Kinetics of Chlorination of Benzophenone-3 in the Presence of

Nov 20, 2015 - *Phone: +33 5 49 45 39 24; fax: +33 5 49 45 37 68; e-mail: [email protected]. ... (2) There are two types of UV filters: (...
0 downloads 0 Views 615KB Size
Article pubs.acs.org/est

Kinetics of Chlorination of Benzophenone‑3 in the Presence of Bromide and Ammonia Pamela Abdallah,† Marie Deborde,*,†,‡ Florence Dossier Berne,† and Nathalie Karpel Vel Leitner† †

Université de Poitiers, Institut de Chimie des Milieux et des Matériaux de Poitiers (IC2MP UMR 7285), Equipe Eaux Géochimie organique Santé (EGS), ENSIP, 1 rue Marcel Doré, Bâtiment B1 TSA 41105, 86073 Poitiers Cedex 9, France ‡ Université de Poitiers, UFR de Médecine et de Pharmacie, 6 rue de la Milétrie, Bâtiment D1, TSA 51115, 86073 Poitiers Cedex 9, France S Supporting Information *

ABSTRACT: The aim of this study was to assess the impact of chlorination on the degradation of one of the most commonly used UV filters (benzophenone-3 (BP-3)) and the effects of bromide and ammonia on the kinetics of BP-3 elimination. Bromide and ammonia are rapidly converted to bromine and chloramines during chlorination. At first, the rate constants of chlorine, bromine and monochloramine with BP-3 were determined at various pH levels. BP-3 was found to react rapidly with chlorine and bromine, with values of apparent second order rate constants equal to 1.25(±0.14) × 103 M−1·s−1 and 4.04(±0.54) × 106 M−1·s−1 at pH 8.5 for kChlorine/BP‑3 and kBromine/BP‑3, respectively, whereas low monochloramine reactivity was observed (kNH2Cl/BP−3 = 0.112 M−1·s−1). To assess the impact of the inorganic content of water on BP-3 degradation, chlorination experiments with different added concentrations of bromide and/or ammonia were conducted. Under these conditions, BP-3 degradation was found to be enhanced in the presence of bromide due to the formation of bromine, whereas it was inhibited in the presence of ammonia. However, the results obtained were pH dependent. Finally, a kinetic model considering 18 reactions was developed using Copasi to estimate BP-3 degradation during chlorination in the presence of bromide and ammonia.



INTRODUCTION The recent public health concern as to the risks related to UV exposure has led to an impressive increase in UV filter production and use in the past decade. Nowadays, UV filters are not only found in sunscreens but are also used in a diverse range of personal care products, such as cosmetics, shampoo, hair spraym, and make up. These compounds are used in combinations and the total amounts can exceed 10% by mass.1 Moreover, UV filters are also used as stabilizers in plastics, paints, textiles and other materials to prevent photodegradation of polymers and pigments.2 There are two types of UV filters: (i) inorganic UV filters such as TiO2 and ZnO; and (ii) the organic UV filters considered in the present study. In sunscreens and personal care products, these UV filters can be used separately or in combination. About 55 UV filters are approved for use in sunscreens worldwide. However, only 10 including benzophenone-3 are uniformly approved.3 Traces of UV filters are found in rivers, seas and lakes, as well as swimming pools, with concentrations ranging up to 0.3 mg/ L, and the highest concentrations are observed in warm seasons.4 Organic UV filters can enter the aquatic environment directly via the direct input of UV filters in lakes, rivers or seawater following rub off during recreational activities (e.g., swimming) or indirectly via effluents from wastewater treatment plants (WWTP). The indirect pathways include renal © XXXX American Chemical Society

excretion after percutaneous uptake and removal of sunscreen residues by rub off or during showering.5 In the environment, these compounds may be subject to transformation, thus contributing to their elimination through biodegradation, chemical, and photochemical degradation.6 Due to their continuous release into the environment along with other personal care products and considering their synergic effects through combined parallel action, even compounds of low persistence might cause unwanted effects in the environment.7,8 Therefore, UV filters are currently considered as emerging contaminants. Several toxicity studies have been conducted concerning the effects of organic UV filters on aquatic organisms as well as on human health, particularly their endocrine disrupting activity.9,10 In a recent study, the presence of six UV filters in human placental tissue was investigated and estrogenic and antiandrogenic activity was confirmed at 1−50 μM concentrations for some compounds through in vitro tests.10 Studies have also shown that prenatal exposure to the UV filter BP-3 was associated with shortened gestational duration.11 Received: July 22, 2015 Revised: November 12, 2015 Accepted: November 20, 2015

A

DOI: 10.1021/acs.est.5b03559 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology Since traces of UV filters were detected in natural waters, and due to their effects on human health as well as on the aquatic environment, a study on the fate of these compounds during water treatment is recommended. According to published data, several UV filters have been found in raw wastewaters in different countries with concentrations ranging from 1.5 to 19 μg/L,12−14 whereas lower concentrations from 0.01 to 2.7 μg/L were observed in treated wastewaters,12,14 indicating the elimination of these compounds in WWTP. The elimination of UV filters during wastewater treatment depends on the physical and chemical properties of each compound and on the treatments considered. Hence, different percentages of elimination of UV filters in WWTP have been reported (68− 96% for benzophenone-3, 18−82% for 4-methylbenzylidene camphor, 97−99% for ethylhexylmethoxycinnamate, and 88− 99% for octocrylene).12 Similarly, the effects of different water treatments used for drinking water production on the elimination of UV filters have been investigated: coagulation/ flocculation, continuous microfiltration, and ozonation. The results revealed maximum removal with ozonation followed by coagulation/flocculation and last microfiltration.13 The aim of this study was to investigate the effects of chlorine on the elimination of one organic UV filter, that is, benzophenone-3 (BP-3) during water disinfection steps. The U.S. Food and Drug Administration (FDA) approved the use of BP-3 (also known as Eusolex 4360, oxybenzone or 2hydroxy-4-methoxyphenone) in sunscreen and other cosmetic preparations and was evaluated as being safe and effective (Category I) at concentrations of 2−6% in over-the-counterproduct use.15,16 BP-3 is the most common UV filter found in swimming pools and surface waters, with concentrations of up to 44 μg/L.4 Chlorine is the most commonly used oxidant for swimming pool disinfection as well as drinking water disinfection. During swimming pool disinfection, chlorination is frequently practiced with the use of organic amines such as cyanuric acid. These organic amines, used as stabilizing agents, protect free chlorine from photodegradation. BP-3 reactivity with chlorine and the possible transformation products have to be examined since BP-3 is found in chlorinated swimming pool water and surface waters used for the production of drinking water. Moreover, natural waters contain a wide variety of ions, such as bromide, ammonia, etc., in addition to dissolved organic matter (DOM) which contributes to chlorine consumption during disinfection. These ions, such as ammonia can quench free available chlorine to produce chloramines, known for being less reactive oxidants with organic compounds.19 Moreover, in the presence of bromide ions reactions with chlorine yielding bromine occur. Bromine is usually more reactive than chlorine with phenolic compounds but is also more rapidly consumed by DOM than chlorine.17 Several studies have shown that the presence of bromide during chlorination significantly increases the yield and rate of formation of THMs.18 In this work, we focused our study BP-3 removal and the impact of bromide and ammonia during chlorination steps used for drinking water production. To attempt this goal, we studied the chlorination, bromination and chloramination of BP-3 at different pH levels. The chlorination of BP-3 under several bromide and ammonia concentrations and pH conditions was then performed and a kinetic model using Copasi was developed. Finally the kinetic model was compared to the experimental data obtained in this work. Considering BP-3, chlorine, bromide, and ammonia concentrations usually found during water treatment production, bromide and ammonia

impact on BP-3 removal during chlorination steps was discussed.



MATERIALS AND METHODS Chemical Reagents. Benzophenone-3 (98% purity) and commercial solution of sodium hypochlorite (13% of active chlorine) were purchased from Alfa Aesar and Acros Organics, respectively. 17α-ethynylestradiol (98% purity) was obtained from Sigma-Aldrich. Phosphate buffers, sodium thiosulfate, and colorimetric agents were supplied by Sigma-Aldrich or Acros Organics and were all of analytical grade. The BP-3 stock solution used for chromatography standards was prepared in 100% methanol due to its low solubility in water. All other stock solutions were prepared using pure water (18.2 MΩ·cm and DOC < 0.1 mg/L) obtained from a Milli-Q water purification system. For chlorination experiments, BP-3 was dissolved in water by stirring for 24 h and the solution was filtered with a 0.45 μm filter before use. Under these conditions, the observed solubility for BP-3 in water was about 35 μM. Bromine stock solutions were obtained by addition of 0.55 mM of bromide to a 0.5 mM solution of ozone at pH 4 (10 mM phosphate) and T = 5 °C. These conditions, (i.e., [O3]/[Br−] of 0.9 and acidic pH conditions) were chosen in order to prevent bromate formation. Monochloramine stock solutions were prepared daily by slowly adding free chlorine to an ammonium chloride solution in a well-stirred reactor at pH 8.5 (nitrogen-to-chlorine ratio = 2 mol/mol). Kinetics Experiments. Kinetics experiments were performed at room temperature (20 ± 2 °C), in aqueous solutions for pH values between 5 and 11 in the presence of phosphate salts (ionic strength 2 × 10−3 M). Under these conditions, no pH variations higher than 0.1 unit was observed. All kinetics experiments were conducted in duplicate. Chlorination and chloramination experiments were conducted under pseudo-first-order kinetic conditions, in the presence of an excess of free available chlorine or monochloramine ([Oxidant]T,0 ≥ 10 [BP-3]T,0) by measuring the decrease in BP-3 concentration by HPLC-UV. Under these conditions, variations of oxidant concentrations were found to be less than 5% over the reaction time. Chlorination experiments were performed in 1 L batch reactors and kinetic runs were started by adding a chlorine stock solution ([Chlorine]T,0 > 15 μM (1.065 mg/L)) to a solution containing 1.5 μM (0.342 mg/L) of BP-3. At different time intervals (0−60 min), 1.5 mL of solution was withdrawn and transferred into vials containing predosed volumes of sodium thiosulfate (8 mM) to quench the residual chlorine and stop the reaction. Chloramination experiments were conducted in 1 L batch reactors containing stock solution of monochloramine ([NH2Cl] = 1.7 mM). Kinetic runs were started by adding BP-3 stock solution ([BP3]T,0 = 1 μM (0.228 mg/L)). Samples were collected at different time intervals (0 to 300 min) in vials containing predosed volumes of sodium thiosulfate (8 mM) to quench residual chloramine. For both oxidants, similar experiments were conducted using L-ascorbic acid (16 mM) instead of sodium thiosulfate to validate the use of both reducing agents without interference with BP-3 detection by liquid chromatography-UV. Since chlorine reacts with Br−, yielding bromine,20 experiments were also conducted with bromine in order to determine the rate constants of bromine with BP-3. The competitive kinetic method was chosen since bromine was found to be much more reactive than chlorine. Different phenolic B

DOI: 10.1021/acs.est.5b03559 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology compounds (e.g., phenol, chlorophenol, and 17α-ethynylestradiol (EE2)) were tested as reference compounds. Among these compounds only EE2 showed similar reactivity with bromine to that of BP-3 over the studied pH range (i.e., 5−11 pH range). This latter compound was thus selected as the competitor for bromination kinetic experiments. Experiments were conducted by adding different doses of bromine (0−15 μM) in 50 mL batch reactors containing equimolar quantities of BP-3 and EE2 (3 μM). The remaining BP-3 and EE2 concentrations were analyzed by HPLC-UV after 24 h to make sure bromine was fully consumed by BP-3 and EE2. Under these conditions, in order to check that no interference due to transformation product formation is observed during HPLC-UV analysis, two additional kinetic experiments were conducted using LC-MS analysis in which the exact mass of BP-3 and EE2 were targeted. For these two experiments, similar results to those obtained using HPLC-UV analysis were observed. Chlorination Experiments in the Presence of Bromide/Ammonia and Modeling. Chlorination experiments in the presence of bromide and/or ammonia were performed in 1 L batch reactors buffered at pH 7 and 8.5 using phosphate salts. For initial chlorine concentrations of 5 μM, the bromide and ammonia concentrations ranged from 0 to 5 μM. The initial BP-3 concentration was 1.5 (±0.2) μM. For these experiments, computer simulations were performed using the Copasi program. The kinetic model was conducted using the LSODA deterministic simulation method after considering all possible reactions taking place in the presence of chlorine as well as bromide ions and ammonia at pH 7 and 8.5. Analytical Methods. Commercial and stock chlorine solutions were standardized by iodometry (APHA, 1995). Bromine stock solution was standardized by measuring the absorbance at 329 nm using a VARIAN Cary 50 UV/vis spectrophotometer (l = 5 cm; ε = 332 M−1cm−1).21 During the chlorination experiments, oxidant concentrations were analyzed by the DPD colorimetric method (APHA, 2000) using an SAFAS spectrophotometer to measure the absorbance at 510 nm (l = 1 cm; ε = 14550 M−1cm−1, LOD = 0.5 μM). The pH was measured using a MeterLab pH-meter equipped with a Radiometer analytical combined electrode. BP-3 and EE2 were quantified using an Alliance Waters 2695 high performance liquid chromatography system (pump and autosampler) with a C18 Kromasil column (250 mm × 4.6 mm, 5 μm), equipped with a UV Visible Waters 2487 detector. The eluent used at a 1 mL/min flow rate was a 0.1% acetic acid methanol/purified water (80/20 or 70/30 for BP-3 and EE2, respectively). The injected volume was 100 μL for each sample and UV detection was set at 290 nm. The limits of detection for BP-3 and EE2 under these conditions were estimated at 0.02 μM.



ln

[BP‐3]T , t [BP‐3]T ,0

= kobst

(3)

where koxidant/BP−3 is the second-order rate constant for a given pH and kobs is the observed pseudo-first-order rate constant with kobs = koxidant/BP−3 × [oxidant]T. koxidant/BP−3 can therefore be determined from the slope of the linear time-course plots of ln([BP-3]T,t/[BP-3]T,0) since the chlorine or chloramine concentration in excess is known and considered stable ([Oxidant]T). The results obtained for the chlorination experiments at different pH levels are shown in Figure S1. The first order reaction relative to chlorine was checked by varying the initial chlorine concentration, as shown in the inset in Figure S1. According to these results, the apparent second order rate constants from 1.05 (±0.28) × 101 to 1.25 (±0.14) × 103 M−1· s−1 for the 5−11 pH range were obtained for BP-3 chlorination. Similar correlations were found in chloramination experiments (R2 = 0.99) (Figure S2). However, only BP-3 chloramination at pH 8.5 was studied due to the low monochloramine reactivity and the probable low BP-3 degradation during water chloramination. An apparent rate constant of 0.112 M−1·s−1 was calculated under these conditions. This result is in the same order of magnitude to those obtained by Cimetière et al. (2009) for phenolic compounds.19 Chlorine reacts with bromide during chlorination, thus producing bromine. Bromine is very reactive with phenolic compounds, with rate constants up to 103-fold higher than with chlorine in drinking water treatment conditions.22 Thus, bromination experiments were conducted in Milli-Q water at pH levels in the 5−11 range and in the presence of EE2 as competitor. This method is commonly applied for the determination of high rate constants (values exceeding 103 M−1·s−123). Under these conditions, the main reactions considered are kOxidant/BP3

BP‐3 + Oxidant ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯→ products

(4)

kOxidant/EE2

EE2 + Oxidant ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯→ products

(5)

Based on second-order reactions, first order relative to the oxidant (active bromine) and first order relative to BP-3 or EE2, the kinetic equations are as follows:

RESULTS AND DISCUSSION Determination of Apparent Rate Constants of Chlorine, Bromine, and Monochloramine with BP-3. All chlorination and chloramination experiments were performed in the presence of a high excess of oxidant (i.e., [Oxidant]T,0 ≥ 10 [BP-3]T,0). Under these conditions, if the BP-3 reaction with the oxidant is a second order reaction (one order relative to each reactant), the following equations apply: d[BP‐3]T = k Oxidant/BP‐3[BP‐3]T [Oxidant]T dt

(2)

and





d[BP‐3]T = kobs[BP‐3]T dt



d[BP‐3]T = k Oxidant/BP‐3[BP‐3]T [Oxidant]T dt

(6)



d[EE2]T = k Oxidant/EE2[EE2]T [Oxidant]T dt

(7)

and k Oxidant/BP − 3 d[EE2]T d[BP‐3]T = [BP‐3]T k Oxidant/EE2 [EE2]T ln

(1) C

[BP‐3]T ,0 [BP‐3]T ,t

=

k Oxidant/BP‐3 k Oxidant/EE2

ln

(8)

[EE2]T ,0 [EE2]T , t

(9)

DOI: 10.1021/acs.est.5b03559 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Figure 1. pH dependence of the apparent second-order rate constants for the reactions of chlorine (a) and bromine (b) with BP-3. Symbols represent measured data and solid lines represent the calculated values of kOxidant/BP‑3 obtained from the model calculations.

Table 1. Elementary Reactions Considered for Chlorination and Bromination of BP-3a no.

a

reactions

chlorination

bromination

eq 10

HOX ⇆ XO− + H+

pKa1

7.5426

8.8922

eq 11

BP‐3 ⇆ BP‐3− + H+

pKa2

9.6b

eq 12

HOX + BP‐3 → products

k1

6.61 (±3.07) M−1·s−1

1.42 (±0.26) × 104 M−1·s−1

eq 13

HOX + BP‐3− → products

k2

1.80 (±0.95) × 105 M−1·s−1

9.60 (±0.87) × 107 M−1·s−1

eq 14

XO− + BP‐3 → products

k3

negligible

negligible

eq 15

XO− + BP‐3− → products

k4

negligible

2.11 (±0.52) × 106 M−1·s−1

(X = Cl or Br). bDetermined in this study by spectrophotometry.

with koxidant/BP−3 and koxidant/EE2 being the apparent second order rate constants for BP-3 and EE2 bromination for a given pH. Linear correlations were obtained for the 5−11 pH range (R2 ≥ 0.98) (Figure S3). Considering the value of koxidant/EE2 (17) , the rate constant koxidant/BP−3 can therefore be calculated. Apparent second order rate constants were calculated for each bromination experiment. Under these conditions, rate constants from 1.49 (±0.03) × 104 to 1.03 (±0.51) × 107 M−1·s−1 were obtained in the 5−11 pH range. pH Dependence Profile and Rate Constant of Elementary Reactions. The variation pattern in the apparent second-order rate constants of chlorine (kChlorine/BP‑3) and bromine (kBromine/BP‑3) versus pH is shown in Figure 1. Similar profiles were obtained for both oxidants with a minimum at pH < 6 and a maximum at pH 8.6 for chlorination and 9.2 for bromination. A kinetic model was developed in order to predict the variation in apparent rate constants with the pH values. The calculated values of kChlorine/BP‑3 and kBromine/BP‑3 were modeled by considering the reactions in Table 1 (where X represents Cl or Br for chlorination or bromination experiments). In this model, only the reactions of HOX and XO− with the neutral and ionized forms of BP-3 were considered. The reactivity of X2O and X2 was assumed to be negligible, since the reaction toward chlorine was a first order reaction and high rate constants were obtained for both chlorine and bromine.24,25 Accordingly, considering that BP-3 reactions with chlorine and bromine are second order reactions:



d[BP‐3] = k Oxidant/BP‐3[BP‐3]T [Oxidant]T dt

(16)

Moreover, based on eqs 12 to 15: −

d[BP‐3] = k1[BP‐3][HOX] + k 2[BP‐3−][HOX] dt + k 3[BP‐3][XO−] + k4[BP‐3−][XO−]

(17)

with [Oxidant]T = [HOX ] + [X O− ]

[BP‐3]T = [BP‐3] + [BP‐3−]

Thus, considering: αHOX and αXO− the protonated and nonprotonated mole fractions of free chlorine or bromine αBP3 and αBP3− the neutral and ionized BP-3 mole fractions BP-3 chlorination or bromination kinetics can be expressed as follows: −

d[BP‐3] = (k1αHOXαBP‐3 + k 2αHOX αBP‐3− + k 3α X O−αBP‐3 dt + k 4α X O−αBP‐3−)[Oxidant]T [BP‐3]T

(18)

Therefore, by combining (eq 17) and (eq 18), the expression of koxidant/BP‑3 can be formulated as D

DOI: 10.1021/acs.est.5b03559 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Figure 2. BP-3 chlorination in the presence of bromide or ammonia at pH 7 ([BP-3]T,0= 1.4 (±0.2) μM, [Chlorine]T,0= 5 (±0.5) μM and T = 20 ± 2 °C) (a,b);[NH4+] = 0 μM and [Br−] = 0−2 μM (c,d)); [Br−] = 0 μM and [NH4+] = 0−5 μM. Symbols represent experimental data and solid lines represent the modeling obtained via Copasi.

0.16 s can be calculated for chlorination and bromination at pH 7, respectively. Concerning chlorination of BP-3, the results obtained in this study can be compared to previously published data.31 The apparent rate constants obtained by Duirk et al. (2013) are of the same order of magnitude as those obtained in this study. However, the elementary rate constant k2 determined by these authors is 1 order of magnitude higher (1.03 × 106 M−1·s−1 versus 1.80 × 105 M−1·s−1 in this work). This could be explained by the higher pKaBP‑3 value considered in this latter study (10.2 versus 9.6 in this study). Thus, for a given pH, the concentration of the ionized form of BP-3 is underestimated compared to our findings. This explains why a comparable apparent rate constant provides a higher elementary rate constant k2 in the study of Duirk et al.31 For both studied oxidants, the highest rate constants obtained for elementary reactions with the ionized form of BP-3 (k2 > k1 for HOX and k4 > k3 for BrO−) could be explained by the higher electron donor character of the phenolate function that favors the electrophilic attack of chlorine or bromine.32 This was confirmed by several authors for various phenolic compounds by comparing the k 2 elementary rate constant to the electron donor character of the substituants represented by the Hammett constants (σo,m,p), which reflect the electronic density on the aromatic ring.22,28,30−33 A comparison between the results obtained in this study for BP-3 and other phenolic compounds considering the classical and the corrected Hammett type correlation suggested by Deborde et von Gunten32 in the case of chlorination is developed in the Supporting Information (Figure S4). In light of these results, a similar reaction mechanism based on

k Oxidant/BP‐3 =

k1[H+]2 + k 2 K a2[H+] + k 3[H+]K a1 + k4K a1K a2 ([H+] + K a2)([H+] + K a1) (19)

The intrinsic second order rate constant values for each elementary reaction in Table 1 were determined by a nonlinear least-squares regression of the experimental data for both oxidation processes by using SigmaPlot 11.0 Software. The chlorine and bromine acidity constants obtained from literature were used for the calculation. In the case of BP-3, since several pKa values (from 7.56 to 10.2) were reported in literature,31,34 experimental pKa value (pKa = 9.6) obtained in this work by spectrophotometry was used for calculation. Figure 1 shows a good correlation between the calculated and experimental values for both chlorination and bromination. This confirms that the elementary reactions presented in Table 1 can explain the global reaction of chlorine and bromine with BP-3. Table 1 presents the elementary rate constant values obtained from this model. The results show that the overall reaction is mainly controlled by the reaction between the protonated form of the oxidant (HOCl and HOBr for both chlorine and bromine, respectively) and the ionized form of BP-3 (eq 13) in the tested pH range (5−11). This is in agreement with the literature data on chlorination and bromination of phenolic compounds.17,22,27−30 Thus, for both oxidants, a maximum apparent rate constant for pH (pKa1 + pKa2)/2 was observed. Finally, bromine was found to be 4 orders of magnitude more reactive with BP-3 than chlorine when comparing both oxidation processes (Figure 1). Considering [Oxidant]T = 14 μM ([Chlorine]T = 1 mg/L; [Bromine]T = 2.2 mg/L), the half-lives of BP-3 of 2.23 min and E

DOI: 10.1021/acs.est.5b03559 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology

Table 2. Rate Constants for the Reactions of HOX with Dissociated BP-3 and Transformation Products and Apparent Rate Constants at pH 7 and 8.5 Used for Kinetic Model apparent rate constant (M−1·s−1)

a

n

reagents

transformation products

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

HOCl + BP-3− HOCl + ClBP-3− HOCl + Cl2BP-3− HOCl + Br− HOBr + BP-3− HOBr + BrBP-3− HOBr + Br2BP-3− HOCl + BrBP-3− HOCl + Br2BP-3− HOCl + BrClBP-3− HOBr + ClBP-3− HOBr + Cl2BP-3− HOBr + BrClBP-3− HOCl + NH3 NH2Cl + H2O HOCl + NH2Cl NHCl2 + H2O NH2Cl + BP-3−

ClBP-3− Cl2BP-3− Cl3BP-3− HOBr + Cl− BrBP-3− Br2BP-3− Br3BP-3− BrClBP-3− Br2ClBP-3− BrCl2BP-3− BrClBP-3− BrCl2BP-3− Br2ClBP-3− NH2Cl + H2O NH3 + HOCl NHCl2 + H2O NH2Cl + HOCl products

b

Σσom,p 0.4036,39 0.5536,39g 0.7836,39 −0.7338,40 −0.3438,40g 0.0538,40 0.5536,39g 0.7836,39 0.7836,39 −0.3638,40g 0.0138,40 0.0338,40

0.4036,39

c

pKaf a 9.6 8.7b 8.0b

9.6a 8.7b 8.0b 8.7b 8.0b 8.0b 8.7b 8.0b 8.0b 9.25

9.6a

−1 −1

k2 (M ·s ) 1.80 4.81 6.56 1.55 9.60 3.08 1.74 4.81 6.56 6.56 3.56 2.33 2.01 3.07 2.11 1.50 6.39

× × × × × × × × × × × × × × × × ×

105a 104c 103c 103(20) 107a 106d 105d 104c 103c 103c 106d 105d 105d 106 41 10−5 41 102 42 10−7 42

pH 7 5.27 7.30 4.63 1.20 3.01 5.95 1.56 7.30 4.63 4.63 6.88 2.09 1.80 1.33

× × × × × × × × × × × × × ×

102a 102e 102e 103e 105a 104e 104e 102e 102e 102e 104e 104e 104e 104e

1.20 × 102e < 0.112a

pH 8.5 1.25 1.84 4.93 1.53 4.04 8.47 9.39 1.84 4.93 4.93 9.79 1.26 1.09 4.58

× × × × × × × × × × × × × ×

103a 103e 102e 102e 106a 105e 104e 103e 102e 102e 105e 105e 105e 104e

1.53 × 101e 0.112a 31 d

This study. Calculated using Sparc Online. k2 calculated using the Hammett correlation for meta-dihydroxybenzene chlorination. k2 calculated using the corrected Hammett correlation for phenol bromination (this study). eApparent rate constant kOxidant/P calculated by considering the reaction between HOX and the ionized form as the major reaction. fpKa values correspond to the compound reacting with the oxidant (HOCl or HOBr). gCalculated by considering the substituent in ortho position.

Figure 3. BP-3 chlorination in the presence of bromide and ammonia ([BP-3]T,0 = 1.4 (±0.1) μM; [Chlorine]T,0= 5.3 (±0.2) μM T = 20 ± 2 °C); (a) pH 7; [Br−] = 0−5 μM; [NH4+] = 0−5 μM; (b) pH 7; [Br−] = [NH4+] = 5 μM (c) pH 8.5; [Br−] = 0−5 μM; [NH4+] = 0−5 μM (d) pH 8.5; [Br−] = [NH4+] = 5 μM. Symbols represent experimental data and solid lines represent the modeling obtained via Copasi.

electrophile substitution on the ortho and/or para sites to the phenol function would be expected during chlorination or bromination of BP-3. The formation of mono-, di-, and trihalogenated derivatives is therefore expected. These trans-

formation products were previously observed by Negreira et al. (2008) and Manasfi et al. (2015).34,35 BP-3 Chlorination in the Presence of Bromide Ions/ Ammonia and Kinetic Model. Experiments were conducted F

DOI: 10.1021/acs.est.5b03559 Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Article

Environmental Science & Technology at pH 7 and 8.5 with 1.5 μM BP-3, 5 μM free chlorine and different bromide or ammonia concentrations (0, 1, 2, and 5 μM). Under these conditions, BP-3 was analyzed for different contact times (Figure 2). As expected by the higher rate constants of bromination compared to chlorination, faster BP-3 degradation was observed in the presence of Br−. After 10 min reaction time, BP-3 decay varied from 57% at pH 7 (80% at pH 8.5) in the absence of bromide to ∼90% in the presence of 1 μM Br−. In the presence of ammonia, the results showed that increasing the ammonia concentration led to a decrease in the BP-3 elimination rate during chlorination (Figure 2c,d). This latter observation could be explained by the fast initial consumption of chlorine by ammonia, yielding chloramines, which are less reactive oxidants. The presence of a high ammonia concentration (5 μM) can lead to complete inhibition of BP-3 chlorination resulting from the complete consumption of free chlorine by ammonia. As chloramines were relatively nonreactive with BP-3, this induced a slight decline (