Degradation of Ketamine and Methamphetamine by the UV / H 2 O 2 System: Kinetics, Mechanisms and Comparison

: The illegal use and low biodegradability of psychoactive substances has led to their introduction to the natural water environment, causing potential harm to ecosystems and human health. This paper compared the reaction kinetics and degradation mechanisms of ketamine (KET) and methamphetamine (METH) by UV / H 2 O 2 . Results indicated that the degradation of KET and METH using UV or H 2 O 2 alone was negligible. UV / H 2 O 2 had a strong synergizing e ﬀ ect, which could e ﬀ ectively remove 99% of KET and METH (100 µ g / L) within 120 and 60 min, respectively. Their degradation was fully consistent with pseudo-ﬁrst-order reaction kinetics (R 2 > 0.99). Based on competition kinetics, the rate constants of the hydroxyl radical with KET and METH were calculated to be 4.43 × 10 9 and 7.91 × 10 9 M − 1 · s − 1 , respectively. The apparent rate constants of KET and METH increased respectively from 0.001 to 0.027 and 0.049 min − 1 with the initial H 2 O 2 dosage ranging from 0 to 1000 µ M at pH 7. Their degradation was signiﬁcantly inhibited by HCO 3 − , Cl − , NO 3 − and humic acid, with Cl − having relatively little e ﬀ ect on the degradation of KET. Ultraperformance liquid chromatography with tandem mass spectrometry was used to identify the reaction intermediates, based on which the possible degradation pathways were proposed. These promising results clearly demonstrated the potential of the UV / H 2 O 2 process for the e ﬀ ective removal of KET and METH from contaminated wastewater.


Introduction
Illicit drugs are nonprescribed or psychostimulant substances which cannot be completely removed by conventional wastewater treatment, resulting in their widespread occurrence in aquatic environments [1,2]. Ketamine (KET) and methamphetamine (METH) were detected most frequently, with concentration levels up to 275 ng/L for KET and 239 ng/L for METH, in surface waters in China [3]. METH removal at most wastewater treatment plants was more than 80%, while the elimination of KET was less than 50% or even negative [4]. It was confirmed that chronic environmental concentrations of METH can lead to health issues in aquatic organisms [5]. Liao et al. [6] also reported that blood circulation and incubation time in medaka fish embryos could be significantly delayed at environmental concentration levels (0.004-40 µM) of KET and METH, which altered the swimming behavior of medaka fish larvae. Thus, there is an urgent need to explore new, efficient methods for eliminating these emerging contaminants in water.

Experimental Section
The experiments were operated in the quartz tubes (25 mm in diameter and 175 mm in length), which were placed in a photochemical reactor (Figure 1, XPA-7, Xujiang Machinery Factory, Nanjing, China). A low-pressure mercury lamp (11 W, emission at 254 nm, Philips Co., Zhuhai, China) was placed in the quartz sleeve. The UV lamp was preheated for 30 min to ensure irradiation stability. The UV fluence rate of 0.1 mW cm −2 was determined using three different methods [14]. The newly configured KET/METH and H2O2 stock solutions were supplemented with appropriate volumes to achieve a 50 mL reaction solution, which was then stirred thoroughly at 300 rpm with Advanced oxidation processes (AOPs) have been employed to destroy illicit drugs due to their high efficiency and lower environmental impact [7,8]. The UV/H2O2 process is one of the AOPs and generates the strong, oxidizing hydroxyl radical (OH, E0 = 2.72 V), which attacks the organic compounds with rate constants ranging from 10 8 to 10 10 M −1 s −1 [9]. Benzoylecgonine (BE), a metabolite of cocaine, was effectively removed by UV/H2O2 from different matrices [10]. The degradation of KET and METH was investigated using various AOPs, but no available report, so far, has addressed OH assisted by UV/H2O2 treatment. After 3 min, 100 μg/L of METH that had been added to deionized water was completely eliminated by TiO2 photocatalysis under UV365nm irradiation [11]. Wei et al. [12] studied the synthesis of a novel sonocatalyst Er 3+ :YAlO3/Nb2O5 and its application for METH degradation. Gu et al. [13] observed that complete removal of KET was achieved by UV/persulfate, and possible transformation pathways were proposed.
To the best of our knowledge, there is little information about the theoretical calculation of the reactivity of KET and METH by radical attack using the UV/H2O2 process. Water constituents in actual wastewater could affect the degradation efficacy; therefore, a comprehensive understanding of the degradation of KET and METH using the UV/H2O2 system is needed. The aim of this study was to investigate the degradation kinetics and mechanisms of KET and METH during the UV/H2O2 process. The influence of various parameters on KET and METH removal was evaluated, including initial H2O2 dosage, pH and water background components. The degradation products were analyzed by ultraperformance liquid chromatography with tandem mass spectrometry (UPLC-MS/MS), and possible transformation paths were proposed.

Experimental Section
The experiments were operated in the quartz tubes (25 mm in diameter and 175 mm in length), which were placed in a photochemical reactor (Figure 1, XPA-7, Xujiang Machinery Factory, Nanjing, China). A low-pressure mercury lamp (11 W, emission at 254 nm, Philips Co., Zhuhai, China) was placed in the quartz sleeve. The UV lamp was preheated for 30 min to ensure irradiation stability. The UV fluence rate of 0.1 mW cm −2 was determined using three different methods [14]. The newly configured KET/METH and H2O2 stock solutions were supplemented with appropriate volumes to achieve a 50 mL reaction solution, which was then stirred thoroughly at 300 rpm with 4846-07-5 9.9 2.07

Experimental Section
The experiments were operated in the quartz tubes (25 mm in diameter and 175 mm in length), which were placed in a photochemical reactor (Figure 1, XPA-7, Xujiang Machinery Factory, Nanjing, China). A low-pressure mercury lamp (11 W, emission at 254 nm, Philips Co., Zhuhai, China) was placed in the quartz sleeve. The UV lamp was preheated for 30 min to ensure irradiation stability. The UV fluence rate of 0.1 mW cm −2 was determined using three different methods [14]. The newly configured KET/METH and H 2 O 2 stock solutions were supplemented with appropriate volumes to achieve a Water 2020, 12, 2999 3 of 13 50 mL reaction solution, which was then stirred thoroughly at 300 rpm with electromagnetic stirrers. Upon UV irradiation, the reaction started at pH 7.0 and room temperature. Specific samples were immediately quenched using a catalase and passed through 0.22 µm nylon filter before further analysis.

Analytical Methods
The concentrations of KET and METH were quantified by UPLC-MS/MS equipped with a Waters Acquity liquid chromatography system and an Xevo T_QS triple quadrupole mass spectrometer (Waters Co., Milford, MA, USA). The analytes were separated by a reverse phase column (Acquity UPLC BEH C18, 1.7 μm, 50 × 2.1 mm, Waters, MA, USA). The mobile phases A and B, with a flow rate of 450 μL min −1 , were 0.1% FA in Milli-Q water and ACN, respectively. Ten percent of phase B was kept for 0.5 min at the initial proportion, linearly increased to 45% at 1.8 min, then increased to 95% within 0.1 min, held for 1.0 min, reverted to 10% at 3.0 min and held for 1.5 min. The injection volume was 5 μL with the column temperature at 40 °C. The chromatograms were recorded in the positive ion multiple reaction monitoring (MRM) mode. Nitrogen was used as the desolvation and nebulizing gas. The capillary voltage was set at 0.5 kV, and the desolvation temperature was 400 °C. Optimized UPLC-MS/MS parameters are given in Table 2.  Figure 2 shows the degradation of KET and METH under different treatment processes. UV or H2O2 alone exhibited negligible effects on their degradation, suggesting that treatment by UV or H2O2 alone was unable to destroy KET and METH. However, nearly complete removal of KET and METH was achieved within 120 and 60 min, respectively, when treated with the combination of UV/H2O2. Similar results were reported regarding ofloxacin degradation, which was drastically increased due to the large amount of hydroxyl radicals (  OH) generated via the breakage of the H2O2

Analytical Methods
The concentrations of KET and METH were quantified by UPLC-MS/MS equipped with a Waters Acquity liquid chromatography system and an Xevo T_QS triple quadrupole mass spectrometer (Waters Co., Milford, MA, USA). The analytes were separated by a reverse phase column (Acquity UPLC BEH C18, 1.7 µm, 50 × 2.1 mm, Waters, MA, USA). The mobile phases A and B, with a flow rate of 450 µL min −1 , were 0.1% FA in Milli-Q water and ACN, respectively. Ten percent of phase B was kept for 0.5 min at the initial proportion, linearly increased to 45% at 1.8 min, then increased to 95% within 0.1 min, held for 1.0 min, reverted to 10% at 3.0 min and held for 1.5 min. The injection volume was 5 µL with the column temperature at 40 • C. The chromatograms were recorded in the positive ion multiple reaction monitoring (MRM) mode. Nitrogen was used as the desolvation and nebulizing gas. The capillary voltage was set at 0.5 kV, and the desolvation temperature was 400 • C. Optimized UPLC-MS/MS parameters are given in Table 2.  Figure 2 shows the degradation of KET and METH under different treatment processes. UV or H 2 O 2 alone exhibited negligible effects on their degradation, suggesting that treatment by UV or H 2 O 2 alone was unable to destroy KET and METH. However, nearly complete removal of KET and METH was achieved within 120 and 60 min, respectively, when treated with the combination of UV/H 2 O 2 . Similar results were reported regarding ofloxacin degradation, which was drastically increased due to the large amount of hydroxyl radicals ( • OH) generated via the breakage of the H 2 O 2 Water 2020, 12, 2999 4 of 13 bond (Equation (1)) [15]. The degradation of KET and METH was consistent with the pseudo-first-order reaction kinetics. The apparent degradation rate constants (k obs ) of KET and METH by UV/H 2 O 2 were 0.027 and 0.049 min −1 , respectively.

Degradation Kinetics of KET and METH
Water 2020, 12, x FOR PEER REVIEW 4 of 15 bond (Equation (1)) [15]. The degradation of KET and METH was consistent with the pseudo-first-order reaction kinetics. The apparent degradation rate constants (kobs) of KET and METH by UV/H2O2 were 0.027 and 0.049 min −1 , respectively.

Determination of Bimolecular Reaction Rate
The generation of  OH in the UV/H2O2 system was proved by the photoluminescence (PL) technique using a probe molecule with terephthalic acid, which tends to react with  OH to form 2-hydroxyterephthalic acid, a highly fluorescent product [16]. The PL intensity of 2-hydroxyterephtalic acid is proportional to the amount of  OH radicals produced in water [17]. Figure 3 shows the PL spectral changes in the 5 × 10 −4 M terephthalic acid solution with a concentration of 2 × 10 −3 M NaOH (excitation at 315 nm), as described by Yu et al. [17]. Similar fluorescence intensity was found in the reaction systems with initial concentrations of 100 and 1000 μM of H2O2, suggesting a constant concentration of  OH with the initial H2O2 dosage ranging from 100 to 1000 μM. The PL signal at 425 nm increased with the irradiation time, which was attributed to the reaction of terephthalic acid with  OH generated in the UV/H2O2 system.

Determination of Bimolecular Reaction Rate
The generation of • OH in the UV/H 2 O 2 system was proved by the photoluminescence (PL) technique using a probe molecule with terephthalic acid, which tends to react with • OH to form 2-hydroxyterephthalic acid, a highly fluorescent product [16]. The PL intensity of 2-hydroxyterephtalic acid is proportional to the amount of • OH radicals produced in water [17]. Figure 3 shows the PL spectral changes in the 5 × 10 −4 M terephthalic acid solution with a concentration of 2 × 10 −3 M NaOH (excitation at 315 nm), as described by Yu et al. [17]. Similar fluorescence intensity was found in the reaction systems with initial concentrations of 100 and 1000 µM of H 2 O 2 , suggesting a constant concentration of • OH with the initial H 2 O 2 dosage ranging from 100 to 1000 µM. The PL signal at 425 nm increased with the irradiation time, which was attributed to the reaction of terephthalic acid with • OH generated in the UV/H 2 O 2 system.  (1)) [15]. The degradation of KET and METH was consistent with the pseudo-first-order reaction kinetics. The apparent degradation rate constants (kobs) of KET and METH by UV/H2O2 were 0.027 and 0.049 min −1 , respectively.

Determination of Bimolecular Reaction Rate
The generation of  OH in the UV/H2O2 system was proved by the photoluminescence (PL) technique using a probe molecule with terephthalic acid, which tends to react with  OH to form 2-hydroxyterephthalic acid, a highly fluorescent product [16]. The PL intensity of 2-hydroxyterephtalic acid is proportional to the amount of  OH radicals produced in water [17]. Figure 3 shows the PL spectral changes in the 5 × 10 −4 M terephthalic acid solution with a concentration of 2 × 10 −3 M NaOH (excitation at 315 nm), as described by Yu et al. [17]. Similar fluorescence intensity was found in the reaction systems with initial concentrations of 100 and 1000 μM of H2O2, suggesting a constant concentration of  OH with the initial H2O2 dosage ranging from 100 to 1000 μM. The PL signal at 425 nm increased with the irradiation time, which was attributed to the reaction of terephthalic acid with  OH generated in the UV/H2O2 system.  The bimolecular reaction rates of KET and METH reacting with • OH were determined through the competition experiments at pH 7 (phosphate buffer solution, 5 mM). BA was used as the reference compound, with which the constant reaction rate of • OH is known to be 5.9 × 10 9 M −1 s −1 [18]. It is important to note that the degradation of KET, METH and BA using UV alone was negligible at less than 9%. Equations (2) and (3) describe the competing kinetics of KET and METH with • OH in the UV/H 2 O 2 oxidation process, through which the bimolecular reaction rates of KET and METH reacting with • OH were 4.43 × 10 9 and 7.91 × 10 9 M −1 s −1 , respectively ( Figure 4).  The bimolecular reaction rates of KET and METH reacting with  OH were determined through the competition experiments at pH 7 (phosphate buffer solution, 5 mM). BA was used as the reference compound, with which the constant reaction rate of  OH is known to be 5.9 × 10 9 M −1 s −1 [18]. It is important to note that the degradation of KET, METH and BA using UV alone was negligible at less than 9%. Equations (2) and (3) describe the competing kinetics of KET and METH with  OH in the UV/H2O2 oxidation process, through which the bimolecular reaction rates of KET and METH reacting with  OH were 4.43 × 10 9 and 7.91 × 10 9 M −1 s −1 , respectively ( Figure 4).

Effect of H 2 O 2 Dosage
The KET and METH degradation under different initial H 2 O 2 dosages were consistent with the pseudo-first-order reaction model (R 2 > 0.99, Figure 5). The k obs of KET and METH increased dramatically from 0.001 min −1 to 0.027 and 0.049 min −1 with the initial H 2 O 2 dosage ranging from 0 to 1000 µM. The reason for this phenomenon is that the production of • OH increased with the initial H 2 O 2 dosage ranging from 0 to 1000 µM, thus accelerating the degradation rate of target compounds [19]. However, the k obs of METH decreased slightly with the initial concentration of H 2 O 2 increased to 2000 µM. A similar phenomenon was observed in a previous report that indicated that the degradation rates of cyclophosphamide and 5-fluorouracil were proportional to the H 2 O 2 dosage and slightly decreased with excess H 2 O 2 [20].
Water 2020, 12, x FOR PEER REVIEW 6 of 15 The KET and METH degradation under different initial H2O2 dosages were consistent with the pseudo-first-order reaction model (R 2 > 0.99, Figure 5). The kobs of KET and METH increased dramatically from 0.001 min −1 to 0.027 and 0.049 min −1 with the initial H2O2 dosage ranging from 0 to 1000 μM. The reason for this phenomenon is that the production of  OH increased with the initial H2O2 dosage ranging from 0 to 1000 μM, thus accelerating the degradation rate of target compounds [19]. However, the kobs of METH decreased slightly with the initial concentration of H2O2 increased to 2000 μM. A similar phenomenon was observed in a previous report that indicated that the degradation rates of cyclophosphamide and 5-fluorouracil were proportional to the H2O2 dosage and slightly decreased with excess H2O2 [20]. An excessive amount of H2O2 would cause the self-scavenging effect of  OH to form HO2 and O2 −  (Equations (4) and (5)) [21], the low reactivity of which could reduce the degradation rate. Similar results were obtained concerning the degradation of ofloxacin [15] and chloramphenicol [22]. Moreover, large amounts of  OH were dimerized to H2O2, and the generated HO2 and O2 −  subsequently participated in other reactions (Equations (6)-(9)) [23]. This negative effect was not observed in this study, probably because the maximum H2O2 dosage (2000 μM) was not high enough to inhibit the KET degradation.     Figure 6 illustrates the KET and METH destruction at different initial pHs, which were adjusted with an H2SO4 or NaOH solution (0.1 M). No buffer was used due to its inhibiting effect on the decomposition of organics [24]. The KET and METH degradation at different initial pHs followed the pseudo-first-order reaction model well. The kobs of KET and METH reached the highest levels in a neutral environment at 0.027 and 0.085 min −1 , respectively. Due to the greater stability of H2O2 at pH 5 and 7, the degradation rates of KET and METH under acidic and neutral conditions were obviously better than those under alkaline conditions. Under alkaline conditions,  OH could be quenched by the HO2 − produced by H2O2 dissociation, thus reducing the yield of  OH in the system.  Figure 6 illustrates the KET and METH destruction at different initial pHs, which were adjusted with an H 2 SO 4 or NaOH solution (0.1 M). No buffer was used due to its inhibiting effect on the decomposition of organics [24]. The KET and METH degradation at different initial pHs followed the pseudo-first-order reaction model well. The k obs of KET and METH reached the highest levels in a neutral environment at 0.027 and 0.085 min −1 , respectively. Due to the greater stability of H 2 O 2 at pH 5 and 7, the degradation rates of KET and METH under acidic and neutral conditions were obviously better than those under alkaline conditions. Under alkaline conditions, • OH could be quenched by the HO 2 − produced by H 2 O 2 dissociation, thus reducing the yield of • OH in the system.

Effect of Water Background Components on Degradation Efficiency of Target Compounds
There are many different substrates in natural water, including different kinds of anions, cations and organic matter. These ions could react with free radicals in advanced oxidation processes, thus inhibiting or promoting the reaction and affecting the overall oxidation effect. Therefore, it is of great significance to study the influence of different ion types and contents on the practical application of advanced oxidation technology.

Effect of HCO3 −
The decomposition of KET and METH was significantly inhibited with the addition of HCO3 − at different initial dosages in the UV/H2O2 oxidation process (Figure 7). When the initial dosage of

Effect of Water Background Components on Degradation Efficiency of Target Compounds
There are many different substrates in natural water, including different kinds of anions, cations and organic matter. These ions could react with free radicals in advanced oxidation processes, thus inhibiting or promoting the reaction and affecting the overall oxidation effect. Therefore, it is of great significance to study the influence of different ion types and contents on the practical application of advanced oxidation technology.

−
The decomposition of KET and METH was significantly inhibited with the addition of HCO 3 − at different initial dosages in the UV/H 2 O 2 oxidation process (Figure 7). When the initial dosage of HCO 3 − ranged from 0 to 10 mM, the reaction rate of KET and METH decreased from 0.027 and 0.049 min −1 to 0.008 and 0.011 min −1 , respectively. The reason for this experimental phenomenon was that HCO 3 − was the quenching agent for • OH which was also consumed by the competing reaction of ionized CO 3 2− (Equations (10)- (13)). Therefore, the inhibitory effect of KET and METH degradation was more obvious with the increase of the HCO 3 − concentration.

Effect of Water Background Components on Degradation Efficiency of Target Compounds
There are many different substrates in natural water, including different kinds of anions, cations and organic matter. These ions could react with free radicals in advanced oxidation processes, thus inhibiting or promoting the reaction and affecting the overall oxidation effect. Therefore, it is of great significance to study the influence of different ion types and contents on the practical application of advanced oxidation technology.

Effect of HCO3 −
The decomposition of KET and METH was significantly inhibited with the addition of HCO3 − at different initial dosages in the UV/H2O2 oxidation process (Figure 7). When the initial dosage of HCO3 − ranged from 0 to 10 mM, the reaction rate of KET and METH decreased from 0.027 and 0.049 min −1 to 0.008 and 0.011 min −1 , respectively. The reason for this experimental phenomenon was that HCO3 − was the quenching agent for  OH which was also consumed by the competing reaction of ionized CO3 2− (Equations (10)- (13)). Therefore, the inhibitory effect of KET and METH degradation was more obvious with the increase of the HCO3 − concentration. −

Effect of Cl −
With the initial concentration of Cl − ranging from 0 to 10 mM, the destruction of KET was dramatically inhibited with the rate constant of KET decreased from 0.027 to 0.018 min −1 (Figure 8), which could be due to the elimination of • OH by Cl − according to Equations (14)-(16) [25]. The degradation reaction rate changed slightly as more Cl − was added. However, the METH degradation was less affected by Cl − , with the reaction rate remaining basically unchanged (0.0446-0.0485 min −1 ).
The influence of NO 3 − on the decomposition of KET and METH is illustrated in Figure 9. With the initial concentration of NO 3 − ranging from 0 to 10 mM, the degradation of both target compounds was obviously inhibited. The reaction rate of KET and METH decreased from 0.027 and 0.049 min −1 to 0.007 and 0.012 min −1 , respectively. The above experimental phenomena were attributed to the following: First, a large amount of • OH could be produced from NO 3 − under UV irradiation (Equations (17)-(18)), which is an important source of • OH in natural water [26]. Second, as a photosensitizer, NO 3 − has a strong absorption in the ultraviolet range, which results in the formation of an internal filter that prevents the effective light transmittance and leads to the decline of • OH production in the UV/H 2 O 2 system [27]. The latter was found to be dominant after the degradation effect of the reaction was analyzed.
Water 2020, 12, x FOR PEER REVIEW 10 of 15

Effect of HA
Due to its complex structure, HA may have uncontrollable effects on the destruction of target compounds. As illustrated in Figure 10, KET and METH degradation was dramatically inhibited once HA was added with different dosages in the UV/H2O2 system. As more HA (0-0.1 mM) was added, the reaction rate of KET and METH declined from 0.027 and 0.049 min −1 to 0.001 and 0.008 min −1 , respectively, while the degradation reaction rate changed slightly with the continued addition of the HA. UV irradiation was absorbed by HA, creating an inner filter ( Figure 11) and significantly inhibiting the UV transmittance for UV photons, thus limiting the generation of  OH in the UV/H2O2 process [28]. Moreover, the degradation of target compounds can be inhibited by the competing reaction of HA with the active radicals [29].

Effect of HA
Due to its complex structure, HA may have uncontrollable effects on the destruction of target compounds. As illustrated in Figure 10, KET and METH degradation was dramatically inhibited once HA was added with different dosages in the UV/H 2 O 2 system. As more HA (0-0.1 mM) was added, the reaction rate of KET and METH declined from 0.027 and 0.049 min −1 to 0.001 and 0.008 min −1 , respectively, while the degradation reaction rate changed slightly with the continued addition of the HA. UV irradiation was absorbed by HA, creating an inner filter ( Figure 11) and significantly inhibiting the UV transmittance for UV photons, thus limiting the generation of • OH in the UV/H 2 O 2 process [28]. Moreover, the degradation of target compounds can be inhibited by the competing reaction of HA with the active radicals [29].
added, the reaction rate of KET and METH declined from 0.027 and 0.049 min −1 to 0.001 and 0.008 min −1 , respectively, while the degradation reaction rate changed slightly with the continued addition of the HA. UV irradiation was absorbed by HA, creating an inner filter ( Figure 11) and significantly inhibiting the UV transmittance for UV photons, thus limiting the generation of  OH in the UV/H2O2 process [28]. Moreover, the degradation of target compounds can be inhibited by the competing reaction of HA with the active radicals [29].  . Figure 11. The ultraviolet-visible spectroscopy of reaction solutions at different concentrations of HA.

Degradation Products and Mechanism
Degradation intermediates and products of METH produced in the UV/H2O2 oxidation process were determined by using UPLC/MS/MS under full scans and product ion scans. During the whole METH degradation process, the mass spectra were compared to identify the intermediates. The structure of the transformation products was analyzed with the specific molecular ions and fragmentation patterns rather than direct comparison with corresponding standards. Figure 12 illustrates the mass spectra and possible structures of the degradation intermediates, based on which the possible transformation pathways of METH during UV/H2O2 are shown in Figure 13. The proposed degradation mechanisms of METH degradation involved in the UV/H2O2 system include hydrogenation, hydroxylation and electrophilic substitution.

Degradation Products and Mechanism
Degradation intermediates and products of METH produced in the UV/H 2 O 2 oxidation process were determined by using UPLC/MS/MS under full scans and product ion scans. During the whole METH degradation process, the mass spectra were compared to identify the intermediates. The structure of the transformation products was analyzed with the specific molecular ions and fragmentation patterns rather than direct comparison with corresponding standards. Figure 12  METH degradation process, the mass spectra were compared to identify the intermediates. The structure of the transformation products was analyzed with the specific molecular ions and fragmentation patterns rather than direct comparison with corresponding standards. Figure 12 illustrates the mass spectra and possible structures of the degradation intermediates, based on which the possible transformation pathways of METH during UV/H2O2 are shown in Figure 13. The proposed degradation mechanisms of METH degradation involved in the UV/H2O2 system include hydrogenation, hydroxylation and electrophilic substitution.

Conclusions
The degradation kinetics and mechanisms of KET and METH using the UV/H2O2 process were investigated in this study. Their degradation in UV photolysis or H2O2 oxidation alone was negligible. However, 99% of KET and METH (100 μg/L) were effectively eliminated by the combination of UV and H2O2 within 120 and 60 min, respectively. According to the competition kinetics, the rate constants of  OH with KET and METH at pH 7 were calculated to be 4.43 × 10 9 and 7.91 × 10 9 M −1 ·s −1 , respectively. The apparent rate constants of KET and METH reached the highest levels in a neutral environment. The degradation of KET and METH was significantly inhibited by HCO3 − , Cl − , NO3 − and HA; however, Cl − had little influence on the METH degradation. Seven reaction intermediates of METH in the UV/H2O2 system were identified by UPLC-MS/MS. Possible transformation mechanisms involved in the KET and METH degradation by UV/H2O2 oxidation system included hydrogenation, hydroxylation and electrophilic substitution. Results demonstrated that UV/H2O2 was an effective technique to remove KET and METH, providing a promising application for the decomposition of trace organic pollutants in natural water.