Evaluation of Pharmaceuticals and Personal Care ... - ACS Publications

Apr 12, 2008 - ketoprofen, and ibuprofen (5/14), caffeine (4/14), and diethyltolua- mide (3/14) were .... One liter was taken on each of days 0, 1, 2,...
0 downloads 0 Views 1MB Size
Environ. Sci. Technol. 2008, 42, 6347–6353

Evaluation of Pharmaceuticals and Personal Care Products as Water-soluble Molecular Markers of Sewage NORIHIDE NAKADA,† KENTARO KIRI,† HIROYUKI SHINOHARA,† ARATA HARADA,‡ KEISUKE KURODA,§ SATOSHI TAKIZAWA,§ AND H I D E S H I G E T A K A D A * ,† Laboratory of Organic Geochemistry (LOG), Institute of Symbiotic Science and Technology, Tokyo University of Agriculture and Technology, Fuchu, Tokyo 183-8509, Japan, Water Quality Team, Water Environment Research Group, Public Works Research Institute, 1-6 minamihara, Tsukuba, Ibaraki 305-8516, Japan, and Department of Urban Engineering, Graduate School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo, Tokyo 113-8656, Japan

Received December 10, 2007. Revised manuscript received February 11, 2008. Accepted February 12, 2008.

We examined the utility of 13 pharmaceuticals and personal care products (PPCPs) as molecular markers of sewage contamination in riverine, groundwater, and coastal environments. The PPCPs were crotamiton, ibuprofen, naproxen, ketoprofen, fenoprofen, mefenamic acid, thymol, triclosan, propyphenazone, carbamazepine, diethyltoluamide, ethenzamide, and caffeine. Measurements in 37 Japanese rivers showed positive correlations of riverine flux of crotamiton (r2 ) 0.85), carbamazepine (r2 ) 0.84), ibuprofen (r2 ) 0.73), and mefenamic acid (r2 ) 0.67) with the population in the catchments. In three surveys in the Tamagawa estuary, crotamiton, carbamazepine, and mefenamic acid behaved conservatively across seasons within a salinity range of 0.4-29‰, suggesting their utility as molecular markers in coastal environments. Removal of ketoprofen and naproxen in the estuary was ascribed to photodegradation. Ibuprofen and thymol were removed from estuarine waters in summer by microbial degradation. Triclosan was removed by a combination of microbial degradation, photodegradation, and adsorption. These results were consistent with those of river water incubated for 8 d at 25 °C in the dark in order to examine the effects of biodegradation and photodegradation. Crotamiton was detected in groundwater from the Tokyo metropolitan area (12 out of 14 samples), suggesting wastewater leakage from decrepit sewers. Carbamazepine, ketoprofen, and ibuprofen (5/14), caffeine (4/14), and diethyltoluamide (3/14) were also detected in the groundwater, whereas the other carboxylic and phenolic PPCPs were not detected and were thought to be removed during their passage through soil. All the data demonstrated the utility of crotamiton and

* Corresponding author e-mail: [email protected]; phone: +8142-367-5825; fax: +81-42-360-8264. † Tokyo University of Agriculture and Technology. ‡ Public Works Research Institute. § The University of Tokyo. 10.1021/es7030856 CCC: $40.75

Published on Web 04/12/2008

 2008 American Chemical Society

carbamazepine as conservative markers in freshwater and coastal environments. We recommend combining these conservative markers with labile PPCPs to detect inputs of poorly treated sewage.

Introduction Organic compounds specific to certain pollution sources can be used as molecular markers. They are useful for identifying the sources of pollutants and tracing their transport pathways (1). Hydrophobic markers (e.g., coprostanol, linear alkylbenzenes) have been investigated and used in many studies to detect sewage particles in marine environments (1). On the other hand, only a few water-soluble markers (e.g., linear alkylbenzene sulfonates [LAS], fluorescent whitening agents [FWAs] (2)) are known, yet they have potential advantages. For example, they could be used to trace the transport of sewage-derived water-soluble components in coastal environments and to detect sewage contamination in groundwater. However, most of the markers so far proposed have some defects as molecular markers. LAS are biodegradable and are not applicable in coastal environments (3). Although FWAs are resistant to microbial degradation (4) and are used in studies of coastal environments (5), they are photodegradable and showed nonconservative behavior in an estuary (6). Also, our latest study (7) suggested that FWAs are readily removed during soil infiltration and are therefore not suitable as molecular markers in groundwater. The identification of persistent water-soluble markers is an important challenge for environmental organic geochemists (1). During the past decade, pharmaceutical and personal care products (PPCPs) in sewage effluent and river water have been extensively surveyed (8, 9). The reports proposed several PPCPs as water-soluble molecular markers. Glassmeyer et al. (8) suggested carbamazepine, diethyltoluamide, and caffeine as potential markers of sewage in river water through their extensive survey of 110 PPCPs in 10 rivers around the USA. Carbamazepine was also proposed as a persistent marker in groundwater because of its conservative behavior during soil passage when treated wastewater was infiltrated into unsaturated soil (10). Although persistent markers are useful for tracing the pathways of treated sewage, labile compounds such as caffeine can be used as indicators of untreated wastewater such as combined sewer overflow (CSO) (11). Buerge et al. (11) demonstrated the utility of caffeine for the quantitative estimation of the contribution from CSO to water-soluble components such as phosphates to lacustrine systems. However, these studies mainly focused on freshwater environments, and very few (12–15) have examined the utility of PPCPs as molecular markers in marine environments. Recently, Benotti and Brownawell (14) investigated the behavior of a range of PPCPs in an estuarine system to demonstrate the usefulness of several persistent compounds (e.g., carbamazepine and sulfamethoxazole) as conservative markers of sewage and that of labile compounds (e.g., caffeine) as indicators of inputs of CSO to a coastal system. Thus, we comprehensively assessed selected PPCPs as molecular markers with special emphasis on their stability in an estuarine system. We focused on 13 PPCPs: crotamiton, ibuprofen, naproxen, ketoprofen, fenoprofen, mefenamic acid, thymol, triclosan, propyphenazone, carbamazepine, diethyltoluamide, ethenzamide, and caffeine. In our previous study (16), we measured these PPCPs in sewage influents and effluents at five municipal sewage treatment plants in Tokyo, Japan. The results highlighted crotamiton and carbamazepine as perVOL. 42, NO. 17, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6347

FIGURE 1. Study area and sampling locations for the surveys of 37 major Japanese rivers (a), the Tamagawa River (R13) and groundwater and spring water (b), and the Tamagawa Estuary (c). The river water used for the incubation experiment was collected from Tamagawa River at the seaward end of the freshwater reach, indicated as a closed circle (Figure 1b). sistent molecular markers of sewage effluent due to their abundance (700 ( 244 and 60 ( 45 ng/L, respectively) and their lower removal efficiencies during secondary treatment (20 ( 22% and 3 ( 52%). In the present study, we examined the utility of the 13 PPCPs as molecular markers in fresh water (37 rivers throughout Japan and 15 groundwater samples from the Tokyo metropolitan area) and brackish water (the Tamagawa estuary).

Experimental Section Field Survey and Incubation Experiment. In a nationwide survey of river water, we collected grab water samples from 37 rivers in winter (November-December) of 2004 and 2005 (Figure 1a). The rivers were selected to represent a range of pollution status of Japanese rivers in terms of biochemical oxygen demand. Sampling locations were set at the seaward end of the freshwater reach of each river. Detailed information on the rivers is presented in Supporting Information Table S1. Surveys were conducted in the Tamagawa estuary, which is 12 km long (Figure 1c) and has no secondary tributaries. Approximately 10 estuarine water samples were collected along the salinity gradient from 0.4 to 29‰ at intervals of 0.1-6‰ in each of September 2005, January 2006, and June 2006. The surveys were conducted at high water of neap tide, and the freshwater flow rates were 46.1, 10.6, and 17.0 m3/s, respectively. The other physical parameters for the surveys were summarized in Supporting Information Table S2. In some cases, depending on season and tidal conditions, secondary effluent reached the middle of the estuary (around station 5) through a creek (station 11) and affected the behaviors of the PPCPs in the estuary. The river and estuarine water was collected from 0.2 m below the surface. Groundwater and spring water samples were collected at 15 locations in the Tokyo metropolitan area (Figure 1b) using a pump or bailer. Some of them were suspected of 6348

9

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

receiving sewage leaked from decrepit sewers. All water samples were collected by stainless steel bucket, stored in 3 L amber glass bottles, transported cool to the laboratory, and filtered through prebaked glass fiber filters (GF/F, Whatman). Part of each filtrate was acidified with 4 M HCl to pH 2, and then all filtrates were stored at 5 °C until extraction (normally