Study of the simulated sunlight photolysis mechanism of ketoprofen: the role of superoxide anion radicals, transformation byproducts, and ecotoxicity assessment

Yingfei Wang a, Wen Deng b, Fengliang Wang a, Yuehan Su a, Yiping Feng a, Ping Chen a, Jingshuai Ma a, Haiying Su a, Kun Yao a, Yang Liu c, Wenying Lv a and Guoguang Liu *a
aSchool of Environmental Science and Engineering, Guangdong University of Technology, No. 100 Waihuan Xi Road, Guangzhou Higher Education Mega Center, Panyu District, Guangzhou 510006, China. E-mail: liugg615@163.com; Fax: +86-20-39322548; Tel: +86-20-39322547
bGuangzhou Municipal Engineering Design & Research Institute, Guangzhou 510098, China
cFaculty of Environmental & Biological Engineering, Guangdong University of Petrochemical Technology, Maoming, 525000, China

Received 8th March 2017 , Accepted 27th June 2017

First published on 4th July 2017


The aim of this study was to investigate the photolysis mechanism of ketoprofen (KET) under simulated sunlight. The results demonstrated that the photolysis of KET aligned well with pseudo first-order kinetics. Radical scavenging experiments and dissolved oxygen experiments revealed that the superoxide anion radical (O2˙) played a primary role in the photolytic process in pure water. Bicarbonate slightly increased the photodegradation of KET through generating carbonate radicals, while DOM inhibited the photolysis via both attenuating light and competing radicals. Moreover, Zhujiang river water inhibited KET phototransformation. Potential KET degradation pathways were proposed based on the identification of products using LC/MS/MS and GC/MS techniques. The theoretical prediction of reaction sites was derived from Frontier Electron Densities (FEDs), which primarily involved the KET decarboxylation reaction. The ecotoxicity of the treated solutions was evaluated by employing Daphnia magna and V. fischeri as biological indicators. Ecotoxicity was also hypothetically predicted through the “ecological structure–activity relationship” (ECOSAR) program, which revealed that toxic products might be generated during the photolysis process.



Environmental impact

Ketoprofen (KET, 2-(3-benzoylphenyl)-propionic acid), as the most popular medicine of Pharmaceutical and Personal Care Products (PPCPs), has been frequently found in aqueous environments and has undesired photoallergic effects on human skin, such as hypersensitivity and myasthenia gravis. Photodegradation is believed to be a critical factor that influences the environmental fates of PPCPs in ambient surface waters. Moreover, the incomplete mineralization of PPCPs during the photolytic process typically leads to the generation of toxic byproducts. This study demonstrated that KET was quite unstable under exposure to simulated light radiation, and some toxic products were generated during the photolysis process.

1. Introduction

The term “Pharmaceutical and Personal Care Products” (PPCPs), which was proposed by G Daughton Christian in 1999 (ref. 1), has become widely accepted. The vast majority of PPCP pollution sources are anthropogenically generated.2,3 PPCPs behave as persistent organic micropollutants that exhibit a “false sustained phenomenon”,4 which is attributed to their continuous input into the environment and relatively stable structure. Moreover, long-term exposure to PPCPs in the environment leads to environmental hormone toxicity (endocrine interference), genetic toxicity (mutagenesis, carcinogenesis, and teratogenesis), and microbial toxicity,5 thereby endangering human health and wildlife. Consequently, the scientific literature6–8 has focused considerable attention on PPCPs as to their potentially deleterious effects on environmental ecosystems and human health.

Ketoprofen (2-(3-benzoylphenyl)-propionic acid) is frequently used in the treatment of rheumatoid arthritis and ankylosing spondylitis, as well as postoperative pain.9 Over the last few years, significant data have revealed the concentrations of KET in various environmental ecosystems. Tokyo is regarded as having the second highest level of PPCPs (on a global scale) in its ambient waterways, with concentrations ranging from 108 to 369 ng L−1.10 Santos JL11 calculated the median loading of KET in the influent of Tokyo WWTPs to be ca. 63–372 mg (d 1000 inh)−1. Further, the loading of KET (428–808 mg (d 1000 inh)−1) in Finland was even higher than that in the WWTP influent reported in Spain and Tokyo.12 In addition, loads of 7.5 mg (d 1000 inh)−1 have been reported in Shanghai.13 It was also revealed that KET was detected in the Wei river at maximum concentrations of 31.35 ng L−1.14 It is noted that exposure to KET in aqueous environments initiates undesired photoallergic effects on human skin, such as hypersensitivity and myasthenia gravis, which are primarily due to benzophenone.9,15,16

Photodegradation is believed to be a critical factor that influences the environmental fates of PPCPs in ambient surface waters.3,17 The photodegradation of PPCPs may be described in terms of direct photolysis via the absorption of solar photons, and indirect photolysis through reactions with reactive oxygen species (ROS) (e.g., O2˙ and ˙OH) from photosensitizers or excited state organisms.18–21 As the potential mechanisms of the photolysis process are disparate, related photodegradation products and pathways might also be different.22 Moreover, the incomplete mineralization of PPCPs during the photolytic process typically leads to the generation of toxic byproducts. Hence, potential toxicity during the photodegradation process is also of great interest.23

Recent research studies have focused on the migration and transformation of KET under ultraviolet degradation. Ayako Nakajima,24 L. L Costanzo,25,26 Francisco Boscá,27 and Tina Kosjek28 investigated the degradation mechanism of KET under UV light irradiation. Several degradation products had been detected and the photodegradation pathways were proposed. However, previous research studies mainly focused on the direct photolysis mechanism of KET as well as its photoproducts under UV light irradiation, and detailed data on KET remain lacking in regard to its self-sensitization photolysis and indirect mechanism in aqueous solutions under simulated solar irradiation. In addition, L. L Costanzo25 also found that reactive oxygen species (ROS) such as ˙OH, 1O2, and O2˙ were involved in the UV photodegradation of KET. However, the contribution rates of ROS to KET photodegradation have not been discussed in their study, and hence, further studies are warranted. The ecotoxicity change of KET during solar exposure has been investigated. Seiya Hanamoto29et al. and Xiao-Huan Wang30et al. reported that the total toxicity of the KET solution for Vibrio fischeri increased under sunlight exposure. However, the toxicity of KET photoproducts has not been further assessed. Consequently, research into the photolysis mechanisms and the toxicity assessment of KET in the ambient environment under simulated sunlight irradiation is urgently required.

The primary objectives of this work were to (1) investigate the simulated sunlight photolytic mechanisms of KET in aqueous solutions; (2) investigate the influences of several water constituents (bicarbonate and DOM) and natural water on the photodegradation of KET; (3) identify the primary photodegradation products and propose possible photodegradation pathways; (4) evaluate changes in ecotoxicity during the photodegradation of KET.

2. Materials and methods

2.1 Chemicals

KET (ketoprofen) (Table S1), 2-(3-benzoylphenyl) propionic acid (98% purity), was purchased from TCI Reagent Co. Ltd. (Shanghai, China). Sodium azide (NaN3) (99% purity) was obtained from Sigma-Aldrich (U.S.). HPLC-grade reagents (acetonitrile, methanol, ethanol, etc.) were purchased from the U.S. ACS Enke Chemistry Co. Ltd. (Guangzhou, China). Other reagents (e.g., ice acetate, isopropanol, p-quinone, etc.) were of analytical grade and used without further purification. Deionized (DI) water (resistance > 18.2 MΩ), obtained from a Milli-Q apparatus (Smart2Pure ultrapure water/water system integration, TKA, Germany), was used in the preparation of all aqueous solutions during the experiment. River water was collected from Zhujiang, Guangzhou. The main properties of water samples are given in Table S2.Vibrio fischeri was obtained from the Institute of Soil Science, Chinese Academy of Sciences (Nanjing, China). Daphnia magna was obtained from Aquatic Biology Research Center, Jinan University (Guangzhou, China).

2.2 Photolysis experiments

Photolysis experiments were performed in a SGY-IIB. Y1 rotary photochemical reactor (Nanjing STO Co. Ltd., China) (Fig. S1). A 350 W Xe-lamp with a 290 nm-cut filter was used as the simulated sunlight source. The Xe-lamp and sunlight spectra as well as the absorption of KET can be observed in Fig. S2. The lamp was encased within a double-walled quartz cooling jacket and connected to a miniature air-cooling system. The reaction solution (20 mL) was introduced into a 25 mL quartz tube ([KET] = 4.0 mg L−1). Throughout the experiments, aliquots were extracted from the rotary photochemical reactor at 30 s intervals, whereafter the residual concentration was immediately analyzed via a high-performance liquid chromatography (HPLC) system. Dark controls were simultaneously conducted under equivalent conditions. Dissolved oxygen experiments were performed by using nitrogen or oxygen as the purging gas during the photolysis process. Each of the experiments was carried out at least in triplicate, with the results reported as the mean ± 95% confidence interval when available.

Radical scavenging experiments were also performed using 50 mM, 100 mM and 200 mM isopropyl alcohol (IPA), 2 mM, 5 mM and 10 mM 4-hydroxy-2,2,6,6-tetramethylpiperidinyloxy (TEMPOL), and 5 mM, 10 mM and 50 mM sodium azide (NaN3) as the radical scavengers.

2.3 Analytical methods

2.3.1 Determination of KET concentration. The concentrations of KET were analyzed by HPLC, and the chromatographic conditions of KET were as follows: column, Zorbax Eclipse XDB-C18, 2.1 × 150 mm, 3.5 μm; temperature, 40 °C; mobile phase, acetonitrile/glacial acetic acid buffer solution (45[thin space (1/6-em)]:[thin space (1/6-em)]55 v/v containing 0.5% glacial acetic acid); flow rates, 1 mL min−1; injection volume, 10 μL; and photodiode array detector, 260 nm wavelength.

The Total Organic Carbon (TOC) of the reaction solutions was measured using a SHIMADZU (TOC-VCPH). The combustion tube was filled with the platinum catalyst (Shimadzu) for the determination of total carbon (kept at 953 K). High purity synthetic air, at a flow rate of 150 mL min−1, and ultrapure Millipore water (Milk-RX 20, Milliq 185) were used. The standard used for the determination of total carbon was potassium acid phthalate p.a. (KHC8H404, Shimadzu), Na2CO3 p.a. and NaHCO3 p.a. for inorganic carbon. Each sample was measured at least three times, whereafter the average value of each sample was determined.

2.3.2 Identification of photolytic by-products. Intermediate photodegradation products were identified using liquid chromatography via tandem mass spectrometry (HPLC-MS/MS). The KET solution (4 mg L−1) was prepared for conducting the photolysis experiments for 180 s, after which the solution was concentrated to 2 mL in a rotary evaporator (RE-5299, YU HUA INSTRUMENT, China), which was subsequently transferred to a sample vial for characterization by HPLC-MS/MS. A Waters AQUITY/UPLC/Q-TOF microsystem (Waters Corporation), which was equipped with a BEH-C18 column (50 × 2.1 mm, 1.7 μm), was used to separate the samples. The eluent, at a flow rate of 0.3 mL min−1, utilized a mixture of 0.5% glacial acetic acid (A) and acetonitrile (B), employing a linear gradient from 45% B to 100% B in 20 min. Mass spectral analysis was conducted in negative mode using an electrospray ionization (ESI) source. The instrumentation parameters were as follows: capillary voltage 3.0 kV, temperature 300 °C, nebulizer pressure 40 psi, and argon (P99.99%), which was used as a collision gas.

The intermediates were also detected by GC/MS (AGILENT 6890-GC/5973i-MS) analyses. GC separation was conducted using a DB-5 MS capillary column (30 m length × 0.25 mm × 0.25 μm). The GC equipment was operated in a temperature programmed mode with an initial temperature of 60 °C, which was then ramped to 270 °C at a 10 °C min−1 rate; electron impact (EI) mass spectra were scanned from 0 to 300 m/z, and helium was used as the carrier gas.

2.4 Ecotoxicity assay

2.4.1 Acute toxicity test. Toxicity tests were performed using the aquatic organism Daphnia magna Straus according to OECD31 guidelines. Sample solutions with 4 mg L−1 KET following irradiation at 0, 30, 60, 90, 120 and 180 s were tested, and a control using only fresh water was also tested. After 24 and 48 h of incubation, the number of dead and immobilized D. magna was calculated. The toxicity was expressed by the D. magna immobilization rate.

The toxicity variations during the photodegradation using V. fischeri as the biological indicator were investigated via a Microtox Model DXY-2 Toxicity Analyzer,32 which evaluated the ability of the photoproducts to inhibit the bioluminescence of the bacterium V. fischeri.33 The luminescence inhibition rate (I%) was calculated as follows: (I = luminescence) eqn (1), where I1 and I0 are the luminosity of the sample and the blank solution, respectively.

 
image file: c7em00111h-t1.tif(1)

2.4.2 Toxicity estimation by ECOSAR. For the purposes of this exercise, the ECOSAR program was employed for predicting the theoretical toxicities of KET and its photolytic transformation products. This software was developed by the US Environmental Protection Agency (EPA) and provides public access to the same methods that are routinely employed by the EPA for evaluating aquatic toxicity, i.e. the method is already being used in a regulatory context.34 The input MOL files of KET and its photolytic transformation products were built by Gaussian 09.

The acute toxicity values at three levels (fish, daphnia, and green algae) of KET and its transformation byproducts were presented as LC50 and EC50. The chronic toxicity values (ChV) of KET and its transformation byproducts on fish, daphnia, and green algae were also calculated. ChV represents the repeated doses of substances that initiated the development of adverse effects.35

2.5 Calculation of the frontier electron density of KET

Molecular orbital calculations were carried out at the single determinant (B3LYP/6-311+G*) level with the optimal conformation having a minimum energy obtained at the same level in a Gaussian 09 program.36 The array of atoms model and atom labelling of KET are depicted in Fig. S3. The frontier electron densities (FEDs) of the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) were determined to predict the initial attack position by reactive species. Bond dissociation energies of C18–O33 and C11–C17 in the KET molecule were calculated at the same level.

3. Results and discussion

3.1 Photodegradation of KET under simulated sunlight

As can be seen in Fig. 1, KET could not be decomposed without exposure to simulated solar radiation. In stark contrast, more than 90% of the KET was degraded after 180 s simulated solar irradiation. The photolysis of KET proceeded in accordance with pseudo-first order kinetics. The effect of different initial concentrations of KET showed that increasing the initial concentration of KET led to a decrease in the kinetic rate constants (Fig. S4). This result might be interpreted by considering the competition for the absorption of the limited population of available photons by the KET.37,38 Interestingly, the quantum yield of KET at 313 nm was determined to be 0.75,25 which was obviously higher than that of other PPCPs.39–41 The result revealed that the investigation of the KET photolysis mechanism was of critical importance.
image file: c7em00111h-f1.tif
Fig. 1 Photodegradation kinetics of KET.

3.2 Mechanism of the photolysis of KET

Previous studies25 have found that a series of reactive oxygen species (e.g., O2˙, ˙OH, and 1O2) were involved in the degradation of PPCPs during the photolytic process. Quenching experiments were carried out to test whether the KET underwent self-sensitization via ROS. Isopropanol (IPA),42 sodium azide (NaN3),43 and 4-hydroxy-2,2,6,6-tetramethylpiperidinyloxy (TEMPOL)44,45 were used as the ˙OH, 1O2 and ˙OH, as well as O2˙ quenchers. As shown in Table 1, the addition of isopropanol had a negligible effect on the degradation of KET. In contrast, sodium azide and TEMPOL exhibited significant inhibitory effects. The results demonstrated that the photodegradation of KET might include a self-sensitization photooxidation process via ˙OH, 1O2, and O2˙ under simulated sunlight. In order to further investigate the photolysis mechanism of KET, the contribution of ROS was estimated as follows:
 
image file: c7em00111h-t2.tif(2)
 
image file: c7em00111h-t3.tif(3)
 
image file: c7em00111h-t4.tif(4)
 
image file: c7em00111h-t5.tif(5)
where R values are the contribution rates of ROS. kisopropanol, kNaN3, and kTEMPOL are the rate constants for the addition of isopropanol, sodium azide, and TEMPOL in pure water. k presents the rate constant of photolysis of KET alone. The contribution rates of ˙OH, 1O2, and O2˙ were calculated to be ca. 14, 8.7, and 61.5%, respectively. And the contribution rate of direct photolysis can be estimated to be 15.8%. The results indicated that the O2˙ played a primary role in the photolysis of KET in pure water.
Table 1 Effects of isopropanol, NaN3, TEMPOL, N2, and O2 on the photodegradation kinetics of KET
Photocatalytic condition Quenching RSs Quencher concentration (mmol L−1) k (s−1) Contribution rate/%
DI water No scavengers 0.0114
Isopropanol Quenching ˙OH 50 0.0099 13.2
100 0.0098 14.0
200 0.0098 13.9
TEMPL Quenching O2˙ 2 0.0048 58.0
5 0.0043 62.2
10 0.0044 61.4
NaN3 Quenching 1O2 and ˙OH 5 0.0088 22.8
10 0.0087 23.6
50 0.0088 22.8
N2 0.0102 10.5
O2 0.0145 −27.2
Zhujiang river water No scavengers 0.0099
Isopropanol Quenching ˙OH 50 0.0866 12.5
TEMPL Quench O2˙ 10 0.0043 62.3
NaN3 Quench 1O2 and ˙OH 10 0.0082 28.4


One can obtain a simplified reaction scheme as follows:

 
image file: c7em00111h-t6.tif(6)
 
image file: c7em00111h-t7.tif(7)
 
3KET* + O2 → KET + 1O2(8)
 
3KET* + O2 → KET + O2˙(9)
 
KET˙ + O2 → KET + O2˙(10)
 
image file: c7em00111h-t8.tif(11)
 
O2˙/1O2/˙OH + KET → products(12)

As can be seen, the ground state of KET can be excited by photons to form an excited KET triplet state (3KET*), followed by direct photodegradation. Subsequently, autoionization reactions were conducted in 3KET*, giving rise to corresponding anion (KET˙)–cation (KET˙+) radical couples.46 Simultaneously, 3KET* with the O2 may undergo energy or electron transfer reactions to photogenerate 1O2 or O2˙.47 Additionally, another pathway for the generation of O2˙ may proceed through electron transfer, from KET˙ to 3O2.48 As is well recognized, H2O2 may be generated by the intermediate hydroperoxyl radical (˙O2H), produced via the protonation of O2˙. Further, H2O2 was capable of forming the stronger HO˙ oxidant upon irradiation.49,50 Finally, the generation of KET˙ and ROS may result in the decomposition of KET.

3.3 Effect of dissolved oxygen on the photodegradation of KET

The study of the photodegradation mechanism of KET reveals that dissolved oxygen in solution plays a significant role in the photodegradation process. In order to further investigate the effects of dissolved oxygen on the photodegradation of KET, experiments were conducted under different oxygen concentration conditions, using nitrogen or oxygen as the purging gas. As shown in Table 1, the higher the dissolved oxygen content the faster the degradation rate, implying O2 played a dual role in KET photolysis. On the one hand, dissolved oxygen in the solution could absorb energy from the excited triplet state KET, which subsequently promotes the self-sensitized degradation of KET through generating more reactive oxygen species.46,51 On the other hand, dissolved oxygen could inhibit the direct photodegradation of KET through quenching the KET molecules from the excited triplet state to the unexcited state.52 In the present study, dissolved oxygen promotes the apparent degradation of KET, indicating that oxidative attack by ROS has a significant role in the photodegradation of KET under simulated sunlight irradiation. Similar results were also reported by Yu-Chen Lin et al.51

3.4 Photodegradation in different water matrices

It has been reported that water constituents present in natural waters (e.g. dissolved organic matter (DOM) and HCO3) could cause indirect photooxidation of organic pollutants by generating reactive oxygen species.53 Meanwhile, these water constituents can also act as the scavengers for ROS and the competitors of photons.53 Experiments were then carried out with addition of some key water matrix species (DOM and HCO3) at different concentrations to investigate their influence on the photodegradation of KET under simulated sunlight irradiation.

As clearly seen in Table 2, the photo-degradation of KET was slightly enhanced in the presence of HCO3. It was demonstrated that a highly selective oxidizing agent, carbonate radicals, can be generated through the reaction of HCO3 with ˙OH. This oxidizing agent can subsequently react with KET, leading to the indirect photodegradation of KET.54 In contrast, the presence of DOM significantly decreased the photolysis rate. On the one hand, as a photosensitizer, DOM can produce ROS and enhance the degradation of pollutants.55 On the other hand, DOM can also play a negative role in photodegradation as a scavenger for ROS and an inner filter to absorb incident light.53 In this study, DOM significantly reduced their degradation rates, indicating that DOM mainly served as an inner filter rather than a photosensitizer.

 
˙OH + HCO3 → CO3˙ + H2O(13)
 
DOM + hv1DOM* → 3DOM*(14)
 
3DOM* + O2 → DOM + 1O2(15)
 
3DOM* + O2 → DOM˙+ + O2˙(16)
 
ROS + KET → products(17)
 
DOM/1DOM*/3DOM*/DOM˙+ + ROS → oxidized DOM(18)

Table 2 Effects of different water matrices on the photodegradation kinetics of KET
Solution matrix Concentration k (s−1) Inhibition rate (%)
DI water 0.0114
DOM 1 mgC L−1 0.0089 21.9
5 mgC L−1 0.0055 51.4
10 mgC L−1 0.0035 69.5
HCO3− 1 mmol L−1 0.0118 −3.3
5 mmol L−1 0.0120 −4.8
10 mmol L−1 0.0125 −9.8
Zhujiang river water 0.0099 12.8


To obtain a further insight into the photodegradation of KET under real water conditions, experiments with Zhujiang river water were carried out. As can be seen, the photolysis of KET in Zhujiang river water was slightly slower than that in DI water. Remarkably, the inhibition rate of Zhujiang river water (12.8%) was found to be significantly lower than that in the presence of 5.0 mgC per L DOM (51.4%). This result indicated that water constituents in the presence of sunlight produce ROS, crucial for the indirect photodegradation of KET in natural waters. In order to primarily investigate the photodegradation mechanism of KET in river water, quenching experiments were carried out. It was found that the addition of IPA, NaN3, and TEMPOL all inhibited the photodegradation of KET, implying that ROS were involved in the KET photolysis in river water under simulated sunlight irradiation. Among them, TEMPOL exhibited significant inhibitory effects on the photodegradation of KET with a contribution rate of 62.3%, indicating that O2˙ also plays a key role in the photolysis of KET under real river water conditions. Hence, further investigation is necessary to elucidate their specific roles.

3.5 Photolysis products and pathways

In order to further investigate the mechanism of KET under simulated light irradiation, the photoproducts were identified using HPLC-MS/MS and LC/MS techniques. The total ion chromatogram (TIC) of HPLC-MS/MS in Fig. S5 shows five major peaks at 3.661, 6.787, 7.689, 9.0841, and 11.568 min, which were assigned to P1, P2, KET, P3, and P4, respectively. Fig. S6(a–d) and Table S3 summarize the corresponding fragmentation pattern of KET and its photoproducts. Product P1 with the m/z 209 molecular ion was assigned as the KET derived decarboxylation product 3-ethylbenzophenone, which was also observed by Kosjek et al.28 and Erzsébet Illés22 during the UV photolysis and O3/UV processes. Product P2 with m/z 207 was regarded as the elimination reaction product, phenyl(3-vinylphenyl)methanone, which was also detected by C. Martínez48 during the photocatalysis experiment. Product P3 with the m/z 225 molecular ion was verified to be the hydroxylation product 3-(1-hydroxyethyl) benzophenone. Product 4 [1-(3-benzoylphenyl)ethanone] with the m/z 223 molecular ion was proposed as the ketonization derivative of m/z 225. Erzsébet Illés22 also identified the P3 and P4 in O3/UV degradation of KET. Moreover, the GC total ion chromatograms (Fig. S7) present four major peaks at 16.9, 18.8, 19.5, and 25.8 min, which were in correspondence with P1, P2, P3, and KET. Fig. S6(e–g) and Table S4 summarize the analysis of GC-MS fragments. Interestingly, the results of the GC-MS were well aligned with those of HPLC-MS/MS. The change of P1–P4 peak areas over time is depicted in Fig. S8. As can be seen, in addition to P2, the relative peak areas of the detected by-products increased gradually with the decomposition of KET.

TOC analysis was then carried out to evaluate the extent of mineralization during the photolysis of KET. As depicted in Fig. 2, ca. 91.43% KET was decomposed under 180 s irradiation, while only ca. 24.87% of the TOC was removed. There was almost no change in the TOC following a long exposure of 240 s, indicating that the majority of KET was transformed into structurally stable intermediate photoproducts without the complete mineralization of CO2 and H2O. The result coincided well with Seiya Hanamoto's study.29


image file: c7em00111h-f2.tif
Fig. 2 Evolution of both TOC removal efficiencies (blue curve, right y-axis) and KET degradation rates (black curve, left y-axis) during the photolysis.

Furthermore, frontier electron densities (FEDs) were calculated to predict the reaction sites for ROS attack. According to the Frontier Orbital Theory, atoms with more positive point charges were more easily attacked by O2˙, while the addition of ˙OH likely took place on atoms with the higher FEDHOMO2 + FEDLUMO2 values.15Table 3 summarizes the FEDs data of KET molecules. C17 and C18 showed more positive point charges than the others, implying that O2˙ nucleophilic reactions occurred in C17 and C18. Previous studies have found that C17 and C18 may be attacked by ROS, resulting in the cleavage of C18–O33 and C11–C17 bonds.24 In order to correctly deduce the degradation pathways, the bond dissociation energies (BDE) of C18–O33 and C11–C17 were calculated. As shown in Table S5, C18–O33 (0.5114 au.) showed a higher dissociation energy than C11–C17 (0.1476 au.). The results demonstrated that the bond cleavage might easily occur in C11–C17, that is, C17 possessed a higher activity than C18 (ref. 9). In addition, the C20 and C21 sites in the aromatic ring had higher FEDHOMO2 + FEDLUMO2 values, which indicated the high probability of hydroxyl formation on the aromatic ring. It should be noted that the detection of the hydroxylation product P3 by LC-MS/MS and GC-MS revealed the existence of ˙OH. However, no hydroxylation products on the aromatic ring were detected by both HPLC-MS/MS and GC/MS, which was likely due to the low concentration and non-selection of ˙OH.

Table 3 Frontier electron densities on the atoms of KET calculated according to the Gaussian 09 program at the B3LYP/6-311+G* level
Atom (number) Point charge FEDHOMO2 + FEDLUMO2 Atom (number) Point charge FEDHOMO2 + FEDLUMO2
1C 0.0403 0.3730 19C −0.3995 0.9585
2C −0.3696 1.0434 20C 0.0313 1.8729
3C −0.0171 1.6546 21C 0.0115 1.8925
4C −0.0913 0.1975 22C −0.1168 0.1279
5C −0.0184 0.1487 24C −0.1093 0.0968
6C 0.0162 0.5571 26C 0.0221 0.1012
11C 0.0255 0.0478 30O −0.9845 0.0049
13C 0.0541 0.0124 31O −0.9172 0.0024
17C 1.4927 0.0446 33O −1.1764 0.2964
18C 1.7112 0.4567


On the basis of the identified photoproducts and the predicted reaction sites, two major photolysis pathways of KET in aqueous solutions under simulated solar irradiation were proposed, as illustrated in Fig. 3. As can be seen, two major pathways including direct photolysis and decarboxylation were involved in the photolytic degradation of KET.


image file: c7em00111h-f3.tif
Fig. 3 Proposed possible photolysis pathways of ketoprofen under simulated sunlight.

Pathway I: pathway I was initiated by O2˙ attacking the C17 atom, resulting in the cleavage of carboxyl, thereby giving rise to a carbon-centered radical.56 Subsequently, the carbon-centered radicals were attacked by HO˙, leading to the generation of product 3. Product 3 can be further oxidized to product 4. Moreover, product 4 may also be generated from carbon-centered radicals via O2˙ oxidation. In addition, carbon-centered radicals underwent protonation, giving rise to P1.

Pathway II: direct photolysis of KET results in the formation of an excited triplet state of KET (KET*). As an important intermediate, the radical anion (KET˙) may be generated via the autoionization of KET*. Bond cleavage of KET˙ subsequently occurred, resulting in the production of intermediate product 2. Similar pathways had been reported in previous research.57

3.6 Toxicity assessment and toxicity predictions

The incomplete mineralization of organic pollutants may lead to the generation of intermediate products with higher toxicity, particularly ketone and hydroxylated products.58 From the above-mentioned results, four intermediates were detected during the photolytic process, which likely caused changes in the toxicity of the KET solution. In the present work, D. magna and V. fischeri were employed to assess the change in toxicity during the photolysis process.

As can be seen in Fig. 4, the original immobilization rates of D. magna were 24.43% and 55.92% at 24 h and 48 h, respectively. After 180 s simulated sunlight irradiation, ca. 91.43% KET was decomposed, while no significant reduction in the immobilization rate of D. magna by the KET solution could be found. In addition, a similar result could be found in the luminescence inhibition rate of V. fischeri by the photolyzed KET solution during the photolysis process. The results suggested that the photoproducts of KET may still have toxicity. This result was consistent with previous reports that the Microtox acute toxicity of the ketoprofen solution increased during the photolysis process.29,30 Therefore, the ecological risks of KET are worthy of attention, as there are higher risks for the generation of toxic byproducts under solar irradiation.


image file: c7em00111h-f4.tif
Fig. 4 (a) Acute toxicity evaluated by D. magna under 24 h and 48 h exposure times; (b) acute toxicity evaluated by V. fischeri for 180 s.

In order to further assess the toxicity of transformation byproducts, the half-maximal effective concentration (LC50 or EC50) for fish, daphnia, and green algae of photolysis byproducts was calculated by the ECOSAR. The predicted values of KET and its transformation by-products in the neutral form are given in Table S6.Fig. 5 summarizes the predicted values of KET and its transformation byproducts. As clearly seen, the prediction of toxicity by the ECOSAR for all the intermediate products may be higher than for KET. Among them, P1 (recognized as the decarboxylation product of KET) might exhibit the highest predicted toxicity, which was likely to be toxic to fish, daphnia, and green algae in terms of both acute and chronic toxicity. The results of mineralization and temporal trends of product peak areas have shown that the majority of KET was transformed into structurally stable intermediate photoproducts, which might be toxic substances that had the capacity to impart harm to aquatic organisms at three trophic levels. Hence, much more attention should be paid to KET photolysis products, as they are considerably more toxic and persistent, in contrast to their precursors.


image file: c7em00111h-f5.tif
Fig. 5 Evolution of acute and chronic toxicities through the degradation pathway.

4. Conclusion

In the current work, we investigated the photolytic mechanisms and ecotoxicity of KET intermediate products under simulated solar irradiation. The observed photodegradation followed pseudo-first order kinetics, whereas the pseudo-first order rate constant decreased with higher initial KET concentrations. The results revealed that ROS plays a primary role during the photolytic process of KET. Bicarbonate slightly increased the photodegradation of KET through generating carbonate radicals, while DOM inhibited the photolysis via both attenuating light and competing radicals. Moreover, Zhujiang river water inhibited KET phototransformation. Indirect photodegradation plays an important role in the photolytic process of KET in natural waters the constituents of which produce ROS in the presence of sunlight. On the basis of byproduct identification using HPLC-MS/MS and GC/MS, as well as the theoretical calculation by FEDs, two primary photochemical transformation pathways of KET, involving direct photolysis and decarboxylation, were proposed. Toxicity evaluation through acute toxicity tests, employing D. magna and V. fischeri as indicator species, and theoretical prediction using the ECOSAR program, revealed that additional toxic products were generated during the photodegradation of KET under solar irradiation. The present work provides important information related to the photolysis behavior of KET in the ambient aqueous environment and establishes the utility of the evaluation of ecological toxicity risks.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (No. 21377031 and 21677040) and the Innovative Team Program of High Education of Guangdong Province (2015KCXTD007). The authors would like to thank the anonymous reviewers and editors for their assistance toward the improvement of this paper.

References

  1. C. G. Daughton and T. A. Ternes, Environ. Health Perspect. Suppl., 1999, 107(Suppl 6), 907–938 CrossRef CAS .
  2. B. J. Richardson, P. K. Lam and M. Martin, Mar. Pollut. Bull., 2005, 50, 913–920 CrossRef CAS PubMed .
  3. G. R. Boyd, H. Reemtsma, D. A. Grimm and S. Mitra, Sci. Total Environ., 2003, 311, 135–149 CrossRef CAS PubMed .
  4. B. E. Erickson, Environ. Sci. Technol., 2002, 36, 140A–145A CrossRef CAS PubMed .
  5. J. Ma, W. Lv, P. Chen, Y. Lu, F. Wang, F. Li, K. Yao and G. Liu, Environ. Sci. Pollut. Res., 2016, 23, 14294–14306 CrossRef CAS PubMed .
  6. T. A. Ternes, M. Meisenheimer, D. Mcdowell, F. Sacher, H. Brauch, B. Haistgulde, G. Preuss, A. Uwe Wilme and N. Zuleiseibert, Environ. Sci. Technol., 2002, 36, 3855–3863 CrossRef CAS PubMed .
  7. Y. Yong and L. Wu, Talanta, 2012, 89, 258–263 CrossRef PubMed .
  8. V. Homem and L. Santos, J. Environ. Manage., 2011, 92, 2304–2347 CrossRef CAS PubMed .
  9. C. Martínez, S. Vilariño, M. I. Fernández, J. Faria, M. Canle L. and J. A. Santaballa, Appl. Catal., B, 2013, 142–143, 633–646 CrossRef .
  10. N. Nakada, T. Tanishima, H. Shinohara, K. Kiri and H. Takada, Water Res., 2006, 40, 3297–3303 CrossRef CAS PubMed .
  11. J. L. Santos, I. Aparicio and E. Alonso, Environ. Int., 2007, 33, 981–987 CrossRef PubMed .
  12. N. Lindqvist, T. Tuhkanen and L. Kronberg, Water Res., 2005, 39, 2219–2228 CrossRef CAS PubMed .
  13. Y. P. Duan, X. Z. Meng, Z. H. Wen and L. Chen, Environ. Sci. Pollut. Res., 2013, 20, 108–116 CrossRef CAS PubMed .
  14. J. L. Cao, J. H. Shi, H. Rui, Y. X. Li and Z. F. Yang, Chin. Sci. Bull., 2010, 55, 3138–3144 CrossRef CAS .
  15. K. A. Musa, J. M. Matxain and L. A. Eriksson, J. Med. Chem., 2007, 50, 1735–1743 CrossRef CAS PubMed .
  16. T. Suzuki, Y. Osanai and T. Isozaki, Photochem. Photobiol., 2012, 88, 884–888 CrossRef CAS PubMed .
  17. A. L. Boreen, W. A. Arnold and K. Mcneill, Aquat. Sci., 2003, 65, 320–341 CrossRef CAS .
  18. Y. Ji, L. Zhou, Y. Zhang, C. Ferronato, M. Brigante, G. Mailhot, X. Yang and J. M. Chovelon, Water Res., 2013, 47, 5865–5875 CrossRef CAS PubMed .
  19. L. Zhou, Y. Ji, C. Zeng, Y. Zhang, Z. Wang and X. Yang, Water Res., 2013, 47, 153–162 CrossRef CAS PubMed .
  20. C. C. Ryan, D. T. Tan and W. A. Arnold, Water Res., 2011, 45, 1280–1286 CrossRef CAS PubMed .
  21. J. J. Inbaraj, P. Bilski and C. F. Chignell, Photochem. Photobiol., 2002, 75, 107–116 CrossRef CAS PubMed .
  22. E. Illés, E. Szabó, E. Takács, L. Wojnárovits, A. Dombi and K. Gajda-Schrantz, Sci. Total Environ., 2014, 472, 178–184 CrossRef PubMed .
  23. P. Chen, W. Lv, Z. Chen, J. Ma, R. Li, K. Yao, G. Liu and F. Li, Environ. Sci. Pollut. Res., 2015, 22, 12585–12596 CrossRef CAS PubMed .
  24. A. Nakajima, M. Tahara, Y. Yoshimura and H. Nakazawa, J. Photochem. Photobiol., A, 2005, 174, 89–97 CrossRef CAS .
  25. L. L. Costanzo, G. G. De, G. Condorelli, A. Cambria and M. Fama, Photochem. Photobiol., 1989, 50, 359–365 CrossRef CAS PubMed .
  26. H. Li, X. He, Z. Kang, H. Hui, Y. Liu, J. Liu, S. Lian, C. H. A. Tsang, X. Yang and S. T. Lee, Angew. Chem., 2010, 122, 4532–4536 CrossRef .
  27. F. Boscá, M. A. Miranda, G. Carganico and D. Mauleón, Photochem. Photobiol., 1994, 60, 96–101 CrossRef .
  28. T. Kosjek, S. Perko, E. Heath, B. Kralj and D. Zigon, J. Mass Spectrom., 2011, 46, 391–401 CrossRef CAS PubMed .
  29. S. Hanamoto, T. Kawakami, N. Nakada, N. Yamashita and H. Tanaka, Environ. Sci.: Processes Impacts, 2014, 16, 2796 CAS .
  30. X. H. Wang and Y. C. Lin, Environ. Pollut., 2014, 186C, 203–215 Search PubMed .
  31. OECD, Methods Mol. Biol., 2013, 947, 37–56 Search PubMed .
  32. B. S. Institution, ISO 11348-2. Water quality. Determination of the inhibitory effect of water samples on the light emission of Vibrio fischeri (Luminescent bacteria test). Part 2. Method using liquid-dried bacteria.
  33. E. Marco-Urrea, M. Pérez-Trujillo, C. Cruz-Morató, G. Caminal and T. Vicent, J. Hazard. Mater., 2010, 176, 836–842 CrossRef CAS PubMed .
  34. N. Burden, S. K. Maynard, L. Weltje and J. R. Wheeler, Toxicol. Lett., 2016, 238, 241–246 Search PubMed .
  35. A. W. K. Wn and K. S. Tay, Env. Sci. Process. Im., 2016, 18, 555–561 RSC .
  36. T. An, H. Yang, G. Li, W. Song, W. J. Cooper and X. Nie, Appl. Catal., B, 2010, 94, 288–294 CrossRef CAS .
  37. A. B. A. Boxall, D. W. Kolpin, B. Hallingsørensen and J. Tolls, Environ. Sci. Technol., 2003, 37, 286A–294A CrossRef CAS PubMed .
  38. J. B. Ellis, Environ. Pollut., 2006, 144, 184–189 CrossRef CAS PubMed .
  39. P. Chen, F. L. Wang, K. Yao, J. S. Ma, F. H. Li, W. Y. Lv and G. G. Liu, Bull. Environ. Contam. Toxicol., 2016, 96, 203–209 CrossRef CAS PubMed .
  40. L. Ge, G. Na, S. Zhang, K. Li, P. Zhang, H. Ren and Z. Yao, Sci. Total Environ., 2015, 527–528, 12–17 CrossRef CAS PubMed .
  41. J. C. Carlson, M. I. Stefan, J. M. Parnis and C. D. Metcalfe, Water Res., 2015, 84, 350 CrossRef CAS PubMed .
  42. G. V. Buxton, C. L. Greenstock, W. P. Helman and A. B. Ross, J. Phys. Chem. Ref. Data, 1988, 17, 513–886 CrossRef CAS .
  43. G. Miolo, G. Viola, D. Vedaldi, F. Dall'Acqua, A. Fravolini, O. Tabarrini and V. Cecchetti, Toxicol. In Vitro, 2002, 16, 683–693 CrossRef CAS PubMed .
  44. W. Wang, Y. Yu, T. An, G. Li, H. Y. Yip, J. C. Yu and P. K. Wong, Environ. Sci. Technol., 2012, 46, 4599–4606 CrossRef CAS PubMed .
  45. L. S. Zhang, K. H. Wong, D. Q. Zhang, C. Hu, J. C. Yu, C. Y. Chan and P. K. Wong, Environ. Sci. Technol., 2009, 43, 7883–7888 CrossRef CAS PubMed .
  46. N. Zhang, G. Liu, H. Liu, Y. Wang, Z. He and G. Wang, J. Hazard. Mater., 2011, 192, 411–418 CAS .
  47. S. Zhang, J. Chen, X. Qiao, L. Ge, X. Cai and G. Na, Environ. Sci. Technol., 2010, 44, 7484–7490 CrossRef CAS PubMed .
  48. C. Martínez, S. Vilarino, M. I. Fernandez, J. Faria, M. L. Canle and J. A. Santaballa, Appl. Catal., B, 2013, 142, 633–646 CrossRef .
  49. L. Shen and H.-F. Ji, J. Photochem. Photobiol., B, 2010, 99, 154–156 CrossRef CAS PubMed .
  50. Y. Zuo and J. Hoigne, Environ. Sci. Technol., 1992, 26, 1014–1022 CrossRef CAS .
  51. Y. C. Lin and M. Reinhard, Environ. Toxicol. Chem., 2005, 24, 1303–1309 CrossRef PubMed .
  52. J. Xu, Z. Hao, C. Guo, Y. Zhang, Y. He and W. Meng, Chemosphere, 2014, 99, 186–191 CrossRef CAS PubMed .
  53. Y. Ji, C. Zeng, C. Ferronato, J. M. Chovelon and X. Yang, Chemosphere, 2012, 88, 644–649 CrossRef CAS PubMed .
  54. J. Huang and S. A. Mabury, Environ. Toxicol. Chem., 2000, 19, 2181–2188 CrossRef CAS .
  55. Y. Du, H. Chen, Y. Zhang and Y. Chang, Chemosphere, 2014, 99, 254–260 CrossRef CAS PubMed .
  56. F. Wang, P. Chen, Y. Feng, Z. Xie, Y. Liu, Y. Su, Q. Zhang, Y. Wang, K. Yao and W. Lv, Appl. Catal., B, 2017, 207, 103–113 CrossRef CAS .
  57. D. Gan, M. Jia, P. P. Vaughan, D. E. Falvey and N. V. Blough, J. Phys. Chem. A, 2008, 112, 2803–2812 CrossRef CAS PubMed .
  58. K. Joonwoo, J. Hyosang, K. Jonggu, H. Ishibashi, M. Hirano, K. Nasu, N. Ichikawa, Y. Takao, R. Shinohara and K. Arizono, J. Health Sci., 2009, 55, 249–258 CrossRef .

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c7em00111h

This journal is © The Royal Society of Chemistry 2017