- Email: [email protected]
chlorine process in the presence of benzophenone-4
Chemical Engineering Journal 351 (2018) 304–311
Contents lists available at ScienceDirect
Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej
Enhanced formation of chlorinated disinfection byproducts in the UV/ chlorine process in the presence of benzophenone-4
T
⁎
Junhe Lua, , Peizeng Yanga, Wei Donga, Yuefei Jia, Qingguo Huangb a b
Department of Environmental Science and Engineering, Nanjing Agricultural University, Nanjing 210095, China Interdisciplinary Toxicology Program, Department of Crop and Soil Sciences, University of Georgia, Griffin, GA 30223, USA
H I GH L IG H T S
was efficiently removed • BP4 chlorine, chloramine, and UV/
G R A P H I C A L A B S T R A C T
by
chlorine treatment.
of BP4 by UV/chlorine was • Removal mostly arose from the reactions with % OH and Cl%.
in the UV/chlorine process • Methanol suppressed BP4 removal but enhanced DBPs formation.
BP4 and O could be formed be• cause of the photochemical activity of ∗
3
1
2
BP4.
responsible for DBP formation • inOthewasUV/chlorine/methanol process. 1
2
A R T I C LE I N FO
A B S T R A C T
Keywords: Benzophenone-4 Chlorine UV/chlorine Disinfection byproducts Photosensitizer
Benzophenone-4 (BP4) is UV-filter that is widely in sunscreens and cosmetics to prevent skin damage from sunlight exposure. Washing off of BP4 from the human body in swimming pools represents a direct source of BP4 in the environment. In this study, we investigated the transformation of BP4 by free chlorine, chloramine, and UV/chlorine, which can be found in swimming pool water. It was found that BP4 can be rapidly removed by both chlorine and chloramine treatment, but only chlorination led to appreciable formation of chlorinated disinfection by-products (DBPs), such as dichloroacetic acid and trichloroacetic acid. However, chloramination of BP4 resulted in higher levels of chlorinated intermediates, which were relatively recalcitrant to further chlorination. The UV/chlorine process was significantly more efficient than the chlorine treatment alone for BP4 removal and resulted in trace chlorinated intermediates and formation of DBPs. Radical scavenger tests revealed that the removal of BP4 in the UV/chlorine process was mostly ascribed to reactions with the Cl% and %OH generated from HClO photolysis. The presence of methanol as a radical scavenger resulted in incomplete removal of BP4 in the UV/chlorine process and enhanced the formation of DBPs. The increase in DBP formation was because the residual BP4 can be exited to a triplet state (3BP4∗) upon UV irradiation. 3BP4∗ subsequently reacted with O2 to form 1O2, a reactive oxygen species that generated DBPs by reacting with chlorinated intermediates in water. All benzophenone-type UV-filters may have similar photochemical activity and thus influence the transformation of other organics in sunlit surface water environments.
⁎
Corresponding author. E-mail address: [email protected] (J. Lu).
https://doi.org/10.1016/j.cej.2018.06.089 Received 4 March 2018; Received in revised form 12 June 2018; Accepted 13 June 2018 1385-8947/ © 2018 Elsevier B.V. All rights reserved.
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
1. Introduction
[34]. The entire process was performed in a cooler at 4 °C. This could maximally inhibited the formation of di- and trichloramine [34,35]. The chloramine concentration was calculated assuming that free chlorine had been completely converted to NH2Cl. Swimming pool water was sampled from the Fitness Center of Nanjing City. The sample was filtered through a 0.45 μM membrane. The residual chlorine and DBPs were analyzed immediately after sampling.
Due to their widespread presence in the environment and the risks associated with their presence, organic UV-filters are considered to be emerging contaminants [1–6]. Benzophenone (BP) has two benzene rings jointed by a carbonyl group. Depending on the number and types of the functional groups attached to the backbone, a total of 12 substituted derivatives of BP are commercially used as UV-filters. Among them, BP3 (2-hydroxyl-4-methoxyphenone) and BP4 (2-hydroxy-4methoxybenzophenone-5-sulphonic acid) are the most widely known and have been approved for use in the EU, the U.S., China, Japan, and many other countries [7,8]. High levels of BP3 and BP4 have been detected in water used for recreational purposes [9–11]. Although BP3 and BP4 are relatively resistant to direct photodegradation and biodegradation, they are readily degraded by advanced oxidation, such as Fenton oxidation, TiO2 catalyzed oxidation, activated persulfate oxidation, and sonication [12–15]. BP3 and BP4 are also highly unstable in the presence of chlorine [16–23]. Under typical water chlorination conditions, the half-lives of BP3 and BP4 range from seconds to a few minutes [7,16,20,21]. Free chlorine reacts with BPs via the electrophilic substitution mechanism, resulting in the stepwise incorporation of Cl atoms into the methoxyphenol moiety [7,16,17]. Cleavage of the chlorinated ring due to further chlorination/ oxidation leads to the formation of a variety of chlorinated disinfection by-products (DBPs), such as chloroform and chloroacetic acids [16,17,20,22]. The reactions between chlorine and BPs have been well investigated. However, few studies have examined the reactions of BPs under swimming pool water conditions, where chlorine can be present in both free and combined forms. In addition, BPs in outdoor swimming pool water can be exposed to UV light. UV disinfection is also applied in swimming pool water treatment systems through which water is cycled and reused. Chlorine rapidly decomposes to Cl% and %OH upon UV irradiation [24–26]. As a highly potent oxidant, %OH reacts with a wide spectrum of organic compounds at nearly diffusion-controlled rates [27]. Cl% is also highly reactive to electron-rich compounds [24,28]. A synergistic effect between UV and chlorine was demonstrated during the degradation of contaminants, which were otherwise less reactive to UV or chlorine alone [29–31]. UV/chlorine has been developed as an advanced water treatment strategy to address micropollutants [24,28,32]. In the present study, we investigated the transformation of BP4 by chlorine, chloramine, and UV/chlorine, which are expected to be found in swimming pool water. We demonstrated that BP4 acted as a photosensitizer, leading to the increased formation of DBPs in the presence of UV irradiation. Such photochemical activity has not been previously reported. We assume that other BP-type UV-filters would have a similar property, which could affect the fate of organic compounds that are copresent in UV-based water treatment processes and sunlit environments.
2.2. Reaction kinetics Chlorination and chloramination of BP4 were performed in 40 mL borosilicate glass vials with polypropylene screw caps as batch reactors. The initial BP4 concentration was 1.0 mg/L. The reaction was initiated by rapid mixing of the BP4 solution with chlorine or chloramine. Phosphate buffer (10 mM) was added to maintain pH 7, and the molar ratio of chlorine or chloramine to BP4 was fixed at 10:1. The reactors were kept in the dark at 20 °C. The same conditions were applied for the UV/chlorine treatment except quartz vials were used as the reactors and placed in a BL-GHX-V photoreactor (Bilang, Shanghai, China) equipped with a low-pressure mercury lamp (ozone free, 15 W) emitting dominantly at 254 nm. The average UV fluence rate through the reaction volume was determined to be 4.68 × 10−7 Einstein∙L−1 s−1 using atrazine as an actinometer. The detailed experimental procedures can be found in a study by Canonica et al. [36]. A more detailed description of the apparatus used for the photo reaction can be found in Fig. S1 in the Supporting Data. Aliquots of the solution were withdrawn from each reactor and quenched immediately with 20 μL of 1 M Na2S2O3. To determine the contribution of the free radicals to BP4 removal in the UV/chlorine treatment, a concurrent experiment was conducted in the presence of 1 mM methanol. All samples were kept in a refrigerator at 4 °C until analysis for residual BP4.
2.3. Identification of the transformation products of BP4 Experiments were performed in 250 mL flasks in the dark. The initial concentration of BP4 was 1 mg/L and pH buffered at 7. The reaction was initiated by rapid mixing of a BP4 solution with an appropriate quantity of chlorine or chloramine stock solution. After 2 h of reaction, excess Na2SO3 was added to quench the residual chlorine or chloramine, and 2 mL of 98% sulfuric acid was added to reduce the pH to < 2. Then, solid phase extraction (SPE) using HyperSep C18 cartridges (Thermo Scientific, Waltham, MA) was performed to purify and concentrate the transformation products. Additional details of the procedure are described in the Supporting Information. For UV/ chlorine, the reaction was performed in quartz vials in a photoreactor. After 2 h, the solution was taken out and quenched with Na2SO3 before being extracted by SPE as described above. The transformation products produced by UV/chlorine in the presence of 1 mM methanol were also collected for comparison. The SPE eluents were analyzed using an Agilent 1200 HPLC coupled with an Agilent G6410B Triple Quad Mass spectrometer with an electron spray ionization (ESI) source operated in the negative mode. The detailed instrumental setup is provided in the Supporting Information.
2. Materials and methods 2.1. Materials A BP4, 3,4,5-trimethylphenol (TMP), furfuryl alcohol (FFA), and haloacetic acid calibration mixture was obtained from Sigma-Aldrich (St. Louis, MO). Sodium hypochloride (NaClO) and N,N-dimethyl-1,4phenylenediamine monohydrochloride (DPD) were purchased from Aladdin (Shanghai, China). Stock solutions of BP4 and NaClO were prepared by dissolving the reagents in Milli-Q water and storing them in a refrigerator. Prior to use, the free chlorine concentration of the NaClO stock solution was measured using the DPD method [33]. The chloramine (NH2Cl) solution was freshly prepared by adding a NaClO stock solution dropwise to the NH4Cl solution to achieve a final NH4Cl/ NaClO molar ratio of 1.2:1 following Schreiber and Mitch’s method
2.4. Formation of chloroacetic acids (CAAs) The formation of CAAs in the chlorination, chloramination, and UV/ chlorine processes was investigated in a series of vials. The initial BP4 concentration was 1 mg/L, and the chlorine/BP4 M ratio was 10. The other conditions were identical to those of the kinetic study. At predetermined time points, a vial was sacrificed by adding 1 mM Na2SO3 (1 mM) to quench the reaction. The quenched reaction solutions were kept in a refrigerator until further treatment and analysis for CAAs. 305
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
2.5. BP4 as a photosensitizer The BP4 solution was pretreated with free chlorine for 30 min and then UV for 15 min. The initial BP4 concentration was 1.0 mg/L, and the HClO/BP4 molar ratio was 5. The pH of the solution was buffered at 7.0. Preliminary experiments demonstrated that no residual BP4 or free chlorine was detected after the chlorination and UV pretreatment. Then, additional BP4 was spiked into the solutions before they were subjected to UV irradiation for an additional 8 h. After the reaction, CAAs were extracted and analyzed. To ascertain the reactive species responsible for the formation of CAAs, 1 mM TMP or FFA was spiked together with BP4 in the middle of the treatment. The other treatments and conditions were identical. The same experiments were also performed on the sample of swimming pool water. The water sample was pretreated with UV for 15 min to decompose the residual chlorine before varying concentrations of BP4 were spiked. Then, the samples were treated with UV for an additional 8 h, and the formation of CAAs was analyzed. Controls spiked with the same quantities of BP4 but without the additional UV treatment were run concurrently. 2.6. Analytical methods The residual BP4 in the samples was analyzed by a Hitachi L-2000 high performance liquid chromatographer (HPLC) with an L-2455 diode array detector (DAD). Separation was performed on a LaChrom C18 reverse phase column (5 μm × 4.6 mm × 250 mm) with an isocratic eluent that consisted of 30% methanol (with 0.1% acetic acid) and 70% water (with 0.1% acetic acid) at a flow rate of 1 mL/min. The injection volume was 10 μL, and quantification was based on the absorbance at 285 nm, which is characteristic of BP4. HAAs were analyzed using the USEPA standard method 552.2. The procedure included liquid-liquid extraction followed by methylation and quantification by an Agilent 7890 gas chromatograph (GC) equipped with an electron capture detector (ECD) and a DB5-fused silica capillary column (30 m × 0.53 mm × 1.5 μm). Additional details can be found in the Supporting Information. The free chlorine in the solution was quantified by the DPD method [33]. An aliquot of 1 mL of the solution was sampled from the reactor at a preset reaction time and immediately mixed with 1 mL of DPD solution (8 mM). The free chlorine concentration was quantified according to the absorbance at 510 nm measured by a Varian Cary 50 spectrophotometer. The standard deviation of this approach was within 5%.
Fig. 1. (a) Removal of BP4 during the chlorination, chloramination, UV/ chlorine, and UV/chlorine/methanol processes. Initial BP4 concentration, 1 mg/L; chlorine or chloramine to BP4 molar ratio, 10; pH 7; temperature 20 °C. (b) Removal of BP4 in UV/chlorine/methanol process at varying chlorine doses. Initial BP4 concentration, 1 mg/L; pH 7; methanol, 1 mM; temperature, 20 °C. The points are experimental data, and he dashed lines represent the fitting of a first-order kinetic model. Error bars represent the standard deviation of 3 replicates.
Thus, the removal of BP4 in the UV/chlorine process can be attributed to reactions with HO%, Cl%, and free chlorine. To determine the contribution of the radicals, the reaction was conducted in the presence of 1 mM methanol, which acted as the radical scavenger. The removal rate was greatly reduced under these conditions. Only a 15% attenuation was observed within 15 s at the initial HClO/BP4 molar ratio of 10, in contrast to the 100% removal in the absence of methanol. Thus, it can be concluded that the radicals (HO% and Cl%) that formed upon free chlorine photolysis contributed to the majority of BP4 removal in the UV/chlorine process.
3. Results and discussion 3.1. Reaction kinetics As shown in Fig. 1a, BP4 was rapidly removed during chlorination. At an HClO/BP4 molar ratio of 10, an approximate 90% removal was achieved in 3 min. In accordance with previous studies [16,17,19–21], BP4 followed pseudo first-order decay with a rate constant of 0.012 s−1. The transformation of BP4 by chloramination was slower than that of chlorination. When the free chlorine was replaced with an equal quantity of chloramine, less than 40% removal was found in 5 min. The reaction could also be modeled by pseudo first-order kinetics
(ln
hv
HCIO→C1% + HO%
Free chlorine also contributed to the small fraction of BP4 removal in the UV/chlorine process, as evidenced by the reduction in the presence of methanol (Fig. 1b). Overall, in the presence of 1 mM methanol, an approximately 35% removal of BP4 was achieved in the UV/chlorine process, but most of the removal occurred during the first minute. In this period, the reaction was slower than that in the chlorination process, but faster than that of chloramination. After that, the reaction slowed down. Almost no degradation of BP4 was observed after 5 min (Fig. 1b). The termination of BP4 removal was coincident with the depletion of free chlorine. As shown in Fig. S2 (Supporting Data), in the UV/chlorine process, HClO was completely degraded in 10 min.
)
= −kt , and the rate constant (k) was 0.002 s−1. UV alone did not cause an appreciable attenuation of BP4 during the investigated time period. However, the removal of BP4 was significantly accelerated in the UV/chlorine process. With an initial HClO/BP4 M ratio of 10 in the presence of UV irradiation, complete removal was found to occur within 15 s, indicating a synergistic effect between chlorine and UV [29–31]. In the UV/chlorine process, HClO undergoes photolysis to form HO% and Cl% (R1) [24,25]. Both radical species can react with aromatics with second-order rate constants ranging from 108–109 M−1 s−1 [27]. c co
(R1)
306
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
electron-rich, thus making Cl-BP4ox, if it was a hydroxylated Cl-BP4, more susceptible to a free chlorine attack. Therefore, we presume that the oxidation occurred on the carbonyl group. It has been reported that chlorinated BP4 can undergo a Baeyer-Villiger type oxidative rearrangement during chlorination, converting diphenylketone to phenyl ester [16,23]. The findings of this study suggest that Cl-BP4ox and Cl2BP4ox were formed via such a mechanism, with Cl-BP4ox undergoing further chlorination to form Cl2-BP4ox. The proposed molecular structures of Cl-BP4ox and Cl2-BP4ox are shown in Table S1. In addition, as shown in Fig. 2b, a product with m/z 295/297/299 was also formed, although its occurrence was less significant. Based on its isotope signature, we infer that this product contains two chlorine atoms. This compound, which could have been generated through desulfonation followed by chlorination, has also been observed in previous studies [16,23]. Although this product appeared to be less significant in MS than other products, its formation should not be overlooked because deprotonation of the molecule becomes more difficult following cleavage of the acidic sulfonate group, which causes a less efficient ionization in the negative ESI mode. Overall, based on the molecular structures of the chlorinated products, the proposed transformation pathways for BP4 chlorination are depicted in Fig. 3. Compared to chlorination, chloramination of BP4 led to the formation of the same suite of transformation products, as illustrated in Fig. 2(c). Interestingly, the evolution of these products as a function of the chlorine and chloramine dosages in both the chlorination and chloramination processes showed similar trends (Supporting Data, Fig. S3). Nevertheless, under the same reaction conditions, the yields of the products from chloramination were always much greater than those from chlorination. Based on these findings, it can be assumed that it was free chlorine that reacted with BP4 during chloramination. Chloramine is formed by the combination of free chlorine and NH4+. The reaction is reversible, thus, while HClO can be slowly released from chloramine, its equilibrium concentration would be extremely low. Thus, the transformation of BP4 was significantly slow and the chlorinated BP4 intermediates could be readily accumulated in the chloramination process. The chlorinated products of BP4 identified in the chlorination and chloramination processes were not found in the UV/chlorine treatment (data not shown). HO% and Cl% that were generated during photolysis of HClO acted as the main reactive species for the BP4 transformation in the UV/chlorine process, as evidenced by the suppressed removal when methanol was used as the radical scavenger (Fig. 1). However, Cl% reacts with organic compounds via an electron transfer mechanism rather than addition or substitution [38], which is unlikely to lead to chlorinated products. Fig. 2(d) also shows that Cl-BP4 and Cl2-BP4ox were
Fig. 2. MS of possible BP4 transformation products generated with a negative ESI. (a) Control; (b) chlorination; (c) chloramination; (d) UV/chlorine/methanol. The initial BP4 concentration was 1 mg/L; HClO or NH2Cl was 10 times (molar ratio) that of the substrate; methanol 1 mM. The reaction time was 2 h for chlorination and chloramination and 20 min for UV/chlorination/methanol; pH 7.
3.2. Identification of reaction products MS was employed to analyze the transformation products of BP4 after the chlorine, chloramine, and UV/chlorine treatments. The data are shown in Fig. 2. Compared to the control sample shown in Fig. 2a, b reveals the disappearance of the signal for BP4 (m/z 307) as well as the appearance of new peaks at m/z 295/297/299, m/z 341/343, m/z 357/ 359, m/z 375/377/379, and m/z 391/393/395. The compound corresponding to m/z 341/343 has a characteristic M + 2 isotopic peak, with an abundance of approximately 1/3 of that of M, indicating that it contains a chlorine atom. In addition, the molecular weight (MW) of this compound is found to be 34 Daltons greater than that of BP4, allowing us to infer that this compound is a mono-chlorine substituted BP4 (Cl-BP4). The compound corresponding to m/z 375/377/379 is 68 Daltons greater than that of BP4, with a characteristic M:(M + 2): (M + 4) ratio of 9:6:1, suggesting that this compound is a derivative of BP4 with two substituted chlorine atoms (Cl2-BP4). Chlorination of phenolic compounds involves electrophilic substitution at the para and ortho positions of the hydroxyl group on the phenolic ring [37]. However, the para site of the BP4 molecule is occupied by a sulfonic group, and one of the ortho sites is connected to the carbonyl group. Thus, we infer that Cl-BP4 is 2-hydroxyl-3-chloro-4-methoxyl benzophenonesulfonic acid. After the para and ortho positions are substituted, the second chlorine is directed to the unoccupied meta site. Therefore, Cl2BP4 is presumed to be 2-hydroxyl-3,6-dichloro-4-methoxyl benzophenone-sulfonic acid. The proposed molecular structures of Cl-BP4 and Cl2-BP4 are shown in Table S1. The formation of chlorine substituted BP4 increased initially and then decreased with the increasing chlorine dose (Supporting Data, Fig. S3). Cl-BP4 reached a maximum concentration at an HClO/BP4 molar ratio of 1, while Cl2-BP4 reached a maximum concentration at an HClO/BP4 molar ratio of 4, suggesting that the transformation of BP4 proceeded from the monochlorinated to the dichlorinated derivative. According to the isotope signature (M:(M + 2) = 3:1) shown in Fig. 2(b), the compound of m/z 357/359 also contains a chlorine atom. Because its MW is 16 Daltons greater than that of Cl-BP4, it can be assumed that this compound is an oxidation product of Cl-BP4 (denoted as Cl-BP4ox). Similarly, the compound with m/z 391/393/395 corresponds to the oxidation product of Cl2-BP4 (denoted as Cl2-BP4ox). ClBP4ox and Cl2-BP4ox appeared to be relatively stable towards free chlorine. Their formation, especially Cl2-BP4ox, increased with the increasing chlorine dosage (Supporting Data, Fig. S4). This stability suggests that it is unlikely that Cl-BP4ox was formed through the hydroxylation of Cl-BP4 because, as an electron-donating functional group, hydroxyl substitution would make the aromatic ring more
Fig. 3. Proposed transformation pathways of BP4 during the chlorination and chloramination processes. 307
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
desulfonicated form. The presence of an electron-withdrawing sulfonic group that decreased the electron density of the phenolic ring made it less vulnerable to chlorine attack and was disadvantageous to the formation of DBPs. Less DBP formation was expected in chloramination than chlorination. Substitution of chlorine with chloramine as the disinfectant is practiced in water treatment to reduce the formation of DBPs. It should be noted, however, that although the formation of DBPs was greatly reduced, chloramination resulted in more chlorinated BP4 derivatives. The potential risks of these chlorinated products to the environment and human health warrant further study. As previously discussed, in the UV/chlorine process, instead of HClO, HO% and Cl% were the main reactive species. These species reacted with organic compounds via one-electron oxidation, which is unlikely to generate chlorinated DBPs[38]. In addition, DBPs, if formed, would be degraded by UV irradiation (Supporting Data, Fig. S5). It is well recognized that given the same chlorine dose, the UV/chlorine treatment generates less halogenation DBPs than chlorination [24,25]. An interesting finding is that substantial CAAs were formed in the UV/chlorine process in the presence of methanol. As shown in Fig. 4(a), this formation was remarkably higher than that of the chlorination process. For instance, at 4 h, the yields of TCAA reached 1.49 and 7.54 μg/L in the chlorination and UV/chlorine/methanol processes, respectively. The yields of DCAA were 0.60 and 4.36 μg/L, respectively. Considering the fact that a significant fraction of BP4 (approximately 65% at this condition) still remained in the solution treated with UV/ chlorine/methanol (Fig. 1(b)), the formation of CAAs in terms of the BP4 molar yield was even higher. This high formation of CAAs in the UV/chlorine/methanol process was unexpected. Since the radicals were mostly scavenged by methanol, the removal of BP4 under these conditions was mainly due to its reaction with free chlorine. However, the majority of the CAAs were formed after 10 min when free chlorine had been completely decomposed (Supporting Data, Fig. S2), and further transformation of BP4 was not observed (Fig. 1(b)). Thus, the CAAs generated after 10 min are presumed to have resulted from the decomposition of the chlorinated BP4 intermediates that formed before the depletion of free chlorine. UV radiation was found to be critical during the formation of HAAs in this scenario. Fig. 4(b) shows CAA formation in the same treatment (UV/chlorine/methanol) but UV was terminated at 30 min. It is observed that the CAA concentrations did not change appreciably after the UV source was turned off. Therefore, it is evident that the majority of CAAs that formed under these conditions were ascribed to reactions driven by UV.
Fig. 4. (a) Formation of CAAs during the chlorination and UV/chlorination/ methanol treatment of BP4. (b) Formation of chlorinated acetic acids in the UV/ chlorination/methanol treatment of BP4 for 30 min followed by a dark incubation. Initial BP4 concentration, 1 mg/L; chlorine or chloramine to BP4 molar ratio, 10; methanol, 1 mM; pH 7; temperature, 20 °C. The error bars represent the standard deviation of 3 replicates.
formed when methanol was present, reinforcing the previous argument that free chlorine contributed to some BP4 removal in the UV/chlorine process, at least during the initial stage of the reaction when HClO was not completely decomposed. However, this contribution might be small compared to that of the radicals. It should be noted that these chlorinated products were not detected in UV/chlorine in the absence of methanol, probably because they were rapidly degraded by free radicals.
3.4. Mechanisms of the formation of CAAs in the UV/chlorine/methanol process Another interesting finding is that besides UV, the residual BP4 in the reaction solution also played a critical role in the formation of CAAs in the UV/chlorine/methanol process. As shown in Fig. 5(a), if the solutions were subsequently kept in the dark, adding additional BP4 to the BP4 solution pretreated with chlorine (no residual BP4 or free chlorine found in the solution when additional BP4 was spiked), did not result in a noticeable influence on the formation of CAAs. When the solutions were incubated under UV subsquently, however, more CAAs were generated. A positive relationship is observed between the formation of CAAs and the quantity of BP4 spiked in the middle of the treatment. Hence, it can be inferred that, in the UV/chlorine/methanol process, the residual BP4 facilitated the UV-driven formation of CAAs. A rational mechanism of CAA formation in the UV/chlorine/methanol process is BP4-sensitized indirect photodegradation of the intermediates resulting from BP4 chlorination. As a UV-filter, BP4 was designed to absorb UV light but be resistant to direct photolysis [39]. To avoid degradation, a photoexcited BP4 molecule can release the energy from a photon in the form of heat or convey the energy to other molecules, i.e., it can act as a photosensitizer. In fact, benzophenone has been used as a photosensitizer to initiate a radical polymerization
3.3. Formation of chloroacetic acids(CAAs) It has been reported that chlorination of BP4 leads to the formation of chlorinated DBPs [16,22]. In this study, only the formation of semivolatile CAAs was quantified (Fig. 4). Since the vials in the photoreactor could not be sealed tightly and head-space remained during the reaction, the volatile chloroform was not analyzed. It was revealed that CAAs were formed in the chlorination process, while almost no CAAs were detected in the chloramination or UV/chlorine treatment (data not shown). In the chlorination process, the yield of trichloroacetic acid (TCAA) was higher than that of dichloroacetic acid (DCAA) and monochloroacetic acid (MCAA) was under the detection limit. Since ClBP4ox and Cl2-BP4ox were relatively stable toward chlorine and chloramine (Supporting Data, Fig. S4), we presumed that CAAs formed as the result of the further chlorination of chlorinated BP4, especially the 308
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
Fig. 5. Influence of the presence of BP4 on the formation of chlorinated acetic acids during the UV treatment process. (a) BP4 was spiked into pre-chlorinated BP4 samples. Initial BP4 concentration, 1 mg/L; chlorine/BP4 molar ratio, 5; pH 7. (b) BP4 was spiked into a swimming pool water sample. No residual free chlorine was presented at the time of additional BP4 spiking. Fig. 6. Effects of TMP and FFA on the formation of (a) DCAA and (b) TCAA due to BP4-mediated indirect photo degradation; 1 mM TMP or FFA was spiked together with BP4 in pre-chlorinated BP4 samples when no free chlorine was left. The sample was subsequently subjected to UV irradiation for 8 h.
reactions in polymer synthesis[40–42]. In environmental chemistry studies, benzophenones are well-defined proxies to assess the photochemical activity of dissolved natural organic matter (NOM). It has been recognized that benzophenone moieties are important chromophores in NOM molecules that can be excited to form a triplet state (3NOM∗), which reacts with dissolved oxygen to generate an oxygen singlet (1O2) in a sunlit environment. Both 3NOM∗ and 1O2 are highly reactive species that can cause the transformation of organic contaminants that are otherwise insensitive to direct photolysis [43–47]. Similarly, upon UV irradiation of (R2), we presume that the residual BP4 in the UV/chlorine/methanol process can be excited to form triplet state BP4 (3BP4∗). 3BP4∗ then further reacts with dissolved oxygen to give rise to 1O2 (R3). The formation of the CAAs was most likely the result of the degradation of chlorinated BP4 intermediates (Cl-BP4, Cl2BP4ox, etc., Fig. 2(d)) by 3BP4∗ and/or an 1O2 attack; i.e., the chlorinated BP4 intermediates underwent indirect photo degradation to break down and release CAAs. To verify this hypothesis, the above experiment was performed in the presence of TMP, which is able to selectively scavenge 3BP4∗ [47]. TMP was spiked along with the additional BP4 in the middle of the treatment process when both free chlorine and BP4 had been depleted. The data shown in Fig. 6 reveal that the formation of CAAs slightly increased when TMP was spiked together with the additional BP4. A comparison of the formation of CAAs in the presence and absence of TMP clearly indicates that 3BP4∗ was involved in the formation of CAAs in the UV/chlorine/methanol
process. Fig. 6 also provides data on the formation of CAAs in the presence of FFA instead of TMP, showing that the addition of FFA also significantly inhibited the formation of CAAs. The yields were at the same levels as those in the presence of TMP. Since FFA selectively quenches 1O2 [48,49], which could only have been generated as the result of the reaction between 3BP4∗ and dissolved oxygen, these data indicate that after the depletion of free chlorine in the UV/chlorine/ methanol process, 1O2 directly reacted with chlorinated BP4 to form CAAs (R4). hv
BP4 → 3BP4∗
(R2)
3BP4∗
+ O2 → BP4 + 1O2
(R3)
+ CIBP4 → →CAAs
(R4)
1O 2
As a photo-sensitizer, we hypothesize that under UV irradiation, BP4 can not only mediate the degradation of chlorinated BP4 intermediates to form DBPs but also that intermediates are derived from reactions between free chlorine and other organic impurities in water. To test this hypothesis, a sample of swimming pool water was collected. Preliminary experiments indicated that there was 1.28 mg/L residual 309
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
chlorine but no BP4 in the water. The sample was subjected to UV irradiation for 15 min to remove the residual chlorine. Then, varying amounts of BP4 were spiked and treated with UV for an additional 8 h. No appreciable degradation of the spiked BP4 was observed after UV treatment. However, as illustrated in Fig. 5(b), a positive relationship was evident between the final concentrations of CAAs and the quantity of spiked BP4. The concentrations of DCAA and TCAA in the sample were 23.51 and 13.75 μg/L, respectively, without the addition of BP4. The formation increased to 79.84 and 92.37 μg/L, respectively, when 1 mg/L of BP4 was spiked. We hypothesize that the increased formation of CAAs was due to the degradation of chlorinated organic matter that was originally in the swimming pool water caused by photo-generated 1 O2. During this process, BP4 acted as the photo-sensitizer. By contrast, when the samples were incubated in the dark, the presence of BP4 did not have an appreciable influence on the concentrations of CAAs (Fig. 5(b)). The concentrations of CAAs were close to their original values, approximately 64.8 and 36.5 μg/L, respectively, for DCAA and TCAA. It should be noted that the concentrations were higher than those in the sample subjected to an additional UV treatment in the absence of BP4 because CAAs were removed by direct photo-degradation (Supporting Data, Fig. S5). The data shown in Figs. 5 and 6 clearly demonstrate that the co-presence of BP4 and UV can significantly increase the formation of DBPs in chlorinated water.
[2] I. Ozaez, J. Luis Martinez-Guitarte, G. Morcillo, Effects of in vivo exposure to UV filters (4-MBC, OMC, BP-3, 4-HB, OC, OD-PABA) on endocrine signaling genes in the insect Chironomus riparius, Sci. Total Environ. 456 (2013) 120–126. [3] C. Liao, K. Kannan, Widespread occurrence of benzophenone-type UV light filters in personal care products from china and the united states: an assessment of human exposure, Environ. Sci. Technol. 48 (2014) 4103–4109. [4] P.Y. Kunz, K. Fent, Estrogenic activity of ternary UV filter mixtures in fish (Pimephales promelas) - An analysis with nonlinear isobolograms, Toxicol. Appl. Pharmacol. 234 (2009) 77–88. [5] M. Krause, A. Klit, M.B. Jensen, T. Soeborg, H. Frederiksen, M. Schlumpf, W. Lichtensteiger, N.E. Skakkebaek, K.T. Drzewiecki, Sunscreens: are they beneficial for health? An overview of endocrine disrupting properties of UV-filters, Int. J. Androl. 35 (2012) 424–436. [6] D.L. Giokas, A. Salvador, A. Chisvert, UV filters: from sunscreens to human body and the environment, Trac-Trends Anal. Chem. 26 (2007) 360–374. [7] A.J.M. Santos, M.S. Miranda, J.C.G. Esteves da Silva, The degradation products of UV filters in aqueous and chlorinated aqueous solutions, Water Res. 46 (2012) 3167–3176. [8] N.A. Shaath, Ultraviolet filters, Photochem. Photobiol. Sci. 9 (2010) 464–469. [9] M.M.P. Tsui, H.W. Leung, T.-C. Wai, N. Yamashita, S. Taniyasu, W. Liu, P.K.S. Lam, M.B. Murphy, Occurrence, distribution and ecological risk assessment of multiple classes of UV filters in surface waters from different countries, Water Res. 67 (2014) 55–65. [10] D.A. Lambropoulou, D.L. Giokas, V.A. Sakkas, T.A. Albanis, M.I. Karayannis, Gas chromatographic determination of 2-hydroxy-4-methoxybenzophenone and octyldimethyl-p-aminobenzoic acid sunscreen agents in swimming pool and bathing waters by solid-phase microextraction, J. Chromatogr. A 967 (2002) 243–253. [11] Y. Tashiro, Y. Kameda, Concentration of organic sun-blocking agents in seawater of beaches and coral reefs of Okinawa Island, Japan, Mar. Pollut. Bull. 77 (2013) 333–340. [12] H. Zuniga-Benitez, J. Soltan, G.A. Penuela, Application of ultrasound for degradation of benzophenone-3 in aqueous solutions, Int. J. Environ. Sci. Technol. 13 (2016) 77–86. [13] H. Zuniga-Benitez, C. Aristizabal-Ciro, G.A. Penuela, Photodegradation of the endocrine-disrupting chemicals benzophenone-3 and methylparaben using Fenton reagent: optimization of factors and mineralization/biodegradability studies, J. Taiwan Inst. Chem. Eng. 59 (2016) 380–388. [14] X. Pan, L. Yan, R. Qu, Z. Wang, Degradation of the UV-filter benzophenone-3 in aqueous solution using persulfate activated by heat, metal ions and light, Chemosphere 196 (2018) 95–104. [15] H. Zuniga-Benitez, C. Aristizabal-Ciro, G.A. Penuela, Heterogeneous photocatalytic degradation of the endocrine-disrupting chemical Benzophenone-3: parameters optimization and by-products identification, J. Environ. Manage. 167 (2016) 246–258. [16] M. Xiao, D. Wei, J. Yin, G. Wei, Y. Du, Transformation mechanism of benzophenone-4 in free chlorine promoted chlorination disinfection, Water Res. 47 (2013) 6223–6233. [17] S. Zhang, X. Wang, H. Yang, Y.F. Xie, Chlorination of oxybenzone: kinetics, transformation, disinfection byproducts formation, and genotoxicity changes, Chemosphere 154 (2016) 521–527. [18] T. Manasfi, V. Storck, S. Ravier, C. Demelas, B. Coulomb, J.-L. Boudenne, Degradation Products of Benzophenone-3 in Chlorinated Seawater Swimming Pools, Environ. Sci. Technol. 49 (2015) 9308–9316. [19] P. Abdallah, M. Deborde, F.D. Berne, N.K.V. Leitner, Kinetics of Chlorination of Benzophenone-3 in the Presence of Bromide and Ammonia, Environ. Sci. Technol. 49 (2015) 14359–14367. [20] S.E. Duirk, D.R. Bridenstine, D.C. Leslie, Reaction of benzophenone UV filters in the presence of aqueous chlorine: kinetics and chloroform formation, Water Res. 47 (2013) 579–587. [21] J. Li, L.-Y. Ma, L. Xu, Transformation of benzophenone-type UV filters by chlorine: kinetics, products identification and toxicity assessments, J. Hazard. Mater. 311 (2016) 263–272. [22] N. Negreira, P. Canosa, I. Rodriguez, M. Ramil, E. Rubi, R. Cela, Study of some UV filters stability in chlorinated water and identification of halogenated by-products by gas chromatography-mass spectrometry, J. Chromatogr. A 1178 (2008) 206–214. [23] N. Negreira, I. Rodriguez, R. Rodil, R. Cela, Assessment of benzophenone-4 reactivity with free chlorine by liquid chromatography quadrupole time-of-flight mass spectrometry, Anal. Chim. Acta 743 (2012) 101–110. [24] J.Y. Fang, Y. Fu, C. Shang, The Roles of Reactive Species in Micropollutant Degradation in the UV/Free Chlorine System, Environ. Sci. Technol. 48 (2014) 1859–1868. [25] T. Li, Y. Jiang, X. An, H. Liu, C. Hu, J. Qu, Transformation of humic acid and halogenated byproduct formation in UV-chlorine processes, Water Res. 102 (2016) 421–427. [26] Y.G. Feng, D.W. Smith, J.R. Bolton, Photolysis of aqueous free chlorine species (NOCI and OCI-) with 254 nm ultraviolet light, J. Environ. Eng. Sci. 6 (2007) 277–284. [27] P. Neta, R. Huie, A. Ross, Rate constants for reactions of inorganic radicals in aqueous solution, J. Phys. Chem. Ref. Data 17 (1988) 1027–1284. [28] X. Yang, J. Sun, W. Fu, C. Shang, Y. Li, Y. Chen, W. Gan, J. Fang, PPCP degradation by UV/chlorine treatment and its impact on DBP formation potential in real waters, Water Res. 98 (2016) 309–318. [29] W.-L. Wang, Q.-Y. Wu, N. Huang, T. Wang, H.-Y. Hu, Synergistic effect between UV and chlorine (UV/chlorine) on the degradation of carbamazepine: influence factors and radical species, Water Res. 98 (2016) 190–198.
4. Conclusions This study demonstrates that BP4 can be removed by chlorine, chloramine, and UV/chlorine treatment. Chlorination was more efficient than chloramine but led to the formation of DBPs, such as CAAs. The degradation of BP4 by UV/chlorine treatment was more efficient than that by chlorine, and the removal was mostly due to reactions with free radicals (HO% and Cl%) that were generated upon photolysis of free chlorine, which resulted in a low formation of DBPs. Adding methanol as the radical quenching agent to the UV/chlorine process caused incomplete removal of BP4, but remarkably increased the formation of CAAs. This enhanced formation of CAAs was due to the photochemical activity of the residual BP4, which, under UV irradiation, was excited to form 3BP4∗. 3BP4∗ reacted with dissolved oxygen to form 1O2, which reacted with the chlorinated BP4 intermediates to form CAAs. We believe that due to their structural similarity, other BP type UV-filters have similar photochemical activities and are thus likely to affect the transformation of other co-existing organic compounds. This photochemical activity of BP type UV-filters should be taken into consideration in studies of their behavior in natural and engineering systems. Acknowledgements This research was supported by the National Natural Science Foundation of China (51578294) and the Priority Academic Program Development (PAPD) of the Jiangsu Higher Education Institute. The authors would like to give their gratitude to Randall David Pierce at the University of Georgia (UGA) for the assistance in polishing the language. J. Lu would like to acknowledge the China Scholar Council for supporting his research at UGA. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cej.2018.06.089. References [1] M.E. Balmer, H.R. Buser, M.D. Muller, T. Poiger, Occurrence of some organic UV filters in wastewater, in surface waters, and in fish from Swiss lakes, Environ. Sci. Technol. 39 (2005) 953–962.
310
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
[40] A.G. Krainev, R.I. Viner, D.J. Bigelow, Benzophenone-sensitized photooxidation of sarcoplasmic reticulum membranes: site-specific modification of the Ca2+-ATPase, Free Radical Biol. Med. 23 (1997) 1009–1020. [41] S. Riyajan, J.T. Sakdapipanich, Y. Tanaka, Controlled degradation of cured natural rubber by encapsulated benzophenone as a photosensitizer, J. Appl. Polym. Sci. 90 (2003) 297–305. [42] K.-I. Fukuka, M. Ueda, Recent progress of photosensitive polyimides, Polym. J. 40 (2008) 281–296. [43] S. Canonica, B. Hellrung, J. Wirz, Oxidation of phenols by triplet aromatic ketones in aqueous solution, J. Phys. Chem. A 104 (2000) 1226–1232. [44] K. McNeill, S. Canonica, Triplet state dissolved organic matter in aquatic photochemistry: reaction mechanisms, substrate scope, and photophysical properties, Environ. Sci. Processes Impacts 18 (2016) 1381–1399. [45] C. Yuan, M. Chakraborty, S. Canonica, L.K. Weavers, C.M. Hadad, Y.-P. Chin, Isoproturon reappearance after photosensitized degradation in the presence of triplet ketones or fulvic acids, Environ. Sci. Technol. (2016). [46] S. Canonica, B. Hellrung, P. Muller, J. Wirz, Aqueous oxidation of phenylurea herbicides by triplet aromatic ketones, Environ. Sci. Technol. 40 (2006) 6636–6641. [47] K. Golanoski, S. Fang, R.D. Vecchio, N.V. Blough, Investigating the mechanism of phenol photooxidation by humic substances, Environ. Sci. Technol. 46 (2012) 3912–3920. [48] W.R. Haag, J. Hoigne, Singlet oxygen in surface waters.3. Photochemical formation and steady-state concentration in various types of waters, Environ. Sci. Technol. 20 (1986) 341–348. [49] M. Jiang, J. Lu, Y. Ji, D. Kong, Bicarbonate-activated persulfate oxidation of acetaminophen, Water Res. 116 (2017) 324–331.
[30] X.R. Zhang, W.G. Li, P.F. Ren, X.J. Wang, Chlorine/UV induced photochemical degradation of total ammonia nitrogen (TAN) and process optimization, RSC Adv. 5 (2015) 63793–63799. [31] X.J. Kong, J. Jiang, J. Ma, Y. Yang, W.L. Liu, Y.L. Liu, Degradation of atrazine by UV/chlorine: efficiency, influencing factors, and products, Water Res. 90 (2016) 15–23. [32] M.J. Watts, E.J. Rosenfeldt, K.G. Linden, Comparative OH radical oxidation using UV-Cl-2 and UV-H2O2 processes, J. Water Supply Res. Technol. AQUA 56 (2007) 469–477. [33] APHA/AWWA/WEF, Standard Methods for the Examination of Water and Wastewater, 21st ed., EPS Group, Inc., 2005. [34] I.M. Schreiber, W.A. Mitch, Influence of the Order of Reagent Addition on NDMA Formation during Chloramination, Environ. Sci. Technol. 39 (2005) 3811–3818. [35] M. Nihemaiti, J. Le Roux, C. Hoppe-Jones, D.A. Reckhow, J.-P. Croué, Formation of haloacetonitriles, haloacetamides, and nitrogenous heterocyclic byproducts by chloramination of phenolic compounds, Environ. Sci. Technol. 51 (2017) 655–663. [36] S. Canonica, L. Meunier, U. von Gunten, Phototransformation of selected pharmaceuticals during UV treatment of drinking water, 42 (2008) 121–128. [37] J. Lu, M.M. Benjamin, G.V. Korshin, H. Gallard, Reactions of the flavonoid hesperetin with chlorine: a spectroscopic study of the reaction pathways, Environ. Sci. Technol. 38 (2004) 4603–4611. [38] K.M. Parker, W.A. Mitch, Halogen radicals contribute to photooxidation in coastal and estuarine waters, PNAS 113 (2016) 5868–5873. [39] E. De Laurentiis, M. Minella, M. Sarakha, A. Marrese, C. Minero, G. Mailhot, M. Brigante, D. Vione, Photochemical processes involving the UV absorber benzophenone-4 (2-hydroxy-4-methoxybenzophenone-5-sulphonic acid) in aqueous solution: reaction pathways and implications for surface waters, Water Res. 47 (2013) 5943–5953.
311
Contents lists available at ScienceDirect
Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej
Enhanced formation of chlorinated disinfection byproducts in the UV/ chlorine process in the presence of benzophenone-4
T
⁎
Junhe Lua, , Peizeng Yanga, Wei Donga, Yuefei Jia, Qingguo Huangb a b
Department of Environmental Science and Engineering, Nanjing Agricultural University, Nanjing 210095, China Interdisciplinary Toxicology Program, Department of Crop and Soil Sciences, University of Georgia, Griffin, GA 30223, USA
H I GH L IG H T S
was efficiently removed • BP4 chlorine, chloramine, and UV/
G R A P H I C A L A B S T R A C T
by
chlorine treatment.
of BP4 by UV/chlorine was • Removal mostly arose from the reactions with % OH and Cl%.
in the UV/chlorine process • Methanol suppressed BP4 removal but enhanced DBPs formation.
BP4 and O could be formed be• cause of the photochemical activity of ∗
3
1
2
BP4.
responsible for DBP formation • inOthewasUV/chlorine/methanol process. 1
2
A R T I C LE I N FO
A B S T R A C T
Keywords: Benzophenone-4 Chlorine UV/chlorine Disinfection byproducts Photosensitizer
Benzophenone-4 (BP4) is UV-filter that is widely in sunscreens and cosmetics to prevent skin damage from sunlight exposure. Washing off of BP4 from the human body in swimming pools represents a direct source of BP4 in the environment. In this study, we investigated the transformation of BP4 by free chlorine, chloramine, and UV/chlorine, which can be found in swimming pool water. It was found that BP4 can be rapidly removed by both chlorine and chloramine treatment, but only chlorination led to appreciable formation of chlorinated disinfection by-products (DBPs), such as dichloroacetic acid and trichloroacetic acid. However, chloramination of BP4 resulted in higher levels of chlorinated intermediates, which were relatively recalcitrant to further chlorination. The UV/chlorine process was significantly more efficient than the chlorine treatment alone for BP4 removal and resulted in trace chlorinated intermediates and formation of DBPs. Radical scavenger tests revealed that the removal of BP4 in the UV/chlorine process was mostly ascribed to reactions with the Cl% and %OH generated from HClO photolysis. The presence of methanol as a radical scavenger resulted in incomplete removal of BP4 in the UV/chlorine process and enhanced the formation of DBPs. The increase in DBP formation was because the residual BP4 can be exited to a triplet state (3BP4∗) upon UV irradiation. 3BP4∗ subsequently reacted with O2 to form 1O2, a reactive oxygen species that generated DBPs by reacting with chlorinated intermediates in water. All benzophenone-type UV-filters may have similar photochemical activity and thus influence the transformation of other organics in sunlit surface water environments.
⁎
Corresponding author. E-mail address: [email protected] (J. Lu).
https://doi.org/10.1016/j.cej.2018.06.089 Received 4 March 2018; Received in revised form 12 June 2018; Accepted 13 June 2018 1385-8947/ © 2018 Elsevier B.V. All rights reserved.
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
1. Introduction
[34]. The entire process was performed in a cooler at 4 °C. This could maximally inhibited the formation of di- and trichloramine [34,35]. The chloramine concentration was calculated assuming that free chlorine had been completely converted to NH2Cl. Swimming pool water was sampled from the Fitness Center of Nanjing City. The sample was filtered through a 0.45 μM membrane. The residual chlorine and DBPs were analyzed immediately after sampling.
Due to their widespread presence in the environment and the risks associated with their presence, organic UV-filters are considered to be emerging contaminants [1–6]. Benzophenone (BP) has two benzene rings jointed by a carbonyl group. Depending on the number and types of the functional groups attached to the backbone, a total of 12 substituted derivatives of BP are commercially used as UV-filters. Among them, BP3 (2-hydroxyl-4-methoxyphenone) and BP4 (2-hydroxy-4methoxybenzophenone-5-sulphonic acid) are the most widely known and have been approved for use in the EU, the U.S., China, Japan, and many other countries [7,8]. High levels of BP3 and BP4 have been detected in water used for recreational purposes [9–11]. Although BP3 and BP4 are relatively resistant to direct photodegradation and biodegradation, they are readily degraded by advanced oxidation, such as Fenton oxidation, TiO2 catalyzed oxidation, activated persulfate oxidation, and sonication [12–15]. BP3 and BP4 are also highly unstable in the presence of chlorine [16–23]. Under typical water chlorination conditions, the half-lives of BP3 and BP4 range from seconds to a few minutes [7,16,20,21]. Free chlorine reacts with BPs via the electrophilic substitution mechanism, resulting in the stepwise incorporation of Cl atoms into the methoxyphenol moiety [7,16,17]. Cleavage of the chlorinated ring due to further chlorination/ oxidation leads to the formation of a variety of chlorinated disinfection by-products (DBPs), such as chloroform and chloroacetic acids [16,17,20,22]. The reactions between chlorine and BPs have been well investigated. However, few studies have examined the reactions of BPs under swimming pool water conditions, where chlorine can be present in both free and combined forms. In addition, BPs in outdoor swimming pool water can be exposed to UV light. UV disinfection is also applied in swimming pool water treatment systems through which water is cycled and reused. Chlorine rapidly decomposes to Cl% and %OH upon UV irradiation [24–26]. As a highly potent oxidant, %OH reacts with a wide spectrum of organic compounds at nearly diffusion-controlled rates [27]. Cl% is also highly reactive to electron-rich compounds [24,28]. A synergistic effect between UV and chlorine was demonstrated during the degradation of contaminants, which were otherwise less reactive to UV or chlorine alone [29–31]. UV/chlorine has been developed as an advanced water treatment strategy to address micropollutants [24,28,32]. In the present study, we investigated the transformation of BP4 by chlorine, chloramine, and UV/chlorine, which are expected to be found in swimming pool water. We demonstrated that BP4 acted as a photosensitizer, leading to the increased formation of DBPs in the presence of UV irradiation. Such photochemical activity has not been previously reported. We assume that other BP-type UV-filters would have a similar property, which could affect the fate of organic compounds that are copresent in UV-based water treatment processes and sunlit environments.
2.2. Reaction kinetics Chlorination and chloramination of BP4 were performed in 40 mL borosilicate glass vials with polypropylene screw caps as batch reactors. The initial BP4 concentration was 1.0 mg/L. The reaction was initiated by rapid mixing of the BP4 solution with chlorine or chloramine. Phosphate buffer (10 mM) was added to maintain pH 7, and the molar ratio of chlorine or chloramine to BP4 was fixed at 10:1. The reactors were kept in the dark at 20 °C. The same conditions were applied for the UV/chlorine treatment except quartz vials were used as the reactors and placed in a BL-GHX-V photoreactor (Bilang, Shanghai, China) equipped with a low-pressure mercury lamp (ozone free, 15 W) emitting dominantly at 254 nm. The average UV fluence rate through the reaction volume was determined to be 4.68 × 10−7 Einstein∙L−1 s−1 using atrazine as an actinometer. The detailed experimental procedures can be found in a study by Canonica et al. [36]. A more detailed description of the apparatus used for the photo reaction can be found in Fig. S1 in the Supporting Data. Aliquots of the solution were withdrawn from each reactor and quenched immediately with 20 μL of 1 M Na2S2O3. To determine the contribution of the free radicals to BP4 removal in the UV/chlorine treatment, a concurrent experiment was conducted in the presence of 1 mM methanol. All samples were kept in a refrigerator at 4 °C until analysis for residual BP4.
2.3. Identification of the transformation products of BP4 Experiments were performed in 250 mL flasks in the dark. The initial concentration of BP4 was 1 mg/L and pH buffered at 7. The reaction was initiated by rapid mixing of a BP4 solution with an appropriate quantity of chlorine or chloramine stock solution. After 2 h of reaction, excess Na2SO3 was added to quench the residual chlorine or chloramine, and 2 mL of 98% sulfuric acid was added to reduce the pH to < 2. Then, solid phase extraction (SPE) using HyperSep C18 cartridges (Thermo Scientific, Waltham, MA) was performed to purify and concentrate the transformation products. Additional details of the procedure are described in the Supporting Information. For UV/ chlorine, the reaction was performed in quartz vials in a photoreactor. After 2 h, the solution was taken out and quenched with Na2SO3 before being extracted by SPE as described above. The transformation products produced by UV/chlorine in the presence of 1 mM methanol were also collected for comparison. The SPE eluents were analyzed using an Agilent 1200 HPLC coupled with an Agilent G6410B Triple Quad Mass spectrometer with an electron spray ionization (ESI) source operated in the negative mode. The detailed instrumental setup is provided in the Supporting Information.
2. Materials and methods 2.1. Materials A BP4, 3,4,5-trimethylphenol (TMP), furfuryl alcohol (FFA), and haloacetic acid calibration mixture was obtained from Sigma-Aldrich (St. Louis, MO). Sodium hypochloride (NaClO) and N,N-dimethyl-1,4phenylenediamine monohydrochloride (DPD) were purchased from Aladdin (Shanghai, China). Stock solutions of BP4 and NaClO were prepared by dissolving the reagents in Milli-Q water and storing them in a refrigerator. Prior to use, the free chlorine concentration of the NaClO stock solution was measured using the DPD method [33]. The chloramine (NH2Cl) solution was freshly prepared by adding a NaClO stock solution dropwise to the NH4Cl solution to achieve a final NH4Cl/ NaClO molar ratio of 1.2:1 following Schreiber and Mitch’s method
2.4. Formation of chloroacetic acids (CAAs) The formation of CAAs in the chlorination, chloramination, and UV/ chlorine processes was investigated in a series of vials. The initial BP4 concentration was 1 mg/L, and the chlorine/BP4 M ratio was 10. The other conditions were identical to those of the kinetic study. At predetermined time points, a vial was sacrificed by adding 1 mM Na2SO3 (1 mM) to quench the reaction. The quenched reaction solutions were kept in a refrigerator until further treatment and analysis for CAAs. 305
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
2.5. BP4 as a photosensitizer The BP4 solution was pretreated with free chlorine for 30 min and then UV for 15 min. The initial BP4 concentration was 1.0 mg/L, and the HClO/BP4 molar ratio was 5. The pH of the solution was buffered at 7.0. Preliminary experiments demonstrated that no residual BP4 or free chlorine was detected after the chlorination and UV pretreatment. Then, additional BP4 was spiked into the solutions before they were subjected to UV irradiation for an additional 8 h. After the reaction, CAAs were extracted and analyzed. To ascertain the reactive species responsible for the formation of CAAs, 1 mM TMP or FFA was spiked together with BP4 in the middle of the treatment. The other treatments and conditions were identical. The same experiments were also performed on the sample of swimming pool water. The water sample was pretreated with UV for 15 min to decompose the residual chlorine before varying concentrations of BP4 were spiked. Then, the samples were treated with UV for an additional 8 h, and the formation of CAAs was analyzed. Controls spiked with the same quantities of BP4 but without the additional UV treatment were run concurrently. 2.6. Analytical methods The residual BP4 in the samples was analyzed by a Hitachi L-2000 high performance liquid chromatographer (HPLC) with an L-2455 diode array detector (DAD). Separation was performed on a LaChrom C18 reverse phase column (5 μm × 4.6 mm × 250 mm) with an isocratic eluent that consisted of 30% methanol (with 0.1% acetic acid) and 70% water (with 0.1% acetic acid) at a flow rate of 1 mL/min. The injection volume was 10 μL, and quantification was based on the absorbance at 285 nm, which is characteristic of BP4. HAAs were analyzed using the USEPA standard method 552.2. The procedure included liquid-liquid extraction followed by methylation and quantification by an Agilent 7890 gas chromatograph (GC) equipped with an electron capture detector (ECD) and a DB5-fused silica capillary column (30 m × 0.53 mm × 1.5 μm). Additional details can be found in the Supporting Information. The free chlorine in the solution was quantified by the DPD method [33]. An aliquot of 1 mL of the solution was sampled from the reactor at a preset reaction time and immediately mixed with 1 mL of DPD solution (8 mM). The free chlorine concentration was quantified according to the absorbance at 510 nm measured by a Varian Cary 50 spectrophotometer. The standard deviation of this approach was within 5%.
Fig. 1. (a) Removal of BP4 during the chlorination, chloramination, UV/ chlorine, and UV/chlorine/methanol processes. Initial BP4 concentration, 1 mg/L; chlorine or chloramine to BP4 molar ratio, 10; pH 7; temperature 20 °C. (b) Removal of BP4 in UV/chlorine/methanol process at varying chlorine doses. Initial BP4 concentration, 1 mg/L; pH 7; methanol, 1 mM; temperature, 20 °C. The points are experimental data, and he dashed lines represent the fitting of a first-order kinetic model. Error bars represent the standard deviation of 3 replicates.
Thus, the removal of BP4 in the UV/chlorine process can be attributed to reactions with HO%, Cl%, and free chlorine. To determine the contribution of the radicals, the reaction was conducted in the presence of 1 mM methanol, which acted as the radical scavenger. The removal rate was greatly reduced under these conditions. Only a 15% attenuation was observed within 15 s at the initial HClO/BP4 molar ratio of 10, in contrast to the 100% removal in the absence of methanol. Thus, it can be concluded that the radicals (HO% and Cl%) that formed upon free chlorine photolysis contributed to the majority of BP4 removal in the UV/chlorine process.
3. Results and discussion 3.1. Reaction kinetics As shown in Fig. 1a, BP4 was rapidly removed during chlorination. At an HClO/BP4 molar ratio of 10, an approximate 90% removal was achieved in 3 min. In accordance with previous studies [16,17,19–21], BP4 followed pseudo first-order decay with a rate constant of 0.012 s−1. The transformation of BP4 by chloramination was slower than that of chlorination. When the free chlorine was replaced with an equal quantity of chloramine, less than 40% removal was found in 5 min. The reaction could also be modeled by pseudo first-order kinetics
(ln
hv
HCIO→C1% + HO%
Free chlorine also contributed to the small fraction of BP4 removal in the UV/chlorine process, as evidenced by the reduction in the presence of methanol (Fig. 1b). Overall, in the presence of 1 mM methanol, an approximately 35% removal of BP4 was achieved in the UV/chlorine process, but most of the removal occurred during the first minute. In this period, the reaction was slower than that in the chlorination process, but faster than that of chloramination. After that, the reaction slowed down. Almost no degradation of BP4 was observed after 5 min (Fig. 1b). The termination of BP4 removal was coincident with the depletion of free chlorine. As shown in Fig. S2 (Supporting Data), in the UV/chlorine process, HClO was completely degraded in 10 min.
)
= −kt , and the rate constant (k) was 0.002 s−1. UV alone did not cause an appreciable attenuation of BP4 during the investigated time period. However, the removal of BP4 was significantly accelerated in the UV/chlorine process. With an initial HClO/BP4 M ratio of 10 in the presence of UV irradiation, complete removal was found to occur within 15 s, indicating a synergistic effect between chlorine and UV [29–31]. In the UV/chlorine process, HClO undergoes photolysis to form HO% and Cl% (R1) [24,25]. Both radical species can react with aromatics with second-order rate constants ranging from 108–109 M−1 s−1 [27]. c co
(R1)
306
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
electron-rich, thus making Cl-BP4ox, if it was a hydroxylated Cl-BP4, more susceptible to a free chlorine attack. Therefore, we presume that the oxidation occurred on the carbonyl group. It has been reported that chlorinated BP4 can undergo a Baeyer-Villiger type oxidative rearrangement during chlorination, converting diphenylketone to phenyl ester [16,23]. The findings of this study suggest that Cl-BP4ox and Cl2BP4ox were formed via such a mechanism, with Cl-BP4ox undergoing further chlorination to form Cl2-BP4ox. The proposed molecular structures of Cl-BP4ox and Cl2-BP4ox are shown in Table S1. In addition, as shown in Fig. 2b, a product with m/z 295/297/299 was also formed, although its occurrence was less significant. Based on its isotope signature, we infer that this product contains two chlorine atoms. This compound, which could have been generated through desulfonation followed by chlorination, has also been observed in previous studies [16,23]. Although this product appeared to be less significant in MS than other products, its formation should not be overlooked because deprotonation of the molecule becomes more difficult following cleavage of the acidic sulfonate group, which causes a less efficient ionization in the negative ESI mode. Overall, based on the molecular structures of the chlorinated products, the proposed transformation pathways for BP4 chlorination are depicted in Fig. 3. Compared to chlorination, chloramination of BP4 led to the formation of the same suite of transformation products, as illustrated in Fig. 2(c). Interestingly, the evolution of these products as a function of the chlorine and chloramine dosages in both the chlorination and chloramination processes showed similar trends (Supporting Data, Fig. S3). Nevertheless, under the same reaction conditions, the yields of the products from chloramination were always much greater than those from chlorination. Based on these findings, it can be assumed that it was free chlorine that reacted with BP4 during chloramination. Chloramine is formed by the combination of free chlorine and NH4+. The reaction is reversible, thus, while HClO can be slowly released from chloramine, its equilibrium concentration would be extremely low. Thus, the transformation of BP4 was significantly slow and the chlorinated BP4 intermediates could be readily accumulated in the chloramination process. The chlorinated products of BP4 identified in the chlorination and chloramination processes were not found in the UV/chlorine treatment (data not shown). HO% and Cl% that were generated during photolysis of HClO acted as the main reactive species for the BP4 transformation in the UV/chlorine process, as evidenced by the suppressed removal when methanol was used as the radical scavenger (Fig. 1). However, Cl% reacts with organic compounds via an electron transfer mechanism rather than addition or substitution [38], which is unlikely to lead to chlorinated products. Fig. 2(d) also shows that Cl-BP4 and Cl2-BP4ox were
Fig. 2. MS of possible BP4 transformation products generated with a negative ESI. (a) Control; (b) chlorination; (c) chloramination; (d) UV/chlorine/methanol. The initial BP4 concentration was 1 mg/L; HClO or NH2Cl was 10 times (molar ratio) that of the substrate; methanol 1 mM. The reaction time was 2 h for chlorination and chloramination and 20 min for UV/chlorination/methanol; pH 7.
3.2. Identification of reaction products MS was employed to analyze the transformation products of BP4 after the chlorine, chloramine, and UV/chlorine treatments. The data are shown in Fig. 2. Compared to the control sample shown in Fig. 2a, b reveals the disappearance of the signal for BP4 (m/z 307) as well as the appearance of new peaks at m/z 295/297/299, m/z 341/343, m/z 357/ 359, m/z 375/377/379, and m/z 391/393/395. The compound corresponding to m/z 341/343 has a characteristic M + 2 isotopic peak, with an abundance of approximately 1/3 of that of M, indicating that it contains a chlorine atom. In addition, the molecular weight (MW) of this compound is found to be 34 Daltons greater than that of BP4, allowing us to infer that this compound is a mono-chlorine substituted BP4 (Cl-BP4). The compound corresponding to m/z 375/377/379 is 68 Daltons greater than that of BP4, with a characteristic M:(M + 2): (M + 4) ratio of 9:6:1, suggesting that this compound is a derivative of BP4 with two substituted chlorine atoms (Cl2-BP4). Chlorination of phenolic compounds involves electrophilic substitution at the para and ortho positions of the hydroxyl group on the phenolic ring [37]. However, the para site of the BP4 molecule is occupied by a sulfonic group, and one of the ortho sites is connected to the carbonyl group. Thus, we infer that Cl-BP4 is 2-hydroxyl-3-chloro-4-methoxyl benzophenonesulfonic acid. After the para and ortho positions are substituted, the second chlorine is directed to the unoccupied meta site. Therefore, Cl2BP4 is presumed to be 2-hydroxyl-3,6-dichloro-4-methoxyl benzophenone-sulfonic acid. The proposed molecular structures of Cl-BP4 and Cl2-BP4 are shown in Table S1. The formation of chlorine substituted BP4 increased initially and then decreased with the increasing chlorine dose (Supporting Data, Fig. S3). Cl-BP4 reached a maximum concentration at an HClO/BP4 molar ratio of 1, while Cl2-BP4 reached a maximum concentration at an HClO/BP4 molar ratio of 4, suggesting that the transformation of BP4 proceeded from the monochlorinated to the dichlorinated derivative. According to the isotope signature (M:(M + 2) = 3:1) shown in Fig. 2(b), the compound of m/z 357/359 also contains a chlorine atom. Because its MW is 16 Daltons greater than that of Cl-BP4, it can be assumed that this compound is an oxidation product of Cl-BP4 (denoted as Cl-BP4ox). Similarly, the compound with m/z 391/393/395 corresponds to the oxidation product of Cl2-BP4 (denoted as Cl2-BP4ox). ClBP4ox and Cl2-BP4ox appeared to be relatively stable towards free chlorine. Their formation, especially Cl2-BP4ox, increased with the increasing chlorine dosage (Supporting Data, Fig. S4). This stability suggests that it is unlikely that Cl-BP4ox was formed through the hydroxylation of Cl-BP4 because, as an electron-donating functional group, hydroxyl substitution would make the aromatic ring more
Fig. 3. Proposed transformation pathways of BP4 during the chlorination and chloramination processes. 307
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
desulfonicated form. The presence of an electron-withdrawing sulfonic group that decreased the electron density of the phenolic ring made it less vulnerable to chlorine attack and was disadvantageous to the formation of DBPs. Less DBP formation was expected in chloramination than chlorination. Substitution of chlorine with chloramine as the disinfectant is practiced in water treatment to reduce the formation of DBPs. It should be noted, however, that although the formation of DBPs was greatly reduced, chloramination resulted in more chlorinated BP4 derivatives. The potential risks of these chlorinated products to the environment and human health warrant further study. As previously discussed, in the UV/chlorine process, instead of HClO, HO% and Cl% were the main reactive species. These species reacted with organic compounds via one-electron oxidation, which is unlikely to generate chlorinated DBPs[38]. In addition, DBPs, if formed, would be degraded by UV irradiation (Supporting Data, Fig. S5). It is well recognized that given the same chlorine dose, the UV/chlorine treatment generates less halogenation DBPs than chlorination [24,25]. An interesting finding is that substantial CAAs were formed in the UV/chlorine process in the presence of methanol. As shown in Fig. 4(a), this formation was remarkably higher than that of the chlorination process. For instance, at 4 h, the yields of TCAA reached 1.49 and 7.54 μg/L in the chlorination and UV/chlorine/methanol processes, respectively. The yields of DCAA were 0.60 and 4.36 μg/L, respectively. Considering the fact that a significant fraction of BP4 (approximately 65% at this condition) still remained in the solution treated with UV/ chlorine/methanol (Fig. 1(b)), the formation of CAAs in terms of the BP4 molar yield was even higher. This high formation of CAAs in the UV/chlorine/methanol process was unexpected. Since the radicals were mostly scavenged by methanol, the removal of BP4 under these conditions was mainly due to its reaction with free chlorine. However, the majority of the CAAs were formed after 10 min when free chlorine had been completely decomposed (Supporting Data, Fig. S2), and further transformation of BP4 was not observed (Fig. 1(b)). Thus, the CAAs generated after 10 min are presumed to have resulted from the decomposition of the chlorinated BP4 intermediates that formed before the depletion of free chlorine. UV radiation was found to be critical during the formation of HAAs in this scenario. Fig. 4(b) shows CAA formation in the same treatment (UV/chlorine/methanol) but UV was terminated at 30 min. It is observed that the CAA concentrations did not change appreciably after the UV source was turned off. Therefore, it is evident that the majority of CAAs that formed under these conditions were ascribed to reactions driven by UV.
Fig. 4. (a) Formation of CAAs during the chlorination and UV/chlorination/ methanol treatment of BP4. (b) Formation of chlorinated acetic acids in the UV/ chlorination/methanol treatment of BP4 for 30 min followed by a dark incubation. Initial BP4 concentration, 1 mg/L; chlorine or chloramine to BP4 molar ratio, 10; methanol, 1 mM; pH 7; temperature, 20 °C. The error bars represent the standard deviation of 3 replicates.
formed when methanol was present, reinforcing the previous argument that free chlorine contributed to some BP4 removal in the UV/chlorine process, at least during the initial stage of the reaction when HClO was not completely decomposed. However, this contribution might be small compared to that of the radicals. It should be noted that these chlorinated products were not detected in UV/chlorine in the absence of methanol, probably because they were rapidly degraded by free radicals.
3.4. Mechanisms of the formation of CAAs in the UV/chlorine/methanol process Another interesting finding is that besides UV, the residual BP4 in the reaction solution also played a critical role in the formation of CAAs in the UV/chlorine/methanol process. As shown in Fig. 5(a), if the solutions were subsequently kept in the dark, adding additional BP4 to the BP4 solution pretreated with chlorine (no residual BP4 or free chlorine found in the solution when additional BP4 was spiked), did not result in a noticeable influence on the formation of CAAs. When the solutions were incubated under UV subsquently, however, more CAAs were generated. A positive relationship is observed between the formation of CAAs and the quantity of BP4 spiked in the middle of the treatment. Hence, it can be inferred that, in the UV/chlorine/methanol process, the residual BP4 facilitated the UV-driven formation of CAAs. A rational mechanism of CAA formation in the UV/chlorine/methanol process is BP4-sensitized indirect photodegradation of the intermediates resulting from BP4 chlorination. As a UV-filter, BP4 was designed to absorb UV light but be resistant to direct photolysis [39]. To avoid degradation, a photoexcited BP4 molecule can release the energy from a photon in the form of heat or convey the energy to other molecules, i.e., it can act as a photosensitizer. In fact, benzophenone has been used as a photosensitizer to initiate a radical polymerization
3.3. Formation of chloroacetic acids(CAAs) It has been reported that chlorination of BP4 leads to the formation of chlorinated DBPs [16,22]. In this study, only the formation of semivolatile CAAs was quantified (Fig. 4). Since the vials in the photoreactor could not be sealed tightly and head-space remained during the reaction, the volatile chloroform was not analyzed. It was revealed that CAAs were formed in the chlorination process, while almost no CAAs were detected in the chloramination or UV/chlorine treatment (data not shown). In the chlorination process, the yield of trichloroacetic acid (TCAA) was higher than that of dichloroacetic acid (DCAA) and monochloroacetic acid (MCAA) was under the detection limit. Since ClBP4ox and Cl2-BP4ox were relatively stable toward chlorine and chloramine (Supporting Data, Fig. S4), we presumed that CAAs formed as the result of the further chlorination of chlorinated BP4, especially the 308
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
Fig. 5. Influence of the presence of BP4 on the formation of chlorinated acetic acids during the UV treatment process. (a) BP4 was spiked into pre-chlorinated BP4 samples. Initial BP4 concentration, 1 mg/L; chlorine/BP4 molar ratio, 5; pH 7. (b) BP4 was spiked into a swimming pool water sample. No residual free chlorine was presented at the time of additional BP4 spiking. Fig. 6. Effects of TMP and FFA on the formation of (a) DCAA and (b) TCAA due to BP4-mediated indirect photo degradation; 1 mM TMP or FFA was spiked together with BP4 in pre-chlorinated BP4 samples when no free chlorine was left. The sample was subsequently subjected to UV irradiation for 8 h.
reactions in polymer synthesis[40–42]. In environmental chemistry studies, benzophenones are well-defined proxies to assess the photochemical activity of dissolved natural organic matter (NOM). It has been recognized that benzophenone moieties are important chromophores in NOM molecules that can be excited to form a triplet state (3NOM∗), which reacts with dissolved oxygen to generate an oxygen singlet (1O2) in a sunlit environment. Both 3NOM∗ and 1O2 are highly reactive species that can cause the transformation of organic contaminants that are otherwise insensitive to direct photolysis [43–47]. Similarly, upon UV irradiation of (R2), we presume that the residual BP4 in the UV/chlorine/methanol process can be excited to form triplet state BP4 (3BP4∗). 3BP4∗ then further reacts with dissolved oxygen to give rise to 1O2 (R3). The formation of the CAAs was most likely the result of the degradation of chlorinated BP4 intermediates (Cl-BP4, Cl2BP4ox, etc., Fig. 2(d)) by 3BP4∗ and/or an 1O2 attack; i.e., the chlorinated BP4 intermediates underwent indirect photo degradation to break down and release CAAs. To verify this hypothesis, the above experiment was performed in the presence of TMP, which is able to selectively scavenge 3BP4∗ [47]. TMP was spiked along with the additional BP4 in the middle of the treatment process when both free chlorine and BP4 had been depleted. The data shown in Fig. 6 reveal that the formation of CAAs slightly increased when TMP was spiked together with the additional BP4. A comparison of the formation of CAAs in the presence and absence of TMP clearly indicates that 3BP4∗ was involved in the formation of CAAs in the UV/chlorine/methanol
process. Fig. 6 also provides data on the formation of CAAs in the presence of FFA instead of TMP, showing that the addition of FFA also significantly inhibited the formation of CAAs. The yields were at the same levels as those in the presence of TMP. Since FFA selectively quenches 1O2 [48,49], which could only have been generated as the result of the reaction between 3BP4∗ and dissolved oxygen, these data indicate that after the depletion of free chlorine in the UV/chlorine/ methanol process, 1O2 directly reacted with chlorinated BP4 to form CAAs (R4). hv
BP4 → 3BP4∗
(R2)
3BP4∗
+ O2 → BP4 + 1O2
(R3)
+ CIBP4 → →CAAs
(R4)
1O 2
As a photo-sensitizer, we hypothesize that under UV irradiation, BP4 can not only mediate the degradation of chlorinated BP4 intermediates to form DBPs but also that intermediates are derived from reactions between free chlorine and other organic impurities in water. To test this hypothesis, a sample of swimming pool water was collected. Preliminary experiments indicated that there was 1.28 mg/L residual 309
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
chlorine but no BP4 in the water. The sample was subjected to UV irradiation for 15 min to remove the residual chlorine. Then, varying amounts of BP4 were spiked and treated with UV for an additional 8 h. No appreciable degradation of the spiked BP4 was observed after UV treatment. However, as illustrated in Fig. 5(b), a positive relationship was evident between the final concentrations of CAAs and the quantity of spiked BP4. The concentrations of DCAA and TCAA in the sample were 23.51 and 13.75 μg/L, respectively, without the addition of BP4. The formation increased to 79.84 and 92.37 μg/L, respectively, when 1 mg/L of BP4 was spiked. We hypothesize that the increased formation of CAAs was due to the degradation of chlorinated organic matter that was originally in the swimming pool water caused by photo-generated 1 O2. During this process, BP4 acted as the photo-sensitizer. By contrast, when the samples were incubated in the dark, the presence of BP4 did not have an appreciable influence on the concentrations of CAAs (Fig. 5(b)). The concentrations of CAAs were close to their original values, approximately 64.8 and 36.5 μg/L, respectively, for DCAA and TCAA. It should be noted that the concentrations were higher than those in the sample subjected to an additional UV treatment in the absence of BP4 because CAAs were removed by direct photo-degradation (Supporting Data, Fig. S5). The data shown in Figs. 5 and 6 clearly demonstrate that the co-presence of BP4 and UV can significantly increase the formation of DBPs in chlorinated water.
[2] I. Ozaez, J. Luis Martinez-Guitarte, G. Morcillo, Effects of in vivo exposure to UV filters (4-MBC, OMC, BP-3, 4-HB, OC, OD-PABA) on endocrine signaling genes in the insect Chironomus riparius, Sci. Total Environ. 456 (2013) 120–126. [3] C. Liao, K. Kannan, Widespread occurrence of benzophenone-type UV light filters in personal care products from china and the united states: an assessment of human exposure, Environ. Sci. Technol. 48 (2014) 4103–4109. [4] P.Y. Kunz, K. Fent, Estrogenic activity of ternary UV filter mixtures in fish (Pimephales promelas) - An analysis with nonlinear isobolograms, Toxicol. Appl. Pharmacol. 234 (2009) 77–88. [5] M. Krause, A. Klit, M.B. Jensen, T. Soeborg, H. Frederiksen, M. Schlumpf, W. Lichtensteiger, N.E. Skakkebaek, K.T. Drzewiecki, Sunscreens: are they beneficial for health? An overview of endocrine disrupting properties of UV-filters, Int. J. Androl. 35 (2012) 424–436. [6] D.L. Giokas, A. Salvador, A. Chisvert, UV filters: from sunscreens to human body and the environment, Trac-Trends Anal. Chem. 26 (2007) 360–374. [7] A.J.M. Santos, M.S. Miranda, J.C.G. Esteves da Silva, The degradation products of UV filters in aqueous and chlorinated aqueous solutions, Water Res. 46 (2012) 3167–3176. [8] N.A. Shaath, Ultraviolet filters, Photochem. Photobiol. Sci. 9 (2010) 464–469. [9] M.M.P. Tsui, H.W. Leung, T.-C. Wai, N. Yamashita, S. Taniyasu, W. Liu, P.K.S. Lam, M.B. Murphy, Occurrence, distribution and ecological risk assessment of multiple classes of UV filters in surface waters from different countries, Water Res. 67 (2014) 55–65. [10] D.A. Lambropoulou, D.L. Giokas, V.A. Sakkas, T.A. Albanis, M.I. Karayannis, Gas chromatographic determination of 2-hydroxy-4-methoxybenzophenone and octyldimethyl-p-aminobenzoic acid sunscreen agents in swimming pool and bathing waters by solid-phase microextraction, J. Chromatogr. A 967 (2002) 243–253. [11] Y. Tashiro, Y. Kameda, Concentration of organic sun-blocking agents in seawater of beaches and coral reefs of Okinawa Island, Japan, Mar. Pollut. Bull. 77 (2013) 333–340. [12] H. Zuniga-Benitez, J. Soltan, G.A. Penuela, Application of ultrasound for degradation of benzophenone-3 in aqueous solutions, Int. J. Environ. Sci. Technol. 13 (2016) 77–86. [13] H. Zuniga-Benitez, C. Aristizabal-Ciro, G.A. Penuela, Photodegradation of the endocrine-disrupting chemicals benzophenone-3 and methylparaben using Fenton reagent: optimization of factors and mineralization/biodegradability studies, J. Taiwan Inst. Chem. Eng. 59 (2016) 380–388. [14] X. Pan, L. Yan, R. Qu, Z. Wang, Degradation of the UV-filter benzophenone-3 in aqueous solution using persulfate activated by heat, metal ions and light, Chemosphere 196 (2018) 95–104. [15] H. Zuniga-Benitez, C. Aristizabal-Ciro, G.A. Penuela, Heterogeneous photocatalytic degradation of the endocrine-disrupting chemical Benzophenone-3: parameters optimization and by-products identification, J. Environ. Manage. 167 (2016) 246–258. [16] M. Xiao, D. Wei, J. Yin, G. Wei, Y. Du, Transformation mechanism of benzophenone-4 in free chlorine promoted chlorination disinfection, Water Res. 47 (2013) 6223–6233. [17] S. Zhang, X. Wang, H. Yang, Y.F. Xie, Chlorination of oxybenzone: kinetics, transformation, disinfection byproducts formation, and genotoxicity changes, Chemosphere 154 (2016) 521–527. [18] T. Manasfi, V. Storck, S. Ravier, C. Demelas, B. Coulomb, J.-L. Boudenne, Degradation Products of Benzophenone-3 in Chlorinated Seawater Swimming Pools, Environ. Sci. Technol. 49 (2015) 9308–9316. [19] P. Abdallah, M. Deborde, F.D. Berne, N.K.V. Leitner, Kinetics of Chlorination of Benzophenone-3 in the Presence of Bromide and Ammonia, Environ. Sci. Technol. 49 (2015) 14359–14367. [20] S.E. Duirk, D.R. Bridenstine, D.C. Leslie, Reaction of benzophenone UV filters in the presence of aqueous chlorine: kinetics and chloroform formation, Water Res. 47 (2013) 579–587. [21] J. Li, L.-Y. Ma, L. Xu, Transformation of benzophenone-type UV filters by chlorine: kinetics, products identification and toxicity assessments, J. Hazard. Mater. 311 (2016) 263–272. [22] N. Negreira, P. Canosa, I. Rodriguez, M. Ramil, E. Rubi, R. Cela, Study of some UV filters stability in chlorinated water and identification of halogenated by-products by gas chromatography-mass spectrometry, J. Chromatogr. A 1178 (2008) 206–214. [23] N. Negreira, I. Rodriguez, R. Rodil, R. Cela, Assessment of benzophenone-4 reactivity with free chlorine by liquid chromatography quadrupole time-of-flight mass spectrometry, Anal. Chim. Acta 743 (2012) 101–110. [24] J.Y. Fang, Y. Fu, C. Shang, The Roles of Reactive Species in Micropollutant Degradation in the UV/Free Chlorine System, Environ. Sci. Technol. 48 (2014) 1859–1868. [25] T. Li, Y. Jiang, X. An, H. Liu, C. Hu, J. Qu, Transformation of humic acid and halogenated byproduct formation in UV-chlorine processes, Water Res. 102 (2016) 421–427. [26] Y.G. Feng, D.W. Smith, J.R. Bolton, Photolysis of aqueous free chlorine species (NOCI and OCI-) with 254 nm ultraviolet light, J. Environ. Eng. Sci. 6 (2007) 277–284. [27] P. Neta, R. Huie, A. Ross, Rate constants for reactions of inorganic radicals in aqueous solution, J. Phys. Chem. Ref. Data 17 (1988) 1027–1284. [28] X. Yang, J. Sun, W. Fu, C. Shang, Y. Li, Y. Chen, W. Gan, J. Fang, PPCP degradation by UV/chlorine treatment and its impact on DBP formation potential in real waters, Water Res. 98 (2016) 309–318. [29] W.-L. Wang, Q.-Y. Wu, N. Huang, T. Wang, H.-Y. Hu, Synergistic effect between UV and chlorine (UV/chlorine) on the degradation of carbamazepine: influence factors and radical species, Water Res. 98 (2016) 190–198.
4. Conclusions This study demonstrates that BP4 can be removed by chlorine, chloramine, and UV/chlorine treatment. Chlorination was more efficient than chloramine but led to the formation of DBPs, such as CAAs. The degradation of BP4 by UV/chlorine treatment was more efficient than that by chlorine, and the removal was mostly due to reactions with free radicals (HO% and Cl%) that were generated upon photolysis of free chlorine, which resulted in a low formation of DBPs. Adding methanol as the radical quenching agent to the UV/chlorine process caused incomplete removal of BP4, but remarkably increased the formation of CAAs. This enhanced formation of CAAs was due to the photochemical activity of the residual BP4, which, under UV irradiation, was excited to form 3BP4∗. 3BP4∗ reacted with dissolved oxygen to form 1O2, which reacted with the chlorinated BP4 intermediates to form CAAs. We believe that due to their structural similarity, other BP type UV-filters have similar photochemical activities and are thus likely to affect the transformation of other co-existing organic compounds. This photochemical activity of BP type UV-filters should be taken into consideration in studies of their behavior in natural and engineering systems. Acknowledgements This research was supported by the National Natural Science Foundation of China (51578294) and the Priority Academic Program Development (PAPD) of the Jiangsu Higher Education Institute. The authors would like to give their gratitude to Randall David Pierce at the University of Georgia (UGA) for the assistance in polishing the language. J. Lu would like to acknowledge the China Scholar Council for supporting his research at UGA. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cej.2018.06.089. References [1] M.E. Balmer, H.R. Buser, M.D. Muller, T. Poiger, Occurrence of some organic UV filters in wastewater, in surface waters, and in fish from Swiss lakes, Environ. Sci. Technol. 39 (2005) 953–962.
310
Chemical Engineering Journal 351 (2018) 304–311
J. Lu et al.
[40] A.G. Krainev, R.I. Viner, D.J. Bigelow, Benzophenone-sensitized photooxidation of sarcoplasmic reticulum membranes: site-specific modification of the Ca2+-ATPase, Free Radical Biol. Med. 23 (1997) 1009–1020. [41] S. Riyajan, J.T. Sakdapipanich, Y. Tanaka, Controlled degradation of cured natural rubber by encapsulated benzophenone as a photosensitizer, J. Appl. Polym. Sci. 90 (2003) 297–305. [42] K.-I. Fukuka, M. Ueda, Recent progress of photosensitive polyimides, Polym. J. 40 (2008) 281–296. [43] S. Canonica, B. Hellrung, J. Wirz, Oxidation of phenols by triplet aromatic ketones in aqueous solution, J. Phys. Chem. A 104 (2000) 1226–1232. [44] K. McNeill, S. Canonica, Triplet state dissolved organic matter in aquatic photochemistry: reaction mechanisms, substrate scope, and photophysical properties, Environ. Sci. Processes Impacts 18 (2016) 1381–1399. [45] C. Yuan, M. Chakraborty, S. Canonica, L.K. Weavers, C.M. Hadad, Y.-P. Chin, Isoproturon reappearance after photosensitized degradation in the presence of triplet ketones or fulvic acids, Environ. Sci. Technol. (2016). [46] S. Canonica, B. Hellrung, P. Muller, J. Wirz, Aqueous oxidation of phenylurea herbicides by triplet aromatic ketones, Environ. Sci. Technol. 40 (2006) 6636–6641. [47] K. Golanoski, S. Fang, R.D. Vecchio, N.V. Blough, Investigating the mechanism of phenol photooxidation by humic substances, Environ. Sci. Technol. 46 (2012) 3912–3920. [48] W.R. Haag, J. Hoigne, Singlet oxygen in surface waters.3. Photochemical formation and steady-state concentration in various types of waters, Environ. Sci. Technol. 20 (1986) 341–348. [49] M. Jiang, J. Lu, Y. Ji, D. Kong, Bicarbonate-activated persulfate oxidation of acetaminophen, Water Res. 116 (2017) 324–331.
[30] X.R. Zhang, W.G. Li, P.F. Ren, X.J. Wang, Chlorine/UV induced photochemical degradation of total ammonia nitrogen (TAN) and process optimization, RSC Adv. 5 (2015) 63793–63799. [31] X.J. Kong, J. Jiang, J. Ma, Y. Yang, W.L. Liu, Y.L. Liu, Degradation of atrazine by UV/chlorine: efficiency, influencing factors, and products, Water Res. 90 (2016) 15–23. [32] M.J. Watts, E.J. Rosenfeldt, K.G. Linden, Comparative OH radical oxidation using UV-Cl-2 and UV-H2O2 processes, J. Water Supply Res. Technol. AQUA 56 (2007) 469–477. [33] APHA/AWWA/WEF, Standard Methods for the Examination of Water and Wastewater, 21st ed., EPS Group, Inc., 2005. [34] I.M. Schreiber, W.A. Mitch, Influence of the Order of Reagent Addition on NDMA Formation during Chloramination, Environ. Sci. Technol. 39 (2005) 3811–3818. [35] M. Nihemaiti, J. Le Roux, C. Hoppe-Jones, D.A. Reckhow, J.-P. Croué, Formation of haloacetonitriles, haloacetamides, and nitrogenous heterocyclic byproducts by chloramination of phenolic compounds, Environ. Sci. Technol. 51 (2017) 655–663. [36] S. Canonica, L. Meunier, U. von Gunten, Phototransformation of selected pharmaceuticals during UV treatment of drinking water, 42 (2008) 121–128. [37] J. Lu, M.M. Benjamin, G.V. Korshin, H. Gallard, Reactions of the flavonoid hesperetin with chlorine: a spectroscopic study of the reaction pathways, Environ. Sci. Technol. 38 (2004) 4603–4611. [38] K.M. Parker, W.A. Mitch, Halogen radicals contribute to photooxidation in coastal and estuarine waters, PNAS 113 (2016) 5868–5873. [39] E. De Laurentiis, M. Minella, M. Sarakha, A. Marrese, C. Minero, G. Mailhot, M. Brigante, D. Vione, Photochemical processes involving the UV absorber benzophenone-4 (2-hydroxy-4-methoxybenzophenone-5-sulphonic acid) in aqueous solution: reaction pathways and implications for surface waters, Water Res. 47 (2013) 5943–5953.
311