Study of Fe3O4/PS System in Degrading BPA in Aqueous Solution

The degradation of bisphenol A (BPA) by Fe3O4/persulfate system was investigated in aqueous solution. The influences of the initial concentrations of Fe3O4, persulfate (PS) and BPA, pH value, and initial reaction temperature on BPA removal were studied. The radical species was investigated by adding excessive dose of scavenger (methanol (MeOH) and tert-butanol (TBA)) into Fe3O4/PS system for the purpose of radical scavenging. The degradation products of BPA were detected by gas chromatography-mass spectrometry (GC-MS). The recyclability of Fe3O4 was also evaluated. The BPA removal rate of 80.7% was achieved under the following conditions: [BPA]0 = 1 mg L, [PS]0 = 0.2 mM, [Fe3O4]0 = 0.1 g L, T0 = 20 ± 1 oC, pH0 = 6.8 ± 0.2. The results confirmed that the main free radicals in the reaction process were sulfate radicals, followed by hydroxyl radicals. Some intermediate products of BPA degradation, such as phenols, benzoquinones and benzoic acid were identified by GC-MS.


Introduction
Bisphenol A (2,2-bis(4-hydroxyphenyl)propane, BPA), a white solid, has been widely used as a ubiquitous intermediate in manufacturing polycarbonate plastics, epoxy resins and polysulfone. 1,2 Therefore, BPA is present in many daily supplies, such as toys, bottles, food and beverage packaging, water supply pipes, and the polymers used in dental treatment. 3,4 BPA is a well-known and studied endocrine disrupting compound (EDC), which can mimic hormones and cause reproductive damage, cancer and other adverse effects on the human body and the ecological environment. 5,6 Previous studies 7,8 have shown that BPA can interfere with the endocrine system of humans and animals even at concentrations below 1 μg m -3 . Due to the discharge of domestic sewage and industrial wastewater, as well as the infiltration of landfill leachate, BPA has frequently been found in surface water and groundwater. 9,10 The study showed that the BPA concentration in some water environments can be as high as 100 μg L -1 . 10 At present, the concentration of BPA in water environment and its harm to aquatic organisms have attracted the attention of researchers who are working on the research of refractory organics removal methods.
Since BPA is a refractory pollutant with two benzene rings, conventional biological processes in wastewater treatment plants are not efficient for complete removal of BPA. 11 There are several alternative approaches that have been used to deal with the BPA over the years, such as adsorption, 12,13 chlorination with sodium hypochlorite, 14 and the advanced oxidation processes (AOPs). The adsorption is restricted owing to the low hydrophobicity (logK ow ) of BPA that limits the adsorption efficiency, 15 and the chlorinated metabolites formed during chlorination with sodium hypochlorite cause some side effects. 16 In contrast, AOPs are popular due to their ability to decompose refractory organics into biodegradable and benign products by powerful free radicals such as hydroxyl radicals (•OH) or sulfate radicals (SO 4 (PS) are commonly used oxidants, which can generate free radicals after being activated by certain ways. 11 4 •− is longer than that of •OH, and the oxidant PS is more stable than O 3 and H 2 O 2 . On the one hand, iron is a common material with benign property, which makes it as an activator to show economic and environmental advantages. On the other hand, from the standpoint of stability, ease of storage, and generation of sulfate radicals, PS is considered to be more suitable as an oxidant in AOPs. Therefore, the use of iron to activate PS to generate free radicals in degradation of refractory organics has become one of the research hotspots in AOPs. The mechanisms by which various forms of iron activate PS are attributed to the redox reactions between Fe II and PS. Based on the existence form of iron, iron-based activator can be classified into homogeneous (such as Fe II ) and heterogeneous (such as Fe 0 and Fe 3 O 4 ). 19 As a homogeneous activator, Fe II is completely dissolved in the solution, which easily causes excessive Fe II , thereby trapping the generated free radicals and reducing the degradation efficiency of organic matter. 20,21 On the contrary, as heterogeneous activates, Fe 0 and Fe 3 O 4 can release Fe II into the water solution at a certain rate to avoid the extinction of free radicals caused by excessive Fe II . Fe 3 O 4 is a preferred activator with mixed valence oxide containing Fe II and Fe III , and it is able to constantly produce Fe II to react with PS and generate free radicals. In addition, Fe 3 O 4 is the main component of magnetite in nature, so it is easy to obtain. Compared with Fe II and Fe 0 , it is easy to recover Fe 3 O 4 from wastewater based on its sub-magnetism. Fe 3 O 4 can also reduce the cost of large volumes of waste effluent treatment and will not result in secondary pollution. 22 Some researchers used Fe II and Fe 0 to activate PS for the treatment of BPA, 22,23 and some researchers studied sulfamonomethoxine, ciprofloxacin hydrochloride degradation in Fe 3 O 4 /PS systems. 22,24,25 However, there are few studies towards BPA removal using Fe 3 O 4 /PS systems, especially the scavenging experiment of free radicals and the intermediate identification of BPA degradation products still need be further studied. In this paper, Fe 3 O 4 magnetic particles have been used as activator to activate PS to produce free radicals for the removal of BPA. The performance of the Fe 3 O 4 /PS system on the degradation of BPA was systematically investigated, the possible effects of environmental factors on BPA removal, and BPA degradation mechanism were discussed in detail. Moreover, the role of active radical species was also explored using scavenging experiments, and the recycling performance of Fe 3 O 4 was evaluated based on Fe 3 O 4 reuse experiment.

Chemicals and equipment
The main chemicals and equipment used in the experiments are present in Tables 1 and 2, respectively. The content of Fe 3 O 4 in the iron oxide material is greater than 99.5%, the particle size of iron oxide is 20 nm, and the specific surface area of iron oxide is 51.46 m 2 g -1 .

Batch experiments
A 1000 mg L -1 of BPA aqueous stock solution was prepared for dilution into a series of concentrations in the batch experiments. The experiments were conducted in the 500-mL beakers placed on a magnetic heating stirrer with was quickly plunge into the beaker to trigger the reaction. The reaction temperature was adjusted by the heater, and the pH was adjusted by adding 0.1 M NaOH or H 2 SO 4 to the solution. The entire reaction time was set to 60 min. Aliquots (0.5 mL) were taken out of the beaker and put into the sample bottle. Then, 0.5 mL ethanol was immediately added into the sample bottle to stop the reaction. The samples were taken every 10 min from the beaker, and the residual concentration of BPA of the final samples were detected by high-performance liquid chromatography (HPLC). For the scavenging of free radicals generated in the reaction, the experiments were performed by adding methanol (MeOH) and tert-butanol (TBA) with preset concentrations in beakers.

Analytical methods
The BPA concentration in samples was examined by a HPLC equipped with a C18 column (4.6 mm × 150 mm, 4 μm) by using a fluorescence detector at 228 nm. The mobile phase was 70% methanol at a flow rate of 0.8 mL min -1 . The excitation and emission wavelengths are 228 and 312 nm, respectively. The temperature of the column compartment was 25 ºC and the injection volume was 20 μL. The retention time of BPA under above conditions was 3.48 min.
The oxidation products of BPA were carried out by a gas chromatography-mass spectrometry (GC-MS). The chromatographic analysis was performed via injection in splitless mode (split ratio of 10:1, injection volume of 1 μL) at 260 ºC using a quartz capillary column (length: 30 m, inner diameter: 0.25 mm, film thickness: 0.25 μm) with a helium flow of 1.0 mL min -1 . The qualitative analysis was performed with electron ionization (EI) at 70 eV using the full scan mode in the m/z range of 45-280. The ion source and quadrupole temperature were maintained at 230 and 150 ºC, respectively.
The removal rate η (%) of BPA is calculated using equation 1, where C 0 is the initial concentration of BPA (mg L -1 ), C t is the BPA concentration after t min treatment (mg L -1 ). The fitting of the kinetic equations in the experiment are based on the first-order reaction kinetic equation, as shown in equation 2, where k is the kinetic constant (min -1 ), b is the y intercept (no unit). (1) (2)

Results and Discussion
The effect of the initial Fe 3   excessive Fe II will extinguish SO 4 •− (equation 5), 27 which will reduce the BPA removal rate. Secondly, the excessive SO 4 •− formed can react between themselves and reduce the number of SO 4 •− (equation 6). 28 Finally, the increase in the amount of Fe 3 O 4 (0.3-0.5 g L -1 ) causes the consumption of PS to increase. At 40 min, PS is exhausted and the number of free radicals decreases, so the degradation rate of BPA increases slowly.
The effect of the initial PS concentration The initial concentration of PS is an important factor affecting the BPA removal by Fe 3 O 4 activation. Figure 2 gives the effect of the PS initial concentration on BPA degradation by Fe 3 O 4 activation. In the absence of PS, it was found that 10.3% of BPA was removed within 60 min, indicating that Fe 3 O 4 could adsorb BPA as an adsorbent. Sun et al. 29 The effect of the initial BPA concentration The concentration of the substrate is also an important factor determining the degradation efficiency. It can be seen from the Figure 3 that when the initial BPA concentration increased from 1 to 5 mg L -1 , after 60 min of reaction, the BPA removal rate dropped from 80.70 to 40.30%. When the initial concentration of BPA is 5 mg L -1 , the kinetic constant for BPA removal is 0.00942 min -1 (ln(C t /C 0 ) = -0.00942t + 0.02804), and the reaction is the slowest. When the initial concentration of BPA was 1 mg L -1 , the kinetic constant for BPA increased to 0.02466 min -1 (ln(C t /C 0 ) = -0.02466t + 0.09686), which was 2.6 times faster than the degradation rate of 5 mg L -1 BPA. As the amount of PS and Fe 3 O 4 remained unchanged, the number of SO 4 •− generated did not increase, and the increase of BPA concentration resulted to a relative shortage of SO 4 •− , which reduced the BPA removal rate. Therefore, in order to improve the degradation efficiency of BPA, the amount of oxidant and catalyst needs to be increased accordingly.

The effect of the initial pH
It is clear that the advanced oxidation reaction in which PS is activated by transition metals is heavily  dependent on solution pH. Therefore, we examined the BPA removal in Fe 3 O 4 /PS system at different pH of 3, 5, 7 and 9 ( Figure 4). According to the BPA removal in descending order, the corresponding pH values were 5, 3, 7, 9, respectively. In other words, the pH value at the maximum BPA removal rate was 5, with the removal rate of 59.2%. Under acidic conditions of pH 3-5, the reaction proceed more efficiently due to the solubilization of the Fe II on the surface of heterogeneous activator Fe 3 O 4 . Therefore, the homogeneous activation of Fe II occurs more effectively for generating SO 4 •− towards removing BPA under acidic conditions. However, under neutral or alkaline conditions, Fe II can hydrolyze to produce the precipitant Fe(OH) 2 which forms a passivation layer on the surface of Fe 3 O 4 , which hinders the migration of Fe II into water, and ultimately slows down the catalytic reaction. In addition, SO 4 •− can react with H 2 O and OH to produce •OH (equation 8) whose oxidizing power is less than SO 4 •− . These reasons resulted in the reduction of BPA oxidation performance under alkaline conditions. This finding was similar to the study of Liu et al. 30 about treatment of landfill leachate biochemical effluent using Fe 3 O 4 /PS system.
The effect of the reaction temperature Temperature is a crucial parameter in the treatment of wastewater as it can affect the rate of chemical reactions. Some scholars have reported experiments that high temperature can activate PS. [31][32][33] We investigated the removal of BPA under four temperature conditions (20,35, 50 and 70 ºC) in Fe 3 O 4 /PS and only PS system. The experimental results with or without Fe 3 O 4 appearing in PS system at different temperatures are shown in Figure 5. In Figure 5, the curve marked with blank is the BPA removal without Fe 3 O 4 (only PS). When only PS was added to the system, the removal rate of BPA increased significantly with the temperature increasing. The highest removal rate occurred at 70 ºC and the lowest removal rate occurred at 20 ºC. When the temperature was 70 ºC, the fastest kinetic constant for BPA removal was 0.00821 min -1 (ln(C t /C 0 ) = -0.00821t + 0.01496), which was 2 times faster than the 20 ºC rate of 0.00414 min -1 (ln(C t /C 0 ) = -0.00414t + 0.00818). Since heat can activate PS to produce SO 4 •− (equation 3), the increase in temperature was conducive to the degradation of BPA. When both PS and Fe 3 O 4 were added to the system, the removal rate of BPA was significantly increased compared to the system with only PS. At 20 and 70 ºC, the removal rate of BPA was 40.46 and 87.10% after 60 min, respectively. This showed high temperature and Fe 3 O 4 have the ability to synergistically catalyze the degradation of BPA by PS. However, the higher the temperature, the more energy is consumed. The actual situation of sewage treatment   should be considered to determine the best temperature from the perspective of energy saving. In addition, as the temperature increased from 20 to 70 ºC, only adding Fe 3 O 4 to the system, the BPA removal rate increased from 10.22 to 12.18% (data not shown). In other words, as the temperature increased, Fe 3 O 4 alone cannot improve the efficiency of its adsorption of BPA.
The scavenging of the free radicals MeOH and TBA are commonly used chemical scavengers to distinguish SO 4 •− and •OH as they react with free radicals with different rate constants. MeOH is an effective scavenger for •OH (reaction rate constant, k = 9.7 × 10 8 M -1 s -1 ) and SO 4 •− (k = 1.1 × 10 7 M -1 s -1 ), while TBA is an effective scavenger for •OH (k = 6.0 × 10 8 M -1 s -1 ) and not for SO 4 •− (k = 9.1 × 10 5 M -1 s -1 ). As observed in Figure 6, without adding the MeOH or TBA, 48.5% of BPA degradation rate was acquired within 60 min. The MeOH and TBA of 80 mM added separately to the system resulted in BPA removal to drop to 25.9 and 45.4%, respectively. As the concentration of MeOH and TBA increased to 800 mM, the degradation rate of BPA further decreased to 12.8 and 39.9%. The addition of MeOH had a more obvious effect on the decrease of BPA degradation rate. Therefore, it was inferred that the SO 4 •− generated in the reaction was the main free radical, followed by the •OH.  Table 3. The detected intermediate products were mainly aromatic compounds, including phenol, p-benzoquinone, p-hydroquinone, 4-(2-hydroxypropan-2-yl) phenol, 4-isopropenylphenol, p-hydroxybenzene propanoic acid and 4-hydroxybenzoic acid. The possible degradation pathways of BPA are shown in Figure 7. The free radicals generated from nano-Fe 3 O 4 /PS system attacked the C-C bond between isopropyl and benzene rings of bisphenol A, forming 4-(2-hydroxypropan-2-yl)phenol and phenol. 34 4-(2-Hydroxypropan-2-yl) phenol was gradually oxidized into 4-isopropenylphenol, p-hydroxyphenylpropionic acid, and 4-hydroxybenzoic acid. 35,36 Phenol was gradually oxidized into p-hydroquinone and p-benzoquinone. Then, all the aromatic compounds produced by two pathways were oxidized into small molecular compounds, 17,37 some of which will be mineralized into CO 2 and H 2 O under the conditions set in Fe 3 O 4 /PS system.

Reusability of Fe 3 O 4
In order to investigate the recyclability of Fe 3 O 4 , a magnet was used to separate Fe 3 O 4 from the solution. After separation, the Fe 3 O 4 was rinsed with deionized water, and dried indoors with natural ventilation for use in the next  experiment. It can be seen from Figure 8  ] 0 = 0.1 g L -1 , T 0 = 20 ± 1 ºC, pH 0 = 6.8 ± 0.2. The oxidation of BPA by Fe 3 O 4 /PS mainly relies on sulfate radicals generated during the reaction, and hydroxyl radicals play a minor role. The possible intermediates of BPA degradation were determined as phenol, p-benzoquinone, p-hydroquinone, 4-(2-hydroxypropan-2-yl) phenol, 4-isopropenylphenol, p-hydroxybenzene propanoic acid and 4-hydroxybenzoic acid. The performance of Fe 3 O 4 was significantly reduced after five times of reuse.