Vous êtes sur la page 1sur 9

Water Research 103 (2016) 215e223

Contents lists available at ScienceDirect

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

Oxidative degradation of triclosan by potassium permanganate:

Kinetics, degradation products, reaction mechanism, and toxicity
Jing Chen, Ruijuan Qu*, Xiaoxue Pan, Zunyao Wang**
State Key Laboratory of Pollution Control and Resources Reuse, School of the Environment, Nanjing University, Jiangsu, Nanjing 210023, PR China

a r t i c l e i n f o a b s t r a c t

Article history: In this study, we systematically investigated the potential applicability of potassium permanganate for
Received 22 February 2016 removal of triclosan (TCS) in water treatment. A series of kinetic experiments were carried out to study
Received in revised form the influence of various factors, including the pH, oxidant doses, temperature, and presence of typical
18 July 2016
anions (Cl, SO2 
4 , NO3 ), humic acid (HA), and fulvic acid (FA) on triclosan removal. The optimal reaction
Accepted 18 July 2016
conditions were: pH ¼ 8.0, [TCS]0:[KMnO4]0 ¼ 1:2.5, and T ¼ 25  C, where 20 mg/L of TCS could be
Available online 20 July 2016
completely degraded in 120 s. However, the rate of TCS (20 mg/L) oxidation by KMnO4
([TCS]0:[KMnO4]0 ¼ 1:2.5) was 1.64  103 mg L1$h1, lower than that at an initial concentration of
20 mg/L (2.24  103 mg L1$h1). A total of eleven products were detected by liquid chromatography-
Permanganate quadrupole-time-of-flight-mass spectrometry (LC-Q-TOF-MS) analysis, including phenol and its de-
Product identification rivatives, benzoquinone, an organic acid, and aldehyde. Two main reaction pathways involving CeO bond
Degradation pathway cleavage (C(8)eO(7)) and benzene ring opening (in the less chlorinated benzene ring) were proposed,
Acute toxicity and were further confirmed based on frontier electron density calculations and point charges. Further-
Frontier electron density more, the changes in the toxicity of the reaction solution during TCS oxidation by KMnO4 were evaluated
by using both the luminescent bacteria Photobacterium phosphoreum and the water flea Daphnia magna.
The toxicity of 20 mg/L triclosan to D. magna and P. phosphoreum after 60 min was reduced by 95.2% and
43.0%, respectively. Phenol and 1,4-benzoquinone, the two representative degradation products formed
during permanganate oxidation, would yield low concentrations of DBPs (STHMFP, 20.99e278.97 mg/mg;
SHAAFP, 7.86  10445.77 mg/mg) after chlorination and chloramination. Overall, KMnO4 can be used as
an effective oxidizing agent for TCS removal in water and wastewater treatment.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction Europe (Halden and Paull, 2005; Pintado-Herrera et al., 2014;

Young et al., 2008). Given the widespread use, this compound has
Triclosan (5-chloro-2-(2,4-dichlorophenoxy)-phenol, TCS) is a been detected in surface water, wastewater, sediment, soil, organ-
broad-spectrum antimicrobial and preservative agent that is widely isms, and even in human milk (Fu et al., 2016; Toms et al., 2011). The
used in a range of consumer products such as toothpastes, anti- high octanol water partitioning coefficient of TCS (log Kow ¼ 4.8 at
septic soaps, detergents, cosmetics, plastic kitchenware, socks, pH ¼ 7) indicates that it can bioaccumulate in the biota and be
carpets and toys (Bedoux et al., 2012; Reiss et al., 2002; Singer et al., biomagnified along the food chain. Triclosan is highly toxic to
2002). The global demand for TCS has continued to increase over freshwater aquatic species like green algae, the water flea Daphnia
the last thirty years, and about 350 tons are consumed every year in magna, and fish (zebrafish, fathead minnows, bluegill sunfish) (Chen
et al., 2014; Dann and Hontela, 2011). Previous researches have
also shown that TCS may cause bacterial resistance, skin irritation,
* Corresponding author. State Key Laboratory of Pollution Control and Resources endocrine disruption, and even increase the formation of carcino-
Reuse, School of the Environment, Xianlin Campus, Nanjing University, Jiangsu, genic by-products (Brausch and Rand, 2011; Dayan, 2007; Novo
Nanjing 210023, PR China. et al., 2013).
** Corresponding author. Over the past few decades, various water treatment techniques,
E-mail addresses: quruijuan0404@nju.edu.cn (R. Qu), wangzy@nju.edu.cn
including ozonation, Fenton, photolysis, and permanganate
(Z. Wang).

0043-1354/© 2016 Elsevier Ltd. All rights reserved.
216 J. Chen et al. / Water Research 103 (2016) 215e223

treatment have been used to degrade TCS (Buth et al., 2009; Suarez phase comprising methanol (80%) and water (20%), a flow rate of
s et al., 2007; Zhang and Huang, 2003). Compared
et al., 2007; Sire 1.0 mL/min, an injection volume of 10 mL, and a detection wave-
with other oxidizing agents, permanganate possesses the attractive length of 230 nm.
characteristics of high efficiency, relatively low cost, comparative Identification of the TCS degradation products was performed
stability, and ease of handling. Jiang et al. (2009) investigated the using the SPE-LC-MS method. At various reaction times, the solu-
kinetics of triclosan degradation by permanganate, revealing the tion was concentrated by using a solid phase extraction (SPE)
catalytic role of MnO2 formed in situ under slightly acidic condi- workstation using CNWBOND LC-C18 SPE cartridges
tions. Wu et al. (2012) found that degradation of TCS by perman- (2.CA0955.0001, 1 g, 6 mL). The procedure was slightly modified
ganate followed pseudo-first order kinetics in a buffered DI water from previous approaches (Loos et al., 2010; Ji et al., 2013). After
system at pH 7.0 and 8.6. In a drinking water system treated with activation with 5 mL methanol and 5 mL ultrapure water, each
permanganate, 2,4-dichlorophenol was identified as the major column was loaded with 3 mL of sample, washed with 5 mL ul-
oxidation product of TCS. Unfortunately, to date, very little infor- trapure water, and then dried under vacuum. After eluting with
mation is available on the transformation products, reaction 3 mL methanol, the degradation products in the extracts were
mechanism and toxicity changes of TCS during oxidative degrada- subjected to LC-MS analysis. A similar SPE method was employed to
tion by KMnO4, though such knowledge is essential for practical enable LC determination of TCS in the kinetics experiment per-
implementation of the permanganate oxidation technique. formed at a low initial concentration (20 mg/L), where the samples
In this study, we systematically investigate the oxidation of TCS were concentrated 200-fold.
by permanganate. A series of operating parameters influencing the LC-MS analysis was conducted using a high resolution hybrid
degradation processes, including the solution pH, temperature, and quadrupole time-of-flight mass spectrometer (Triple TOF 5600, AB
the presence of typical inorganic anions, humic, and fulvic acid, are Sciex, Foster City, CA) equipped with an electrospray ion source. A
evaluated to explore the optimum reaction conditions. The degra- 5 mL sample was injected into the source using an Agilent 1260
dation products are identified by liquid chromatography-mass infinity high performance liquid chromatography (HPLC) system.
spectrometric analysis, and the reaction pathways are thus pro- Chromatographic separation was performed on a Thermo BDS
posed. Frontier electron density (FED) calculations and point Hypersil C18 column (2.1 mm  100 mm, particle size: 2.4 mm;
charges analysis are also employed to further confirm the identity Thermo Fisher Scientific, Waltham, MA) maintained at 30  C, and
of the intermediate products. Changes in the toxicity of triclosan the mobile phase consisting of 0.3% formic acid in water (A) and
during the oxidation process are evaluated using the luminescent methanol (B) was eluted at a flow rate of 200 mL/min. The linear
bacteria Photobacterium phosphoreum and water flea Daphnia gradient was decreased from 90% A (2 min) to 10% A in 1 min,
magna. In addition, the generation of disinfection-by-products from maintained for 23 min, then returned to the starting condition in
triclosan degradation products after implementing chlorination 1 min, followed by an 8 min equilibration. Mass spectra (m/z
and chloramination in a full-scale treatment plant are discussed 70e1000) were recorded in negative ion mode by using an elec-
herein. trospray ion (ESI) source with the following ion source parameters:
ion source gas 1, 55 psi; ion source gas 2, 55 psi; curtain gas, 35
2. Materials and methods psi; temperature, 550  C; ionspray voltage floating, 4500 V;
declustering potential, -80 V; collision energy, 10 V. Nitrogen gas
2.1. Chemicals was used throughout. Following the MS analysis, the individual
parent ions of the possible transformation products of TCS were
Triclosan (purity  99%) was purchased from Aladdin (Shanghai, manually selected for MS/MS analysis. A sweeping collision energy
China). The HPLC grade methanol and formic acid were supplied by setting of 30 ± 10 V was applied for collision-induced dissociation.
Merck (Darmstadt, Germany). Other chemicals were of analytical The high-resolution LC-MS data were acquired with Analyst TF
grade and were used as received. Stock solutions of TCS, KMnO4 and software (Version 1.6, AB Sciex) and processed using PeakView
NaNO2 were prepared with ultrapure water at concentrations of software (Version 1.2, AB Sciex).
500 mg/L, 5 g/L, and 30 g/L, respectively, and were all stored in
brown bottles at 4  C in a refrigerator. 2.4. Toxicity assay

2.2. Oxidation procedures Changes in the toxicity of the triclosan solution during KMnO4
oxidation were assessed by using the luminescent bacteria
The oxidation reactions were initiated by adding 30 mL of po- P. phosphoreum and water flea D. magna. The lyophilized lumines-
tassium permanganate solution (5 g/L) to centrifuge tubes con- cent bacteria P. phosphoreum was supplied by the Institute of Soil
taining 3 mL of TCS solution (20 mg/L). The mixture solution was Science, Chinese Academy of Sciences (Nanjing, China). The
agitated with a vortex stirrer to ensure a uniform reaction. After 10, D. magna was purchased from the Research Center from Eco-
30, 60, 90, 120, 180, and 300 s, permanganate was quenched with Environmental Sciences, Chinese Academy of Sciences (Wuhan,
30 mL of NaNO2 stock solution (the reason for choosing NaNO2 as China). For the toxicity texts, the reaction solutions were centri-
quenching agent is shown in Text S1). All experiments were per- fuged and filtered to remove the insoluble product MnO2. The
formed in triplicate. The pH of the solutions was adjusted using HCl bioluminescence test was carried out by following the ISO standard
or NaOH (Merck). For product identification, all the sample solu- method. After activation and incubation, the bacteria were exposed
tions were concentrated by solid phase extraction (SPE) to elimi- to the reaction solutions for 15 min at 25  C, and the biolumines-
nate interference from MnO2 precipitation and certain other ions cence was measured on a Tecan Infinite 200® PRO multimode
such as Naþ, MnO  
4 , NO2 , and Cl . microplate reader (Tecan, Switzerland). After 15 min exposure to
the reaction solutions at pH 8.0 and 25  C, the luminescence in-
2.3. Analytical methods tensity was measured; the result is expressed as the inhibition
percentage versus the black control. Six replicate tests were per-
The concentration of TCS was quantified by using a HPLC system formed for all bioluminescence experiments. The toxicity of NaNO2
(Agilent 1100) equipped with a C18 column (Agilent Zorbax 300SB) to P. phosphoreum was considered negligible since the concentra-
with UV detection. The analysis was conducted using a mobile tion used in this study is much smaller than the non-lethal
J. Chen et al. / Water Research 103 (2016) 215e223 217

concentration (1.1 g/L). Acute toxicity testing with D. magna was of MnO 4 decreased as compared to that in acidic medium, the re-
conducted in accordance with the OECD guideline. Seven daphnia action rate obtained at pH ¼ 8.0 was still very high. At this pH point,
neonates (<24 h) were exposed to 20 mL of sample in a 50 mL the percentage of ionic TCS was almost identical to that of the
beaker, which was held at 20 ± 2  C in an illuminating incubator molecular species. The strong electron-donating ability of the
with a 14 h: 10 h light: dark photoperiod. D. magna was not fed deprotonated hydroxyl groups makes TCS easily oxidized by MnO 4
during the experimental period. After 24 h exposure, the number of (Jiang et al., 2009; Qiao et al., 2014; Waldemer and Tratnyek, 2006).
dead organisms (dead number) was recorded and the inhibition Consequently, the degradation kinetics of TCS at different pH can
rate was calculated for each sample. All sample tests and control result from a combination of the effects of the pH on TCS speciation
experiments for D. magna were performed in triplicate. A reagent and the oxidizing properties of MnO 4 . Because the highest removal
control with the same concentration of NaNO2 was also assessed to efficiency was achieved at pH 8.0, all subsequent experiments were
eliminate the possible influence of NaNO2. performed at this pH.

2.5. Chlorination and chloramination 3.1.2. Effect of oxidant doses

It is well known that the oxidant dose dramatically influences
The trihalomethane formation potential (THMFP) and haloacetic the degradation of organic substances (Abdelraheem et al., 2015;
acids formation potential (HAAFP) measurements were performed Wacławek et al., 2016). In order to achieve an efficient degrada-
according to Standard Method 5710B. Monochloramine stock so- tion rate at low cost, the oxidation of triclosan by permanganate
lution was prepared by reference to the method of Chen and Young was evaluated for initial KMnO4 concentrations of ranging from
(2008). The chlorine dosage for each water sample was determined 10 mg/L (TCS:KMnO4 ¼ 1:0.5) to 150 mg/L (TCS:KMnO4 ¼ 1:7.5).
such that the final residual concentration of chlorine in the sample Fig. 1B shows that the removal efficiency increased with an
after 5 days of incubation at 25  C was 3e5 mg/L. Samples were increasing amount of permanganate. After 300 s, triclosan was
adjusted to a pH of 7.0 ± 0.2 using H2SO4 and NaOH. The neutralized degraded by 42.5%, 82.3%, and nearly 100% at the TCS:KMnO4 ratios
solution was then buffered with a phosphate solution prior to in- of 1:0.5, 1:1.5, and 1:2.5, respectively. When the TCS:KMnO4 ratio
cubation in amber bottles at 25  C for five days. At the end of the changed from 1:2.5 to 1:7.5, the influence of permanganate on the
incubation period, the samples were dechlorinated using sodium removal efficiency was not so obvious, and nearly complete
sulfite (Na2SO3). The chlorinated samples were extracted with removal was achieved at 120 s for the three oxidant dosages. From
methyl-tert butyl ether (MTBE) to analyze four kinds of tri- the economic perspective, the ratio of 1:2.5 was considered as the
halomethanes (THMs), i.e. Chloroform (TCM), bromodichloro- most suitable KMnO4 dosage for triclosan removal.
methane (BCDM), dibromochloromethane (DBCM), and
bromoform (TBM) using a modified EPA method 551.1. Five species 3.1.3. Effect of temperature
of haloacetic acids (HAAs), i.e. monochloroacetic acid (MCAA), The degradation of TCS was evaluated at various temperatures
dichloroacetic acid (DCAA), trichloroacetic acid (TCAA), mono- in the range of 15e35  C to examine the influence of temperature.
bromoacetic acid (MBAA), and dibromoacetic acid (DBAA), were As shown in Fig. 1C, the rate of triclosan degradation by perman-
analyzed in accordance with EPA method 552.2. MTBE was used as ganate increased with increasing temperature. Specifically, the
the sole extracting solvent. The concentrations of THMs and HAAs remaining triclosan after 10 s decreased from 57.5% to 19.4%, when
were measured using a gas chromatograph (Agilent 6890A) with an the reaction temperature increased from 15  C to 35  C. The
electron capture detector (ECD). temperature-dependence is not unexpected as the well-established
Arrhenius law suggests that the number of activated reactant
3. Results and discussion molecules will increase at a higher temperature. On the whole, the
removal of triclosan was fast at the three temperatures, with
3.1. Kinetic study degradation ratios of 94.4e100.0% in 90 s. To avoid energy con-
sumption associated with heating, 25  C was selected as the best
3.1.1. Effect of initial solution pH reaction temperature.
The pH of the aqueous solution may greatly influence the
oxidative degradation of TCS by potassium permanganate, due to 3.1.4. Effect of the typical anions and humic acid
changes in the existing species of TCS (pKa ¼ 8.1) (Kim et al., 2013) Since industrial wastewaters and raw waters are complex sys-
and the oxidation-reduction potential (E0) of KMnO4 (Yan and tems, the influence of many inorganic ions and organic species
Schwartz, 1999). The following reactions proceed under the should be examined for treatment of real TCS-containing water
stated conditions, under acidic conditions: MnO þ
4 þ 8H þ 5e #
with permanganate. In this study, the typical anions (Cl, SO2 4 ,

Mn þ 4H2O, E0 ¼ 1.51 V; under neutral or slightly alkaline con- NO 3 ), humic acid (HA), and fluvic acid (FA) exerted a slight effect on
ditions: MnO e 
4 þ 2H2O þ 3e # MnO2 þ 4OH , E0 ¼ 0.58 V; under the degradation of triclosan (Fig. S1). The fast reaction between TCS
alkaline conditions, MnO e 2
4 þ e # MnO4 , E0 ¼ 0.56 V. In this and KMnO4 (nearly 100% removal in 90 s) may account for the
study, the solution pH was varied from 2.0 to 10.0 to evaluate the insignificant influence of the investigated water components.
effect of the pH on the TCS degradation efficiency. As shown in
Fig. 1A, the removal of TCS is highly dependent on the solution pH. 3.1.5. Kinetic study in a low initial concentration
Under acidic conditions (pH ¼ 2.0e4.0), the oxidation efficiency To get a better understanding of the oxidation efficiency of tri-
decreased with increasing pH. The lowest degradation was closan in realistic environmental conditions, the tests were per-
observed at pH ¼ 4.0, where 22% TCS remained after 5 min of formed at a relatively low initial concentration (20 mg/L) with
oxidation. In contrast, the oxidation efficiency increased rapidly different dosages of the oxidizing agent. The concentration varia-
with an increase of the pH from 4.0 to 8.0, and TCS was almost tion of TCS at five different ratios of TCS:KMnO4 (1:2.5, 1:5, 1:10,
completely degraded in 120 s at pH ¼ 8.0. With a further increase of 1:50, and 1:100) was shown in Fig. 1D. The result at low initial
the pH to 10.0, the removal rate decreased to some degree. For the concentration (20 mg/L) showed that the degradation of TCS at a
studied pH range, the degradation rate was relatively high at ratio of TCS:KMnO4 ¼ 1:2.5 was in accordance of the pseudo-first-
pH ¼ 2.0, which is attributed to the high oxidation-reduction po- order reaction kinetics. That is, the reaction rate is proportional to
tential of KMnO4 (Dash et al., 2009); although the oxidizing ability the substrate concentration. The rate of TCS (20 mg/L) oxidation by
218 J. Chen et al. / Water Research 103 (2016) 215e223

1.0 pH = 2 1.0 [TCS]0:[KMnO4]0 = 1:0.5

pH = 3 [TCS]0:[KMnO4]0 = 1:1.5
0.8 pH = 4
pH = 5
0.8 [TCS]0:[KMnO4]0 = 1:2.5
pH = 6

0.6 pH = 7 0.6
pH = 8 [TCS]0:[KMnO4]0 = 1:5

pH = 9 [TCS]0:[KMnO4]0 = 1:7.5
0.4 pH = 10 0.4

0.2 0.2

0.0 0.0
0 50 100 150 200 250 300 0 50 100 150 200 250 300
Time (s) Time (s)
1.0 1.0 [TCS]0:[KMnO4]0 = 1:2.5
25 [TCS]0:[KMnO4]0 = 1:5
0.8 35 0.8 [TCS]0:[KMnO4]0 = 1:10
[TCS]0:[KMnO4]0 = 1:50
0.6 0.6

[TCS]0:[KMnO4]0 = 1:100

0.4 0.4

0.2 0.2

0.0 0.0
0 50 100 150 200 250 300 0 10 20 30 40 50
Time (s) Time (h)

Fig. 1. Degradation kinetics of triclosan by permanganate. (A) Effect of pH. Experimental conditions: [TCS]0 ¼ 20 mg/L, [TCS]0:[KMnO4]0 ¼ 1:2.5, T ¼ 25  C. (B) Effect of oxidant
doses. Experimental conditions: [TCS]0 ¼ 20 mg/L, pH ¼ 8.0, T ¼ 25  C. (C) Effect of temperature. Experimental conditions: [TCS]0 ¼ 20 mg/L, pH ¼ 8.0, [TCS]0:[KMnO4]0 ¼ 1:2.5. (D)
Removal of TCS at a low initial concentration. Experimental conditions: [TCS]0 ¼ 20 mg/L, pH ¼ 8.0, T ¼ 25  C.

KMnO4 ([TCS]0:[KMnO4]0 ¼ 1:2.5) was 1.64  103 mg L1$h1, a total of eleven new peaks were observed with evident changes in
compared with that at an initial concentration of 20 mg/L the relative intensity at different reaction times. Notably, the six
(2.24  103 mg L1$h1). Thus, permanganate effectiveness for TCS products with precursor ions at m/z 330.94, 274.96, 272.98, 230.96,
degradation is lower at a low initial TCS concentration that is only 204.95 and 148.95 were observed for the first time. Typical frag-
one in a thousand of high initial concentration. However, the ment losses like CO, CO2, COeCO, COeCO2 and CO2eCO2 were
elimination rate also increased with increasing oxidant dosage. The observed from the MS/MS analysis, implying the presence of
remaining TCS after 48 h was 30% and 20% at the respective carbonyl, carboxylic, and aldehyde groups in the degradation
TCS:KMnO4 ratios of 1:2.5 and 1:5, whereas complete removal was products.
observed in 12 h, 4 h, and 30 min at the ratios of 1:10, 1:50 and P2, P4: The chromatographic peak with a retention time of
1:100, respectively. Thus, a typical dosage of 2 mg/L KMnO4 12.13 min comprised an ion cluster with m/z 274.96/276.96/278.96,
([TCS]0:[KMnO4]0 ¼ 1:100) was required for efficient removal of indicating that this compound (P2) contained two Cl atoms. In the
TCS at the mg/L level in water and wastewater treatment (Liu et al., product ion spectra (Fig. 3B), the fragment ion peaks with m/z
2009). This can be used as a cost-effective technique for reducing 230.96 and 186.95 indicate sequential losses of CO2 from the parent
the operational fee to less than 4.56  103 US dollars per ton. In ion (274.96), which provides direct evidence of the presence of two
addition, the removal of TCS in real wastewater at mg/L levels was carboxyl groups in P2. Notably, loss of C2H2 from the species with
also investigated. Select physico-chemical properties of the test m/z 186.95 generates a fragment ion peak at m/z 160.95, which is
wastewater are presented in Table S1. As illustrated in Fig. S2, no consistent with the parent ion of P6 (C6H3Cl2O, M-H). Furthermore,
obvious difference in the decomposition of TCS was observed for the fragment ion at m/z 124.98 results from cleavage of the CeCl
ultrapure water versus real waste water, implying that the waste- bond in the m/z 160.95 species. Hence, it can be assumed that
water components exerted no influence on the TCS oxidation. oxidation takes place in the hydroxyl-containing benzene ring.
Similar phenomena were also observed in the presence of some At 12.13 min, an ion cluster with m/z 230.96/232.96/234.96 was
typical ions and dissolved organic matter (DOM; see Fig. S1). detected, having a chlorine isotope ratio of 100.0%/63.9%/10.2%,
suggesting the presence of two Cl atoms in the compound P4. As
shown in Fig. 3C, the fragment ions at m/z 202.96, 174.97, 160.95 and
3.2. Exploration of reaction mechanism
124.98 correspond to the loss of CO, CO, CH2 and Cl, respectively.
Thus, P4 was identified as C9H6Cl2O3, the structure of which is
3.2.1. Identification of oxidation products
shown in Fig. 3C.
The products of triclosan oxidation by permanganate were
The presence of two Cl atoms in the molecules of P1, P3, P8, and
detected by LC-Q-TOF-MS analysis. The total ion current (TIC)
P9 was also confirmed by the chlorine isotopic distribution. The
chromatograms of TCS solution before and after oxidation are
fragmentation spectra of these compounds are similar to that of P2
shown in Fig. 2. From the reaction sample at 0 min, we can see that
and P4 (Fig. S3). Likewise, the fragment peaks at m/z 160.95 and
the peak of TCS appears at near 17.2 min. This substrate peak
124.96 implied the occurrence of the 2,4-DCP unit in the structures.
gradually disappeared with progress of the oxidation reaction, and
J. Chen et al. / Water Research 103 (2016) 215e223 219

Fig. 2. Total ion current (TIC) chromatograms of the oxidized samples at different reaction time.

Fig. 3. Product ion spectrum of TCS (A), P2 (B) and P4 (C) obtained by HPLC-TOF-MS/MS and proposed fragmentation pathway.

The most probable molecular formula of P1, P3, P8, and P9 are triclosan. A large amount of the three isomers was detected in the
proposed as C12H5Cl3O3, C11H8Cl2O4, C12H6Cl2O7 and C7H4Cl2O3, present study, while the product P5 was produced in a low con-
respectively. centration. Based on the fragment ions at m/z 284.93, 248.95,
P5eP7, P10, P11: Compound P6 was eluted at 14.43 min. An ion 194.96, 184.98, 160.95 and 124.98, we speculate that P7 (m/z
isotopic cluster was observed at m/z 160.96/162.96/164.96 in the 320.92) may be a 2,4-dichlorophenol adduct, plausibly derived
spectrum, suggesting that two chlorines are present in the product. from the coupling of two 2,4-dichlorophenol molecules (Fig. S3).
The fragments at m/z 124.98, 89.00, and 61.00 confirm the occur- Similarly, P11 is assumed to be generated from the coupling of 2,4-
rence of Cl, Cl, and CO, respectively (Fig. S3). The product was dichlorophenol and TCS.
identified as 2,4-dichlorophenol (2,4-DCP), which has two isomers
(2,6-DCP, 3,5-DCP) with chromatographic retention times of 12.13
3.2.2. Elucidation of oxidation pathways
and 15.00, respectively. 2,4-DCP has been detected as a trans-
Many studies have suggested different oxidation pathways for
formation product of triclosan in many studies involving oxidative
triclosan based on various treatment techniques. During the free-
treatment via manganese oxides (Zhang and Huang, 2003), ferrate,
chlorine-mediated oxidation process, triclosan can undergo ether
and ozonation (Chen et al., 2012; Yang et al., 2011), and also in
cleavage resulting in the production of 2,4-dichlorophenol, or can
aqueous photochemical experiments (Latch et al., 2005). Kim et al.
be chlorinated to produce one of two chlorophenoxyphenol in-
(2011) also observed this compound in the biodegradation of
termediates: 5,6-dichloro-2-(2,4-dichlorophenoxy)phenol and 4,5-
220 J. Chen et al. / Water Research 103 (2016) 215e223

dichloro-2-(2,4-dichlorophenoxy)phenol. Chen et al. (2012) re- 3.2.3. Studies of frontier electron densities (FEDs)
ported that the hydroxylation reaction took place in the less chlo- Frontier electron densities theory has been widely used for
rinated benzene ring when TCS was subjected to ozonation. Yang evaluating the oxidative degradation behaviors of organic sub-
et al. (2011) proposed some reaction pathways for the oxidation stances (Dai et al., 2015; Li et al., 2016). According to the frontier
of triclosan by Fe(VI), involving scission of the ester bond, bond- orbital theory, atoms with a larger 2FED2HOMO are probable attack
breaking of the ether linkage, coupling reaction, phenoxy radical points for electron extraction. In addition, the oxidation reaction is
addition reaction, and phenol ring opening reaction. Based on these more likely to occur at the site with more negative point charges
previous research findings and the identified degradation products, (Han et al., 2013).
two main reaction pathways are proposed herein for the degra- To support our conclusion that the degradation of triclosan by
dation of TCS by permanganate (Fig. 4). permanganate is initiated by ether bond oxidation and the ring
KMnO4 is inclined to attack the hydroxyl and the para-position opening reaction, molecular orbital calculations were carried out
in the less chlorinated phenol moiety, yielding 2-chloro-5-(2,4- for the triclosan molecule using the Gaussian 09 program. The
dichlodichlorophenoxy)-[1,4] benzoquinone (P1) via two-electron Mulliken atomic charges, the 2FED2HOMO values, and the isosurface
transfer. Meanwhile, some ring-opening products are generated, of the highest occupied molecular orbital (HOMO) are displayed in
including 2-(2,4-dichlorophenoxy)maleic acid (P2), (E)-2-(2,4- Fig. 5a5d. As shown in Fig. 5a, the oxygen in the hydroxyl (O(16))
dichlorophenoxy)-4-formylbut-2-enoic acid (P3) and 2-(2,4- group had the most negative Mulliken atomic charge, followed by
dichlorophenoxy)malonaldehyde (P4). Subsequently, (E)-3-(2,4- the oxygen in the ether bond. Thus, the two oxygen atoms would be
dichlorophenoxy)-2,5-dioxohex-3-enedioic acid (P8) was gener- easily oxidized. The highest 2FED2HOMO value in the TCS molecule
ated as the ring opening product of P1, while 2,4-dichlorophenyl was also found to be localized at the oxygen in the hydroxyl, second
hydrogen carbonate (P9) was produced from P2, P3, and P4 to which is the C(8) atom, followed by C(10), C(12), and O(7)
through the decarbonylation and decarboxylation reaction. (Fig. 5b). These results indicated that the oxygen in the hydroxyl
Another possible reactive site is the ether bond, the cleavage of group was the most vulnerable site for oxidation, and that cleavage
which results in the formation of 2,4-dichlorophenol (P6) and 3- of the ether bond (C(8)eO(7)) was reasonable. Moreover, the
chlorophenol (P5). This reaction pathway was also observed in C(10) and C(12) atoms are also likely to be oxidized. Consistent with
the oxidation of TCS by free chlorine, ferrate, O3 or MnO2 (Chen the 2FED2HOMO calculation results, the HOMO isosurface in Fig. 5c
et al., 2012; Rule et al., 2005; Yang et al., 2011; Zhang and Huang, and d also shows that the orbital is mainly distributed over the
2003). Thus, 2,4-dichlorophenol can be further transformed to 2- oxygen in the hydroxyl, then at the C(8), C(10), C(12), and O(7)
(2,4-dichlorophenoxy)-4,6-dichlorophenol (P7) and 2,6-bis(2,4- atoms. These calculations collectively suggest that oxygen in the
dichlorophenoxy)-3-chlorophenol (P11) by coupling with residual hydroxyl group was the most possible site of the oxidation reaction,
triclosan or another 2,4-dichlorophenol. Furthermore, 2- but positions around the C(8), C(10), C(12), and O(7) atoms are also
chloromaleic acid (P10) was produced from P6 by ring opening. susceptible. Thus, the formation of the intermediate P6 (m/
In brief, the reaction pathways for TCS oxidation by perman- z ¼ 160.95) shown in Fig. 4 can be ascribed to cleavage of the ether
ganate mainly include cleavage of the ether bond (C(8)eO(7)) bond (Pathway I). Moreover, the benzoquinone product P4 (m/
and benzene ring-opening (in the less chlorinated benzene ring). z ¼ 300.93) and the benzene ring opening products P2 (m/
z ¼ 274.96) and P3 (m/z ¼ 272.98) were generated by attack on the

Pathway II Cl
+ P6 O Cl
coulping reaction

P5 P6 P7 Cl
126.99/128.99 160.95/162.95 320.92/322.91/324.91/326.91/328.90
Cl OH /164.95
Cl Cl
Cl Cl
Cl C O O Cl

Pathway I P10 OH P11 Cl

148.95/150.95 446.90/448.90/450.90/452.89/454.89/456.89

Cl OH Cl OH Cl
Cl O H
Cl C Cl C Cl O
Cl Cl P3 H2
P1 P2 OH P4
300.93/302.93/304.93/306.93 274.96/276.96/278.96 272.98/274.98/276.98 230.96/232.96/234.96

Cl O Cl
Cl Cl
O P9
330.94/332.94/334.93 204.95/206.95/208.94

Fig. 4. Proposed degradation pathways of triclosan by KMnO4 oxidation.

J. Chen et al. / Water Research 103 (2016) 215e223 221

Fig. 5. Mulliken atomic charges, computed FED and highest occupied molecular orbital (HOMO) shown as isosurfaces of TCS calculated by Gaussian 09 program at the B3LYP/6-
311G** level: (a) Mulliken atomic charges of TCS, (b) 2FED2HOMO data of each atom, (c) isodensity surfaces of HOMO with an isovalue of 0.05, (d) isodensity surfaces of HOMO with
an isovalue of 0.08. The arrows imply positions for electron extraction.

O(16), C(10), and C(16) atoms (Pathway II). the toxicity trend was similar to that of D. magna, exhibiting a rapid
decrease at the initial reaction stage followed by a slight change.
3.3. Toxicity studies The initial TCS solution exerted high toxicity with 100% lumines-
cence inhibition even at 2-fold dilution, and an inhibition ratio of
To assess the potential risks of the intermediates on aquatic 53.4% at 8-fold dilution. After 10 s, the inhibition rate was 99.3%,
ecosystems, changes in the toxicity of the reaction solutions were 70.9%, 43.3%, and 20.9% for initial sample, 2-fold, 4-fold, and 8-fold
evaluated by using two aquatic model test species P. phosphoreum dilution samples, respectively. As the reaction time increased to
and D. magna. Since P. phosphoreum was completely inhibited 3 min, the inhibition ratio decreased to 57.1%, 13.9%, 9.0% and 1.0%,
within the first 1 min, 2-fold, 4-fold and 8-fold dilution were respectively, suggesting that at a low concentration, the in-
applied for the reaction samples to evaluate the true toxicity vari- termediates do not reduce the activity of P. phosphoreum. Notably,
ation trend. the toxicity of TCS (initial concentration ¼ 20 mg/L) to D. magna and
As shown in Fig. 6, the toxicity of the reaction solution to P. phosphoreum after 60 min was reduced by 95.2% and 43.0%,
D. magna followed a generally decreasing trend. One hundred respectively. In addition, the reaction solution has no inhibitory
percent immobilization was observed for the raw solution, while effect on D. magna at an initial TCS concentration of 5000 ng/L or
the immobilization ratio decreased to 33.3%, 14.3% and 4.8% after 0.5 mg/L. For P. phosphoreum, TCS reaction solution at an initial
10 s, 3 min and 60 min of reaction, respectively. For P. phosphoreum, concentration of 2 mg/L showed no toxicity after 60 min.

120 100
A P. phosphoreum B D. magna (24h)
Initial TCS concentration 20 mg/L Initial TCS concentration 20 mg/L
100 P. phosphoreum1/2 D. magna (24h)
P. phosphoreum1/4 80 Initial TCS concentration 0.5 mg/L
Inhibition Rate (%)

80 P. phosphoreum1/8 D. magna (24h)

Inhibition Rate (%)

P. phosphoreum Initial TCS concentration 5000 ng/L

Initial TCS concentration 2 mg/L 60

40 40

20 20

0 0.17 0.5 1 1.5 2 3 5 10 30 60 0 0.17 0.5 1 1.5 2 3 5 10 30 60
Processing Time (min) Processing Time (min)

Fig. 6. Inhibition rate of reaction solutions to P. phosphoreum and D. magna during TCS oxidation by KMnO4. Experimental conditions: pH ¼ 8.0, [TCS]0:[KMnO4]0 ¼ 1:2.5.
222 J. Chen et al. / Water Research 103 (2016) 215e223

It is reported that the aquatic toxicity of chlorinated compounds benzoquinone produced THMs easier than phenol during chlori-
depends on the number of aromatic rings and chlorine groups (Del nation and chloramination processes, while a contrary tendency
Castillo et al., 2012). Levy et al. (1999) further pointed out that the occurred in the formation of HAAs. In addition, compared with
antibacterial activity of TCS was mainly correlated with the chlorination, chloramination decreased the formation of THMs and
hydroxyl-containing benzene ring. As discussed above, degradation HAAs greatly; the concentrations of the THMs (phenol and 1,4-
of TCS by KMnO4 includes cleavage of the ether bond and benzene benzoquinone) formed during chloramination were only 7.9 and
ring-opening. This oxidative transformation of triclosan would 26.75% of those formed during chlorination, and the concentrations
reduce the toxicity of the reaction solutions, leading to a relatively of HAAs were 0.003 and 0.18% of those formed during chlorination.
low ecological risk. Thus, KMnO4 is an environmentally friendly This is consistent with previous results (Cowman and Singer, 1995;
oxidant that can be widely used in water and wastewater Lu et al., 2009).

4. Conclusions
3.4. Chlorination and chloramination
Potassium permanganate treatment is demonstrated to be
In full-scale treatment plants, permanganate treatment is al- effective for the removal of TCS in water. Under the optimum re-
ways used in combination with chlorination and chloramination action conditions of pH ¼ 8.0, [TCS]0:[KMnO4]0 ¼ 1:2.5 and
(Rodríguez et al., 2007; Chen and Yeh, 2005; Guan et al., 2010). T ¼ 25  C, TCS at a concentration of 20 mg/L can be completely
Phenol and 1,4-benzoquinone are the two major kinds of degra- degraded in 120 s. However, a high KMnO4 dosage is required for
dation products formed during oxidation of TCS by permanganate. the efficient removal of TCS at mg/L levels in water and wastewater
Thus, chlorination and chloramination is performed with phenol treatment.
and benzoquinone to detect the possible formation of disinfection- Mass spectrometry analysis reveals the formation of 11 prod-
by-products (DBPs). ucts, including phenol, derivatives of phenol, benzoquinone,
In the present study, phenol and 1,4-benzoquinone were organic acids and aldehydes. Six products with precursor ion at m/z
completely removed after chlorination and chloramination (anal- 330.94, 274.96, 272.98, 230.96, 204.95 and 148.95 were reported for
ysis method in Text S2), while the THM and HAA species detected the first time. Two main reaction pathways involving CeO bond
included four species of THMs and monochloroacetic acid. Thus, cleavage and benzene ring opening were proposed, which was
safety evaluation is necessary when phenol and benzoquinone are further confirmed by frontier electron density calculations and
present in source water as micropollutants, since they can act as point charge analysis. Toxicity tests with P. phosphoreum and
precursors to generate THMs and HAAs. The specific yields of THMs D. magna reveal that the treatment by permanganate reduces the
and HAAs (expressed as mg/mg C) are shown in Fig. 7 AeD (see toxicity of solutions containing triclosan, leading to relatively low
Table S2 for specific values). Results indicate that chlorination and ecological risk. Chlorination and chloramination experiments
chloramination of phenol and 1,4-benzoquinone would yield low indicate that chlorination and chloramination of phenol and 1,4-
concentrations of DBPs (STHMFP, 20.99e278.97 mg/mg; SHAAFP, benzoquinone would yield low concentrations of DBPs (STHMFP,
7.86  10445.77 mg/mg). As the precursors of DBPs, 1,4- 20.99e278.97 mg/mg; SHAAFP, 7.86  10445.77 mg/mg). Based on

Fig. 7. The formation potential of THMs and HAAs in subsequent treatment of chlorination and chloramination.
J. Chen et al. / Water Research 103 (2016) 215e223 223

these findings, we suggest that KMnO4 oxidation can be used as an 2011. Triclosan susceptibility and co-metabolismea comparison for three aer-
obic pollutant-degrading bacteria. Bioresour. Technol. 102 (3), 2206e2212.
environmentally friendly and cost-effective technique for TCS
Kim, D., Han, J., Choi, Y., 2013. On-line solid-phase microextraction of triclosan,
removal in water and wastewater treatment. bisphenol A, chlorophenols, and selected pharmaceuticals in environmental
water samples by high-performance liquid chromatographyeultraviolet
Acknowledgements detection. Anal. Bioanal. Chem. 405 (1), 377e387.
Latch, D.E., Packer, J.L., Stender, B.L., VanOverbeke, J., Arnold, W.A., McNeill, K., 2005.
Aqueous photochemistry of triclosan: formation of 2, 4-dichlorophenol, 2, 8-
This research was financially supported by the National Natural dichlorodibenzo-p-dioxin, and oligomerization products. Environ. Toxicol.
Science Foundation of China (No. 21577063, 21377051) and the Chem. 24 (3), 517e525.
Levy, C.W., Roujeinikova, A., Sedelnikova, S., Baker, P.J., Stuitje, A.R., Slabas, A.R.,
Major Science and Technology Program for Water Pollution Control Rice, D.W., Rafferty, J.B., 1999. Molecular basis of triclosan activity. Nature 398
and Treatment of China (No. 2012ZX07506-001). (6726), 383e384.
Li, Y., Yang, S., Sun, C., Wang, L., Wang, Q., 2016. Aqueous photofate of crystal violet
under simulated and natural solar irradiation: kinetics, products, and pathways.
Appendix A. Supplementary data Water Res. 88 (3), 173e183.
Liu, C., Qiang, Z., Adams, C., Tian, F., Zhang, T., 2009. Kinetics and mechanism for
Supplementary data related to this article can be found at http:// degradation of dichlorvos by permanganate in drinking water treatment. Water
Res. 43 (14), 3435e3442.
dx.doi.org/10.1016/j.watres.2016.07.041. Loos, R., Locoro, G., Contini, S., 2010. Occurrence of polar organic contaminants in
the dissolved water phase of the Danube River and its major tributaries using
References SPE-LC-MS 2 analysis. Water Res. 44 (7), 2325e2335.
Lu, J.F., Zhang, T., Ma, J., Chen, Z.L., 2009. Evaluation of disinfection by-products
formation during chlorination and chloramination of dissolved natural
Abdelraheem, W.H., He, X., Duan, X., Dionysiou, D.D., 2015. Degradation and
organic matter fractions isolated from a filtered river water. J. Hazard. Mater
mineralization of organic UV absorber compound 2-phenylbenzimidazole-5-
162 (1), 140e145.
sulfonic acid (PBSA) using UV-254 nm/H2O2. J. Hazard. Mater 282, 233e240.
Novo, A., Andre , S., Viana, P., Nunes, O.C., Manaia, C.M., 2013. Antibiotic resistance,
Bedoux, G., Roig, B., Thomas, O., Dupont, V., Le Bot, B., 2012. Occurrence and toxicity
antimicrobial residues and bacterial community composition in urban waste-
of antimicrobial triclosan and by-products in the environment. Environ. Sci.
water. Water Res. 47 (5), 1875e1887.
Pollut. Res. 19 (4), 1044e1065.
Pintado-Herrera, M.G., Gonza lez-Mazo, E., Lara-Martín, P.A., 2014. Determining the
Brausch, J.M., Rand, G.M., 2011. A review of personal care products in the aquatic
distribution of triclosan and methyl triclosan in estuarine settings. Chemo-
environment: environmental concentrations and toxicity. Chemosphere 82
sphere 95, 478e485.
(11), 1518e1532.
Qiao, X., Zheng, X., Xie, Q., Yang, X., Xiao, J., Xue, W., Chen, J., 2014. Faster photo-
Buth, J.M., Grandbois, M., Vikesland, P.J., McNeill, K., Arnold, W.A., 2009. Aquatic
degradation rate and higher dioxin yield of triclosan induced by cationic sur-
photochemistry of chlorinated triclosan derivatives: potential source of poly-
factant CTAB. J. Hazard. Mater 275, 210e214.
chlorodibenzo-p-dioxins. Environ. Toxicol. Chem. 28 (12), 2555e2563.
Reiss, R., Mackay, N., Habig, C., Griffin, J., 2002. An ecological risk assessment for
Chen, J.J., Yeh, H.H., 2005. The mechanisms of potassium permanganate on algae
triclosan in lotic systems following discharge from wastewater treatment plants
removal. Water Res. 39 (18), 4420e4428.
in the United States. Environ. Toxicol. Chem. 21 (11), 2483e2492.
Chen, W.H., Young, T.M., 2008. NDMA formation during chlorination and chlor-
Rodríguez, E., Onstad, G.D., Kull, T.P., Metcalf, J.S., Acero, J.L., von Gunten, U., 2007.
amination of aqueous diuron solutions. Environ. Sci. Technol. 42 (4), 1072e1077.
Oxidative elimination of cyanotoxins: comparison of ozone, chlorine, chlorine
Chen, X., Richard, J., Liu, Y., Dopp, E., Tuerk, J., Bester, K., 2012. Ozonation products of
dioxide and permanganate. Water Res. 41 (15), 3381e3393.
triclosan in advanced wastewater treatment. Water Res. 46 (7), 2247e2256.
Rule, K.L., Ebbett, V.R., Vikesland, P.J., 2005. Formation of chloroform and chlori-
Chen, Z.F., Ying, G.G., Liu, Y.S., Zhang, Q.Q., Zhao, J.L., Liu, S.S., Pan, C.G., 2014. Tri-
nated organics by free-chlorine-mediated oxidation of triclosan. Environ. Sci.
closan as a surrogate for household biocides: an investigation into biocides in
Technol. 39 (9), 3176e3185.
aquatic environments of a highly urbanized region. Water Res. 58, 269e279.
Singer, H., Müller, S., Tixier, C., Pillonel, L., 2002. Triclosan: occurrence and fate of a
Cowman, G.A., Singer, P.C., 1995. Effect of bromide ion on haloacetic acid speciation
widely used biocide in the aquatic environment: field measurements in
resulting from chlorination and chloramination of aquatic humic substances.
wastewater treatment plants, surface waters, and lake sediments. Environ. Sci.
Environ. Sci. Technol. 30 (1), 16e24.
Technol. 36 (23), 4998e5004.
Dai, Y., Song, Y., Wang, S., Yuan, Y., 2015. Treatment of halogenated phenolic com-
Sires, I., Oturan, N., Oturan, M.A., Rodríguez, R.M., Garrido, J.A., Brillas, E., 2007.
pounds by sequential tri-metal reduction and laccase-catalytic oxidation. Water
Electro-Fenton degradation of antimicrobials triclosan and triclocarban. Elec-
Res. 71, 64e73.
trochim. Acta 52 (17), 5493e5503.
Dann, A.B., Hontela, A., 2011. Triclosan: environmental exposure, toxicity and
Suarez, S., Dodd, M.C., Omil, F., Gunten, U.V., 2007. Kinetics of triclosan oxidation by
mechanisms of action. J. Appl. Toxicol. 31 (4), 285e311.
aqueous ozone and consequent loss of antibacterial activity: relevance to
Dash, S., Patel, S., Mishra, B.K., 2009. Oxidation by permanganate: synthetic and
municipal wastewater ozonation. Water Res. 41 (12), 2481e2490.
mechanistic aspects. Tetrahedron 65 (4), 707e739.
Toms, L.M.L., Allmyr, M., Mueller, J.F., Adolfsson-Erici, M., McLachlan, M., Murby, J.,
Dayan, A.D., 2007. Risk assessment of triclosan [Irgasan®] in human breast milk.
Harden, F.A., 2011. Triclosan in individual human milk samples from Australia.
Food Chem. Toxicol. 45 (1), 125e129.
ndez, P., Lafuente, A., Rodríguez-Llorente, I.D., Caviedes, M.A., Chemosphere 85 (11), 1682e1686.
Del Castillo, I., Herna 
Wacławek, S., Antos, V., Hraba k, P., Cerník, M., Elliott, D., 2016. Remediation of
Pajuelo, E., 2012. Self-bioremediation of cork-processing wastewaters by
hexachlorocyclohexanes by electrochemically activated persulfates. Environ.
(chloro) phenol-degrading bacteria immobilised onto residual cork particles.
Sci. Pollut. Res. 23 (1), 765e773.
Water Res. 46 (6), 1723e1734.
Waldemer, R.H., Tratnyek, P.G., 2006. Kinetics of contaminant degradation by per-
Fu, Q., Sanganyado, E., Ye, Q., Gan, J., 2016. Meta-analysis of biosolid effects on
manganate. Environ. Sci. Technol. 40 (3), 1055e1061.
persistence of triclosan and triclocarban in soil. Environ. Pollut. 210, 137e144.
Wu, Q., Shi, H., Adams, C.D., Timmons, T., Ma, Y., 2012. Oxidative removal of selected
Guan, X.H., He, D., Ma, J., Chen, G.H., 2010. Application of permanganate in the
endocrine-disruptors and pharmaceuticals in drinking water treatment sys-
oxidation of micropollutants: a mini review. Front. Environ. Sci. Eng. 4 (4),
tems, and identification of degradation products of triclosan. Sci. Total Environ.
439, 18e25.
Halden, R.U., Paull, D.H., 2005. Co-occurrence of triclocarban and triclosan in US
Yan, Y.E., Schwartz, F.W., 1999. Oxidative degradation and kinetics of chlorinated
water resources. Environ. Sci. Technol. 39 (6), 1420e1426.
ethylenes by potassium permanganate. J. Contam. Hydrol. 37 (3), 343e365.
Han, J., Meng, S., Dong, Y., Hu, J., Gao, W., 2013. Capturing hormones and bisphenol
Yang, B., Ying, G.G., Zhao, J.L., Zhang, L.J., Fang, Y.X., Nghiem, L.D., 2011. Oxidation of
A from water via sustained hydrogen bond driven sorption in polyamide
triclosan by ferrate: reaction kinetics, products identification and toxicity
microfiltration membranes. Water Res. 47 (1), 197e208.
evaluation. J. Hazard. Mater. 186 (1), 227e235.
Ji, Y., Zhou, L., Zhang, Y., Ferronato, C., Brigante, M., Mailhot, G., Chovelon, J.M., 2013.
Young, T.A., Heidler, J., Matos-Pe rez, C.R., Sapkota, A., Toler, T., Gibson, K.E.,
Photochemical degradation of sunscreen agent 2-phenylbenzimidazole-5-
Schwab, K.J., Halden, R.U., 2008. Ab initio and in situ comparison of caffeine,
sulfonic acid in different water matrices. Water Res. 47 (15), 5865e5875.
triclosan, and triclocarban as indicators of sewage-derived microbes in surface
Jiang, J., Pang, S.Y., Ma, J., 2009. Oxidation of triclosan by permanganate (Mn (VII)):
waters. Environ. Sci. Technol. 42 (9), 3335e3340.
importance of ligands and in situ formed manganese oxides. Environ. Sci.
Zhang, H., Huang, C.H., 2003. Oxidative transformation of triclosan and chlorophene
Technol. 43 (21), 8326e8331.
by manganese oxides. Environ. Sci. Technol. 37 (11), 2421e2430.
Kim, Y.M., Murugesan, K., Schmidt, S., Bokare, V., Jeon, J.R., Kim, E.J., Chang, Y.S.,