Nicotine Derivatives in Wastewater and Surface Waters - American

Aug 2, 2008 - Typical concentrations of cotinine and 3′- hydroxycotinine were ∼1-10 μg/L in untreated wastewater, but clearly less in treated was...
0 downloads 0 Views 177KB Size
Download PDF
Environ. Sci. Technol. 2008, 42, 6354–6360

Nicotine Derivatives in Wastewater and Surface Waters: Application as Chemical Markers for Domestic Wastewater IGNAZ J. BUERGE,* MAREN KAHLE, HANS-RUDOLF BUSER, MARKUS D. MÜLLER, AND THOMAS POIGER Plant Protection Chemistry, Swiss Federal Research Station (Agroscope), CH-8820 Wa¨denswil, Switzerland

Received February 14, 2008. Revised manuscript received June 16, 2008. Accepted July 02, 2008.

Nicotine is extensively metabolized in the human body to a number of compounds, which may enter natural waters via discharge of domestic wastewater. However, little is known on exposure of and potential effects on the aquatic environment. In this study, two major urinary metabolites, cotinine and 3′hydroxycotinine, as well as a further tobacco alkaloid, N-formylnornicotine, were measured in wastewater and water from Swiss lakes using an analytical procedure based on SPE and LC-MS/MS SRM with cotinine-d3 as internal standard (LOQs, 1.0-1.5 ng/L). Typical concentrations of cotinine and 3′hydroxycotinine were ∼1-10 µg/L in untreated wastewater, but clearly less in treated wastewater (∼0.01-0.6 µg/L), corresponding to elimination efficiencies of 90-99%. N-Formylnornicotine,however,wasfoundatsimilarconcentrations in untreated and treated wastewater (0.02-0.15 µg/L). Its apparent persistence during wastewater treatment was further confirmed by incubation experiments with activated sludge. In lakes, cotinine, 3′-hydroxycotinine, and N-formylnornicotine were detected at concentrations up to 15, 80, and 6 ng/L, respectively. Concentrations in lakes correlated with the expected anthropogenic burden by domestic wastewater (ratio population per water throughflow), demonstrating the suitability of these nicotine derivatives as hydrophilic, anthropogenic markers. In small receiving waters with significant wastewater discharges, concentrations of a few hundred ng/L may be expected. Possible ecotoxicological risks associated with such environmental concentrations, can, however, not be assessed at present as data on effects on aquatic organisms are very limited, in particular on long-term effects.

consumption of nicotine can be estimated at ∼5 × 104 tons. Nicotine would thus rank as a high production volume chemical if it were a synthetic compound. The alkaloid is toxic by inhalation, ingestion, and skin contact. The lethal oral dose for man is stated to be ∼40-60 mg (4). In the liver, nicotine is extensively metabolized by oxidative enzymatic transformations to cotinine, trans-3′hydroxycotinine, and further metabolites (Figure 1 and ref 3). These metabolites may eventually enter natural waters via discharge of domestic wastewater, but surprisingly little is known on exposure of and potential effects on the aquatic environment. Some data are available for cotinine, which has been detected in U.S. and Canadian surface waters at concentrations in the low ng/L range (5–11). Analyses of untreated and treated wastewater in the U.S. and in Spain indicated a moderate elimination of cotinine during wastewater treatment of ∼50% (concentrations in the low µg/L range 12, 13). For other nicotine derivatives, however, no data were found on occurrence and fate in the environment. Ecotoxicological studies are, as for many other environmental chemicals, even less published (14–16). Nicotine is toxic to insects. In the past, the compound was widely used as an insecticide against aphids, thrips, whiteflies, and other sucking insects on a range of crops, including fruits, vines, vegetables, and ornamentals (4). Nicotine derivatives may, therefore, be harmful particularly to arthropods. In this study, the two major human urinary metabolites, cotinine and 3′-hydroxycotinine, as well as a further minor tobacco alkaloid, N-formylnornicotine, were measured in samples from Swiss wastewater treatment plants (WWTPs) and various surface waters using an analytical procedure based on solid-phase extraction (SPE) and liquid chromatography-tandem mass spectrometry (LC-MS/MS). Similar analytical methods have been described previously in the literature (5, 12, 13, 17, 18). The biodegradability of the compounds was investigated with activated sludge incubation experiments as well as with analyses of 24-h flowproportional composite samples of untreated and treated wastewater. A comparison of per-capita loads determined in WWTP effluents and in the outflows of lakes allowed a further assessment of their elimination/persistence in surface waters, and even suggested additional inputs other than with treated wastewater. Due to the widespread consumption of tobacco, cotinine has been suggested as a potential chemical marker for domestic wastewater contamination of surface waters (6, 9, 12, 19). The compound is already used as an indicator of tobacco smoke exposure (20). The suitability of cotinine, 3′-hydroxycotinine, and N-formylnornicotine as quantitative

Introduction Nicotine is a well-investigated, highly toxic alkaloid found in tobacco and other nightshade plants (1). As a constituent of tobacco smoke, nicotine is regularly consumed by a large fraction of the world’s population. Moreover, humans are also exposed unintentionally to nicotine by passive smoking. Approximately 5.5 trillion cigarettes are produced globally each year by the tobacco industry (2). Considering a typical nicotine content of ∼8 mg/cigarette (3), the annual global * Corresponding author phone: ++41 44 783 63 83; fax: ++41 44 780 63 41; e-mail: [email protected]. 6354

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 17, 2008

FIGURE 1. Structures of nicotine and its urinary metabolites (simplified metabolic pathway) and structure of the tobacco alkaloid N-formylnornicotine. 10.1021/es800455q CCC: $40.75

 2008 American Chemical Society

Published on Web 08/02/2008

markers for domestic wastewater was therefore assessed in this study by analyses of water samples from several lakes with widely differing anthropogenic burdens. Nicotine itself was not investigated as this compound may enter surface waters not only via discharge of domestic wastewater, but also directly such as with cigarette stubs or from tobacco cultivation.

Experimental Section Chemicals. (-)-Nicotine ((S)-3-(1-methyl-2-pyrrolidinyl)pyridine; 1.0 mg/mL in methanol; purity, 99%) was obtained from Sigma, St. Louis, MO; (-)-cotinine ((S)-1-methyl-5-(3pyridyl)-2-pyrrolidinone, 98%) was obtained from SigmaAldrich,Steinheim,Germany;trans-3′-hydroxycotinine((3′R,5′S)3′-hydroxycotinine, 98%) and N-formylnornicotine (Nformyl-2-(3-pyridyl)pyrrolidine, 98%) were purchased from Toronto Research Chemicals, North York, Canada; and cotinine-d3 (1.0 mg/mL in methanol, 99%), used as internal standard, was from Isotec, Miamisburg, OH. Ibuprofen (2(4-isobutylphenyl)propionic acid, 99%), a reference compound in activated sludge incubation experiments, was obtained from Aldrich, Milwaukee, WI. Water used as solvent after SPE (see below) and as mobile phase in LC-MS was from Carl Roth, Karlsruhe, Germany (LC-MS grade). Ammonium acetate (99.0%) and ammonium carbonate (30-33% as NH3) were purchased from Fluka, Buchs, Switzerland. Water Samples. Wastewater samples were obtained from 10 municipal WWTPs in the region of Zu ¨ rich, Switzerland, serving populations of between 10,000 and 370,000 (Table 1). These installations operate with four stages: mechanical, biological (activated sludge, mostly with nitrification and partially with denitrification), and chemical treatment (phosphate precipitation by iron salts, no chlorination), and subsequent sand filtration. Influent and effluent wastewater were collected flow-proportionally during 24 h. Generally, influent samples were taken after the primary sedimentation basin (except in WWTPs Zu ¨ rich and Ma¨nnedorf, where raw wastewater was collected), and effluent samples were taken after sand filtration. Samples were stored in polyethylene bottles at ∼4 °C and extracted within the same day. Surface water was collected from 8 lakes located in the Swiss Midland region and from a remote Alpine mountain lake (Table 2). Grab samples were taken at the outflow of the lakes at 0-1 m depth. In winter, when the lakes are well mixed, these samples can be considered representative for the whole water body (21). “Fossil” groundwater (former pumping station Aqui, Zu ¨rich (22), was used for blank samples to study potential contamination from reagents and the experimental procedure as well as cross-contamination between samples. This groundwater, with an age of several thousand years, is not expected to be contaminated with anthropogenic compounds. The groundwater was also used for recovery experiments. Water samples were filled into glass bottles, stored at ∼4 °C, and extracted within one week. Solid-Phase Extraction. Extraction of nicotine derivatives was done with reusable glass columns containing ∼10 mL of a macroporous polystyrene divinylbenzene adsorbent (BioBeads SM-2, 20-50 mesh, Bio-Rad Laboratories, Hercules, CA). This adsorbent has previously been used for other medium polar compounds such as caffeine, cyclophosphamide, and ifosfamide (21, 23). Separate adsorbent columns were used for surface water and treated and untreated wastewater samples in order to minimize potential crosscontamination. Columns were reused up to 20 times. Prior to extraction, the columns were properly cleaned using one column volume of dichloromethane, methanol, and fossil groundwater as solvents. Thereafter, volumes of 1.0 L lake water or 0.5 L wastewater were fortified with internal standard, cotinine-d3 (∼0.1 ng/µL), to spike levels of ∼10 and 20 ng/L, respectively, and were then passed through the

SPE columns at ∼10 mL/min. The analytes were recovered with 5 mL of methanol, which also removed residual water from the SPE material, and then with 10 mL of dichloromethane. The combined eluates were shaken vigorously and the phases were allowed to separate. The dichloromethane phase was transferred into a glass vial. Additional volumes of 10 and 5 mL of dichloromethane were passed through the column, partitioned with the methanolic phase, and transferred to the same vial. After complete evaporation of dichloromethane at room temperature under a gentle draft of air, the residues were taken up in 5 mL of water. Extracts from untreated wastewater contained some suspended material and were therefore filtered (0.45 µm Chromafil RC45/25, Macherey-Nagel, Du ¨ ren, Germany). Online SPE-Liquid Chromatography-Tandem Mass Spectrometry. Nicotine derivatives were analyzed with LC-MS/MS (Agilent 1100 Series HPLC (binary pump, microvacuum degasser), Palo Alto, CA; API 4000 triple quadrupole mass spectrometer, Applied Biosystems, Foster City, CA). Further enrichment of the analytes as well as a cleanup of the samples (removal of highly hydrophilic compounds) was achieved by online coupling of a SPE cartridge (cartridge precolumn packed with Gemini C18 4 × 3 mm, 5 µm particle size, Phenomenex, Torrance, CA) to the analytical HPLC column (Gemini C18 2 × 150 mm, 5 µm) using a column switching technique with a dual 6-port valve system (HTS PAL autosampler, CTC Analytics, Zwingen, Switzerland). First, sample volumes of 0.5 mL were injected into a 2 mL PEEK loop. The valves were then switched so that the sample was transferred to the SPE cartridge during 90 s with 1 mM ammonium acetate buffer as mobile phase (flow rate, 0.5 mL/min, delivered by an auxiliary HPLC pump, Jasco PU980, Gross-Umstadt, Germany). After transfer was complete, the valves were switched to allow elution of the enriched compounds in the opposite direction to the analytical column. The LC conditions were as follows: linear gradient from 95% 1 mM ammonium acetate/5% methanol to 100% methanol within 20 min, followed by an isocratic phase of 5 min; flow rate, 0.2 mL/min. The mass spectrometer was equipped with a turbo ion spray source, operated in positive mode (ion spray voltage, 3 kV, 400 °C) and, for trace analysis, selected reaction monitoring (SRM) with the following ion transitions: nicotine, m/z 163f132 with a collision energy of 22 eV (and for confirmatory purposes: m/z 163f84, 28 eV); cotinine, 177f80, 36 eV (177f98, 30 eV); 3′-hydroxycotinine, 193f134, 26 eV (193f80, 42 eV); N-formylnornicotine, 177f159, 25.5 eV (177f80, 36 eV); and cotinine-d3, 180f80, 36 eV. An important fragmentation reaction under the selected conditions was the cleavage of the pyridine-pyrrolidine/pyrrolidone bond leading to the successor ion m/z 80 for all compounds and m/z 84 for nicotine, m/z 98 for cotinine and N-formylnornicotine. The primary ion transitions for nicotine, 3′-hydroxycotinine, and N-formylnornicotine resulted from loss of 31 (-CH3NH2), 59 (-C2H3O2), and 18 (-H2O), respectively. Loss of H2O is usually not a very selective MS/ MS transition, but we found clearly less interference with the slightly earlier eluting cotinine having three ion transitions in common with N-formylnornicotine (see Figure 2 and discussion later). Concentrations were determined from peak area ratios relative to the internal standard and in reference to suitable standard solutions (not extracted with SPE, corresponding to 0-5, 0-10, and 0.0-0.2 µg/L cotinine, 3′-hydroxycotinine, and N-formylnornicotine, respectively, in samples). Precision, Matrix Effects, Recoveries, and LOQs. The precision of the analytical method was determined with four replicate SPE extractions and LC-MS/MS analyses of two representative samples (untreated wastewater and surface water). Relative standard deviations were 4-5, 8-10, and VOL. 42, NO. 17, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6355

10-16% for cotinine, N-formylnornicotine, and 3′-hydroxycotinine, respectively. Matrix effects were investigated in SPE extracts from untreated wastewater and fossil groundwater. The absolute ion suppression of the internal standard, cotinine-d3, added in these experiments after SPE to the final extracts, was ∼75% in untreated wastewater (compared with standards in LC-MS grade water), whereas no suppression was found in extracts from fossil groundwater. Relative ion suppression was studied by standard additions to SPE extracts. The corresponding slopes (peak area ratio of analyte/internal standard vs concentration ratio analyte/internal standard) obtained in SPE extracts were compared to those in standards. In untreated wastewater, the ion suppression relative to the internal standard was ∼90% for cotinine, ∼62% for Nformylnornicotine, and ∼136% for 3′-hydroxycotinine. In fossil groundwater, the relative ion suppression was small for all compounds (90-99%). Fortified quality control samples showed acceptable recoveries of 92-96% for cotinine (spike levels, 5-1000 ng/ L) and 85-116% for N-formylnornicotine (5-100 ng/L). However, recoveries of the more polar 3′-hydroxycotinine were not satisfactory, probably due to insufficient retention on the SPE adsorbent. After extraction of 500 mL of water with concentrations of 100-1000 ng/L (procedure for wastewater), only 44-45% could be recovered. After extraction of 1000 mL with 5 ng/L (procedure for lake water), the recovery was only 25%. Concentrations for 3′-hydroxycotinine were, therefore, corrected for recoveries, but nevertheless have to be considered as semiquantitative. Blank samples extracted in series with surface water samples contained traces of 0.3 ( 0.2 ng/L cotinine and 0.2 ( 0.1 ng/L N-formylnornicotine (mean ( SD). Corresponding limits of quantification (LOQ) were ∼1.5 and 1.0 ng/L, respectively (defined as meanblank + 6 SDblank). For 3′hydroxycotinine, which was not detected in blank samples, the LOQ was determined from a signal-to-noise ratio of 6:1 (∼1 ng/L or ∼5 ng/L when corrected for low recoveries after SPE, see above). Corresponding LOQs in untreated and treated wastewater, calculated in the same way, were ∼16 and 7 ng/L cotinine, ∼9 and 6 ng/L N-formylnornicotine, and ∼50 and 20 ng/L 3′-hydroxycotinine, respectively. Nicotine was not analyzed in lake water and (treated) wastewater due to blank problems (possibly from smoking laboratory personnel, not involved in the study). The compound was measured only in activated sludge incubation experiments at high ng/L concentrations. Activated Sludge Incubations. The biodegradability of nicotine, cotinine, 3′-hydroxycotinine, and N-formylnornicotine in activated sludge was investigated under laboratory conditions. For that, 0.8 L of untreated wastewater from the primary sedimentation basin of WWTP Wa¨denswil was mixed with 0.8 L of return-sludge, thus in a ratio similar to that under typical operating conditions. The suspension was stirred at 15 °C in the dark and aerated with water-saturated, compressed air through a glass frit. Two incubation experiments were performed. One suspension was fortified with ∼1 µg/L of nicotine to study the potential formation of N-formylnornicotine, the other was fortified with ∼1 µg/L of N-formylnornicotine to follow its biodegradability. To both suspensions, ∼10 µg/L of ibuprofen was added at 0 and 8 h of incubation as a marker for the biological activity of the sludge under laboratory conditions. Ibuprofen is readily biodegradable in activated sludge (24). Periodically, 5 mL samples were removed, fortified with internal standard, filtered (0.45 µm Chromafil RC-45/25), and analyzed by online SPE-LC-MS/MS as described above, except that 1 mM ammonium carbonate was used as mobile phase. The higher pH buffer was required for a sufficient chromatographic 6356

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 17, 2008

retention of nicotine (weak base with a pKa of 8.2 (4)). The analysis of ibuprofen is described elsewhere (25).

Results and Discussion High Elimination of Cotinine and 3′-Hydroxycotinine in WWTPs. The two major urinary nicotine metabolites cotinine and 3′-hydroxycotinine were detected in all WWTPs investigated. Concentrations in untreated wastewater ranged from 0.8 to 2.9 and 1.1 to 9.5 µg/L, respectively (24-h flowproportional composite samples, Table 1, typical chromatograms in Figure 2a). Corresponding loads, normalized for the population discharging wastewater to the respective plants, showed less variation (0.88 ( 0.29 and 2.10 ( 1.02 mg person-1d-1, respectively, mean ( SD, n ) 17), but were consistently higher in WWTP Zu ¨ rich (cotinine, 0.99-1.23 mg person-1d-1) than in other installations (0.39-0.94 mg person-1d-1). Presumably, these higher per-capita loads can be explained by the large number of commuters that work in the city of Zu ¨ rich. In fact, daily analyses in WWTP Zu ¨ rich between November 13 and 20, 2007 showed somewhat higher loads on weekdays (1.11-1.23 mg person-1d-1) than during the weekend (0.99-1.04 mg person-1d-1). Typical concentrations of cotinine in a smoker’s urine amount to ∼1.6 mg/L (data from Germany, n ) 1600 (26)). Assuming a mean urine production of ∼1.5 L/d, the daily cotinine excretion of a smoker can be estimated at ∼2.4 mg person-1d-1. This value, compared to the mean per-capita load in wastewater, suggests a percentage of ∼37% smokers in the population of the canton of Zu ¨ rich or less, if further inputs of cotinine to the sewer system are of importance (e.g., cigarette stubs). Anyhow, this estimate is rather close to statistical data for Switzerland (29% smokers of the 14-65 years old population, 2006 (27)). In treated wastewater, concentrations of cotinine and 3′hydroxycotinine were considerably lower (0.01-0.11 and 0.06-0.61 µg/L, respectively, Table 1). Consequently, elimination of these compounds in WWTPs was high, on average 98 ( 2 and 93 ( 6%, respectively. Typical chromatograms are shown in Figure 2b. Identification of N-Formylnornicotine in Wastewater. SRM chromatograms of the ion transition for cotinine (m/z 177f80) consistently showed a second peak at a slightly higher retention time (∆tR ≈ 0.5 min, Figure 2) that was also present in two qualifier ion transitions (m/z 177f98 and 177f159). The peak area ratio of the ion transitions was, however, different from that of cotinine, suggesting an isomeric, with cotinine related compound present in these samples. After consultation of a database of nicotine derivatives (Toronto Research Chemicals), the unknown compound was eventually identified as N-formylnornicotine (structure, see Figure 1), based on its retention time and the ratio of ion transitions in comparison to a reference standard. N-formylnornicotine is a minor alkaloid present in tobacco in amounts of ∼0.04-0.21 mg/g dry weight (for comparison, tobacco contains ∼10-21 mg/g nicotine, ∼0.4-0.9 mg/g anatabine, ∼0.2-0.6 mg/g nornicotine, ∼0.04-0.20 mg/g cotinine, and further minor alkaloids (28)). In untreated wastewater, concentrations of N-formylnornicotine were about 2-3 orders of magnitude lower (0.02-0.15 µg/L, Table 1) than those of the two major urinary metabolites of nicotine. The typical concentration ratio cotinine/3′-hydroxycotinine/ N-formylnornicotine was ∼50:120:1. The somewhat higher concentration of N-formylnornicotine in WWTP Ma¨nnedorf may be due to a contamination such as from smoking personnel. In treated wastewater, the concentration ratio was then clearly changed (∼1:8:1), in favor of N-formylnornicotine (compare Figure 2a with 2b) that was measured at concentrations of 0.02-0.07 µg/L, hence, similar to that in untreated wastewater (Table 1). Mean influent and effluent per-capita

VOL. 42, NO. 17, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6357

Feb 08, 2007 May 22, 2007 May 30, 2007 May 30, 2007 May 30, 2007 Feb 06, 2007 Feb 08, 2007 Feb 06, 2007 Feb 08, 2007 May 22, 2007 Nov 13-20, 2007

sampling date 11800 21800 17900 10300 23000 20000 36000 19500 25500 369900

population served 6240 10900 9000 5170 11980a no data 27150 10500 12080 172010 134510-308020

18.6 5.5 7.0 11.8 3.9 5.4 6.3 no data 11.3 10.4

sludge age [d]b

0.88 ( 0.29

0.92 0.78 0.94 1.52 1.03 1.23 1.24 1.29 1.45 2.65 1.48-2.88

influent [µg/L]

effluent [µg/L]

0.018 ( 0.015 98 ( 2

0.010 0.075 0.024 0.032 0.045 0.046 0.042 0.112 0.038 0.023 0.016-0.043

cotinine

2.10 ( 1.02

2.37 1.85 2.05 2.30 1.90 4.06 3.36 1.07 3.98 4.95 3.00-9.53

influent [µg/L]c

0.16 ( 0.11 93 ( 6

0.19 0.24 0.06 0.11 0.14 0.14 0.18 0.59 0.24 0.25 0.32-0.61

effluent [µg/L] c

3′-hydroxycotinine

0.018 ( 0.016

0.020 0.048 0.016 0.149 0.018 0.028 0.022 0.031 0.028 0.062 0.020-0.042

influent [µg/L]

0.020 ( 0.009 (-13 ( 113)

0.028 0.049 0.015 0.022 0.033 0.028 0.025 0.044 0.029 0.070 0.034-0.065

effluent [µg/L]

N-formylnornicotine

Sep 17, 07 Jan 23, 07 Jan 31, 07 Jan 15, 07 Jan 31, 07 Jan 31, 07 Jan 31, 07 Jan 23, 07 Feb 09, 07 Jan 23, 07 Feb 09, 07 1.1

112900 c

b

no data 1.4 3.4 1.4 15 3.8 5.6 2.6

0 20700 b 106200 b 356100 b 9300 b 19000 b 11500 b 9600 c

mean water residence time, τ [years]d

4.1 (2006f)

no data 48 (2006e) 103 (2004-2006e) 82 (2006e) 1.3 (1978-2005e) 2.4 (d) 1.3 (1978-2005 e) 0.7 (d)

mean water throughflow (reference years), Q [m3 s-1]

0.321

0.000 0.005 0.012 0.050 0.084 0.093 0.103 0.168

population per water throughflow, P/Q [persons m-3 d]

0.040 ( 0.011

2.6 2.6 3.0 6.5 6.9 8.4 8.9 13.5 13.5 14.8 15.4

cotinine [ng/L]

c

0.20 ( 0.05

(3) h (4) h 7 14 12 17 26 41 49 77 51

3′-hydroxycotinine [ng/L]g

0.018 ( 0.003

(0.6) h (0.5) h (0.6) h 2.2 2.2 2.0 2.7 4.4 4.2 6.2 6.1

N-formylnornicotine [ng/L]

Map with sampling sites, see Figure 1 in ref 21. Federal Office for the Environment, Berne, Switzerland, 2005/2006, personal communication, P. Scha¨r. Office for Waste, Water, Energy, and Air of the Canton of Zurich, 2005/2006, personal communication, U. Holliger. d Ref 36. e Federal Office for the Environment, Berne, Switzerland, www.hydrodaten.admin.ch. f Office for Waste, Water, Energy, and Air of the Canton of Zurich, www.hochwasser.zh.ch. g Corrected for recoveries, semiquantitative data (see Experimental Section). h Below limit of quantification. i Determined from slope of linear correlations shown in Figure 3.

a

per-capita export load [mg person-1d-1]i

Greifensee

Guraletschsee Walensee Vierwaldsta¨tter Zu¨richsee Sempachersee Hallwilersee Baldeggersee Pfa¨ffikersee

lakea

sampling date

population in catchment area, P [persons]

TABLE 2. Concentrations of Cotinine, 3′-Hydroxycotinine, and N-Formylnornicotine in Swiss Lakes

a Influent; 10236 m3 d-1 in effluent. b Sludge retention times, mean estimates from Office for Waste, Water, Energy, and Air of the Canton of Zurich, personal communication, U. Holliger. c Corrected for recoveries, semiquantitative data (see Experimental Section).

per-capita load [mg person-1d-1], mean ( SD, n ) 17 elimination [%], mean ( SD

Gossau Horgen Ku¨snacht Ma¨nnedorf Meilen Thalwil Uster Wa¨denswil Wetzikon Zu¨rich

WWTP

wastewater throughput [m3 d-1]

TABLE 1. Concentrations and Loads of Cotinine, 3′-Hydroxycotinine, and N-Formylnornicotine in WWTPs, Canton of Zu1 rich, Switzerland

FIGURE 2. LC-MS/MS chromatograms acquired in SRM mode showing the elution of 3′-hydroxycotinine, cotinine, and N-formylnornicotine in samples of untreated (a) and treated (b) wastewater (WWTP Wa¨denswil), and from a Swiss lake (Vierwaldsta¨ttersee, c). SRM transitions, see Experimental Section. Signal intensities are normalized to 100%.

FIGURE 3. Concentrations of cotinine, 3′-hydroxycotinine, and N-formylnornicotine in Swiss lakes in winter 2007 plotted vs population in the catchment area per throughflow of water. loads indicated a high stability of N-formylnornicotine during wastewater treatment (0.018 ( 0.016 and 0.020 ( 0.009 mg person-1d-1, respectively). High Stability of N-Formylnornicotine in Activated Sludge. Degradation of nicotine, cotinine, 3′-hydroxycotinine, and N-formylnornicotine was further studied with laboratory incubation experiments using untreated wastewater and activated sludge from a municipal WWTP (for details, see Experimental Section). The first three compounds were readily degraded with half-lives of less than 1 h, consistent with the high elimination efficiencies found in WWTPs (Table 1). A sterile control was not considered necessary as it is reasonable to assume that sorption to sludge is negligible for these quite hydrophilic compounds and thus the observed dissipation due to degradation. In a recently published study, 6358

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 17, 2008

cotinine was also shown to be quickly biodegraded in river sediment microcosm experiments (19). In contrast, N-formylnornicotine showed no dissipation within 24 h of incubation, confirming the persistence of this compound observed in WWTPs. Furthermore, the incubation experiment with spiked nicotine did not provide any indication for the formation of N-formylnornicotine from nicotine in activated sludge. Occurrence in Surface Waters at Low ng/L Concentrations. Cotinine, 3′-hydroxycotinine, and N-formylnornicotine were analyzed in samples from various Swiss Midland lakes. During winter, these lakes are typically well mixed, both laterally and vertically, so that samples from the outflow can be considered representative for the whole water body. The three compounds were detected in all Midland lakes investigated, at concentrations up to 15 ng/L cotinine, 77 ng/L 3′-hydroxycotinine, and 6 ng/L N-formylnornicotine (Table 2). The highest values were found in Greifensee, a lake with a densely populated catchment area. Typical concentration ratios were ∼2:11:1, thus slightly different from WWTP effluents (∼1:8:1, Discussion, see later). Representative chromatograms are shown in Figure 2c. Surprisingly, cotinine was also detected in a remote Alpine mountain lake at 2.6 ng/L, and thus clearly above LOQ derived from blank values (see Experimental Section). Nicotine itself was previously analyzed in air samples of Italian cities at low ng/m3 concentrations and was also detected in rural environments (29). Moreover, nicotine was shown to undergo photodegradation upon irradiation with artificial UV light to a number of metabolites, among others cotinine (30). Aerial transport, photodegradation, and deposition may thus explain low ng/L concentrations of cotinine in surface waters, even in remote locations. Application as Anthropogenic Markers for Domestic Wastewater. Concentrations of the three nicotine derivatives in lakes correlated reasonably well with the ratio of population per water throughflow (P/Q, Figure 3). P/Q expresses the expected anthropogenic burden of a lake by domestic

wastewater, which increases with the population in the catchment area and decreases with the water throughflow (dilution effect). These linear correlations indicate that the compounds are suitable anthropogenic markers for domestic wastewater contamination of surface waters, not only in a qualitative (6, 9), but also quantitative sense. Similar correlations have previously been observed for caffeine, polycyclic musks, or methyl triclosan (21, 31, 32). In another study, cotinine concentrations were shown to inversely correlate with the salinity in an urban estuary (Jamaica Bay, New York), receiving sewage effluent as the major source of freshwater during dry weather conditions (12). The slopes of the regression lines in Figure 3 are a measure of the mean per-capita loads in the outflows of these lakes (0.040 ( 0.011, 0.20 ( 0.05, and 0.018 ( 0.003 mg person-1d-1 for cotinine, 3′-hydroxycotinine, and N-formylnornicotine, respectively, Table 2). For N-formylnornicotine, per-capita export loads in lakes were thus very similar to those in untreated and treated wastewater (Table 1), indicating a high persistence of this compound also in the investigated surface waters. Inputs with Combined Sewer Overflows. For cotinine and 3′-hydroxycotinine, however, mean per-capita loads were higher in the outflows of lakes than in treated wastewater, suggesting additional inputs to lakes. As these compounds are largely eliminated in WWTPs (>90%, see above), discharges of untreated wastewater may considerably contribute to their overall inputs to surface waters. Assuming neither further direct inputs (Iother ) 0, see formula below), nor significant degradation/dissipation of the compounds in the investigated surface waters (R ) 0), the importance of inputs with untreated wastewater can be estimated using a simple mass balance approach: Itotal ) f × Iuntreated + (1 - f) × Itreated + Iother ) E + R (units: mg person-1d-1) (1) where Itotal is the total per-capita input to surface waters; Iuntreated and Itreated are the mean per-capita loads in untreated and treated wastewater, respectively (Table 1); f is the fraction of discharges of untreated wastewater; and E is the mean per-capita export load in lakes (Table 2). With this mass balance, fractions of discharges of untreated wastewater were estimated at ∼2.5 and 2.2% using the markers cotinine and 3′-hydroxycotinine, respectively. As in the canton of Zu ¨ rich, more than 99% of the households are connected to WWTPs (33), these inputs were most likely due to combined sewer overflows during rain events. For lake Greifensee, a similar fraction of untreated wastewater of 2-3% had previously been derived using the marker caffeine (34). In that study, it could be shown that the loads of caffeine in the receiving waters were considerably higher during rainy weather, due to combined sewer overflows. In another study in the Jamaica Bay estuary, cotinine concentrations decreased under heavy precipitation conditions (12). As removal efficiencies in WWTPs discharging to this estuary were, on average, only ∼50%, the dilution effect by rainwater was higher than the increased load of cotinine bypassing treatment due to combined sewer overflows. Environmental Relevance. In this study, three nicotine derivates were found in surface waters at concentrations in the low ng/L range. In small receiving waters with significant wastewater discharges, concentrations of a few hundred ng/L may be expected. Possible ecotoxicological risks associated with such environmental concentrations, can, however, not be assessed at present. While ample data on effects of nicotine on aquatic organisms exist (35), much less data are available for the compounds investigated here, with the exception of a few studies on (acute) effects of cotinine on frog embryos (FETAX teratogenesis assay (14)), hepatocytes of rainbow trout (15), and in the hemolymph of mussels (16). In these

studies, effects were observed only at mg/L concentrations. However, nothing is currently known on potential long-term effects of these compounds at low exposure levels.

Acknowledgments This research project was sponsored by the Federal Office for the Environment (BAFU, Berne, Switzerland). Interesting discussions were held with B. Hitzfeld and A. Weber and are kindly acknowledged. We thank the personnel of the WWTPs for providing wastewater samples and A. Ba¨chli for her help in surface water sampling and in the laboratory.

Literature Cited (1) Benowitz, N. L. Nicotine Safety and Toxicity; Oxford University Press: New York, 1998; p 203. (2) World Health Organization. The Tobacco Atlas; Myriad Editions: Brighton, 2002; p 128. (3) Gorrod, J. W.; Wahren, J. Nicotine and Related Alkaloids. Absorption, Distribution, Metabolism and Excretion; Chapman & Hall: London, 1993; p 299. (4) Tomlin, C. D. S. The e-Pesticide Manual, 13th ed., version 3.2.; The British Crop Protection Council: Hampshire, 2005. (5) Kolpin, D. W.; Furlong, E. T.; Meyer, M. T.; Thurman, E. M.; Zaugg, S. D.; Barber, L. B.; Buxton, H. T. Pharmaceuticals, hormones, and other organic wastewater contaminants in U.S. streams, 1999-2000: a national reconnaissance. Environ. Sci. Technol. 2002, 36, 1202–1211. (6) Metcalfe, C. D.; Miao, X.-S.; Koenig, B. G.; Struger, J. Distribution of acidic and neutral drugs in surface waters near sewage treatment plants in the lower Great Lakes, Canada. Environ. Toxicol. Chem. 2003, 22, 2881–2889. (7) Kolpin, D. W.; Skopec, M.; Meyer, M. T.; Furlong, E. T.; Zaugg, S. D. Urban contribution of pharmaceuticals and other organic wastewater contaminants to streams during differing flow conditions. Sci. Total Environ. 2004, 328, 119–130. (8) Stackelberg, P. E.; Furlong, E. T.; Meyer, M. T.; Zaugg, S. D.; Henderson, A. K.; Reissman, D. B. Persistence of pharmaceutical compounds and other organic wastewater contaminants in a conventional drinking-water-treatment plant. Sci. Total Environ. 2004, 329, 99–113. (9) Glassmeyer, S. T.; Furlong, E. T.; Koplin, D. W.; Cahill, J. D.; Zaugg, S. D.; Werner, S. L.; Meyer, M. T.; Kryak, D. D. Transport of chemical and microbial compounds from known wastewater discharges: potential use as indicators of human fecal contamination. Environ. Sci. Technol. 2005, 39, 5157–5169. (10) Haggard, B. E.; Galloway, J. M.; Green, W. R.; Meyer, M. T. Pharmaceuticals and other organic chemicals in selected NorthCentral and Northwestern Arkansas streams. J. Environ. Qual. 2006, 35, 1078–1087. (11) Hua, W.; Bennett, E. R.; Letcher, R. J. Ozone treatment and the depletion of detectable pharmaceuticals and atrazine herbicide in drinking water sourced from the upper Detroit River, Ontario, Canada. Water Res. 2006, 40, 2259–2266. (12) Benotti, M. J.; Brownawell, B. J. Distributions of pharmaceuticals in an urban estuary during both dry- and wet-weather conditions. Environ. Sci. Technol. 2007, 41, 5795–5802. (13) Huerta-Fontela, M.; Galceran, M. T.; Ventura, F. Ultraperformance liquid chromatography-tandem mass spectrometry analysis of stimulatory drugs of abuse in wastewater and surface waters. Anal. Chem. 2007, 79, 3821–3829. (14) Dawson, D. A.; Fort, D. J.; Smith, G. J.; Newell, D. L.; Bantle, J. A. Evaluation of the developmental toxicity of nicotine and cotinine with frog embryo teratogenesis assay: Xenopus. Teratog. Carcinog. Mutagen. 1988, 8, 329–338. (15) Gagne´, F.; Blaise, C.; Andre´, C. Occurrence of pharmaceutical products in a municipal effluent and toxicity to rainbow trout (Oncorhynchus mykiss) hepatocytes. Ecotoxicol. Environ. Saf. 2006, 64, 329–336. (16) Gagne´, F.; Blaise, C.; Fournier, M.; Hansen, P. D. Effects of selected pharmaceutical products on phagocytic activity in Elliptio complanata mussels. Comp. Biochem. Phys. C 2006, 143, 179–186. (17) Cahill, J. D.; Furlong, E. T.; Burkhardt, M. R.; Kolpin, D.; Anderson, L. G. Determination of pharmaceutical compounds in surfaceand ground-water samples by solid-phase extraction and highperformance liquid chromatography-electrospray ionization mass spectrometry. J. Chromatogr. A 2004, 1041, 171–180. VOL. 42, NO. 17, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6359

(18) Xu, X.; Iba, M. M.; Weisel, C. P. Simultaneous and sensitive measurement of anabastine, nicotine, and nicotine metabolites in human urine by liquid chromatography - tandem mass spectrometry. Clin. Chem. 2004, 50, 2323–2330. (19) Bradley, P. M.; Barber, L. B.; Kolpin, D. W.; McMahon, P. B.; Chapelle, F. H. Biotransformation of caffeine, cotinine, and nicotine in stream sediments: implications for use as wastewater indicators. Environ. Toxicol. Chem. 2007, 26, 1116–1121. (20) Haufroid, V.; Lison, D. Urinary cotinine as a tobacco smoke exposure index: a minireview. Int. Arch. Occup. Environ. Health 1998, 71, 162–163. (21) Buerge, I. J.; Poiger, T.; Mu ¨ ller, M. D.; Buser, H.-R. Caffeine, an anthropogenic marker for wastewater contamination of surface waters. Environ. Sci. Technol. 2003, 37, 691–700. (22) Buser, H.-R. Atrazine and other s-triazine herbicides in lakes and in rain in Switzerland. Environ. Sci. Technol. 1990, 24, 1049– 1058. (23) Buerge, I. J.; Buser, H.-R.; Poiger, T.; Mu ¨ ller, M. D. Occurrence and fate of the cytostatic drugs cyclophosphamide and ifosfamide in wastewater and surface waters. Environ. Sci. Technol. 2006, 40, 7242–7250. (24) Buser, H.-R.; Poiger, T.; Mu ¨ ller, M. D. Occurrence and environmental behavior of the chiral pharmaceutical drug ibuprofen in surface waters and in wastewater. Environ. Sci. Technol. 1999, 33, 2529–2535. (25) Kahle, M.;Buerge, I. J.;Hauser, A.;Mu ¨ ller, M. D.;Poiger, T. Azole fungicides: occurrence and fate in wastewater and surface waters Environ. Sci. Technol.,submitted. (26) Umweltbundesamt, Dessau-Rosslau, Germany. Available at http://www.umweltbundesamt.de/survey/us98/nikotin.htm. (27) Keller, R.; Krebs, H.; Radtke, T.; Hornung, R. Der Tabakkonsum der Schweizer Wohnbevo¨lkerung in den Jahren 2001 bis 2006, Zusammenfassung des Forschungsberichts 2007; Universita¨t Zu ¨ rich: Zu ¨ rich, 2007; p 11.

6360

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 17, 2008

(28) Piade, J. J.; Hoffmann, D. Chemical studies on tobacco smoke LXVII. Quantitative determination of alkaloids in tobacco by liquid chromatography. J. Liq. Chromatogr. 1980, 3, 1505–1515. (29) Cecinato, A.; Balducci, C. Detection of cocaine in the airborne particles of the Italian cities Rome and Taranto. J. Sep. Sci. 2007, 30, 1930–1935. (30) Benner, C. L.; Bayona, J. M.; Caka, F. M.; Tang, H.; Lewis, L.; Crawford, J.; Lamb, J. D.; Lee, M. L.; Lewis, E. A.; Hansen, L. D.; Eatough, D. J. Chemical composition of environmental tobacco smoke. 2. Particulate-phase compounds. Environ. Sci. Technol. 1989, 23, 688–699. (31) Buerge, I. J.; Buser, H.-R.; Mu ¨ ller, M. D.; Poiger, T. Behavior of the polycyclic musks HHCB and AHTN in lakes, two potential anthropogenic markers for domestic wastewater in surface waters. Environ. Sci. Technol. 2003, 37, 5636–5644. (32) Balmer, M. E.; Poiger, T.; Droz, C.; Romanin, K.; Bergqvist, P.A.; Mu ¨ ller, M. D.; Buser, H.-R. Occurrence of methyltriclosan, a transformation product of the bactericide triclosan, in fish from various lakes in Switzerland. Environ. Sci. Technol. 2004, 38, 390–395. (33) Office for Waste, Water, Energy, and Air of the Canton of Zu ¨ rich, Switzerland (AWEL). Available at http://www.hochwasser.zh.ch. (34) Buerge, I. J.; Poiger, T.; Mu ¨ ller, M. D.; Buser, H.-R. Combined sewer overflows to surface waters detected by the anthropogenic marker caffeine. Environ. Sci. Technol. 2006, 40, 4096– 4102. (35) PAN Pesticides Database. Available at http://www.pesticideinfo.org. (36) Liechti, P. Der Zustand der Seen in der Schweiz; Bundesamt fu ¨r Umwelt, Wald und Landschaft (BUWAL): Bern, 1994; p 159.

ES800455Q