Drugs of Abuse in Surface and Tap Waters of the Tagus River Basin: Heterogeneous Photo-Fenton Process Is Effective in Their Degradation

Environment International 41 (2012) 35–43

Contents lists available at SciVerse ScienceDirect

Environment International

journal homepage: www.elsevier.com/locate/envint

Drugs of abuse in surface and tap waters of the Tagus River basin: Heterogeneous photo-Fenton process is effective in their degradation

Y. Valcárcel a,⁎, F. Martínez b, S. González-Alonso a, Y. Segura b, M. Catalá c, R. Molina b, J.C. Montero-Rubio d, N. Mastroianni e, M. López de Alda e, C. Postigo e, D. Barceló e a Department of Preventive Medicine, Public Health, Immunology and Medical Microbiology. Faculty of Health Sciences, Rey Juan Carlos University. Avda. Atenas s/n. E-28922 Alcorcón (Madrid), Spain b Department of Chemical and Environmental Technology. School of Experimental Sciences & Technology, Rey Juan Carlos University. Avda. Tulipán s/n. E-28933 Móstoles (Madrid), Spain c Department of Biology and Geology. School of Experimental Sciences & Technology, Rey Juan Carlos University. Avda. Tulipán s/n. E-28933 Móstoles (Madrid), Spain d Health Sciences Institute of Castilla-La Mancha, Talavera de la Reina (Castilla-La Mancha), Spain e Department of Environmental Chemistry. Institute of Environmental Assessment and Water Research (IDAEA-CSIC). Jordi Girona 18–26. E-08034 Barcelona, Spain article info abstract

Article history: This work investigates for the ﬁrst time the occurrence of drugs of abuse and metabolites in surface waters Received 21 October 2011 from the Tagus River on its way through the province of Toledo (downstream Madrid metropolitan area) Accepted 23 December 2011 and in drinking waters in two nearby cities. Some of the studied drugs are used for therapeutic purposes Available online xxxx but they can also be consumed as illicit drugs. The results of this preliminary study have revealed the presence of 12 out of 22 drugs of abuse analyzed in ﬂuvial water at concentrations ranging from 1.14 to Keywords: 40.9 ng/L. The largest concentrations corresponded to the anxiolytics diazepam and lorazepam, the cocaine Drugs of abuse Surface water metabolite benzoilecgonine, the amphetamine-like compound ephedrine, and the methadone metabolite Tap water EDDP. All these substances, except for lorazepam, were detected in all the sampling points. Traces of metha- Photo-Fenton done and ephedrine were detected in some samples of tap water. Despite the low concentrations of these Tagus River pollutants, effects on wildlife or human health cannot be disregarded, especially on vulnerable population. Spain Thus, the treatment of these substances using a heterogeneous photo-Fenton process has been evaluated, rendering a remarkable effectiveness for their degradation. © 2011 Elsevier Ltd. All rights reserved.

1. Introduction

In recent years a great interest has been aroused in the so called Abbreviations: ACN, acetonitrile; ALCs, amphetamine-like compounds; ALP, alpraz- “emerging pollutants” (EPs), which comprise very different substances olam; AM, amphetamine; AOP, advanced oxidation process; 6ACM, 6-acetylmorphine; such as endocrine disruptors, pharmaceuticals, drugs of abuse (DAs), BAS, biological activated sludge; BE, benzoylecgonine; CE, cocaethylene; COC, cocaine; COD, chemical oxygen demand; DAs, drugs of abuse; DIAZ, diazepam; DWTP, drinking and their metabolites are included. The recent development and ap- water treatment plant; DW, drinking water; EDDP, 2-ethylidene-1,5-dimethyl-3,3- plication of advanced analytical methodologies for the detection of diphenylpyrrolidine; EPs, emerging pollutants; ESI, electrospray ionization; GAC, DAs in urban wastewaters have permitted to establish that these granular activated carbon; HER, heroin; LC-MS/MS, liquid chromatography-tandem substances are not effectively eliminated in sewage treatment plants mass spectrometry; IS, internal standard; LODet, method limit of determination; (STPs). As a consequence, the presence of DAs has been reported in LOR, lorazepam; LSD, lysergic acid diethylamide; MA, methamphetamine; MDMA, 3,4-methylenedioxymethamphetamine; METH,methadone;MOR,morphine;MR, different European surface waters of Italy (Zuccato et al., 2008), UK Madrid region; NI, negative ionization; nor-BE, nor-benzoylecgonine; nor-LSD, nor- (Kasprzyk-Hordern et al., 2008), Germany (Hummel et al., 2006), LSD and nor-iso LSD; OH-THC, 11-hydroxy-Δ9-tetrahydrocannabinol; O-H-LSD, 2- Belgium (Gheorghe et al., 2008; van Nuijs et al., 2010), Ireland oxo-3-hydroxy-LSD; PCBs, polychlorinated biphenyls; PI, positive ionization; RSD, (Bones et al., 2007)andCroatia(Terzic et al., 2010). relative standard deviation; RPS, river point sample; SPE, solid phase extraction; SRM, selected reaction monitoring; STP, sewage treatment plant; THC, Δ9-tetrahydrocannabinol; In Spain, several works have studied the occurrence of DAs in waste- THC-COOH, 11-nor-9-carboxy-Δ9-tetrahydrocannabinol; TF, tertiary ﬁltration; TOC, total waters and surface waters from Catalonia (Boleda et al., 2007, 2009; organic carbon. Huerta-Fontela et al., 2007), the Ebro River basin (Postigo et al., 2010) ⁎ Corresponding author at: Ediﬁcio departamental I. Departamento de Medicina (Northeastern Spain), Galicia (Northwestern Spain) (Gonzalez-Marino Preventiva, Salud Pública, Inmunología y Microbiología Médicas. Facultad de Ciencias et al., 2010), the Henares River (Central Spain) (Martinez Bueno de la Salud. Avda Atenas s/n. E-28922 Alcorcón, Madrid, Spain. Tel.: +34 94914888891; fax: +34 914888955. et al., 2010), and L'Albufera National Park in Valencia (Eastern E-mail address: [email protected] (Y. Valcárcel). Spain) (Vazquez-Roig et al., 2010). Meanwhile, Huerta-Fontela and

Boleda et al. (Boleda et al., 2009; Huerta-Fontela et al., 2011)have treatment plants (DWTPs), which supply the finished drinking water also investigated the presence of illicit drugs at very low concentra- to the population. Currently, neither the Spanish nor the European tions in treated drinking waters. legislations demand the analysis of these EPs in treated wastewater The UNODC (United Nations Office on Drugs and Crime, 2008) has for regeneration uses, surface waters or drinking water. recently published the World Drug Report 2011, indicating that, in These EPs have been shown to be very recalcitrant to elimination 2009, between 149 and 272 million people (between 3.3% and 6.1% by physicochemical and biological treatments at conventional STPs. of the population aged 15–64) worldwide have used illicit drugs at In this respect, advanced oxidation processes (AOPs) have emerged least once. While the total number of illicit drug users has increased as a promising alternative for their removal from waters (Boleda et since the late 1990s, the prevalence rates have remained largely al., 2011). The efficiency of these processes is based on the strong stable. Cannabis has been reported to be by far the most widely used oxidizing power of the hydroxyl radicals, which can be generated illicit drug type, followed by amphetamine-type stimulants (mainly using different methods such as photocatalytic systems, hydrogen methamphetamine, amphetamine and ecstasy), opioids (including peroxide catalyzed processes or ozonation (Balcioglu et al., 2003). opium, heroin and prescription opioids) and cocaine. Data for 2009 Among the AOPs assisted by UV–VIS irradiation, the photo-Fenton indicated annual prevalence rates of 2.8–4.5% for cannabis and 0.3– process may be a compelling alternative, since it has proved highly 0.5% for cocaine. efficient in the elimination of multiple polycyclic/aromatic organic A recent Spanish survey on alcohol and drugs (EDADES, 2010) compounds such as phenolic derivatives, chlorinated pesticides, aryl confirms cannabis as the most consumed illicit substance with a and alkyl sulfonates and stains, and EPs like estrogens and pharmaceu- prevalence rate of ca. 32.1% of persons that have ever used this drug tical compounds (Ternes et al., 2003). during their life, which places Spain in the fourth position, after In this context, the main objectives of the present preliminary Denmark, France and the United Kingdom, in the ranking of European study were (i) to assess for the first time the occurrence of 22 DAs countries in the use of cannabis. The second most consumed illicit and metabolites in surface water samples from the Tagus River basin drug is cocaine. 10.2% of the Spanish population has taken cocaine at and in drinking waters from the province of Toledo (center of Spain), some stage and its level of consumption is similar to that of the USA and (ii) to evaluate the potential application of heterogeneous photo- in 2009. Other stimulants, such as ecstasy, amphetamines or several Fenton treatment as an AOP for elimination of the substances detected hallucinogenics show low and stable consumptions, with prevalence in the water samples. rates lower than 1%. Heroin, in addition to having a low prevalence rate (0.6%), shows a constantly decreasing consumption trend (EDADES, 2. Material and methods 2010). DAs are substances that, due to their elevated volumes of produc- 2.1. Sampling tion and consumption, are continuously discharged into the aquatic environment (Jones et al., 2005). These compounds, with an elevated The province of Toledo belongs to the Autonomous Community of pharmacological potential, may have negative effects on wildlife or Castilla-La Mancha and houses its capital, Toledo city. The Tagus River human health (Pomati et al., 2006). Thus, continuous monitoring of enters the province of Toledo by Añover del Tajo (Northeast), crosses the chemical quality of the effluents discharged from urban STPs the province from east to west flowing by Toledo and Talavera de la would be desirable in order to ensure complete elimination of these Reina cities, and leaves it by Azutan (Fig. 1). The main geographical substances and to avoid undesirable effects on the environment. and demographic data of the aforementioned urban cities are includ- This concern is even more imperative in the case of drinking water ed in Fig. 1 (INEbase, 2009).

Fig. 1. Map of Toledo province: cities involved in the study and demographic data (2009). Y. Valcárcel et al. / Environment International 41 (2012) 35–43 37

Table 1 Codes of the sampling points and characteristics of the STPs monitored. The data were obtained from the water quality network of the Tagus Hydrographic Confederation (http:// www.chtajo.es/redes/calidad/calidad.htm), Toledo Town Hall and Talavera de la Reina Town Hall (http://www.aqualia.es/talavera).

Sample Type of Name of STP upstream Efﬂuent of STP Industrial Population Actual Treatment code sample (m3/year) component equivalent population Water Sludge served

RPS2 Fluvial Santamaría de la Bequerencia 7,300,000 20–30% 120,000 19,000 TF d+s RPS3 Fluvial Toledo city 8,760,000 0% 120,000 75,000 BAS d+s RPS4 Fluvial Talavera de la Reina 18,000,000 n.a. 126,000 n.a. BAS s

Abbreviations: TF: tertiary ﬁltration; BAS: biological activated sludge; d: anaerobic digestion; s: spin drying; n.a.: not available.

Surface water samples from the Tagus River were collected at the the heroin metabolic product 6-acetylmorphine (6ACM), the synthetic entry (Añover de Tajo, RPS1) and the exit (Azután, RPS5) of Toledo opioid-agonist methadone (METH), which is usually employed to treat province. The latter, in Azutan, was more specifically collected from heroin addiction, and its main excretion product 2-ethylidene-1,5- a reservoir used for irrigation. Surface waters were also taken approx- dimethyl-3,3-diphenylpyrrolidine (EDDP). imately 100 m downstream of the discharge of two STPs located in High purity (>97%) standard solutions of the target compounds Toledo city (RPS2 and RPS3) and the STP of Talavera de la Reina and their deuterated analogs (used as surrogate standards (SS) for (RPS4). Table 1 summarizes the features of the selected sampling quantification) were obtained from Cerilliant (Round Rock, TX, USA) points. as solutions in methanol or acetonitrile. Additionally, three tap water samples were collected from public places supplied by different DWTPs from Toledo city (DW1 and DW2) 2.3. Analytical method and Talavera de la Reina (DW3). Table 2 indicates the main characteristics of the investigated DWTPs. Water from the Cazalegas reservoir, Upon reception, samples were vacuum filtered through 1-μmglass which is near the inlet of Talavera de la Reina DWTP, was also evaluated fiber filters and 0.45-μm nylon membrane filters (Whatman Interna- (RES). tional Ltd., Maidstone, England) and stored in amber polyethylene Grab water samples (1 L) were collected in pre-rinsed amber glass terephthalate (PET) bottles at −20 °C in the dark until analysis. bottles on June 28th, 2010. The pH and the temperature were imme- The methodology applied to the analysis of illicit drugs and their diately measured, and the samples were thereafter sent to the labora- metabolites in the collected water samples is a variation of a previ- tory for analysis within the next 24 h. ously described fully automated method based on on-line solid phase extraction-liquid chromatography-tandem mass spectrometry (on-line 2.2. Target compounds SPE-LC-MS/MS) (Postigo et al., 2008). Modifications of the previous method include: Up to 22 drugs and metabolites belonging to 6 different chemical (i) the use of a different on-line sample processor system, a classes – amphetamine-like compounds (ALCs), benzodiazepines, can- Symbiosis-Pico, instead of a Prospekt-2 system (both from nabinoids, cocainics, lysergic acid diethylamide (LSD) and metabolites, Spark Holland, Emmen, The Netherlands), for SPE of samples. and opioids – were monitored in the collected waters (see Table 3). (ii) the addition of a few more drugs, namely, methadone, EDDP, The group of ALCs investigated included amphetamine (AM), meth- alprazolam, lorazepam and diazepam (with their correspond- amphetamine (MA), 3,4-methylenedioxymethamphetamine (MDMA ing deuterated analogs for internal standard quantification) or ecstasy), and ephedrine (EPH). The selected benzodiazepines were to the list of target compounds, and alprazolam (ALP), diazepam (DIAZ) and lorazepam (LOR). Target canna- (iii) the use of a modified mobile phase consisting of acetonitrile binoids were the main psychoactive component of the cannabis plant, (ACN) and 10 mM ammonium formate aqueous solution for Δ9-tetrahydrocannabinol (THC), and two metabolic products, 11-nor- analysis of the target compounds in the positive ionization 9-carboxy-Δ9-tetrahydrocannabinol (THC-COOH) and 11-hydroxy-Δ9- (PI) mode. tetrahydrocannabinol (OH-THC). Target cocainics included cocaine (COC), its major metabolite benzoylecgonine (BE), the minor metabo- Prior to extraction, all water samples were spiked with a standard lite nor-benozylecgonine (nor-BE) and the trans-esterification product mixture of the deuterated compounds at a concentration of 100 ng/L cocaethylene (CE) formed when cocaine and ethanol are simulta- for BE-d8, COC-d3,EPH-HCl-d3,andMOR-d3, 50 ng/L for the ALCs neously consumed. The lysergic compounds investigated were the AM-d5,MA-d14 and MDMA-d5 and the benzodiazepines DIAZ-d5, most potent known hallucinogenic substance, LSD, and two of its me- LOR-d4,ALP-d5, and 20 ng/L for the rest of compounds (CE-d3,THC-d3, tabolites, nor-LSD and nor-iso LSD (nor-LSD) and 2-oxo-3-hydroxy- OH-THC-d3, THC-COOH-d3,LSD-d3,HER-d9,6ACM-d6,METH-d3,EDDP- LSD (O-H-LSD). Opioids surveyed were morphine (MOR), heroin (HER), d3,). Five milliliters of spiked water samples were preconcentrated

Table 2 Characteristics of the drinking water samples. The data were obtained from the water quality network of the Tagus Hydrographic Confederation (http://www.chtajo.es/redes/ calidad/calidad.htm), Toledo Town Hall and Talavera de la Reina Town Hall (http://www.aqualia.es/talavera).

Sample Town Population Water treated Treatments Water supply area Sampling site code served (m3/day) PO C/F D Fi S

DW1 Toledo 75,000 12,100 Cl2 and/or O3 Aluminum polychloride Dynamic Sand Cl2 Toledo city Regional (pulsator) administration site

DW2 Toledo 19,000 6600 O3 Aluminum polychloride Static Sand Cl2 Santa María de la Pharmacy Bequerencia

DW3 Talavera de 100,000 248,000 Cl2 and active carbon Dkﬂoc wp 310 Al2(SO4)3 Static Sand Cl2 Talavera Regional

la Reina (partially O3) anionic polyelectrolite administration site

Abbreviations: PO: preoxidation; C/F: coagulation/ﬂocculation; D: decantation; Fi: ﬁltration; S: sterilization. 38 Y. Valcárcel et al. / Environment International 41 (2012) 35–43

Table 3 Target analytes, CAS number, SRM transitions, and method recovery, repeatability and sensitivity.

Chemical group Drug CAS number Parent ion→product ion Absolute (relative) RSDb LODetc 1/product ion 2 recoveries (%)a (%) (ng/L)

ALCs Amphetamine 300-62-9 136.2→119.0/91.0 72 (92) 14.5 4.28 Ephedrine 24221-86-1 166.2→148.0/133.0 78 (106) 4.2 0.45 MDMA 42542-10-9 194.3→163.0/105.0 85 (97) 5.6 1.26 Methamphetamine 4846-07-5 150.2→91.0/119.0 93 (98) 11.0 1.28 Benzodiazepines Alprazolam 28981-97-7 309.0→281.0/205.0 84 (105) 8.2 3.60 Diazepam 439-14-5 285.0→193.0/222.0 84 (99) 9.1 1.40 Lorazepam 846-49-1 321.0→303.0/275.0 199 (110) 17.9 14.60 Cannabinoids THC 1972-08-3 313.5→245.1/191.0 22 (111) 2.30 6.81 THC-COOHd 56354-06-4 343.5→299.5/191.2 57 (98) 3.4 4.19 OH-THCd 36557-05-8 329.5→311.2/268.0 59 (96) 8.9 1.26 Cocainics Benzoilecgoninee 519-09-5 290.3→168.0/77.0 116 (116) 2.3 0.60 Cocaethylenee 529-38-4 304.4→182.0/77.0 80 (100) 14.8 1.47 Cocaine 50-36-2 318.4→196.0/77.0 86 (86) 10.0 0.13 nor-benzoylecgoninee 60426-41-7 276.0→154.0/136.0 85 (85) 14.9 0.96 LSD and metabolites LSD 50-37-3 324.4→208.0/223.0 84 (124) 7.6 1.80 nor-LSD/nor-iso-LSDf 35779-43-2/71953-76-9 310.4→193.0/209.0 70 (85) 12.9 0.86 O-H-LSDf Not available 356.4→237.0/222.0 101 (103) 5.7 0.16 Opioids 6ACMg 2784-73-8 328.4→165.0/152.0 105 (97) 10.7 0.64 EDDPh 66729-78-0 278.0→234.0/249.0 93 (98) 6.8 0.57 Methadone 76-99-3 310.0→268.0/105.0 63 (101) 16.7 0.62 Morphine 57-27-2 286.3→152.0/128.0 78 (111) 3.1 1.45 Heroin 561-27-3 370.4→268.0/165.0 73 (94) 9.2 1.66

a Absolute recovery was calculated from the peak areas obtained in on-line analysis of spiked (50 ng/L) water as percentages of the peak areas obtained from direct chromatographic injection (5 μL) of equivalent amounts of the standards in methanol. Relative recovery: recovery of the analyte relative to the recovery of the associated deuterated analog used as surrogate standard. b RSD: relative standard deviation, spiking concentration: 50 ng/L (n=3). c LODet: limit of determination, minimum concentration that can be quantified (above the limit of quantification of SRM1) and confirmed (above the limit of detection of SRM2). d Metabolite of THC. e Metabolite of cocaine. f Metabolite of LSD. g Metabolite of heroin. h Metabolite of methadone.

onto previously conditioned (1 mL of ACN and 1 mL of LC-grade water) exception of the LSD metabolites and nor-BE, for which LSD-d3 and on-line SPE cartridges. PLRPs cartridges (Spark Holland, Emmen, The BE-d8 were used, respectively. The method was shown to be linear, Netherlands) were used to isolate the compounds measured in the with correlation coefficients higher than 0.995 for all compounds PI mode (all but cannabinoids), and Oasis HLB cartridges (Waters, within the concentration range 0.1–500 ng/L. LC-grade water sam- Barcelona, Spain) were used to preconcentrate the analytes mea- ples were analyzed after the calibration curve and every four real sured in the negative ionization (NI) mode (cannabinoids). Elution water samples in order to check for potential carry over effects. of the analytes from the cartridges to the LC system was done with Additionally, several quality control samples (LC-grade water spiked the chromatographic mobile phase pumped by the Symbiosis Pico with the standard mixture at 50 ng/L) were included in the sample system. batch to control the performance of the analytical instrumentation. Chromatographic separation was achieved with a Purospher Star Overall, analyte absolute recoveries were satisfactory for all com- RP-18 end-capped column (125×2 mm, particle size 5 μm) preceded pounds but lorazepam and THC, which presented though satisfactory by a guard column of the same packing material, both supplied by relative recoveries (110 and 111%, respectively). Method repeatabil- Merck (DaMRstadt, Germany). The mobile phase consisted of a linear ity, calculated as the relative standard deviation (RSD) of the repli- gradient of ACN/water in the NI mode and a linear gradient of ACN/ cate (n=3) analysis of spiked (50 ng/L) water samples was also 10 mM ammonium formate in the PI mode, at a constant flow rate satisfactory, with RSD values below 23%. of 0.3 mL/min. MS/MS analyses were performed with a 4000 QTRAP Average method limits of determination (LODet, minimum con- hybrid triple quadrupole-linear ion trap mass spectrometer equipped centration of a compound that can be quantified and confirmed) with a turbo ion spray source (Applied-Sciex, Foster City, CA). For re- were between 0.13 ng/L (COC) and 14.60 ng/L (amphetamine). For liable quantitative and confirmatory analysis, data acquisition was more details about LODet calculations, see Postigo et al., (Postigo et performed in the selected reaction monitoring (SRM) mode, record- al., 2008). ing the transitions between the precursor ion and the two most abundant product ions for each target analyte, thus achieving 4 iden- 2.5. Photo-Fenton experiments for degradation of drugs tification points per compound. Photo-Fenton experiments were performed with a 150 W medium 2.4. Quality control/quality assurance pressure mercury lamp (Heraeus TQ-150). The lamp was surrounded by a quartz jacket in which a copper sulfate aqueous solution circulates The analytical methodology was validated in terms of linearity, to block radiation at wavelengths shorter than 313 nm. Typically, a repeatability, accuracy and sensitivity. Further details on their calcu- cylindrical glass vessel was filled with 1 L of the water sample. The lation can be found in Postigo et al. (Postigo et al., 2008). Quantifica- catalyst was suspended into the aqueous solution (0.6 g/L) and the tion, based on peak areas, was performed by the internal standard pH was initially adjusted to ca. 3 with H2SO4 (0.1 M) and left uncon- (IS) method, which allowed correcting potential matrix effects that trolled during the treatment. The initial hydrogen peroxide concen- usually occur in electrospray ionization (ESI). Each analyte was tration used for the treatments was calculated according to the quantified using as IS its corresponding deuterated analog, with the stoichiometric amount for the complete mineralization of the total Y. Valcárcel et al. / Environment International 41 (2012) 35–43 39

3. Results and discussion organic carbon towards CO2 and H2O. In this case a powder silica- supported iron oxide (Fe2O3/SBA-15) was used as heterogeneous The study was focused on the analysis of different DAs in surface waters of the photo-Fenton-like catalyst facilitating its separation after the treat- Tagus River or reservoirs used for abstraction of drinking water or for irrigation pur- ment. This catalyst was prepared following a method described else- poses. Moreover, contamination by these substances in three tap water samples supplied by different DWTPs of Toledo province was also tackled. The removal of these where (Lazar et al., 2004). Fe O /SBA-15 is characterized by containing 2 3 DAs in surface water samples was evaluated by a photo-Fenton process based on the crystalline iron oxides, in the form of hematite, supported on a mesos- application of a heterogeneous iron-containing catalyst (Fe2O3-SBA-15) in the pres- tructured SBA-15 silica support. The properties of the silica matrix are ence of hydrogen peroxide and UV–vis irradiation. characteristic of a 2-dimensional hexagonally mesostructured SBA-15 material, with a BET surface area of approximately 470 m2/g and a nar- row pore diameter distribution centered at ca. 7 nm. The total iron con- 3.1. Drugs in surface waters tent of the bulk material is about 16 wt.%. The concentrations of the investigated DAs, as well as some more typical charac- The temperature of the water samples in the cylindrical vessel was terization data of surface water samples such as pH, conductivity, COD, TOC and the set at ca. 22 °C. The solution was stirred to ensure the complete dis- presence of some metals are depicted in Table 4. All samples showed a weak acid pH persion of the catalyst within the reaction medium. The duration of ranging from 5.25 to 5.95 and fairly similar conductivity, between 1.63 and 2.01 mS/ the treatments was long-stand up to 6 h until the hydrogen peroxide cm. The metal concentrations are within the water quality regulations, most of them was totally consumed. After 6 h, the samples were filtered through being at concentrations below the detection limit of the equipment (ca. 0.5 mg/L). μ However, TOC and COD parameters evidenced higher differences depending on the 0.22 m nylon membranes to remove the heterogeneous catalyst sampling site. prior to analysis. Samples taken at that time were submitted to quan- In these collected fluvial waters 12 out of the 22 drugs analyzed were detected, tification of DAs concentrations as well as some other macroscopic with concentrations ranging from 1.14 ng/L to 40.90 ng/L. In terms of the overall parameters such as total organic carbon (TOC), chemical oxygen drug contamination, RPS1, RPS2 and RPS3 samples were the most polluted with concentrations of ca. 131, 149, and 134 ng/L, respectively. In turn, RPS4 and RPS5 samples demand (COD) and conductivity. The total hydrogen peroxide con- showed lower contents with concentrations of ca. 40 and 20 ng/L, respectively (Table 4). sumption after treatment was confirmed using an iodometric titra- From these preliminary results, it can be concluded that DAs contributed very little to the tion method. TOC and COD levels measured and hence that the waters investigated contained many other dissolved substances. This aspect is important because the organic content is known to play a crucial role in the degradation of drugs by advanced oxidation processes due to competition for the oxidizing radical species (Radjenovic et al., 2009). The largest 2.6. Statistics amount of drugs of RPS2 (149 ng/L) corresponds to the sampling point located downstream of the discharge of “Santamaría de la Bequerencia” STP (Toledo municipality DAs concentrations were analyzed by computing indices of central with an actual served population of 19,000 inhabitants and 20–30% industrial compo- tendency and dispersion (minimum and maximum). The first value nent). The following point with the largest summed concentrations of DAs was RPS3 (134 ng/L), which corresponds to the site downstream of the discharge of the above the limit of quantification was considered the minimum value, Toledo city STP. The third most polluted point was RPS1, with an overall concentration and only values above the limit of quantification were used for median of 131 ng/L, corresponding to Añover del Tajo, the point where the Tagus River enters calculation. Toledo province from the Madrid region.

Table 4 Concentration of the target DAs and metabolites in surface waters (ng/L).

Drugs RPS1 RPS2 RPS3 RPS4 RPS5 Freq (%) Median (min-max)

ALCs Amphetamine n.d. n.d. n.d. n.d. n.d. – n.d. Ephedrine 15.00 14.70 13.40 3.17 1.76 100 13.40 (1.76–14.70) MDMA 2.51 0.63a 1.63 n.d. n.d. 40 2.07 (0.63a–2.51) Methamphetamine n.d. n.d. 3.22 n.d. n.d. 20 3.22 Benzodiazepines Alprazolam n.d. n.d. n.d. n.d. n.d. –– Diazepam 15.10 40.90 32.70 3.96 4.70 100 15.10 (3.96–40.90) Lorazepam 29.00 39.80 21.90 19.80a 10.30a 60 29 (10.30a–39.80) Cannabinoids THC n.d. n.d. n.d. n.d. n.d. –– THC-COOHb 6.51 n.d. n.d. n.d. n.d. 20 6.51 OH-THCb n.d. n.d. n.d. n.d. n.d. –– Cocainics Benzoilecgoninec 40.60 34.10 29.20 5.38 1.83 100 29.20 (1.83–40.6) Cocaethylenec 0.64a 0.29a 1.90a n.d. n.d. 40 – Cocaine n.d. n.d. n.d. n.d. n.d. –– Nor-benzoylecgoninec 9.44 7.39 7.39 1.86 n.d. 80 7.39 (1.86–9.44) LSD and metabolites LSD n.d. n.d. n.d. n.d. n.d. –– nor-LSD/nor-iso-LSDd n.d. n.d. n.d. n.d. n.d. –– O-H-LSDd n.d. n.d. n.d. n.d. n.d. –– Opioids 6ACMe n.d. n.d. n.d. n.d. n.d. –– EDDPf 7.57 8.92 13.80 4.40 1.56 100 7.57 (1.56–13.80) Methadone 3.06 2.09 9.01 1.14 n.d. 80 2.58 (1.14–9.01) Morphine 1.74 n.d. n.d. n.d. n.d. 20 1.74 Heroin n.d. n.d. n.d. n.d. n.d. –– Total concentration of analyzed DAs 131.17 148.82 134.15 39.71 20.15 – 131.17 (20.15–148.82) pH 5.25 5.56 5.43 5.72 5.92 – 5.56 (5.25–5.92) Conductivity (mS/cm) 1.63 1.96 2.01 1.91 1.87 – 1.91 (1.63–2.01)

COD (mg O2/L) 23 19 25 35 43 – 25 (19–43) TOC (mg C/L) 9.6 11.3 13.5 6.8 18.5 – 11.30 (6.8–18.5) Tb (°C) 23.7 26.2 25.6 26.9 27.5 – 26.20 (23.70–27.50) n.d.: not detected. a Not conﬁrmed by second transition (SRM2) or SRM1/SRM2 ratio outside range. b Metabolite of THC. c Metabolite of cocaine. d Metabolite of LSD. e Metabolite of heroin. f Metabolite of methadone. 40 Y. Valcárcel et al. / Environment International 41 (2012) 35–43

The most abundant DAs were the anxiolytics DIAZ (40.9 ng/L) and LOR (39.8 ng/L), the concentration 11.8 ng/L), followed by MA (14% of positive samples, maximum concen- cocaine metabolite BE (40.6 ng/L), the amphetamine-like compound EPH (14.7 ng/L), and tration 0.7 ng/L) and AM (7% of positive samples, maximum concentration 12.1 ng/L). the methadone metabolite EDDP (13.8 ng/L). Except for LOR, these substances were also Fairly similar trends, with MDMA as the most ubiquitous and usually also the most the most ubiquituous compounds, being detected in all analyzed samples (Table 4). These abundant ALC found (concentrations up to 6 ng/L), and MA and AM as less frequent concentrations are comparable to the levels reported in river basins of other European and minor ALCs (most often below method determination limits, concentrations up countries, such as Belgium (Gheorghe et al., 2008), Italy (Zuccato et al., 2008), UK to 3.4 ng/L), have been observed by Vázquez-Roig et al. (Vazquez-Roig et al., 2010) (Kasprzyk-Hordern et al., 2008) and the USA (Bartelt-Hunt et al., 2009)(Table 5). in the Natural Park of L'Albufera in Valencia (Spain) and by Zuccato et al. (Zuccato et Within the group of ALCs, EPH was, as previously mentioned, detected in all the al., 2008) in Italian and British surface waters (EPH was not measured in either sampling sites whereas MDMA and MA were found in two and one sampling sites, re- study). Otherwise, higher concentrations and detection frequencies were reported by spectively, at low concentrations (up to 3.22 ng/L), and AM was not detected at all. The Kasprzyk-Hordern et al. (Kasprzyk-Hordern et al., 2008) for AM in southern Wales median level of EPH was 13.40 ng/L. In a recent study carried out in another Spanish (around 83–87% positive samples, concentrations up to 21 ng/L (other ALCs were not river basin (the Ebro River, (Postigo et al., 2010)), EPH was also detected in 100% of measured)). The different ranges of concentration of these drugs are normally associ- the samples with a median level of 5.4 ng/L but reaching concentrations up to ated to multiple factors, such as the hydrologic, demographic and sociological habits 145 ng/L. The ubiquity and abundance of EPH in surface waters is likely due to its ther- of the population as well as the inﬂuence of the technical operation units of the STPs apeutic use (in addition to the illicit one) in pharmaceutical formulations to treat inﬂu- in the depuration of wastewaters. enza, asthma, hypotension or congestion, its high percentage of excretion without The benzodiazepine anxiolytics DIAZ and LOR, were found in 100% and 60% of the metabolization (between 22% and 99%), and its incomplete elimination in STPs sampling sites, respectively, whereas ALP was not detected. DIAZ and LOR maximum (around 70%) (Postigo et al., 2010). In the aforementioned study, MDMA was also concentrations were very similar and high (40.90 and 39.80 ng/L, respectively) as com- the second (after EPH) most ubiquitous ALC (64% of positive samples, maximum pared to the other target compounds monitored in this study, although in general LOR

Table 5 Concentrations of DAs detected in recent published studies (ng/L).

Compound Spain Italy Belgium UK USA

ALCs Amphetamine 6.8a;NAb;3.38c;NAd b0.65e;b0.65f; NAi,j,k b0.65l;14m; NDp,q,r b0.65g;b0.65h 21n Methamphetamine 0.7a;NAb;NDc;NAd 1.7e; 2.1f; NAi,j,k b0.41 2.3p; 62.6q;14r;ND b0.41g; b0.41h Ephedrine 145a;NAb;NAc;NAd NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r MDMA 11.8a;NAb; 2.48c;NAd 1.7e; 1.1f; NAi,j,k 4NAp,q,r 0.4g; 1.4h Benzodiazepines Alprazolam NAa,b,c,d NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r Lorazepam NAa,b,c,d NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r Diazepam NAa;NAb;NAc; 90/21d NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r Cannabinoids THC NDa;7b; NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r NDc;NAd OH-THC 0.4a;NAb;NAc;NAd NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r THC-COOH 5.5a; 79.5b;NDc;NAd b0.48e; 3.7f; NAi,j,k 1NAp,q,r 0.5g;1h Cocainics Benzoilecgonine 346a;NAb; 78.71c;NAd 183e;50f; 191 17l; 123m; NAp,q,r 5.1g; 37.2h 84n Cocaine 59.2a;NAb; 4.43c;NAd 44e;15f; 26 6l;7m; NAp,q,r 0.8g; 2.9h b0.3n Nor-benzoylecgonine NAa,b,c,d 8.4e; 3.2f; NAi,j,k NDl,m,n NAp,q,r 0.4g; 2.5h Cocaethylene 6.8a;NAb;NAc;NAd 1.3e; 0.2f; NAi,j,k NDl,m,n NAp,q,r b0.07g; 0.1h LSD and metabolites LSD 0.7a;NAb;NAc;NAd NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r Oxo-LSD 0.3a;NAb;NAc;NAd NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r Nor-LSD 0.9a;NAb;NAc;NAd NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r Opioids Morphine 10.8a; 30.9b; 11.7c;NAd 38e; 3.5f; NAi,j,k 42 NAp,q,r b0.05g; 4.7h 6ACM NDa;NAb;NDc;NAd b0.93e; b0.93f; NAi,j,k NDl,m,n NAp,q,r b0.31g; b0.93h Methadone NAa; 18.1b; 0.84c;NAd 8.6e; 3.4f; NAi,j,k NDl,m,n NAp,q,r 0.8g; 10.1h EDDP NAa; 53.6b;NAc;NAd 18e; 9.9f; NAi,j,k NDl,m,n NAp,q,r 1.9g; 6.6h Heroin NDa; 1.5b;NDc;NAd NAe,f,g,h NAi,j,k NAl,m,n NAp,q,r

ND: not detected; NA: not analyzed. a Ebro River (Postigo et al. (2010)). b Llobregat River (Boleda et al. (2009)). c National Park of L'Albufera (Vazquez-Roig et al. (2010)). d Rivers of Madrid Region (Martinez Bueno et al., 2010; Valcarcel et al. (2011)). e Olona River. f Lambro River. g Po River. h Arno River (Zuccato et al. (2008)). i Grote River. j Demer River. k Senette River (Gheorghe et al. (2008)). l Thames River. m Taff River. n Ely River (Kasprzyk-Hordern et al. (2008)). p West fork of the Big Blue River. q Wood River. r Salt Creek River (Bartelt-Hunt et al. (2009)). Y. Valcárcel et al. / Environment International 41 (2012) 35–43 41

was found at higher levels than DIAZ (median concentrations 29.00 ng/L and 15.10 ng/L, efficient than heterogeneous TiO2 photocatalysis. The present work was focused on respectively). Similar concentrations were detected in the main rivers of the nearby region of the evaluation of a heterogeneous photo-Fenton system, in this case for the removal Madrid by (Martinez Bueno et al., 2010)and(Valcarcel et al., 2011) in a previous study. The of a wide variety of drug substances that have been positively detected in surface highest concentration of DIAZ in fluvial water has been reported by (Ternes, 2001), 0.88 μg/L. water samples, such as amphetamine-like compounds (EPH, MDMA, MA), benzodi- This substance has also been detected in the Po and Lambro Rivers in Italy (Calamari et al., azepines(DIAZ,LOR),cannabinoids(THC-COOH,BE,CE,Nor-BE),andcocainics 2003)orinRumania(Moldovan et al., 2007). LOR is extensively metabolized to its glucoro- (EDDP, METH, MOR). nide conjugate, with negligible amounts excreted unchanged (Ghasemi and Niazari, 2005). The efficiency of heterogeneous photo-Fenton treatment for the degradation of these In river waters of Spain, LOR has been detected at concentrations of ca. 40 ng/L (Gros et al., illicit drugs was really noteworthy as compared to those found for conventional activated 2010)andintreatedsewageeffluentsatlevelsupto200ng/L(Coetsier et al., 2009). sludge secondary treatments in STPs (Petrovic et al., 2009). In addition, the influence of Regarding cannabinoids, neither Δ9-tetrahydrocannabinol THC nor OH-THC were the water matrix does not seem to have a crucial role in the efficiency of the hetero- detected. THC-COOH was detected in only one sampling point, at a concentration of geneous photo-Fenton treatment for the removal of DAs, as it has also been reported 6.51 ng/L. THC and OH-THC are metabolized to THC-COOH (Baselt, 2004), which could ex- in other works for the application of AOPs in the degradation of other EPs (Rodriguez- plain why the former are not detected in surface river waters. The presence of cannabi- Gil et al., 2010). noids has also been reported in rivers of Catalonia by Boleda et al. (Boleda et al., 2009) Concerning other macroscopic characterization parameters, the degradation of and Postigo et al. (Postigo et al., 2010) and by (Zuccato et al., 2008)inItaly. DAs to oxygenated by-products of lower molecular weight can be attested to, even As for cocainics, the COC metabolite BE is the most widespread compound, found with partial mineralization to carbon dioxide and water, since COD and TOC concentra- in all sampling points. The largest BE concentration was 40.6 ng/L, corresponding to tions were reduced up to conversions ranging from 37 to 91% and 9 to 71%, respectively RPS1 (entry of Tagus River from Madrid province to Toledo province). The second (Table 6). The highest TOC reductions were exhibited for RPS1 and RPS5 samples, with most abundant cocainic compound was NorBE, with a maximum concentration of TOC reductions around 70%. Unlike the results related to the removal of DAs, the 9.44 ng/L in RPS1 and present in 80% of the sampling sites. In Spain, to the best of our water matrix shows an important influence on the efficiency of the heterogeneous knowledge, this is the first work which analyzed and detected NorBE in surface water. photo-Fenton treatment in terms of the COD and TOC parameters as can be seen in In comparison to other studies, comparatively larger concentrations of COC (up to the different values of COD/TOC of the treated samples. It must also be noted that 59.2 ng/L) and its metabolites BE (up to 346.0 ng/L) and CE (6.8 ng/L) were detected the COD/TOC ratio of as-received water samples (between 1.7 and 5.1, median at in the Ebro River (Spain) by Postigo et al. (Postigo et al., 2010). Also in Spain, in sur- 2.3) decreased after the treatment (from 0.7 to 1.7, median at 1.4), which indicates face waters of the Natural Park of L'Albufera, Vazquez-Roig et al. (Vazquez-Roig et al., the increase of the partial oxidation of the organic matter towards low molecular 2010) reported the presence of COC, BE, and another major COC metabolite, Ecgonine Meth- weight oxygenated by-products. yl Ester (ECGME), at concentrations up to 4.43 ng/L, 78.71 ng/L and 1.35 ng/L, respectively, demonstrating that these substances are ubiquitous in surface waters. Only 3.3. DAs in tap waters and in the Cazalegas drinking water reservoir 1–9% of the consumed COC is eliminated unchanged; the rest is mainly excreted as its major metabolites BE (35–54%) or ECGME (32–49%) (Postigo et al., 2008). More- There are virtually no studies on the occurrence of DAs in drinking waters. Only over, the degradation of COC to these and other transformation products after excre- four studies in Spain have addressed this issue. These studies investigated the pres- tion in the STPs and in the aquatic environment, where their degradation can ence, fate and behavior of various DAs through the drinking water treatment process continue under the natural weather conditions (Castiglioni et al., 2006; Gheorghe in a plant located in the last stretch of the Llobregat River, which serves the Barcelona et al., 2008), explains the relatively low levels of COC often found in surface waters metropolitan area (Boleda et al., 2009; Huerta-Fontela et al., 2011). In these studies, and their decrease with time in favor of other transformation products. the only substances found to be recalcitrant and to persist in the finished water, even The most widespread opioids EDDP (METH metabolite) and METH were found in after considerable treatment, were BE, found at a maximum concentration of 130 ng/L, 100% and 60% of the sampling sites, with median values of 7.57 ng/L and 2.58 ng/L, re- and METH and EDDP, found at concentrations up to 1.7 and 2.9 ng/L, respectively. In spectively. In both cases, the largest concentrations were found in RPS2, which corre- this respect, it may be worth pointing out that this DWTP, due to the considerably bad sponds to the sampling point downstream of the discharge of the second largest STP quality of the surface water used as source, is being subjected to a very comprehensive of the Toledo city municipality. Morphine was only detected in one point (RPS1) at line of treatment that includes prechlorination, coagulation, sand filtration, ozona- very low concentration (1.74 ng/L), and neither heroine nor its metabolite 6- tion, granular activated carbon (GAC) filtration and post-chlorination. acetylmorphine (6ACM) were detected. Boleda et al. (Boleda et al., 2009)alongthe In the present study, the number of compounds detected in the finished drinking Llobregat River basin and Zuccato et al. (2008) reported similar values in Italian sur- water was comparatively larger. Anxiolytics and LSD and its metabolites were not face waters to those presented herein. The occurrence of METH and its main metab- detected in any of the drinking water and reservoir samples. Within the group of olite EDDP in surface waters results from the high use of METH as therapeutic drug ALCs, EPH (not investigated in the aforementioned studies) was found in 2 out of in the substitution treatment for opioids in Spain, their significant excretion rates (5–50% and 3–25%, respectively) (Melis et al., 2011), their poor or non-removal at all in STPs (Postigo et al., 2011a), and their high persistence in the aquatic Table 7 environment. Concentration of the target DAs in drinking waters (ng/L).

Drugs DW1 DW2 DW3 RES

3.2. Evaluation of photo-Fenton treatment for the removal of DAs from surface waters ALCs Amphetamine n.d. n.d. n.d. n.d. Ephedrine 0.12a n.d. 0.90 0.27a The results of the photo-Fenton treatment of surface water samples are summa- MDMA n.d. n.d. 1.51 n.d. rized in Table 6. The levels of all target DAs after 6 h of heterogeneous photo-Fenton Methamphetamine n.d. n.d. n.d. n.d. treatment were below the method detection limits, which demonstrates the potential Benzodiazepines Alprazolam n.d. n.d. n.d. n.d. application of this technology for the removal of these substances. Recently, two solar Diazepam n.d. n.d. n.d. n.d. photocatalytic processes, one of them based on heterogeneous photocatalysis using titani- Lorazepam n.d. n.d. n.d. n.d. um dioxide, and the other one based on a homogeneous photo-Fenton system using iron Cannabinoids THC n.d. 5–53a n.d. n.d. (II) salts in dissolution, have been evaluated by the authors at a pilot-plant scale for the THC-COOHb n.d. n.d. n.d. n.d. removal of COC and METH (Postigo et al., 2011b, 2011c). Preliminary results for both pro- OH-THCb 0.49 n.d. n.d. n.d. cesses revealed for the first time the potential application of photocatalytic processes for Cocainics Benzoilecgoninec n.d. n.d. n.d. 2.47 theremovalofthesesubstances.Theresults of the two studied photocatalytic sys- Cocaethylenec n.d. n.d. n.d. n.d. tems revealed that the homogeneous photo-Fenton treatment was slightly more Cocaine n.d. n.d. n.d. 2.11 Nor-benzoylecgoninec n.d. n.d. n.d. n.d. LSD and metabolites LSD n.d. n.d. n.d. n.d. nor-LSD/nor-iso-LSDd n.d. n.d. n.d. n.d. Table 6 O-H-LSDd n.d. n.d. n.d. n.d. Results of the photo-Fenton treatment applied to surface waters. Opioids 6ACMe n.d. n.d. n.d. n.d. EDDPf n.d. n.d. n.d. 0.34a Samplesa RPS1 RPS2 RPS3 RPS4 RPS5 Median Methadone 0.99 n.d. n.d. 0.47a (min–max) Morphine n.d. n.d. n.d. n.d. TOC conversion (%) 71 34 42 9 69 42 (9–71) Heroin n.d. n.d. n.d. n.d. COD conversion (%) 91 37 68 74 77 74 (37–91) n.d.: not detected. COD/TOC (before photo-Fenton 2.4 1.7 1.8 5.1 2.3 2.3 (1.7–5.1) a Not confirmed by second transition (SRM2) or SRM1/SRM2 ratio outside range. treatment) b Metabolite of THC. COD/TOC (after photo-Fenton 0.7 1.6 1.0 1.4 1.7 1.4 (0.7–1.7) c Metabolite of cocaine. treatment) d Metabolite of LSD. a The levels of all monitored drug substances were below the method determination e Metabolite of heroin. limits. f Metabolite of methadone. 42 Y. Valcárcel et al. / Environment International 41 (2012) 35–43 the 3 tap water samples (DW1 and DW3) and in the Cazalegas reservoir (RES) at (PI 2007/28), the Spanish Ministry of Science and Innovation for sub-ng/L levels (Table 7). Sample DW3 was also positive for MDMA (1.5 ng/L). Both its financial support through the projects TRAGUA (Consolider 2010 EPH and MDMA presented higher concentrations in sample DW3, coming from the Tala- vera de la Reina DWTP, than in the Cazalegas reservoir (RES), which is the reservoir of this CSD2006-44), SCARCE (Consolider-Ingenio 2010 CSD2009-00065) and DWTP. This could be attributed to the analysis of grab water samples and should therefore CEMAGUA (CGL2007-64551/HID). Raquel Feito and Juan Antonio be confirmed with a specifically designed study using composite samples. If confirmed, a Ibañez are acknowledged for their assistance in sampling. Nicola reversion to the parent compound of non-analyzed metabolites or by-products could be Mastroianni acknowledges the CSIC-JAE predoctoral fellowship. Merck happening, similarly to what has been described for some pharmaceuticals (deconjuga- is acknowledged for the gift of LC columns. tion by enzymatic processes in STPs (Calisto and Esteves, 2009)) and also for THC-COOH (Postigo et al., 2010). The cannabinoid THC and its metabolite OH-THC were each detected in only one References sample at 5.53 ng/L and 0.49 ng/L, respectively. Because of their physical–chemical properties (log Kow>5), cannabinoids are expected to partition to solids to a higher BalciogluIA,AlatonIA,OtkerM,BaharR,Bakar N, Ikiz M. Application of advanced extent than other DAs. In fact, this is the argument most extensively used to explain oxidation processes to different industrial wastewaters. J Environ Sci Health A their relatively low detection frequencies and concentrations in raw wastewaters, Tox Hazard Subst Environ Eng 2003;38(8):1587–96. where levels according to consumption would be expected to be higher. Neverthe- Bartelt-Hunt SL, Snow DD, Damon T, Shockley J, Hoagland K. The occurrence of illicit less, in the case of relatively clean waters, such as some natural surface waters, and therapeutic pharmaceuticals in wastewater effluent and surface waters in ground water or drinking water, another possible reason may lie in their method de- Nebraska. Environ Pollut 2009;157(3):786–91. tection limits, which are in most instances larger than for other DAs, and consequent- Baselt RC. Disposition of Toxic Drugs and Chemicals in Man. 7th ed. Foster City, CA: Bio- ly insufficient to detect them. In the abovementioned studies carried out in the medical Publication; 2004. DWTP of Barcelona, cannabinoids were either not detected in the source water Boleda MR, Galceran MT, Ventura F. Trace determination of cannabinoids and opiates (Boleda et al., 2011) or detected (14.7 ng/L of THC-COOH and 12.0 ng/L of THC in in wastewater and surface waters by ultra-performance liquid chromatography- 1 out of 18 samples) and further eliminated after the first steps of prechlorination tandem mass spectrometry. J Chromatogr A 2007;1175(1):38–48. and sand filtration (Boleda et al., 2009). Boleda MA, Galceran MA, Ventura F. Monitoring of opiates, cannabinoids and their fi Within the group of cocainics, only COC and its metabolite BE were detected and metabolites in wastewater, surface water and nished water in Catalonia, Spain. – just in the reservoir sample (RES) at concentrations of 2.11 ng/L and 2.47 ng/L, respec- Water Res 2009;43(4):1126 36. Boleda MA, Galceran MA, Ventura F. Behavior of pharmaceuticals and drugs of abuse in tively. In the studies carried out by Huerta-Fontela et al. (Huerta-Fontela et al., 2008) a drinking water treatment plant (DWTP) using combined conventional and ultra- and Boleda et al. (Boleda et al., 2011), these two compounds were among the most re- filtration and reverse osmosis (UF/RO) treatments. Environ Pollut 2011;159(6): calcitrant ones being either not completely eliminated (BE) or eliminated only near 1584–91. completion of the process after progressive reduction of the concentrations throughout Bones J, Thomas KV, Paull B. Using environmental analytical data to estimate levels of the various treatment steps (COC). community consumption of illicit drugs and abused pharmaceuticals. J Environ As for opioids, methadone was detected in only one of the analyzed tap water Monit 2007;9(7):701–7. samples (DW1) and in the reservoir sample (RES), where it was found together Calamari D, Zuccato E, Castiglioni S, Bagnati R, Fanelli R. Strategic survey of therapeutic with EDDP at concentrations below the determination limits. As previously men- drugs in the rivers Po and Lambro in northern Italy. Environ Sci Technol 2003;37: tioned, these two compounds, detected in the finished drinking water of Barcelona, 1241–8. seem to be quite resistant to purification (91 and 87% elimination after prechlorina- Calisto V, Esteves VI. Psychiatric pharmaceuticals in the environment. Chemosphere tion, coagulation, sand filtration, ozonation, GAC filtration and post-chlorination 2009;77(10):1257–74. (Boleda et al., 2009)), although the last published study reported their complete Castiglioni S, Zuccato E, Crisci E, Chiabrando C, Fanelli R, Bagnati R. Identification and elimination after GAC filtration and also after a newly implemented process of re- measurement of illicit drugs and their metabolites in urban wastewater by liquid – verse osmosis. In the present preliminary study, the occurrence of METH, EDDP, chromatography-tandem mass spectrometry. Anal Chem 2006;78(24):8421 9. Coetsier CM, Spinelli S, Lin L, Roig B, Touraud E. Discharge of pharmaceutical products COC and BE in the reservoir sample (RES) but not in the corresponding tap water (PPs) through a conventional biological sewage treatment plant: MECs vs PECs? (DW3) would indicate that the Talavera de la Reina DWTP effectively eliminates Environ Int 2009;35(5):787–92. them, although the very low concentrations measured do not allow us to reach firm EDADES. Encuesta Domiciliaria sobre Alcohol y Drogas en España (EDADES) 2009/2010, conclusionsinthisrespect. Delegación del Gobierno para el Plan Nacional sobre Drogas; 2010. Disponible en: http://www.mspsi.es. Accesed in Febrary 2011. Ghasemi J, Niazari A. Two and three way chemometrics methods applied for spec- 4. Conclusions trophotometric determination of lorazepam in pharmaceutical formulations and biological fluids. Anal Chim Acta 2005;533:169–77. The results obtained confirm the presence of DAs in the Tagus Gheorghe A, van NA, Pecceu B, Bervoets L, Jorens PG, Blust R, et al. Analysis of cocaine and its principal metabolites in waste and surface water using solid-phase extrac- River along the Toledo province, downstream of the Madrid region. tion and liquid chromatography-ion trap tandem mass spectrometry. Anal Bioa- The anxiolytics DIAZ and LOR, the cocaine metabolites BE and nal Chem 2008;391(4):1309–19. NorBE, the ALCs EPH and MDMA, the opioid METH and its metabo- Gonzalez-Marino I, Rodriguez I, Quintana JB, Cela R. Matrix solid-phase dispersion followed by gas chromatography–mass spectrometry for the determination of lite EDDP, and the THC metabolite THC-COOH were all found in flu- triclosan and methyl triclosan in sludge and sediments. Anal Bioanal Chem vial waters in concentrations ranging from 1.56 ng/L to 40.9 ng/L. 2010;398(5):2289–97. The highest concentrations of these substances were detected in Gros M, Petrovic M, Ginebreda A, Barcelo D. Removal of pharmaceuticals during wastewater treatment and environmental risk assessment using hazard indexes. Environ the RPS2 sample, collected downstream of the discharge of the Int 2010;36(1):15–26. Toledo city STP. Most of these substances were not found in tap Huerta-Fontela M, Galceran MT, Ventura F. Ultraperformance liquid chromatography- water, although some of them were detected at very low concen- tandem mass spectrometry analysis of stimulatory drugs of abuse in wastewater and surface waters. Anal Chem 2007;79(10):3821–9. trations, such as METH (0.99 ng/L) in Toledo city tap water (DW1) and Huerta-Fontela M, Galceran MT, Ventura F. Stimulatory drugs of abuse in surface EPH (0.90 ng/L) in Talavera de la Reina tap water (DW3). Despite the waters and their removal in a conventional drinking water treatment plant. low concentrations of these pollutants, the risk to human health and Environ Sci Technol 2008;42(18):6809–16. Huerta-Fontela M, Galceran MT, Ventura F. Occurrence and removal of pharmaceuti- the environment is nowadays under consideration and further cals and hormones through drinking water treatment. Water Res 2011;45(3): studies on their occurrence, ecotoxicology and effects on public 1432–42. health (especially on vulnerable populations) should be carried Hummel D, Loffler D, Fink G, Ternes TA. Simultaneous determination of psychoactive out. The application of heterogeneous photo-Fenton treatment as drugs and their metabolites in aqueous matrices by liquid chromatography mass spectrometry. Environ Sci Technol 2006;40(23):7321–8. advanced oxidation process has shown a remarkable effectiveness INEbase. National Statistics Institute on-line database; 2009. Accessed in February for the degradation of these substances, making this technique a 2011. feasible technology for implementation in drinking water treatment Jones OA, Lester JN, Voulvoulis N. Pharmaceuticals: a threat to drinking water? Trends Biotechnol 2005;23(4):163–7. plants. Kasprzyk-Hordern B, Dinsdale RM, Guwy AJ. The occurrence of pharmaceuticals, personal care products, endocrine disruptors and illicit drugs in surface water in South Wales, UK. Water Res 2008;42(13):3498–518. Acknowledgments Lazar K, Calleja G, Melero JA, Martínez F, Molina R. Influence of synthesis routes on the state of iron species in SBA-15 mesoporous materials. Stud Surf Sci Catal 2004;154: 805–12. The authors would like to thank the Foundation for Health Stud- Martinez Bueno M, Hernando M, Herrera S, Gómez M, Fernández-Alba A, Bustamante ies in Castilla-la Mancha (FISCAM) for the founding of the project I, et al. Pilot survey of chemical contaminants from industrial and human Y. Valcárcel et al. / Environment International 41 (2012) 35–43 43

activities in river waters in Spain. Int J Environ Anal Chem 2010;90(3): irradiation: kinetics, characterization of major intermediate products and toxicity 321–43. evaluation. Water Res 2011c;45(16):4816–26. Melis M, Castiglioni S, Zuccato E. Metabolism and excretion of illicit drugs in humans. Radjenovic J, Jelic A, Petrovic M, Barceló D. Determination of pharmaceuticals in sew- In: Castiglioni S, Zuccato E, Fanelli R, editors. Illicit Drugs in the Environment: age sludge by pressurized liquid extraction (PLE) coupled to Liquid Chromatogra- Ocurrence, Analysis, and Fate using Mass Spectrometry. Hobokenm New Yersey: phy-tandem Mass Spectrometry (LC-MS/MS). Anal Bioanal Chem 2009;393(6–7): John Wiley & Sons, Inc; 2011. p. 31–52. 1685–95. Moldovan Z, Schmutzer G, Tusa F, Calin R, Alder AC. An overview of pharmaceuticals Rodriguez-Gil JL, Catala M, González Alonso S, Romo Maroto R, Valcarcel Y, Segura Y, and personal care products contamination along the river Somes watershed, et al. Heterogeneous photo-Fenton treatment for the reduction of pharmaceutical Romania. J Environ Monit 2007;9(9):986–93. contamination in Madrid rivers and ecotoxicological evaluation by a miniatur- Petrovic M, de Alda MJ, az-Cruz S, Postigo C, Radjenovic J, Gros M, et al. Fate and removal of ized fern spores bioassay. Chemosphere 2010;80(4):381–8. pharmaceuticals and illicit drugs in conventional and membrane bioreactor Ternes T. Analytical metholds for the determination of pharmaceuticals in aqueous wastewater treatment plants and by riverbank filtration. Philos Transact A environmental samples. TrAC Trends Anal Chem 2001;20(8):419–34. Math Phys Eng Sci 2009;367(1904):3979–4003. Ternes TA, Stuber J, Herrmann N, McDowell D, Ried A, Kampmann M, et al. Ozonation: a Pomati F, Castiglioni S, Zuccato E, Fanelli R, Vigetti D, Rossetti C, et al. Effects of a com- tool for removal of pharmaceuticals, contrast media and musk fragrances from plex mixture of therapeutic drugs at environmental levels on human embryonic wastewater? Water Res 2003;37(8):1976–82. cells. Environ Sci Technol 2006;40(7):2442–7. Terzic S, Senta I, Ahel M. Illicit drugs in wastewater of the city of Zagreb (Croatia)— Postigo C, de Alda MJ, Barcelo D. Fully automated determination in the low nanogram estimation of drug abuse in a transition country. Environ Pollut 2010;158(8): per liter level of different classes of drugs of abuse in sewage water by on-line 2686–93. solid-phase extraction-liquid chromatography-electrospray-tandem mass spec- UNODC (United Nations Office on Drugs and Crime). World Drug Report 2008; 2008. trometry. Anal Chem 2008;80(9):3123–34. Accessed in February 201. Postigo C, Lopez de Alda MJ, Barcelo D. Drugs of abuse and their metabolites in the Ebro Valcarcel Y, Gonzalez AS, Rodriguez-Gil JL, Gil A, Catala M. Detection of pharmaceuti- River basin: occurrence in sewage and surface water, sewage treatment plants cally active compounds in the rivers and tap water of the Madrid Region (Spain) removal efficiency, and collective drug usage estimation. Environ Int 2010;36(1): and potential ecotoxicological risk. Chemosphere 2011;84(10):1336–48. 75–84. van Nuijs AL, Tarcomnicu I, Simons W, Bervoets L, Blust R, Jorens PG, et al. Optimization Postigo C, Lopez de Alda MJ, Barcelo D. Ocurrence of illicit drugs in wastewater in Spain. and validation of a hydrophilic interaction liquid chromatography-tandem mass In: Castiglioni S, Zuccato E, Fanelli R, editors. Illicit Drugs in the Environment: spectrometry method for the determination of 13 top-prescribed pharmaceuticals Ocurrence, Analysis, and Fate using Mass Spectrometry. Hobokenm New Yersey: in influent wastewater. Anal Bioanal Chem 2010;398(5):2211–22. John Wiley & Sons, Inc; 2011a. p. 117–36. Vazquez-Roig P, Andreu V, Blasco C, Pico Y. SPE and LC-MS/MS determination of 14 Postigo C, Sirtori C, Oller I, Malato S, Maldonado MI, Lopez de Alda M, et al. Solar illicit drugs in surface waters from the Natural Park of L'Albufera (Valencia, degradation and photocatalytic treatment of cocaine: kinetics, characterization Spain). Anal Bioanal Chem 2010;397(7):2851–64. of major intermediate products and toxicity evaluation. Appl CatalB Environ Zuccato E, Castiglioni S, Bagnati R, Chiabrando C, Grassi P, Fanelli R. Illicit drugs, a novel 2011b;104(1–2):37–48. group of environmental contaminants. Water Res 2008;42(4–5):961–8. Postigo C, Sirtori C, Oller I, Malato S, Maldonado MI, Lopez de Alda M, et al. Photolytic and photocatalytic transformation of methadone in aqueous solutions under solar