Oxidation kinetics of cyclophosphamide and methotrexate by ozone in drinking water

Oxidation kinetics of cyclophosphamide and methotrexate by ozone in drinking water

Chemosphere 79 (2010) 1056–1063 Contents lists available at ScienceDirect Chemosphere journal homepage: www.elsevier.com/locate/chemosphere Oxidati...

330KB Sizes 278 Downloads 61 Views

Chemosphere 79 (2010) 1056–1063

Contents lists available at ScienceDirect

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

Oxidation kinetics of cyclophosphamide and methotrexate by ozone in drinking water Araceli Garcia-Ac a, Romain Broséus b,*, Simon Vincent b, Benoit Barbeau b, Michèle Prévost b, Sébastien Sauvé a a

Department of Chemistry, Université de Montréal, P.O. Box 6128, Succ. Centre-ville, Montréal, QC, Canada H3C 3J7 NSERC Industrial Chair on Drinking Water, Civil, Geological and Mining Engineering Department, École Polytechnique de Montreal, CP 6079, Succ. Centre-ville, Montréal, QC, Canada H3C 3A7 b

a r t i c l e

i n f o

Article history: Received 15 January 2010 Received in revised form 21 March 2010 Accepted 22 March 2010 Available online 18 April 2010 Keywords: Water treatment Drinking water Ozone Cytostatic drugs Mass spectrometry

a b s t r a c t This study investigates the aqueous degradation by ozone of two target cytostatic drugs, cyclophosphamide and methotrexate. A column switching technique for on-line solid phase extraction (SPE) coupled to electro-spray ionization-tandem mass spectrometry (LC–ESI-MS/MS) was used for the simultaneous detection of the trace contaminants. The second-order kinetic rate constants for the reaction of cyclophosphamide with molecular ozone and hydroxyl radicals were determined in bench-scale experiments at pH 8.10. The molecular ozone oxidation kinetics was studied in buffered ultrapure water and compared to the oxidation kinetics in natural water from a municipal drinking water treatment plant in the province of Quebec (Canada). For cyclophosphamide, the degradation rate constant with molecular ozone in ultrapure water was low (kO3 = 3.3 ± 0.2 M1 s1) and the extent of oxidation was linearly correlated to the ozone exposure. The impact of water quality matrix on oxidation efficacy was not significant during direct ozone reaction (kO3 = 2.9 ± 0.3 M1 s1). The rate constant with hydroxyl radicals was higher at 2.0  109 M1 s1. Methotrexate reacted quickly with molecular ozone at dosages typically applied in drinking water treatment (kO3 > 3.6  103 M1 s1). Overall, the results confirmed that organic compounds reactivity with ozone was dependent of their chemical structure. Ozone was very effective against methotrexate but high oxidant concentration  contact time (CT) values were required to completely remove cyclophosphamide from drinking water. Further studies should be conducted in order to identify the ozonation by-products and explore the impact of ozone on their degradation and toxicity. Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction Pharmaceuticals and their degradation by-products are routinely detected in surface waters (Heberer et al., 1997; Calisto and Esteves, 2009; Kummerer, 2009a,b; Mompelat et al., 2009) and even drinking waters (Stackelberg et al., 2004; Jones et al., 2005; Snyder et al., 2007; Benotti et al., 2009). A significant fraction of the pharmaceuticals released into the aquatic environment is attributed to their incomplete removal through conventional wastewater treatment (Halling-Sorensen et al., 1998; Daughton and Ternes, 1999; Joss et al., 2004; Clara et al., 2005; Segura et al., 2007; Carballa et al., 2008; Zhang et al., 2008; Matamoros et al., 2009). Among various classes of pharmaceuticals, cytostatic chemotherapy drugs (also called antineoplastic agents) are of particular environmental concern because they are potentially carcinogenic, mutagenic and genotoxic, even at low concentrations (Zounkova et al., 2007). Cyclophosphamide and methotrexate are * Corresponding author. Tel.: +1 514 340 4711x2194; fax: +1 514 340 5918. E-mail address: [email protected] (R. Broséus). 0045-6535/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2010.03.032

widely used at high dose for the chemotherapy of various forms of cancer (bronchial, breast and ovarian cancer, lymphomas, leukemias, etc.) and at low dose for the treatment of autoimmune diseases (e.g., rheumatoid arthritis), and also as immunosuppressants after organ transplantations (e.g., bone marrow transplantations) (Jolivet et al., 1983; Kuo et al., 2003; Buerge et al., 2006). The excretion of those unchanged parent molecules in patients appears to be around 10–20% (Johnson et al., 2008). According to Buerge et al. (2006), these compounds may reach the aquatic environment via hospital or domestic wastewater collectors on their way to wastewater treatment plants through excretions in the urine and feces of patients under medical treatment. Concentrations of cyclophosphamide, ifosfamide, 5-fluorouracil, anthracyclines and cancerostatic platinum compounds range from 6 ng L1 to 145 lg L1 in hospital wastewater effluents (Kummerer et al., 2000; Lenz et al., 2005; Mahnik et al., 2007). Cyclophosphamide, ifosfamide and methotrexate have been analyzed in untreated and treated municipal wastewater at concentrations in the ng L1 range (Kummerer et al., 1997; StegerHartmann et al., 1997; Ternes et al., 1998; Buerge et al., 2006; Garcia-Ac et al.,

A. Garcia-Ac et al. / Chemosphere 79 (2010) 1056–1063

2009). Antineoplastic agents have also been detected in source waters receiving treated wastewater effluents (650 pg L1– 10 ng L1) (Zuccato et al., 2000; Buerge et al., 2006) and occasionally in tap water (613 ng L1) (Aherne et al., 1990). More recently, a risk assessment study showed that although the risks about the potential concentrations of cyclophosphamide in drinking water may be negligible for healthy adults, more concern may be associated with special subgroup populations, such as pregnant women, their foetuses, and breast-feeding infants, due to their developmental vulnerability (Rowney et al., 2009). Due to the increase in the demand for chemotherapy drugs in developed countries and their common administration in outpatient treatment departments (Johnson et al., 2008), the concentration of cytostatic agents in municipal wastewaters is expected to increase. Their potential presence in surface waters points to the need to verify the efficacy of drinking water treatment processes to remove such compounds. Among biological, physical and chemical treatments, chemical oxidation using ozone treatment has demonstrated its effectiveness for a wide spectrum of organic micropollutants in both wastewater and drinking water and at various levels of experiments (bench-, pilot- and full-scale) (Ternes et al., 2002, 2003; Huber et al., 2005; Westerhoff et al., 2005; Ikehata et al., 2006; Snyder et al., 2006; Vieno et al., 2007; Gagnon et al., 2008). Ozone reacts with organic contaminants either by the direct reaction of molecular ozone or by the reaction of free radicals (mostly hydroxyl radicals OH) produced by the decomposition of ozone. The rate of OH formation depends on the water matrix, especially the pH, alkalinity and type and content of natural organic matter in drinking water (von Gunten, 2003). Molecular ozone reacts selectively with unsaturated bonds, aromatic systems and amino groups whereas OH radicals react much more indiscriminately (Ljubic and Sabljic, 2002). To predict the potential removal by ozonation, it is important to determine rate constants of the reaction of micropollutants with ozone and OH radicals. Some rate constants for direct (kO3 ) and indirect (kOH) ozone oxidation have been published and can serve to predict the performance of oxidation (Huber et al., 2003; von Gunten, 2003). Those rate constants can also be estimated by various quantitative models for estimating degradation of chemicals in various environmental compartments (Sabljic and Peijnenburg, 2001). However, few studies have evaluated the efficacy of ozone for the oxidation of cytostatics, which vary in chemical structures and properties that will affect their ability to be oxidized. PérezRey et al. (1999) studied the degradation of 5-fluorouracil, cytarabine, azathioprine and methotrexate with ozone in water using higher concentrations (1.1–2.2  103 M) than those detected in surface waters. The removal of cyclophosphamide and cisplatin by ozone and advanced oxidation processes (AOP) has been investigated (Venta et al., 2005; Chen et al., 2008; Hernandez et al., 2008) but only the pseudo first-order constants were estimated. Chen et al. (2008) reported the values of the second-order rate constants of irinotecan, tamoxifen and cyclophosphamide in ultrapure water only. A similar study conducted to assess the degradability of 30 pharmaceuticals by AOP showed that O3 process and O3based/UV-based AOP could efficiently remove a variety of the selected compounds. However, some pharmaceuticals such as cyclophosphamide showed a relatively low degradability (Kim et al., 2008). The determination of the reaction rate constants in ultrapure and natural water for two commonly used cytostatics, cyclophosphamide and methotrexate, will provide important information for drinking water treatment selection and adjustment. To the best of our knowledge, this is the first paper using very low concentrations of cytostatics representing nicely the environmental concentrations, and presenting the impact of ozone oxidation on cytostatics in a natural water matrix.

1057

2. Objectives of study In the research project presented hereafter, three objectives were sought: 1. To determine the efficacy of ozone for the degradation of two target cytostatic drugs at bench-scale. 2. To determine the second-order rate constants for the reaction of the selected compounds with molecular ozone and OH radicals in buffered ultrapure water. 3. To assess the impact of water quality on molecular ozone oxidation kinetics using bench-scale assays and natural water matrices representative of conventionally treated surface water. 3. Experimental section 3.1. Chemicals and reagents Cyclophosphamide and methotrexate (certified purity P99%) were purchased from Sigma–Aldrich Canada (Oakville, ON, Canada). Their chemical structures and their main physico-chemical properties are presented in Table 1. The pKa value of cyclophosphamide is given as a range because no exact value can be found in the literature (Wang et al., 2009). Methotrexate is a weak organic dicarboxylic acid (Reid et al., 1993). Both substances are hydrophilic (log Kow < 2: low sorption potential). LC–MS grade acetonitrile, water and methanol were obtained from J.T. Baker (Phillipsburg, NJ, USA). Acetic acid (from FisherScience, Fair Lawn, NJ, USA) was HPLC-grade reagent and formic acid 98% was purchased from Sigma–Aldrich Canada. L-ascorbic acid (99.0%) and hydrogen peroxide, previously standardized by KMnO4 titration (Harris, 2003), were purchased from Fisher Scientific (Far Lawn, NJ). Ultrapure water (18 MX cm) was produced with a Millipore™ apparatus. Phosphate buffer and tert-butanol (final concentrations: 0.2 M and 50 mM, respectively) were prepared by dissolution of the commercial compounds in ultrapure water.

3.2. Preparation of standards Stock solutions (10 mg L1) of each cytostatic compound were prepared by weighting and dissolving the corresponding pure powders in ultrapure water. Stock solutions were stored and kept at 15 °C before use in order to prevent microbial degradation. Mixed working solutions containing 50 lg L1 of the compounds were prepared each time by dilution of the stock solutions in ultrapure water and kept at 4 °C.

3.3. Instrumentation The on-line SPE procedure for samples and standards to be analyzed are based on the method described by Garcia-Ac et al. (2009). The Environmental Quantification System (EQuan™, Thermo Fisher Scientific, Waltham, MA, USA) was used to carry out the analysis. Also two quaternary LC pumps, an autosampler with a 1 mL-loop (Thermo Fisher Scientific, Waltham, MA, USA) and a TSQ Quantum Mass Spectrometer equipped with an Ion Max API Source (Thermo, Waltham, MA) and an electro-spray ionization (ESI) multimode source using compound-specific fragments. Tube lens and collision energies which are compound-specific appear in Table 2. Method detection limits were 3 and 6 ng L1 for cyclophosphamide and methotrexate, respectively.

1058

A. Garcia-Ac et al. / Chemosphere 79 (2010) 1056–1063

Table 1 Physico-chemical properties and molecular structure of target drugs. MW (g mol1)

pKa

log Kow

Cyclophosphamide C7H15Cl2N2O2P

260.02

4.5–6.51

0.631

Methotrexate C20H22N8O5

454.45

4.8 and 5.52

1.283

Compound

Structure

MW: molecular weight; pKa: dissociation constant; log Kow: partition coefficient (octanol/water). 1 Reference Wang et al. (2009). 2 Reference Reid et al. (1993). 3 Reference Sanderson and Thomsen (2009).

Table 2 Collision energies, tube lens, precursor and fragment ions of the compounds of interest generated by electro-spray ionization (ESI). Cited references give explanations for the ESI conditions of the observed fragments. Analyte

Cyclophosphamide Methotrexate 1 2

ESI CE (V)

Tube lens

23 21

81.1 101.6

Observed precursor ion

Observed fragment and reference

[M + H]+ m/z 261 [M + H]+ m/z 455

[M–PNHO2]+ m/z 1401 [M–NHC3H5(COOH)2]+ m/z 3082

References Castiglia et al. (2005), Liu et al. (2005) and Barbieri et al. (2006). Reference Barbieri et al. (2006).

3.4. Natural water samples Natural filtered water samples were collected before the ozonation process from a single municipal drinking water treatment plant (DWTP) in the province of Quebec drawing water from the St-Lawrence River. This water was sampled in 10-L polypropylene carboys that had been thoroughly washed and rinsed successively with distilled and ultrapure waters. Samples were filtered through 0.45-lm polyethersulfone membranes and kept at 4 °C prior to the ozone experiments. The filtered water had a pH of around 8.04, low turbidity (<0.15 NTU), low natural organic matter (2.55 mg L1 dissolved organic carbon (DOC)) and moderate alkalinity (80 mg L1 CaCO3). 3.5. Bench-scale ozonation experiments Bench-scale ozone experiments were performed at 20 ± 1 °C. In all experiments, water samples were spiked with target compounds to achieve concentrations of cytostatics between 100 and 200 ng L1 depending on the compound and the detection limit of the method of determination (LOD) (LOD methotrexate = 6 ng L1 and LOD cyclophosphamide = 3 ng L1). During direct reaction with molecular ozone, ultrapure water was buffered to pH 8.10 with a phosphate buffer to account for the pH of the natural water, while the pH of the natural water

was left unbuffered. Tert-butanol was added in buffered ultrapure water and in natural water to quench OH radicals generated by ozone decomposition. This allowed the determination of the second-order rate constants for the reaction of molecular ozone with the target cytostatic drugs (kO3 ) in buffered ultrapure and natural water, and assessing the impact of water quality on molecular ozone oxidation kinetics. Stock ozone solutions (50–60 mg L1 O3) were prepared by diffusing gaseous ozone, produced with an oxygen-fed generator (Ozone services, BC, CA) through a waterjacketed flask containing ultrapure water chilled at 4 °C. The ozonation experiments were initiated by mixing aqueous ozone (6–15 mg L1) and solutions of cytostatic drugs (100–200 ng L1). Aliquots of ozone solutions were injected via a syringe into a continuously stirred 1-L glass reactor containing the water sample and equipped with a floating Teflon lid to prevent degassing. Aliquots of 4 mL were then withdrawn from the test water at regular time intervals for ozone residual analysis. For target compounds analysis, aliquots of 40 mL were withdrawn and transferred into vials containing 400 lL of ascorbic acid (5 g L1) to quench the residual ozone and prevent microbial degradation (Westerhoff et al., 2005). Target compounds concentrations were analyzed for contact times from 0 to 120 min. When required, hydrogen peroxide was dosed in buffered ultrapure water at a concentration of 0.25 mg H2O2 per mg O3 prior to the injection of ozone (10 mg L1) in order to determine the

1059

A. Garcia-Ac et al. / Chemosphere 79 (2010) 1056–1063

second-order rate constant for the reaction of cytostatics with OH radicals (kOH) and due to the resistance of one of the target compounds to direct ozonation. This concentration was selected as it is in the optimal range to promote AOP (Acero and Von Gunten, 2001). Experiments were also conducted with natural water free from tert-butanol and hydrogen peroxide in order to verify the natural promotion of OH radicals when ozone was applied (10 mg L1). Ozone oxidation reaction was then a combination of direct and indirect reactions with molecular ozone and OH radicals, respectively. In these previous experiments and after ozone injection, samples were collected for ozone residual analysis following the method previously described. Moreover, for pCBA analysis, a sub-sample of 2 mL was collected and analyzed using a HPLC system. Initial ozone consumption was measured as the difference between the ozone dosage and the ozone residual after 10 s. Initial ozone consumption (mg L1) was then converted in percentage of the ozone dosage for subsequent comparison between ozone experiments. 3.6. Ozone residual and pCBA analysis At bench-scale, ozone stock solution concentrations and ozone residuals in water were determined according to the standard colorimetric method 4500-O3 using indigo trisulfonate (e600 nm = 20 000 M1 cm1). Samples were analyzed at 600 nm with a Varian spectrophotometer (Cary 100, Victoria, Australia) in a 1-cm or 2-cm quartz cell. Para-chlorobenzoic acid (pCBA) was analyzed using a HPLC system (column characteristics: Nucleosil 100-5, RP-C18, 150 mm length, 4.6 lm i.d., from Macherey–Nagel). The mobile phase consisted of 60% water (acidified at pH 2 with H2SO4) and 40% methanol (HPLC grade). The detection of pCBA was made using diode array detection at 236 nm (Elite LaChrom, Hitachi). A method detection limit of 0.2 lg L1 was obtained.

tween the ozone residual and the time (min) and k0 = ozone first-order decay constant (min1). CT values in mg min L1 were converted in M s for a subsequent comparison of reaction kinetics (M1 s1) with data from the literature. 3.8. Reaction kinetics with OH radicals A competition kinetics method was used to determine the second-order rate constants for the reaction between the cytostatic drugs and hydroxyl radicals (kOH). The rate constants were calculated by following the concentration of an OH-probe compound, which can indirectly provide the concentration of OH radicals during the ozonation process following the pCBA method (Elovitz and von Gunten, 1999). Final integration of data yields:

ln



½CDt ½CD0



  ¼  kOH Rct þ kO3

Z

t

½O3 dt

ð4Þ

0

where kOH and kO3 are the second-order rate constants for the reactions of cytostatic drugs CD with OH and O3, respectively; Rct is defined as the ratio between the OH radicals and molecular O3 exposure; ozone exposure (CT value) was calculated as described in Section 3.7. 3.9. Data analysis The results were analyzed using Statistica software version 7.1 (Statistica 7.0). Unless otherwise mentioned, standard deviations are used to characterize the uncertainty in kinetic constants and statistical significance was set at p = 0.05. 4. Results and discussion 4.1. Analytical determination of the chemotherapy drugs

3.7. Determination of rate constants for the reaction with molecular ozone The value of the rate constant with molecular ozone was determined in buffered ultrapure water at pH 8.10 and at ambient pH in natural water, in the presence of tert-butanol as hydroxyl radical scavenger. Von Gunten (2003) reported that the kinetics of ozone reactions with organic and inorganic compounds is typically second-order, i.e. first-order both with ozone and the contaminant concentration. The degradation of cytostatics with molecular ozone can be described by:



d½CD ¼ kO3 ½O3 ½CD dt

ð1Þ

where CD = target cytostatic drug and kO3 = reaction rate constant with molecular ozone. The rate constant is obtained from the integration of Eq. (1):

ln



½CDt ½CD0

 ¼ kO3

Z

t

½O3 dt

ð2Þ

0

R

where [O3]dt is the time-integrated ozone concentration or ozone exposure CT (oxidant concentration  contact time). In this study, CT values (mg min L1) were calculated using the integrated CT concept (Barbeau et al., 2005), for which the effective CT at time t (min) is equal to the area under the decay curve at that time. CT values were calculated using the ozone concentration profiles (Eq. (3)) assuming a simple first-order decay:

CTeffective ¼

Z

The normalized selected reaction monitoring (SRM) chromatograms for a DWTP natural water sample spiked with methotrexate and cyclophosphamide are illustrated in Fig. 1. Fig. 1a illustrates the good separation of both analytes in an initial solution spiked with 200 ng L1. Since the analysis was conducted using SRM, quantification of each specified ion was not subjected to interferences from the others. After two minutes of reaction with a 10 mg L1 ozone dosage, methotrexate practically disappeared (97% conversion), while for cyclophosphamide a small peak was still observed equivalent to a removal rate of 87% (Fig. 1b). Fig. 1c shows a chromatogram of the same sample taken at 30 min of reaction time. At this time, none of the two compounds were detected taking into account the detection limit of the method of determination of the compounds. From those chromatograms, methotrexate was easier to oxidize than cyclophosphamide. To obtain accurate concentrations of the selected chemotherapy drugs, calibration curves in the range of 50–500 ng L1 were established for both compounds (R2 > 0.98, data not shown).

CðtÞ  dt ¼

Co 0 0 ½1  expðk  tÞ k

ð3Þ

where C = ozone residual (mg L1); Co = initial ozone residual (mg L1) determined from the exponential fit of the relation be-

4.2. Ozone decomposition Examples of dissolved ozone decay curves in buffered ultrapure water (pH = 8.10) and in DWTP natural water samples are presented in Fig. 2. Assays were performed under the following conditions: (#1) natural water (10 mg L1 O3), (#2) natural water spiked with tert-butanol (10 mg L1 O3), (#3) natural water (10 mg L1 O3) spiked with hydrogen peroxide (2.5 mg L1 H2O2) and (#4) buffered ultrapure water (pH = 8.10) spiked with tert-butanol (10 mg L1 O3). This experimental design enables the calculation of the second-order reaction rates kO3 and kOH in buffered ultrapure water and in natural water and the comparison of the efficacy of

1060

A. Garcia-Ac et al. / Chemosphere 79 (2010) 1056–1063

(a) 0 min 2.90

100

100

NL: 1.55E4

90

(c) 30 min

(b) 2 min 100

NL: 4.18E2

90

90

80

80

80

70

70

70

60

60

60

50

50

40

40

30

30

30

20

20

20

10

10

10

0

0

0

100

50

2.91

40

100

3.7

90

NL: 1.26E5

NL: 1.67E4

80

70

70

70

60

60

60

50

50

50

40

40

40

30

30

30

20

20

20

10

10

10

0

0 2

3

4

5

6

7

8

NL: 1.94E3

90

80

1

2.90

100 3.77

90

80

0

NL: 3.59E2

3.7

0 0

1

2

3

Time (min)

4

5

Time (min)

6

7

8

0

1

2

3

4

5

6

7

8

Time (min)

Fig. 1. Chromatograms of methotrexate (upper graphs) and cyclophosphamide (lower graphs) during ozonation in natural water samples: (a) initial concentration of methotrexate and cyclophosphamide (200 ng L1); (b) results at 2 min of reaction time; (c) results at 30 min of reaction time (ordinate axis: relative abundance).

9.0

Ozone residual (mg L-1)

8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0

0

10

20

30

40

50

60

70

Reaction time (min) Fig. 2. Examples of ozone residual decay for ozonation in buffered ultrapure water (pH = 8.10) and in DWTP natural water samples (pH = 8.04); h natural water – 10 mg L1 O3; j natural water + tert-butanol – 10 mg L1 O3; s natural water – 10 mg L1 O3 + 2.5 mg L1 H2O2; d buffered ultrapure water + tert-butanol – 10 mg L1 O3.

dosages were chosen in order to correctly assess second-order kinetics. In waters spiked with tert-butanol, ozone was, as expected, more stable (Fig. 2). Ozone stability in the natural water spiked with tert-butanol was slightly higher than the one in the buffered ultrapure water, which could be explained by the moderate alkalinity (80 mg L1 CaCO3) of the natural water as an additional presence of carbonate as an OH scavenger. In natural waters free of tert-butanol, the impact of OH radicals formation on ozone is clearly visible. Finally, the addition of H2O2 (0.25 mg H2O2 per mg O3) in natural water significantly accelerated ozone decomposition and no residual ozone was detected beyond 1 min. Ozone decay was modeled with the first-order kinetic (R2 > 0.94). Ozone decay rates (k0 ) varied over two orders of magnitude (from 0.024 to 5.15 min1) depending on ozonation conditions. Ozone half-lives were derived from ozone decay rates. They varied from 5.5 min in natural water to 21.1 (±4.2) min and 26.7 (±3.6) min in buffered ultrapure water and in natural water both spiked with tert-butanol, respectively. The addition of hydrogen peroxide (0.25 mg H2O2 per mg O3) lowered ozone half-life to <0.5 min. Overall, initial ozone consumption varied between 33.4% and 54.4% of the ozone dosage. 4.3. Ozone oxidation of target compounds

conventional ozonation with advanced oxidation in natural water. Ozone dosages typically used in drinking water treatment practice are lower than 1 mg O3 per mg TOC (Speitel et al., 1993). Due to the high resistance of one of the target compounds to ozonation, high

Methotrexate reacted rapidly with molecular ozone (10 mg L1 O3) (data not shown) with removals exceeding 1.4 log. This drug was never detected in any ozonated water samples. Additional

A. Garcia-Ac et al. / Chemosphere 79 (2010) 1056–1063

experiments were conducted by spiking 2 lg L1 of methotrexate and injecting ozone dosages 62 mg L1 O3. However, removals exceeded 2.0 log at 0.80 mg min L1 (kapp > 3.6  103 M1 s1, where kapp is the apparent rate constant). Methotrexate contains amino groups, an aromatic ring and two N-containing aromatic rings, which are moieties known to quickly react with ozone. Because of the fast oxidation, it was not possible to determine the reaction rate constants with molecular ozone or OH radicals using the same experimental methods applied for cyclophosphamide. Rey et al. (1999) showed that ozonation was effective in removing methotrexate using an initial drug concentration varying between 454 and 1000 mg L1 (pH 3.0 or 7.0, T = 25 ± 0.1 °C) during semi-batch experiments. According to von Gunten (2003), compounds that react rapidly with molecular ozone at dosages typically used for drinking water treatment have an oxidation kinetics with molecular ozone kO3 > 104 M1 s1. A competition kinetics method is under development to confirm these preliminary data and determine the degradation rate using a reference compound. But still, given the fast kinetics, methotrexate represents less environmental and human health issues given its ease of removal. Removal of cyclophosphamide as a function of ozone exposure (CT) is presented in Fig. 3. In order to estimate the kinetic rate constant kO3 for the reaction between molecular ozone and cyclophosphamide in buffered ultrapure water, tert-butanol was added as a scavenger to exclude the effects of OH radicals formed due to the ozone decomposition. From these experiments, it was observed that the oxidation of cyclophosphamide was limited, therefore indicating a slow reaction between this compound and molecular ozone. The linear relationship (R2 = 0.90) yields a calculated rate constant of kO3 = 3.3 ± 0.2 M1 s1 (p = 6  105) in buffered ultrapure water suggesting that molecular ozone alone would be ineffective in removing this drug unless very high ozone CT values are provided. Under typical water treatment conditions (i.e., pH  6–8), molecular ozone reactions will play a minor role in the degradation mechanism of cyclophosphamide. Molecular ozone reaction is a selective process and cyclophosphamide does not contain unsaturated bonds and aromatic moieties but two electron-withdrawing groups (chloro-substituted double-bond) that lower reactivity of organic molecules toward ozone (von Gunten, 2003). There are no functional groups present in the cyclophosphamide molecule that would be expected to exhibit any more than minimal reactivity toward ozone. Chen et al. (2008) observed similar trends but reported value of the second-order rate

CT (mg min L-1) 0.0

0

50

100

150

200

250

300

350

-0.5

ln (C/C0)

-1.0 -1.5 -2.0 -2.5 -3.0 -3.5 -4.0 Fig. 3. Evolution of normalized cyclophosphamide concentration with CT value. Symbols: h natural water (10 mg L1 O3); j natural water spiked with tert-butanol (10 mg L1 O3); d buffered ultrapure water spiked with tert-butanol (10 mg L1 O3).

1061

constant two orders of magnitude higher (143 M1 s1) than the rate constant calculated in this study. The applied ozone concentration was 2 mg L1 at pH 7.1 with an initial drug concentration between 260 and 1041 lg L1 while using a 10 mM concentration of tert-butanol. However, the temperature used during the experiments is unknown. The cause for such important difference in second-order rate constants remains unanswered at this time. In natural water spiked with tert-butanol (R2 = 0.90), the kinetic rate constant for the reaction between molecular ozone and the drug cyclophosphamide was 2.9 ± 0.3 M1 s1 (p = 8  103) and was not statistically different than in buffered ultrapure water with tert-butanol (p > 0.05) (Fig. 3). Therefore, the reaction rates determined in ultrapure water can be applied to assess the behaviour of the target compound in natural waters exposed to molecular ozone. This result was previously demonstrated (Huber et al., 2003, 2005) where the authors studied the reaction of selected pharmaceuticals with ozone in natural waters. However, studies with chlorine showed faster cyanotoxins degradation rates in natural waters when compared with deionised water (Xagoraraki et al., 2006; Daly et al., 2007), tending to show that the reaction of oxidants and natural organic matter could produce more reactive species than chlorine alone. From the rate constants determined here, it can be predicted that the half-life of cyclophosphamide would have varied from 164 to 190 min with an ozone concentration of about 1 mg L1 (21 lM). In natural water (without tert-butanol or hydrogen peroxide), under which condition oxidation reaction is a combination of molecular ozone and OH radicals reactions, the depletion of the drug was higher (Fig. 3). However, a CT value of 45 mg min L1 was required to remove 1.4 log (96%) of cyclophosphamide. This result clearly demonstrated that ozonation only would not be sufficient to completely remove this drug. A CT value of approximately 1.0 mg min L1 is appropriate for primary disinfection of Giardia (1 log, T 6 1 °C) and virus inactivation (2 log, T 6 1 °C) with ozone. Using ozone in buffered solutions and without tertbutanol Venta et al. (2005) showed that only 4% and 30% of cyclophosphamide was degraded in aqueous solutions after 6 min of reaction time at pH 7 and pH 9, respectively. The greater degradation at pH 9 was explained by the formation of OH radicals by ozone decomposition. Several studies have previously shown the effectiveness of combining ozone with hydrogen peroxide to enhance the oxidation of trace micropollutants (Ning and Graham, 2008). In our experiments, when OH radicals formation was promoted with H2O2 addition, the oxidation of cyclophosphamide was indeed very fast (Fig. 2) in accordance with the results of Venta et al. (2005), who used 6 and 15 min of reaction time with a O3/H2O2 molar concentration ratio of 3:1. However, it was not possible to calculate the second-order rate constant for the reaction of cyclophosphamide with OH radicals in ultrapure water due to the limited number of ozone residual and pCBA data. Therefore, the kinetic rate constant for the reaction between OH radicals and cyclophosphamide (kOH) was assessed using the ozonation experiments performed in natural water (without tert-butanol). The Rct obtained, defined as the ratio between the OH radicals and molecular O3 exposure, was 7.5  108 at 20 °C. The kinetic rate constant for the reaction between OH radicals and cyclophosphamide was therefore calculated as 2.0  109 M1 s1. Chen et al. (2008) calculated that the secondorder rate constant was equal to 2.1  109 M1 s1. These values fall well within the typical range of reported OH rate constants, which vary between 108 M1 s1 and 1010 M1 s1 (Westerhoff et al., 2005). Further research is necessary to examine the efficiency of drinking water treatment processes, and especially ozonation, to remove a range of relevant cytostatic drugs as well as their by-products and the toxicity of the ozonated mixture. As an example, an

1062

A. Garcia-Ac et al. / Chemosphere 79 (2010) 1056–1063

ozonation by-product of cisplatin was shown to be recalcitrant to reaction with molecular ozone (Hernandez et al., 2008). Moreover, other studies showed the formation of persistent ozonation byproducts of two antibiotics (Radjenovic et al., 2009). 5. Conclusion The oxidation kinetics of cyclophosphamide with molecular ozone and OH radicals were investigated at bench-scale. The second-order rate constant for direct ozone reaction with cyclophosphamide was determined as 3.3 ± 0.2 M1 s1 in ultrapure water buffered to pH 8.10. The impact of the natural water matrix on the efficacy was not statistically significant during direct ozone reaction (2.9 ± 0.3 M1 s1). Overall process performance was improved for cyclophosphamide in natural water, because of the combination of direct and indirect reactions pathways. The kinetic rate constant for reaction of cyclophosphamide with hydroxyl radicals was estimated at 2.0  109 M1 s1. In contrast, methotrexate reacted quickly with molecular ozone (kapp > 3.6  103 M1 s1). This study showed that ozone was very effective to oxidize methotrexate but high CT values would be required to remove cyclophosphamide in natural water matrix. The chemical structure and toxicity of the reaction by-products should be determined. Acknowledgments The authors would like to acknowledge the technical staff from the CREDEAU and the NSERC Industrial Chair on Drinking Water for their support in the laboratory work and their assistance with sample collection. The NSERC Industrial Chair on Drinking Water of Ecole Polytechnique and its partners (City of Montreal, City of Laval and John Meunier Inc.) provided funding for this project. The financial support of the CONACYT (Mexico), the Canadian Foundation for Innovation (equipment) and the Natural Sciences and Engineering Research Council of Canada is also acknowledged. References Acero, J.L., Von Gunten, U., 2001. Characterization of oxidation processes: ozonation and the AOP O3/H2O2. J. Am. Water Works Assoc. 93, 90–100. Aherne, G.W., Hardcastle, A., Nield, A.H., 1990. Cytotoxic drugs and the aquatic environment – estimation of bleomycin in river and water samples. J. Pharm. Pharmacol. 42, 741–742. Barbeau, B., Desjardins, R., Mysore, C., Prevost, M., 2005. Impacts of water quality on chlorine and chlorine dioxide efficacy in natural waters. Water Res. 39, 2024– 2033. Barbieri, A., Sabatini, L., Indiveri, P., Bonfiglioli, R., Lodi, V., Violante, F.S., 2006. Simultaneous determination of low levels of methotrexate and cyclophosphamide in human urine by micro liquid chromatography/ electrospray ionization tandem mass spectrometry. Rapid Commun. Mass Sp. 20, 1889–1893. 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, 597–603. Buerge, I.J., Buser, H.R., Poiger, T., Muller, M.D., 2006. Occurrence and fate of the cytostatic drugs cyclophosphamide and ifosfamide in wastewater and surface waters. Environ. Sci. Technol. 40, 7242–7250. Calisto, V., Esteves, V.I., 2009. Psychiatric pharmaceuticals in the environment. Chemosphere 77, 1257–1274. Carballa, M., Omil, F., Lema, J.M., 2008. Comparison of predicted and measured concentrations of selected pharmaceuticals, fragrances and hormones in Spanish sewage. Chemosphere 72, 1118–1123. Castiglia, L., Miraglia, N., Pieri, M., Genovese, G., Simonelli, A., Basilicata, P., Sannolo, N., Acampora, A., 2005. Mono- and di-iodocyclophosphamide as possible internal standards for cyclophosphamide quantification: characterization by ion trap multi-stage mass spectrometry and effects of iodine–chlorine substitution on the fragmentation pattern. Rapid Commun. Mass Sp. 19, 1858–1866. Chen, Z., Park, G., Herckes, P., Westerhoff, P., 2008. Physicochemical treatment of three chemotherapy drugs: irinotecan, tamoxifen, and cyclophosphamide. J. Adv. Oxid. Technol. 11, 254–260. Clara, M., Kreuzinger, N., Strenn, B., Gans, O., Kroiss, H., 2005. The solids retention time – a suitable design parameter to evaluate the capacity of wastewater treatment plants to remove micropollutants. Water Res. 39, 97–106.

Daly, R.I., Ho, L., Brookes, J.D., 2007. Effect of chlorination on Microcystis aeruginosa cell integrity and subsequent microcystin release and degradation. Environ. Sci. Technol. 41, 4447–4453. Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personal care products in the environment: agents of subtle change? Environ. Health Persp. 107, 907– 938. Elovitz, M.S., von Gunten, U., 1999. Hydroxyl radical ozone ratios during ozonation processes. I-The R-ct concept. Ozone: Sci. Eng. 21, 239–260. Gagnon, C., Lajeunesse, A., Cejka, P., Gagne, F., Hausler, R., 2008. Degradation of selected acidic and neutral pharmaceutical products in a primary-treated wastewater by disinfection processes. Ozone: Sci. Eng. 30, 387–392. Garcia-Ac, A., Segura, P.A., Gagnon, C., Sauve, S., 2009. Determination of bezafibrate, methotrexate, cyclophosphamide, orlistat and enalapril in waste and surface waters using on-line solid-phase extraction liquid chromatography coupled to polarity-switching electrospray tandem mass spectrometry. J. Environ. Monit. 11, 830–838. Halling-Sorensen, B., Nielsen, S.N., Lanzky, P.F., Ingerslev, F., Lutzhoft, H.C.H., Jorgensen, S.E., 1998. Occurrence, fate and effects of pharmaceutical substances in the environment – a review. Chemosphere 36, 357–394. Harris, D.C., 2003. Quantitative Chemical Analysis. W.H. Freeman & Co., New York. Heberer, T., Dunnbier, U., Reilich, C., Stan, H.J., 1997. Detection of drugs and drug metabolites in ground water samples of a drinking water treatment plant. Fresen. Environ. Bull. 6, 438–443. Hernandez, C., Ramos, Y., Fernandez, L.A., Ledea, O., Bataller, M., Veliz, E., Besada, V., Rosado, A., 2008. Ozonation of cisplatin in aqueous solution at pH 9. Ozone: Sci. Eng. 30, 189–196. Huber, M.M., Canonica, S., Park, G.Y., Von Gunten, U., 2003. Oxidation of pharmaceuticals during ozonation and advanced oxidation processes. Environ. Sci. Technol. 37, 1016–1024. Huber, M.M., Gobel, A., Joss, A., Hermann, N., Loffler, D., McArdell, C.S., Ried, A., Siegrist, H., Ternes, T.A., von Gunten, U., 2005. Oxidation of pharmaceuticals during ozonation of municipal wastewater effluents: a pilot study. Environ. Sci. Technol. 39, 4290–4299. Ikehata, K., Naghashkar, N.J., Ei-Din, M.G., 2006. Degradation of aqueous pharmaceuticals by ozonation and advanced oxidation processes: a review. Ozone: Sci. Eng. 28, 353–414. Johnson, A.C., Jurgens, M.D., Williams, R.J., Kummerer, K., Kortenkamp, A., Sumpter, J.P., 2008. Do cytotoxic chemotherapy drugs discharged into rivers pose a risk to the environment and human health? An overview and UK case study. J. Hydrol. 348, 167–175. Jolivet, J., Cowan, K.H., Curt, G.A., Clendeninn, N.J., Chabner, B.A., 1983. The pharmacology and clinical use of methotrexate. N. Engl. J. Med. 309, 1094– 1104. Jones, O.A., Lester, J.N., Voulvoulis, N., 2005. Pharmaceuticals: a threat to drinking water? Trends Biotechnol. 23, 163–167. Joss, A., Andersen, H., Ternes, T., Richle, P.R., Siegrist, H., 2004. Removal of estrogens in municipal wastewater treatment under aerobic and anaerobic conditions: consequences for plant optimization. Environ. Sci. Technol. 38, 3047–3055. Kim, I.H., Tanaka, H., Iwasaki, T., Takubo, T., Morioka, T., Kato, Y., 2008. Classification of the degradability of 30 pharmaceuticals in water with ozone, UV and H2O2. Water Sci. Technol. 57, 195–200. Kummerer, K., 2009a. Antibiotics in the aquatic environment – a review – Part I. Chemosphere 75, 417–434. Kummerer, K., 2009b. Antibiotics in the aquatic environment – a review – Part II. Chemosphere 75, 435–441. Kummerer, K., StegerHartmann, T., Meyer, M., 1997. Biodegradability of the antitumour agent ifosfamide and its occurrence in hospital effluents and communal sewage. Water Res. 31, 2705–2710. Kummerer, K., Al-Ahmad, A., Bertram, B., Wiessler, M., 2000. Biodegradability of antineoplastic compounds in screening tests: influence of glucosidation and of stereochemistry. Chemosphere 40, 767–773. Kuo, C.Y., Wu, H.L., Kou, H.S., Chiou, S.S., Wu, D.C., Wu, S.M., 2003. Simultaneous determination of methotrexate and its eight metabolites in human whole blood by capillary zone electrophoresis. J. Chromatogr. A 1014, 93–101. Lenz, K., Hann, S., Koellensperger, G., Stefanka, Z., Stingeder, G., Weissenbacher, N., Mahnik, S.N., Fuerhacker, M., 2005. Presence of cancerostatic platinum compounds in hospital wastewater and possible elimination by adsorption to activated sludge. Sci. Total Environ. 345, 141–152. Liu, Z.F., Chan, K.K., Wang, J.J., 2005. Tandem mass spectrometric analysis of cyclophosphamide, ifosfamide and their metabolites. Rapid Commun. Mass Sp. 19, 2581–2590. Ljubic, I., Sabljic, A., 2002. Theoretical study of the mechanism and kinetics of gasphase ozone additions to ethene, fluoroethene, and chloroethene: a multireference approach. J. Phys. Chem. A 106, 4745–4757. Mahnik, S.N., Lenz, K., Weissenbacher, N., Mader, R.M., Fuerhacker, M., 2007. Fate of 5-fluorouracil, doxorubicin, epirubicin, and daunorubicin in hospital wastewater and their elimination by activated sludge and treatment in a membrane-bio-reactor system. Chemosphere 66, 30–37. Matamoros, V., Hijosa, M., Bayona, J.M., 2009. Assessment of the pharmaceutical active compounds removal in wastewater treatment systems at enantiomeric level. Ibuprofen and naproxen. Chemosphere 75, 200–205. Mompelat, S., Le Bot, B., Thomas, O., 2009. Occurrence and fate of pharmaceutical products and by-products, from resource to drinking water. Environ. Int. 35, 803–814. Ning, B., Graham, N.J.D., 2008. Ozone degradation of iodinated pharmaceutical compounds. J. Environ. Eng. ASCE 134, 944–953.

A. Garcia-Ac et al. / Chemosphere 79 (2010) 1056–1063 Radjenovic, J., Godehardt, M., Petrovic, M., Hein, A., Farre, M., Jekel, M., Barcelo, D., 2009. Evidencing generation of persistent ozonation products of antibiotics roxithromycin and trimethoprim. Environ. Sci. Technol. 43, 6808– 6815. Reid, T., Yuen, A., Catolico, M., Carlson, R.W., 1993. Impact of omeprazole on the plasma-clearance of methotrexate. Cancer Chemoth. Pharm. 33, 82–84. Rey, R.P., Padron, A.S., Leon, L.G., Pozo, M.M., Baluja, C., 1999. Ozonation of cytostatics in water medium. Nitrogen bases. Ozone: Sci. Eng. 21, 69–77. Rowney, N.C., Johnson, A.C., Williams, R.J., 2009. Cytotoxic drugs in drinking water: a prediction and risk assessment exercise for the Thames catchment in the United Kingdom. Environ. Toxicol. Chem. 28, 2733–2743. Sabljic, A., Peijnenburg, W., 2001. Modeling lifetime and degradability of organic compounds in air, soil, and water systems – (IUPAC technical report). Pure Appl. Chem. 73, 1331–1348. Sanderson, H., Thomsen, M., 2009. Comparative analysis of pharmaceuticals versus industrial chemicals acute aquatic toxicity classification according to the United Nations classification system for chemicals. Assessment of the (Q)SAR predictability of pharmaceuticals acute aquatic toxicity and their predominant acute toxic mode-of-action. Toxicol. Lett. 187, 84–93. Segura, P.A., Garcia-Ac, A., Lajeunesse, A., Ghosh, D., Gagnon, C., Sauve, S., 2007. Determination of six anti-infectives in wastewater using tandem solid phase extraction and liquid chromatography–tandem mass spectrometry. J. Environ. Monit. 9, 307–313. Snyder, S.A., Wert, E.C., Rexing, D.J., Zegers, R.E., Drury, D.D., 2006. Ozone oxidation of endocrine disruptors and pharmaceuticals in surface water and wastewater. Ozone: Sci. Eng. 28, 445–460. Snyder, S.A., Wert, E., Lei, H.D., Westerhoff, P., Yoon, Y., 2007. Removal of EDCs and pharmaceuticals in drinking and reuse treatment processes. American Water Works Association Research Founfation (AWWARF), Denver, Colorado, USA. Speitel, G.E., Symons, J.M., Diehl, A.C., Sorensen, H.W., Cipparone, L.A., 1993. Effect of ozone dosage and subsequent biodegradation on removal of DBP precursors. J. Am. Water Works Assoc. 85, 86–95. Stackelberg, P.E., Furlong, E.T., Meyer, M.T., Zaugg, S.D., Henderson, A.K., Reissman, D.B., 2004. Persistence of pharmaceutical compounds and other organic wastewater contaminants in a conventional drinking-water treatment plant. Sci. Total Environ. 329, 99–113. Statistica 7.0, S., Inc., Okla., USA.

1063

StegerHartmann, T., Kummerer, K., Hartmann, A., 1997. Biological degradation of cyclophosphamide and its occurrence in sewage water. Ecotox. Environ. Safe. 36, 174–179. Ternes, T.A., Hirsch, R., Mueller, J., Haberer, K., 1998. Methods for the determination of neutral drugs as well as betablockers and beta(2)-sympathomimetics in aqueous matrices using GC/MS and LC/MS/MS. Fresen. J. Anal. Chem. 362, 329– 340. Ternes, T.A., Meisenheimer, M., McDowell, D., Sacher, F., Brauch, H.J., Gulde, B.H., Preuss, G., Wilme, U., Seibert, N.Z., 2002. Removal of pharmaceuticals during drinking water treatment. Environ. Sci. Technol. 36, 3855–3863. Ternes, T.A., Stuber, J., Herrmann, N., McDowell, D., Ried, A., Kampmann, M., Teiser, B., 2003. Ozonation: a tool for removal of pharmaceuticals, contrast media and musk fragrances from wastewater? Water Res. 37, 1976–1982. Venta, M.B., Castro, C.H., Garcia, L.A.F., Marzo, A.L., Lorenzo, E.V., Alvarez, C.A., 2005. Effect of O3/H2O2 molar concentration ratio at different pH values on cyclophosphamide degradation. J. Water Supply Res. T. 54, 403–410. Vieno, N.M., Harkki, H., Tuhkanen, T., Kronberg, L., 2007. Occurrence of pharmaceuticals in river water and their elimination a pilot-scale drinking water treatment plant. Environ. Sci. Technol. 41, 5077–5084. von Gunten, U., 2003. Ozonation of drinking water: Part I. Oxidation kinetics and product formation. Water Res. 37, 1443–1467. Wang, L., Albasi, C., Faucet-Marquis, V., Pfohl-Leszkowicz, A., Dorandeu, C., Marion, B., Causserand, C., 2009. Cyclophosphamide removal from water by nanofiltration and reverse osmosis membrane. Water Res. 43, 4115–4122. Westerhoff, P., Yoon, Y., Snyder, S., Wert, E., 2005. Fate of endocrine-disruptor, pharmaceutical, and personal care product chemicals during simulated drinking water treatment processes. Environ. Sci. Technol. 39, 6649–6663. Xagoraraki, I., Harrington, G.W., Zulliger, K., Zeier, B., Krick, W., Karner, D.A., Standridge, J.H., Westrick, J., 2006. Inactivation kinetics of the cyanobacterial toxin microcystin-LR by free chlorine. J. Environ. Eng. ASCE 132, 818–823. Zhang, Y.J., Geissen, S.U., Gal, C., 2008. Carbamazepine and diclofenac: removal in wastewater treatment plants and occurrence in water bodies. Chemosphere 73, 1151–1161. Zounkova, R., Odraska, P., Dolezalova, L., Hilscherova, K., Marsalek, B., Blaha, L., 2007. Ecotoxicity and genotoxicity assessment of cytostatic pharmaceuticals. Environ. Toxicol. Chem. 26, 2208–2214. Zuccato, E., Calamari, D., Natangelo, M., Fanelli, R., 2000. Presence of therapeutic drugs in the environment. Lancet 355, 1789–1790.