Mechanistic insight into the generation of reactive oxygen species in sulfite activation with Fe(III) for contaminants degradation

Journal of Hazardous Materials xxx (xxxx) xxxx

Contents lists available at ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

Mechanistic insight into the generation of reactive oxygen species in sulfite activation with Fe(III) for contaminants degradation Hongyu Donga,b,c, Guangfeng Weia,d,*, Daqiang Yina,b,c, Xiaohong Guana,b,c,* a

State Key Laboratory of Pollution Control and Resources Reuse, Tongji University, Shanghai 200092, People’s Republic of China Shanghai Institute of Pollution Control and Ecological Security, Shanghai 200092, People’s Republic of China c International Joint Research Center for Sustainable Urban Water System, Tongji University, Shanghai 200092, People’s Republic of China d Shanghai Key Laboratory of Chemical Assessment and Sustainability, School of Chemical Science and Engineering, Tongji University, Shanghai 200092, China b

G R A P H I C A L A B S T R A C T

A R T I C LE I N FO

A B S T R A C T

Editor: R. Debora

Since the reactive species during the sulfite activation by Fe(III) (Fe(III)/sulfite process) had not been directly determined and the role of in-situ generated Fe(II) was overlooked, this study evaluated the oxidation performance of the Fe(III)/sulfite process, identified the reactive species, and investigated the role of in-situ generated Fe(II) in this process. The results demonstrated that carbamazepine (CBZ) could be degraded at different sulfite concentrations. Compared to the single-dosing mode, sulfite applied with multiple-dosing mode was beneficial to CBZ removal in this process when the same amount of sulfite was dosed. Fe(II) was rapidly generated and then decayed in this process, which were consistent with the trends of CBZ degradation and sulfite consumption. Electron paramagnetic resonance and scavenging experiments showed that SO4%− was a major oxidant, while HO% also played a significant role in CBZ degradation in this process. The tert-butyl alcohol assay indicated that the generation and re-oxidation of Fe(II) was accompanied with the generation of reactive species. Besides sulfite dosage, CBZ degradation was also affected by initial pH, Fe(III) dosage, and CBZ concentration. Cl− showed little inhibition on CBZ degradation while humic acid inhibited CBZ degradation in this process. This study advances the application of this oxidation system.

Keywords: Sulfate radical Hydroxyl radical Fe(III)/Fe(II) cycle Advanced oxidation process Mechanism

1. Introduction The

⁎

transition-metal

catalyzed

oxidation

of

S(IV)

(sulfite,

predominantly HSO3− at pH 1.86–7.2 and SO32− at pH > 7.2) in the presence of O2 has been an object of research for more than a century (Brandt and van Eldik, 1995). The most commonly used transition-

Corresponding authors at: State Key Laboratory of Pollution Control and Resources Reuse, Tongji University, Shanghai 200092, People’s Republic of China. E-mail addresses: [email protected] (G. Wei), [email protected] (X. Guan).

https://doi.org/10.1016/j.jhazmat.2019.121497 Received 17 July 2019; Received in revised form 17 October 2019; Accepted 17 October 2019 0304-3894/ © 2019 Elsevier B.V. All rights reserved.

Please cite this article as: Hongyu Dong, et al., Journal of Hazardous Materials, https://doi.org/10.1016/j.jhazmat.2019.121497

Journal of Hazardous Materials xxx (xxxx) xxxx

H. Dong, et al.

et al., 2019) and later UV radiation was introduced to this system to enhance the degradation of organic contaminants (Xu et al., 2016; Guo et al., 2013). In the Fe(III)/sulfite and UV-Fe(III)/sulfite systems, SO4%− was proposed as the main reactive oxidative species (ROS) based on radical quenching experiments (Yu et al., 2016; Zhou et al., 2015; Xu et al., 2016; Guo et al., 2013). However, no direct evidence was offered for the generation of SO4%− and/or HO% in the Fe(III)/sulfite and UV-Fe (III)/sulfite systems. Additionally, these literatures did not mention the transformation of iron species in the Fe(III)/sulfite and UV-Fe(III)/sulfite systems, which may be crucial to radicals generation in these systems. Therefore, it is essential to determine the origin and sink of radicals in order to understand the chemistry of Fe(III)/sulfite system for degrading various organic pollutants so as to develop strategies to improve the performance of this system. Therefore, taking carbamazepine (CBZ), a typical antiepileptic drug and one of the most frequently detected emerging contaminants in aquatic environment, as a target contaminant, this study (i) assessed the performance of CBZ degradation in the Fe(III)/sulfite process; (ii) identified ROS responsible for CBZ degradation and quantified the contributions of ROS to contaminants degradation in the Fe(III)/sulfite process; (iii) determined the origin and yields of ROS in the Fe(III)/ sulfite process; and (iv) evaluated the influence of various operating factors (i.e. pH, Fe(III) dosage, CBZ dosage, Cl−, and humic acid (HA)) on CBZ degradation in the Fe(III)/sulfite process.

metals include Fe(III), Fe(II), Mn(II), and Co(II) (Brandt et al., 1994; Zhang et al., 2013; Berglund et al., 1993; Karatza et al., 2010). Meanwhile, Fe(III)-catalyzed oxidation of sulfite by O2 has received considerable attention in atmospheric chemistry since it has significant implications for sulfur transformation, acid precipitation, and aerosol nucleation (Kotronarou and Sigg, 1993; Kuo et al., 2006a). Due to the complexity of Fe(III)/sulfite reactions, different mechanisms have been proposed for the Fe(III)/sulfite reactions: non-radical mechanism, radical mechanism, and combined non-radical and radical mechanism (Brandt et al., 1994; Conklin and Hoffmann, 1988; Martin et al., 1991). Up to date, radical mechanism is frequently applied to explain the chemistry of Fe(III)/sulfite reactions where the redox cycling of Fe(III) is fundamental, as depicted in Eqs. 1–11 (Brandt et al., 1994; Yermakov and Purmal, 2003; Ziajka et al., 1994). In general, the initiation step is the formation of a Fe(III)-sulfito complex which decomposed spontaneously to produce Fe(II) and SO3%− (Eqs. 1, 2) (Brandt et al., 1994; Kuo et al., 2006a). Then the generated SO3%− reacts rapidly with O2 to form SO5%− (Eq. 3) (Buxton et al., 1996; Neta et al., 1988), which in turn leads to the formation of other reactive oxysulphur intermediates (Eqs. 4–7) (Neta et al., 1988; Das, 2001; Fischer and Warneck, 1996). The reduced Fe(II) is subsequently re-oxidized to Fe(III) by reactive oxysulphur intermediates via Eqs. 8–10 (Herrmann et al., 1996; Gilbert and Stell, 1990; Lee and Rochelle, 1987).

Fe3 + + HSO−3 ⇌ FeSO+3 + H+ (K=600)

(1)

FeSO+3

(2)

2. Materials and methods

(3)

2.1. Reagents

(4)

The chemical list is provided in Text S1 of Supporting Information (SI). All chemicals were used as received and all solutions were prepared in deionized water (> 18.2 MΩ cm resistivity; Millipore Milli-Q system).

⇌

SO•3−

Fe2 +

+

( k=

0.19s−1)

SO•3− + O2 → SO•5− ( k= 1.0-2.5 × 109M-1s-1) SO•5−

SO•5−

+

HSO−3

+

HSO−3

→

HSO5−

→

SO24−

+

+

SO•3−

SO•4−

( k= 8.6 ×

+

H+

103-3.0

( k= 3.6 ×

×

105M-1s-1)

102-3.0

×

105M-1s-1) (5)

HSO− 5

+

HSO−3

→

2SO24−

+ 2H+ ( k= 1.0 × 103M-1s-1)

(6) 2.2. Kinetics experiments

SO•4− + HSO−3 → SO24− + SO•3− + H+ (k> 2.0 × 109M-1s-1)

(7)

7 -1 -1 Fe2 + + SO•5− + H+ → Fe3 + + HSO− 5 ( k= (4.3 ± 2.4) × 10 M s )

(8)

3 + + SO• − + OH− ( k= 1.0 × 103M-1s-1) Fe2 + + HSO− 5 → Fe 4

(9)

Fe2 + + SO•4− → Fe3 + + SO24− (k=8.6 × 108M-1s-1)

The batch experiments were performed in 250 mL wide-mouth bottles at 20 ± 1 °C (temperature controlled with water bath) with continuously magnetic stirring to monitor changes in the concentrations of target organic contaminant (CBZ), Fe(II), and sulfite during reactions initiated at varying solution conditions in the Fe(III)/sulfite process. Working solution containing Fe(III) and target contaminant(s) was initially prepared and adjusted to the desired initial pH value (pHini) using H2SO4 or NaOH. The NaHSO3 stock solution was pre-adjusted to the same pHini as the working solution. To initiate the reaction, aliquot of NaHSO3 stock solution was added to the working solution. Preliminary experiments demonstrated that the target organic contaminant involved in this study had negligible reaction with Fe(III) alone. At designated time intervals, 5.0 mL aliquots were sampled and immediately quenched using excess sodium thiosulfate stock solution (100 μL, 0.50 M) and filtered (0.22 μm membrane) to subsequently analyze the target organic compound. To quantify the changes in Fe(II) and sulfite concentrations during Fe(III)/sulfite reactions, separate samples were collected and examined immediately without quenching. Experiments were also conducted in N2-sparged solutions to assess the influence of dissolved oxygen (DO) on reactions, the procedure of which was similar to that described above except that the stock solutions were sparged with N2 for 15 min (DO < 0.1 mg L−1) before initiating the reaction and the working solution was always sparged with N2 during the reaction. Selected experiments were also conducted to evaluate the potential influence of methanol (MeOH), tert-butyl alcohol (t-BuOH), carbon tetrachloride (CCl4), Cl-, or HA on reaction kinetics. Unless otherwise noted, batch experiments were conducted in solutions open to the air and performed at least in duplicate. The average values of obtained data with standard deviations are displayed.

(10)

During the investigation of desulfurization, researchers found that organic compounds such as sobrerol inhibited the autoxidation of S(IV), which was attributed to the scavenging of radical and/or complexing of Fe(III) by organic compounds (Ziajka et al., 1994; Lee and Rochelle, 1987; Grgić et al., 1998; Pasiuk-Bronikowska et al., 2003). Meanwhile, the organic compounds were degraded when acting as scavenging species and the scavenged radical mainly referred to SO4%− in the literatures (Ziajka et al., 1994; Lee and Rochelle, 1987). Besides SO4%− proposed in the literatures, HO% could be possibly produced from the reaction of SO4%− with water in the Fe(III)/sulfite process (Eq. 11) (Neta et al., 1988). Both HO% and SO4%− are powerful oxidants with high redox potential (E(SO4%−/SO42−) = 2.5–3.1 V vs NHE, E(HO%/ H2O) = 2.7 V vs NHE under acidic conditions and E(HO%/OH-) = 1.8 V vs NHE under neutral conditions) (Neta et al., 1988; Buxton et al., 1988). HO% is nonselective and reacts rapidly with various organic substrates with second-order rate constants between 106 and 1011 M-1 s1 (Buxton et al., 1988). SO4%− is more selective and generally oxidizes various organic compounds with reaction rate constants between 105 and 109 M-1 s-1 (Neta et al., 1988).

SO•4− + H2 O→ SO24− + HO• + H+ (k< 6.0 × 101M-1s-1)

(11)

In recent years, the Fe(III)/sulfite system was applied to degrade organic contaminants in water (Yu et al., 2016; Zhou et al., 2015; Yuan 2

Journal of Hazardous Materials xxx (xxxx) xxxx

H. Dong, et al.

When sulfite concentration was further elevated to 1.5 mM, CBZ removal within 1200s did not increase but dropped. This indicated that at the fixed initial Fe(III) concentration, CBZ removal firstly increased and then decreased with increasing sulfite concentration, which was due to the fact that both sulfite and in-situ generated Fe(II) competed with CBZ for ROS generated in the Fe(III)/sulfite process. However, the CBZ degradation rate decreased with increasing sulfite concentration, which was consistent with the change of Fe(II) and sulfite concentrations as displayed in Fig. 1b and c. As shown in Fig. 1b, Fe(II) was produced rapidly within the first 10 s and then Fe(II) concentration declined to almost zero with further increasing reaction time, which indicated that the produced Fe(II) was re-oxidized in the Fe(III)/sulfite process. The results in Fig. 1c showed that sulfite was consumed during Fe(III)/ sulfite reactions, independent of the initial sulfite concentrations. Meanwhile, CBZ with the applied concentration had negligible effects on the change of Fe(II) concentration and the consumption of sulfite in the Fe(III)/sulfite process (Fig. S2). The initial decomposition kinetics of CBZ removal, Fe(II) re-oxidation, and sulfite consumption were fit with pseudo-first-order rate law and the resulting rate constants (kCBZ (s−1); kFe(II) (s−1); ksulfite (s−1)) were shown in Fig. S3. The rate of CBZ degradation was positively correlated with these of Fe(II) re-oxidation and sulfite consumption. Additionally, the rate of Fe(II) re-oxidation was also found to be positively correlated with that of sulfite consumption. These above results demonstrated that ROS were generated and the in-situ formed Fe(II) played a critical role in the generation of ROS responsible for CBZ degradation in the Fe(III)/sulfite process. The dosing mode of sulfite had significant influence on CBZ degradation, change of Fe(II) concentration, and sulfite consumption in the Fe(III)/sulfite process, as depicted in Fig. 2. CBZ removal was greatly improved by changing the sulfite dosing mode from singledosing mode (Fig. 1a) to multiple-dosing mode (Fig. 2a), indicating that the multiple-dosing mode was beneficial to CBZ removal in the Fe(III)/ sulfite process when same amount of sulfite was applied in total. This may be resulted from the generation and decay of Fe(II) and the consumption of sulfite. In Fig. 2a, CBZ degradation increased with reaction time and reached a steady state with reaction time longer than 180 s when 0.20 mM sulfite was dosed at the beginning of the reaction. Further degradation of CBZ occurred when 0.20 mM and 0.10 mM sulfite was dosed at 5.5 min and 10.5 min, respectively. Correspondingly, the generation and decay of Fe(II) and consumption of sulfite took place and the evolution trends were consistent with that of CBZ degradation, which implied that the in-situ generated Fe(II) was crucial to the generation of ROS in the Fe(III)/sulfite process. This also demonstrated that Fe(III) in the Fe(III)/sulfite process could be recycled for contaminants degradation, which was significant for the application of this process in water treatment. Based on above results, the generation and re-oxidation of Fe(II)

2.3. Analytical methods Concentrations of CBZ, methyl phenyl sulfoxide (PMSO), and methyl phenyl sulfone (PMSO2) were determined with UPLC (Waters ACQUITY UPLC H-Class). The compound was separated with a BEH C18 column (2.1 × 100 mm, 1.7 μm; Waters Co.) in an isocratic mode of elution at 35 ± 1 °C with UV–vis detector. The mobile phase consisted of 0.1 % formic acid solution-methanol (40/60 v/v for CBZ), and 0.1 % formic acid solution-acetonitrile (72/28 v/v for PMSO and PMSO2). The flow rate and injection volume were 0.2 mL/min and 10 μL, respectively. The Fe(II) concentration was quantified with the 1,10-phenanthroline colorimetric method at 510 nm using an ultraviolet–visible spectrophotometer (TU-1901, Purkinje General Instrument) (Fan et al., 2018). A modified spectrophotometric method was applied to determine the concentration of sulfite (Humphrey et al., 1970) and the brief procedures followed our previous study (Dong et al., 2019). Formaldehyde was analyzed by UPLC with UV–vis detector after derivatization with 2,4-dinitrophenylhydrazine (DNPH) (Feng et al., 2017). Briefly, 2.5 mL sample was added to the tube containing 100 μL DNPH (1 g/L in acetonitrile, 3 % H3PO4) and 2.5 mL deionized water. The mixture was allowed to react for 30 min at 60 ± 1 °C controlled with water bath and then analyzed by UPLC (0.1 % formic acid solution-acetonitrile, 72/28 v/v, 365 nm). Electron paramagnetic resonance (EPR) spectra were collected using a Bruker EMX plus 10/12 spectrometer to identify the radicals in the Fe (III)/sulfite process. DMPO was applied as a spin-trapping agent for HO% and sulfur-centered radicals detection. Aliquots of Fe(III) and NaHSO3 stock solutions were first mixed and then DMPO was injected into the mixed solutions at determined time intervals. Then the reaction solutions were transferred to a capillary tube and inserted into the cavity of the spectrometer for analysis. The EPR spectra were obtained under the following conditions: a center field of 3350 Gs, a sweep width of 100 Gs, a microwave power of 1.0 mW, a modulation amplitude of 0.30 Gs, and a sweep time of 41.96 s. Residual peroxymonosulfate (HSO5−) was detected with a modified iodometric oxidation method and detailed procedure was provided in Text S2 of SI.

3. Results and discussion 3.1. CBZ degradation by Fe(III)/sulfite process Control experiments showed that no CBZ was degraded with the treatment of Fe(III), Fe(II), or sulfite alone (Fig. S1). As shown in Fig. 1a, CBZ degradation occurred in the Fe(III)/sulfite process with varying initial sulfite concentrations. CBZ removal within 1200s was enhanced as sulfite concentration increased from 0.10 mM to 0.50 mM.

Fig. 1. Influence of sulfite concentration on the (a) kinetics of CBZ degradation, (b) Fe(II) generation and consumption, and (c) sulfite consumption in the Fe(III)/ sulfite process. Solid lines represent pseudo-first-order kinetics model fits of the initial reaction kinetics data. Reaction conditions: [Fe(III)]0 = 100 μM, [CBZ]0 = 5.0 μM, pHini = 3.0. 3

Journal of Hazardous Materials xxx (xxxx) xxxx

H. Dong, et al.

Fig. 2. Influence of sulfite applied with multiple-dosing mode on the (a) kinetics of CBZ degradation, (b) Fe(II) generation and consumption, and (c) sulfite consumption in the Fe(III)/sulfite process. Reaction conditions: [Fe(III)]0 = 100 μM, [CBZ]0 = 5.0 μM, pHini = 3.0. Sulfite of 0.20 mM was dosed at 0 min and 5.5 min, respectively, and sulfite of 0.10 mM was dosed at 10.5 min.

should be involved during Fe(III)/sulfite reactions and thus the catalytic cycle was completed. As shown in Eqs. 1–2, the generation of Fe (II) initially derived from the reduction of Fe(III) by sulfite. Then, the generated Fe(II) could be re-oxidized to Fe(III) by SO5%−, HSO5−, SO4%−, and/or HO% (Eqs. 8–10, 12) proposed in the literatures (Herrmann et al., 1996; Gilbert and Stell, 1990; Lee and Rochelle, 1987; Stuglik and PawełZagórski, 1981). However, there was no direct evidence for the generation of SO4%− and HO% in the Fe(III)/sulfite process. It is not clear which species (SO5%−, HSO5−, SO4%−, and/or HO%) was responsible for re-oxidation of Fe(II) and how Fe(II) played a significant part in the generation of ROS in the Fe(III)/sulfite process. These issues will be discussed in the following sections.

Fe2 + + HO• → Fe3 + + OH− (k=3.2 × 108M-1s-1)

(12)

3.2. Identification of ROS in the Fe(III)/sulfite process EPR experiments were performed to directly identify the radical species generated in the Fe(III)/sulfite process using DMPO as a spin trap. EPR spectra were obtained at different reaction time in the Fe(III)/ sulfite system. As shown in Fig. 3, when DMPO was dosed at 0 s (DMPO and sulfite were mixed first and Fe(III) was transferred to the mixed solution), a EPR spectrum with the hyperfine coupling constants (αN = 14.9 G, αβ-H = 15.3 G) was observed, which was in agreement with that of DMPO/SO3%− adducts reported previously (Ranguelova and Mason, 2011). This result indicated that SO3%− was generated during the reaction between Fe(III) and sulfite. When DMPO was introduced at 120 s, the EPR spectrum characteristic of the mixture of DMPO/SO3%− adducts and DMPO/HO% adducts was observed, which was very similar to the signal displayed in the literature (Zamora and Villamena, 2012). The signal of DMPO/HO% adducts (αN = 14.8 G, αβH = 14.6 G) was also observed when the DMPO was applied at 270 s. At the same time, the EPR signal appeared with the hyperfine coupling constants (αN = 14.6 G, αβ-H = 10.0 G, αγ-H1 = 1.14 G, αγ-H2 = 0.90 G) consistent with that of DMPO/SO4%− (Zamora and Villamena, 2012). These results demonstrated that HO% and SO4%− were generated from the transformation of SO3%− in the Fe(III)/sulfite process since the generated SO3%− can readily react with O2 to form SO5%− at near diffusion-limited rates (Eq. 3) and then the formed SO5%− reacted with sulfite to generate HO% and SO4%− through a series of reactions (Eqs. 4–9 and 11). It should be noted that DMPO/SO4%− adducts readily undergo nucleophilic substitution reactions with H2O/OH− to form stable DMPO/HO% adducts (Davies et al., 1992; Timmins et al., 1999). The half-life of DMPO/SO4%− adduct was 95 s in water while the halflife of DMPO/HO% adduct was up to 2.6 h (Davies et al., 1992; Finkelstein et al., 1980). The signal for DMPO/HO% adducts was even detected in typical SO4%−-based process, i.e. Co(II)/peroxymonosulfate

Fig. 3. EPR spectra obtained at different reaction time in the Fe(III)/sulfite system. Reaction conditions: [Fe(III)]0 = 100 μM, [sulfite]0 = 0.50 mM, [CBZ]0 = 5.0 μM, [DMPO]0 = 110 mM, pHini = 3.0. Fe(III) and sulfite were mixed first and DMPO was transferred to the mixed solution at different reaction time. (★ indicates DMPO/SO3%− adduct; ♦ indicates the mixture of DMPO/ SO3%− adduct and DMPO/HO% adduct; ● indicates DMPO/HO% adduct; ▼ indicates DMPO/SO4%− adduct.).

system (Dong et al., 2019). In Fig. 3, the EPR signal of SO4%− was much weaker than that of HO% even though the EPR signal of SO4%− was observed at 270 s. In contrast, no EPR signal of SO4%− was observed at 120 s, which may be ascribed to the ready transformation of DMPO/ SO4%− adducts to DMPO/HO% adducts and shadowing of the EPR signal of SO4%− by those of SO3%− and HO%. Consequently, EPR spectra could not offer the information of the relative abundances of SO4%− and HO% in the Fe(III)/sulfite process. The role of SO3%− in CBZ degradation in the Fe(III)/sulfite process was ruled out based on the evidence that the EPR spectrum for SO3%− was also detected in deoxygenated solutions where Fe(III)/sulfite reactions were occurring, but where negligible CBZ degradation was observed (Fig. S4). In addition, the possible involvement of O2%− in the Fe(III)/sulfite process was examined through the specific quenching experiments. As shown in Fig. S5, excess amount of carbon tetrachloride (CCl4) had negligible influence on the CBZ degradation in the Fe(III)/sulfite process, which excluded the role of O2%− in CBZ degradation in the Fe(III)/sulfite process since O2%− could degrade CCl4 at 4

Journal of Hazardous Materials xxx (xxxx) xxxx

H. Dong, et al.

high reaction rate (1.6 × 1010 M−1 s−1) (Fu et al., 2016). The literatures reported that ferryl ion (Fe(IV)) could be produced from the reaction of Fe(II) with H2O2 at near neutral pH or peroxydisulfate at pH 3.0–5.0, which was identified based on the fact that the oxidation product of sulfoxides (e.g., methyl phenyl sulfoxide (PMSO), and methyl p-tolyl sulfoxide (TMSO)) by Fe(IV) was markedly different from that by HO% and SO4%− (Bataineh et al., 2012; Wang et al., 2018). PMSO could be oxidized to methyl phenyl sulfone (PMSO2) through an oxygen atom transfer step by Fe(IV) rather than HO% and SO4%− (Wang et al., 2018). Thus, PMSO was selected to examine whether Fe(IV) was formed in the Fe(III)/sulfite process due to the significance of in-situ generated Fe(II) in this process as mentioned in Section 3.1. In Fig. S6, about 2.8 μM PMSO was degraded in the Fe(III)/sulfite process while no PMSO2 was formed, which indicated that Fe(IV) was not generated in this process.

excess could also scavenge SO4%− besides HO% in the Fe(III)/sulfite process. Therefore, the value of ck was used to compare the competitive capacity of the probes for SO4%− and HO%, where c is the concentration of the probes and k is the reaction rate constant of the probes with SO4%− or HO%. Since the ck values of 250 mM MeOH for HO% and SO4%− were about 4432–5682 times and 52–651 times greater than these of 5.0 μM CBZ for HO% and SO4%−, respectively, 250 mM MeOH was enough to scavenge HO% and SO4%− in the Fe(III)/sulfite process. The inhibition of 250 mM MeOH could represent the sum contribution of HO% and SO4%− and thus the contribution of HO% and SO4%− to CBZ degradation was calculated to be 89.04 %. Correspondingly, the contribution of other species mainly referring to SO5%− was 10.96 %. Because the ck value of 10 mM t-BuOH for HO% was about 86–173 times greater than that of 5.0 μM CBZ, 10 mM t-BuOH was enough to scavenge HO% in the Fe(III)/sulfite process. Meanwhile, the ck value of 5.0 μM CBZ for SO4%− was 1.05–2.4 times greater than that of 10 mM tBuOH, which indicated 10 mM t-BuOH could not scavenge SO4%− in the Fe(III)/sulfite process. Thus, 10 mM t-BuOH could be used to differentiate the contributions between HO% and SO4%− and the contributions of HO% and SO4%− to CBZ degradation were estimated to be 23.29 % and 65.75 %, respectively. These results demonstrated that SO4%− was a major oxidant, and HO% also played a significant role in the CBZ degradation in the Fe(III)/sulfite process.

3.3. The contribution of different radicals to contaminant degradation in the Fe(III)/sulfite process Based on the results in Section 3.2, both HO% and SO4%− could be generated in the Fe(III)/sulfite process. Meanwhile, SO5%− may also contribute to the CBZ degradation in the Fe(III)/sulfite process since it could be produced (Eq. 3) and is an oxidative species with a redox potential of E0(SO5%−/HSO5−) = 1.1 V vs NHE at pH 7.0 (Huie and Neta, 1984). MeOH could be applied to quench both HO% and SO4%− due to its high reactivity towards both radicals (7.8 × 108-1.0 × 109 M−1 s−1 for HO% and 2.0 × 106-2.5 × 107 M−1 s−1 for SO4%−) (Neta et al., 1988; Buxton et al., 1988). t-BuOH is an effective agent for quenching HO% because the second-order rate constant of t-BuOH with HO% (3.8 × 108-7.6 × 108 M−1 s−1) is approximately 1000-fold larger than that with SO4%− (4.0 × 105–9.1 × 105 M−1 s−1) (Neta et al., 1988; Buxton et al., 1988). Meanwhile, SO5%− is relatively inert toward alcohols due to its reaction rate with alcohols less than 103 M−1 s−1 (Neta et al., 1988). Thus, MeOH could be used to differentiate the contributions of SO5%− and SO4%−/HO%, while t-BuOH could be introduced to differentiate the contributions of SO4%− and HO%. In Fig. 4, different concentrations of MeOH or t-BuOH were introduced to the Fe(III)/sulfite system and the results showed that the inhibitions of MeOH or t-BuOH on CBZ degradation increased with increasing the concentrations of MeOH or t-BuOH. When MeOH and tBuOH were applied at the same concentration, the inhibition of MeOH on CBZ degradation was greater than that of t-BuOH, which indicated that both SO4%− and HO% had contributions to the CBZ degradation in the Fe(III)/sulfite process. The initial decomposition kinetics of CBZ removal was fit with pseudo-first-order rate law and the resulting rate constants (kCBZ; s−1) are summarized in Table S1. The relative contributions of radicals were calculated based on the difference in the reaction rate as shown in Table S1. However, this could not actually reflect the relative contributions of these radicals since t-BuOH in large

3.4. Critical role of generated Fe(II) in the radical yields during Fe(III)/ sulfite reactions Since the generation and re-oxidation of Fe(II) played a critical role in the generation of ROS, and SO4%− and HO% were the major oxidants in the Fe(III)/sulfite process, it was essential to reveal whether the generation of radicals (SO4%− and HO%) had correlation with the generated Fe(II) in this process. The t-BuOH assay was initially used to quantify the HO% yield based on the determination of formaldehyde formation, where the HO% yield is about twice of the formaldehyde yield (Flyunt et al., 2003). Later, the t-BuOH assay was also developed to estimate the SO4%− yield and it was established that the SO4%− yield was also about twice of the formaldehyde yield (Yang et al., 2015). Therefore, 1.0 M t-BuOH was applied to completely scavenge SO4%− and HO% in the Fe(III)/sulfite process. As shown in Fig. S7, addition of excess t-BuOH had negligible influence on the generation and re-oxidation of Fe(II) in the Fe(III)/sulfite process. Meanwhile, initial Fe(III) concentration exerted influence on the generation and re-oxidation of Fe(II) and the peak concentration of Fe(II) generated in this process (Fig. S8). Thus, the formaldehyde yield versus the peak concentration of generated Fe(II) was obtained as displayed in Fig. 5a, which showed that the formaldehyde generation was positively correlated with the peak concentration of Fe(II) generated in the Fe(III)/sulfite process. Correspondingly, the generation of radicals (SO4%− and HO%) was well correlated to the peak concentration of Fe(II) generated in this process

Fig. 4. Influence of (a) MeOH and (b) t-BuOH on CBZ degradation kinetics in the Fe(III)/sulfite system. Solid lines stand for pseudo-first-order kinetics model fits of the initial reaction kinetics data. Reaction conditions: [Fe(III)]0 = 100 μM, [sulfite]0 = 0.50 mM, [CBZ]0 = 5.0 μM, pHini = 3.0. 5

Journal of Hazardous Materials xxx (xxxx) xxxx

H. Dong, et al.

Fig. 5. The formation of (a) formaldehyde and (b) radical as a function of the highest Fe(II) concentration generated in the Fe(III)/sulfite process by the t-BuOH assay. Reaction conditions: [Fe(III)]0 = 10, 20, 30, 50 and 100 μM, [sulfite]0 = 0.50 mM, [t-BuOH]0 = 1.0 M, pHini = 3.0.

absorbance bands at 522 nm (Fig. S10). Meanwhile, since Fe(III) was reduced to Fe(II) and then generated Fe(II) was re-oxidized to Fe(III) (Figs. 1b and 2 b), Fe(III) had catalytic role in the Fe(III)/sulfite process.

(Fig. 5b). Thus, the relationship between the radical generation and the peak concentration of generated Fe(II) could be expressed in Eq. 13. The radical generated in the Fe(III)/sulfite process was detected to be 38.11 μM using the t-BuOH assay when the initial concentrations of Fe (II) and sulfite were 100 μM and 0.20 mM, respectively, at pHini 3.0. According to Eq.13, the peak concentration of generated Fe(II) was calculated to be 19.39 μM, which was close to the detected peak concentration of generated Fe(II) (19.19 μM) as shown in Fig. 1b. Furthermore, the radical generated in the Fe(III)/sulfite process was determined to be 79.16 μM using the t-BuOH assay under the conditions identical to that shown in Fig. 2. Then the peak concentration of generated Fe(II) was calculated to be 53.35 μM according to Eq. 13, which was close to the sum peak concentrations of generated Fe(II) (51.98 μM) detected experimentally. These results further confirmed the good correlation of radical generation with the in-situ generated Fe(II) in the Fe(III)/sulfite process.

[Radical] = 1.21[Fe(II)] + 14.68

3.5. Factors affecting CBZ degradation in the Fe(III)/sulfite process 3.5.1. Effect of pH Influence of pHini on CBZ degradation in the Fe(III)/sulfite process was examined and the results were shown in Fig. 6. CBZ could be degraded at pHini 3.0–7.0 and the rate of CBZ degradation decreased with increasing pHini. The change of CBZ concentration with reaction time was accompanied with the corresponding increase in H+ concentration, regardless of pHini. Since sulfite mainly exists as HSO3− under the conditions employed in this study and Fe(III) becomes less redox-reactive with increasing pH (Kuo et al., 2006b), the reaction of Fe(III) with sulfite was more favored at lower pH. More prominent lag phase at pHini 5.0–7.0 was observed at higher pHini since the un-catalyzed oxidation HSO3− by O2 could result in H+ generation and the rate of pH drop increased with decreasing pHini (Wilkinson et al., 1993). Moreover, the formation of SO4%− in the Fe(III)/sulfite process was derived from the reaction of SO5%− with sulfite and that of HSO5− with Fe(II) while HO% was produced from the reaction of SO4%− with H2O. It was well documented that pH affected the generation of SO4%− from the reaction of HSO5− with Fe(II) (Rastogi et al., 2009). Furthermore, pH influenced the transformation of SO4%− to HO since SO4%− could also react with HO− to generate HO% with reaction rate constant ∼104 times greater than that of reaction between SO4%− and H2O (Liang and Su, 2009). Therefore, pH plays a complicated influence on the generation of reactive species in this system and can affect multiple radical chain reactions besides the speciation of Fe(III) and sulfite. In addition, 88–90% of CBZ was removed at equilibrium when pHini was in the range of 4.0–6.0 while less CBZ was removed at equilibrium at pHini 3.0 and 7.0 (Fig. 6). The final pH increased slightly from 3.3 to 3.6 when pHini was elevated from 4.0 to 6.0 and thus the similar removal of CBZ at pHini 4.0–6.0 may be ascribed to the similar pH level at the end of reaction. However, the final pH values were 2.9 and 4.5, respectively, when pHini were 3.0 and 7.0. Since both the process of generating radicals and the competition of CBZ with non-target reductant for radicals were pH dependent, CBZ removal at equilibrium was pH-dependent.

(13)

Based on the above results, the generation and re-oxidation of Fe(II) was accompanied with the generation of radicals (SO4%− and HO%). In the Fe(III)/sulfite process, Fe(III) was firstly reduced to Fe(II) by sulfite to generate SO3%− through a one-electron transfer (Figs. 3 and S4). The generated SO3%− reacted rapidly with O2(aq) to form SO5%− and thus the secondary radicals were generated, which was confirmed by transformation of SO3%− to SO4%− and HO% and the larger intensity of the EPR spectrum for SO3%− without the presence of DO than that with the presence of DO (Figs. 3 and S4b). The formed SO5%− reacted with sulfite through two branching reactions: one was to form HSO5− (Eq. 4) and the other was to form SO4%− (Eq. 5). The amounts of HSO5− generated immediately after sulfite was consumed in the Fe(III)/sulfite process was detected at different initial sulfite concentrations, as summarized in Table S2. The generated HSO5− could react with Fe(II) to form SO4%−, which was verified by the EPR spectrum in Fig. S9a. Additionally, there was no radical formation during the reaction of HSO5− with sulfite as shown in Fig. S9b. Thus, the formation of SO4%− in the Fe(III)/sulfite process was derived from the reaction of SO5%− with sulfite and that of HSO5− with Fe(II). HO% was produced from the reaction of SO4%− with H2O. Consequently, SO4%−, HO, SO5%−, and HSO5− were the plausible species responsible for the re-oxidation of Fe(II) in the Fe(III)/sulfite process. As displayed in Fig. S7, excess t-BuOH had negligible on the generation and re-oxidation of Fe(II) in the Fe(III)/sulfite process, suggesting the negligible role of SO4%− and HO% in the re-oxidation of Fe(II) in this process. Therefore, SO5%− and HSO5− were the major oxidants responsible for the re-oxidation of Fe(II) in the Fe(III)/sulfite process. Moreover, the critical role of generated Fe(II) in the ROS generation in the Fe(III)/sulfite process was further convinced by the inhibition of CBZ degradation with the presence of bipyridyl (BPY) which could form a color complex with Fe(II) with characteristic

3.5.2. Effect of initial Fe(III) and CBZ concentrations Influence of initial Fe(III) dosage on the kinetics of CBZ degradation in the Fe(III)/sulfite process was investigated and the results are shown in Fig. 7a. When the initial Fe(III) dosage increased from 10 μM to 100 μM, the rate of CBZ degradation in this system increased, which was due to the significance role of in-situ generated Fe(II) in the Fe(III)/ sulfite process since the rate of Fe(II) generation and decay increased with increasing initial Fe(III) concentration, as demonstrated in Fig. S8. 6

Journal of Hazardous Materials xxx (xxxx) xxxx

H. Dong, et al.

concentration. 3.5.3. Influence of Cl− and HA Fig. 7c displays the influence of Cl− on CBZ removal in the Fe(III)/ sulfite system. Cl− exerted slight inhibition on CBZ removal as Cl− concentration increased from 1.0 mM to 20.0 mM in the Fe(III)/sulfite system, which was ascribed to the reactions of Cl− with SO4%− and/or HO%. Cl− can react with HO% to generate ClOH%− (Eq. 14) (Jayson et al., 1973), which will dissociate into Cl% and H2O rapidly (Eq. 15) (Jayson et al., 1973). Cl− can also react with SO4%− to form Cl% (Eq. 16) (Mcelroy, 1990). Cl% showed considerable reactivity towards CBZ compared with SO4%− and/or HO% since the second-order rate constants of CBZ with SO4%−, HO%, and Cl% were reported to be (1.92 ± 0.01)×109 M-1 s-1, (8.8 ± 1.2)×109 M-1 s-1, and (5.6 ± 1.6)×1010 M-1 s-1, respectively (Matta et al., 2011; Huber et al., 2003; Wang et al., 2016). Therefore, Cl− showed little inhibiting effect on CBZ degradation in the Fe(III)/sulfite system.

HO• + Cl− → ClOH• − k=4.3 × 109M-1s-1

(14)

ClOH• − + H+ → Cl• + H2 O k=2.1 × 1010M-1s-1

(15)

SO•4− + Cl− → Cl• + SO24− k=2.7 × 108M-1s-1

(16)

As shown in Fig. 7d, HA depressed CBZ removal to a greater extent with increasing HA concentration from 0 to 5.0 mg/L in the Fe(III)/ sulfite system. HA, as a significant constituent of natural organic matter (NOM), could compete with CBZ for SO4%− and HO since SO4%− and HO% could effectively react with HA at rate constants of 6.8 × 103 L·mg C−1 s−1 and 1.4 × 104 L·mg C−1 s−1, respectively (Lutze et al., 2015). 4. Conclusions In this study, the performance of Fe(III)/sulfite process was investigated and the underlying mechanisms were elucidated. CBZ degradation occurred in the Fe(III)/sulfite process with varying initial sulfite concentrations. Fe(II) was rapidly generated and then decayed in the Fe(III)/sulfite process. Moreover, sulfite applied with the multipledosing mode was beneficial to CBZ removal in the Fe(III)/sulfite process than the single-dosing mode. SO4%− and HO% were found to be the primary and secondary oxidants, respectively, for CBZ degradation in the Fe(III)/sulfite process based on the EPR spectra, quenching experiments, and using PMSO as a probe substrate. The t-BuOH assay showed the correlation of radical generation with the in-situ generated Fe(II) in the Fe(III)/sulfite process, confirming the crucial role of Fe(II) in the radicals (SO4%− and HO%) generation in this process. SO4%− in the Fe(III)/sulfite process was generated from the reaction of SO5%− with sulfite and that of HSO5− with Fe(II). HO% was produced from the reaction of SO4%− with H2O. SO5%− and HSO5− were the major oxidants responsible for the re-oxidation of Fe(II) in the Fe(III)/sulfite process. Meanwhile, Fe(III) played a catalytic role in the Fe(III)/sulfite process. Moreover, CBZ degradation was closely related to pHini, initial Fe(III) dosage, and CBZ dosage besides initial sulfite concentration. The presence of Cl− showed slight effect on CBZ degradation while HA at elevated concentration inhibited the CBZ degradation in the Fe(III)/ sulfite process. The results in this study clearly elucidated the Fe(III)/Fe (II) cycle in the Fe(III)/sulfite process and the mechanism of this system, which would advance the application of this oxidation system. The performance of the Fe(III)/sulfite process can be improved by facilitating the cycle of Fe(III) and Fe(II) and optimizing the dosing mode of sulfite.

Fig. 6. Kinetics of CBZ degradation and changes in H+ concentration during CBZ removal in the Fe(III)/sulfite system at different pHini. Reaction conditions: [Fe(III)]0 = 100 μM, [sulfite]0 = 0.50 mM, [CBZ]0 = 5.0 μM.

However, CBZ removal at initial Fe(III) dosages of 10 μM, 50 μM, and 100 μM was close at equilibrium, which was due to the fact that sulfite and in-situ generated Fe(II) competed with CBZ for reactive radicals in the Fe(III)/sulfite process. In Fig. 7b, the amount of CBZ removed increased with increasing initial CBZ concentration from 5.0 μM to 20.0 μM, which was due to the fact that more CBZ competed for reactive species (mainly SO4%− and HO%) with increasing CBZ

Declaration of Competing Interest We declare that we have no conflict of interest. 7

Journal of Hazardous Materials xxx (xxxx) xxxx

H. Dong, et al.

Fig. 7. Effect of (a) initial Fe(III) concentration on kinetics of CBZ degradation in the Fe(III)/ sulfite system; Influence of (b) CBZ concentration, (c) Cl−, and (d) HA on CBZ removal in the Fe(III)/sulfite system. Reaction conditions: [Fe(III)]0 = 100 μM (for panels b–d), [sulfite]0 = 0.50 mM, [CBZ]0 = 5.0 μM (for panels a, c and d), pHini = 3.0, reaction time 20 min for panel b, reaction time 10 min for panels c and d.

Acknowledgments

Dong, H., Chen, J., Feng, L., Zhang, W., Guan, X., Strathmann, T.J., 2019. Degradation of organic contaminants through activating sulfite by cerium (IV): a sulfate radicalpredominant oxidation process. Chem. Eng. J. 357, 328–336. Fan, P., Sun, Y., Qiao, J., Lo, I.M.C., Guan, X., 2018. Influence of weak magnetic field and tartrate on the oxidation and sequestration of Sb(III) by zerovalent iron: batch and semi-continuous flow study. J. Hazard. Mater. 343, 266–275. Feng, Y., Lee, P.H., Wu, D., Shih, K., 2017. Rapid selective circumneutral degradation of phenolic pollutants using peroxymonosulfate-iodide metal-free oxidation: role of iodine atoms. Environ. Sci. Technol. 51, 9410. Finkelstein, E., Rosen, G.M., Rauckman, E.J., 1980. Spin trapping of superoxide and hydroxyl radical: practical aspects. Arch. Biochem. Biophys. 200, 1–16. Fischer, M., Warneck, P., 1996. Photodecomposition and photooxidation of hydrogen sulfite in aqueous solution. J. Phys. Chem. 100, 15111–15117. Flyunt, R., Leitzke, A., Mark, G., Mvula, E., Reisz, E., Schick, R., Sonntag, C.V., 2003. Determination of •OH, O2•−, and hydroperoxide yields in ozone reactions in aqueous solution. J. Phys. Chem. B 107, 7242–7253. Fu, X., Gu, X., Lu, S., Miao, Z., Xu, M., Zhang, X., Danish, M., Cui, H., Farooq, U., Qiu, Z., 2016. Enhanced degradation of benzene by percarbonate activated with Fe(II)-glutamate complex. Environ. Sci. Pollut. Res. 23, 6758–6766. Gilbert, B.C., Stell, J.K., 1990. Mechanisms of peroxide decomposition. An ESR study of the reactions of the peroxomonosulphate anion (HOOSO3–) with TiIII, FeII, and αoxygen-substituted radicals. J. Chem. Soc. Perkin Trans. I 27, 1281–1288. Grgić, I., Dovžan, A., Berčič, G., Hudnik, V., 1998. The effect of atmospheric organic compounds on the Fe-catalyzed S(IV) autoxidation in aqueous solution. J. Atmos. Chem. 29, 315–337. Guo, Y., Lou, X., Fang, C., Xiao, D., Wang, Z., Liu, J., 2013. Novel photo-sulfite system: toward simultaneous transformations of inorganic and organic pollutants. Environ. Sci. Technol. 47, 11174–11181. Herrmann, H., Jacobi, H.W., Raabe, G., Reese, A., Zellner, R., 1996. Laser-spectroscopic laboratory studies of atmospheric aqueous phase free radical chemistry. Fresenius J. Anal. Chem. 355, 343–344. Huber, M.M., Canonica, S., Park, G.-Y., von Gunten, U., 2003. Oxidation of pharmaceuticals during ozonation and advanced oxidation processes. Environ. Sci. Technol. 37, 1016–1024. Huie, R.E., Neta, P., 1984. Chemical behavior of SO3•− and SO5•− radicals in aqueous solutions. J. Phys. Chem. 88, 5665–5669. Humphrey, R.E., Ward, M.H., Hinze, W., 1970. Spectrophotometric determination of sulfite with 4,4′-dithio-dipyridine and 5,5′-dithiobis(2-nitrobenzoic acid). Anal. Chem. 42, 698–702. Jayson, G.G., Parsons, B.J., Swallow, A.J., 1973. Some simple, highly reactive, inorganic chlorine derivatives in aqueous solution. Their formation using pulses of radiation and their role in the mechanism of the fricke dosimeter. J. Chem. Soc. Faraday Trans 1: Phys. Chem. 69, 1597–1607. Karatza, D., Prisciandaro, M., Lancia, A., Musmarra, D., 2010. Sulfite oxidation catalyzed

This work was supported by the National Natural Science Foundation of China (Grant No. 21976133 and 21603165). The EPR spectra were collected on the Steady High Magnetic Field Facilities, High Magnetic Field Laboratory, CAS. Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jhazmat.2019.121497. References Bataineh, H., Pestovsky, O., Bakac, A., 2012. pH-induced mechanistic changeover from hydroxyl radicals to iron(IV) in the Fenton reaction. Chem. Sci. 3, 1594–1599. Berglund, J., Fronaeus, S., Elding, L.I., 1993. Kinetics and mechanism for manganesecatalyzed oxidation of sulfur(IV) by oxygen in aqueous solution. Inorg. Chem. 32, 4527–4538. Brandt, C., van Eldik, R., 1995. Transition metal-catalyzed oxidation of sulfur(IV) oxides. Atmospheric-relevant processes and mechanisms. Chem. Rev. 95, 119–190. Brandt, C., Fabian, I., van Eldik, R., 1994. Kinetics and mechanism of the iron(III)-catalyzed autoxidation of sulfur(IV) oxides in aqueous solution. Evidence for the redox cycling of iron in the presence of oxygen and modeling of the overall reaction mechanism. Inorg. Chem. 33, 687–701. Buxton, G.V., McGowan, S., Salmon, G.A., Williams, J.E., Wood, N.D., 1996. A study of the spectra and reactivity of oxysulphur-radical anions involved in the chain oxidation of S(IV): a pulse and γ-radiolysis study. Atmos. Environ. 30, 2483–2493. 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, 513–886. Conklin, M.H., Hoffmann, M.R., 1988. Metal ion-sulfur(IV) chemistry. 3. Thermodynamics and kinetics of transient iron(III)-sulfur(IV) complexes. Environ. Sci. Technol. 22, 899. Das, T.N., 2001. Reactivity and role of SO5•− radical in aqueous medium chain oxidation of sulfite to sulfate and atmospheric sulfuric acid generation. J. Phys. Chem. A 105, 9142–9155. 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. I 2, 333–335.

8

Journal of Hazardous Materials xxx (xxxx) xxxx

H. Dong, et al. by cobalt ions in flue gas desulfurization processes. J. Air Waste Manage. Assoc. 60, 675–680. Kotronarou, A., Sigg, L., 1993. SO2oxidation in atmospheric water: Role of Fe(II) and effect of ligands. Environ. Sci. Technol. 27, 2725. Kuo, D.T.F., Kirk, D.W., Jia, C.Q., 2006a. The chemistry of aqueous S(IV)-Fe-O2 system: state of the art. J. Sulfur Chem. 27, 461–530. Kuo, D.T., Kirk, D.W., Jia, C.Q., 2006b. The chemistry of aqueous S(IV)-Fe-O2 system: state of the art. J. Sulfur Chem. 27, 461–530. Lee, Y.J., Rochelle, G.T., 1987. Oxidative degradation of organic acid conjugated with sulfite oxidation in flue gas desulfurization: products, kinetics, and mechanism. Environ. Sci. Technol. 21, 266–272. Liang, C., Su, H.-W., 2009. Identification of sulfate and hydroxyl radicals in thermally activated persulfate. Ind. Eng. Chem. Res. 48, 5558–5562. Lutze, H.V., Stephanie, B., Insa, R., Nils, K., Rani, B., Melanie, G., Clemens, V.S., Schmidt, T.C., 2015. Degradation of chlorotriazine pesticides by sulfate radicals and the influence of organic matter. Environ. Sci. Technol. 49, 1673. Martin, L.R., Hill, M.W., Tai, A.F., Good, T.W., 1991. The iron catalyzed oxidation of sulfur(IV) in aqueous solution: differing effects of organics at high and low pH. J. Geophys. Res. Atmos. 96, 3085–3097. Matta, R., Tlili, S., Chiron, S., Barbati, S., 2011. Removal of carbamazepine from urban wastewater by sulfate radical oxidation. Environ. Chem. Lett. 9, 347–353. Mcelroy, W.J., 1990. A laser photolysis study of the reaction of SO4•− with Cl− and the subsequent decay of Cl2•− in aqueous solution. J. Phys. Chem. 94, 2435–2441. 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, 1027–1284. Pasiuk-Bronikowska, W., Bronikowski, T., Ulejczyk, M., 2003. Inhibition of the S(IV) autoxidation in the atmosphere by secondary terpenic compounds. J. Atmos. Chem. 44, 97–111. Ranguelova, K., Mason, R.P., 2011. The fidelity of spin trapping with DMPO in biological systems. Magn. Reson. Chem. 49, 152–158. Rastogi, A., Al-Abed, S.R., Dionysiou, D.D., 2009. Sulfate radical-based ferrous–peroxymonosulfate oxidative system for PCBs degradation in aqueous and sediment systems. Appl. Catal., B: Environ. 85, 171–179. Stuglik, Z., PawełZagórski, Z., 1981. Pulse radiolysis of neutral iron(II) solutions: oxidation of ferrous ions by OH radicals. Radiat. Phys. Chem. 17, 229–233. 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-1-pyrroline-

N-oxide and 5-diethoxyphosphoryl-5-methyl -1-pyrroline-N-oxide as spin traps for HO• and SO4•−. Free Radical Biol. Med. 27, 329–333. Wang, Z., Jiang, J., Pang, S., Zhou, Y., Guan, C., Gao, Y., Li, J., Yang, Y., Qiu, W., Jiang, C., 2018. Is sulfate radical really generated from peroxydisulfate activated by iron(II) for environmental decontamination? Environ. Sci. Technol. 52, 11276–11284. Wang, W.-L., Wu, Q.-Y., Huang, N., Wang, T., Hu, H.-Y., 2016. Synergistic effect between UV and chlorine (UV/chlorine) on the degradation of carbamazepine: influence factors and radical species. Water Res. 98, 190–198. Wilkinson, P.M., Doldersum, B., Cramers, P.H.M.R., Van Dierendonck, L.L., 1993. The kinetics of uncatalyzed sodium sulfite oxidation. Chem. Eng. Sci. 48, 933–941. Xu, J., Ding, W., Wu, F., Mailhot, G., Zhou, D., Hanna, K., 2016. Rapid catalytic oxidation of arsenite to arsenate in an iron(III)/sulfite system under visible light. Appl. Catal. B: Environ. 186, 56–61. Yang, Y., Jiang, J., Lu, X., Ma, J., Liu, Y., 2015. Production of sulfate radical and hydroxyl radical by reaction of ozone with peroxymonosulfate: a novel advanced oxidation process. Environ. Sci. Technol. 49, 7330–7339. Yermakov, A.N., Purmal, A.P., 2003. Iron-catalyzed oxidation of sulfite: from established results to a new understanding. Prog. React. Kinet. Mec. 28, 189–256. Yu, Y., Li, S., Peng, X., Yang, S., Zhu, Y., Chen, L., Wu, F., Mailhot, G., 2016. Efficient oxidation of bisphenol A with oxysulfur radicals generated by iron-catalyzed autoxidation of sulfite at circumneutral pH under UV irradiation. Environ. Chem. Lett. 14, 527–532. Yuan, Y., Luo, T., Xu, J., Li, J., Wu, F., Brigante, M., Mailhot, G., 2019. Enhanced oxidation of aniline using Fe(III)-S(IV) system: role of different oxysulfur radicals. Chem. Eng. J. 362, 183–189. 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, 7210–7218. Zhang, Y., Zhou, J., Li, C., Guo, S., Wang, G., 2013. Reaction kinetics and mechanism of iron(II)-induced catalytic oxidation of sulfur(IV) during wet desulfurization. Ind. Eng. Chem. Res. 51, 1158–1165. Zhou, D., Yuan, Y., Yang, S., Gao, H., Chen, L., 2015. Roles of oxysulfur radicals in the oxidation of acid orange 7 in the Fe(III)-sulfite system. J. Sulfur Chem. 36, 373–384. Ziajka, J., Beer, F., Warneck, P., 1994. Iron-catalysed oxidation of bisulphite aqueous solution: evidence for a free radical chain mechanism. Atmos. Environ. 28, 2549–2552.

9