (Fe 0 dosage and H 2 O 2 concentration) on the performance of WMF-Fe 0 /H 2 O 2 system for BPA removal was analyzed with response surface methodology. A complete 3 × 3 experimental design [29] was performed using Fe 0 dosage ranging from 3.5 mM to 10.5 mM, and H 2 O 2 concentrations from 4 mM to 8 mM. The codes corresponding to the real values of the variables in the factorial design of experiments were displayed in Table S4. The initial pH was adjusted to 5.0 and the initial concentration of BPA was 0.2 mM. The effects of Fe 0 dosage and H 2 O 2 concentrations were ascertained by calculating degradation efficiency of BPA at 5, 10, 20, 30, 40, and 60 min.
Based on the factorial design of experiments, second-order polynomial equations (Equations (7)-(12)) were used to predict the degradation efficiency (1 − C/C 0 ) of BPA in this study.
Water 2021, 13, 1724 6 of 12 (12) where C BPA , the removal efficiency of BPA, equals (C 0 − C)/C 0 . The experimental results and the calculated data, as well as the standard deviation, are presented in Table S5. As shown in Figure S4, there is an excellent relationship between the experimental data and the predicted values calculated by Equations (7)- (12). The influence of the combination and independent variables is related to the coefficients of the polynomial expressions. The coefficient of Fe 0 dosage term (X) was negative at an initial reaction time (10 min) and conversely became positive along with the reaction. This is because the release of Fe 2+ from the fresh Fe 0 is fast and a high concentration of Fe 0 and Fe 2+ decreased the concentration of radicals (Equation (S8) in Table S1). However, with the reaction proceeding, the release rate of Fe 2+ decreased as part of surface of Fe 0 was covered by Fe oxides. For the H 2 O 2 dosage term (Y), the coefficient was negative after 5 min, indicating that the increase of H 2 O 2 dosage under the concentration range of this study has an adverse effect on BPA degradation. In addition to the precursor of HO • , H 2 O 2 also consumes radicals and depresses the degradation of BPA (Equation (S9) in Table  S1). The value of the coefficient decreased firstly followed by an increase. The increased suppression of H 2 O 2 at the initial stage was due to the decreased release rate of Fe 2+ . By the end of the reaction, the adverse effect of H 2 O 2 was alleviated due to the decrease of the concentration of H 2 O 2 . This suggests that keeping the concentration of H 2 O 2 under an appropriate range during the reaction might be an effective method for optimizing the removal of pollutants in the WMF-Fe 0 /H 2 O 2 system. The positive coefficients of combined variables (XY, X 2 Y, and XY 2 ) indicate the important role of the synergistic effect of Fe 0 and H 2 O 2 . The response curves of degradation efficiency issuing from Equations (7)-(12) are presented in Figure 2. The convex curved surfaces suggest that the optimum dosages of Fe 0 and H 2 O 2 for BPA degradation were within the range of the values studied.
where , the removal efficiency of BPA, equals (C0 − C)/C0. The experimental results and the calculated data, as well as the standard deviation, are presented in Table S5. As shown in Figure S4, there is an excellent relationship between the experimental data and the predicted values calculated by Equations (7)- (12). The influence of the combination and independent variables is related to the coefficients of the polynomial expressions. The coefficient of Fe 0 dosage term (X) was negative at an initial reaction time (10 min) and conversely became positive along with the reaction. This is because the release of Fe 2+ from the fresh Fe 0 is fast and a high concentration of Fe 0 and Fe 2+ decreased the concentration of radicals (Equation (S8) in Table S1). However, with the reaction proceeding, the release rate of Fe 2+ decreased as part of surface of Fe 0 was covered by Fe oxides. For the H2O2 dosage term (Y), the coefficient was negative after 5 min, indicating that the increase of H2O2 dosage under the concentration range of this study has an adverse effect on BPA degradation. In addition to the precursor of HO • , H2O2 also consumes radicals and depresses the degradation of BPA (Equation (S9) in Table S1). The value of the coefficient decreased firstly followed by an increase. The increased suppression of H2O2 at the initial stage was due to the decreased release rate of Fe 2+ . By the end of the reaction, the adverse effect of H2O2 was alleviated due to the decrease of the concentration of H2O2. This suggests that keeping the concentration of H2O2 under an appropriate range during the reaction might be an effective method for optimizing the removal of pollutants in the WMF-Fe 0 /H2O2 system. The positive coefficients of combined variables (XY, X 2 Y, and XY 2 ) indicate the important role of the synergistic effect of Fe 0 and H2O2. The response curves of degradation efficiency issuing from Equations (7)-(12) are presented in Figure 2. The convex curved surfaces suggest that the optimum dosages of Fe 0 and H2O2 for BPA degradation were within the range of the values studied.

Characterization of the Corrosion Production of Fe 0
To explore the corrosion mechanisms of Fe 0 during BPA degradation by the Fentonlike system, the corrosion products of Fe 0 in the absence or presence of WMF were characterized with XRD and SEM. XRD spectra was shown in Figure 3A. The Fe 0 sample before

Characterization of the Corrosion Production of Fe 0
To explore the corrosion mechanisms of Fe 0 during BPA degradation by the Fenton-like system, the corrosion products of Fe 0 in the absence or presence of WMF were characterized with XRD and SEM. XRD spectra was shown in Figure 3A. The Fe 0 sample before corrosion has diffraction peaks at 2θ = 45 • , 66 • , and 83 • , while the corrosion product had new characteristic peaks at 2θ = 13 • , 28 • , and 37 • , corresponding to the lepidocrocite. The intensity of the characteristic peaks of lepidocrocite in the WMF-Fe 0 /H 2 O 2 process is higher than that in the Fe 0 /H 2 O 2 process. In addition, the XRD patterns of magnetite were also detected in the corrosion product at WMF-Fe 0 /H 2 O 2 systems. The significant difference in the crystalline structures of iron oxides should be ascribed to the interference of magnetic field gradient force and Lorentz force [30]. corrosion has diffraction peaks at 2θ = 45°, 66°, and 83°, while the corrosion product had new characteristic peaks at 2θ = 13°, 28°, and 37°, corresponding to the lepidocrocite. The intensity of the characteristic peaks of lepidocrocite in the WMF-Fe 0 /H2O2 process is higher than that in the Fe 0 /H2O2 process. In addition, the XRD patterns of magnetite were also detected in the corrosion product at WMF-Fe 0 /H2O2 systems. The significant difference in the crystalline structures of iron oxides should be ascribed to the interference of magnetic field gradient force and Lorentz force [30]. The morphologies evolution of Fe 0 particles was analyzed by monitoring the SEM images, as illustrated in Figure 3(B1-B3). The shape of Fe 0 particles coarsened slightly in Fe 0 /H2O2 systems. However, the surface morphology of Fe 0 samples collected under WMF irradiation became rough and irregular, and some of the surface of Fe 0 particles cracked. This might be attributed to the formed iron hydroxide/oxide and the proton (H + ) dissolution of the FexOy layer [23]. The different morphologies of the corroded Fe confirmed the promotion of WMF on Fe 0 corrosion.

Identification of Intermediate Products and Speculation of Degradation Path
GC-MS was employed to analyze the oxidation intermediates of BPA in the WMF-Fe 0 /H2O2 systems at pH 5.0. Five compounds were discriminated with the GC-MS analysis ( Figure S5-S9), and the results are shown in Table S6.
The transformation of organics by HO • is primarily via hydrogen atom abstraction from a C-H bond and HO − addition to a C-C double bond [31]. Single electron transfer predominated in the reactions of Cl • and Cl2 •− with phenolic compounds. Based on the intermediates listed in Table S6 and the results reported by other researchers [32,33], as well as the oxidation mechanisms of organic pollutants by HO • and RCS, the possible degradation pathway of BPA in WMF-Fe 0 /H2O2 systems was speculated, as shown in Figure 4. On the one hand, the radicals attack the C-C bond between the benzene ring and the isopropyl to form some primary oxidation products, such as 3,4-di-tert-butylphenol (DTB) and 4-isopropenylphenol (IPP) [34]. Then, DTB and IPP were converted by radicals into phenol and hydroquinone. The oxhydryl of the opposite position in phenol is reactive to oxidants and easily oxidized to p-benzoquinone. Further, 2-butoxypent-1-ene (BPE) formed through the oxidative ring-opening reaction. On the other hand, the radicals attack the C-O bond between the benzene ring and the hydroxyl radical to form 1,1-diphenyl propylene (DPE). Styrene was generated by degradation of DPE, followed by the formation of maleic acid. The small molecular compounds were eventually mineralized into CO2 and H2O. The morphologies evolution of Fe 0 particles was analyzed by monitoring the SEM images, as illustrated in Figure 3(B1-B3). The shape of Fe 0 particles coarsened slightly in Fe 0 /H 2 O 2 systems. However, the surface morphology of Fe 0 samples collected under WMF irradiation became rough and irregular, and some of the surface of Fe 0 particles cracked. This might be attributed to the formed iron hydroxide/oxide and the proton (H + ) dissolution of the Fe x O y layer [23]. The different morphologies of the corroded Fe confirmed the promotion of WMF on Fe 0 corrosion.

Identification of Intermediate Products and Speculation of Degradation Path
GC-MS was employed to analyze the oxidation intermediates of BPA in the WMF-Fe 0 /H 2 O 2 systems at pH 5.0. Five compounds were discriminated with the GC-MS analysis ( Figures S5-S9), and the results are shown in Table S6.
The transformation of organics by HO • is primarily via hydrogen atom abstraction from a C-H bond and HO − addition to a C-C double bond [31]. Single electron transfer predominated in the reactions of Cl • and Cl 2 •− with phenolic compounds. Based on the intermediates listed in Table S6 and the results reported by other researchers [32,33], as well as the oxidation mechanisms of organic pollutants by HO • and RCS, the possible degradation pathway of BPA in WMF-Fe 0 /H 2 O 2 systems was speculated, as shown in Figure 4. On the one hand, the radicals attack the C-C bond between the benzene ring and the isopropyl to form some primary oxidation products, such as 3,4-di-tert-butylphenol (DTB) and 4-isopropenylphenol (IPP) [34]. Then, DTB and IPP were converted by radicals into phenol and hydroquinone. The oxhydryl of the opposite position in phenol is reactive to oxidants and easily oxidized to p-benzoquinone. Further, 2-butoxypent-1-ene (BPE) formed through the oxidative ring-opening reaction. On the other hand, the radicals attack the C-O bond between the benzene ring and the hydroxyl radical to form 1,1-diphenyl propylene (DPE). Styrene was generated by degradation of DPE, followed by the formation of maleic acid. The small molecular compounds were eventually mineralized into CO 2 and H 2 O.
The chlorinated products were not detected during the reaction, suggesting addition of Cl-is not the main reaction mechanisms of BPA degradation. As the ated compounds are thought to be highly toxic, the WMF-Fe 0 /H2O2 is expected to d the toxicity of the BPA solution.  Figure 5 shows the influence of Cl − on BPA degradation in the WMF-Fe 0 /H cess. Concentrations of Cl − of 0.002 and 0.1 M were selected as they are common water and industrial wastewater [26]. As can be seen, the degradation rate of creased with increasing Cl − concentration. Phenol and benzoic acid were also ad the reaction solution to analyze the distribution of radicals, and the correspondin dation kinetics are presented in Figure S10. The concentrations of HO • , Cl • , and C calculated with competition kinetic methods. As shown in Table S3, the concentr HO • increased with increasing Cl − dosage despite the transformation of HO • by RCS. In addition, the percentage of RCS compared to the total radicals increas improving Cl − concentration.

Influence of Cl − on BPA Removal and Chlorate Formation in the WMF-Fe 0 /H2O2 Pr
Previous studies have demonstrated that the copresence of WMF and Cl − acc the Fe 0 corrosion and Fe 2+ release [22,25]. As mentioned above, the transmission was driven by the magnetic field gradient force and the Lorentz force. Based on t ciple of local electro neutrality, the positive charges must be balanced with n charges to keep local charge neutrality [25]. Therefore, Cl − is expected to move with Fe 2+ toward the magnetic poles. Therefore reactive sites of Fe0 centers were The chlorinated products were not detected during the reaction, suggesting that the addition of Cl-is not the main reaction mechanisms of BPA degradation. As the chlorinated compounds are thought to be highly toxic, the WMF-Fe 0 /H 2 O 2 is expected to decrease the toxicity of the BPA solution.  Figure 5 shows the influence of Cl − on BPA degradation in the WMF-Fe 0 /H 2 O 2 process. Concentrations of Cl − of 0.002 and 0.1 M were selected as they are common in fresh water and industrial wastewater [26]. As can be seen, the degradation rate of BPA increased with increasing Cl − concentration. Phenol and benzoic acid were also added into the reaction solution to analyze the distribution of radicals, and the corresponding degradation kinetics are presented in Figure S10. The concentrations of HO • , Cl • , and Cl 2

•−
were calculated with competition kinetic methods. As shown in Table S3, the concentration of HO • increased with increasing Cl − dosage despite the transformation of HO • by Cl − into RCS. In addition, the percentage of RCS compared to the total radicals increased with improving Cl − concentration.
Considering the rapid reaction of H2O2 with Fe 2+ (Equation (1)), the rise of Fe 2+ concentra tion should be ascribed to the exhaustion of H2O2. In the absence of Cl − , the concentration of Fe 2+ was kept low under the reaction time of this study, suggesting that the corrosion rate of Fe 0 was low and H2O2 was not completely consumed within the reaction time (Fig  ure S11). Therefore, the increase of the concentration of radicals with increasing Clwas mainly due to the enhanced corrosion of Fe 0 . ClO3 − is a harmful chemical which potentially forms in the AOPs in the presence o Cl − [35]. The health reference level of ClO3 − is 210 µg/L according to the U.S. Environmen tal Protection Agency and the drinking water standards are 200 and 700 µg/L in Switzer land and China, respectively. As shown in Figure 6, ~1.9 mg/L of ClO3 − was detected dur ing BPA degradation in the WMF-Fe 0 /H2O2 process in the presence of 2 mM Clafter a 60 min reaction. According to the study of Hou et al. [35], Cl • is the intermediate during the conversion of Cl − to ClO3 − . It is supposed that the substance which influences the conver sion of Cl • may alter the generation of ClO3 − . HCO3 − , a common ion present in real water reacts rapidly with Cl • (Equation (S13) in Table S1). In addition, HCO3 − can also diminish Cl • concentrations by consuming Cl2 •-as there is an equilibrium among Cl • and Cl2 • (Equations (S16), (S17), and (S19) in Table S1). Thus, HCO3 − is anticipated to influence the formation of ClO3 − . Tests were carried out at two HCO3 − concentrations (1 and 10 mM). As shown in Figure 6, the detected concentration of ClO3 − decreased to 0.50 and 0.34 mg/L with HCO3 − dosing of 1 and 10 mM, respectively. Previous studies have demonstrated that the copresence of WMF and Cl − accelerated the Fe 0 corrosion and Fe 2+ release [22,25]. As mentioned above, the transmission of Fe 2+ was driven by the magnetic field gradient force and the Lorentz force. Based on the principle of local electro neutrality, the positive charges must be balanced with negative charges to keep local charge neutrality [25]. Therefore, Cl − is expected to move together with Fe 2+ toward the magnetic poles. Therefore reactive sites of Fe0 centers were exposed in the solution. However, in the system without Cl − , the accumulation of Fe 2+ in the magnetic poles inhibited the subsequent transmission of Fe 2+ due to the electrostatic exclusion. The concentration of Fe 2+ as the function of time in the WMF-Fe 0 /H 2 O 2 process in the presence of different concentrations of Cl − is shown in Figure S11. Compared with Figure 5, the concentration of Fe 2+ was quite low during the degradation of BPA and rose after that. Considering the rapid reaction of H 2 O 2 with Fe 2+ (Equation (1)), the rise of Fe 2+ concentration should be ascribed to the exhaustion of H 2 O 2 . In the absence of Cl − , the concentration of Fe 2+ was kept low under the reaction time of this study, suggesting that the corrosion rate of Fe 0 was low and H 2 O 2 was not completely consumed within the reaction time ( Figure S11). Therefore, the increase of the concentration of radicals with increasing Clwas mainly due to the enhanced corrosion of Fe 0 . ClO 3 − is a harmful chemical which potentially forms in the AOPs in the presence of Cl − [35]. The health reference level of ClO 3 − is 210 µg/L according to the U.S. Environmental Protection Agency and the drinking water standards are 200 and 700 µg/L in Switzerland and China, respectively. As shown in Figure 6,~1.9 mg/L of ClO 3 − was detected during BPA degradation in the WMF-Fe 0 /H 2 O 2 process in the presence of 2 mM Clafter a 60 min reaction. According to the study of Hou et al. [35], Cl • is the intermediate during the conversion of Cl − to ClO 3 − . It is supposed that the substance which influences the conversion of Cl • may alter the generation of ClO 3 − . HCO 3 − , a common ion present in real water, reacts rapidly with Cl • (Equation (S13) in Table S1). In addition, HCO 3 − can also diminish Cl • concentrations by consuming Cl 2 •as there is an equilibrium among Cl • and Cl 2 •− (Equations (S16), (S17), and (S19) in Table S1). Thus, HCO 3 − is anticipated to influence the formation of ClO 3 − . Tests were carried out at two HCO 3 − concentrations (1 and 10 mM). As shown in Figure 6, the detected concentration of ClO 3 − decreased to 0.50 and 0.34 mg/L with HCO 3 − dosing of 1 and 10 mM, respectively.
Water 2021, 13, x FOR PEER REVIEW In addition to RCS, HCO3 − also reacts with HO • and produces CO3 •− (Equatio in Table S1). Therefore, HCO3 − changed the distribution of radicals in the WMF-F process and may influence the degradation of BPA. As shown in Figure S12, the a of HCO3 − slightly accelerated the degradation of BPA in the presence of 2 mM C positive effect of HCO3 − could be explained by two reasons: One reason is that t tivity of CO3 •− towards BPA is very high (k = 2.23 × 10 8 M −1 s −1 ) [36]. The second r that the steady state concentrations of CO3 •− (~10 −12 M level) are generally much than HO • and Cl • [36]. Consequently, the contribution of CO3 •− for BPA oxidatio the loss of HO • and RCS. The reaction ended after 30 min regardless of whethe presence or absence of HCO3 − and the removal efficiencies are similar, suggesting negligibly influenced the corrosion of Fe 0 .
It should be noted that CO3 •− is a selective oxidant and tends to oxidation com containing electron-donating groups, such as NH2and HO-containing aromat pounds [37]. The second-order-rate constants of CO3 •− with organic compounds w ported to range widely from 10 2 to 10 9 M −1 s −1 [37]. Thus, the synergy effects of HCO3 -on AOPs, especially for the degradation kinetics of pollutants, are expecte pend on the structures of compounds which need further study.

Conclusions
This study demonstrated the substantial enhancement of WMF on BPA degr by the Fe 0 /H2O2 process. The corrosion rate of Fe 0 decreased significantly with inc pH which limits the application of the Fe 0 /H2O2 process. WMF extended the avail range of the Fe 0 /H2O2 process, saving the dosage of acid for pH adjustment whic main cost of the Fenton process. Due to the derivational move of Fe 2+ to the pol particles under magnetic forces, the Fe 0 will not be covered by the iron oxides, k the reactivity of Fe 0 during the reaction. Radicals, rather than reactive Fe intermed responsible for BPA degradation. Clshifts the distribution of radicals from HO • t Cl • and Cl2 •− which are highly reactive towards BPA. However, the undesired Cl generated in the presence of Cl − . HCO3 − could diminish the formation of ClO3 − b forming the radicals into CO3 •− . Due to the high reactivity of CO3 •− towards BPA, dition of HCO3 -also promoted BPA degradation by the WMF-Fe 0 /H2O2/Cl − system the widespread ions, Cl − and HCO3 − , in water are beneficial for BPA remova Fe 0 /H2O2-WMF process. The WMF used in this study was produced by permane nets and did not require energy input. Under WMF, commercial Fe 0 materials c In addition to RCS, HCO 3 − also reacts with HO • and produces CO 3 •− (Equation (S20) in Table S1). Therefore, HCO 3 − changed the distribution of radicals in the WMF-Fe 0 /H 2 O 2 process and may influence the degradation of BPA. As shown in Figure S12, the addition of HCO 3 − slightly accelerated the degradation of BPA in the presence of 2 mM Cl -. The positive effect of HCO 3 − could be explained by two reasons: One reason is that the reactivity of CO 3 •− towards BPA is very high (k = 2.23 × 10 8 M −1 s −1 ) [36]. The second reason is that the steady state concentrations of CO 3 •− (~10 −12 M level) are generally much higher than HO • and Cl • [36]. Consequently, the contribution of CO 3 •− for BPA oxidation offset the loss of HO • and RCS. The reaction ended after 30 min regardless of whether in the presence or absence of HCO 3 − and the removal efficiencies are similar, suggesting HCO 3 negligibly influenced the corrosion of Fe 0 . It should be noted that CO 3 •− is a selective oxidant and tends to oxidation compounds containing electron-donating groups, such as NH 2 -and HO-containing aromatic compounds [37]. The second-order-rate constants of CO 3 •− with organic compounds were reported to range widely from 10 2 to 10 9 M −1 s −1 [37]. Thus, the synergy effects of Cland HCO 3 on AOPs, especially for the degradation kinetics of pollutants, are expected to depend on the structures of compounds which need further study.

Conclusions
This study demonstrated the substantial enhancement of WMF on BPA degradation by the Fe 0 /H 2 O 2 process. The corrosion rate of Fe 0 decreased significantly with increasing pH which limits the application of the Fe 0 /H 2 O 2 process. WMF extended the available pH range of the Fe 0 /H 2 O 2 process, saving the dosage of acid for pH adjustment which is the main cost of the Fenton process. Due to the derivational move of Fe 2+ to the pole of Fe 0 particles under magnetic forces, the Fe 0 will not be covered by the iron oxides, keeping the reactivity of Fe 0 during the reaction. Radicals, rather than reactive Fe intermediate, are responsible for BPA degradation. Clshifts the distribution of radicals from HO • towards Cl • and Cl 2 •− which are highly reactive towards BPA. However, the undesired ClO 3 − was generated in the presence of Cl − . HCO 3 − could diminish the formation of ClO 3 − by transforming the radicals into CO 3 •− . Due to the high reactivity of CO 3 •− towards BPA, the addition of HCO 3 also promoted BPA degradation by the WMF-Fe 0 /H 2 O 2 /Cl − system. Thus, the widespread ions, Cl − and HCO 3 − , in water are beneficial for BPA removal in the Fe 0 /H 2 O 2 -WMF process. The WMF used in this study was produced by permanent magnets and did not require energy input. Under WMF, commercial Fe 0 materials could be directly acceptable, without expensive and inconvenient pretreatment methods for removing surface iron oxides. Therefore, applying WMF to enhance the performance of Fe 0 /H 2 O 2 process for BPA removal in real water treatment was a promising and environmentally friendly approach.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/w13131724/s1, Figure S1: Diagram of experimental device for degradation of BPA; Figure S2: Effect of WMF on the removal of BPA by the Fe 0 /H 2 O 2 process at different pH; Figure S3: Degradation of PMSO and generation of PMSO 2 in the WMF-Fe 0 /H 2 O 2 process; Figure S4: Accuracy of predicted data with respect to the experimental results of BPA removal; Figure S5: GC-MS spectrum of maleic acid; Figure S6: GC-MS spectrum of 4-isopropenylphenol; Figure S7: GC-MS spectrum of 2-butoxypent-1-ene; Figure S8: GC-MS spectrum of 3,4-di-tert-butylphenol; Figure S9: GC-MS spectrum of bisphenol A; Figure S10: Influence of Clon the degradation kinetics of phenol and benzoic acid by the WMF-Fe 0 /H 2 O 2 process; Figure S11: Effect of Clon the concentration of Fe 2+ in the reaction solution during BPA degradation by the WMF-Fe 0 /H 2 O 2 process; Figure S12: Influence of HCO 3 − on BPA degradation in the WMF/Fe 0 /H 2 O 2 process; Table S1: Principal reactions in the Fenton system in the presence of Cland HCO 3 − ; Table S2: Second-order-rate constants of radicals towards probe compounds; Table S3: Influence of Clon the distribution of radicals in the WMF-Fe 0 /H 2 O 2 process; Table S4: Factorial design of experiments; Table S5: Descriptive statistics of 3D response surface analysis; Table S6: Mass spectra data of intermediate products.