Oxidation of the antibacterial agent norfloxacin during sodium hypochlorite disinfection of marine culture water

Chemosphere 182 (2017) 245e254

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

Oxidation of the antibacterial agent norfloxacin during sodium hypochlorite disinfection of marine culture water Yuanyuan Zhang a, b, Chuan Rong a, c, Yanqun Song a, c, Yinghui Wang a, b, Jiying Pei a, b, Xinying Tang a, b, Ruijie Zhang a, b, Kefu Yu a, b, d, * a

School of Marine Sciences, Guangxi University, Nanning 530004, China Coral Reef Research Center of China, Guangxi University, Nanning 530004, China School of Environment, Guangxi University, Nanning 530004, China d Guangxi Laboratory on the Study of Coral Reefs in the South China Sea, Nanning 530004, China b c

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


Norfloxacin was oxidized to two new Br-DBPs during sodium hypochlorite disinfection of marine culture water.
The addition of bromide ions accelerates the reaction rate of norfloxacin with sodium hypochlorite.
The carboxyl of NOR was oxidized by sodium hypochlorite when bromide ions were present or it was stable.
The bio-accumulative properties of Br-DBPs were found to be more substantial than those of norfloxacin.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 December 2016 Received in revised form 2 May 2017 Accepted 3 May 2017 Available online 7 May 2017

Chlorination disinfection and antibiotic addition are two universal processes of marine culture. The generation of disinfection byproducts (DBPs) is unavoidable. Antibiotic residue not only pollutes water but also acts as a precursor to the production of new DBPs. The fate of antibiotic norfloxacin (NOR) in chlorination disinfection was investigated. It was observed that NOR could be oxidized by disinfection agent sodium hypochlorite, but the oxidation rate varied considerably with the type of disinfected water. For fresh water, marine culture water and sea water, the reaction rate constant was 0.066 min1, 0.466 min1 and 1.241 min1, respectively. The difference was primarily attributed to the promotion role of bromide ions in seawater and marine culture water. Moreover, the bromide ions could result in the generation of brominated DBPs (Br-DBPs). The kinetics, products, reaction centers and mechanisms were investigated. The active site of NOR was found to be the N4 atom on piperazinyl in fresh water. During marine culture water and sea water disinfection, the carboxyl on NOR was oxidized and two Br-DBPs were formed. This was attributed to the lowering of the reaction's required activation energy when performed in the presence of bromide ions. The Br-DBPs were also confirmed in real shrimp pond brackish water. Quantitative structure activity relationships and the total organic halogen analysis showed that the DBPs in marine culture water possessed stronger toxicological properties than the DBPs in fresh water. The toxicity increase was attributed to the production of Br-DBPs in the disinfection process of marine culture water. © 2017 Elsevier Ltd. All rights reserved.

Handling Editor: Xiangru Zhang Keywords: Norfloxacin Sodium hypochlorite Bromide ion Disinfection Br-DBPs

* Corresponding author. No. 100 East Daxue Road, Xixiangtang District, Nanning, Guangxi Autonomous Region, China. E-mail address: [email protected] (K. Yu). http://dx.doi.org/10.1016/j.chemosphere.2017.05.023 0045-6535/© 2017 Elsevier Ltd. All rights reserved.

246

Y. Zhang et al. / Chemosphere 182 (2017) 245e254

1. Introduction Recirculating aquaculture systems have been widely used in marine culture in order to reduce the consumption of seawater and fresh water resources (Hambly et al., 2015; Krom et al., 2014; Kenneth et al., 2011). In the recirculating system, disinfection processes are essential in preventing epidemic diseases (Gong et al., 2016; Chu et al., 2013; Lyon et al., 2014). The chemical disinfectants used in this process are usually chlorine containing compounds (Park et al., 2016; Liu et al., 2016a,b; Kahler et al., 2016). Chlorine disinfection eliminates pathogens, however chlorine also reacts with other organic substances in water to generate halogenated disinfection byproducts (DBPs) (Liu et al., 2016a,b; Jeong et al., 2015). Halogenated DBPs such as trihalomethanes (THMs) (Han et al., 2015; Pan et al., 2014) and haloacetic acids (HAAs) (Cardador and Mercedes, 2016; Smith et al., 2016) have attracted much concern due to their potential health risks. Antibacterial agents are frequently added to fish feeds in order to promote their growth (Ellsworth et al., 2006; Robert et al., 2008; Sun et al., 2012; Carvalho et al., 2014). These antibiotics, however, cannot be completely absorbed by the animals. As a result, many of them entered into the aquaculture water (Zhang et al., 2013a). Indeed, various antibiotics, such as ciprofloxacin, enrofloxacin, ofloxacin, flumequine, lomefloxacin, and norfloxacin (NOR) have been repeatedly detected at concentrations ranging from 142 to 10,000 mg/L in farm wastewater in China (Zhang et al., 2012). Antibiotic residue in water not only causes marine culture pollution, but also acts as a special precursor that reacts with disinfectants to produce new DBPs (Huber et al., 2005; Navalon et al., 2008). For example, Dodd et al. (2005) reported that several DBPs were originated from the reactions of ciprofloxacin and enrofloxacin with chlorine from the disinfection process. Wang et al. (2010a) reported the oxidation of fluoroquinolone antibiotics by chlorine dioxide, in this process, fluoroquinolone's piperazinyl N4 atom was attacked by chlorine dioxide leading to dealkylation, hydroxylation and intramolecular ring closure at the piperazine moiety. This was then followed by the production of new DBPs through the reaction of fluoroquinolone antibiotics with chlorine dioxide. These investigations were conducted using freshwater samples, and only chlorinated DBPs (Cl-DBPs) have been observed to date. Considering that there are high concentrations of bromide ions (65 mg L1) in seawater, residual antibiotics actually coexist with bromide ions in brackish marine culture water (a mixture of seawater and fresh water at a ratio of approximately 1:2). Hypochlorite with strong oxidation properties can oxidize bromide ions to hypobromite which can react more rapidly with many organic pollutants than hypochlorite (Ichihashi et al., 1999). Moreover, bromine substitution processes are faster than those of chlorine. Therefore, the chlorine disinfection process of marine culture wastewater may possibly generate brominated DBPs (Br-DBPs). It has been reported that the toxicity of Br-DBPs is much higher than that of Cl-DBPs (Liu and Zhang, 2014; Yang and Zhang, 2013). For example, Plewa et al. (2002) reported that bromoacetic acid was 18.4 and 89.8 times more cytotoxic than chloroacetic acid in Salmonella typhimurium and Chinese hamster ovary cells, respectively. Through studies of chronic cytotoxicity and DNA damage (SCGE assay) in CHO cells, including an analysis of the structureeactivity relationships of DBPs, Richardson et al. (2007) revealed that bromoacetic acid was the most genotoxic on mammalian cells of the regulated DBPs. The toxic nature of Br-DBPs encouraged us to investigate the reactions of sodium hypochlorite (NaClO) with antibacterial agents in the presence of bromide ions. As an initial investigation, this paper will be primarily focused on the kinetics and pathways of the reaction between NaClO and

antibacterial agents. Special attention will be paid to the identification of the reaction products. As shown in Fig. 1, NOR is a typical fluoroquinolone antibiotic agent with piperazine ring, nalidixic ring and lateral chain. It is widely used in aquaculture and frequently detected at concentrations of 2.5e35000 mg L1 in shrimp ponds located along the coast of Guangxi, China. Therefore, NOR was selected as the target antibiotic in the initial investigation. The aim of this paper is to evaluate the activity and the health risk of BrDBPs in the chlorine disinfection process of marine culture water. 2. Materials and methods 2.1. Chemicals Analytical grade NOR was obtained from the Dalian Meilun biological technology company and was used without further purification. Norfloxacin stocks were prepared in pure water. NaClO, NaCl, NaBr and Na2S2O3 were obtained from the Guangzhou Chemical Reagent Company, China. The available chlorine content of NaClO amounted to 10% of the total chlorine content. All reagent solutions were prepared using water from a Millipore Milli-Q Ultrapure Gradient A10 purification system. Real brackish water samples of shrimp ponds were obtained from Qinzhou, Guangxi province, China. 2.2. Analytical methods An Agilent 1050 high performance liquid chromatography (HPLC) system equipped with a Zorbax RX-C18 column (4.6 mm  250 mm, 5 mm), fluorescence detector, and singlewavelength UV detector was used to monitor NOR in NaClO disinfection process experiments. The mobile phase used was trifluoroacetic acid and acetonitrile at a flow rate of 0.30 ml min1. The detection wavelength was set to 278 nm for NOR. The residual NaClO after the reaction was determined by way of iodometry. 2.3. Batch kinetic experiments To avoid missing some of the transformed products, the initial concentration of NOR was set to 0.1 mM. Excess NaClO was added to initiate the reaction. In the kinetics studies, the pH value of the reaction solution was maintained by 10 mM phosphate (pH ¼ 6e8) and borate buffer (pH ¼ 9). Samples were collected at specified time intervals during the reaction period by adding Na2SO3 to terminate the reaction. Concentrations of residual NOR were analyzed by HPLC. All of the kinetic experiments were conducted at room temperature (298.15 K) and duplicate at least twice. Pseudofirst-order rate constants for losses of NOR were obtained from the slopes of fitted linear plots of Ln ([NOR]) vs. time (0.97 > r2 > 1). 2.4. Product identification Reaction products were analyzed based on the accurate mass,

Fig. 1. Structure of NOR.

Y. Zhang et al. / Chemosphere 182 (2017) 245e254

isotope patterns, and MS/MS patterns. These were obtained using a Thermo Fisher LC-MS/MS (Q-Exactive) system with a Zorbax SBC18 column (2.4 mm  150 mm, 5 mm), a diode-array UV/vis detector, and a mass spectrometer. Acetonitrile and 0.2% formic acid at a flow rate of 0.30 ml min1 was used for gradient elution. Products with a mass scan range of m/z 50e1000 were analyzed by electrospray ionization in a positive mode (ESIþ). The full width half maximum (FWHM) resolution of the scan mode was set to 0.0001 m/z. Halogen compounds were analyzed by mass spectrometry. Isotopic ion peaks were found in the mass spectra of organic compounds with non-single isotopic compositions, and especially for compounds containing bromine and chlorine elements. The fragmentation and formation mechanisms of all identifiable products were obtained by Mass Frontier 7.0. 3. Results and discussion 3.1. Reaction kinetics of NOR with NaClO The reaction of NOR with NaClO was conducted in fresh water, synthetic brackish marine culture water and seawater, respectively. Fig. 2a shows that NOR was mostly oxidized within approximately 10 min in both marine culture water and seawater. In contrast, the concentration of NOR remained at 17 mg L1 after 10 min reaction in fresh water, it took approximately 2 h to completely degrade. These results indicate that bromide ions promote the reaction of NOR with NaClO. Although the oxidation process of NOR is complex, the main consecutive reactions can be roughly described as follows:

HClO þ Br /HBrO þ Cl

(I)

HBrO þ NOR/products

(II)

The next direct simultaneously:

oxidation

of

NOR

by

HClO þ NOR/products

HClO

occurred

(III)

Therefore, the rate equation of the consecutive reactions should be:


 h ib d½HClO d Br ¼ ¼ k1 ½HClOa Br  dt dt d½NOR ¼ k2 ½HBrOc ½NORd dt The rate equation of reaction III should be:

(1)

(2)

d½NOR ¼ k3 ½HClOe ½NORf dt

247

(3)

where k is the reaction rate constant and a, b, c, d, e and f represent the reaction order with respect to the corresponding matter. In actual marine culture, the concentration of antibiotic is generally much lower than that of the disinfectant. Therefore, the initial mole ratio of NaClO to NOR was set at more than 10:1. Moreover, although the concentration of bromide ions in seawater (0.28 mM) and the marine culture water (0.1 mM) is higher than that in fresh water (negligible), it is considerably less than that of NaClO. The complete oxidation of bromide ions and NOR should thus consume a small amount of NaClO. The concentration of NaClO was determined to be 0.226 M after the reaction, which was slightly lower than the initial concentration of 0.24 M. The concentration of HClO can be approximated as constant in the whole reaction process. According to the reported reference (Ichihashi et al., 1999), the kinetics of reaction (I) are first order with respect to both HClO and Br, and the rate is rapid with k1 > 2  103 M1 S1. Thus, eq. (1) can be simplified to eq. (4):

h i d½HClO d½Br ¼ ¼ k01 Br dt dt

(4)

where k01 ¼ k1 ½HClO. Hypobromous acid was detected using ion chromatography. It was found that the concentrations of HBrO laid approximately within the range of 2e4 mg L1 during the reaction. As the intermediate product in the consecutive reaction, HBrO at a low concentration can be considered invariable or in its stable state. A similar assumption can be found in many kinetic studies (Deng et al., 2008). Using this assumption, eq. (2) could be described as follows:

d½NOR ¼ k02 ½NORd dt

(5)

where k02 ¼ k2 ½HBrOc . Eq. (3) could be described as follows:

d½NOR ¼ k03 ½NORf dt

(6)

where k03 ¼ k3 ½HClOe . The kinetics data revealed that the reactions were all pseudo first order for NOR. The reaction rate k2 was significantly smaller than k1 (Ichihashi et al., 1999). As the mathematical treatment of consecutive reaction is too complicated, approximate processing

Fig. 2. Oxidation of NOR by NaClO in different water samples and the linear fit of the kinetic data ([NOR]0 ¼ 0.5 mM, 20 ml and [NaClO]0 ¼ 0.24 M, 1 ml, pH ¼ 7, T ¼ 298.15 K).

248

Y. Zhang et al. / Chemosphere 182 (2017) 245e254

Fig. 3. Oxidation of NOR by NaClO at different concentrations of Br1 and the linear fit of the kinetic data ([NOR]0 ¼ 0.5 mM, 20 ml, [NaClO]0 ¼ 0.24 M, 1 ml, [Cl1] ¼ 6.6 mg L1, pH ¼ 7, T ¼ 298.15 K).

methods are generally used in the kinetics studies. The step with the lowest reaction rate is considered as the rate determining step. Therefore, reaction (II) with the lowest reaction rate is the rate determining step. In other words, the rates of the two consecutive reactions (I) and (II) were mainly determined by the reaction of HBrO with NOR. The rate of reaction III was the same as that of the reaction in the fresh water and lower than that in sea water and marine culture water. The reaction rates of NOR oxidation in various mediums followed the order seawater > marine culture water > fresh water (Fig. 2b). Their rate constants calculated by the slopes of the curves were 0.066 min1, 0.466 min1 and 1.241 min1, respectively. We observed that the rate constant was considerably dependent on the concentration of bromide ions. For example, when Cl concentration was fixed at 6.6 mg L1 and concentrations of Br were changed from 0 mg L1 to 65 mg L1, the reaction rate increased from 0.0765 min1 to 0.6667 min1. The latter is 8.7 times the

former (Fig. 3). This suggests that the oxidation of NOR by NaClO is strongly promoted by the presence of bromide ions. The mechanism of this promotion is an acceleration of the rate determining step (II) (Amy et al., 2000). The reason may be attributed to that the reactivity of bromine with organic precursors is stronger than that of chlorine (Chang et al., 2001). For most antibiotics including NOR, the presence of ionizable group such as carboxylic and amidocyanogen means that the antibiotic molecule will assume various dissociation states depending on the pH (Gallard and Gunten, 2002). Scheme 1 shows the dissociation states of NOR across the entire pH range. The pH value was shown to have obvious effects on the reaction of NOR with disinfectant. For example, as reported by Zhang and Huang, 2005, the specific reaction rate constants of cationic, neutral, and anionic species of NOR with ClO2 were <1.0  104 M1 s1, 10.1 M1 s1, and 1.76  102 M1 s1, respectively. These results indicated that the anionic NOR species was most reactive with

Scheme 1. Speciation patterns of NOR.

Fig. 4. Oxidation of NOR by NaClO at different pH values and the linear fit of the kinetic data ([NOR]0 ¼ 0.1 mM, 20 ml and [NaClO]0 ¼ 0.24 M, 1 ml, [Cl1]0 ¼ 6.6 g L1, [Br1]0 ¼ 22 mg L1, pH ¼ 7, T ¼ 298.15 K).

Y. Zhang et al. / Chemosphere 182 (2017) 245e254

oxidants and the cationic species was insignificant in this regard. Fig. 4 presents the variations in the rate of NOR oxidation in synthetic marine culture water with different pH values. The rate constants increased from 0.367 min1 to 0.591 min1 when the pH value was raised from 6 to 9. This trend should mainly be attributed to NOR speciation, which may be heavily influenced by solution pH. pKa is the pH at which a group dissociated in half of the total concentration. Moreover, macroscopic Ka1 and Ka2 constants are often linked to carboxylate groups and piperazinyl N4 atoms (Wang et al., 2010a), respectively. For NOR, pKa1 and pKa2 have been reported to be 6.22 and 8.51, respectively (Wang et al., 2010b). Faster reaction rates at high pH values indicated that the main reaction active site of NOR was the piperazinyl N4 atom corresponding to pKa2. The pKa values for HClO and HBrO were found to be 7.5 and 8.62, respectively. In the reactions of both aqueous bromine (HOBr/ OBr) and chlorine (HOCl/OCl) with organic matter, the halogen dissociation effect could be precluded (Westerhoff et al., 2004). Therefore, NOR functional dissociation may account for the effect of pH values on the reaction rates in this research. 3.2. Product and reaction center identification As the chromatogram of products showed no difference after 2 h reaction, the final transformation products of NOR's reaction with excess NaClO were determined to be stable. The composition and structure of the compounds were analyzed based on the exact molecular weight obtained from a high resolution LC/MS/MS spectrum and fragment ions as shown in Fig. S1 and Table S1. To examine the effects of Cl and Br on product formation, the DBPs in fresh water, simulated brackish marine culture water and seawater were analyzed. These products were designated p1, p2

249

and p3, respectively, and product mass losses or gains from the parent compound are given in the following figures and schemes. Scheme 2 shows that the piperazine ring of NOR was oxidized by NaClO in fresh water without Cl and Br, the N4 and C atoms were removed while the N1 atom remained. Chlorine substitution occurred at the C8 position, generating a mono-chlorinated compound p1. The appearance of product p1 indicated that in the absence of Cl and Br, the main reaction centers of NOR with NaClO include the N4 atom on the piperazine ring and the C8 atom on the nalidixic ring. However, the water conditions of the simulated brackish marine culture water and seawater differed from those of the fresh water. The mass spectrometry isotope peak clusters indicated that both the mono-chlorinated-bromide compound p2 and dibromide compound p3 emerged in the brackish marine culture water and seawater. Combined with the accurate mass and MS/MS pattern of their fragmentation (Fig. S1), the structures of p2 and p3 are shown in Scheme 2. The carboxyl group in the nalidixic ring was oxidized and replaced by a bromine atom. Apart from the piperazine ring and C8 atom, the carboxyl group in the quinolone ring became the other primary reaction site in the presence of Cl and Br. This finding may be attributed to the bromide ions which were oxidized to BrO by ClO in the simulated brackish marine culture water and seawater. Although the oxidation activity of BrO (standard electrode potential 1.574 V) is slightly weaker than that of ClO (standard electrode potential 1.611 V) (Milica et al., 2015), the activation energy required for the reaction of NaClO/Br with NOR is lower than that of NaClO with NOR. This means that the reaction should occur more easily and be comparatively faster. The Arrhenius equation reflects the relationship between the chemical reaction rate constant and temperature change (Zhang et al., 2013b). Associated quantitative laws can be

Scheme 2. Final products of NOR reactions with NaClO in different water samples ([NOR]0 ¼ 0.1 mM, 20 ml and [NaClO]0 ¼ 0.24 M, 1 ml, pH ¼ 7).

250

Y. Zhang et al. / Chemosphere 182 (2017) 245e254

Fig. 5. The reaction activation energy calculation for NOR with NaClO in fresh water (a. Linear fit of kinetic data at different reaction temperatures, b. Linear of Lnk and 1/T) ([NOR]0 ¼ 0.1 mM, 20 ml and [NaClO]0 ¼ 0.24 M, 1 ml, pH ¼ 7).

Fig. 6. The reaction activation energy calculation of NOR with NaClO in simulated brackish marine culture water (a. Linear fit of kinetic data at different reaction temperatures, b. Linear of Lnk and 1/T) ([NOR]0 ¼ 0.1 mM, 20 ml and [NaClO]0 ¼ 0.24 M, 1 ml, [Cl1]0 ¼ 6.6 g L1, [Br1]0 ¼ 22 mg L1, pH ¼ 7).

described as follows:

  Ea k ¼ A exp RT

(7)

respectively. Thus, the reaction of NaClO with NOR occurred at an accelerated rate in the presence of bromide ions. Moreover, the carboxyl group in the NOR was oxidized. 3.3. Proposed reaction pathways

where k is the reaction rate constant, Ea and A are the reaction activation energy and pre-exponential factors, respectively, R is the molar gas constant and T is the reaction thermodynamic temperature. Activation energies were obtained from the kinetic data at different reaction temperatures as shown in Figs. 5 and 6. Pseudofirst-order rate constants of NOR with NaClO were obtained both for the fresh water and simulated brackish marine culture water. The energies corresponding to the reaction of NaClO with NOR and NaClO/Br with NOR were calculated to be 104.99 KJ and 33.96 KJ,

In fresh water. Intermediate products of NOR reacting with excess NaClO were determined to infer the proposed reaction pathways. Mass spectral data and fragmentation information for products are provided in Fig. S2 and Table S1. As shown in Scheme 3, the dashed boxes and solid boxes represent the intermediate products and the final products, respectively. A piperazine moiety fracture was found in the transformation product pf1 with m/ z ¼ 294.1250 after a 10 min reaction in fresh water without Cl1 and Br1. Following this, the mono-chlorine compound pf2 with m/

Scheme 3. Products and proposed pathways for NOR reactions with NaClO in fresh water ([NOR]0 ¼ 0.1 mM, 20 ml, [NaClO]0 ¼ 0.24 M, 1 ml, pH ¼ 8).

Y. Zhang et al. / Chemosphere 182 (2017) 245e254

251

Scheme 4. Products and proposed pathways for NOR reactions with NaClO in the simulated brackish marine culture water and seawater ([NOR]0 ¼ 0.1 mM, 20 ml and [NaClO]0 ¼ 0.24 M, 1 ml, pH ¼ 7).

z ¼ 328.0860 appeared. When Na2SO3 was added to consume the residual NaClO and terminate the reaction, the stable product pf2 detected during the reaction verified that chlorine atoms are not associated with reducible N-Cl bonds. The reason can be attributed to that SO2 3 can reduce chlor-amine to its parent amine structures (Qiang et al., 2006). When the spectral data was combined with fragmentation information, structures of pf2 were confirmed and p1 with m/z ¼ 285.0639 was generated through the further oxidation of N4 atoms. In simulated brackish marine culture water and seawater. Intermediate products obtained from simulated brackish marine culture water (where concentrations of Cl and Br were 6.6 g L1 and 22 mg L1, respectively) and seawater (where concentrations of Cl and Br were 19.5 g L1 and 65 mg L1, respectively) were analyzed. At the same concentration ratio between Cl and Br, the reaction processes of NOR with NaClO in the above two environments were essentially the same with the exception of their reaction rates. As shown in Scheme 4, the peak of NOR was still present after reacting for 2 min in the simulated brackish marine culture water. At the same time, the appearance of a new peak at m/z 354.0608 indicated that a new product pb1 was generated. According to the structural analysis of pb1, decarboxylation which was a phenomenon

Table 1 Characteristics of real brackish marine culture water from a shrimp pond (mg/L). pH

Cl

Br

CODMn

NO 3

F

HCO 3

SO24

8.1

7100

24.3

4.42

<0.3

0.5

52

857

different from the reaction in fresh water occurred first. Norfloxacin was undetectable after 30 min, the new m/z peaks at 362.0070, 431.9723, 405.9565, 328.0457 and 285.0026 indicated that products pb2-pb6 had been generated. All of the LC/MS-ESIþ spectra and fragmentation information for the identifiable products pb1-

Table 2 The concentrations of THMs, HAAs and HANs analyzed by GC/MS. t ¼ 2 h, T ¼ 303.15 K, [NaClO]0 ¼ 0.24 M, pH ¼ 7 TTHM, mg/l TCM DCM MCM BCM TBM DBM MBM BDCM DBCM P TTHM

THAA, mg/l 4.6 N/A N/A 1.4 118.6 N/A N/A 10.8 18.5 154

MCAA MBAA DCAA TCAA BCAA DBAA BDCAA DBCAA TBAA P THAA

THAN, mg/l 10.2 3.7 4.9 2.6 N/A 24.5 N/A 21.4 41.7 109

DCAN TCAN BCAN DBAN P THAN

N/A N/A 4 N/A 4

N/A:not available. THMs: (chloroform[TCM], dichloromethane[DCM], bromodichloromethane [BDCM], monochloromethane [MCM], bromochloromethane[BCM], dibromochloromethane [DBCM], and bromoform [TBM], dibromomethane [DBM], monobromomethane [MBM]). HAAs: (monochloroacetic acid[MCAA], dichloroacetic acid[DCAA], trichloroacetic acid[TCAA], bromochloroacetic acid[BCAA], monobromoacetic acid [MBAA], dibromoacetic acid [DBAA], bromodichloroacetic acid[BDCAA], dibromochloroacetic acid [DBCAA], tribromoacetic acid[TBAA]). HANs: (dichloroacetonitrile [DCAN], trichloroacetonitrile [TCAN], bromochloroacetonitrile [BCAN], dibromoacetonitrile [DBAN]).

252

Y. Zhang et al. / Chemosphere 182 (2017) 245e254

pb6 are given in Fig. S3 and Table S1. From these structures, it can be seen that the piperazine ring on NOR was attacked by oxidative substances and fractured. Moreover, halogenation occurred at the C8 atom of the nalidixic ring. As halogenated amine formation was impossible after the addition of Na2SO3, the latter halogenation was a substitution for the carboxyl. Apart from the peaks of 318.9571 and 362.9137, no new product appeared after reaction of 2 h. At the same time, the peaks of pb1-pb6 disappeared indicating that the final products were p2 and p3. 3.4. Products identified in real brackish marine culture water  In addition to the chloride and bromide ions, SO2 4 , HCO3 and other inorganic anions containing in seawater may influence reactions between NOR and NaClO. The organic matter, other than NOR will react with NaClO to generate other typical DBPs such as THMs and HAAs. We thus examined the reaction in real brackish marine culture water collected from a shrimp pond in Qinzhou, Guangxi (features of the water sample are shown in Table 1). The typical DBPs were determined by GC/MS as shown in Table 2. The concentration of THMs, HAAs and HANs were 154 mg L1, 109 mg L1 and 4 mg L1, respectively. The content of Br-DBPs was significantly higher than that of Cl-DBPs due to the high bromide concentration in marine culture water. Tribromomethane which contains three

bromine atoms accounted for 77% of the total five THMs. In the seven detected HAAs, the content of TBAA which also contains three bromine atoms was the highest. Bromochloroacetonitrile as a kind of haloacetonitrile was detected to be 4 mg L1. For reaction of NOR with NaClO in the simulated brackish water, the number of product peaks increased when initial concentrations of NOR and NaClO were the same as in the simulated water samples (as shown in Fig. 7a). However, the main products p2 and p3 still appeared in the chromatogram at 8.28 min and 8.48 min, respectively. Our detection of p2 and p3 (Fig. 7b) indicated the possibility of the reaction between NOR, NaClO, Cl, and Br in actual marine culture water. Quantitative structure activity relationships (QSAR) models are usually used to link chemical activities with molecular structure and composition. Recently, QSAR models are developed to investigate DBP behaviors. The Hammett equation is a classic algorithm which is often applied to compounds with different substituents on the aromatic group. As shown in equation (8), the Hammett models can be made by single linear relationship equations.

logðYÞ ¼ rs þ constant

(8)

where Y is the activity of compound, r is the electronic effect coefficient and s is the electronic effect constant.

Fig. 7. a. Chromatogram of products reacted in real brackish marine culture water. b. LC/MS-ESIþ spectra for products appeared in the chromatogram at 8.28 min (p2) and 8.48 min (p3).

Y. Zhang et al. / Chemosphere 182 (2017) 245e254 Table 3 The substituent R1 and R2 of the three DBPs.

253

showed that the DBPs in marine culture water were more toxic than the DBPs in fresh water. These results indicated that for the disinfection of marine culture water containing the bromide ions, the use of disinfection agent NaClO should be minimized. Acknowledgments

As shown in the equation, there is a positive exponential relationship between Y and s. That is, the larger the value of s, the greater the activity of the compound. For P1, P2 and P3, the chemical structure formula can be described uniformly. The difference between the three DBPs was the substituent of R1 and R2 as shown in Table 3. The electronic effect constant s of the substituent was 0.37, 0.37 and 0.39 for Cl, COOH and Br, respectively. For P1 and P2, the only difference was the substituent on R1. As the s value of the carboxyl group of P1 was smaller than the bromine atom of P2, the Y value of P1 was smaller than P2. For the comparison between P2 and P3, the substituent on R2 caused the difference in the activity. The s value of bromine atom on P3 was bigger than chlorine atom on P2, so the activity of P3 was greater than P2. These results showed that the activity of the DBPs followed the order P3 > P2 > P1. The bioaccumulation properties of P1, P2 and P3 may be stronger than NOR. This may be attributed to the fact that for P2 and P3, the removal of hydrophilic carboxyl groups enhanced hydrophobicity levels. In addition, as halogen non-bonding electrons interacted with p electron ions, the polarity of the carbon-halogen bond declined, and thus halogen atoms significantly improved hydrophobicity levels. In the future, we will examine the absolutely numerical simulation of the DBPs by QSAR model to identify unknown characteristics of the new DBPs. Total organic halogen analysis (Li et al., 2017) which is a good surrogate and a toxicity indicator for the overall halogenated DBPs in a disinfected water sample was also determined. The data obtained for the fresh water, marine culture water and sea water were 3.14 mg L1, 15.62 mg L1 and 48.32 mg L1, respectively. These results demonstrated that DBPs produced in the marine culture water were more toxic than the DBPs in the fresh water. Because NOR is relatively stable in water for an extended period of time, the concentration of NaClO will decrease with time, DBPs P2 and P3 can partly be avoided when NOR is not immediately injected after NaClO disinfection. 4. Conclusions This report described the first investigation of the fate of NOR in the chlorination disinfection of marine culture water. Data showed that NOR could be oxidized by the disinfection agent NaClO, but that the oxidation rate was highly dependent on the type of the disinfected water. For reactions performed in the presence of a fixed concentration of Cl1, rate constants increased with the increase of Br1 concentration. In fresh water, the main reaction centers of NOR were the N4 atom on the piperazine ring and the C8 atom on the nalidixic ring. In addition to these centers, the carboxyl group in the quinolone ring was oxidized in marine culture water and seawater. Reaction activation energies of NaClO and NaClO/Br with NOR were measured to be 104.99 KJ and 33.96 KJ, respectively, indicating that the oxidation of NOR occurred more easily in the presence of bromide ions. Moreover, the two new Br-DBPs simultaneously found in the simulated marine culture water and seawater were confirmed to exist in actual brackish marine culture water from a shrimp pond. QSAR and total organic halogen analysis

This study was supported by the National Natural Science Foundation of China (51668005, 21367003, 41473118, and 41273139), the Natural Science Foundation of Guangxi Province (2015GXNSFBB139009, 2014GXNSFAA118296), the China Postdoctoral Science Foundation (No. 2016M592607) and the BaGui Scholars Program Foundation (2014). Extra grammatical support was provided by Mr Hainian Yu from the University of Queensland, Australia. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.chemosphere.2017.05.023. References Amy, G., Douville, C., Daw, B., Sohn, J., Galey, C., Gatel, D., Cavard, J., 2000. Bromate formation under ozonation conditions to inactivate cryptosporidium. Water Sci. Technol. 41, 61e66. Cardador, M.J., Mercedes, G., 2016. Static headspace-gas chromatography-mass spectrometry for the simultaneous determination of trihalomethanes and haloacetic acids in canned vegetables. J. Chromatogr. 1454, 9e14. Carvalho, A.L., Ferreira, B.F., Martins, C.H.G., Nassar, E.J., Nakagaki, S., Machado, G.S., 2014. Tetracarboxyphenylporphyrinekaolinite hybrid materials as efficient catalysts and antibacterial agents. J. Phys. Chem. C 118, 24562e24574. Chang, E.E., Lin, Y.P., Chiang, P.C., 2001. Effects of bromide on the formation of THMs and HAAs. Chemosphere 43, 1029e1034. Chu, W.H., Gao, N.Y., Yin, D.Q., Krasner, S.W., 2013. Formation and speciation of nine haloacetamides, an emerging class of nitrogenous DBPs, during chlorination or chloramination. J. Hazard. Mater. 260, 806e812. Deng, Y.H., Jiang, J.Y., Zhang, Y.Y., Lin, X.P., Du, C.M., Xiong, Y., 2008. FeVO4 as a highly active heterogeneous Fenton-like catalyst towards the degradation of Orange II. Appl. Catal. B Environ. 84, 468e473. Dodd, M.C., Shah, A.D., Gunten, U.V., Huang, C.H., 2005. Interactions of fluoroquinolone antibacterial agents with aqueous chlorine: reaction kinetics, mechanisms, and transformation pathways. Environ. Sci. Technol. 39, 7065e7076. Ellsworth, E.L., Tran, T.P., Hollis, H.D., Joseph, S., 2006. 3-Aminoquinazolinediones as a new class of antibacterial agents demonstrating excellent antibacterial activity against wild-type and multidrug resistant organisms. J. Med. Chem. 49, 6435e6438. Gallard, H., Gunten, U.V., 2002. Chlorination of phenols: kinetics and formation of chloroform. Environ. Sci. Technol. 36, 884e890. Gong, T.T., Tao, Y.X., Xian, Q.M., 2016. Selection and applicability of quenching agents for the analysis of polar iodinated disinfection byproducts. Chemosphere 163, 359e365.
ska-Sobecka, B., Hambly, A.C., Arvin, E., Pedersen, L.F., Pedersen, P.B., Seredyn Stedmon, C.A., 2015. Characterising organic matter in recirculating aquaculture systems with fluorescence EEM spectroscopy. Water Res. 83, 112e120. Han, Q., Yan, H., Zhang, F., Xue, N., Wang, Y., Chu, Y.B., Gao, B.Y., 2015. Trihalomethanes (THMs) precursor fractions removal by coagulation and adsorption for bio-treated municipal wastewater: molecular weight, hydrophobicity/hydrophily and fluorescence. J. Hazard. Mater. 297, 119e126. Huber, M.M., Korhonen, S., Ternes, T.A., Gunten, U.V., 2005. Oxidation of pharmaceuticals during water treatment with chlorine dioxide. Water Res. 39, 3607e3617. Ichihashi, K., Teranishi, K., Iehimura, A., 1999. Brominated trihalomethane formation in halogenation of humic acid in the coexistence of hypoehlorite and hypobromite. Water Res. 33, 477e483. ~ as, B.J., Jeong, C.H., Postigo, C., Richardson, S.D., Simmons, J.E., Kimura, S.Y., Marin Barcelo, D., Liang, P., Wagner, E.D., Plewa, M.J., 2015. Occurrence and comparative toxicity of haloacetaldehyde disinfection byproducts in drinking water. Environ. Sci. Technol. 49, 13749e13759. Kahler, A.M., Cromeans, T.L., Metcalfe, M.G., 2016. Aggregation of adenovirus 2 in source water and impacts on disinfection by chlorine. Food Environ. Virol. 8, 148e155. Kenneth, Schneider, Sher, Yonatan, Erez, Jonathan, van Rijn, Jaap, 2011. Carbon cycling in a zero-discharge mariculture system. Water Res. 45, 2375e2382. Krom, M.D., Ben David, A., Ingall, E.D., Benning, L.G., Clerici, S., Bottrell, S., Davies, C., Potts, N.J., Mortimer, R.J.G., Rijn, J., 2014. Bacterially mediated removal of phosphorus and cycling of nitrate and sulfate in the waste stream of a “zero-

254

Y. Zhang et al. / Chemosphere 182 (2017) 245e254

discharge”recirculating mariculture system. Water Res. 56, 109e121. Li, Y., Zhang, X.R., Yang, M.T., Liu, J.Q., Li, W.X., Graham, Nigel J.D., Li, X.Y., Yang, B., 2017. Three-step effluent chlorination increases disinfection efficiency and reduces DBP formation and toxicity. Chemosphere 168, 1302e1308. Liu, P., Farre, M.J., Keller, J., 2016a. Reducing natural organic matter and disinfection by-product precursors by alternating oxic and anoxic conditions during engineered short residence time riverbank filtration: a laboratory-scale column study. Sci. Total Environ. 565, 616e625. Liu, W., Wei, D.B., Liu, Q., Du, Y.G., 2016b. Transformation pathways and acute toxicity variation of 4-hydroxylbenzophenone in chlorination disinfection process. Chemosphere 154, 491e498. Liu, J.Q., Zhang, X.R., 2014. Comparative toxicity of new halophenolic DBPs in chlorinated saline wastewater effluents against a marine alga: halophenolic DBPs are generally more toxic than haloaliphatic ones. Water Res. 65, 64e72. Lyon, B.A., Cory, R.M., Weinberg, H.S., 2014. Changes in dissolved organic matter fluorescence and disinfection by product formation from UV and subsequentchlorination/chloramination. J. Hazard. Mater. 264, 411e419. Milica, J., Fernando, C.S., Andreas, L., 2015. Electrochemical detection of free chlorine at inkjet printed silver electrodes. J. Electroanal. Chem. 756, 171e178. Navalon, S., Alvaro, M., Garcia, H., 2008. Reaction of chlorine dioxide with emergent water pollutants: product study of the reaction of three beta -lactam antibiotics with ClO2. Water Res. 42, 1935e1942. Pan, S.L., An, W., Li, H.Y., Su, M., Zhang, J.L., Yang, M., 2014. Cancer risk assessment on trihalomethanes and haloacetic acids in drinking water of China using disability-adjusted life years. J. Hazard. Mater. 280, 288e294. Park, K.Y., Choi, S.Y., Lee, S.H., 2016. Comparison of formation of disinfection byproducts by chlorination and ozonation of wastewater effluents and their toxicity to Daphnia magna. Environ. Pollut. 215, 314e321. Plewa, M.J., Kargalioglu, Y., Vankerk, D., Minear, A.R., Wagner, D.E., 2002. Mammalian cell cytotoxicity and genotoxicity analysis of drinking water disinfection byproducts. Environ. Mol. Mutagen 40, 134e142. Qiang, Z.M., Macanley, J.J., Mormile, M.R., 2006. Treatment of antibiotics and antibiotic resistant bacteria in swine wastewater with free chlorine. J. Agric. Food Chem. 54, 8114e8154. Richardson, S.D., Plewa, M.J., Wagner, E.D., 2007. Occurrence, genotoxicity and

carcinogenicity of regulated and emerging disinfection byproducts in drinking water: a review and road map for research. Mutat. Res. 636, 178e242. Robert, M., Guenther, S., Reinhard, W., Rolando, C.S., Tahira, S., Lukas, T., Isabell, F., 2008. Proanthocyanidins: target compounds as antibacterial agents. J. Agric. Food Chem. 56, 6959e6966. Smith, R.B., Edwards, S.C., Best, N., 2016. Birth weight, ethnicity, and exposure to trihalomethanes and haloacetic acids in drinking water during pregnancy in the born in bradford cohort. Environ. Health Perspect. 124, 681e689. Sun, B., Slomberg, D.L., Chudasama, S.L., Lu, Y., Schoenfisch, M.H., 2012. Nitric oxidereleasing dendrimers as antibacterial agents. Biomacromolecules 13, 3343e3354. Wang, P., He, Y.L., Huang, C.H., 2010a. Oxidation of fluoroquinolone antibiotics and structurally related amines by chlorine dioxide: reaction kinetics, product and pathway evaluation. Water Res. 44, 5989e5998. Wang, P., He, Y.L., Huang, C.H., 2010b. Reactions of tetracycline antibiotics with chlorine dioxide and free chlorine. Water Res. 45, 1838e1846. Westerhoff, P., Chao, P., Mash, H., 2004. Reactivity of natural organic matter with aqueous chlorine and bromine. Water Res. 38, 1502e1513. Yang, M.T., Zhang, X.R., 2013. Comparative developmental toxicity of new aromatic halogenated DBPs in a chlorinated saline sewage effluent to the marine polychaete Platynereis dumerilii. Environ. Sci. Technol. 47, 10868e10876. Zhang, R.J., Tang, J.H., Li, J., Cheng, Z.N., Chakra, C., Liu, D.Y., Zheng, Q., Zhang, G., 2013a. Occurrence and risks of antibiotics in coastal aquatic environment of the Yellow Sea, North China. Sci. Total Environ. 450, 197e204. Zhang, R.J., Tang, J.H., Li, J., Zheng, Q., Liu, D., Chen, Y.J., Zou, Y.D., Chen, X.X., Luo, C.L., Zhang, G., 2012. Antibiotics in the offshore waters of the bohai sea and the Yellow Sea in China: occurrence, distribution and ecological risks. Environ. Pollut. 174, 71e77. Zhang, H., Huang, C.H., 2005. Oxidative transformation of fluoroquinolone antibacterial agents and structurally related amines by manganese oxide. Environ. Sci. Technol. 39, 4474e4483. Zhang, Y.Y., Xiong, Y., Tang, Y.K., Wang, Y.H., 2013b. Degradation of organic pollutants by an integrated photo-Fenton-like catalysis/immersed membrane separation system. J. Hazard. Mater. 245, 758e764.