Environ. Sci. Technol. 2007, 41, 5314-5322
Quinoline and Derivatives at a Tar Oil Contaminated Site: Hydroxylated Products as Indicator for Natural Attenuation? A N N E - K I R S T E N R E I N E K E , †,⊥ THOMAS GO ¨ E N , ‡,| A L F R E D P R E I S S , § A N D J U L I A N E H O L L E N D E R * ,⊥ Institute of Hygiene and Environmental Medicine, RWTH Aachen, Pauwelsstr. 30, D-52074 Aachen, Germany, Institute and Out-patient Clinic for Occupational Medicine, RWTH Aachen, Pauwelsstr. 30, D-52074 Aachen, Germany, Fraunhofer Institute for Toxicology and Experimental Medicine ITEM, Nikolai-Fuchs-Str. 1, D-30625 Hannover, Germany, and Eawag, Swiss Federal Institute of Aquatic Science and Technology, 8600 Du ¨ bendorf, Switzerland
LC-MS-MS analysis of groundwater of a tar oil contaminated site showed the occurrence of the N-heterocycles quinoline and isoquinoline as well as their hydroxylated and hydrogenated metabolites. The concentrations of the hydroxylated compounds, 2(1H)-quinolinone and 1(2H)isoquinolinone, were significantly higher than those of the nonsubstituted parent compounds. Therefore, exclusive quantification of the parent compounds leads to an underestimation of the amount of N-heterocycles present in the groundwater. Microbial degradation experiments of quinoline and isoquinoline with aquifer material of the site as inocculum showed the formation of hydroxylated and hydrogenated products under sulfate-reducing conditions, the prevailing conditions in the field. However, since analyses of seven tar products showed that these compounds are also primary constituents, their detection in groundwater is found to be a nonsufficient indicator for the occurrence of biological natural attenuation processes. Instead, the ratio of hydroxylated to parent compound (Rmetabolite) is proposed as a useful indicator. We found that 65-83% of all groundwater samples showed Rmetabolite for 2(1H)-quinolinone, 1(2H)-isoquinolinone, 3,4-dihydro-2(1H)-quinolinone, and 3,4dihydro-1(2H)-isoquinolinone, which was higher than the highest ratio found in tar products. With respect to the observed partition coefficient between tar oil and water of 3.5 for quinoline and isoquinoline and 0.3 for 2(1H)-quinolinone and 1(2H)-isoquinolinone, the ratio in groundwater would be approximately 10 times higher than the ratio in tar oil. When paying attention to these two parameters, 19-31% of groundwater samples exceed the highest tar oil ratio. This indicates that biological processes take place in the
* Corresponding author phone: +41(0)44 823 54 93; fax: +41(0)448235893; e-mail:
[email protected]. † Institute of Hygiene and Environmental Medicine. ‡ Institute and Out-patient Clinic for Occupational Medicine. § Fraunhofer Institute for Toxicology and Experimental Medicine ITEM. ⊥ Eawag, Swiss Federal Institute of Aquatic Science and Technology. | Current address: Occupational Social and Environmental Medicine and Policlinic, Friedrich-Alexander University, Erlangen-Nu ¨ rnberg, Schillerstr. 29, D-91054 Erlangen, Germany. 5314
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 15, 2007
aquifer of the site and Rmetabolite is an applicable indicator for natural attenuation.
Introduction Tar oils have been produced for various applications such as wood impregnation or as raw material for chemical products. They consist of BTEX, PAHs, phenols, and also heterocyclic compounds. Because of leakage and spillage, the subsurface and groundwater at former gas plants, coke manufacturing, and wood preserving facilities are often contaminated with tar products. These sites are usually large, and a complete decontamination is not possible because of limitations in money and technology, so other strategies, such as natural attenuation, need to be considered. Natural attenuation (NA) at contaminated sites is defined as the reduction in toxicity, mass, and/or mobility of a contaminant without human intervention. This includes biological as well as abiotic processes (e.g., dilution, sorption, volatilization, and precipitation), but only destructive processes such as biodegradation should be considered as an appropriate decontamination strategy (1). Differentiation between the processes at contaminated sites is a very demanding task. The following approaches can be used to obtain information about the biodegradation potential: (1) A detailed mapping of the plume can demonstrate NA due to the mass reduction of dissolved contaminants. However, microbial degradation is not always easly to distinguish from other processes. (2) Consumption of electron acceptors and production of reduced species in a plume can be used to indicate general biodegradative activity. (3) More direct evidence of degradation can be the formation of metabolic products, but metabolites can only be used to document degradation of the parent compounds if they are sufficiently specific. An ideal indicator of in situ biodegradation is a compound that could definitively be related to the metabolism of a specific compound. In addition, the compound should be an intermediate of mineralization rather than a dead-end product (2). (4) Under certain circumstances stable isotope fractionation analysis is another powerful tool to assess biodegradation. Prerequisite is the confirmation of an isotopic fractionation during the microbial degradation process which has to be proven by laboratory studies with pure or mixed cultures. There exists only minor knowledge about the fate of heterocyclic compounds such as high water soluble two ring N-heterocycles. Quinoline (QUI) and 2(1H)-quinolinone (2OH-QUI) as well as their isomeric compounds isoquinoline (ISOQUI) and 1(2H)-isoquinolinone (1-OH-ISOQUI) (Table 1; physicochemical data) have been detected in various aquifers of tar oil contaminated sites (3-7). The hydroxylated compounds 2-OH-QUI and 1-OH-ISOQUI as well as the hydrogenated compound 3,4-dihydro-2(1H)-quinolinone (DIHYDRO-QUI) are known metabolites of the anaerobic degradation of QUI and ISOQUI (4, 8-9). Following the hydroxylation the compounds tautomerize into the more stable keto-form in water (in favor of readability they are termed hydroxylated compounds in the text). The mutagenity of QUI is well-known, while ISOQUI was found to be nonmutagenic (10-11). 2-OH-QUI and 1-OH-ISOQUI are reported as nonmutagenic in the Ames test (10, 12). Data on ecotoxicity showed the ecotoxic potential of QUI and ISOQUI (13, 14), while data for hydroxylated quinoline compounds are absent until now. The detection of 2-OH-QUI, DIHYDRO-QUI and 1-OHISOQUI in the groundwater of tar oil contaminated sites may 10.1021/es070405k CCC: $37.00
2007 American Chemical Society Published on Web 07/03/2007
TABLE 1. Physicochemical Data of QUI and Its Derivatives
a (39) SciFinder. ACD software. Supporting Information)
b
(31) Estimation Programs Interface Suite. c Determined in this study by HPLC-DAD or HPLC-MS-MS (see
indicate biodegradation, because these compounds are reported as metabolites of anaerobic degradation of QUI and ISOQUI (4, 5, 9, 15). However, this interpretation is complicated because the compounds may be present in the original complex mixture of tar oils as postulated elsewhere (5, 16). However, a clear proof for the occurrence of 2-OHQUI, DIHYDRO-QUI, and 1-OH-ISOQUI in tar oil is absent in literature. The aim of this study was to investigate whether hydroxylated QUI and ISOQUI can be used as indicators for natural attenuation. Therefore, the following approaches were carried out: (i) The fate of QUI and ISOQUI as well as hydroxylated metabolites in groundwater was studied in detail at a tar oil contaminated site. (ii) The formation of metabolites of QUI and ISOQUI was investigated with aquifer material of the site in microcosms under sulfate-reducing redox conditions. (iii) The occurrence of 2-OH-QUI, DIHYDRO-QUI, 1-OH-ISOQUI, and DIHYDRO-ISOQUI in several tar products was investigated. The results are discussed and take into account further parameters which may influence the spatial distribution of compounds at the field site such as dilution, volatilization, sorption, dissolution from tar oils in groundwater, and microbial degradation.
Materials and Methods Chemicals. Commercially available compounds of highest purity were used (for more information see the Supporting Information). Tar oil no. 1 (type WEI B), no. 2 (type WEI C), and no. 3 (type WEI C) originated from an active wood-impregnation site. Tar oil no. 4 (type WEI B) and coal tar no. 5 were a gift of Ru ¨ ttgerswerke (Castrop-Rauxel, Germany). Tar oil no. 6 originated from a former impregnation site and was skimmed from groundwater and soil. Coal tar 7 was purchased from Alfa Aesar (Karlsruhe, Germany).
The Castrop-Rauxel Field Site and Collection of Groundwater Samples. Groundwater and aquifer material were collected from a former coal mine with coking plant which is located in Castrop-Rauxel, North Rhine Westphalia, Germany (Figure 1). From the beginning of the 20th century to 1972, the site was used for the production of coke and various tar oil products (benzene, naphthalene, and tar distillate, among others) (17). Then, the production factory was destroyed and the ground was filled with the resulting building rubble. The contamination of groundwater in quaternary deposits as well as in marl rocks results from discharging waste into the ground and because of leakage and destruction during the Second World War. Real data about the date and points of contamination are scarce. The main sources of the contamination of the aquifer are the plants to extinguish coke and coke purification, as well as the benzene plant in the southern part of the field site. In this study the contamination of the quaternary aquifer was investigated: The groundwater flow in quaternary deposits is estimated to be 27-31 m/a as a result from tracer experiments (18). The monitoring of groundwater was repeated in intervals of several months since December 2003 and shows a flow of groundwater in the westnorthwest direction. The resulting plume of PAHs is estimated to be of ∼325 m in length, of 250 m in breadth and of 50 m in depth (17). The texture of the quaternary aquifer is mainly contributed by fine and medium sand (95%), with some silty and gritty samples (Supporting Information: Figure S1). The organic carbon content is
