Sonolytic degradation of bisphenol S: Effect of dissolved oxygen and peroxydisulfate, oxidation products and acute toxicity

Water Research 165 (2019) 114969

Contents lists available at ScienceDirect

Water Research journal homepage: www.elsevier.com/locate/watres

Sonolytic degradation of bisphenol S: Effect of dissolved oxygen and peroxydisulfate, oxidation products and acute toxicity Xiaohui Lu a, Jingnan Zhao b, Qun Wang b, **, Da Wang a, Haodan Xu a, Jun Ma a, *, Wei Qiu a, Tao Hu b a b

State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology, Harbin, 150090, PR China Faculty of Geosciences and Environmental Engineering, Southwest Jiaotong University, Chengdu, 610031, PR China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 January 2019 Received in revised form 26 July 2019 Accepted 8 August 2019 Available online 9 August 2019

In this paper, the kinetics of bisphenol S (BPS) degradation in the presence of peroxydisulfate (PDS) or dissolved oxygen (DO) in ultrasound (US) system were investigated. For PDS (US/PDS), increased PDS concentration result in faster BPS degradation, but the enhancement was not remarkable with multiplying PDS dosages. Therefore, heterogeneous PDS activation model based on a Langmuir-type adsorption mechanism was proposed to explain the trait of BPS abatement. The equilibrium constant of PDS (KPDS) was calculated to be 2.91
104/mM, which was much lower than that of BPS, suggesting that PDS was hard to adsorb on the gas-liquid interface of the cavitation bubble following by activation. Besides,  the formation of OH and SO 4 in US/PDS system was reinvestigated. The result showed that SO4 rather than OH was the predominant radical, which was quite different from previous study. Dissolved oxygen largely improve the degradation of BPS in US system and OH rather than O 2 was proved to be the main reactive oxygen species (ROS). The improvement of OH generation possibly caused by the reaction of DO with H so that it cannot recombine with OH. The transformation of the BPS in US system mainly included BPS radical polymerization, hydroxylation and hydrolysis. Frustratingly, the acute toxicity assay of Vibrio fischeri suggests that the degradation products of BPS are more toxic. These results will improve the understanding on the activation mechanisms of PDS and the role of dissolved oxygen play in US. Further investigations may need to explore other treatment ways of BPS and evaluate the acute toxicity of degradation products. © 2019 Elsevier Ltd. All rights reserved.

Keywords: Bisphenol S (BPS) Peroxydisulfate (PDS) Dissolved oxygen (DO) Kinetics Degradation pathway Acute toxicity

1. Introduction Bisphenol S (BPS; 4, 4-sulfonyldiphenol) was gradually concerned due to its estrogenic activity and potential geno-toxicity (Eladak et al., 2015; Grignard et al., 2012; Ullah et al., 2016). Unfortunately, BPS was ubiquitous in foodstuffs (Liao and Kannan, 2014), paper products, currency bills (Liao et al., 2012a), personal care products, water bodies (Guo et al., 2016; Jin and Zhu, 2016) and even was detected in human urine (Liao et al., 2012b) with the wide application. Besides, it was reported that BPS have a longer half-life but lower biodegradability compared to BPA (Danzl et al., 2009; Ike et al., 2006). Recently, the oxidation of BPS by different treatment ways were reported (Gao et al., 2017; Li et al., 2018; Wang et al., 2017; Xu et al., 2018; Yang et al., 2019), but the post-treatment

* Corresponding author. ** Corresponding author. E-mail address: [email protected] (J. Ma). https://doi.org/10.1016/j.watres.2019.114969 0043-1354/© 2019 Elsevier Ltd. All rights reserved.

acute toxicity of BPS was poorly acknowledged. Over the past three decades, advanced oxidation processes (AOPs) have been continuously developed for treatment of hazardous organic compounds in water (Matilainen and Sillanpaa, 2010). Free radicals can be generated by several means including ozone (Staehelin and Hoigne, 1985; Wang et al., 2018), UV (Shen et al., 1995), ultrasound (US) (Goel et al., 2004), peroxydisulfate (PDS) (Anipsitakis and Dionysiou, 2004), peroxymonosulfate (PMS) (Anipsitakis and Dionysiou, 2003), hydrogen peroxide (Lin and Gurol, 1998) and periodate (PI) (Bokare and Choi, 2015; Lee et al., 2014; Weavers et al., 1997) based AOP systems. Ultrasound was paid considerable attention in recent ten years for water/wastewater treatment due to the hydroxyl radical (OH) generation by the collapse of the bubbles that produced via sonication of aqueous solution (Ince, 2018). As the extreme condition (i.e., high temperature and pressure) of “hot spots” forms with the collapse of bubbles (McNamara. et al., 1999), US combined with many oxidants such as ozone (Barik and Gogate, 2017; Guo et al., 2015), PDS (Chen and Su, 2012), PMS (Yin et al., 2018) and PI (Hamdaoui and

2

X. Lu et al. / Water Research 165 (2019) 114969

Merouani, 2017; Lee et al., 2016) have shown good synergistic effect. Recently, PDS has attracted considerable attention for in situ chemical oxidation (ISCO) technology. It has been confirmed that ultrasound could accelerate the decomposition of PDS (Price and Clifton, 1996), but the studies about the activation of PDS by US were still limited (Darsinou et al., 2015; Ferkous et al., 2017; Li et al., 2013; Wang and Zhou, 2016). Since no addition of extra activator, ultrasound has the potential for PDS activation in ISCO technology and water treatment. Wei et al. investigated the mechanism of activated PDS in US by in situ electron paramagnetic resonance (EPR) spin trapping technique (Wei et al., 2017). The results revealed that OH was dominant in US/PDS rather than the sulfate radicals (SO 4 ) due to the  hydrolysis of SO might be 4 . However, the hydrolysis of SO4 overestimated in this study. The author estimated the yield of radicals (i.e., OH and SO 4 ) according to the steady state concentration of DMPO (5, 5-dimethyl-1-pyrroline-N-oxide) adduct (i.e., DMPO-OH and DMPO-SO 4 ), but the nucleophilic substitution reaction of SO 4 eDMPO yielding OHeDMPO was neglected (Davies et al., 1992; KIRINO et al., 1981; Timmins et al., 1999; Zamora and Villamena, 2012). Obviously, the formation of DMPOOH by DMPO-SO 4 hydrolysis would slow down the decline of DMPO-OH, as a result of higher steady state yield of DMPO-OH than DMPO-SO 4. The positive influence of oxygen on contaminants degradation in US system was acknowledged for transforming into reactive oxygen species (ROS) (Berberidou et al., 2007). But the function of oxygen was not very clear up to now. In this study, firstly, different BPS degradation experiments were conducted at various conditions to evaluate the role of DO in US, and how it transformed in US system. Secondly, the dominating ROS and kinetic orders in US/PDS system were further studied. Heterogeneous PDS activation model was proposed due to the discrepancy in kinetic orders between US and homogenous thermal activation system. Finally, the degradation products of BPS were identified by LC e QTOF e MS/MS and the post-treatment toxicity of BPS was measured. 2. Materials and methods 2.1. Reagents PDS (99% pure) was purchased from Sigma-Aldrich Co., Ltd (Shanghai, China). BPS, benzoic acid (BA, 99% pure), 4-chloro-7nitrobenzo-2-oxa-1,3-diazole (NBD-Cl, 98% pure), Humic acid (HA), and tetrachloromethane (CCl4, 99% pure) were purchased from Aladdin Chemical Reagent Co., Ltd., China. Titanium potassium oxalate (K2TiO(C2O4)2, 99% pure), phosphate (K2HPO4 and KH2PO4, 99% pure), Potassium chloride (KCl, 99% pure), and Potassium bicarbonate (KHCO3, 99% pure) were obtained from Sinopharm Chemical Reagent Company Ltd. (Shanghai, China). HPLC grate methanol, acetonitrile and acetic acid used for BPS and intermediates determination were obtained from Tedia, USA. All of the stock solutions were prepared using Milli-Q deionized water (>18.2 MU cm). 2.2. Experimental procedure 2.2.1. Sonication experiment The ultrasound experimental setup was shown in Fig. S1. All experiments were conducted in a 500-mL titanium alloy reactor with an ultrasonic transducer (28 kHz) at the bottom. The thermostat (THD-2015, Tianheng, China) was used to maintain temperature at 20 ± 1.0  C. The solutions were mixed evenly before the

reaction without extra stir during the sonication due to the enhanced mass transfer of ultrasound. Reactions were initiated by switching on the ultrasonic generator. At given time intervals, 1 mL of water samples were withdrawn and quenched with 200 mL methyl alcohol for BPS analysis. In order to figure out the effects of dissolved oxygen (DO), four experimental conditions were employed as follows: 1. Oxic conditions. The ultrapure water with 9.20 mg/L DO as the background solution. 2. Anoxic conditions. The reaction solutions were purged by nitrogen stripping in 250 mL flask for 15 min with magnetic stirring and then monitored by DO meter (HACH, USA) (about 98% of DO was removed, shown in Fig. S2). Then, the solutions were transferred into the titanium alloy reactor with constant nitrogen bubbling. Meanwhile, the diffuser disc above the titanium alloy reactor aerated the gas space of reactor with nitrogen ensuring anoxic conditions during the reactions. 3. Oxygen aeration. Oxygen aerating within titanium aerator was conducted for the entire reaction time. 4. Nitrogen aeration. In addition to the set of steps in condition 2, nitrogen aerating within titanium aerator was conducted for the entire reaction period. 2.2.2. Thermo-activated persulfate experiment The thermo-activated persulfate oxidation experimental procedure has been shown in detail in our previous study (Wang et al., 2017). 2.2.3. UV radiation experiment The UV device and experimental procedures were shown in Fig. S3 and Text S1. 2.3. Analytical methods All organic contaminants were determined on highperformance liquid chromatograph (HPLC) (Waters e2695) equipped with a Symmetry C18 column (4.6 mm
150 mm
5 mm, Waters) and dual l UVevis detector (Waters 2998) at the column temperature of 35  C. The eluent solution consisted of water (0.1% acetic acid) (phase A) and methanol (phase B) at the flow rate of 1.0 mL/min. The phase ratio of 40:60 (A/B) was used for BPS analysis and 50:50 (A/B) for BA and pHBA analysis, respectively. The concentration of chloride ions content was measured by an ion chromatograph (Dionex ICSe3000), and 20 mM KOH was used as isocratic eluent at a flow rate of 1.0 mL min1, and the sup-pressor current was set at 50 mA. High performance liquid chromatography combined with an ABSciex QTrap 5500 MS at ESI negative ionization mode was used to BPS degradation intermediates evolution. The mobile phase consisted of pure water (A) and acetonitrile (B) was maintained at a rate of 0.5 ml min1, and the gradient was as follows: 0e5 min, 5% B; 5e15 min, phase B increased linearly from 5% to 95%; 15e25 min, 95% B; 25e30 min, phase B return to 5%; 30e35 min, 5% B. The ESI source parameters were listed as follows: spray voltage, 5500 V; source temperature, 500  C; curtain gas, gas I and gas II were set at 35, 50 and 50 arbitrary units, respectively; declustering potential (DP), 80V; collision energy (CE), 10V; MS scan (m/z) range, 100e800. H2O2 concentration was determined by using the titanium potassium oxalate method (Sellers, 1980). The 2 ml samples were added into the test tube with 2 ml titanium potassium oxalate solution (8 mM) and 100 mL H2SO4 (4 M), then, the samples were measured at 400 nm with a UVevis spectrophotometer. The NBDeCl was used for determination of superoxide radicals (O 2 ) (Heller and Croot, 2010). The reaction product of NBDeCl with O 2 has a characteristic absorbance at 470 nm. The Fluorescence Spectrometer (F-6500, JASCO, Japan) was used for qualitative analysis of O 2 , excitation wavelength and emission wavelength

X. Lu et al. / Water Research 165 (2019) 114969

were set at 470 nm and 555 nm, respectively. 100 mM concentration of NBDeCl was used in different US experiments. Samples in different US system were withdrawn at desired time intervals and then used for fluorometric analysis. The acute toxicity of BPS and post-treatment samples was carried out based on measuring the inhibition of bioluminescence emitted by the photobacterium Vibrio fischeri (see Text S2). The luminescence intensity of samples was measured by Promega GloMax®-Multi Jr after 15 min (Fernandez-Alba et al., 2002; Olmez-Hanci et al., 2013). All bioassays were run in triplicate.

3. Results and discussion 3.1. Degradation efficiency of BPS in different US systems The enhanced degradation of BPS was achieved in the absence of cooling water (insert of Fig. 1) because of the higher power density of solution (The cooling water system may dissipate energy). In addition, without cooling water system, the solution temperature would continue to increase as high as 55  C in 60 min due to the heating effect of ultrasonic transducer. Almost complete degradation of BPS was achieved in the presence of 1.0 mM PDS due to the thermal activation of PDS. Therefore, the cooling water system was indispensable in this study in order to maintain suitable solution temperature for mechanism study. Fig. 1 showed the degradation curves of BPS in different US systems. No buffer solutions were added in this study unless otherwise noted. As can be seen, the removal of BPS was negligible in the presence of 100 mM methyl alcohol at oxic conditions, indicating that pyrolysis in US contribute little to BPS disappearance and BPS degradation was mainly caused by the ROS generated in US. As shown, 28.67% removal of BPS was observed in US at oxic condition, but only 11.97% BPS was degraded when DO was wiped off. Comparatively, in the presence of PDS, the removal rate of BPS slightly increased from 28.67% to 33.35%. Apparently, the presence of DO (0.29 mM in US system) was more effective than PDS (1.0 mM in US system) for BPS degradation, although PDS (ES2O82-/ SO42- ¼ 2.01 V) possesses much higher oxidation potential than DO (EO2/OH- ¼ 0.40 V). For the above-mentioned phenomena, the influences of DO and PDS in US for pollutants degradation were studied respectively.

3

3.2. Role of DO for BPS degradation in US system 3.2.1. Kinetics study of BPS degradation in different US systems As shown in Fig. 2a, four conditions described in previous section with different BPS concentration (10e80 mM) were conducted to explore the function of DO in US system. The BPS degradation rate increased with increasing BPS concentration in all conditions and the order was condition 3 > condition 4 > condition 1 > condition 2 When BPS concentration fall below 80 mM, the solution with DO seem to possess higher BPS removal efficiency than without DO under different circumstances. Specifically, the BPS degradation rate of the condition 4 was only slightly higher than that of condition 1 when 80 mM of BPS was added. Generally, the causes of gases have a major influence on sonochemical activity come from the following three aspects: gas specific heat ratio, gas thermal conductivity, and gas solubility. The first two factors influenced the collapse properties of cavitation bubble (i.e. temperatures and pressures), and the third affected the amount of cavitation bubbles (i.e. higher soluble gases provided more nucleation sites for cavitation.). Different aeration processes not only introduced gases with different solubility (O2>N2), but revealed the importance of O2. The zero-order reactions for pollutants were observed in thermo-activated persulfate system as the concentration of the generated SO 4 was much lower than that of pollutants (Liang and Bruell, 2008; Wang et al., 2017) (i.e. the degradation rates were independent on the concentration of pollutants). In US, the pollutants degradation rates were increased with their increasing initial concentration due to the presence of cavitation bubbles and local high OH. More cavitation bubbles and higher local OH make US system more heterogeneous and may improve the degradation of high concentration pollutants. Therefore, as shown in Fig. 2b, the Dr was calculated from Fig. 2a (Eq. (1)) to evaluate the degree of heterogeneity in the four systems. At first, the values ofDr with aeration were both higher than that ofDr without gas bubbled, perhaps due to the intensification of hydrolysis, pyrolysis, and more cavitation bubbles generated when aerating, result in the increased BPS abatement at high BPS concentration. Secondly, and more importantly, the condition 1 possess two times theDr of condition 2 without introduction of gas, possibly caused by higher local ROS concentration.

Dr ¼

1
r20  r10 r40  r20 r80  r40  þ þ 3 10 20 40

(1)

where, ri0 (i ¼ 1, 2, 4 and 8) were the BPS degradation rate at given BPS concentration (i.e. 10, 20, 40 and 80 mM). 3.2.2. The generation of O 2 and OH in different US systems In US, many research ascribed the improved ultrasonic activity with DO to the enhanced formation of reactive species, but the detailed mechanism was still unclear. Superoxide radical ion (O 2 ), which can be generated from sonolysis (Hayyan et al., 2016), was first taken into consideration. We conjectured that O2 react with the pollutants via direct electron transfer reaction is the possible pathway of O 2 generation. (Eq. (2)). As shown in Fig. 3a, the yield of O 2 at condition 1 was higher than condition 2, but the fluorescence intensity ratio is inconsistent with the concentration ratio of DO. Therefore, the generation of O 2 was not rely on the DO, and O 2 was not likely the dominating ROS in the presence of DO (the possible pathway of O 2 generation was discussed in the end of this section). Fig. 1. Influence of PDS, DO and MeOH on BPS degradation. Conditions: [PDS]0 ¼ 1.0 mM, [BPS]0 ¼ 10 mM, [MeOH] ¼ 100 mM, [O2]0 ¼ 0.29 mM, ultrasound power, 60.0 W.

O2 þ R/Rþ þ O, 2

(2)

4

X. Lu et al. / Water Research 165 (2019) 114969

Fig. 2. Influence of different BPS concentration on BPS degradation rates in various conditions (a), the values of Dr in different conditions (b). Conditions: [BPS]0 ¼ 10e80 mM, [BA]0 ¼ 100 mM, ultrasound power, 60.0 W.

The conversion of benzoic acid (BA) to p-hydroxybenzoic acid (p-HBA) by OH as a probe reaction for quantitative analysis of OH has been frequently-used (Joo et al., 2005). To assess the relative contribution of OH for pollutants degradation, thereby understanding the ratio of OH to ROS at different conditions, the oxidation of BA to p-HBA was applied in this experiment. As shown in Fig. 3b, p-HBA generated at all conditions with the abatement of BA, and increased BA degradation rates were accompanied by more p- HBA production except condition 4, suggesting that the increasing BA degradation did not rely on OH at condition 4. Interestingly, as shown in Fig. 3c, the value of DBA/Dp-HBA (demand of BA for per mole p-HBA generation) for condition 1 and 2 were almost equal, which indicate that the OH remained the predominant ROS for pollutants degradation in the presence of DO. Hua et al. studied the hydrolysis of p-Nitrophenyl acetate in ambient conditions and US system, the results shown that the formation of transient supercritical water was crucial for the acceleration of the hydrolysis in the presence of ultrasound (Hua et al., 1995). They predicted that approximately 0.15% of the irradiated water was in the supercritical state by a simple heat-transfer model. Li et al. proposed radical reaction mechanism for oxygen role in Supercritical Water Oxidation (SCWO) (Eqs. (3)e(6)). Oxygen transformed into hydrogen peroxide by hydrogen abstraction along with the pollutants destruction, followed by conversion of hydrogen peroxide through pyrolysis to hydroxyl radicals oxidizing pollutants quickly (Li et al., 1991).

RH þ O2 /R, þ HO2 ,

(3)

RH þ HO2 ,/R, þ H2 O2

(4)

H2 O2 / 2HO,

(5)

RH þ HO,/R, þ H2 O

(6)

As shown in Fig. 3d, hydrogen peroxide was generated in US in the order condition 3 > 1>2 > 4. The mechanism proposed by Li et al. may not be reliable in the US because the limited hydrogen peroxide concentration was unlikely to be the source of OH. Besides, HO2 and H were not observed in US at O2-saturated condition by ESR experiments in previous study (Kohno et al., 2011), suggesting that H was consumed due to the presence of DO and

Eq. (3) or Eq. (7) could be neglected. In order to verify the scavenging role of oxygen towards H, CCl4 was chosen as competitive scavenger of oxygen to conduct experiments at condition 1 and 2. Due to the different reactivity of CCl4 with H and OH (Eqs. (8) and (9)), the dechlorination should be different. As shown in Fig. 3e, the condition 2 produced slightly higher concentrations of chloride ion than condition 1 at 5, 10 and 15 min, indicating that the oxygen quenched the H indeed. Fischer et al. studied the isotopic distribution of O2 and H2O2 in sonication of water in the presence of 18,18 O2. The results shown that although possible isotopic O2 and H2O2 were observed, the 16,16O-containing molecules were the domination (Fischer et al., 1986). So, the possible explanation in this study was that DO as hydrogen radical scavenger contributed to the OH formation by reducing the recombination of H and OH and leaving OH reactive in the system (Eq. (10)).

,H þ O2 /HO2 , ,H þ CCl4 /,CCl3 þ Hþ þ Cl ,OH þ CCl4 /no reaction

(7) k ¼ 3:8
107 M 1 S1

(8) (9)

,H þ O2 /HO, þ O

(10)

O2 42O

(11)

2 , H þ O/H2 O

(12)

The superoxide radical ion may be formed as a result of reaction of OH and hydrogen peroxide in US (Eq. (13) and (14)) (Buxton et al., 1988). In order to verify this assumption, the data of Fig. 3b were used in quantitative analysis according to the fluorescence intensity ratio of two conditions (1 and 2). Specifically, the ratio of formed p-HBA in two conditions was equal to that of steady state concentration of OH in this two conditions and also two experiments of superoxide radical ion determination (Text S3). As the hydrogen peroxide formed mainly by the conjunction of OH, the rate of superoxide radical ion formation was proportional to the cube of OH concentration. Accordingly, the ratio of superoxide radical ion formation rate was theoretically equal to 2.42 (i.e., (0.616/0.459)3), which was quite close to the value of 2.37 (i.e., the

X. Lu et al. / Water Research 165 (2019) 114969

5

Fig. 3. The fluorescence intensity of adduct of NBD-Cl with O2 in different time (a), the BA degradation and corresponding p-HBA formation in various conditions (b), the values of DBA/Dp-HBA in different conditions (c), the yield of H2O2 in various conditions (d), the concentration of chloride ion in condition 1 and 2 (e). Conditions: [NBD-Cl]0 ¼ 100 mM, [CCl4]0 z 200 mM, oxygen-free or oxygen enrichment, ultrasound power, 60.0 W.

ratio of 144.25 to 60.75 in Fig. 3a), indicating the reasonableness of the assumption.

þ HO , þ H2 O2 /O, 2 þ H2 O þ H

(13)

2HO , /H2 O2

(14)

3.3. Mechanism of PDS activation in US system 3.3.1. Determination of rate constants for the reactions of BPS with OH and SO 4 Since the second order rate constants of BPS with OH (kOH, BPS)  and SO 4 (k SO4 , BPS) affected the investigation of PDS effects in US system, the competition kinetics were applied to determine the unknown second order rate constants. Take benzoic acid (BA) as the reference substance in the UV/PDS and UV/H2O2 process for kOH,  BPS and kSO4, BPS measurement, and the second order rate constants  9 1 1 of BA with OH (kOH, BA) and SOs 4 (kSO4, BA) are 5.9
10 M 9 1 1 and 1.2
10 M s (Neta et al., 1988), respectively. More details

6

X. Lu et al. / Water Research 165 (2019) 114969

are shown in Text S1. According to Eqs. (15) and (16), the kOH, BPS  and k SO4, BPS could be calculated by the kOH, BA, kSO4, BA and the slopes that obtained from Fig. S4.

      k,OH;BPS ½BPSt ½BAt  kUV;BPS t ¼ ln  kUV;BA t ln ½BPS0 k,OH;BA ½BA0

(15)

      kSO4,;BPS ½BPSt ½BAt  kUV;BPS t ¼ ln  kUV;BA t ln ½BPS0 kSO4,;BA ½BA0 (16) where, kUV, BPS and kUV, BA are the pseudo first-order reaction rate constants of BPS and BA direct photolysis. As a result, the second order rate constants of BPS with OH and SO 4 calculated from this experiment were 5.5
109 M1 s1 and 3.1
109 M1 s1, respectively. 3.3.2. Effect of PDS concentration on BPS degradation in US system For PDS activation mechanism by US, thermal dissociation of PDS at localized “hot spots” that generated by collapse of cavitation bubble was generally accepted. Another possible activation mechanism involved the OH generated from water dissociation in US, that was, OH reacted with PDS forming SO 4 , as described Eq. (17) (Neppolian et al., 2010). The reaction occurred surely in presence of PDS in US, but could be negligible for PDS activation due to the reaction did not increase the amount of ROS in US. In this experiment, the value of kROS, BPSC[BPS] is three times larger than that kROS, PDSC[PDS], which means more than 75% of ROS were consumed for BPS degradation when 16 mM PDS was applied in US (kOH, BPS and  k SO4, BPS were determined in this work, and kOH, PDS and kSO4, PDS were about 106 M1 s1 and 6.6
105 M1 s1, respectively.).

1 2 , þ S2 O2 8 þ ,OH/H þ SO4 þ SO4 þ O2 2

(17)

In order to explore the kinetics of PDS activation by US, different concentration PDS were applied at anoxic conditions in US for BPS degradation. It was clearly seen from Fig. 4 that the increased PDS concentration result in faster BPS degradation, but potentiation was not remarkable with multiplying PDS dosages. The degradation rate of BPS was expressed by average removal concentration in unit time. As shown in insert of Fig. 4, the ln [kUS/PDS - kUS] of BPS

Fig. 4. Influence of initial PDS concentration on kinetics of BPS degradation. Insert: the kinetic orders of PDS for BPS degradation in US. Conditions: [BPS]0 ¼ 10 mM, oxygenfree, ultrasound power, 60.0 W.

degradation by PDS activation was linearly correlated to the ln [PDS]0 with a slope of 0.563. Given the PDS thermal activation mechanism in US, the slope of 0.563 conflicts with the result obtained from thermo-activated persulfate (TAP) system, where a slope of 1.0 was observed (SI Fig. S5). Obviously, PDS activation in US was not effective as in TAP system, which may be due to the heterogeneity of cavitation bubbles in US system. 3.3.3. Heterogeneous PDS activation model The cavitation bubbles in US, led to the non-uniform distribution of ROS and target compound. Consequently, the degradation of target compound in US was concentration dependent, i.e., increased target compound usually exhibits more rapid degradation rate. Many heterogeneous reaction kinetics models were proposed in order to simulate the target compound behavior during sonication (Chiha et al., 2010; Hamdaoui and Naffrechoux, 2007; Okitsu et al., 2005), and a Langmuir-type mechanism was used widely. The target compound decomposition was supposed to occur in a layer region, i.e., so called gas-liquid interface of the cavitation bubble. The assumptions of Langmuir-type kinetic model and derivation of formulas were shown in Text S4. A linear relationship between the 1/r and 1/C0 was formulated:

r ¼ kq ¼

kKC0 1 þ KC0

(18)

As shown in Fig. S6 (a), the kinetic degradation of BPS in US at oxic condition was conducted over a range of concentrations from 4.0 mM to 15.0 mM. As can been seen from Fig. S6 (b), the linear relationship between 1/r and C0 was in accordance with Eq. (18), and the equilibrium constant K of BPS was calculated to be 0.079 mM1 according to the intercept and slope, which was close to that of trans-isomer 4-methylcyclohexanemethanol (MCHM) in previous research (Cui et al., 2017). Analogously, the heterogeneity of PDS distribution was proposed to describe the behavior of different PDS concentration for BPS degradation as mentioned in the previous section. As detailed in SI Text S5, the function relationship between the BPS degradation rate and the PDS initial concentration was established in heterogeneous PDS activation model:

1 2ðk0 þ k1 Þ 2ðk0 þ k1 Þ ¼ þ r kk0 kKPDS C½PDS0

Fig. 5. 1/r0 as a function of 1/C0.

(19)

X. Lu et al. / Water Research 165 (2019) 114969

A plot of 1/r vs. C[PDS]0 showed excellent linearity with a correlation coefficient R2 ¼ 0.998 (Fig. 5), suggesting that the proposed heterogeneous PDS activation model was reasonable in the US system. It was noteworthy that the value of KPDS was 2.91
104 mM1, which was calculated to be 2 orders of magnitude lower than that of BPS. As the K equals to the ratio of compounds adsorption to desorption rate, it represents the affinity of compounds to the gas-liquid interface reasonably. That is, compounds with high K accumulated at the gas-liquid interface more easily. As the collapse of the bubbles, the “hot spots” were generated with high temperature and pressure (about 5000 K and 1000 atm), such extreme conditions far surpass the critical point of water (647 K and 218atm). Therefore, the gas-liquid interface (reactive zone) can be considered as the supercritical water layer that links the hot-spot regions and the bulk solution (Hua et al., 1995). Because of the high hydrophobicity of supercritical water layer, PDS was hardly compatible with the reactive zone, while most of organics were hydrophobic, result in the very low K of PDS compared with BPS and other target compounds.

Fig. 6. The fitting curve of DMPO-OH adduct formation with sonication time (data obtained from (Wei et al., 2017). Environ Sci Technol.) (Experimental conditions: pH ¼ 7.4, DMPO ¼ 40 mM, Power Density ¼ 1.5 W/L).

3.3.4. Reinvestigate the OH and SO 4 formation in US/PDS In previous study, the OH was considered the dominating radical compared with SO 4 , due to the yield of DMPO  OH in the presence of 100 mM PDS being nearly 3 times larger than DMPO  SO 4 in US/PDS system and DMPO  OH in the single US system only (Wei et al., 2017). The authors attributed the result to the hydrolysis of SO 4 . However, the radicals yield should not be judged by the DPMO adducts steady state concentrations due to the  hydrolysis of DMPO  SO 4 . The reactions between SO4 , OH and DMPO in solution were shown in Table 1. As detailed in SI Text S6, the concentrations of DMPO  OH and DMPO  SO 4 with time can be expressed by following equations:

½DPMO,  OH ¼

K2 h 1 kDO  ekDO t

i

considered to be constant for fitting of K1 and K2 in the US/PDS system (the value of kDS, 0.65 min1, was the average of the two values obtained fitting curves of DMPO  SO 4 , shown in Fig. S7). According to Eqs. (20) and (21), all the DMPO  OH concentrations would be steady after a period of time and the values in US and US/ PDS system were K2/kDO and (K1 þ K2)/kDO, respectively. Given the fixed concentration of DMPO, the K1 and K2 at different conditions can reflect the corresponding radical concentrations (assumed that k1 z k2). In the presence of 1.0 mM PDS, the value of K2 decreased from 2.01 mM/min to 1.82 mM/min and the value of K1 was 2.15 mM/ min, suggesting the decrease of OH concentration and the DMPO  SO 4 hydrolysis accounted for more than half of DMPO  OH steady concentration. Moreover, the value of K1 was 5 times higher than that of K2 when 100.0 mM PDS was used, indi cating the predominance of SO 4 and great role of DMPO  SO4 hydrolysis for higher DMPO  OH steady concentration. Identifying the relative contributions of OH and SO 4 in PDS or PMS activation were always difficult since they were generated simultaneously. Besides, very high response value of DMPO  OH compared to DMPO  SO 4 was observed in ESR experiments. Some reaction rate constants unknown and complicated reaction conditions make it harder to calculate radical concentrations. Although DMPO  SO 4 was transformed into DMPO  OH by nucleophilic substitution (Davies et al., 1992; KIRINO et al., 1981; Timmins et al., 1999; Zamora and Villamena, 2012), clarifying that the DMPO adducts reactions and utilizing the data of DMPO adducts with time can obtain more direct and accurate radical contributions. The method used in our study may be applicable for finding out the ratio of OH or SO in some PDS and PMS 4 activation.

ðIn US systemÞ (20)

½DPMO,  OH ¼ i  ekDS t

i h K1 þ K2 h K1 1  ekDO t þ ekDO t kDO kDO  kDS

(21) 

 K1 h 1 DPMO,  SO 4 ¼ kDS  ekDS t

i

7

ðIn US=PDS systemÞ (22)

where K1 ¼ k1 [DMPO] [SO4 ], K2 ¼ k2 [DMPO] [OH]. As shown in Fig. 6, the DMPO  OH concentrations with sonication time in the presence or absence of PDS were used to obtain the fitting curves by derived equations. The value of kDO and kDS were obtained firstly by the fitted equations, which were

3.4. Influence of different water matrices on BPS degradation The influence of common anion (Cl and HCO 3 ) and humic acid

Table 1 Reactions of DMPO in solutions with SO 4 and OH. reaction

rate constant

 DMPO þ SO, 4 / DMPO,  SO4

SO 4

þ H2 O / DMPO,  OH þ DMPO ,  DMPO þ ,OH / DMPO,  OH DMPO ,  OH / Product

SO2 4

þ

Hþ

k1 (M1 s1) kDS (s1) k2 (M1 s1) kDO (s1)

8

X. Lu et al. / Water Research 165 (2019) 114969

(HA) in US on BPS (0.5 mM and 8.0 mM) degradation were studied. As can been seen in Fig. 7c and f, the BPS degradation were retarded by HA, whatever BPS and HA concentration. HA generally inhibit pollutants degradation in AOPs due to radicals or oxidants depletion. The presence of Cl also inhibit the BPS degradation (Fig. 7a and d), which may be caused by consuming OH and forming less reactive chlorine radicals/free chlorine (ClOH/HOCl, Cl 2 , and Cl) (Luo et al., 2015). Interestingly, HCO 3 retarded BPS degradation when the concentration was 8 mM but improve the degradation of 0.5 mM BPS  (Fig. 7b and e). HCO 3 can react with OH forming CO3 (Eq. (35)) (Buxton et al., 1988). Although the second-order rate constant of reaction between CO 3 and organic pollutants being lower that of OH with organic pollutants, CO 3 may play a positive role on pollutants degradation in OH e based AOPs (Merouani et al., 2010; Petrier et al., 2010) due to its higher steady state concentration compared with OH (Luo et al., 2015). In the US system, the HCO 3 play dual role on BPS degradation which depends on BPS concentration. As mentioned above, lower BPS concentration in US presented a lower degradation rates. For example, the BPS degradation rate decreased from 0.0431 mM/min to 0.00908 mM/min when the BPS concentration was switched from 8.0 mM to 0.5 mM. More OH self-reacted (Eq. (14)) due to less OH was consumed by BPS. Therefore, in the presence of HCO 3 , Eq. (35) may enhanced by more available OH, which is possibly favorable for BPS degradation when low BPS concentration was applied. The likely explanation for positive role of HCO 3 on low concentration of BPS degradation is as follows: 1. HCO 3 react with OH at gas-liquid interface forming CO 3 has little negative influence on BPS degradation at gas-liquid interface due to low BPS concentration. 2. Higher stability of CO 3 compared to OH (Eq. (24) and (25)) (Buxton et al., 1988; Neta et al., 1988) result in higher steady state CO 3 concentration and enable CO 3 to spread to bulk solution, which bring bout considerable BPS degradation in bulk solution.

, ,OH þ HCO 3 /H2 O þ CO3

8:6
106 M1 s1 (23) 2k ¼ 1:1
1010 ,M1 ,s1

2HO , /H2 O2

(24) 2 2CO, 3 / CO2 þ C2 O4 7

¼ 4:0
10 ,M

1

,s

2k

1

(25)

3.5. BPS degradation products and acute toxicity evaluation A total of five oxidation products of BPS were detected and analyzed by LC-QTOF-MS/MS in US system, and their constitutional formulas were identified by the accurate mass and fragment peaks of degradation intermediates in the second- order mass spectrum. Fig. 8 showed the mass spectra of BPS and five Oxidation products. OP-497, which also detected in our previous study (Wang et al., 2017), can be formed by combination of BPS radicals and considered as the function of SO 4 due to its electron abstraction reaction. In US system, the dipolymer of BPS was also formed in the absence of SO 4 , while the dipolymer of BPA was not found in the US/PDS system (Darsinou et al., 2015), suggesting that BPS more likely form dipolymer than BPA. OP-265, the hydroxylated product of BPS, was produced by the OH generated in US. OP-173, p-hydroxybenzenesulfonic acid, can be formed by combination of OH and 4-hydroxybenzenesulfinate radicals, formed by the b-Scission of BPS radicals, it can also be produced by hydrolysis of BPS in US. The relative concentration of BPS and five oxidation products during reactions were shown in Fig. S8.

Fig. 7. Effects of (a) (d) chloride, (b) (e) bicarbonate, and (c) (f) HA on BPS oxidation in US system. Conditions: [BPS]0 ¼ 0.5/8.0 mM, oxygen enrichment, ultrasound power, 60.0 W.

X. Lu et al. / Water Research 165 (2019) 114969

9

Fig. 8. The tandem mass spectra for BPS and its degradation products.

The photobacterium of Vibrio fischeri was used in posttreatment acute toxicity evaluation of BPS in US. As shown in Fig. 9, L0 was the luminescence intensity at 25 mM BPS without treatment and L was the luminescence intensity of simples at different degradation levels or conditions. The acute toxicity of simples can be evaluated by the relative inhibition values of luminescence intensity (L/L0), where a higher value means lower acute toxicity, and vice versa. The dashed line represented the acute toxicity of samples that only have different BPS concentration. The value of L/L0 when no BPS existed was 3.03, suggesting the percent relative inhibition value of 25 mM BPS equal 33.0%. The value was below 58% of Olmez-Hanci et al., (2013), where 88 mM BPA was applied, indicating that BPS is more toxic than BPA. Annoyingly, the treated samples of BPS all become more toxic compared with originally BPS. For example, the inhibition effects of luminescence intensity up to 50%, where only approximately 25% BPS removal was realized. Then, 34% degradation of BPS led to 96% inhibition of luminescence intensity. When over 40% BPS was degraded, the

luminescence intensity of samples was inhibited completely. These results showed that more toxic products were generated in process of BPS degradation induced by US. Therefore, further research about BPS removal and its detoxification efficiency are needed considering the wide application of BPS.

4. Conclusions The treatment of BPS by US was studied in this experiment by examining the effects of dissolved oxygen and peroxydisulfate, oxidation products, and the post-treatment acute toxicity. The main conclusions are summarized as follows: (i) The dissolved oxygen improve BPS degradation by enhancing OH generation while formation of O 2 was excluded. The DO as scavenger of H played important role in intensified oxidation ability of US.

10

X. Lu et al. / Water Research 165 (2019) 114969

Fig. 9. Influence of pre-and post-treatment BPS on Vibrio fischeri luminescence. Conditions: [BPS]0 ¼ 25 mM.

(ii) The relationship between BPS degradation rates and PDS concentration fit the Langmuir-type kinetic model well, and the low value of KPDS indicates that PDS shows low activation efficiency in US system due to its high dielectric constant. The formation of OH and SO 4 in US/PDS system were reinvestigated, the result show that SO 4 rather than OH was the predominant radical. (iii) BPS mainly undergoes three degradation pathways in US system including BPS radical polymerization, hydroxylation and hydrolysis, the representative products were OP-497, OP-265 and OP-173, respectively. The acute toxicity evaluation experiments reveal that more toxic products were formed in the process of BPS degradation induced by US.

Conflicts of interest We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the manuscript entitled ‘Sonolytic degradation of Bisphenol S: Effect of dissolved oxygen and peroxydisulfate, oxidation products and acute toxicity’. Acknowledgments This work was financially supported by the National Key R&D Program of China (2017YFA0207203) and Open Project of State Key Laboratory of Urban Water Resource and Environment, Harbin Institute of Technology(No. HC201812). Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.watres.2019.114969. References Anipsitakis, G.P., Dionysiou, D.D., 2003. Degradation of organic contaminants in water with sulfate radicals generated by the conjunction of peroxymonosulfate with cobalt. Environ. Sci. Technol. 37 (20), 4790e4797. Anipsitakis, G.P., Dionysiou, D.D., 2004. Radical generation by the interaction of transition metals with common oxidants. Environ. Sci. Technol. 38 (13), 3705e3712. Barik, A.J., Gogate, P.R., 2017. Degradation of 2,4-dichlorophenol using combined

approach based on ultrasound, ozone and catalyst. Ultrason. Sonochem. 36, 517e526. Berberidou, C., Poulios, I., Xekoukoulotakis, N.P., Mantzavinos, D., 2007. Sonolytic, photocatalytic and sonophotocatalytic degradation of malachite green in aqueous solutions. Appl. Catal. B Environ. 74 (1e2), 63e72. Bokare, A.D., Choi, W., 2015. Singlet-oxygen generation in alkaline periodate solution. Environ. Sci. Technol. 49 (24), 14392e14400. Buxton, G.V., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988. Critical Review of rate constants for reactions of hydrated electrons, hydrogen atoms and hydroxyl radicals OH/,O in Aqueous Solution. J. Phys. Chem. Ref. Data 17 (2), 513e886. Chen, W.S., Su, Y.C., 2012. Removal of dinitrotoluenes in wastewater by sonoactivated persulfate. Ultrason. Sonochem. 19 (4), 921e927. Chiha, M., Merouani, S., Hamdaoui, O., Baup, S., Gondrexon, N., Petrier, C., 2010. Modeling of ultrasonic degradation of non-volatile organic compounds by Langmuir-type kinetics. Ultrason. Sonochem. 17 (5), 773e782. Cui, D., Mebel, A.M., Arroyo-Mora, L.E., Holness, H., Furton, K.G., O'Shea, K., 2017. Kinetic, product, and computational studies of the ultrasonic induced degradation of 4-methylcyclohexanemethanol (MCHM). Water Res. 126, 164e171. Danzl, E., Sei, K., Soda, S., Ike, M., Fujita, M., 2009. Biodegradation of bisphenol A, bisphenol F and bisphenol S in seawater. Int. J. Environ. Res. Public Health 6 (4), 1472e1484. Darsinou, B., Frontistis, Z., Antonopoulou, M., Konstantinou, I., Mantzavinos, D., 2015. Sono-activated persulfate oxidation of bisphenol A: kinetics, pathways and the controversial role of temperature. Chem. Eng. J. 280, 623e633. Davies, M.J., Gilbert, B.C., Stell, J.K., Whitwood, A.C., 1992. Nucleophilic substitution reactions of spin adducts. Implications for the correct identification of reaction intermediates by EPR/spin trapping. J. Chem. Soc. Perkin Trans. 2 (3), 333e335. Eladak, S., Grisin, T., Moison, D., Guerquin, M.J., N'Tumba-Byn, T., Pozzi-Gaudin, S., Benachi, A., Livera, G., Rouiller-Fabre, V., Habert, R., 2015. A new chapter in the bisphenol A story: bisphenol S and bisphenol F are not safe alternatives to this compound. Fertil. Steril. 103 (1), 11e21. Ferkous, H., Merouani, S., Hamdaoui, O., Petrier, C., 2017. Persulfate-enhanced sonochemical degradation of naphthol blue black in water: evidence of sulfate radical formation. Ultrason. Sonochem. 34, 580e587. Fernandez-Alba, A.R., Hernando, D., Agüera, A., Caceres, J., Malato, S., 2002. Toxicity assays: a way for evaluating AOPs efficiency. Water Res. 36 (17), 4255e4262. Fischer, C.H., Hart, E.J., Henglein, A., 1986. Ultrasonic irradiation of water in the presence of 18, 18O2: isotope exchange and isotopic distribution of H202. J. Phys. Chem. 90 (9), 1954e1956. Gao, Y., Jiang, J., Zhou, Y., Pang, S.Y., Ma, J., Jiang, C., Yang, Y., Huang, Z.S., Gu, J., Guo, Q., Duan, J.B., Li, J., 2017. Chlorination of bisphenol S: kinetics, products, and effect of humic acid. Water Res. 131, 208e217. Goel, M., Hongqiang, H., Mujumdar, A.S., Ray, M.B., 2004. Sonochemical decomposition of volatile and non-volatile organic compounds–a comparative study. Water Res. 38 (19), 4247e4261. Grignard, E., Lapenna, S., Bremer, S., 2012. Weak estrogenic transcriptional activities of Bisphenol A and Bisphenol S. Toxicol. In Vitro 26 (5), 727e731. Guo, H., Li, H., Liang, N., Chen, F., Liao, S., Zhang, D., Wu, M., Pan, B., 2016. Structural benefits of bisphenol S and its analogs resulting in their high sorption on carbon nanotubes and graphite. Environ. Sci. Pollut. Res. Int. 23 (9), 8976e8984. Guo, W.Q., Yin, R.L., Zhou, X.J., Du, J.S., Cao, H.O., Yang, S.S., Ren, N.Q., 2015. Sulfamethoxazole degradation by ultrasound/ozone oxidation process in water: kinetics, mechanisms, and pathways. Ultrason. Sonochem. 22, 182e187. Hamdaoui, O., Naffrechoux, E., 2007. Modeling of adsorption isotherms of phenol and chlorophenols onto granular activated carbon. Part I. Two-parameter models and equations allowing determination of thermodynamic parameters. J. Hazard Mater. 147 (1e2), 381e394. Hamdaoui, O., Merouani, S., 2017. Improvement of sonochemical degradation of Brilliant blue R in water using periodate ions: implication of iodine radicals in the oxidation process. Ultrason. Sonochem. 37, 344e350. Hayyan, M., Hashim, M.A., AlNashef, I.M., 2016. Superoxide ion: generation and chemical implications. Chem. Rev. 116 (5), 3029e3085. Heller, M.I., Croot, P.L., 2010. Application of a superoxide (O(2)(-)) thermal source (SOTS-1) for the determination and calibration of O(2)(-) fluxes in seawater. Anal. Chim. Acta 667 (1e2), 1e13. Hua, I., Hoechemer, R.H., Hoffmann, M.R., 1995. Sonolytic hydrolysis of p-nitrophenyl acetate: the role of supercritical water. J. Phys. Chem. 99 (8), 2335e2342. Ike, M., Chen, M.Y., Danzl, E., Sei, K., Fujita, M., 2006. Biodegradation of a variety of bisphenols under aerobic and anaerobic conditions. Water Sci. Technol. 53 (6), 153e159. Ince, N.H., 2018. Ultrasound-assisted advanced oxidation processes for water decontamination. Ultrason. Sonochem. 40, 97e103. Jin, H., Zhu, L., 2016. Occurrence and partitioning of bisphenol analogues in water and sediment from liaohe river basin and taihu lake, China. Water Res. 103, 343e351. Joo, S.H., Feitz, A.J., Sedlak, D.L., 2005. Quantification of the oxidizing capacity of nanoparticulate zero-valent iron. Environ. Sci. Technol. 39 (5), 1263e1268. KIRINO, Y., OHKUMA, T., KWAN, T., 1981. Spin trapping with 5, 5-dimethylpyrrolineN-oxide in aqueous solution. Chem. Pharm. Bull. 29 (1), 29e34. Kohno, M., Mokudai, T., Ozawa, T., Niwano, Y., 2011. Free radical formation from sonolysis of water in the presence of different gases. J. Clin. Biochem. Nutr. 49 (2), 96e101. Lee, H., Yoo, H.Y., Choi, J., Nam, I.H., Lee, S., Lee, S., Kim, J.H., Lee, C., Lee, J., 2014. Oxidizing capacity of periodate activated with iron-based bimetallic nanoparticles. Environ. Sci. Technol. 48 (14), 8086e8093.

X. Lu et al. / Water Research 165 (2019) 114969 Lee, Y.C., Chen, M.J., Huang, C.P., Kuo, J., Lo, S.L., 2016. Efficient sonochemical degradation of perfluorooctanoic acid using periodate. Ultrason. Sonochem. 31, 499e505. Li, B., Li, L., Lin, K., Zhang, W., Lu, S., Luo, Q., 2013. Removal of 1,1,1-trichloroethane from aqueous solution by a sono-activated persulfate process. Ultrason. Sonochem. 20 (3), 855e863. Li, J., Pang, S.Y., Zhou, Y., Sun, S., Wang, L., Wang, Z., Gao, Y., Yang, Y., Jiang, J., 2018. Transformation of bisphenol AF and bisphenol S by manganese dioxide and effect of iodide. Water Res. 143, 47e55. Li, L., Chen, P., Gloyna, E.F., 1991. Generalized kinetic model for wet oxidation of organic compounds. AIChE J. 37 (11), 1687e1697. Liang, C.J., Bruell, C.J., 2008. Thermally activated persulfate oxidation of trichloroethylene: experimental investigation of reaction orders. Ind. Eng. Chem. Res. 47 (9), 2912e2918. Liao, C., Kannan, K., 2014. A survey of bisphenol A and other bisphenol analogues in foodstuffs from nine cities in China. Food Addit. Contam. Part A Chem Anal Control Expo Risk Assess 31 (2), 319e329. Liao, C., Liu, F., Kannan, K., 2012a. Bisphenol s, a new bisphenol analogue, in paper products and currency bills and its association with bisphenol a residues. Environ. Sci. Technol. 46 (12), 6515e6522. Liao, C., Liu, F., Alomirah, H., Loi, V.D., Mohd, M.A., Moon, H.B., Nakata, H., Kannan, K., 2012b. Bisphenol S in urine from the United States and seven Asian countries: occurrence and human exposures. Environ. Sci. Technol. 46 (12), 6860e6866. Lin, S.-S., Gurol, M.D., 1998. Catalytic decomposition of hydrogen peroxide on iron oxide: kinetics, mechanism, and implications. Environ. Sci. Technol. 32 (10), 1417e1423. Luo, C., Ma, J., Jiang, J., Liu, Y., Song, Y., Yang, Y., Guan, Y., Wu, D., 2015. Simulation and comparative study on the oxidation kinetics of atrazine by UV/H(2)O(2), UV/HSO(5)(-) and UV/S(2)O(8)(2)(-). Water Res. 80, 99e108. Matilainen, A., Sillanpaa, M., 2010. Removal of natural organic matter from drinking water by advanced oxidation processes. Chemosphere 80 (4), 351e365. McNamara, W.B., Didenko, Y.T., Suslick, K.S., 1999. Sonoluminescence temperatures during multi-bubble cavitation. Nature 401 (6775), 772e775. Merouani, S., Hamdaoui, O., Saoudi, F., Chiha, M., Petrier, C., 2010. Influence of bicarbonate and carbonate ions on sonochemical degradation of Rhodamine B in aqueous phase. J. Hazard Mater. 175 (1e3), 593e599. Neppolian, B., Doronila, A., Ashokkumar, M., 2010. Sonochemical oxidation of arsenic(III) to arsenic(V) using potassium peroxydisulfate as an oxidizing agent. Water Res. 44 (12), 3687e3695. Neta, P., Huie, R.E., Ross, A.B., 1988. Rate constants for reactions of inorganic radicals in aqueous solution. J. Phys. Chem. Ref. Data 17 (3), 1027e1284. Okitsu, K., Iwasaki, K., Yobiko, Y., Bandow, H., Nishimura, R., Maeda, Y., 2005. Sonochemical degradation of azo dyes in aqueous solution: a new heterogeneous kinetics model taking into account the local concentration of OH radicals and azo dyes. Ultrason. Sonochem. 12 (4), 255e262. Olmez-Hanci, T., Arslan-Alaton, I., Genc, B., 2013. Bisphenol A treatment by the hot persulfate process: oxidation products and acute toxicity. J. Hazard Mater. 263 Pt 2, 283e290.

11

Petrier, C., Torres-Palma, R., Combet, E., Sarantakos, G., Baup, S., Pulgarin, C., 2010. Enhanced sonochemical degradation of bisphenol-A by bicarbonate ions. Ultrason. Sonochem. 17 (1), 111e115. Price, G.J., Clifton, A.A., 1996. Sonochemical acceleration of persulfate decomposition. Polymer 37 (37), 3971e3973. Sellers, R.M., 1980. Spectrophotometric determination of hydrogen peroxide using potassium titanium(IV) oxalate. Analyst 105 (1255), 950. Shen, Y.S., Ku, Y., Lee, K.C., 1995. The effect of light absorbance on the decomposition of chlorophenols by ultraviolet radiation and u.v./H2O2 processes. Water Res. 29 (3), 907e914. Staehelin, J., Hoigne, J., 1985. Decomposition of ozone in water in the presence of organic solutes acting as promoters and inhibitors of radical chain reactions. Environ. Sci. Technol. 19 (12), 1206e1213. Timmins, G.S., Liu, K.J., Bechara, E.J.H., Kotake, Y., Swartz, H.M., 1999. Trapping of free radicals with direct in vivo EPR detection: a comparison of 5,5-dimethyl-1pyrroline-N-oxide and 5-diethoxyphosphoryl-5-methyl-1-pyrroline-N-oxide as spin traps for HO* and SO4*-. Free Radical Biol. Med. 27 (3e4), 329e333. Ullah, H., Jahan, S., Ain, Q.U., Shaheen, G., Ahsan, N., 2016. Effect of bisphenol S exposure on male reproductive system of rats: a histological and biochemical study. Chemosphere 152, 383e391. Wang, D., Xu, H., Ma, J., Lu, X., Qi, J., Song, S., 2018. Strong promoted catalytic ozonation of atrazine at low temperature using tourmaline as catalyst: influencing factors, reaction mechanisms and pathways. Chem. Eng. J. 354, 113e125. Wang, Q., Lu, X., Cao, Y., Ma, J., Jiang, J., Bai, X., Hu, T., 2017. Degradation of Bisphenol S by heat activated persulfate: kinetics study, transformation pathways and influences of co-existing chemicals. Chem. Eng. J. 328, 236e245. Wang, S., Zhou, N., 2016. Removal of carbamazepine from aqueous solution using sono-activated persulfate process. Ultrason. Sonochem. 29, 156e162. Weavers, L.K., Hua, I., Hoffmann, M.R., 1997. Degradation of triethanolamine and chemical oxygen demand reduction in wastewater by photoactivated periodate. Water Environ. Res. 69 (6), 1112e1119. Wei, Z., Villamena, F.A., Weavers, L.K., 2017. Kinetics and mechanism of ultrasonic activation of persulfate: an in situ EPR spin trapping study. Environ. Sci. Technol. 51 (6), 3410e3417. Xu, H., Wang, D., Ma, J., Zhang, T., Lu, X., Chen, Z., 2018. A superior active and stable spinel sulfide for catalytic peroxymonosulfate oxidation of bisphenol S. Appl. Catal. B Environ. 238, 557e567. Yang, T., Wang, L., Liu, Y., Huang, Z., He, H., Wang, X., Jiang, J., Gao, D., Ma, J., 2019. Comparative study on ferrate oxidation of BPS and BPAF: kinetics, reaction mechanism, and the improvement on their biodegradability. Water Res. 148, 115e125. Yin, R., Guo, W., Wang, H., Du, J., Zhou, X., Wu, Q., Zheng, H., Chang, J., Ren, N., 2018. Enhanced peroxymonosulfate activation for sulfamethazine degradation by ultrasound irradiation: performances and mechanisms. Chem. Eng. J. 335, 145e153. Zamora, P.L., Villamena, F.A., 2012. Theoretical and experimental studies of the spin trapping of inorganic radicals by 5,5-dimethyl-1-pyrroline N-oxide (DMPO). 3. Sulfur dioxide, sulfite, and sulfate radical anions. J. Phys. Chem. A 116 (26), 7210e7218.