Journal of Contaminant Hydrology 126 (2011) 181–194 Contents lists available at SciVerse ScienceDirect Journal of Contaminant Hydrology journal homepage: www.elsevier.com/locate/jconhyd Review Article Importance of heterocylic aromatic compounds in monitored natural attenuation for coal tar contaminated aquifers: A review Philipp Blum a,⁎, Anne Sagner b, Andreas Tiehm b, Peter Martus c, Thomas Wendel d, Peter Grathwohl d a b c d Karlsruhe Institute of Technology (KIT), Institute for Applied Geosciences (AGW), Kaiserstraße 12, 76131 Karlsruhe, Germany Water Technology Center (TZW), Environmental Biotechnology, Karlsruher Straße 84, Karlsruhe, Germany URS Deutschland GmbH, Sutelstraße 2, 30659 Hannover, Germany University of Tübingen, Center for Applied Geoscience (ZAG), Sigwartstraße 10, 72076 Tübingen, Germany a r t i c l e i n f o Article history: Received 21 March 2011 Received in revised form 12 August 2011 Accepted 15 August 2011 Available online 22 August 2011 Keywords: Natural attenuation Coal tar creosote Plume lengths Biodegradation BTEX PAH NSO heterocycles a b s t r a c t NSO heterocycles (HET) are typical constituents of coal tars. However, HET are not yet routinely monitored, although HET are relatively toxic coal tar constituents. The main objectives of the study is therefore to review previous studies and to analyse HET at coal tar polluted sites in order to assess the relevance of HET as part of monitored natural attenuation (MNA) or any other long-term monitoring programme. Hence, natural attenuation of typical HET (indole, quinoline, carbazole, acridine, methylquinolines, thiophene, benzothiophene, dibenzothiophene, benzofuran, dibenzofuran, methylbenzofurans, dimethylbenzofurans and xanthene) were studied at three different field sites in Germany. Compound-specific plume lengths were determined for all main contaminant groups (BTEX, PAH and HET). The results show that the observed plume lengths are site-specific and are above 250 m, but less than 1000 m. The latter, i.e. the upper limit, however mainly depends on the level of investigation, the considered compound, the lowest measured concentration and/or the achieved compoundspecific detection limit and therefore cannot be unequivocally defined. All downstream contaminant plumes exhibited HET concentrations above typical PAH concentrations indicating that some HET are generally persistent towards biodegradation compared to other coal tar constituents, which results in comparatively increased field-derived half-lives of HET. Additionally, this study provides a review on physicochemical and toxicological parameters of HET. For three well investigated sites in Germany, the biodegradation of HET is quantified using the centre line method (CLM) for the evaluation of bulk attenuation rate constants. The results of the present and previous studies suggest that implementation of a comprehensive monitoring programme for heterocyclic aromatic compounds is relevant at sites, if MNA is considered in risk assessment and for remediation. © 2011 Elsevier B.V. All rights reserved. Contents 1. 2. 3. Introduction . . . . . . . . . . . . . . . . . . . . . . . . Physicochemical properties and toxicology of NSO heterocycles Determination of natural attenuation at 3 field sites . . . . . . 3.1. Field sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ⁎ Corresponding author at: Karlsruhe Institute of Technology (KIT), Germany. Tel.: +49 721 608 47612; fax: +49 721 606 279. E-mail address: philipp.blum@kit.edu (P. Blum). 0169-7722/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jconhyd.2011.08.004 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 183 185 185 182 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 3.1.1. Hamburg (North Germany) . . . . . . . . . . . . . . 3.1.2. ‘Testfeld Süd’ (South Germany) . . . . . . . . . . . . 3.1.3. Karlsruhe (South-West Germany) . . . . . . . . . . . 3.2. Analytical methods . . . . . . . . . . . . . . . . . . . . . . 3.3. Comparison of the compound-specific plume lengths . . . . . . 3.4. Comparison of compound-specific bulk attenuation rate constants 4. Monitored natural attenuation of NSO heterocylic aromatic compounds . 5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. Introduction An enormous number of coal tar or creosote contaminated sites exist worldwide, which are generally related to former wood impregnation plants, manufactured-gas plants (MGP) and waste disposal sites. In the USA, for example, up to 2000 MGP represent a significant environmental legacy (Landmeyer et al., 1998). In Germany, more than 1400 coal tar contaminated sites were identified and only 20% have been seriously considered for remediation so far (KORA, 2008). These sites are usually contaminated with monoaromatic hydrocarbons (MAH such as BTEX), polycyclic aromatic hydrocarbons (PAH), as well as phenolic and heterocyclic aromatic compounds containing nitrogen, sulfur or oxygen. The latter, NSO heterocycles (HET), however, are currently not routinely monitored (e.g. Neuhauser et al., 2009), although up to 40% of the water-soluble fraction of the coal tar consists of HET (Licht et al., 1996). In addition, HET are generally among the most toxic coal tar constituents (e.g. Eastmond et al., 1984; Eisentraeger et al., 2008; Hartnik et al., 2007; Johansen et al., 1998; Neuwoehner et al., 2009; Sagner et al., 2006; Seymour et al. 1997; Tiehm and Sagner, 2008). For example, Eisentraeger et al. (2008) showed that some HET such as quinoline, 6methylquinoline and xanthene are mutagenic using the Salmonella/microsome test. Biodegradation of HET and their metabolites was studied on various scales from laboratory to field scale (e.g. Broholm and Arvin, 2001; Safinowski et al., 2006; Sagner, 2009; Tiehm et al., 2008). Several laboratory experiments were carried out to study the biodegradability of HET and their transformation products under different redox conditions using contaminated groundwater and soils (e.g. Althoff et al., 2001; Dyreborg et al., 1996, 1997; Gai et al., 2008; Mundt et al., 2003). Several studies could demonstrate that HET can have an inhibitory effect on the biodegradation of creosote compounds such as PAH and benzene (e.g. Dyreborg et al., 1996, 1997; Meyer and Steinhart, 2000). Dyreborg et al. (1997), for example, studied various HET under four redox conditions (aerobic, denitrifying, sulphate-reducing, and methanogenic) over a period of more than 2 years using laboratory batch microcosms. They generally concluded that anaerobic degradation for HET was significantly slower than aerobic degradation. At three creosote contaminated sites in Denmark, Broholm and Arvin (2001) showed that some HET compounds such as thiophene and benzothiophene may also persist in very low concentrations throughout the contaminant plume. Safinowski et al. (2006) investigated the anaerobic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 185 185 186 187 187 190 192 192 192 degradation of some PAH and HET using laboratory and field studies at a site in South Germany (‘Testfeld Süd’), which is also considered in this study. Both data from laboratory and field indicate that biodegradation of PAH and HET appears to be limited not only by the availability of electron donors, electron acceptors, and appropriate bacterial strains, but also by the existence of potential inhibitors. The study provides an example for anaerobic cometabolic transformation of PAH and HET. Reineke et al. (2007) studied the N-heterocycles quinoline and isoquinoline with their hydroxylated and hydrogenated metabolites at a field site in Germany (Castrop-Rauxel). They showed that the ratio of hydroxylated to parent compound can be used as potential indicator of in situ biodegradation. However, the proposed ratio still has to be tested for other coal tar contaminated sites. Although detailed investigated field studies on the spatial distribution and biodegradation of HET mainly located in Denmark, Germany and Canada were conducted (e.g. Blum et al., 2008; D'Affonseca et al., 2008; Fraser et al., 2008; Herold et al. 2009; Johansen et al., 1997; Pereira and Rostad, 1986; Reineke, 2008), only few field studies exist worldwide (e.g. Hamburg, ‘Testfeld Süd’, Karlsruhe; D'Affonseca et al., 2011; Griebler et al., 2004; Tiehm and Schulze, 2003), where physicochemical properties, biodegradation, toxicity, propagation and long term risk assessment of HET are available and were comprehensively considered in context of MNA. The latter is part of the present review. Blotevogel et al. (2006, 2008) suggested an assessment procedure for the identification of HET priority substances by including physicochemical properties, biodegradability, toxicity and groundwater propagation. They considered concentrations of HET at three different control planes at five different sites in Germany and one in Denmark (Johansen et al., 1997); two of these sites are also considered here (Karlsruhe and ‘Testfeld Süd’). The main objective of their studies was to derive a priority list of HET. However, the suggested list of key compounds varies and is predominantly based on substance-specific properties rather than site-specific observations (Kern et al., 2008). A comprehensive synopsis and review of toxicology, persistence, spreading and quantification of biodegradation of HET in the field is not yet available. The main objective of the current study is to review existing data on physicochemical and toxicological properties of HET and also to include aspects on quantification of biodegradation and long-term propagation of HET, and to discuss the importance of assessing and monitoring HET for a successful implementation of MNA in coal tar contaminated aquifers. Furthermore, compound-specific plume lengths for all main P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 contaminant groups (BTEX, PAH and HET) and quantification of bulk attenuation rate constants of HET using the centre line method (CLM) are analysed in the present study on three representative field sites in Germany (Hamburg, ‘Testfeld Süd’ and Karlsruhe). 2. Physicochemical properties and toxicology of NSO heterocycles For the assessment and discussion on the implementation of MNA for HET, the most relevant transport parameters such as molecular structure, molecular weight (MW), aqueous solubility (Sw) and hydrophobicity using octanol–water partition coefficients (Kow) are provided for 14 frequently analyzed HET (Table 1; Fig. 1). Parameters of other typical coal tar constituents such as benzene, phenol and naphthalene are also provided for comparison. These parameters may differ significantly depending on chosen reference or database (e.g. Blotevogel et al., 2008; Melber et al., 2004). For example, the aqueous solubility of benzofuran varies according to Melber et al. (2004), which refer to the Chemfinder database, between 100 and 1000 mg/l at 18 °C. In general, aqueous solubilities of HET are much or even orders of magnitude higher than the equivalent PAH and/or BTEX, which is mainly caused by the increased polarity of the HET due to the heteroatoms in the molecules. For example, the aqueous solubility of quinoline is more than two orders of magnitude higher than the solubility of naphthalene and that of pyrrole is 33 times higher than of benzene (Table 1). For the fate of HET the knowledge on their sorption behaviour is important (e.g. Bi et al., 2006, 2007, 2009; Broholm et al. 1999, 2000; Hassett et al., 1980; Reineke et al., 2007). Using the approach of soil column chromatography, Bi et al. (2007) studied the effects of several environmental factors such as temperature, ionic strength, and dissolved/sorbed ion composition on the sorption of HET. Bi et al. (2006) could demonstrate that the latter can be described by a conceptual sorption model, which includes partitioning to soil organic matter, cation exchange and an additional sorption process (most likely surface complexation of neutral species). For the non-linear sorption of O- and SHET partitioning to soil organic matter is the dominant sorption mechanism (Bi et al., 2007). Hence, organic carbon normalized partition coefficients (Koc), which are well correlated with octanol–water partition coefficients (Kow) or the inverse of their aqueous solubility (e.g. Razzaque and Grathwohl, 2008), are provided in Table 1. Koc is often also predicted using the linear free energy relationships (LFERs). However, a sorbate-specific linear multiple regression analysis is required for the development of a polyparameter LFERs (e.g. Endo et al., 2009). Thus, for the current study distribution coefficients (Kd) were determined based on organic carbon contents (foc) and listed in Table 2. Although sorption is crucial for understanding contaminant transport, it is however not relevant under steady state conditions and thus not considered an efficient process for NA. Thus, sorption is not particularly evaluated for the so-called three lines of evidence approach, which for example in the USA requires (1) an observed decrease in contaminant concentrations and mass, (2) geochemical indicators and (3) evidences for microbial activity to demonstrate active biodegradation. 183 This approach is used in several countries such as USA, UK, Netherlands and Denmark (Rügner et al., 2006). Hence, sorption of pollutants results in retardation, but not in a decrease in pollutant mass, and therefore is not considered an efficient process for MNA. More critical for the acceptance of MNA and the identification of suitable HET key pollutants is an understanding of the toxicological effects of these heterocyclic aromatic compounds in ecosystems and towards humans. For example, Hartnik et al. (2007) found that PAH, which comprise up to 85% of pure creosote, account for only 13% of total toxicity in the creosote-contaminated groundwater, while methylated benzenes, phenolics (e.g. 2,6Dimethylphenol) and N-HET accounted for approximately 80% of the total measured toxicity. Eisentraeger et al. (2008) demonstrated that S-HET are more toxic than N- and O-HET and that quinoline, 6-methylquinoline and xanthene are even mutagenic using the Ames fluctuation test. Recently, it was demonstrated that the O-heterocycles benzofuran and 2,3-dimethylbenzofuran also have a dioxin-like activity in two bioassays (Hinger et al., 2011). Neuwoehner et al. (2009) studied the ecotoxicity of quinoline and its metabolites using genotoxicity (SOS-Chromotest) and also Ames fluctuation tests. They confirmed the previous results that quinoline shows genotoxicity (potentially mutagenic or carcinogenic). However, the hydroxylation of quinoline, which might be a product of biodegradation, leads to a detoxification of the genotoxic potential (Neuwoehner et al., 2009). Tiehm et al. (2008) demonstrated that BTEX, PAH, and HET present in polluted groundwater could be microbiologically degraded in an aerobic bioreactor. Biodegradation of the pollutants corresponded to the reduction of groundwater toxicity in the Vibrio fischeri test. To provide an overview of the toxic effects of specific HET in more detail, the results of three toxicity bioassays from previously performed studies, using the ubiquitous soil bacterium Pseudomonas fluorescens, the marine bacterium V. fischeri and the planktonic crustacean Daphnia (water flea) are reviewed here (Bundy et al., 2001; Eisentraeger et al., 2008; Hartnik et al., 2007). The outcome of these bioassays is an effective concentration (EC), which is a rough estimate of ecotoxicity of these compounds. For example, the median EC50 value is the concentration of a chemical that is required to result in a 50% effect on the studied organism. Thus, the smaller the EC value, the higher the toxicity of the studied compound. The results of the toxicity tests are summarised in Table 1. Comparisons between the various studied tests show generally very good agreement, although EC50 and EC80 values cannot be directly compared with each other. The highest correlation was found between the Microtox (V. fischeri) (Hartnik et al., 2007) and Daphnia magna immobilization tests (Eisentraeger et al., 2008) with a correlation coefficient of 0.98 for the 7 pairs (Table 1). Furthermore, out of 14 studied HET dibenzothiophene and dibenzofuran are the most ecotoxic HET in all three examined toxicity tests (Fig. 2). A high ecotoxicity was also found for methyldibenzofuran using the program EPI Suite™ from the U.S. EPA with a EC50 value of 0.45 mg/L (KORA, 2008). Although, for naphthalene very low EC50 values were determined in comparison to their equivalent HET such as quinoline, 1benzothiophene and benzofuran, N-hetereocyclics still accounts for around 80% of the total measured toxicity in the creosote-contaminated groundwater (Hartnik et al., 2007; 184 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 QSAR regression developed by Bundy et al. (2001) it is however not possible to adequately predict the toxicity of HET by log Kow (Fig. 1), although a correlation between log Kow and EC80 values exists (Fig. 2). Loibner et al., 2004). Quantitative structure–activity relationship (QSAR) models are also used to predict toxicity, carcinogenicity and biodegradability of HET (e.g. Bundy et al., 2001; Philipp et al., 2007; Zhang et al., 1992). Using the Table 1 Physicochemical and three effective concentrations (EC) based three different toxicity bioassays for the 14 studied NSO heterocyclic compounds and five other typical organic compounds of coal tars (benzene, ethlybenzene, phenol, naphthalene and acenaphthene). Aqueous solubility at 25 °C (mg/L) log Kow EC50 (Vibrio fischeri) 15 min (mg/L) EC50 (Daphnia) (mg/L)c EC80 (P. fluorescens) (mg/L)d 67.1 58,800b 0.75b – – 4707 Indole 117.0 1875e 2.00e 2.39f 1.3 172 Quinoline 129.2 6718a 2.03e 31.8d 14.7 115 Carbazole 167.2 1.2g 3.72g 11.57i 3.4 – Acridine 179.2 46.6g 3.48g 7.47i 4.6 27 6-Methylquinoline 143.2 631h 2.57h 9.03i – – 84.1 3600j 1.81j – – 753 1-Benzothiophene 134.2 130e 3.12e – n.d. 59 Dibenzothiophene 184.3 4.38e 0.12i 0.2 0.13 Benzofuran 118.1 2.67e 3.49f 2.1 174 Dibenzofuran 168.2 4.12e 1.09i 0.6 1.10 2-Methylbenzofuran 132.2 3.22h – 7.4 – 2,3-Dimethylbenzofuran 146.2 62.2j 3.63j – 6.2 – Xanthene 182.2 1.0h 4.23h – 2.3 – Benzene 78.1 1780b 2.12b 102.78i – 773 106.2 152b 3.13b 9.69i – – 94.1 81,300e 1.46e 25.61l – – Group Compound N-HET Pyrrole S-HET O-HET BTEX Thiophene Ethylbenzene Molecular structure Molecular weight (g/mol)a,b 1.5e 678k 4.8e 160h PKX Phenol PAH Naphthalene 128.2 31.7b 3.37b 0.93i 1.89m – n.d. Acenaphthene 154.2 3.9b 3.39b 0.74i – – P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 3. Determination of natural attenuation at 3 field sites In the current study three well investigated coal tar contaminated sites were selected in Germany (Hamburg, ‘Testfeld Süd’ and Karlsruhe), which were previously investigated in great detail and where HET were analysed. In addition to these previous studies, compound-specific plume lengths and bulk attenuation rate constants were evaluated as part of the current study. Table 2 summarises site-specific data such as operation time, average groundwater velocities, hydrogeological parameters, source widths and detected redox conditions. 185 sampling campaigns over a time period of more than 2 years (31.10.2003, 08.06.2004, 22.11.2004 and 16.06.2005) along the centre line of the plume. The measured organic compounds demonstrate similar decreasing concentrations with increasing distance to the source. The phenol concentrations along the central flow path decrease by a factor of 420, PAH by a factor of 330 and BTEX by a factor of 80. Importantly, the repetitive measurements indicate that the present contaminant plume is at steady-state and that the present biodegradation is sufficient to compensate the continuous contaminant mass flux from the source across control planes (Blum et al. 2009; D'Affonseca et al., 2008). 3.1. Field sites 3.1.1. Hamburg (North Germany) The Hamburg field site, which is a former wood impregnation plant, is located 40 km south of Hamburg, Germany. The operation, which closed in 1986, resulted in a coal tar contamination of the unsaturated and saturated zones down to 50 m depth (Table 2). The geology is mainly comprised of glacial sands with a local thickness of approximately 130 m separated into an upper and a lower aquifer. The regional upper aquifer can be subdivided in four different sediments: (1) medium to coarse sands from 63 m up to 43 m above sea level (asl); (2) fine sands from 43 m up to 23 m asl; (3) coarse to medium sands from 23 m up to 14 m asl and (4) silty fine sands from 14 m asl. The groundwater level in the upper unconfined aquifer is between 44 m and 45 m asl. The lower confined aquifer, which is not impacted by the contamination, is formed mainly by silt and sand deposits. Based on detailed site investigation campaigns a comprehensive three-dimensional conceptual sedimentological, hydrogeological and hydrogeochemical site model could be developed (Blum et al. 2009; D'Affonseca et al., 2008) and a reactive transport model was performed, which also includes redox-dependent carbon isotope fractionation (D'Affonseca et al., 2011). The average hydraulic conductivity of the lower part of the upper aquifer (medium to coarse sands between 40 and 49 mbs), in which the contaminant plume is predominantly located, was 4.2 × 10 − 4 m/s. The average seepage velocity in this layer is approximately 90 m/a using an average observed hydraulic gradient of 0.0015 and an average effective porosity of 21% (Blum et al., 2009). The contaminant plume of the most soluble PAH, i.e. naphthalene, is approximately 150 m wide and more than 450 m long. The longitudinal spreading of the organic compounds was studied using mean concentrations of four groundwater Note to Table 1: n.d. = not detectable. a Mueller et al. (1989). b Melber et al. (2004). c Eisentraeger et al. (2008). d Bundy et al. (2001). e Broholm et al. (2000). f Kaiser and Palabrica (1991). g Johansen et al. (1997). h Meylan and Boethling (1996). i Hartnik et al. (2007). j Verschueren (1996). k Blotevogel et al. (2008). l Ghosh and Doctor (1992). m Loibner et al. (2004). 3.1.2. ‘Testfeld Süd’ (South Germany) The second field site representing a former gaswork, which operated for almost 100 years (Table 2), caused a substantial subsurface contamination (e.g. Zamfirescu and Grathwohl, 2001). The shallow Quaternary aquifer with a thickness of around 3.5 m is extremely heterogeneous and mainly consists of sand and gravel. The aquifer is underlined by a deeper fractured aquifer system, which locally discharges to the shallow aquifer. The average hydraulic gradient and hydraulic conductivity are 0.002 and 1.7×10− 3 m/s, respectively. Assuming an average effective porosity of 15%, the average groundwater velocity results in 2.0 m/day (Table 2). Herold et al. (2009) studied the contaminant mass flow rates by a series of integral pumping tests (IPTs) and in particular for the main contaminants of concern such methylbenzofurans, dimethylbenzofurans and acenaphthene. They found that the highest mass flow rates at the most distal control plane were determined for dimethylbenzofurans with 1.11 g/day and for methylbenzofurans with 0.17 g/day. Other previous studies focused on the assessment of in situ biodegradation using compound specific isotope analysis (CSIA) for benzene, toluene, and o-xylene (Griebler et al., 2004; Mak et al. 2006; Peter et al., 2004). 3.1.3. Karlsruhe (South-West Germany) The third study site was a former sand pit area, located in Karlsruhe, South-West Germany, filled with municipal and industrial waste including slags, incineration ash and gasworks residues in the years 1925–1956. The disposal site covers an area of about 1500 m 2. The bottom of the dump reaches the water table in about 8 m depth. In the northern part of the dump, tar oil was found as non-aqueous phase liquid in the fluctuation zone of the groundwater table. The aquifer consists of porous sand and gravel representing a typical composition for the Upper Rhine Valley. The 186 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 5 different electron acceptors and analysis of the microbial community demonstrated the impact of redox conditions on biodegradation rates of PAH and BTEX (Schulze and Tiehm, 2004). N heterocycles S heterocycles O heterocycles Benzene Phenol Naphthalene 4 3.2. Analytical methods log Kow 3 2 1 0 0 1 2 3 4 5 log Sw (mg/L) Fig. 1. Relationship between log aqueous solubility (Sw) and log octanol– water partition coefficient (Kow) for the 14 studied NSO heterocyclic compounds and three other typical organic compounds of coal tars (benzene, phenol and naphthalene). Values are taken from Table 1. average groundwater velocity is 0.8 m/day with an average hydraulic conductivity of about 2 × 10 − 3 m/s (Table 2) (Schulze and Tiehm 2004; Tiehm and Schulze, 2003). Schulze and Tiehm (2004) could demonstrate different dominating redox zones in the plume by analysing the electron acceptors and respiration products in the field. They observed a succession of redox processes downgradient of the pit as well as a significant decrease in pollutant concentration with increasing distance to the disposal site. The changing ratio of naphthalene and acenaphthene along the plume pointed to active biodegradation processes, because naphthalene has a higher solubility than acenaphtene, but decreases more rapidly in the aqueous phase than the more hydrophobic and more sorptive acenaphthene (Tiehm and Schulze, 2003). Microcosm studies in the presence of Coal tar or creosotes are complex organic mixtures — creosote is obtained by the fractional distillation of crude coal tars. Coal tar is a mixture of several hundred, probably thousands of chemicals and by far not all of them have been identified. Six major classes of compounds can be distinguished: (1) aromatic hydrocarbons, including polycyclic aromatic hydrocarbons (PAHs) and alkylated PAHs, (2) tar acids/phenolics, (3) tar bases/nitrogen-containing heterocycles, (4) aromatic amines, (5) sulfur-containing heterocycles and (6) oxygen-containing heterocycles. HET and their metabolites are typically analysed by gas chromatography mass spectrometry (GC–MS) and high pressure liquid chromatography (HPLC) with diode array detector (DAD) or MS(–MS)-detectors (e.g. Johansen et al., 1996; Mundt and Hollender, 2005). In the current study 13 typical HET (indole, quinoline, carbazole, acridine, 6-methylquinoline, thiophene, 1-benzothiophene, dibenzothiophene, benzofuran, dibenzofuran, methylbenzofurans, 2,3-dimethylbenzofuran and xanthene) were enriched from specific groundwater samples (Hamburg, 15.06.2005; ‘Testfeld Süd’, 19.01.2006) by solid-phase-extraction (SPE) and liquid–liquid-extraction (LLE), respectively. The SPE-cartridges were extracted with ethyl acetate. PAHs were extracted from groundwater samples using cyclohexane as a solvent. Ethyl acetate extracts and cyclohexane extracts were analysed for NSOheterocyclic aromatic compounds and 16 EPA-PAHs by GC–MS. NSO-heterocyclic aromatic compounds were analysed by GC– MS in the full-scan mode, the PAHs were analysed in the selected ion monitoring (SIM) mode. BTEX analyses of the groundwater samples were performed using Purge and Trap Concentrator connected to GC–MS. For the LLE 45 mL of groundwater were extracted with 5 mL MTBE for 20 min. After separation and adding the internal standard, the organic phase was analysed with GC–MS. The latter Table 2 Summary of site-specific data. Site name Production/ operation time Average groundwater velocity, va [m/day] Transverse dispersivity, αT [m] Hamburga 1904–1986 0.25a 0.84 ‘Testfeld Süd’ Karlsruhe g d 1875–1970 1925–1956 2.0 d 0.80 n.a. = not analysed. a D'Affonseca et al. (2008). b Blum et al. (2009). c D'Affonseca et al. (2011). d Bösel et al. (2000). e Griebler et al. (2004). f Zamfirescu and Grathwohl (2001). g Tiehm and Schulze (2003). 30 0.30 g 0.03 Source width, Ws [m] 160 g 40 Average organic carbon content, foc [−] Detected redox conditions 0.0010 ± 0.0020 Aerobic, methanogenic, nitrate-, iron-, manganese and sulfate-reducingb,c Strongly reducing conditions, but no detailed delineation of the redox zonese,f Aerobic, methanogenic, nitrate-, iron-, manganese and sulfate-reducingg n.a. 0.0011 ± 0.0005 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 187 was performed with Agilent 6890N gas chromatograph equipped with an Agilent 5973 mass spectrometer. An Optima 5 capillary column (25 m×0.25 mm; 0.25 μm film thickness, Macherey & Nagel, Düren, Germany) was used for compound separation and helium was used as carrier gas. The temperature program was as follows: 40 °C (4 min isothermal), 40°–300 °C increasing with a rate of 5 °C/min, and 300 °C (12 min isothermal). The MS was operated in SIM mode. This method was used for the groundwater samples at the site in Karlsruhe. 26.1 mg/L close to the source (5 m). In a distance of 260 m from the source only methylbenzofurans could be detected with 0.5 μg/L and all other 10 HET were below the detection limit of 0.1 μg/L. Hence, the study of the compound-specific plume lengths for all three sites indicates that HET concentrations at the tip of contaminant plumes are generally high and frequently above PAH concentrations indicating that some HET are generally more persistent than other coal tar constituents. 3.3. Comparison of the compound-specific plume lengths 3.4. Comparison of compound-specific bulk attenuation rate constants In the current study plume lengths were determined for BTEX, PAH and NSO heterocycles along the centre line of the contaminant plumes using the maximum observed concentration with increasing distance from the source (Figs. 3 and 4). At the Hamburg site the total concentration of 11 measured HET at the edge of the contaminant plume, which include all HET in Table 1 apart from pyrrole, thiophene and dibenzothiophene, amounts to 56 μg/L at a distance of around 450 m from the source showing the highest total concentrations of all considered contaminant groups (Fig. 3). 1-Benzothiophene showed the highest concentration with 52 μg/L of all analyzed HET. At the ‘Testfeld Süd’ site the PAH concentrations along the centre line were mainly higher than the HET concentration towards the edge of the contaminant plume at a distance of around 420 m from the source, the total HET concentrations were 0.25 μg/L (only methylbenzofurans) and thus slightly above the PAH concentration (only acenaphthene and fluorene) at 0.17 μg/L. The comparatively high concentrations of methylbenzofurans, acenaphthene and fluorene demonstrate the persistence of these specific compounds. At this site the detection limit was 0.01 μg/L (PAH) and 0.1 μg/L (HET). At the Karlsruhe site the same 11 HET as at the site in Hamburg were analysed. However, here BTEX showed the highest observed concentrations followed by HET, which is probably due to the initial high concentrations of BTEX with Biodegradation was found to be effective at all three sites using various assessment methods such as hydrochemical analysis, microcosm biodegradation studies, microbial community analysis, signature metabolite analysis (SMA) and/or CSIA (e.g. Martus and Schaal, 2010; Peter et al. 2004; Tiehm and Schmidt, 2007; Tiehm and Schulze, 2003). For example, using CSIA in situ biodegradation could be qualitatively demonstrated for benzene, toluene, o-xylene, m/p-xylene, naphthalene and 1methylnaphthalene at the ‘Testfeld Süd’ (Griebler et al., 2004) and at the site in Hamburg for o-xylene and naphthalene (Blum et al., 2009). However, CSIA was not yet performed for HET. Hence, to quantitatively assess the in situ attenuation of HET a classical first order decay model was applied along the centre line of the plumes to determine bulk attenuation rate constants (e.g. Beyer et al., 2006; Blum et al., 2009; Zamfirescu and Grathwohl, 2001): λ¼− N heterocycles EC 50 (Vibrio fischeri ) S heterocycles O heterocycles N heterocycles EC 50 (Daphnia ) S heterocycles O heterocycles N heterocycles S heterocycles EC 80 (P. fluorescens ) O heterocycles Benzene Phenol Naphthalene QSAR (Bundy et al., 2001) 2 log EC 50,80 (mg/L) ð1Þ The bulk attenuation rate constant λ is obtained by linear regression plotting the logarithmic concentrations change C(x)/C0 over travel time. The travel time was derived from travel distance Δx divided by the average groundwater 3 1 0 -1   va C ðxÞ ln Δx C0 1 2 3 4 5 log Kow (mg/L) Fig. 2. Relationship between log octanol–water partition coefficient (Kow) and the effective concentration (EC50 and EC80) using three different toxicity bioassays (Vibrio fischeri, Daphnia, Pseudomonas fluorescens) for the 14 studied NSO heterocyclic compounds and three other typical organic compounds of coal tars (benzene, phenol and naphthalene) including the quantitative structure–activity relationship (QSAR) regression by Bundy et al. (2001). 188 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 a) 100,000 Hamburg (Germany) PAH (EPA) 10,000 BTEX HET 1,000 100 Groundwater flow 10 1 0 100 200 300 400 500 Distance from the source [m] b) 100,000.0 Testfeld Süd (Germany) PAH (EPA) 10,000.0 BTEX 1,000.0 HET 100.0 10.0 1.0 0.1 Groundwater flow 0.0 0 100 200 300 400 500 600 Distance from the source [m] c) 100,000 Karlsruhe (Germany) PAH (EPA) 10,000 BTEX HET 1,000 100 10 Groundwater flow 1 0 0 100 200 300 Distance from the source [m] Fig. 3. Concentrations along the centre line of the plumes for the a) Hamburg, b) ‘Testfeld Süd’ (please note that at the final well all measured HET concentrations are below the detection limit of 0.1 μg/l and therefore no data point is shown) and c) Karlsruhe sites in Germany. velocity va (Table 2). Half-lives (t1/2 = 0.693 / λ) for all three sites are summarised in Table 3 and illustrated in Fig. 5. Due to very low concentrations close or even below the detection limit it was not always possible to evaluate the bulk attenuation rate constants or half-lives, respectively (Table 3). The overall lowest bulk rate attenuation constant was determined for 2,3-dimethylbenzofuran with 0.40 a− 1 at the Hamburg site, which in comparison to the other two sites also has the lowest bulk rate attenuation constants. A previously performed sensitivity analysis at the Hamburg site for the evaluation 189 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 a) 1,000 Concentration [µg/l] Hamburg (Germany) 100 10 Acenaphthene Ethylbenzene Groundwater flow Benzothiophene 1 0 100 200 300 400 500 400 500 Distance from the source [m] b) 1,000.0 Concentration [µg/l] Testfeld Süd (Germany) 100.0 10.0 Acenaphthene Ethylbenzene 1.0 Methylbenzofurans Groundwater flow 0.1 0 100 200 300 Distance from the source [m] c) 100,000.0 Karlsruhe (Germany) Concentration [µg/l] 10,000.0 1,000.0 Groundwater flow 100.0 10.0 Acenaphthene Benzene 1.0 Methylbenzofurans 0.1 0 100 200 300 Distance from the source [m] Fig. 4. Highest compound-specific concentrations along the centre line of the plumes for the a) Hamburg, b) ‘Testfeld Süd’ and c) Karlsruhe sites in Germany. of the bulk rate attenuation constants showed that the groundwater velocity (hydraulic conductivity and porosity) has the greatest influence followed by other input parameters such as source concentration, source width, longitudinal and transversal dispersions (Blum et al., 2007; 2009). Although first-order decay rates or half-lives do not reflect the relevant biodegradation processes, this approach is suitable to compare the overall elimination rates for different pollutants or different sites. Very high half-lives were determined for 1-benzothiophene and methylbenzofurans, which also showed high concentrations downgradient of the source indicating the persistence of these two compounds (Table 3, Fig. 4). The bulk rate attenuation 190 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 plume and therefore lower measured concentrations at the tip of the plume. The latter indicates that with ongoing knowledge of the delineation of the contaminant plume, the evaluation of bulk rate attenuation constants might also change in time and space. Table 3 Half-lives based on the determined bulk attenuation rate constants for all three studied sites including the Pearson product–moment correlation coefficients (PMCC) in brackets between the measured and simulated concentrations using a first order decay model (Eq. (1)). Compound Half-life t1/2 [days] and PMCC in brackets 4. Monitored natural attenuation of NSO heterocylic aromatic compounds Hamburg ‘Testfeld Süd’ Karlsruhe N-HET Pyrrole Indole Quinoline Carbazole Acridine 6-Methylquinoline S-HET Thiophene 1-Benzothiophene Dibenzothiophene O-HET Benzofuran Dibenzofuran Methylbenzofurans 2,3-Dimethylbenzofuran Xanthene BTEX Benzene Ethylbenzene PAH Naphthalene Acenaphthene n.a. 154 (1.0) n.p.b 204 (0.8) 329 (1.0) n.p. n.a. 538 (0.8) 230 (0.5) 460 (1.0) 198 (0.9) 538 (0.9) 632 (0.6) n.p. 347 (0.6) 329 (0.7) 182 (1.0) 316 (0.9) n.a. n.a. n.p.c 13 (1.0) n.p.c n.a. n.p.d 11 (1.0) 24 (0.8) n.p.e n.a. 35 (0.9) n.p.f n.p.e 7 (1.0) 9 (1.0) 1 (1.0) 21 (0.9) n.a. 63 (–)a n.p.d 25 (0.9) 8 (–)a n.p.d n.a. 22 (1.0) 67 (0.9) 13 (1.0) 33 (0.8) 44 (1.0) 60 (0.9) 13 (–)a 26 (0.8) 81 (0.7) 16 (1.0) 90 (0.9) For the successful implementation of MNA two main different concepts are currently applied worldwide: (1) risk-based MNA concepts as used, for example, in the USA, and (2) MNA concepts that rely on a precautionary principle of soil and groundwater protection as developed, for example, in Germany (Rügner et al., 2006). To comprehensively assess MNA at a site, several properties such as toxicology, persistence, spreading and biodegradation of the contaminants are usually considered (Tiehm and Schmidt, 2007). Hence, based on the present review Table 4 summarises the top three HET regarding physicochemical, toxicological properties, spreading and natural attenuation. In addition, 9 minimal recommended priority HET are provided for the assessment of MNA at coal tar contaminated sites, which are based on the current review and two previously performed studies (Table 4). The study by Johansen et al. (1997) at three creosote contaminated sites in Denmark showed that pyrrole was only detected once with a very low concentration of 0.22 μg/L. Quinoline, also with a high aqueous solubility, was measured at several sites in Germany and Denmark (Blotevogel et al., 2008; Johansen et al., 1997). At the three sites of the present study the highest concentration was measured at the ‘Testfeld Süd’ with 1.6 μg/L, though only close to the source. In all other samples quinoline was always below the detection limit. Thiophene, the third most soluble HET (Table 4), was measured at the Danish site by Johansen et al. (1997), could be detected only close to the contaminant sources (b55 m) with a maximum concentration of 9.2 μg/L. Although, pyrrole, quinoline and thiophene have rather high aqueous solubilities, they are hardly present n.a. = not analysed; n.p. = no best fit is possible. a Only two concentrations. b All concentrations below the detection limit (b 0.8 μg/L). c Only two concentrations above the detection limit (b 0.5 μg/L). d Concentrations below the detection limit (b0.1 μg/L). e Only one concentrations above the detection limit (b 0.5 μg/L). f Initial concentration lower than first concentration downgradient. constants for the ‘Testfeld Süd’ determined in this study are generally higher in comparison with previous studies by Zamfirescu and Grathwohl (2001), Bockelmann et al. (2003) and Piepenbrink et al. (2005), who applied an integral approach, mainly due to ongoing site investigations at the tip of the plume resulting in a better delineation of the contaminant Hamburg Karlsruhe Testfeld Süd Half-life [days] 1000 100 10 en e e N ap ht h al en Be nz en e nz hy ot h nz 1- Be lb e io ba ph en e zo le e ar C en ap ht he n en e Ac ot en z ib D M et hy lb en zo hi op h fu r an s 1 Et Group Compound Fig. 5. Comparison of selected field-derived half-lives for all three studied sites. 191 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 Table 4 Important parameters for the implementation of monitored natural attenuation (MNA) and the top three NSO heterocyclic compounds. Parameter NSO heterocyclic compounds Top 1 Top 2 Top 3 Physicochemical properties Highest aqueous solubility Pyrrole Quinoline Thiophene Toxicological properties Lowest EC50 (Daphnia) Dibenzothiophenea Dibenzofurana Acridineb Spreadingc Highest concentration at the tip of the plume Benzothiophene Methylbenzofurans Natural attenuationc Highest half-life 2,3-Dimethylbenzofuran Methylbenzofurans/benzothiophene a b c Dibenzothiophene Using EC50 by Hartnik et al. (2007). using EC80 by Bundy et al. (2001). Site-specific data from the three studied sites. in the studied aquifers and if present only close to the source and in low concentrations. Thus, we do not recommend them as priority HET (Table 5). The review of several toxicity studies shows that dibenzothiophene and dibenzofuran are the most ecotoxic HET. In addition, both HET are also present in high concentrations in contaminant plumes. Furthermore, dibenzothiophene shows one of the lowest bulk attenuation rate constants in all three studied sites. Thus, both HET should be carefully considered for a successful implementation of MNA at a given site. For the studied sites the highest concentrations of HET at the tip of the plumes were observed for benzothiophene and methylbenzofurans. Similar observations were reported for the Danish sites (Johansen et al., 1997) and four other German sites (Blotevogel et al., 2008), where both contaminants also showed generally high concentrations close to the source, and also downgradient of the source. In addition to the previous observations, 2,3-dimethylbenzofuran shows very low bulk attenuation rate constants indicating a high persistence. Based on physicochemical and toxicological properties, microbial degradation and site-specific concentrations of HET at 6 coal tar contaminated sites Blotevogel et al. (2008) derived a priority list of 20 HET for the investigation of the contaminant plumes (Table 5), which also contains substituted derivates and transformation products. The assessment, which is based on substance-specific properties and site-specific observations, is biased towards substance-specific properties with an average ratio of 2.3 between substance- and site-specific properties. In the current study 9 key HET were identified (Table 4), 7 of them are also part of the previously derived priority list of 20 HET. Dibenzothiophene with the lowest EC50 value and a low bulk attenuation rate constant and two other HET (pyrrole Table 5 Derived priority NSO heterocycles based on three different studies by Johansen et al. (1997), Blotevogel et al. (2008), this study and minimal recommended priority NSO heterocycles for the assessment of MNA for coal tar contaminated aquifers. Group Priority NSO heterocycles Johansen et al. (1997) Blotevogel et al. (2008) This study Recommendation N-HET – – – – Quinoline Isoquinoline Isoquinoline-1-on Quinoline-2-on 4-Methylquinoline-2-on 2,4-Dimethylquinoline 1-Methylisoquinoline 2-Methylquinoline Carbazole Acridine – – Benzothiophen – 3-Methylbenzothiophen 2-Hydroxybiphenyl Benzofuran Dibenzofuran Methylbenzofurans Dimethylbenzofurans Xanthenon Pyrrole Quinoline – – – – – – – – Acridine – Thiophene Benzothiophene Dibenzothiophene – – – Dibenzofuran Methylbenzofurans Dimethylbenzofurans – – – – – – – – – 2-Methylquinoline Carbazole Acridine – – Benzothiophene Dibenzothiophene – – Benzofuran Dibenzofuran Methylbenzofurans Dimethylbenzofurans – – S-HET O-HET – – – Carbazole – Hydroxylquinolines Thiophene Benzothiophene – – – Benzofuran Dibenzofuran – – – 192 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 and thiophene) however are not included in this priority list and in the 6 recommended HET (carbozole, thiophene, benzothiophene, benzofuran, dibenzofuran and hydroxylquinolines) by Johansen et al. (1997). Several studies including the present study (e.g. Blotevogel et al., 2008; Blum et al., 2008; Johansen et al., 1997, Kern et al. 2008), wherein physicochemical properties, biodegradation, toxicity and spreading of HET were comprehensively considered in the field, could evidently demonstrate the importance of HET for MNA. Since 1997 the 16 EPA priority pollutants for PAH are widely accepted to assess their environmental impact. How many HET should be considered and regularly monitored for the long term risk management of coal tar or creosote contaminated aquifers, however is still under debate and might also depend on the country-specific MNA approach. Nevertheless, based on the current knowledge we recommend a minimum of 9 HET that should be regularly monitored for a successful implementation of MNA (Table 5). 5. Conclusions In the present study substance-specific properties such as physicochemical and toxicological properties and sitespecific observations such as compound-specific plume lengths and half-lives of HET were investigated. The major findings of the study can be summarised as follows: • The aqueous solubilities of HET are generally several times higher than those of equivalent PAH and/or BTEX, which is mainly caused by the increased polarity of the HET due to the hetero-atoms in the molecules; • The few currently available toxicity studies indicate the high toxicity of NSO heterocycles — out of 14 studied HET dibenzothiophene and dibenzofuran are the most ecotoxic HET in all three examined toxicity assays; • Among the studied HET, 1-benzothiophene and methylbenzofurans show the highest concentrations relative to other compounds at the tip of the plumes; • The relative concentrations and transport distances of some HET in comparison to other typical coal tar constituents (PAH and BTEX) clearly demonstrate the persistence of such compounds in contaminant plumes; • Estimated field-derived half-lives of HET are usually higher than their equivalent PAH and/or BTEX. • A minimum of 9 HET (methylquinoline, carbazole, acridine, benzothiophene, dibenzothiophene, benzofuran, dibenzofuran, methylbenzofurans and dimethylbenzofurans) are recommended for the assessment of MNA in coal tar contaminated aquifers. Our analysis of the literature data and our results on the plume length and bulk attenuation rates clearly indicate that – due to their comparatively high aqueous solubility, low octanol–water partition coefficients, high toxicity, relative long and persistence plumes – some NSO heterocycles form relatively long and persistence plumes that are comparable to other typical coal tar constituents such as PAH and BTEX, in relation to their size. Thus, at least a number of key NSO heterocycles should be regularly assessed and monitored for a successful implementation of MNA in coal tar or creosote contaminated aquifers. Hence, there is an urgent need to derive a priority list of HET, which is generally accepted, applicable and appropriate for any chosen MNA concept. Acknowledgments We would like to thank Christoph Olk and Rainer Domalski from the Rütgers GmbH for their support. The financial support by the Rütgers GmbH for the field studies at the Hamburg site is highly appreciated. Financial support of the German Ministry of Education and Research (BMBF, grants no. 02WT9957/9, 02WN0361 and 02WN0362) for studying the Karlsruhe and ‘Testfeld Süd’ sites is also gratefully acknowledged. Finally, we also thank three anonymous reviewers for their helpful comments and suggestions that improved this manuscript. References Althoff, K., Mundt, M., Eisentraeger, A., Dott, W., Hollender, J., 2001. Microcosms-experiments to assess the potential for natural attenuation of contaminated groundwater. Water Res. 35 (3), 720–728. Beyer, C., Bauer, S., Kolditz, O., 2006. Uncertainty assessment of contaminant plume length estimates in heterogeneous aquifers. J. Contam. Hydrol. 87, 73–95. Bi, E., Schmidt, T.C., Haderlein, S.B., 2006. Sorption of heterocyclic organic compounds to reference soils: column studies for process identification. Environ. Sci. Technol. 40 (19), 5962–5970. Bi, E., Schmidt, T.C., Haderlein, S.B., 2007. Environmental factors influencing sorption of heterocyclic aromatic compounds to soil. Environ. Sci. Technol. 41 (9), 3172–3178. Bi, E., Zhang, L.L., Schmidt, T.C., Haderlein, S.B., 2009. Simulation of nonlinear sorption of N-heterocyclic organic contaminates in soil columns. J. Contam. Hydrol. 107 (1–2), 58–65. Blotevogel, J., Reineke, A.K., Hollender, J., Held, T., 2006. Monitoring of NSOheterocycles at creosote-contaminated sites. Grundwasser 11 (4), 295–297 (in German). Blotevogel, J., Reineke, A.K., Hollender, J., Held, T., 2008. Identification of NSO-heterocyclic priority substances for investigating and monitoring creosote-contaminated sites. Grundwasser 13 (3), 147–157 (in German). Blum, P., Kamkar, P., Melzer, R., 2007. Sensitivitätsanalyse von Natural Attenuation anhand analytischer Transportmodelle. Altlasten Spektrum 2, 74–81. Blum, P., Barker, J.F., Fraser, M., Sagner, A., Tiehm, A., Melzer, R., Grathwohl, P., 2008. Natural Attenuation of NSO Heterocycles in Coal Tar Contaminated Aquifers, 324. IAHS Publication, pp. 327–334. Blum, P., Hunkeler, D., Weede, M., Beyer, C., Grathwohl, P., Morasch, B., 2009. Quantification of biodegradation for o-xylene and naphthalene using first-order, Michaelis–Menten kinetics and stable carbon isotopes. J. Contam. Hydrol. 105, 118–130. Bockelmann, A., Zamfirescu, D., Ptak, T., Grathwohl, P., Teutsch, G., 2003. Quantification of mass fluxes and natural attenuation rates at an industrial site with a limited monitoring network: a case study. J. Contam. Hydrol. 60, 97–121. Bösel, D., Herfort, M., Ptak, T., Teutsch, G., 2000. Design, performance, evaluation and modelling of a natural gradient multitracer transport experiment in a contaminated heterogeneous porous aquifer. In: Dassargues, A. (Ed.), Tracers and Modelling in Hydrogeology, 262. IAHS Publication, pp. 45–51. Broholm, M.M., Broholm, K., Arvin, E., 1999. Sorption of heterocyclic compounds on natural clayey till. J. Contam. Hydrol. 39 (3–4), 183–200. Broholm, K., Nilsson, B., Sidle, R.C., Arvin, E., 2000. Transport and biodegradation of creosote compounds in clayey till, a field experiment. J. Contam. Hydrol. 41 (3–4), 239–260. Broholm, K., Arvin, E., 2001. Biodegradation of creosote compounds: comparison of experiments at different scales. Ground Water Monit. Rem. 21 (4), 101–108. Bundy, J.G., Morriss, A.W.J., Durham, D.G., Campbell, C.D., Paton, G.I., 2001. Development of QSARs to investigate the bacterial toxicity and biotransformation potential of aromatic heterocylic compounds. Chemosphere 42 (8), 885–892. D'Affonseca, F.M., Blum, P., Finkel, M., Melzer, R., Grathwohl, P., 2008. Field scale characterisation and modelling of contaminant release from a coal tar source zone. J. Contam. Hydrol. 102, 120–139. P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 D'Affonseca, F.M., Prommer, H., Finkel, M., Blum, P., Grathwohl, P., 2011. Modeling of the long-term and transient evolution of biogeochemical and isotopic signatures in coal tar contaminated aquifers. Water Resour. Res. 47, W05518. Dyreborg, S., Arvin, E., Broholm, K., 1996. Effects of creosote compounds on the aerobic bio-degradation of benzene. Biodegradation 7 (3), 191–201. Dyreborg, S., Arvin, E., Broholm, K., 1997. Biodegradation of NSO-compounds under different redox-conditions. J. Contam. Hydrol. 25 (3–4), 177–197. Eastmond, D.A., Booth, G.M., Lee, M.L., 1984. Toxicity, accumulation and elimination of polycyclic aromatic sulfur heterocycles in Daphnia magna. Arch. Environ. Contam. Toxicol. 13, 105–111. Eisentraeger, A., Brinkmann, C., Hollert, H., Sagner, A., Tiehm, A., Neuwoehner, J., 2008. Heterocyclic compounds: toxic effects using algae, daphnids, and the Salmonella/microsome test taking methodical quantitative aspects into account. Environ. Toxicol. Chem. 27 (7), 1590–1596. Endo, S., Grathwohl, P., Haderlein, S.B., Schmidt, T.C., 2009. LFERs for soil organic carbon - water distribution coefficients (KOC) at environmentally relevant Sorbate concentrations. Environ. Sci. Techn. 43 (9), 3094–3100. Fraser, M., Barker, J.F., Butler, B., Blaine, F., Joseph, S., Cooke, C., 2008. Natural attenuation of a plume from an emplaced coal tar creosote source over 14 years. J. Contam. Hydrol. 100 (3–4), 101–115. Gai, Z., Yu, B., Wang, X., Deng, Z., Xu, P., 2008. Microbial transformation of benzothiophenes, with carbazole as the auxiliary substrate, by Sphingomonas sp. strain XLDN2-5. Microbiology 154 (12), 3804–3812. Ghosh, S.K., Doctor, P.B., 1992. Toxicity screening of phenol using microtox. Environ. Toxicol. Water Qual. 7 (2), 157–163. Griebler, C., Safinowski, M., Vieth, A., Richnow, H.H., Meckenstock, R.U., 2004. Combined application of stable carbon isotope analysis and specific metabolites determination for assessing in situ degradation of aromatic hydrocarbons in a tar oil-contaminated aquifer. Environ. Sci. Technol. 38, 617–631. Hartnik, T., Norli, H.R., Eggen, T., Breedveld, G.D., 2007. Bioassay-directed identification of toxic organic compounds in creosote-contaminated groundwater. Chemosphere 66 (3), 435–443. Hassett, J.J., Means, J.C., Banwart, W.L., Wood, S.G., Ali, S., Khan, A., 1980. Sorption of dibenzothiophene by soils and sediments. J. Environ. Qual. 9, 184–186. Herold, M., Ptak, T., Bayer-Raich, M., Wendel, T., Grathwohl, P., 2009. Integral quantification of contaminant mass flow rates in a contaminated aquifer: conditioning of the numerical inversion of concentration-time series. J. Contam. Hydrol. 106 (1–2), 29–38. Hinger, G., Brinkmann, M., Bluhm, K., Sagner, A., Takner, H., Eisenträger, A., Braunbeck, T., Eengwall, M., Tiehm, A., Hollert, H., 2011. Some heterocyclic aromatic compounds are Ah receptor agonists in the DR-CALUX assay and the EROD assay with RTL-W1 cells. Environ. Sci. Pollut. Res. doi:10.1007/s11356-011-0483-7 Johansen, S.S., Hansen, A.B., Mosbæk, H., Arvin, E., 1996. Method development for trace analysis of heteroaromatic compounds in contaminated groundwater. J. Chromatogr. A 738 (2), 295–304. Johansen, S.S., Hansen, A.B., Mosbæk, H., Arvin, E., 1997. Identification of heteroaromatic and other organic compounds in ground water at creosotecontaminated sites in Denmark. Ground Water Monit. Rem. 17 (2), 106–115. Johansen, S.S., Arvin, E., Mosbæk, H., Hansen, A.B., 1998. Heteroaromatic compounds and their biodegradation products in creosotecontaminated groundwater. Toxicol. Environ. Chem. 66 (1–4), 195–228. Kaiser, K.L.E., Palabrica, V.S., 1991. Photobacterium phosphoreum toxicity data index. Water Poll. Res. J. Can. 26, 361–431. Kern, F., Möhser, H., Reinhard, M., Sagner, A., Sorg, K.-P., Tiehm, A., 2008. NSO-Heterozyklen – Vorkommen, Analytik, Beurteilung – Hinweise für die Praxis, Schriftenreihe des Altlastenforums Baden-Württemberg, Heft 12, E. Schweizerbart'sche Verlagsbuchhandlung, Stuttgart. KORA, 2008. Leitfaden – Natürliche Schadstoffminderungen bei Teerölaltlasten. Beiträge zu Abfallwirtschaft und Altlasten, Band 58. Technische Universität Dresden. 270 pp. Landmeyer, J.E., Chapelle, F.H., Petkewich, M.D., Bradley, P.M., 1998. Assessment of natural attenuation of aromatic hydrocarbons in groundwater near a former manufactured-gas plant, South Carolina. USA. Environ. Geol. 34 (4), 279–292. Licht, D., Ahring, B.K., Arvin, E., 1996. Effects of electron acceptors, reducing agents, and toxic metabolites on anaerobic degradation of heterocyclic compounds. Biodegradation 7, 83–90. Loibner, A.P., Szolar, O.H.J., Braun, R., Hirmann, D., 2004. Toxicity testing of 16 priority polycyclic aromatic hydrocarbons using Lumistox®. Environ. Toxicol. Chem. 23 (3), 557–564. Mak, K.S., Griebler, C., Meckenstock, R.U., Liedl, R., Peter, A., 2006. Combined application of conservative transport modelling and compound-specific carbon isotope analyses to assess in situ attenuation of benzene, toluene, and o-xylene. J. Contam. Hydrol. 88, 306–320. 193 Martus, P., Schaal, W., 2010. Metabolite analysis as direct proof of biodegradation: experience from monitored natural attenuation (MNA) projects. Environ. Forensics 11 (1–2), 94–101. Melber, C., Kielhorn, J., Mangelsdorf, I., 2004. Coal tar creosote. Concise International Chemical Assessment Document, 62. World Health Organization, Geneva. Meyer, S., Steinhart, H., 2000. Effects of heterocyclic PAHs (N, S, O) on the biodegradation of typical tar oil PAHs in a soil/compost mixture. Chemosphere 40, 359–367. Meylan, M.W., Boethling, R.S., 1996. Improved method for estimating water solubility from octanol/water partition coefficient. Environ. Toxicol. Chem. 15 (2), 100–106. Mueller, J.G., Chapman, P.J., Pritchard, P.H., 1989. Creosote-contaminated sites. Their potential for bioremediation. Environ. Sci. Technol. 23, 1197–1201. Mundt, M., Althoff, K., Dott, W., Hollender, J., 2003. Microbial degradation of tar oil compounds under different redox conditions. Acta Hydroch. Hydrob. 31 (3), 204–212. Mundt, M., Hollender, J., 2005. Simultaneous determination of NSOheterocycles, homocycles and their metabolites in groundwater of tar oil contaminated sites using LC with diode array UV and fluorescence detection. J. Chromatogr. A 1065 (2), 211–218. Neuhauser, E.F., Ripp, J.A., Azzolina, N.A., Madsen, E.L., Mauro, D.M., Taylor, T., 2009. Monitored natural attenuation of manufactured gas plant tar mono- and polycyclic aromatic hydrocarbons in ground water: a 14year field study. Ground Water Monit. Rem. 29 (3), 66–76. Neuwoehner, J., Reineke, A.K., Hollender, J., Eisentraeger, A., 2009. Ecotoxicity of quinoline and hydroxylated derivatives and their occurrence in groundwater of a tar-contaminated field site. Ecotoxicol. Environ. Saf. 72 (3), 819–827. Philipp, B., Hoff, M., Germa, F., Schink, B., Beimborn, D., Mersch-Sundermann, V., 2007. Biochemical interpretation of quantitative structure–activity relationships (QSAR) for biodegradation of N-heterocycles: a complementary approach to predict biodegradability. Environ. Sci. Technol. 41 (4), 1390–1398. Pereira, W.E., Rostad, C.E., 1986. Investigations of organic contaminants derived from wood-treatment processes in a sand and gravel aquifer near Pensacola, Florida. US Geol. Surv. Water Supply Pap. 2290, 65–80. Peter, A., Steinbach, A., Liedl, R., Ptak, T., Michaelis, W., Teutsch, G., 2004. Assessing microbial degradation of o-xylene at field-scale from the reduction in mass flow rate combined with compound-specific isotope analyses. J. Contam. Hydrol. 71 (1–4), 127–154. Piepenbrink, M., Krüger, M., Ptak, T., Grathwohl, P., 2005. Natural Attenuation of N–S–O Heteroaromatic Compounds in Groundwater: Chemical Analyses, Reactive Transport Experiments and Field Studies. IAHS-AISH Publication, pp. 338–345. Razzaque, M.M., Grathwohl, P., 2008. Predicting organic carbon–water partitioning of hydrophobic organic chemicals in soils and sediments based on water solubility. Water Res. 42 (14), 3775–3780. Reineke, A.K., Goen, T., Preiss, A., Hollender, J., 2007. Quinoline and derivatives at a tar oil contaminated site: hydroxylated products as indicator for natural attenuation? Environ. Sci. Technol. 41 (15), 5314–5322. Reineke, A.K., 2008. Heterocyclic compounds in tar oil contaminated groundwater: occurrence, fate, and indications for natural attenuation. PhD-thesis, Rheinisch-Westfälischen Technischen Hochschule (RWTH) Aachen Rügner, H., Finkel, M., Kaschl, A., Bittens, M., 2006. Application of monitored natural attenuation in contaminated land management. A review and recommended approach for Europe. Environ. Sci. Pol. 9, 568–576. Safinowski, M., Griebler, C., Meckenstock, R.U., 2006. Anaerobic cometabolic transformation of polycyclic and heterocyclic aromatic hydrocarbons: evidence from laboratory and field studies. Environ. Sci. Technol. 40 (13), 4165–4173. Sagner, A., Brinkmann, C., Eisenträger, A., Hinger, G., Hollert, H., Tiehm, A., 2006. Vorkommen und Ökotoxizität von heterozyklischen Kohlenwasserstoffen (NSO-HET). Workshop Themenverbund, 2. KORA, pp. 37–41. Sagner, A., 2009. Vorkommen, Toxizität und mikrobiologischer Abbau heterozyklischer Kohlenwasserstoffe., Band 41. TWZ-Schriftenreihe, Karlsruhe. ISSN 1434-5765. Schulze, S., Tiehm, A., 2004. Assessment of microbial natural attenuation in groundwater polluted with gasworks residues. Water Sci. Technol. 50 (5), 347–353. Seymour, D.T., Verbeek, A.G., Hrudey, S.E., Fedorak, P.M., 1997. Acute toxicity and aqueous solubility of some condensed thiophenes and their microbial metabolites. Environ. Toxicol. Chem. 16, 658–665. Tiehm, A., Müller, A., Alt, S., Jacob, H., Schad, H., Weingran, C., 2008. Development of a groundwater biobarrier for the removal of PAH, BTEX, and heterocyclic hydrocarbons. Water Sci. Technol. 58 (7), 1349–1355. Tiehm, A., Sagner, A., 2008. NSO-Heterocyclic aromatic compounds (NSOHET): Ecotoxicity and biodegradability. Proceedings (CD) of the 10th 194 P. Blum et al. / Journal of Contaminant Hydrology 126 (2011) 181–194 international UFZ-Deltares/TNO Conference on Soil–Water Systems, Milano, Italy, 3–6 June 2008: Theme D, pp. 153–159. Tiehm, A., Schmidt, K., 2007. Methods to evaluate biodegradation at contaminated sites. In: Knödel, K., Lange, G., Voigt, H.-J. (Eds.), Environmental Geology, Handbook of Field Methods and Case Studies. Springer-Verlag, pp. 876–920. Tiehm, A., Schulze, S., 2003. Intrinsic aromatic hydrocarbon biodegradation for groundwater remediation. Oil Gas Sci. Technol. Rev. IFP 58 (4), 449–462. Verschueren, K., 1996. Handbook of Environmental Data on Organic Chemicals. Wiley-Interscience, New York. Zamfirescu, D., Grathwohl, P., 2001. Attenuation rates of specific organic compounds in the groundwater plume at a former gasworks site. J. Contam. Hydrol. 53, 407–427. Zhang, L., Sannes, K., Shusterman, A.J., Hansch, C., 1992. The structure–activity relationship of skin carcinogenicity of aromatic hydrocarbons and heterocycles. Chem. Biol. Interact. 81 (1–2), 149–180.