Sulfate radical induced degradation of β2-adrenoceptor agonists salbutamol and terbutaline: Phenoxyl radical dependent mechanisms

Water Research 123 (2017) 715e723

Contents lists available at ScienceDirect

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

Sulfate radical induced degradation of b2-adrenoceptor agonists salbutamol and terbutaline: Phenoxyl radical dependent mechanisms Lei Zhou a, b, Mohamad Sleiman b, Corinne Ferronato a, Jean-Marc Chovelon a, *, Pascal de Sainte-Claire b, Claire Richard b, ** a b

Univ Lyon, Universit
e Claude Bernard Lyon 1, CNRS, IRCELYON, F-69626, 2 Avenue Albert Einstein, Villeurbanne, France Universit
e Clermont Auvergne, CNRS, Sigma-Clermont, Institut de Chimie de Clermont-Ferrand, F-63178, Aubi ere, France

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 March 2017 Received in revised form 20 June 2017 Accepted 12 July 2017 Available online 14 July 2017

The present study investigated the reactivity and oxidation mechanisms of salbutamol (SAL) and ter
butaline (TBL), two typical b2-adrenoceptor agonists, towards sulfate radical (SO4) by using photoactivated persulfate (PS). The reaction pathways and mechanisms were proposed based on products identification using high resolution HPLC-ESI-MS, laser flash photolysis (LFP) and molecular orbital
calculations. The results indicated that SO4 was the dominant reactive species in the UV/PS process. The second-order rate constants of sulfate radical reaction with SAL and TBL were measured as (3.7 ± 0.3)  109 and (4.2 ± 0.3)  109 M1 s1 by LFP, respectively. For both SAL and TBL, phenoxyl radicals were found to play key roles in the orientation of the primary pathways. For SAL, a benzophenone derivative was generated by oxidation of the phenoxyl radical. However, in the case of TBL, the transformation of the phenoxyl radical into benzoquinone was impossible. Instead, the addition of eOSO3H on the aromatic ring was the major pathway. The same reactivity pattern was observed in the case of TBL structural analogs resorcinol and 3,5-dihydroxybenzyl alcohol. Our results revealed that basic conditions inhibited the decomposition of SAL and TBL, while, increasing PS dose enhanced the degradation. The present work could help for a better understanding of the difference in oxidation reactivity of substituted phenols widely present in natural waters. © 2017 Published by Elsevier Ltd.

Keywords: Photo-activated PS Laser flash photolysis Substituted phenols Reactivity and photoproducts

1. Introduction Pharmaceuticals and personal care products (PPCPs) are emerging contaminants that have drawn extensive attention recently due to their potential hazardous effect on ecological system and human beings, even at trace concentrations (Yang et al., 2011; Kosma et al., 2014). In the last few years, b2-adrenergic receptor agonists have been increasingly used for human beings as well as for livestock (Slotkin and Seidler, 2013; Rhodes et al., 2004). Salbutamol (SAL, also known as albuterol) and terbutaline (TBL) are selected as the target subjects of this study due to their widely uses. SAL is clinically the most widely used in treatment of bronchial asthma, it has been detected in several European countries from rivers and sewage treatment plants, and the concentration ranges

* Corresponding author. ** Corresponding author. E-mail addresses: [email protected] (J.-M. Chovelon), [email protected] (C. Richard). http://dx.doi.org/10.1016/j.watres.2017.07.025 0043-1354/© 2017 Published by Elsevier Ltd.

from 1.14 to 471 ng/L (Calamari et al., 2003; Bound and Voulvoulis, 2006; Ternes, 1998). TBL is considered as one of the major feed additive medicines to increase the proportion of lean meat of livestock (Slotkin and Seidler, 2013), it is also used as many as 1 million pregnancies annually to arrest preterm labor (Rhodes et al., 2004). b2-adrenergic receptor agonists possess a common bphenyl- b-ethanol amine group, the structure makes them hydrophilic and not easily biodegradable (Huerta-Fontela et al., 2011). Conventional sewage treatment plants are not able to degrade residues of these chemicals completely from waste streams (Huerta-Fontela et al., 2011), thus, developing effective treatment technology for elimination of these chemicals in waters is of great interest. Advanced oxidation process (AOPs) based on the generation of reactive species are becoming increasingly popular as a technology for the degradation of contaminants with the final goal of wastewater decontamination (Oller et al., 2011; Oturan and Aaron, 2014). Among the AOPs, hydrogen peroxide (H2O2) and persulfate (S2O2 8 , PS) are among the most popular oxidants (Zhou et al., 2013). The

716

L. Zhou et al. / Water Research 123 (2017) 715e723

AOPs based on PS for the remediation of contaminated soils, groundwater and sediments have received considerable attention in recent years (Dahmani et al., 2006; Fang et al., 2013a). PS can be activated by UV, heat, base or transition metals to generate sulfate
radical (SO4-), a strong one-electron oxidant (E0 z 2.6 V) (House,
1962). SO4 is capable of oxidizing a large number of contaminants such as polychlorinated biphenyls (PCBs), polycyclic aromatic hydrocarbons (PAHs), perfluoroalkyl compounds (PFCs) and azo dyes (Fang et al., 2012; Zhao et al., 2013; Lee et al., 2009; Xu and Li,
2010). Mechanisms of SO4- reactions differ somewhat from those of

hydroxyl radical ( OH) reactions. SO4- reacts more readily by oneelectron transfer than
OH but slower by H-abstraction and addi
tion (Neta et al., 1988; Lutze et al., 2015). Additionally, SO4- is less impacted by soluble microbial products and natural organic matter which are ubiquitously present in waters than
OH (He et al., 2014;
Gara et al., 2009). Thus, SO4--based oxidation could complement the
more common
OH-based AOPs technologies. SO4- has been shown effective in treating pharmaceuticals in waters (Nfodzo and Choi, 2011; Ahmed et al., 2012; Tan et al., 2013), many pharmaceuticals
can react with SO4- at comparable rates as with
OH (Rickman and Mezyk, 2010; Matta et al., 2011). However, the reaction rates and
pathways of SO4- with numerous pharmaceuticals including SAL and TBL have not been studied. Sulfate radical is a strong electrophile, and in the reaction with electron-rich compounds, such as phenolic compounds (ArOH), electron transfer reaction is a preferred reaction pathway (Lutze et al., 2015; Zhang et al., 2016). For electron transfer reactions in the case of ArOH, intermediate radical cations are proposed to be generated (Eqs. (2) and (3)). They would deprotonate subsequently to form phenoxyl radicals (Brede et al., 2002; Ganapathi et al., 2000). Both of SAL and TBL are substituted phenols, the formation of phenoxyl radicals in the reaction process as well as their transformation pathways need further investigation. $ S2 O2 8 þ hv/2SO4

(1)

2 $þ ArOH þ SO$ 4 /SO4 þ Ar  OH

(2)

Ar  OH$þ /ArO$ þ Hþ

(3)

In the present study, we evaluated the efficacy of simulated sunlight activated PS processes in eliminating SAL and TBL. The main objectives of this study are (1) to identify the main oxidation products of SAL and TBL and propose the oxidation mechanism of them by the reactive species involved in the processes; (2) to investigate the reactivity and predominant reactive species responsible for the degradation of SAL and TBL.

2.2. Laser flash photolysis (LFP) Transient absorption experiments were carried out using a nanosecond laser-flash photolysis spectrometer from Applied Photophysics (LKS.60) with laser excitation at 266 nm from a Quanta Ray GCR130-1 Nd:YAG. A 32 bits RISC processor kinetic spectrometer workstation was used to analyze the digitized signal. Appropriate volumes of chemical stock solutions (PS, SAL and TBL) were mixed just before the experiments to obtain the desired mixtures and concentrations (PS ranging from 0 to 24 mM). The excitation energy was set as low as possible to avoid biphotonic photoionization of the phenolic compounds and therefore minimize the generation of the phenoxyl radicals through their direct photolysis. LFP experiments of phenols alone in the same conditions were also conducted, the results confirmed that the generation of the phenoxyl radicals was negligible in the case of phenols alone. The decay of the sulfate radical was monitored at 450 nm in the experiments with SAL at maximum absorption of this radical but at 500 nm in the case of TBL because the phenoxyl radical of TBL absorbs until ~480 nm. In the presence of SAL or TBL, the decay rate of the sulfate radical was increased. By using a linear regression of the decay rate of the sulfate radical versus the SAL or TBL concentration, we could estimate the second order rate constant for the reactions of sulfate radical with SAL or TBL. All the experiments were performed at ambient temperature (295 ± 2 K) and in aerated solution, unless otherwise stated. 2.3. Steady irradiation experiments The sulfate radicals were generated by the activation of PS in a solar simulator Suntest CPSþ (HERAEUS, Hanau, France) equipped with a 1.5 kW xenon arc and an ultraviolet filter cutting off wavelengths shorter than 290 nm. Reaction solutions were poured in a cylindrical Pyrex reactor (i.d. ¼ 5 cm, H ¼ 15 cm, V ¼ 150 mL) maintained at 20  C with a circulating water system. A scheme of the applied reactor and the light intensity are given in Figs. S1 and S2 (Supplementary Material), respectively. Specific aliquot of substrates (SAL or TBL) and appropriate volumes of PS stock solution were added to achieve 50 mL reaction solutions (PS ranging from 0 to 24 mM). Control experiments with substrates alone and dark experiments with substrate-PS were carried out under identical conditions. They confirmed that the removal of the substrates was negligible in these conditions. Initial pH of reaction solutions was adjusted using 0.01 M H2SO4 and 0.01 M NaOH. Aliquot samples of 0.3 mL were withdrawn at predetermined time intervals and quenched immediately with the same volume of methanol, a scavenger for the sulfate radicals and the hydroxyl radicals, to stop the chemical oxidation reaction. All the experiments were performed at least in duplicate.

2. Materials and methods

2.4. Analyses

2.1. Chemicals

The concentrations of SAL, TBL and NB were monitored using a Shimadzu 10A series high performance liquid chromatography (HPLC, Shimadzu) system equipped with a Model 7725i injector with a 20 mL loop and coupled with a LC-10AT binary pump and a SPD-M10ADAD. Samples were separated with an Yperspher BDS C18 column (5 mm, 125 mm*4.0 mm, i.d.) (Interchim, France) at 40  C. The mobile phase was a mixture of acetonitrile and acidifiedwater (3‰ formic acid by volume, pH 3.0). An overview of the HPLC parameters is given together in Table S1. High resolution mass spectrometry (HR-MS) analyses were performed to identify the oxidation products, a Waters Q-TOF mass spectrometer in negative ion (ES) and positive ion (ESþ) ionization modes was used, which was coupled to an HPLC Waters Alliance

Salbutamol (SAL, 98%), terbutaline (TBL, 98%), nitrobenzene (NB, 99%) and potassium persulfate (K2S2O8, 99%) were purchased from Sigma-Aldrich. Suwannee River fulvic acid (SRFA, 1R101F) of reference grade was obtained from the International Humic Substance Society (IHSS). HPLC or LC-MS grade acetonitrile (ACN), ethanol (EtOH) and tert-butyl alcohol (TBA) were supplied by Fisher Chemical. The other reagents were at least of analytical grade and used as received without further purification. Milli-Q water (18 MU cm) was prepared from a Millipore Milli-Q system. All the solutions were prepared by dissolving the chemical agents into Milli-Q water without further preparations.

L. Zhou et al. / Water Research 123 (2017) 715e723

system consisting of a 2695 module for compounds separation and a 2998 photodiode array detector. An Yperspher BDS C18 column was also used, and the flow rate was set at 0.2 mL min1. The pH of the solutions was measured with a combined glass electrode connected to a PHM 210 Standard pH meter (Radiometer, Copenhagen).

2.5. DFT calculations The Density Functional Theory (DFT) calculations were performed using the Gaussian 09 software package (Frisch et al., 2010). The ground state geometries of the investigated species were optimized using the hybrid density functional B3LYP method with the 6e31 þ G(d,p) basis set. The absorption spectra were calculated with the time-dependent DFT method (TD-DFT) at the B3LYP/631 þ G(d,p) level. The vibrational progression in the absorption spectra was obtained from analysis of electronic excitations between vibrational levels of the ground and excited states. This basis set was shown to give accurate results for the phenoxyl radical (the 6-311 þ G(d,p) or cc-pVTZ basis sets yielded similar absorption energies) (Santoro et al., 2011). In an attempt to check whether the solvent affected our results, we completed B3LYP/6-31 þ G(d,p) calculations by simulating the solvent explicitly. Two water molecules were added in the model in the vicinity of the phenoxyl substituent. Optimized structures for the ground (D0) and excited (D3) states are given in the Supplementary Material (Text S1, S2 and Fig. S6). However, the maximum wavelength of absorption changed by less than 3 nm, indicating that the effect of water molecules on the absorption spectrum was negligible. It was thus decided to perform calculations without the solvent.

717

3.1.1. Transient signals from SAL and assignment of its absorption bands Fig. 1 also shows the transient species generated when PS is photolyzed in the presence of SAL (250 mM). At the end of the pulse, a broad absorption band extending from 350 to more than 500 nm is observed (Fig. 1A). The decay of the transient absorbance between 410 and 500 nm is fully achieved within 10 ms after the pulse (Fig. 1C), while the absorbance between 350 and 410 nm decay much more slowly (Fig. 1B). This indicates the presence of at least two species. The short-lived species is assigned to the sulfate radical and the long-lived one to the phenoxyl radical generated from SAL, in accordance with the consecutive peak structure between 360 and 420 nm of the absorption spectrum (Feitelson et al., 1973; Chen et al., 2011; Silverstein et al., 2013). This series of equally spaced peaks is mainly due to the vibrational progressions, the peak at the longer wavelength corresponding to the 0-0 transition. Very similar spectra were found in the irradiated solutions of SAL structural analogs 2-hydroxybenzyl alcohol (2-HBA) and 4hydroxybenzyl alcohol (4-HBA) with PS (Fig. S4). 3.1.2. Transient signals from TBL and assignment of its absorption bands LFP experiments were also carried out for TBL and its structural analogs, resorcinol and 3,5-dihydroxybenzyl alcohol (3,5-DHBA). Again, the decay of the sulfate radical at 450 nm is fast, while longlived transients are detected between 350 and 480 nm. In the case of resorcinol (Fig. 2A), the transient spectrum shows two bands at 405 and 425 nm (band 1 and 2) similar to those assigned to the phenoxyl radical (Ganapathi et al., 2000; Roginsky et al., 1998). For

3. Results and discussion 3.1. Formation and identification of transient species The transient absorption spectrum of the sulfate radical formed after 266 nm laser excitation of PS (44 mM) (eq (1)) shows a maximum at 450 nm (Fig. 1A) and decay in about 100 ms To investigate the reaction mechanisms of SAL/TBL with the sulfate radical, LFP experiments were carried out to identify the new generated transient species involved in the oxidation processes and to measure the reaction kinetics.

Fig. 1. LFP of PS and SAL. A: Transient spectra measured at the pulse end (A0) and 20 ms after (A20). B: Time profile absorbance monitored at 400 nm. C: Time profile absorbance monitored at 450 nm (266 nm, [SAL] ¼ 250 mM, [PS] ¼ 44 mM, pH ¼ 6).

Fig. 2. Transient absorption spectra of aqueous resorcinol (A), TBL (B) and 3,5-DHBA (C) in the presence of PS produced by LFP (the insert bands represent the phenoxyl radical absorption bands), and the time profiles (20 ms) of TBL/PS transient species at 410 nm (D1), 425 nm (D2), 440 nm (D3) and 500 nm (D4) (266 nm, [TBL] ¼ [3,5DHBA] ¼ [resorcinol] ¼ 250 mM, [PS] ¼ 44 mM, pH ¼ 6).

718

L. Zhou et al. / Water Research 123 (2017) 715e723

3.3.1. Effect of pH Solution pH might play a complex role in the simulated sunlight
activated PS process. Indeed, it had been demonstrated that SO4- is transformed to
OH under neutral or basic conditions (Eqs. (4) and (5)) (Liang et al., 2007; Fang et al., 2013b). In real world applications, solutions to be treated can show a very large pH range, it is therefore important to investigate the effect of solution pH. Thus, in the present work, the effect of initial pH (from 5.0 to 9.0) on the degradation efficiencies of SAL and TBL was evaluated. 2 þ SO$ 4 þ H2 O/SO4 þ $OH þ H 2  SO$ 4 þ OH /SO4 þ $OH

Fig. 3. Vibrational progression of the UVeVis absorption spectra of the resorcinol phenoxyl radical and the 3,5-DHBA phenoxyl radical obtained by DFT calculations at the B3LYP/6-31 þ G(d,p) level.

both TBL (Fig. 2B) and 3,5-DHBA (Fig. 2C) a third absorption band with a maximum at 440 nm (band 3) is observed. As this additional band shows the same decay rate as the two others (Fig. 2D1 to D3), we concluded that it belonged to the same species. Based on the DFT calculation results, the new band was also assigned to the feature of the phenoxyl radical. As seen in Fig. 3, the calculation indicated that the maximum absorption of resorcinol phenoxy radical appeared at 395 nm, while it was located at 440 nm for 3,5DHBA. Therefore, the CH2OH substituent on the ring induces a significant red-shift of the absorption spectrum of the phenoxyl radical.

3.2. Reactivity of sulfate radical towards SAL and TBL To measure the rate constant of reaction between the sulfate radical and SAL, we monitored the transient absorbance decay at
450 nm for SAL concentrations ranging from 0 to 250 mM. SO4- was 5 1 observed with a decay rate at 1.5  10 s in the absence of SAL. After the addition of SAL, the rate of sulfate radical decay begun to increase, evidencing the high reactivity of the sulfate radical towards SAL (Fig. 4A). By using a linear regression of the decay rate of sulfate radical versus SAL concentration, we could estimate the second order rate constant at (3.7 ± 0.3)  109 M1 s1 (Fig. 4B). The same procedure was used for TBL and the reaction rate constant of
SO4- with TBL was estimated at (4.2 ± 0.3)  109 M1 s1 (Fig. S3). The rate constants of reaction of
OH with SAL and TBL were also determined by the competition kinetic methods, using nitrobenzene as the molecular probe (SM, Text S1). Values of (5.3 ± 0.7)  109 M1 s1 and (5.9 ± 0.8)  109 M1 s1, respectively, were found. For both SAL and TBL, the reaction rate constants with
SO4- were comparable to those with
OH, demonstrating that photo
activated persulfate to generate SO4- could be an effective way to remove SAL and TBL from water solutions.

k < 2  103 s1

k5 z6:5  107 M1 s1

(4) (5)

The observed rate constants (kobs) under various initial pH values are presented in Table 1. The rates of SAL and TBL degradation increased as pH declined, acidic conditions being favorable for the oxidations. In addition, the pH values in all these experiments decreased during the reaction process, mostly due to the decomposition of persulfate and the formation of acidic by-products. The possible explanation of the pH effect is the base-activated PS decomposition. Sulfate radical can be generated at high pH values (Furman et al., 2010) (eq (6)), however, the yield of this decomposition reaction is much lower than that at acidic conditions from photon activated PS. $ 2 $ þ 2S2 O2 8 þ 2H2 O/SO4 þ 3SO4 þ O2 þ 4H

(6)

Zhang et al., have determined the steady-state concentrations of

SO4- and
OH ([SO4-]ss and [
OH]ss) by UV/PS under different pH
values (Zhang et al., 2015). Both [SO4-]ss and [
OH]ss values were lower at basic (pH 9) than that at acidic condition (pH 6).

3.3.2. Effect of PS dose The rate constants of SAL and TBL (kobs) degradation under different PS concentrations are illustrated in SM, Fig. S5. As reported in many articles, increasing the initial PS concentration significantly enhances the degradation rates of SAL and TBL (Nie et al., 2014; Zhou et al., 2017). This result suggests that higher PS concentrations yield higher level of reactive species by UV activation. Moreover, a linear relationship between kobs and PS concentration can be clearly observed, indicating that degradation rates
are closely related with the total amount of oxidants, such as SO4and
OH in the solution.

3.3.3. Identification of the involved reactive species
SO4- and
OH were considered to be the main oxidizing species during the decomposition processes involving SAL and TBL. Thus, to obtain a further insight into the reaction mechanism, experiments were conducted to identify the reactive species by using specific scavengers. Here, EtOH and TBA were used as the scavengers, as this method was intensively used in previous studies and proved to be reliable (Anipsitakis and Dionysiou, 2004). 7 1 1 EtOH þ SO$ 4 /intermediates; k7 ¼ ð1:6  7:7Þ  10 M s

(7) 3.3. UV/PS oxidation under simulated solar light In simulated solar light, irradiations for 2h of SAL and TBL in the presence of PS (6 mM) led to a removal of SAL and TBL of more than
94%, showing that using activated persulfate to generate SO4- is an effective way to remove SAL and TBL under experimental conditions.

EtOH þ $OH/intermediates; k8 ¼ ð1:2  2:8Þ  109 M1 s1 (8) 5 1 1 TBA þ SO$ 4 /intermediates; k9 ¼ ð4  9:1Þ  10 M s

(9)

L. Zhou et al. / Water Research 123 (2017) 715e723

719

Fig. 4. Decay of the sulfate radical monitored at 450 nm as a function of SAL concentration. Transient absorbance were obtained upon LFP (266 nm) for PS at 44 mM (A); Linear plot
of the pseudo-first order decay rate constant of SO4 (k, s1) versus the concentration of SAL (B).

Table 1 Pseudo first-order degradation rate constants of SAL and TBL in different conditions ([SAL]0 ¼ [TBL]0 ¼ 120 mM, [PS]0 ¼ 6 mM). Initial pH Scavenger SAL TBL concentration (mM) kobs (104 s1) kobs (104 s1) Scavenger effect 7.0 7.0 7.0 7.0 7.0 pH effect 5.0 6.0 7.0 8.0 9.0

0 EtOH 12 EtOH 120 TBA 12 TBA 120 e e e e e

2.18 1.02 0.33 1.92 1.57 2.30 2.27 2.18 2.11 1.80

± ± ± ± ± ± ± ± ± ±

0.13 0.06 0.02 0.06 0.05 0.09 0.05 0.13 0.09 0.10

2.51 2.14 0.71 2.27 2.10 2.68 2.61 2.51 2.36 1.86

± ± ± ± ± ± ± ± ± ±

0.08 0.03 0.02 0.07 0.02 0.06 0.05 0.08 0.06 0.11

h i SO$ ¼ 4

½$OH ¼

rSO$ 4   k5 OH þ kSO$ ;SAL ½SAL þ k9 ½TBA 4

r$OH k$OH;SAL ½SAL þ k10 ½TBA

TBA þ $OH/intermediates; k10 ¼ ð3:8  7:6Þ  108 M1 s1 (10) As can be seen from Eqs. (7)e(10), the reaction rate constant of
EtOH with
OH is approximately 50 fold higher than that with SO4-, whereas the reaction of
OH with alcohols that do not contain a
hydrogen, like TBA, is nearly 1000 times greater than that with SO4-.
Thus, the role played by OH during the degradation could be demonstrated by adding an excessive amount of TBA in the reaction system. Then by comparing the difference between the oxidation rates with an excess in TBA or EtOH it could be possible to evaluate
the SO4- participation. Table 1 displays the degradation rates of SAL and TBL by simulated sunlight activated PS process in the presence of EtOH and TBA, respectively. Comparing the initial rates in the absence and in the
presence of scavengers allows to determine the contribution of SO4
and OH in the reaction. Using the quasi-steady-state approxima
tion, one gets the generation and consumption rate of both SO4- and
OH:

rSO$ ¼ Ia F 4

h i h i h i i  $ , $ SO ½SAL þ k SO ½TBA þ k SO ½OH ¼ kSO, 5 9 ;SAL 4 4 4 4 (11) (12)

(13)

(14)

Based on these, one gets:

h i $ SO ½SAL rðSALÞTBA ¼ k$OH;SAL ½$OH½SAL þ kSO$ ;SAL 4 4 ¼

pKa for the phenol group of SAL and TBL are 10.12 and 10.64, respectively.

r$OH ¼ k$OH;SAL ½$OH½SAL þ k10 ½$OH½TBA

where Ia is the light intensity absorbed by PS and F, is the quantum yield of sulfate radical formation. Thus, the quasi-steady-state
concentration of SO4- and
OH can be written as:

k$OH;SAL ½SAL   r$OH k5 OH þ k$OH;SAL ½SAL þ k10 ½TBA þ

kSO$ ;SAL ½SAL 4 kSO$ ;SAL ½SAL þ k9 ½TBA 4

rSO$ 4

(15)

Here we used the mean values of k9 and k10 from Eqs. (9) and
(10) to calculate the contribution of SO4- and
OH. For instance, when [TBA]/[SAL] ¼ 100, r(SAL)TBA equals to:

rðSALÞTBA ¼ 0:085r$OH þ 0:98rSO$ 4

(16)

By comparing rSAL (in the absence of TBA) with r(SAL)TBA ([TBA]/ [SAL] ¼ 100), we could calculate that approximately 10% of SAL
removal was attributed to
OH, and SO4-, played a dominant role during the oxidation process. Similarly, for TBL, we calculated that
OH only contributed to about 9% of the total oxidation which is in a
good agreement with the fact that at pH 7 SO4- is the main reactive species. 3.4. Oxidation products 3.4.1. Oxidation products of SAL High resolution HPLC-MS analysis under both full scan and product ion scan mode were employed to characterize the structures of the oxidation products. In the case of SAL, four groups of oxidation products were identified in ESþ (Table 2). The product with m/z ¼ 238.1433 (corresponding to [MþH]þ) has lost 2 H compared to SAL. It was therefore assigned to 2-(tert-Butylamino)1-(4-hydroxy-3-hydroxymethylphenyl) ethanone, in which the CHOH group is replaced into a C]O group. Three peaks with m/ z ¼ 254.1382 were also detected. They have lost 2 H atoms and gained one O atom compared to SAL. We thus concluded that these

720

L. Zhou et al. / Water Research 123 (2017) 715e723

Table 2 Proposed structure of SAL and TBL (120 mM) degradation products in the presence of PS (12 mM) after irradiation of 60 min (in positive mode).

SAL

TBL

RT (min)

Measured exact mass [MþH]þ

Theoretical exact mass [MþH]þ

Dppm

Formula of neutral structure

0.90

240.1584

240.1594

4.1

C13H22O3N

0.72

132.1015

132.1019

2.8

C6H14O2N

1.70

238.1433

238.1438

3.3

C13H20O3N

1.17;1.60;1.77

254.1382

254.1387

2.6

C13H20O4N

1.74;1.93;2.10

376.1748

376.1755

3.0

C20H26O6N

0.90

226.1435

226.1438

2.7

C12H20O3N

1.37;1.95

322.0949

322.0955

2.1

C12H20O7NS

Proposed structure

R ¼ CH2NHC(CH3)3.

compounds are the hydroxylated ethanones for which the addition of eOH can take place at three different positions of aromatic ring. A product with m/z ¼ 376.1755 was found. Compared to SAL, it has gained C7H4O3. This compound is likely an adduct of the hydroxylated ethanone with the aromatic moiety of SAL constituted by the ring bearing the OH and CH2OH substituents. Three different peaks were observed, corresponding to the different possible isomers. In a good agreement, we detected a peak at m/z ¼ 132.1019 that is assumed to be the 2-(tert-butylamino)-acetic acid, i.e., the oxidized lateral chain of SAL. 3.4.2. Oxidation products of TBL Only two peaks at m/z ¼ 322.0949 were observed in the total ion chromatogram (TIC). They correspond to the gain of SO4 on the parent compound. The plausible structure is that given in Table 2, in which eOSO3H is added to the ring. This is the first time stable adducts produced by sulfate addition to the aromatic ring are detected. In most cases, such an addition occurs at the beginning of
the reaction with SO4-, afterwards the elimination of the sulfate group takes place to form other intermediates, since sulfate is an excellent leaving group (Aravindakumar et al., 2003). To further confirm this result, oxidation products of resorcinol and 3,5-DHBA (structurally simlar to TBL) with sulfate radical were also identified. For resorcinol, two peaks at m/z ¼ 204.9805 were observed from the total ion chromatogram (TIC, negative mode, SM Table S3), corresponding to the addition of 96 mass units to the parent

compound at three different positions on the ring. For 3,5-DHBA, three peaks of the sulfate addition products were also observed (m/ z ¼ 234.9916 in negative mode, SM Table S2). All these results
indicate that SO4- reacted with resorcinol and related compounds, TBL and 3,5-DHBA in a similar way. 3.5. Mechanisms of oxidation pathways The analytical study reveals that the oxidation products of SAL and TBL are very different. The initial step in the UV-PS oxidation is the formation of the phenoxyl radicals. Therefore, the two phenoxyl radicals rearrange differently. For TBL, the dominant pathway is the addition of the sulfate radical on the ring (Scheme 1A). As a metasubstituted phenol, the transformation of TBL into p-quinonemethides through an H atom abstraction driven by oxygen, for instance, is inaccessible. Consequently, the phenoxyl radical of TBL is only able to recombine with the other radicals present in the solution. The main recombination reaction is with the sulfate radical but recombination with another phenoxyl radical to form a dimer could be also possible. Yet, such dimeric products were not detected by HPLC-MS. In the case of SAL, the ortho and para position of CH2OH and CHOHR substituents with respect with the OH function makes possible the formation of a quinone-methide structure. Such quinone-methide structure could arise from an H abstraction by

oxygen or SO4-, and the addition of the SO4- on the ring could also

L. Zhou et al. / Water Research 123 (2017) 715e723

721

Scheme 1. Proposed oxidation pathways of TBL (A) and SAL (B) by UV/PS process.

take place (Scheme 1B). However, the main detected primary photoproduct is the carbonylated compound (m/z ¼ 238) and the mechanism of the quinone-methide rearrangement into this product needs to be clarified. With this goal, we performed DFT calculations, using 2,4-bis(hydroxymethyl)-phenol as a simplified model. Three intermediates can be proposed (Fig. 5), including the sulfate addition on ortho position (I) or para position (II), as well as the H atom abstraction to form the quinone-methide structure (III). The generation of species I is thermodynamically less favorable due to its relatively high energy level (38.3 kcal mol1) with respect to II. In contrast, there is no electronic energy barrier in the entrance channel to form the sulfate additional intermediate II. However, the HPLC-MS data did not reveal the existence of such an additional product. Although there is a very high energy barrier from II/III, we expect that the excess energy is sufficient to overcome the energy barrier towards the thermodynamically favored product III (Intrinsic Reaction Coordinate calculations showed that the Transition State at 15.2 kcal/mol in Fig. 5 connected II and III). Note that the proximity of the reactive groups in II (OSO 3 and H in CH2) may favor reaction II/III before thermal equilibrium is reached for II and thus benefit advantageously of the excess energy. Then, the intermediate III moves to the final products, species IV (quinone methide) or species V (ketone, aldehyde in the simulations). In Fig. 5, energies of IV and V correspond to separated products. Thermodynamically, V is more stable than IV. The transformation of IV to V has been demonstrated in the literature, this process being mediated by H2O (Chiang et al., 2002). Including
the solvatation energy of SO4- would lower significantly the

energies of IV and V. Thus, based on LFP, HPLC-MS and DFT calculation results, the degradation mechanism of SAL by the sulfate radical shown in Scheme 1B can be proposed. The main photoproduct deriving from the phenoxyl radical is the carbonylated product (m/z ¼ 238). This latter can undergo different reactions subsequently, either through attack of a sulfate radical or through photolysis. The formation of the hydroxylated product is a classic pathway of the sulfate radical attack on the phenolic compounds, which is similar to the reaction
of phenol with SO4-, resulting in the production of catechol or hydroquinone(Anipsitakis et al., 2006). Alternatively, direct electrophilic attack by
OH on the aromatic ring could also contribute to the production of the hydroxylated product. Both hydroxyl and hydroxymethylene group can act as electron donors, thus, the addition of eOH can occur on the three positions remaining on the ring. Moreover, the carbonyl compound can undergo a Norrish type I reaction in which the a-carbon bond is cleaved to give acyl and phenyl radicals. This reaction is likely the result of photolysis rather than an oxidation reaction by sulfate radical. Acyl radical could in turn be oxidized to generate 2-(tert-butylamino)-acetic acid, while the phenyl radical, would react with the eOH addition product (m/z 254) to produce the adduct (m/z ¼ 376). 4. Conclusion In summary, the feasibility of photo-activated persulfate for the removal of b2-adrenoceptor agonists SAL and TBL was verified, and UV/PS was found to be an effective approach for their degradation

722

L. Zhou et al. / Water Research 123 (2017) 715e723

Fig. 5. Reactions of the phenoxyl radical from SAL. Energy levels are shown in the inset (energies in kcal/mol). The reference energy level is that of the reactants. The bold and dashed lines represent local minima and transition states, respectively (energies of IV and V correspond to separated products. Including the solvation energy of
OSO 3 would lower significantly the energies of IV and V).

under natural water conditions (pH 5e8). The LFP technique provided the transient kinetic data such as second-order rate constants, which should be further more reliable than the classic steady-state photolysis method. Our results highlight the potential utility of using LFP. This work also revealed the difference in oxidation mechanisms related to the phenoxyl radicals reactivity. Two mechanisms, namely, eOSO3H addition (TBL) versus quinonemethide formation (SAL), were proposed by using LFP, HPLC-MS as well as the model substituted phenols (2-HBA and 4-HBA for SAL, 3,5-DHBA and resorcinol for TBL). This is the first time one demonstrates the selectivity of sulfate radical with different substituted phenols. Considering that a large variety of substitued phenols are present in natural waters, the results of this work could help better understand their fates during AOPs treatment.

Acknowledgements The authors gratefully acknowledge the financial support from the China Scholarship Council for Lei Zhou to study at IRCELYON in France.

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.watres.2017.07.025. References Ahmed, M.M., Barbati, S., Doumenq, P., Chiron, S., 2012. Sulfate radical anion oxidation of diclofenac and sulfamethoxazole for water decontamination. Chem. Eng. J. 197, 440e447. Anipsitakis, G.P., Dionysiou, D.D., 2004. Radical generation by the interaction of transition metals with common oxidants. Environ. Sci. Technol. 38 (13), 3705e3712. Anipsitakis, G.P., Dionysiou, D.D., Gonzalez, M.A., 2006. Cobalt-mediated activation of peroxymonosulfate and sulfate radical attack on phenolic compounds. Implications of chloride ions. Environ. Sci. Technol. 40 (3), 1000e1007. Aravindakumar, C.T., Schuchmann, M.N., Rao, B.S., von Sonntag, J., von Sonntag, C., 2003. The reactions of cytidine and 2-deoxycytidine with SO 4_ revisited. Pulse radiolysis and product studies. Org. Biomol. Chem. 1, 401e408. Bound, J.P., Voulvoulis, N., 2006. Predicted and measured concentrations for selected pharmaceuticals in UK rivers: implications for risk assessment. Water Res. 40 (15), 2885e2892. Brede, O., Kapoor, S., Mukherjee, T., Hermann, R., Naumov, S., 2002. Diphenol radical cations and semiquinone radicals as direct products of the free electron transfer from catechol, resorcinol and hydroquinone to parent solvent radical cations. Phys. Chem. Chem. Phys. 4 (20), 5096e5104.

L. Zhou et al. / Water Research 123 (2017) 715e723 Calamari, D., Zuccato, E., Castiglioni, S., Bagnati, R., Fanelli, R., 2003. Strategic survey of therapeutic drugs in the rivers Po and Lambro in northern Italy. Environ. Sci. Technol. 37 (7), 1241e1248. Chen, X., Larsen, D.S., Bradforth, S.E., van Stokkum, I.H., 2011. Broadband spectral probing revealing ultrafast photochemical branching after ultraviolet excitation of the aqueous phenolate anion. J. Phys. Chem. A 115 (16), 3807e3819. Chiang, Y., Kresge, A., Zhu, Y., 2002. Flash photolytic generation and study of pquinone methide in aqueous solution. An estimate of rate and equilibrium constants for heterolysis of the carbon-bromine bond in p-hydroxybenzyl bromide. J. Am. Chem. Soc. 124 (22), 6349e6356. Dahmani, M.A., Huang, K., Hoag, G.E., 2006. Sodium persulfate oxidation for the remediation of chlorinated solvents (USEPA superfund innovative technology evaluation program). Water, Air, & Soil Pollut. Focus 6 (1e2), 127e141. Fang, G.D., Dionysiou, D.D., Wang, Y., Al-Abed, S.R., Zhou, D.M., 2012. Sulfate radicalbased degradation of polychlorinated biphenyls: effects of chloride ion and reaction kinetics. J. Hazard. Mater. 227, 394e401. Fang, G.D., Dionysiou, D.D., Zhou, D.-M., Wang, Y., Zhu, X.-D., Fan, J.-X., Cang, L., Wang, Y.J., 2013a. Transformation of polychlorinated biphenyls by persulfate at ambient temperature. Environ. Sci. Technol. 90 (5), 1573e1580. Fang, G., Gao, J., Dionysiou, D.D., Liu, C., Zhou, D., 2013b. Activation of persulfate by quinones: free radical reactions and implication for the degradation of PCBs. Environ. Sci. Technol. 47 (9), 4605e4611. Feitelson, J., Hayon, E., Treinin, A., 1973. Photoionization of phenols in water. Effects of light intensity, oxygen, pH, and temperature. J. Am. Chem. Soc. 95 (4), 1025e1029. Frisch, M.J., Trucks, G.W., Schlegel, H.B., Scuseria, G.E., Robb, M.A., Cheeseman, J.R., Scalmani, G., Barone, V., Mennucci, B., Petersson, G.A., Nakatsuji, H., Caricato, M., Li, X., Hratchian, H.P., Izmaylov, A.F., Bloino, J., Zheng, G., Sonnenberg, J.L., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Vreven, T., Montgomery, J.A., Peralta Jr., J.E., Ogliaro, F., Bearpark, M., Heyd, J.J., Brothers, E., Kudin, K.N., Staroverov, V.N., Keith, T., Kobayashi, R., Normand, J., Raghavachari, K., Rendell, A., Burant, J.C., Iyengar, S.S., Tomasi, J., Cossi, M., Rega, N., Millam, J.M., Klene, M., Knox, J.E., Cross, J.B., Bakken, V., Adamo, C., Jaramillo, J., Gomperts, R., Stratmann, R.E., Yazyev, O., Austin, A.J., Cammi, R., Pomelli, C., Ochterski, J.W., Martin, R.L., Morokuma, K., Zakrzewski, V.G., Voth, G.A., Salvador, P., Dannenberg, J.J., Dapprich, S., Daniels, A.D., Farkas, O., Foresman, J.B., Ortiz, J.V., Cioslowski, J., Fox, D.J., 2010. Gaussian 09, Revision C.01. Gaussian, Inc., Wallingford CT. Furman, O.S., Teel, A.L., Watts, R.J., 2010. Mechanism of base activation of persulfate. Environ. Sci. Technol. 44 (16), 6423e6428. Ganapathi, M.R., Hermann, R., Naumov, S., Brede, O., 2000. Free electron transfer from several phenols to radical cations of non-polar solvents. Phys. Chem. Chem. Phys. 2 (21), 4947e4955. Gara, P.M.D., Bosio, G.N., Gonzalez, M.C., Russo, N., del Carmen Michelini, M.,
rtire, D.O., 2009. A combined theoretical and experimental study Diez, R.P., Ma on the oxidation of fulvic acid by the sulfate radical anion. Photochem. Photobiol. Sci. 8 (7), 992e997. He, X., Mezyk, S.P., Michael, I., Fatta-Kassinos, D., Dionysiou, D.D., 2014. Degradation kinetics and mechanism of b-lactam antibiotics by the activation of H2O2 and Na2S2O8 under UV-254nm irradiation. J. Hazard. Mater. 279, 375e383. House, D.A., 1962. Kinetics and mechanism of oxidations by peroxydisulfate. Chem. Rev. 62 (3), 185e203. Huerta-Fontela, M., Galceran, M.T., Ventura, F., 2011. Occurrence and removal of pharmaceuticals and hormones through drinking water treatment. Water Res. 45 (3), 1432e1442. Kosma, C.I., Lambropoulou, D.A., Albanis, T.A., 2014. Investigation of PPCPs in wastewater treatment plants in Greece: occurrence, removal and environmental risk assessment. Sci. Total Environ. 466e467 (0), 421e438. Lee, Y.C., Lo, S.L., Chiueh, P.T., Chang, D.G., 2009. Efficient decomposition of perfluorocarboxylic acids in aqueous solution using microwave-induced persulfate. Water Res. 43 (11), 2811e2816. Liang, C., Wang, Z.S., Bruell, C.J., 2007. Influence of pH on persulfate oxidation of TCE at ambient temperatures. Chemosphere 66 (1), 106e113.

723

Lutze, H.V., Bircher, S., Rapp, I., Kerlin, N., Bakkour, R., Geisler, M., von Sonntag, C., Schmidt, T.C., 2015. Degradation of chlorotriazine pesticides by sulfate radicals and the influence of organic matter. Environ. Sci. Technol. 49 (3), 1673e1680. Matta, R., Tlili, S., Chiron, S., Barbati, S., 2011. Removal of carbamazepine from urban wastewater by sulfate radical oxidation. Environ. Chem. Lett. 9 (3), 347e353. 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. Nfodzo, P., Choi, H., 2011. Sulfate radicals destroy pharmaceuticals and personal care products. Environ. Eng. Sci. 28 (8), 605e609. Nie, M., Yang, Y., Zhang, Z., Yan, C., Wang, X., Li, H., Dong, W., 2014. Degradation of chloramphenicol by thermally activated persulfate in aqueous solution. Chem. Eng. J. 246, 373e382.
nchez-Pe
rez, J., 2011. Combination of advanced oxidation Oller, I., Malato, S., Sa processes and biological treatments for wastewater decontaminationda review. Sci. Total. Environ. 409 (20), 4141e4166. Oturan, M.A., Aaron, J.J., 2014. Advanced oxidation processes in water/wastewater treatment: principles and applications. A review. Crit. Rev. Env. Sci. Tec. 44 (23), 2577e2641. Rhodes, M.C., Seidler, F.J., Qiao, D., Tate, C.A., Cousins, M.M., Slotkin, T.A., 2004. Does pharmacotherapy for preterm labor sensitize the developing brain to environmental neurotoxicants? Cellular and synaptic effects of sequential exposure to terbutaline and chlorpyrifos in neonatal rats. Toxicol. Appl. Pharm. 195 (2), 203e217. Rickman, K.A., Mezyk, S.P., 2010. Kinetics and mechanisms of sulfate radical oxidation of b-lactam antibiotics in water. Chemosphere 81 (3), 359e365. Roginsky, V.A., Pisarenko, L.M., Bors, W., Michel, C., Saran, M., 1998. Comparative pulse radiolysis studies of alkyl-and methoxy-substituted semiquinones formed from quinones and hydroquinones. J. Chem. Soc. 94 (13), 1835e1840. Santoro, F., Cappelli, C., Barone, V., 2011. Effective time-independent calculations of vibrational resonance raman spectra of isolated and solvated molecules including Duschinsky and Herzberg-Teller effects. J. Chem. Theory Comput. 7, 1824e1839. Silverstein, D.W., Govind, N., van Dam, H.J., Jensen, L., 2013. Simulating one-photon absorption and resonance raman scattering spectra using analytical excited state energy gradients within time-dependent density functional theory. J. Chem. Theory. Comput. 9 (12), 5490e5503. Slotkin, T.A., Seidler, F.J., 2013. Terbutaline impairs the development of peripheral noradrenergic projections: potential implications for autism spectrum disorders and pharmacotherapy of preterm labor. Neurotoxicol. Teratol. 36, 91e96. Tan, C., Gao, N., Deng, Y., Zhang, Y., Sui, M., Deng, J., Zhou, S., 2013. Degradation of antipyrine by UV, UV/H2O2 and UV/PS. J. Hazard. Mater. 260, 1008e1016. Ternes, T.A., 1998. Occurrence of drugs in German sewage treatment plants and rivers1. Water Res. 32 (11), 3245e3260. Xu, X.R., Li, X.Z., 2010. Degradation of azo dye Orange G in aqueous solutions by persulfate with ferrous ion. Sep. Purif. Technol. 72 (1), 105e111. Yang, X., Flowers, R.C., Weinberg, H.S., Singer, P.C., 2011. Occurrence and removal of pharmaceuticals and personal care products (PPCPs) in an advanced wastewater reclamation plant. Water Res. 45 (16), 5218e5228. Zhang, R., Sun, P., Boyer, T.H., Zhao, L., Huang, C.H., 2015. Degradation of pharmaceuticals and metabolite in synthetic human urine by UV, UV/H2O2, and UV/ PDS. Environ. Sci. Technol. 49 (5), 3056e3066. Zhang, T., Chen, Y., Leiknes, T., 2016. Oxidation of refractory benzothiazoles with PMS/CuFe2O4: kinetics and transformation intermediates. Environ. Sci. Technol. 50 (11), 5864e5873. Zhao, D., Liao, X., Yan, X., Huling, S.G., Chai, T., Tao, H., 2013. Effect and mechanism of persulfate activated by different methods for PAHs removal in soil. J. Hazard. Mater. 254, 228e235. Zhou, L., Zheng, W., Ji, Y., Zhang, J., Zeng, C., Zhang, Y., Wang, Q., Yang, X., 2013. Ferrous-activated persulfate oxidation of arsenic (III) and diuron in aquatic system. J. Hazard. Mater. 263, 422e430. Zhou, L., Ferronato, C., Chovelon, J.M., Sleiman, M., Richard, C., 2017. Investigations of diatrizoate degradation by photo-activated persulfate. Chem. Eng. J. 311, 28e36.