Effects of acetylacetone on the photoconversion of pharmaceuticals in natural and pure waters

Environmental Pollution 225 (2017) 691e699

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Effects of acetylacetone on the photoconversion of pharmaceuticals in natural and pure waters* Guoyang Zhang, Bingdang Wu, Shujuan Zhang* State Key Laboratory of Pollution Control and Resource Reuse, School of the Environment, Nanjing University, Nanjing, 210023, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 August 2016 Received in revised form 7 January 2017 Accepted 9 January 2017 Available online 8 April 2017

Acetylacetone (AcAc) has proven to be a potent photo-activator in the degradation of color compounds. The effects of AcAc on the photochemical conversion of five colorless pharmaceuticals were for the first time investigated in both pure and natural waters with the UV/H2O2 process as a reference. In most cases, AcAc played a similar role to H2O2. For example, AcAc accelerated the photodecomposition of carbamazepine, oxytetracycline, and tetracycline in pure water. Meanwhile, the toxicity of tetracyclines and carbamazepine were reduced to a similar extent to that in the UV/H2O2 process. However, AcAc worked in a way different from that of H2O2. Based on the degradation kinetics, solvent kinetic isotope effect, and the inhibiting effect of O2, the underlying mechanisms for the degradation of pharmaceuticals in the UV/ AcAc process were believed mainly to be direct energy transfer from excited AcAc to pharmaceuticals rather than reactive oxygen species-mediated reactions. In natural waters, dissolved organic matter (DOM) played a crucial role in the photoconversion of pharmaceuticals. The role of H2O2 became negligible due to the scavenging effects of DOM and inorganic ions. Interestingly, in natural waters, AcAc first accelerated the photodecomposition of pharmaceuticals and then led to a dramatic reduction with the depletion of dissolved oxygen. Considering the natural occurrence of diketones, the results here point out a possible pathway in the fate and transport of pharmaceuticals in aquatic ecosystems. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Acetylacetone PPCPs Photolysis Tetracycline Toxicity

1. Introduction The widespread use of pharmaceuticals and personal care products (PPCPs) and their incomplete elimination by conventional water treatment have resulted in their undesirable accumulation in the environment (Gao et al., 2012; Zhang et al., 2014a). A significant number of PPCPs have been frequently detected in various water bodies (wastewater, surface water, drinking water, ground water) and solids (sludge, soil, and sediments) (Benotti et al., 2009; Musolff et al., 2009; Gobel et al., 2005; Kim and Carlson, 2007; Sun et al., 2016; Kuemmerer, 2009a, 2009b; Li et al., 2015; Li, 2014). Although the acute and chronic effects of PPCPs on the ecosystem and human health are not yet fully understood, longterm and low-dose exposure to PPCPs will inevitably induce irreversible adversity by increasing the drug resistance of bacteria, and subsequently threatening the health of human beings (Andreozzi et al., 2006; Costanzo et al., 2005; Dodgen et al., 2013; Schmitt

*

This paper has been recommended for acceptance by Charles Wong. * Corresponding author. E-mail address: [email protected] (S. Zhang).

http://dx.doi.org/10.1016/j.envpol.2017.01.089 0269-7491/© 2017 Elsevier Ltd. All rights reserved.

et al., 2004). Considering the risks of PPCPs in the ecosystem, the study of their occurrence, toxicity, distribution, and fate becomes necessary. Photolysis under solar irradiation has been considered as one of the most important ways for the degradation of PPCPs in the natural aquatic environment (Boreen et al., 2003; Bohn et al., 2013). The UV/H2O2 process is one of the most intensively investigated processes in water treatment because of the efficient formation of strongly oxidative hydroxyl radicals (Kim et al., 2009). Small molecular diketones are natural products of biofermentation. Some of them exist in all brewery sewage (White and Wainwright, 1975). Among the diketones, acetylacetone (2,4-pentanedione, AcAc) has been widely used in organic synthesis as precursors or catalysts (metal complexes) and in chemical analysis due to its strong chelating ability (Zhou et al., 2008). AcAc is also an additive in gasoline, lubricants, inks, and dyes (Budavari et al., 2001). It is reported that the content of AcAc in one printing ink was as high as 6% (w/w) (Rastogi, 1991). AcAc could be directly released into the environment from various waste streams. Furthermore, AcAc is also a possible semi-oxidation product in water treatment, such as the ozonation of sludge-press liquors (Boyle and McCullough, 1996).

692

G. Zhang et al. / Environmental Pollution 225 (2017) 691e699

Therefore, the coexistence of diketones and pharmaceuticals in the environment is highly possible. AcAc has been reported as an excellent alternative to H2O2 in the photobleaching of dyes (Liu et al., 2014; Wang et al., 2013). In terms of decolorization, the UV/AcAc process was much more efficient than the UV/H2O2 process because of the formation of dye-AcAc exciplexes (Zhang et al., 2014b). Therefore, the strong light absorbing ability of dyes, which usually reduces the efficiency of the UV/H2O2 process due to the inner filter effect, became a useful characteristic in the UV/AcAc process. Unlike dyes, pharmaceuticals are usually colorless. We were curious whether the good targetselectivity of the UV/AcAc process still works for pharmaceuticals. Therefore, in the present work, the photodegradation of several pharmaceuticals were systematically investigated side-by-side with the UV/AcAc and UV/H2O2 processes. The main objectives of this study are: (i) to compare the efficiency of the two processes in degradation of PPCPs, (ii) to explore the effect of AcAc and oxygen on the photolysis mechanism, (iii) to investigate the influence of water matrices on the photochemical behavior of PPCPs in the UV/AcAc and UV/H2O2 processes, and (iv) to assess the toxicity evolution in the photochemical processes. 2. Materials and methods 2.1. Chemicals AcAc, 2,3-butanedione (denoted as BD), and H2O2 of analytical grade were purchased from Shanghai Reagent Station, China. NaOH and HCl of analytical grade were obtained from Nanjing Reagent Station, China. Methanol, acetonitrile, and formic acid of chromatographical grade were purchased from Sigma-Aldrich and were used as received. LB-broth was supplied from Qingdao Hope BioTechnology Co., Ltd., China. Catalase (2000e5000 u/mg), tetracycline hydrochloride (TC) (95%, CAS number: 64-75-5) and rose bengal (RB) were purchased from Sigma-Aldrich. Oxytetracycline hydrochloride (OTC) (97%, CAS number: 2058-46-0) was obtained from J&K, China. Carbamazepine (CBZ) (98%, CAS number: 298-464), ciprofloxacin hydrochloride (CIP) (98%, CAS number: 86393-320), and chloramphenicol (CHL) (98%, CAS number: 56-75-7) were purchased from Aladdin, China. More detailed information about the tested PPCPs is listed in Table S1. Deuteroxide (D2O) was purchased from the Sigma Chemical Corporation. 2,2,6,6-Tetramethylpiperidine (TEMP) (98%) was obtained from J&K Co., China. KO2 (96.5%) was purchased from Alfa Aesar, USA. Nitro blue tetrazolium chloride (NBT) was purchased from Aladdin, China. High purity N2 (99.999%) and N2O (99.95%) were used to purge the solutions whenever needed. All experiments were carried out under ambient conditions unless otherwise stated. 2.2. Water quality measurements Natural waters were sampled from three local water bodies: Yangtze River, Jiuxiang River, and Xuanwu Lake, China. The sampling locations are available in our previous work (Zhou et al., 2015). After sampling, the waters were immediately filtered through a 0.45 mm cellulose nitrate filter (Shimadzu, Japan) and stored at 4
C until use. The total organic carbon (TOC), total carbon (Tot C), total nitrogen (Tot N) were determined with a Multi N/C TOC apparatus (TOC-L, Shimadzu, Japan). The pH, electronic conductivity (EC), dissolved oxygen (DO), oxidation-reduction potential (ORP) were determined with an HQ30d apparatus (HACH, United States). The water quality parameters were measured in duplicate or triplicate and are listed in Table S2.

2.3. Photoirradiation experiments The UV irradiation experiments were carried out in a rotating disk photoreactor (Nanjing StoneTech Electric Equipment, China), which is shown in Fig. S1a. A medium pressure mercury (UV) lamp (300 W) with a maximum light emission at 365 nm or a Xenon (Xe) lamp (350 W) was vertically placed in a cooling water jacket. Sample solutions containing 80 mM target compound and 0.5 mM AcAc or H2O2 were parallelly arranged in a quartz tube around the lamp. The distance between the sample tube and the lamp was 5 cm. The sample holder revolves around the lamp and the tubes self-rotate. The light intensity reaching the solution was measured using a radiometer (Photoelectric Instrument Factory of Beijing Normal University, China), equipped with a sensor with peak sensitivity at 365 nm (Fig. S1b). Considering that the pH of natural waters is usually in range of 6e9 (Brezonik and Arnold, 2011), the initial solution pH was adjusted to 7.0 with HCl or NaOH in pure water unless otherwise stated. As for the natural water matrices, there was no pH adjustment during the irradiation. 2.4. Analytical methods The UV-Vis spectra were analyzed with a double beam spectrophotometer (UV-2700, Shimadzu, Japan). The concentrations of pharmaceuticals were determined with a high performance liquid chromatography (HPLC) system (Dionex U3000, United States) equipped with an Agilent C18 reversed phase column (100 mm  4.6 mm, 3.5 mm) at 25
C. The detailed HPLC conditions are listed in Table S1. The concentration of AcAc was determined by the HPLC equipped with a C8 reversed phase column (150 mm  4.6 mm, 5 mm) at a flow rate of 0.6 mL/min at 274 nm. The mobile phase (60/40, v/v %) was methanol/1 mM CuCl2 solution adjusted with CH3COOH to pH 4.0. Reaction intermediates and products were identified with a liquid chromatography-mass spectrometer (LC-MS) equipped with a ThermoFinnigan LCQ Advantage MAX mass spectrometer with an electrospray ionization (ESI) interface source. The LC system was equipped with an Agilent 4.6  150 mm, 5 mm ZORBAX Eclipse Plus C18 column, and the flow rate was 0.2 mL min1. A sample volume of 10 mL was injected by using an auto sampler. MS settings were as follows: capillary temperature: 300
C, spray voltage: 4.5 kV, sheath gas flow: 35 and auxiliary gas: 5 (arbitrary units), capillary voltage: 25 V, and tube lens offset: 100 V. The mass spectral data were obtained in the negative ion mode between m/z 50 and 800. Electron paramagnetic resonance (EPR) experiments were performed with a Bruker EMX-10/12 spectrometer. The settings for the EPR spectrometer were as follow: center field: 3480 G; sweep width: 200.0 G; microwave frequency: 9.785 GHz; temperature 296 K, microwave power: 20 mW, field modulation: 0.1 mT at 100 kHz, scan time: 83.8 s. The light source for EPR determination was a 180 W mercury lamp. Quartz capillary tubes with an inner diameter of 1 mm were used in the UV irradiation experiments. 2.5. Toxicity assay The antimicrobial activities of the two tetracyclines and CBZ during the photodegradation process in pure water were evaluated by the bacterial growth assay. The standard test bacterial species, Bacillus subtilis, was grown in LB broth. The assays were conducted in a 96 well microtiter plate. Each well contained 60 mL bacterial suspension (103 CFU/well), 40 mL samples, and 200 mL LB medium with a final volume of 300 mL. The plates were incubated at 32
C. The optical density at 550 nm (O.D. 550 nm) was measured using a microtiter reader (Synergy H1M microplate reader). At the selected photoreaction time, a specific amount of reaction solution was

G. Zhang et al. / Environmental Pollution 225 (2017) 691e699

693

transferred to the microplate for the toxicity assay. Prior to analysis, each sample solution was diluted two times, and the pH was adjusted to 7.0 with 1 M NaOH for better growth of bacteria. During the 11 h incubation period, the light absorption at 550 nm (O.D. 550 nm) was measured at about one hour interval using a microplate reader. In the case of the UV/H2O2 process, the excess H2O2 in the solution was removed using catalase. All analysis was conducted in triplicate. 3. Results and discussion 3.1. Photolysis of PPCPs in pure water 3.1.1. Photodegradation kinetics and absorption spectra The degradation of the five selected PPCPs in both the UV/AcAc and UV/H2O2 processes followed pseudo-first-order kinetics with R2 values greater than 0.98 in most cases (Fig. S2). As depicted in Fig. 1, the pseudo-first-order degradation rate constants (k1 values) for the five PPCPs in the UV process ranged from 6.2  104 min1 (CBZ) to 1.6  101 min1 (CHL). In other words, CBZ was the most difficult to be photolyzed whereas CHL could be easily decomposed by UV light. The k1 values for all the investigated PPCPs were significantly increased by the addition of H2O2. The addition of AcAc also enhanced the photodecomposition of these PPCPs except CHL and CIP. For both TC and OTC, the k1 values in the UV/AcAc process were roughly equal to those in the UV/H2O2 process, indicating that AcAc could serve as an effective photo-activator in the degradation of tetracyclines. The UV-Vis spectra of the selected PPCPs are shown in Fig. 2. The light absorption of CHL was the weakest among the five PPCPs, while its k1 in the UV process was the largest, demonstrating that the UV absorption was not predominant in the photolysis. For CHL, the quantum yield or other possible photo-sensitization process is of the essence. 3.1.2. Effect of solution pH Note that there were inflexion points in the kinetic profiles of the photodegradation of CIP (Fig. S2 and Fig. S3), which might be related to the changes of solution pH and DO (Fig. S4a). In the UV/ AcAc process, the DO was rapidly exhausted and the solution pH was reduced to 4.0. Meanwhile, in the UV and UV/H2O2 processes, the DO was merely reduced about 20% and the solution pH was stabilized at around 5.0. Fig. S4b shows the degradation kinetic profiles of CIP in the UV

Fig. 1. The k1 values of the PPCPs in the three photochemical processes. PPCPs: 80 mM, AcAc: 0.5 mM, H2O2: 0.5 mM, initial pH: 7.0, light intensity: 4.35 mW cm2 (365 nm).

Fig. 2. The UV-Vis spectra of PPCPs (a) and AcAc (b) aqueous solutions and the emission spectrum of the medium pressure mercury lamp (b). PPCPs: 80 mM, AcAc: 0.5 mM.

process at initial pHs of 5.0, 7.0 and 8.0, respectively. The degradation rate constants of CIP at pH 7.0 and 8.0 were obviously larger than that of the pH 5.0 counterpart. Considering the pKa values of CIP (pKa1 ¼ 6.2, pKa2 ¼ 8.8) (Bobu et al., 2013), there might be some relevance between the acid-base properties of the PPCPs and the pH effect. To validate this possibility, the photodegradation of the five PPCPs were conducted at two pHs (7.0, 8.0) and are shown in Fig. S5. An obvious relevance was observed between the pKa values of the PPCPs and the pH effects: the closer the pKa value to the solution pH, the greater the pH effect. For TC and OTC, their pKa values were in the range of 7.0e8.0. Therefore, a shift in their absorption spectra and an increased k1 were observed when the solution pH was increased from 7.0 to 8.0. For those PPCPs whose pKa values were far from the solution pH, there was almost no pH effect. The above results demonstrate that the protonated PPCPs was more resistant to photodegradation. This conclusion was in agreement with the finding of others (Mella et al., 2001; Vasconcelos et al., 2009; Jiao et al., 2008a, 2008b). 3.1.3. Effect of dissolved oxygen O2 is usually an important factor in photo-oxidation. The photoreactions of many PPCPs involved self-sensitized photo-oxidation process via 1O2 or other reactive oxygen species (ROS) (Boreen et al., 2003). To evaluate the role of DO in the degradation of PPCPs in the three photochemical processes, the solutions of the five PPCPs were saturated with N2 or N2O during the UV irradiation (Fig. S3 and Fig. S6). The k1 ratios for the PPCPs in the N2 or N2O-saturated solutions to the ambient control (air-equilibrated) are listed in Table 1. In the UV process, all the k1 values except CHL were increased by the deoxygenation of the solution (kN2/kair > 1). For the UV/H2O2 process, the effects of O2 were not consistent. For CBZ and CIP, the kN2/

694

G. Zhang et al. / Environmental Pollution 225 (2017) 691e699

Table 1 The k1 ratios of the PPCPs in N2 or N2O-purged aqueous solutions to those of airequilibrated solutions. PPCP

CBZ TC CIP OTC CHL

UV

UV/H2O2

UV/AcAc

kN2/kair

kN2O/kair

kN2/kair

kN2O/kair

kN2/kair

kN2O/kair

4.781 1.374 2.925 1.071 0.821

4.490 1.230 3.123 1.210 0.741

1.322 0.743 3.030 0.792 0.921

1.184 0.580 2.920 0.894 1.194

1.999 1.196 2.926 0.927 1.174

2.165 1.190 2.754 0.860 1.431

kair were larger than 1, while the situation was reversed for TC (kN2/ kair ¼ 0.743), CHL (kN2/kair ¼ 0.921), and OTC (kN2/kair ¼ 0.792). For the UV/AcAc process, the participation of DO inhibited the photolytic reactions of all the PPCPs except OTC (kN2/kair ¼ 0.927). This seems that O2 is generally an inhibitory factor in the UV/AcAc process. Interestingly, the degradation products in the N2-saturated solutions were similar to those in the N2O-saturated solutions, but the populations of these degradation products were significantly different from those of the ambient control, even though the solutions under the three conditions had similar degradation extents (Fig. S7), indicating that contributions of the major photolysis pathways in these solutions were different. N2O is known as a good scavenger for e-aq and H$ (Buxton et al., 1988). Through reactions 1 and 2, it could convert the reductive e-aq and H$ to the strongly oxidative $OH.  e aq þ N2 O þ H2 O/$OH þ OH þ N2

k2 ¼ 9:1  109 ,M1 s1 (1)

$H þ N2 O/$OH þ N2

k2 ¼ 2:1  106 ,M1 s1

(2)

The similar effects of N2O on the photodegradation of the PPCPs to those of the inert N2 demonstrate that the roles of e-aq and H$ in three photochemical processes were negligible.

3.1.4. Effect of light source Xe lamps are usually employed to simulate solar radiation in photodegradation study. As shown in Fig. S8a, there was some overlap between the absorption spectrum of AcAc and the spectral irradiance of the Xe lamp whereas the overlap between the absorption spectrum of H2O2 and the spectral irradiance of the Xe lamp was limited. In the UVC band (<280 nm), the molar absorption coefficient of AcAc was about three orders of magnitude larger than that of H2O2. With the Xe lamp as the light source, the k1 values of CBZ in the three processes (Fig. S8b) were much lower than the corresponding k1 values with the medium pressure mercury lamp as the irradiation source (Fig. S2). However, the Xe/AcAc process became more efficient than the Xe/H2O2 process. Therefore, although the low wavelength UV light that reaches the Earth's surface is a small fraction (less than 5%) of the solar radiation, AcAc would exert a considerable influence on the pharmaceuticals. As for the different effects of H2O2 and AcAc on the photodegradation of these PPCPs, a quantitative structure-activity relationship is warranted for the further explanation. Quantum chemical calculation and collection of sufficient experimental data for statistical analysis are on-going in our group. The work here preliminarily demonstrates that AcAc also had an obvious influence on some colorless compounds. Considering the natural occurrence of diketones (White and Wainwright, 1975) and the wide existence of PPCPs, the results here point out a possible pathway in the fate and transport of PPCPs.

3.2. Photolysis of PPCPs in natural water As aforementioned, the UV/AcAc process has proven to be a promising approach for the treatment of dyeing wastewater (Zhang et al., 2014b). The introduction of AcAc into water treatment might lead to the residue of AcAc. Considering that AcAc had an enhancing effect on the photodegradation of TCs, the effects of water matrices on the photodegradation of TC and OTC were further studied by dissolving the PPCPs into three natural waters from Jiuxiang River, Xuanwu Lake, and Yangtze River. There was no significant difference among the three natural waters in terms of pH, DO, and ORP. The TOC of the Jiuxiang River water was nearly doubled to those of the other two water samples (Table S2), indicating a higher DOM content in Jiuxiang River water. As a result, the Jiuxiang River water had a stronger UV absorption band in the wavelength range of 230e400 nm (Fig. S9). As indicated by the EC and IC values, the ionic strength of Jiuxiang River water was also higher than other two samples (Table S2). DOM, bicarbonates, nitrates, and chlorides in natural waters have been reported as photosensitizers in indirect photolysis (Ge et al., 2010; Lin and Reinhard, 2005; Pestotnik et al., 2014). When exposed to sunlight, these substrates may produce reactive species, react with target pollutants in surface waters, and eventually affect the photolysis rate and environmental occurrence of variant pollutants (Pestotnik et al., 2014; Giokas et al., 2007; Santos et al., 2012; Wang and Lin, 2014). Similar to the experiments in pure water, the degradation kinetics of TC and OTC could be described by the pseudo-first-order model (Fig. 3). The k1 values in the control experiments (in the absence of H2O2/AcAc) with the three natural waters as the background matrices were about 2e3 times larger than their counterparts in pure water and there was no distinct difference among the three samples. The addition of H2O2 had nearly no effect on the k1 values, suggesting that the formed $OH radicals in the UV/H2O2 process might be completely scavenged by the DOM and inorganic ions under the given conditions. Interestingly, the k1 values in the UV/AcAc process were initially slightly larger than those in the UV/H2O2 process, whereas the k1 became smaller in the latter stage for all the three natural water samples (Fig. 3). The inflexion between the two stages generally appeared at ln(C/C0) lower than 1. Such a two-stage kinetics has been observed in the UV/AcAc process for the decolorization of dyes (Liu et al., 2014; Wang et al., 2013). Because of the formation of acidic intermediates, such as acetic acid and pyruvic acid, the solution pH in the UV/AcAc system was rapidly decreased with the increase of irradiation time (Zhang et al., 2014b). The enol form of AcAc was regarded as the crucial activator in the UV/AcAc process (Zhang et al., 2014b; Song et al., 2014). Low solution pH is favorable for the enol form in the keto-enol tautomerization. Therefore, the reduction of solution pH was believed to play an important role in the self-acceleration in the UV/AcAc process. In pure water, there was indeed a rapid decrease of solution pH in the UV/AcAc process (Fig. 4a). The reduction of solution pH was somewhat inhibited in natural waters due to the buffering capacity of DOM and inorganic salts. Therefore, the inflection in the photodecomposition profile of TCs in the UV/AcAc process with natural waters as the background matrices (Fig. 4) could not be explained by the evolution of solution pH. The excited DOM (DOM*) could be quenched by DO through either energy transfer to generate singlet molecular oxygen (1O2) or through charge transfer to form superoxide (O$2 ), which are reactive species to pollutants (Zepp et al., 1985; Wenk et al., 2011). The increased k1 values in the natural waters (as compared to those in pure water) suggest that DOM might play an important role in the photoconversion of TCs. The presence of AcAc significantly

G. Zhang et al. / Environmental Pollution 225 (2017) 691e699

695

Fig. 3. Photodegradation profiles of OTC (a) and TC (b) in three natural waters. Scatter: experimental data, line: linear fitting. OTC: 80 mM, TC: 80 mM, AcAc: 0.5 mM, H2O2: 0.5 mM.

Fig. 4. The evolution of solution pH (a) and DO (b) in the photodegradation of TC in pure water (DI) and Jiuxiang River water (JR).

accelerated the consumption of DO (Fig. 4b). As a result, in the latter stage of the UV/AcAc process, DOM lost its photosensitization effect due to the depletion of DO. Consequently, the photolytic degradation of TC and OTC was retarded.

3.3. Photolysis mechanism of TC in the UV/AcAc process AcAc had a strong absorption cross section in the UV band

(Fig. 2). On one side, AcAc might compete with PPCPs for photons, leading to an inner filter effect. On the other side, AcAc might enhance the generation of 1O2 or other ROS. Furthermore, photoinduced energy and electron transfer between excited AcAc and PPCPs might occur, resulting in an enhanced decomposition of PPCPs. To fully understand the effects of AcAc on the photoconversion of PPCPs, TC was selected as a representative for further discussion.

696

G. Zhang et al. / Environmental Pollution 225 (2017) 691e699

3.3.1. Inner filter effect The inner filter effects caused by AcAc and H2O2 were calculated as the ratio of the number of photons absorbed by the solution in the presence of co-existing matter to that of the control (Text S1). The ratio of absorbed photons was defined as correction factor (CF). The smaller the CF was, the stronger was the inner filter effect. As shown in Table S3, the CF values for the UV/H2O2 process were close to 1, indicating that the inner filter effect caused by H2O2 was negligible. By the contrast, the CF values for the UV/AcAc process were obviously less than 1, especially for CBZ (CF ¼ 0.834) and CHL (CF ¼ 0.845), indicating that there existed significant inner filter effects. By dividing the k1 with CF, the obtained k10 could be used to reflect the effects from the factors other than the inner filter effect. As listed in Table S3, the corrected k1 values for the UV/H2O2 process (k1,UV/H2O20 ) and the UV/AcAc process (k1,UV/AcAc’) were all larger than the k1 values in the UV process (k1,UV) with the exception of CHL and CIP. The k1,UV/AcAc’ for CIP was nearly the same as the k1,UV value, suggesting that the inner filter effect was exclusively accountable for the negative effect of AcAc on the photodegradation of CIP. For CHL, AcAc not only led to an inner filter effect but also interfered with the photodegradation of CHL through other ways, such as competing with CHL for ROS. 3.3.2. Role of 1O2 Miskoski et al. (1998) reported that upon direct photoirradiation TCs could generate 1O2 with a quantum yield of 0.03.1O2 has a longer lifetime in D2O than in H2O (Rodgers and Snowden, 1982). Surprisingly, the photodegradation of TC was retarded by using D2O as the solvent (Fig. 5a). This result suggests that the kinetic isotope effect (mainly on the excited states) (Shafirovich et al., 1995) dominate over the positive role of 1O2. Replacing H2O with D2O only slightly enhanced the photodegradation in the UV/AcAc process (Fig. 5a). Furthermore, with TEMP as a spin trap for 1O2, no signal except the background noise was observed for the UV-irradiated AcAc solution (Fig. S10), indicating that 1O2 was not the decisive species in the UV/AcAc process. 3.3.3. Role of O2·NBT is a typical indicator of superoxide anion (O2$-). It reacts with O2$- to produce NBTþ$ and O2 (reaction 3). The subsequent disproportionation of two NBTþ$ results in one stable intermediate form (reaction 4), monoformazan (MFþ), with an absorption peak at 530 nm (Bielski et al., 1980). þ$ NBT2þ þ O$ þ O2 2 /NBT

(3)

NBTþ$ þ NBTþ$ þ H/MFþ þ NBT2þ

(4) -

In aqueous solutions, KO2 could directly generate O2$ and react with NBT to form MFþ. As shown in Fig. 5b, TC generated a considerable amount of O2$- upon UV irradiation. The addition of AcAc to the TC solution, to some extent, reduced the concentration of O2$-. These results demonstrate that O2$- might play an important role in the photodegradation of TC and the addition of AcAc consumed the available O2$-. Even though, the k1 of TC in the UV/ AcAc process was still much higher than that in the UV process (Fig. 1), suggesting that other pathways dominated in the UV/AcAc process. 3.3.4. Role of other ROS BD, the simplest a-diketone, can readily break into acetyl radicals under UV irradiation and further generate acetylperoxyl radicals through reaction with O2 (Neevel et al., 1990, 1992). The photodegradation of TC in the UV/AcAc and UV/BD processes with or without N2-purging are shown in Fig. S11. BD enhanced the degradation of TC as AcAc did. However, the effect of N2 on the UV/ BD process was completely different from that of the UV/AcAc process. As discussed above, N2 purging led to an increased k1 of TC in the UV/AcAc process (Table 1), whereas the enhancement effect caused by BD completely disappeared under N2 purging (Fig. S11). In our previous work (Zhang et al., 2014b), the generation of acetyl radicals and acetylperoxyl radicals in the UV/AcAc process was found much weaker than that in the UV/BD process. The N2 effects here together with the previous EPR results (Zhang et al., 2014b) demonstrate that the role of other ROS, such as acetyl radicals and acetylperoxyl radicals, could be excluded as the main mechanism in the UV/AcAc process. These results also indicate that other diketones, not just AcAc, might impact the photochemical behaviors of PPCPs in aquatic environment. 3.3.5. Degradation products and pathway In the literature, the photodegradation of TC is believed as a free radical process with the photodissociation of the 4-dimethylamino group as the first step, which then reacts rapidly with oxygen (formed peroxyl radical) or water (H atom transfer) under anoxic conditions, and then undergoes H2O elimination to form a red product which has a quinoid structure (Davies et al., 1979; Moore et al., 1983). The degradation products of TC in the UV and UV/ AcAc processes were identified by LC-MS. As listed in Table S4, besides these reported products (P1-P4 in Scheme S1), several more products of higher mass to charge ratio (m/z: 446, 458, 474, 479, 481 and 483) than that of the TC (443) were observed in the

Fig. 5. (a) Degradation profiles of TC in the UV and UV/AcAc processes with H2O or D2O as the solvent. (b) The UV-Vis spectra of NBT solutions in the presence of KO2, TC and/or AcAc. NBT: 0.2 mM, AcAc: 0.5 mM, TC: 80 mM, KO2: 80 mM, irradiation time: 40 min.

G. Zhang et al. / Environmental Pollution 225 (2017) 691e699

UV/AcAc system, which might be the decomposition products from the TC-AcAc complexes. As aforementioned, the presence of AcAc enhanced the photodegradation of TC (k1 of TC was increased from 0.0200 min1 to 0.0545 min1, Fig. 6a). Meanwhile, under the otherwise identical conditions, the presence of TC significantly inhibited the decomposition of AcAc (k1 of AcAc was decreased from 0.0533 min1 to 0.01138 min1, Fig. 6b). By taking the inner filter effect caused by TC (CF ¼ 0.503) into account, the k1’ of AcAc in the presence of TC (0.0274 min1) was still only half of that in the AcAc-only solution. These results, together with the solvent kinetic isotope effect (Fig. 5a), suggest that there might be energy transfer from the excited AcAc to TC. O2 is a quenching agent for excited triplets (Gijzeman et al., 1973; Abdel-Shafi and Wilkinson, 2000). The inhibiting effects of DO on the photodegradation of the four PPCPs, as listed in Table 1, strongly indicate the role of the triplet-excited states in the transformation of these PPCPs. Based on the above discussions, the degradation pathways for TC in the UV/AcAc process are schematically shown in Scheme 1. O2 could either quench the triplet TC (TC*) through energy dissipation or through electron transfer to form O2$-. The roles of AcAc include energy transfer to TC (I), formation of exciplexes (II), and consumption of the generated O2$- (III). Pathways I and II were positive to the degradation of TC whereas pathway III led to a negative effect. Therefore, the relative contributions of the three pathways (IIII) determined the observed effect. If the consumption of O2$dominated over energy transfer, the k1 of PPCPs in the UV/AcAc process would be lower than that of the UV process, as what we observed in the photodegradation of CIP and CHL. The calculation of excited states might be helpful to explain the differences among the five tested PPCPs. 3.4. Toxicity of degradation products in the irradiation processes The toxicity of certain pollutants might be increased during oxidation processes due to the combined effect of all the products in solution (Jiao et al., 2008a, 2008b; Lopez-Penalver et al., 2010). For example, Jiao et al. (2008a) examined the photolysis of TC and evaluated the toxicity of the photolysis products with luminescent bacteria. Their results revealed that the toxicity of the irradiated solution was increased. On the contrary, Mboula et al. (2012) reported that the toxicity of TC and its products was reduced after photocatalysis through a dehydrogenase inhibition test. Palominos et al. (2009) found that TC oxidation products did not present antibacterial activity against Staphylococcus aureus. These contradictions suggest that the toxicity of products depends on the oxidation process due to the possible different reaction

697

Scheme 1. The roles of AcAc and O2 in the photodegradation of TC.

mechanisms involved, and also on the organism used in the toxicity test. The toxicity of the two TCs and their degradation products during the three photochemical processes in pure water was determined with the traditional bacterial growth method. As shown in Fig. S12, AcAc at the given concentration had no effect on the growth of Bacillus subtilis. Although H2O2 could inhibit the growth of Bacillus subtilis, the addition of catalase completely removed the residue of H2O2. Therefore, the growth profiles of Bacillus subtilis (Fig. 7) exactly reflected the toxicity evolutions of the examined solutions. Both TC and OTC significantly inhibited the growth of Bacillus subtilis (Fig. 7a and b). The direct UV irradiation only led to a slight reduction in the inhibiting effect (Fig. 7a and b). After treatment with either the UV/H2O2 or UV/AcAc process, the inhibition of bacteria's growth was significantly reduced (Fig. 7c and d). Similar results were observed for the irradiated CBZ solutions (Fig. S13). In our previous work, the toxicity of irradiated dye solutions was evaluated with a phytotoxicity study using rice seeds as the probing target (Zhang et al., 2014b). The toxicity of the UV/AcAc treated solution was comparable to or even lower than that of the UV/H2O2 counterpart. Besides, the degradation products of AcAc had good biocompatibility (Wu et al., 2016). These results demonstrate that although AcAc worked in a different way to that of H2O2 in the photoconversion of PPCPs, the toxicity evolution caused by the UV/

Fig. 6. Photodegradation of AcAc (a) and TC (b) in single and binary solutions. AcAc: 0.5 mM, TC: 80 mM, light intensity: 4.53 mW cm2 (365 nm).

698

G. Zhang et al. / Environmental Pollution 225 (2017) 691e699

Fig. 7. The growth profiles of Bacillus subtilis bacteria in (a) UV treated OTC, (b) UV treated TC, (c) UV/H2O2 or UV/AcAc treated OTC, and (d) UV/H2O2 or UV/AcAc treated TC solutions. Assays were conducted in triplicate.

AcAc process was comparable to that in the UV/H2O2 process.

4. Conclusions As emerging environmental pollutants, PPCPs have attracted a great deal of concern. Photochemical reactions in surface waters are likely a major pathway in the fate of many PPCPs. The results in this work demonstrated that AcAc, like H2O2, was also able to affect the photolysis of PPCPs. When natural waters were used as the background matrices, the photodecomposition profiles of TCs in the UV/ AcAc and UV/H2O2 processes were obviously different, indicating that the underlying mechanisms were different. AcAc could accelerate the photodecomposition of three of the five selected PPCPs whereas it slightly inhibited the photoconversion of CHL and CIP. Interestingly, among the five PPCPs, CHL was the most reactive one in direct photolysis whereas its light absorption was the weakest. The inhibiting effect of O2 in the photodegradation of PPCPs strongly suggests that the excited triplet state play a key role in the transformation of PPCPs. AcAc interacted with TC through direct energy transfer rather than ROS. A detailed analysis of the degradation pathways and the establishment of a quantitative-structure activity relationship are warranted for a thorough understanding of the photodegradation of PPCPs in natural waters.

Acknowledgements The work was supported by the National Natural Science Foundation of China (21522702) and the Fundamental Research Funds for the Central Universities.

Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.envpol.2017.01.089.

References Abdel-Shafi, A.A., Wilkinson, F., 2000. Charge transfer effects on the efficiency of singlet oxygen production following oxygen quenching of excited singlet and triplet states of aromatic hydrocarbons in acetonitrile. J. Phys. Chem. A 104, 5747e5757. Andreozzi, R., Canterino, M., Lo Giudice, R., Marotta, R., Pinto, G., Pollio, A., 2006. Lincomycin solar photodegradation, algal toxicity and removal from wastewaters by means of ozonation. Water Res. 40, 630e638. Benotti, M.J., Trenholm, R.A., Vanderford, B.J., Holady, J.C., Stanford, B.D., Snyder, S.A., 2009. Pharmaceuticals and endocrine disrupting compounds in us drinking water. Environ. Sci. Technol. 43, 597e603. Bielski, B.H.J., Shiue, G.G., Bajuk, S., 1980. Reduction of nitro blue tetrazolium by CO-2 and O-2 radicals. J. Phys. Chem. 84, 830e833. Bobu, M., Yediler, A., Siminiceanu, I., Zhang, F., Schulte-Hostede, S., 2013. Comparison of different advanced oxidation processes for the degradation of two fluoroquinolone antibiotics in aqueous solutions. J. Environ. Sci. Health Part aToxic/Hazardous Subst. Environ. Eng. 48, 251e262. € rklund, E., Krogh, K.A., Hansen, M., 2013. Abiotic degradation Bohn, P., Bak, S.A., Bjo of antibiotic ionophores. Environ. Pollut. 182, 177e183. Boreen, A.L., Arnold, W.A., McNeill, K., 2003. Photodegradation of pharmaceuticals in the aquatic environment: a review. Aquat. Sci. 65, 320e341. Boyle, L.L., McCullough, N.H., 25e30 August 1996. 212th National Meeting of the American Chemical Society Division of Environmental Chemistry. Orlando, Florida. Preprints extended abstract 36, 4. Brezonik, P.L., Arnold, W.A., 2011. Water chemistry: an introduction to the chemistry of natural and engineered aquatic systems. In: Part III: Inorganic Chemical Equilibria and Kinetics. Oxford University Press, p. 347. Budavari, S., O'Neil, M.J., Smith, A., Heckelman, P.E., Obenchain Jr., J.R., Gallipeau, J.A.R., D'Arecea, M.A., 2001. The Merck Index, an Encyclopedia of Chemicals, Drugs, and Biologicals, 13th ed. Merck & Co., Inc., White house Station, NJ. Buxton, G.V., Greenstock, C.L., Helman, W.P., Ross, A.B., 1988. Critical review of rate constants for reactions of hydrated electrons, hydrogen atom, and hydroxyl radicals ($OH/$O-) in aqueous solutions. J. Phys. Chem. Reference Data 17, 513e886. Costanzo, S.D., Murby, J., Bates, J., 2005. Ecosystem response to antibiotics entering the aquatic environment. Mar. Pollut. Bull. 51, 218e223. Davies, A.K., McKellar, J.F., Phillips, G.O., Reid, A.G., 1979. Photochemical oxidation of tetracycline in aqueous solution. J. Chem. Soc. Perkin Trans. 2, 369e375. Dodgen, L.K., Li, J., Parker, D., Gan, J.J., 2013. Uptake and accumulation of four PPCP/ EDCs in two leafy vegetables. Environ. Pollut. 182, 150e156. Gao, P., Ding, Y., Li, H., Xagoraraki, I., 2012. Occurrence of pharmaceuticals in a municipal wastewater treatment plant: mass balance and removal processes. Chemosphere 88, 17e24. Ge, L.K., Chen, J.W., Wei, X.X., Zhang, S.Y., Qiao, X.L., Cai, X.Y., Xie, Q., 2010. Aquatic photochemistry of fluoroquinolone antibiotics: kinetics, pathways, and

G. Zhang et al. / Environmental Pollution 225 (2017) 691e699 multivariate effects of main water constituents. Environ. Sci. Technol. 44, 2400e2405. Gijzeman, O.L., Kaufman, F., Porter, G., 1973. Oxygen quenching of aromatic tripletstates in solution. Part 1. J. Chem. Soc. Faraday Trans. 2 Mol. Chem. Phys. 69, 708e720. Giokas, D.L., Salvador, A., Chisvert, A., 2007. UV filters: from sunscreens to human body and the environment. Trends Anal. Chem. 26, 360e374. Gobel, A., Thomsen, A., McArdell, C.S., Joss, A., Giger, W., 2005. Occurrence and sorption behavior of sulfonamides, macrolides, and trimethoprim in activated sludge treatment. Environ. Sci. Technol. 39, 3981e3989. Jiao, S.J., Meng, S.R., Yin, D.Q., Wang, L.H., Chen, L.Y., 2008a. Aqueous oxytetracycline degradation and the toxicity change of degradation compounds in photoirradiation process. J. Environ. Sciences-China 20, 806e813. Jiao, S.J., Zheng, S.R., Yin, D.Q., Wang, L.H., Chen, L.G., 2008b. Aqueous photolysis of tetracycline and toxicity of photolytic products to luminescent bacteria. Chemosphere 73, 377e382. Kim, S.C., Carlson, K., 2007. Temporal and spatial trends in the occurrence of human and veterinary antibiotics in aqueous and river sediment matrices. Environ. Sci. Technol. 41, 50e57. Kim, I., Yamashita, N., Tanaka, H., 2009. Photodegradation of pharmaceuticals and personal care products during UV and UV/H2O2 treatments. Chemosphere 77, 518e525. Kuemmerer, K., 2009a. Antibiotics in the aquatic environment - a review - Part I. Chemosphere 75, 417e434. Kuemmerer, K., 2009b. The presence of pharmaceuticals in the environment due to human use - present knowledge and future challenges. J. Environ. Manag. 90, 2354e2366. Li, W.C., 2014. Occurrence, sources, and fate of pharmaceuticals in aquatic environment and soil. Environ. Pollut. 187, 193e201. Li, Z., Xiang, X., Li, M., Ma, Y.P., Wang, J.H., Liu, X., 2015. Occurrence and risk assessment of pharmaceuticals and personal care products and endocrine disrupting chemicals in reclaimed water and receiving groundwater in China. Ecotoxicol. Environ. Saf. 119, 74e80. Lin, A.Y.C., Reinhard, M., 2005. Photodegradation of common environmental pharmaceuticals and estrogens in river water. Environ. Toxicol. Chem. 24, 1303e1309. Liu, X.T., Song, X.J., Zhang, S.J., Wang, M.S., Pan, B.C., 2014. Non-hydroxyl radical mediated photochemical processes for dye degradation. Phys. Chem. Chem. Phys. 16, 7571e7577. Lopez-Penalver, J.J., Sanchez-Polo, M., Gomez-Pacheco, C.V., Rivera-Utrilla, J., 2010. Photodegradation of tetracyclines in aqueous solution by using UV and UV/ H2O2 oxidation processes. J. Chem. Technol. Biotechnol. 85, 1325e1333. Mboula, V.M., Hequet, V., Gru, Y., Colin, R., Andres, Y., 2012. Assessment of the efficiency of photocatalysis on tetracycline biodegradation. J. Hazard. Mater. 209, 355e364. Mella, M., Fasani, E., Albini, A., 2001. Photochemistry of 1-cyclopropyl-6-fluoro-1, 4dihydro-4-oxo-7-(piperazin-1-yl) quinoline-3-carboxylic acid (¼ ciprofloxacin) in aqueous solutions. Helvetica Chim. Acta 84, 2508e2519. Miskoski, S., Sanchez, E., Garavano, M., Lopez, M., Soltermann, A.T., Garcia, N.A., 1998. Singlet molecular oxygen-mediated photo-oxidation of tetracyclines: kinetics, mechanism and microbiological implications. J. Photochem. Photobiol. B-Biology 43, 164e171. Moore, D.E., Fallon, M.P., Burt, C.D., 1983. Photooxidation of tetracycline - a differential pulse polarographic study. Int. J. Pharm. 14, 133e142. Musolff, A., Leschik, S., Moder, M., Strauch, G., Reinstorf, F., Schirmer, M., 2009. Temporal and spatial patterns of micropollutants in urban receiving waters. Environ. Pollut. 157, 3069e3077. Neevel, J.G., Vanbeek, H.C.A., Denouden, H.H.I., Vandegraaf, B., 1990. Photooxidation of azo dyes in the presence of biacetyl and oxygen. J. Soc. Dye. Colour. 106, 176e181. Neevel, J.G., Vanbeek, H.C.A., Vandegraaf, B., 1992. Products and intermediates

699

formed during the biacetyl-sensitized photooxidation of some azo dyes. J. Soc. Dye. Colour. 108, 489e492. Palominos, R.A., Mondaca, M.A., Giraldo, A., Penuela, G., Perez-Moya, M., Mansilla, H.D., 2009. Photocatalytic oxidation of the antibiotic tetracycline on TiO2 and ZnO suspensions. Catal. Today 144, 100e105. Pestotnik, K., Kosjek, T., Heath, E., 2014. In: Lambropoulou, D.A., Nollet, L.M. (Eds.), Transformation Products of Personal Care Products: UV Filters Case Studies, in Transformation Products of Emerging Contaminants in the Environment: Analysis, Processes, Occurrence, Effects and Risks. Wiley, pp. 471e504. Rastogi, S.C., 1991. Levels of organic solvents in printer's inks. Archives Environ. Contam. Toxicol. 20, 543e547. Rodgers, M.A.J., Snowden, P.T., 1982. Lifetime of O2 (1Dg) in liquid water as determined by time-resolved infrared luminescence measurements. J. Am. Chem. Soc. 104, 5541e5543. Santos, A.J.M., Miranda, M.S., Esteves da Silva, J.C.G., 2012. The degradation products of UV filters in aqueous and chlorinated aqueous solutions. Water Res. 46, 3167e3176. Schmitt, H., Van Beelen, P., Tolls, J., Van Leeuwen, C.L., 2004. Pollution-induced community tolerance of soil microbial communities caused by the antibiotic sulfachloropyridazine. Environ. Sci. Technol. 38, 1148e1153. Shafirovich, V.Y., Courtney, S.H., Ya, N.Q., Geacintov, N.E., 1995. Proton-coupled photoinduced electron-transfer, deuterium isotope effects, and fluorescence quenching in noncovalent benzo[a]pyrenetetraol-nucleoside complexes in aqueous solutions. J. Am. Chem. Soc. 117, 4920e4929. Song, X.J., Wu, B.D., Zhang, S.J., 2014. Decoloration of alizarin red (an anthraquinone dye) with the UV/acetylacetone process. Acta Chim. Sin. 72, 461. Sun, Q., Li, M.Y., Ma, C., Chen, X.Q., Xie, X.Q., Yu, C.P., 2016. Seasonal and spatial variations of PPCP occurrence, removal and mass loading in three wastewater treatment plants located in different urbanization areas in Xiamen, China. Environ. Pollut. 208, 371e381. Vasconcelos, T.G., Henriques, D.M., Konig, A., Martins, A.F., Kummerer, K., 2009. Photo-degradation of the antimicrobial ciprofloxacin at high pH: identification and biodegradability assessment of the primary by-products. Chemosphere 76, 487e493. Wang, X.H., Lin, A.Y.C., 2014. Is the phototransformation of pharmaceuticals a natural purification process that decreases ecological and human health risks? Environ. Pollut. 186, 203e215. Wang, M.S., Liu, X.T., Pan, B.C., Zhang, S.J., 2013. Photodegradation of acid orange 7 in a UV/acetylacetone process. Chemosphere 93, 2877e2882. Wenk, J., von Gunten, U., Canonica, S., 2011. Effect of dissolved organic matter on the transformation of contaminants induced by excited triplet states and the hydroxyl radical. Environ. Sci. Technol. 45, 1334e1340. White, F.H., Wainwright, T., 1975. Occurrence of diketones and a-acetohydroxyacids in fermentation. J. Inst. Brew. 81, 46e52. Wu, B.D., Zhang, G.Y., Zhang, S.J., 2016. Fate and implication of acetylacetone in photochemical processes for water treatment. Water Res. 101, 233e240. Zepp, R.G., Schlotzhauer, P.F., Sink, R.M., 1985. Photosensitized transformations involving electronic-energy transfer in natural waters role of humic substances. Environ. Sci. Technol. 19, 74e81. Zhang, D.Q., Gersberg, R.M., Ng, W.J., Tan, S.K., 2014a. Removal of pharmaceuticals and personal care products in aquatic plant-based systems: a review. Environ. Pollut. 184, 620e639. Zhang, S.J., Liu, X.T., Wang, M.S., Wu, B.D., Pan, B.C., Yang, H., Yu, H.Q., 2014b. Diketone-mediated photochemical processes for target-selective degradation of dye pollutants. Environ. Sci. Technol. Lett. 1, 167e171. Zhou, S., Barnes, I., Zhu, T., Bejan, I., Albu, M., Benter, T., 2008. Atmospheric chemistry of acetylacetone. Environ. Sci. Technol. 42, 7905e7910. Zhou, L.X., Zhu, D.X., Zhang, S.J., Pan, B.C., 2015. A settling curve modeling method for quantitative description of the dispersion stability of carbon nanotubes in aquatic environments. J. Environ. Sciences-China 29, 1e10.