Trimethylbenzoic Acids As Metabolite Signatures in the Biogeochemical Evolution of an Aquifer Contaminated with Jet Fuel Hydrocarbons

Home , Alkylbenzenes, Tetramethylbenzenes

Journal of Contaminant Hydrology 67 (2003) 177–194 www.elsevier.com/locate/jconhyd

Trimethylbenzoic acids as metabolite signatures in the biogeochemical evolution of an aquifer contaminated with jet fuel hydrocarbons

J.A. Namocatcata,1, J. Fanga,*, M.J. Barcelonaa,2, A.T.O. Quibuyenb, T.A. Abrajano Jr.c

a National Center for Integrated Bioremediation Research and Development, Department of Civil and Environmental Engineering, The University of Michigan, Ann Arbor, MI 48109, USA b Institute of Chemistry, University of the Philippines, Diliman, Quezon City 1101, Philippines c Department of Earth and Environmental Sciences, Rensselaer Polytechnic Institute, Troy, NY 12180, USA Received 24 May 2002; accepted 21 March 2003

Abstract

Evolution of trimethylbenzoic acids in the KC-135 aquifer at the former Wurtsmith Air Force Base (WAFB), Oscoda, MI was examined to determine the functionality of trimethylbenzoic acids as key metabolite signatures in the biogeochemical evolution of an aquifer contaminated with JP-4 fuel hydrocarbons. Changes in the composition of trimethylbenzoic acids and the distribution and concentration profiles exhibited by 2,4,6- and 2,3,5-trimethylbenzoic acids temporally and between multilevel wells reflect processes indicative of an actively evolving contaminant plume. The concentration levels of trimethylbenzoic acids were 3–10 orders higher than their tetramethylbenzene precursors, a condition attributed to slow metabolite turnover under sulfidogenic conditions. The observed degradation of tetramethylbenzenes into trimethylbenzoic acids obviates the use of these alkylbenzenes as non-labile tracers for other degradable aromatic hydrocarbons, but provides rare field evidence on the range of high molecular weight alkylbenzenes and isomeric assemblages amenable to anaerobic degradation in situ. The coupling of actual tetramethylbenzene loss with trimethylbenzoic acid production and the general decline in the concentrations of these compounds

* Corresponding author. Present address: Department of Geological and Atmospheric Sciences, Iowa State University, 3010 Agronomy Building, Ames, IA 50011-3212, USA. Tel.: +1-515-294-6583; fax: +1-515-294- 6049. E-mail address: [email protected] (J. Fang). 1 Present address: Environmental Science Program, University of the Philippines, Diliman, Quezon City 1101, Philippines. 2 Present address: Department of Chemistry, Western Michigan University, 3442 Wood Hall, Kalamazoo, MI 49008, USA.

0169-7722/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0169-7722(03)00067-6 178 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 demonstrate the role of microbially mediated processes in the natural attenuation of hydrocarbons and may be a key indicator in the overall rate of hydrocarbon degradation and the biogeochemical evolution of the KC-135 aquifer. D 2003 Elsevier Science B.V. All rights reserved.

Keywords: Groundwater; Biodegradation; Intrinsic bioremediation; Jet fuel hydrocarbons; Biogeochemistry

1. Introduction

Variability in hydrogeochemical conditions influences the type of microbial niches found in the subsurface (Haack and Bekins, 1988). Microbial processes, in turn, play a key role in subsurface reactions (McMahon and Chapelle, 1991) and are largely responsible for the observed biogeochemical changes occurring in aquifers contaminated with fuel hydrocarbons (Bennett et al., 1993; Baedecker et al., 1993; Eganhouse et al., 1993; Kelly et al., 1997). The surfeit organic carbon sources in fuel-contaminated aquifers typically deplete the dissolved oxygen supply (Lovley, 1997; Anderson et al., 1998) and poise the system for alternate anaerobic processes. The energetics of these processes suggests that utilization of alternate electron acceptors proceeds sequentially from nitrate reduction followed by iron reduction and so forth (Champ et al., 1979; Baedecker et al., 1993; Lovley and Chapelle, 1995). This simplified notion of spatial partitioning of redox processes occurs largely on a macro-scale, considering that concurrent redox reactions may also occur at the micro-scale as was shown for sulfate reduction occurring in the oxygenated surface regions of some microbial mats (Canfield and Des Marais, 1991; Fru¨nd and Cohen, 1991). Such concurrent redox processes are also observed in some subsurface investigations, albeit in strictly anoxic conditions as a function of electron donor availability in the case of competing sulfidogenic and methanogenic processes (Vroblesky et al., 1996), complexity of contaminant plumes (Albrechtsen et al., 1999), or due to spatial variations within the hydrogeochemical framework (Ludvigsen et al., 1998). In general, redox processes in contaminant plumes affect biodegradation rates of aromatic hydrocarbons (Vroblesky and Chapelle, 1994). The shift from aerobic to anaerobic conditions is characterized by the disappearance of 2 oxidants such as dissolved O2,NO3 , and SO4 and the subsequent evolution of reduced species not previously found that is coupled with the degradation of hydrocarbons (Bennett et al., 1993; Baedecker et al., 1993; Borden et al., 1994; Prommer et al., 1999). These species represent varied redox regimes predominant at a particular time and space during the geochemical evolution of the aquifer. These redox regimes are also associated with changes in microbial community structure. For example, shift to ferridogenic conditions is accompanied by increase in population of iron bacteria responsible for the solubilization of ferric hydroxides that is coupled with the degradation of organic compounds (Lovley et al., 1989; Lovley, 1991; Anderson et al., 1998). Similarly, shift to sulfidogenic conditions stimulated the activity of sulfate-reducing bacteria that degrade benzene (Anderson and Lovley, 2000), substantiated by increase in cyclopropyl fatty acids (Fang et al., 1997; Fang and Barcelona, 1998), biomarkers of J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 179 sulfate-reducing bacteria (Dowling et al., 1986; Vestal and White, 1989; Findlay and Dobbs, 1993). The extent of microbial role in the degradation of organic contaminants in situ is difficult to quantify (Reinhard et al., 1997), but most studies suggest that intrinsic bioremediation is an important process in the natural attenuation of aromatic hydrocarbons in the subsurface (Baedecker et al., 1993; Eganhouse et al., 1993; Eganhouse et al., 1996; Curtis and Lammey, 1998; Gieg et al., 1999). No single factor is considered sufficient to demonstrate the occurrence of intrinsic bioremediation in the field (NRC, 1993). Instead, multiple strategies are required to show that microbial attenuation occurs. However, these strategies are constrained by the inherently complex hydrogeochemical settings of the aquifer and the compositional heterogeneity of the dissolved contaminant mixtures. While typical groundwater geochemical signatures demonstrate the occurrence of intrinsic bioremediation processes (Borden et al., 1994; McNab, 1999), production of metabolic intermediates in the contaminant plume concurrent with hydrocarbon losses provides unequivocal evidence that hydrocarbons are being actively degraded in situ (Fang et al., 1997; Cozzarelli et al., 1990). Organic acid metabolites such as aromatic acids typically observed in the anoxic zone of the contaminant plume are considered biogeochemical indicators in the degradation of aromatic hydrocarbons in situ (Fang et al., 1997; Cozzarelli et al., 1990; Beller, 2000). These compounds were neither observed in pristine areas of the aquifer nor were they components of the original hydrocarbon mass (Cozzarelli et al., 1990, 1994, 1995). Because aromatic acids persist in the anoxic plume longer than alicylic or aliphatic acids, aromatic acids are ideal metabolite signatures to document the microbially mediated, contaminant breakdown reactions, processes which characterize the biogeochemical evolution of the aquifer. Reports on trimethylbenzoic acids as products of anaerobic metabolism of tetramethylbenzenes are exceedingly rare (Beller, 2000). This paper investigates the role of trimethylbenzoic acids as potential metabolite signatures in the biogeochemical evolution of KC- 135 aquifer at the former Wurtsmith Air Force Base (WAFB), Oscoda, MI with a history of JP-4 fuel contamination. Trimethylbenzoic acid production is traced to the degradation of tetramethylbenzene precursors (Cozzarelli et al., 1990). All three tetramethylbenzene isomers—prenitene (1,2,3,4-), izodurene (1,2,3,5-), and durene (1,2,4,5-) found active as genotoxic compounds (Janik-Spiechowicz and Wyszynska, 1999) were detected in this aquifer. Although toluene, xylenes, ethylbenzene, and trimethylbenzenes are the major hydrocarbon components of the contaminant plume in KC-135 with tetramethylbenzenes comprising only a minor fraction, trimethylbenzoic acids are the dominant metabolic intermediates. This paper also reports the detection of additional trimethylbenzoic acids with unknown methylation, which were not documented in other contaminant plumes.

2. Field and laboratory methods

This study was part of the multidisciplinary investigation at the KC-135 Crash Site in the decommissioned Wurtsmith Air Force Base (WAFB) in Oscoda, MI. The aquifer was contaminated with JP-4 fuel resulting from the tragic crash of a KC-135 fuel tanker in October 1988, spilling at least 3000 gal of jet fuel. Unknown quantity of the fuel burned as a result of the crash, while the remainder percolated into the ground. This area was 180 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 previously uncontaminated and represents a rare example of a site with a single known contaminant source.

2.1. Site description

The KC-135 Crash Site is located near the main runway of WAFB (Fig. 1). Previous site characterization indicates that the aquifer is unconfined and consists of two distinct units: medium to coarse-grained sand/gravel underlain by a low permeability lacustrine clay layer. The water table is approximately 3.4 m below land surface (bls). WAFB is bounded by a line of 24-m-high bluffs in the west, by the Van Etten Lake to the northeast, and by the Au Sable River to the south, which flows eastward and discharges into Lake Huron. Recent groundwater level measurements and historic data (USGS, 1990) indicate that groundwater flow at the KC-135 Crash Site is consistent with the regional groundwater flow direction, that is southeast toward the Au Sable River, with a hydraulic gradient of 0.0024 m/m. The hydrocarbon contaminant plume (Fig. 1) is approximately 0.45 ha in coverage at the surface, extending to the southeast from the point where the plane crash occurred. Site geology and hydrogeological parameters were described further in Fang and Barcelona (1998).

2.2. Field sampling

Multilevel wells (ML-13, ML-14, and ML-15) (Fig. 2) were installed in November 1995 at the site using either a Geoprobe piston corer (Geoprobe Systems, Salina, KS) or a hollow stem auger without lubricants. ML-16, f 10 m downgradient from ML-14 was installed in June 1998. The wells (2.5 cm internal diameter) were emplaced above and below the water table with screen lengths of 0.3 m. Depth-discrete changes in soil gas and ground water composition were evaluated using drive-point techniques at upgradient, source, and downgradient locations. Quarterly groundwater samples for the analysis of volatile organic carbons (VOCs) and metabolites were collected from each multilevel well using a peristaltic pump (Master Flex

Fig. 1. Field site showing the KC-135 Crash Site at the former Wurtsmith Airforce Base (WAFB) in Oscoda, MI. J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 181

Fig. 2. Reconstructed multilevel well installations at KC-135 aquifer showing parallel screened depths along upgradient, source, and downgradient locations. 5 = Screened interval, depth descriptors not to scale; z = water table at 3.4 m (bls).

Speed Controller, Cole Parmer, IL). Before sampling, each well was purged to remove sand grains and then connected to a calibrated flow cell (Purge Saver, QED Environmental Systems, MI) for simultaneous analysis of dissolved oxygen (DO), pH, conductivity, temperature, and oxidation–reduction potential. Once the readings had stabilized (Barce- lona et al., 1995), the flow cell was disconnected and sampling began immediately. Samples for VOC analyses were collected in 40-ml VOA vials (fitted with Teflon-lined screw caps) with no headspace. VOC samples were immediately preserved with 1 ml of 40% NaHSO4 solution. Metabolite samples were collected in 1.0 l amber glass containers and preserved with KOH. Samples for nitrate, sulfate, and CH4 (without headspace) determination were also collected in 40-ml VOC vials. All samples were immediately stored on ice. After thorough documentation, samples were shipped to NCIBRD Laboratory at Ann Arbor, MI and stored at 5 jC for subsequent processing.

2.3. Chemical analysis

Volatile organic compounds in ground water were analyzed on an OI Analytical 4552 purge and trap autosampler (College Station, TX) connected to a Hewlett Packard 5890 Series II gas chromatograph (GC) and HP 5972 mass selective detector. Groundwater samples were spiked with internal standards o-xylene-d10 and naphthalene-d8 before analysis. Separation of individual components was accomplished using a DB-624 column: 60 m 0.25 mm i.d. (film thickness df = 1.8 mm) (J&W Scientific, Folsom, CA). Injector and detector temperatures were both set at 250 jC. The oven temperature program began at 50 jC and held for 6 min and ramped at 15 jC min 1 to a final temperature of 240 jC. The GC-MSD was calibrated by injection of calibration standards at five different concentration 182 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 levels (10, 50, 100, 200, and 400 mg/l). Response factors were obtained for all compounds including internal standards. Compounds were identified based on relative retention time and mass spectra. Concentrations of each compound were calculated using the internal standard method, and are reported as nanograms per milliliter of ground water. Organic acids were extracted using the classical liquid–liquid extraction technique. One hundred milliliters of field samples was delivered into a silanized separatory funnel, spiked with a surrogate standard (o-methylbenzoic acid-d7), acidified to pH 2.0, saturated with NaCl, swirled to homogeneity, and extracted four times with acid-free dichloro- methane. The extract was then derivatized using pentafluorobenzylbromide with acetoni- trile as the reaction solvent and potassium carbonate as the base catalyst. One microliter of the derivative was injected in splitless mode into the HP 5980 Series II Gas Chromato- graph interfaced to HP 5972 Mass Selective Detector. Chromatographic separation was achieved on HP-1 (60 m 0.25 mm 0.25 mm film thickness) capillary column. Carrier gas is high-purity grade helium at a linear velocity of 24.3 cm s 1 connected to a moisture trap and oxygen scrubber. Injector and detector temperatures were set at 280 and 300 jC, respectively. The oven temperature program began at 50 jC and ramped at three levels to 110 jCat10jC min 1, 225 jCat3jC min 1, and 300 jCat30jC min 1, with a total run time of 53.0 min. Linearity was achieved via a five-point calibration prepared as follows: 50, 100, 250, 500, and 1000 ng/ml. Calibration graphs and equations were constructed for each compound using the default data analyzer (HP Chemstation). Response factors for each compound were obtained relative to those of the internal standard (decafluorobiphenyl). Reproducibility was examined by spiking deionized water and field samples with 50 and 100 ng/ml of the calibration standards at n = 11. Selected ion monitoring (SIM) was used for detection and quantitation of analytes, by setting the base peak as the quantitation ion and the secondary peak and another fragment ion as the qualifier peaks. Chromatograms were collected under time-scheduled SIM using acquisition windows tailored for each group of compounds. Peaks were confirmed by injecting a pure standard, coinjection, matching of retention times, and mass spectra interpretation. Field determinations of dissolved oxygen, alkalinity (as HCO3 ), sulfide, and soluble iron were made using CHEMetrics Kit (CHEMetrics, Calverton, VA) or Hach Kit (Hach). 2+ 2 Laboratory analyses of NO3 ,Ca , and SO4 in ground water were carried out using a Dionex DX-100 Ion Chromatograph (IC, Dionex, Sunnyvale, CA) equipped with an AS40 autosampler. A Dionex AS-14S column (250 4 mm) with a guard column (50 4 mm) was used for the analysis. Each sample of 2.5 ml was injected into the IC. Results were obtained as average of duplicate analyses.

3. Results and discussion

3.1. Evolution of trimethylbenzoic acids in KC-135 aquifer

We detected five trimethylbenzoic acid isomers in the anoxic zone of the jet fuel plume out of six potential metabolic intermediates that can be traced to the anaerobic degradation of tetramethylbenzene precursors (Fig. 3)—two were identified as 2,4,6- and 2,3,5- J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 183

Fig. 3. Tetramethylbenzene isomers detected in KC-135 aquifer and their potential degradation products via methyl oxidation. 184 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 trimethylbenzoic acids. The selective ion recordings (Fig. 4) were characterized by the + base peak m/z 163 [C10H11O2] indicative of the original analyte (MW = 164) and the + pentafluorotropylium ion m/z 181 (secondary peak) [C7H2F5], characteristic of the cleaved derivative moiety of the pentafluorobenzyl ester (MW = 344) (Cline et al., 1990; Brill et al., 1991). Due to lack of authentic standards and available data on retention times, the assignment of isomers for the unknown trimethylbenzoic acids was not possible. These

Fig. 4. Selective ion recordings highlighting the trimethylbenzoic acid range (as pentafluorobenzyl esters) from samples collected in ML-14 ! 4.79 m on (a) Jun-96, (b) Feb-98, and (c) Mar-99; 246-TMB = 2,4,6- trimethylbenzoic acid; 2,3,5-TMB = 2,3,5-trimethylbenzoic acid; *= unknown methylation; **= coeluted with an unknown compound. Note the decline in ion abundance of m/z 163 = base peak, and m/z 181 = secondary peak. [M+] < 8% of base peak (not shown). J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 185 compounds are addition to the range of trimethylbenzoic acids previously documented in the subsurface contaminated with diesel (Cozzarelli et al., 1990), gasoline (Cozzarelli et al., 1995), and jet fuel (Gieg et al., 1999). Detection of five trimethylbenzoic acid isomers, presumably from the three tetramethylbenzene isomers found in this aquifer provides rare field evidence on the range of higher alkylbenzenes and isomeric assemblages amenable to anaerobic degradation in situ and demonstrates the potential degradability of higher alkylbenzenes under dynamic anaerobic field conditions. Both 2,4,6- and 2,3,5-trimethylbenzoic acids are putative metabolites from the degradation of 1,2,3,5-tetramethylbenzene (Fig. 3) (Cozzarelli et al., 1990). In anaerobic microcosm studies using 1,2,3,4- and 1,2,3,5-tetramethylbenzenes as substrates, the 2,3,6- and 2,4,6-trimethylbenzoic acids were identified as oxidized compounds (Cozzar- elli et al., 1994). Recently, 2,4,5-trimethylbenzoic acid was reported as an intermediary metabolite from the administration of 1,2,4,5-tetramethylbenzene in rats (Ligocka et al., 2001). Based on these studies and the unique and unequivocal biochemical relationship of trimethylbenzoic acids to parent hydrocarbons, we suggest that 2,4,6- and 2,3,5- trimethylbenzoic acids documented in this aquifer are breakdown products from the degradation of 1,2,3,5-tetramethylbenzene. The unidentified isomers may have originated from the oxidation of 1,2,3,4- and 1,2,4,5-tetramethylbenzene as the oxidative pathways indicated (Fig. 3). The mechanisms that generate trimethylbenzoic acids in situ are unknown but these organic acids may be generated under aerobic conditions (Bayly and Barbour, 1984; Gibson and Subramanian, 1984; Smith, 1990) and under anaerobic processes where oxygen is sourced from the water molecule (Vogel and Grbic´-Galic´, 1986). Under hypoxic conditions, Kukor and Olsen (1996) reported functional dioxygenases that catalyze the degradation of toluene to its metabolic intermediates. The anaerobic degradation of tetramethylbenzene may be analogous to toluene (Grbic´-Galic´ and Vogel, 1987; Schocher et al., 1991; Altenschmidt and Fuchs, 1992; Seyfried et al., 1994), involving hydroxylation of one methyl group to yield a trimethylbenzyl alcohol, followed by oxidation to a trimethylbenzaldehyde and subsequently to a trimethylbenzoic compound. In mammalian metabolism of 1,2,4,5-tetramethylbenzene, two routes are pro- posed—main route is oxidation of one methyl group leading to the formation of 2,4,5- trimethylbenzoic acid; the second route is conjugation with reduced glutathione (Ligocka et al., 2001).

3.2. Trimethylbenzoic acids vs. tetramethylbenzenes

Under field conditions, one caveat in relating metabolic intermediates with their hydrocarbon precursors is the lack of mass balance between these compounds. One plausible reason is that a specific product of biotransformation may be generated by several oxidation pathways, for example, benzoate (Beller, 2000). Broadly, this may also be attributed to the general complexity of contaminant mixtures and hydrogeochemical heterogeneity of the aquifer. In KC-135, the actual amount of jet fuel that percolated into the aquifer is not known, including the residual hydrocarbons that are possibly trapped in the vadose zone, hence, a mass-balance approach involving metabolic intermediates and parent substrates may not be plausible to explain in situ bioremediation. 186 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194

Analysis of KC-135 samples suggests that there is no mass correlation between 1,2,3,5- tetramethylbenzene and 2,4,6-/2,3,5-trimethylbenzoic acids as the total concentrations of the latter were 3–10 orders of magnitude higher than the parent substrates (Fig. 5). The ‘lack of mass balance’ probably results from accumulation of trimethylbenzoic acids, where metabolite production exceeds mineralization rate. In a deep aquifer of the Atlantic Coastal Plain, McMahon and Chapelle (1991) observed that accumulation of acetate and formate from microbial fermentation of organic matter was due to microbial fermentation outpacing microbial respiration. The mineralization of trimethylbenzoic acids to CO2 and H2O may also be affected by increase in acidity and inherent toxicity of aromatic acid metabolites (Long and Aelion, 1999), the terminal electron-accepting processes involved and the position of the methyl substituent on the aromatic ring (Eganhouse et al., 1996; Berry et al., 1987; Kuhn et al., 1988). Phospholipid ester-linked fatty acid analysis of the microbial community structure at KC-135 has indicated the shift to sulfate-reducing condition (Fang et al., 1997; Fang and Barcelona, 1998). This TEAP condition is a reasonable predictor of metabolite persistence in the anoxic plume.

3.3. Variability in trimethylbenzoic acid production

Fig. 6 presents the trimethylbenzoic acid data in the multilevel sampling wells (ML-14, ML-16) located within the source zone of the contaminant plume (Figs. 1 and 2). Trimethylbenzoic acids were neither detected upgradient nor downgradient of the plume. In general, the concentration profiles of the five trimethylbenzoic acids dominated by both the 2,4,6- and 2,3,5-isomers indicated a declining trend. At the core of the anoxic plume (ML-14 ! 4.79 m), which is the well closest to the center of the plume body (Figs. 1 and 2), total trimethylbenzoic acid production was f 9.0 mg/l in Jun-96. By Feb-98, the concentration was reduced by 90%. Between Feb-98 and Mar-99, production of trimethylbenzoic acids was characterized by short-term variations that reflect hydrogeochemical changes in the aquifer. For example, concentrations of trimethylbenzoic acids increased in Sep-98, approaching the Feb-98 levels. This upsurge in concentration coincided with fluctuation of the water table.

Fig. 5. Concentration profile of 1,2,3,5-tetramethylbenzene relative to 2,4,6-trimethylbenzoic acid and 2,3,5- trimethylbenzoic acid in ML-14 ! 4.79 m. *= 10. J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 187

Fig. 6. Vertical profile of trimethylbenzoic acids (ng/ml) in (a) ML-14 and (b) ML-16, within the source zone of the contaminant plume. *= data for ML-14 ! 4.79 m only, where 2,4,6-/2,3,5-isomers = 4; a = unknown methylation, b = coeluted with an unknown compound. Depth descriptors not to scale.

In ML-14, trimethylbenzoic acids were generally observed from 4.14 to 5.24 m (bls) but maximum at 4.79 m depth. In ML-16, which is immediately downgradient of ML-14 (Fig. 2), trimethylbenzoic acids were noted only in 4.34 m (bls) and in one instance at 4.94 m depth. The five trimethylbenzoic acid isomers were only detected between Jun-96 and Feb-98. From Jun-98 to Mar-99, only 2,4,6- and 2,3,5-trimethylbenzoic acids remained in the contaminant plume. Because of water level fluctuation, no samples were collected in ML-14 ! 4.14 m after Jun-98. Variations in the composition and distribution of trimethylbenzoic acids reflect the dynamics of anaerobic processes and the chemical and physical heterogeneity of the aquifer that can be introduced by stochastic variations. The most abundant tetramethylbenzene isomer in the aquifer was prenitene (1,2,3,4-), followed by izodurene (1,2,3,5-), and durene (1,2,4,5-) in the shallow depths of ML-14 but maximum in 4.79 m (bls) (Fig. 7). The concentrations of prenitene and durene at 188 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194

Fig. 7. Distribution of tetramethylbenzene isomers in the shallow depths of ML-14 in the source zone of the contaminant plume. Depth descriptors not to scale, TeMB = tetramethylbenzene, z = water table at f 3.4 m (bls).

4.79 m depth suggest decreasing trends, while that of izodurene remained unchanged in Sep-98 relative to the Jun-98 value. However, the concentration levels of 2,4,6- and 2,3,5-trimethylbenzoic acid, putative metabolites of izodurene degradation increased in Sep-98 (Fig. 5). The relatively unchanged concentration of izodurene in Sep-98 may be expected from solubilization of residual phases trapped in the vadose zone by infiltrating J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 189 rainwater or snowmelt. This hydrogeologic phenomenon may have increased tetramethylbenzene concentrations available for microbial metabolism. In general, the concentration profile of izodurene in ML-14 ! 4.79 m appeared to coincide with 2,4,6- and 2,3,5-trimethylbenzoic acids in Sep-98 (Fig. 5); the concentrations increased as the water level recedes.

3.4. Biogeochemical implications

The concentration profiles of trimethylbenzoic acids across spatial and temporal gradients relative to their tetramethylbenzene precursors elucidate the importance of intrinsic bioremediation as an important fate process in KC-135 aquifer. Evolution, accumulation, and mineralization of trimethylbenzoic acids are processes characteristic of a system that has evolved geochemically in response to heavy influx of potentially oxidizable substrates at a scale magnifying the role of microbial processes in contaminant mass removal. The evolved state of the aquifer is characterized by depressed levels of 2 2+ dissolved oxygen (DO), NO3 ,SO4 (Fig. 8a) but elevated concentrations of Fe ,CH4, H2S in the contaminant plume (Fang et al., 1997). Concurrently, the pH decreased resulting from the release of CO2 from hydrocarbon degradation accompanied by increase in alkalinity and concentration of Ca2+ and dissolved inorganic carbon (DIC) (Fig. 8b). Although geochemical markers cited above demonstrate microbial degradation (Bor- den et al., 1994; McNab, 1999), aromatic acid metabolites are considered specific biogeochemical indicators of such process (Fang et al., 1997; Cozzarelli et al., 1990; Beller, 2000) since evolution of metabolites suggests breakdown of aromatic hydrocarbons. Microbial degradation initiates a series of biogeochemical reactions (Baedecker et al., 1993; Bennett et al., 1993; Eganhouse et al., 1993) crucial to the overall evolution of the aquifer. The evolution of KC-135 aquifer into a state essentially dominated by trimethylbenzoic acids suggests that the contaminant plume is being actively degraded. Chen et al. (1997), in its preliminary modeling investigation of the KC-135 aquifer, reported that the BTEX plume appears to be shrinking, an indication that biogeochemical processes limit the transport of contaminant mass downgradient to areas of biological significance. In the early stage of plume evolution, metabolic intermediates from BTEX degradation dominated the metabolite pool (Fang and Barcelona, 2000). Following the decline of BTEX contaminant mass and the evolving aquifer geochemistry, the concentrations of methylbenzoic and dimethylbenzoic acids decreased and varied considerably. Since there are no discernible trends that could be derived from the breakdown products of BTEX degradation owing to their variability and volatility in this aquifer, trimethylbenzoic acids are potential metabolite signatures of biogeochemical processes since they are detected against variable hydrogeochemical conditions. The concentration levels of trimethylbenzoic acids can also be directly related to the concentration and occurrence of other aromatic acids. When trimethylbenzoic acids were observed at concentrations greater than 2.0 mg/l, aromatic acids such as benzoic acid, tolylacetic acids, toluic acids, and dimethylbenzoic acids (data not shown) were detected. When the concentration falls below 100 ng/ml, most low molecular weight aromatic acids were no longer detectable 190 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194

Fig. 8. Relationship of trimethylbenzoic acid production with S Tetramethylbenzenes and (a) DO, NO3 , and 2 2+ SO4 and (b) DIC, alkalinity, and Ca in ML-14 along a vertical profile. Metabolite production is restricted in the surficial, anoxic layers of the aquifer where metabolite fluxes may be influenced by a number of stochastic variations. Depth descriptors not to scale, z = water table at f 3.4 m (bls). J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 191 and only trimethylbenzoic acids persist. This suggests that trimethylbenzoic acids are compounds that could also serve as useful tracers to monitor the degradation or occurrence of other organic acid metabolites in this aquifer.

4. Conclusion

This investigation has demonstrated the utility of using trimethylbenzoic acids as metabolite signatures to document the production, accumulation, and mineralization of organic acid metabolites, key processes in the biogeochemical evolution of the KC-135 aquifer contaminated with jet fuel. Spatial and temporal patterns exhibited by trimethylbenzoic acids demonstrate the dynamics of anaerobic processes and hydrogeochemical heterogeneity that can be introduced by stochastic variations. Decline in the concentration of trimethylbenzoic acids and tetramethylbenzenes suggests that the contaminant plume in general is shrinking and being actively degraded and that intrinsic bioremediation is the predominant fate process in this aquifer. This work also represents field corroboration of recent reports on some aromatic acids as specific biogeochemical indicators of in situ alkylbenzene degradation in anoxic environments.

Acknowledgements

This study was funded by a cooperative agreement from the US Environmental Protection Agency and the Department of Defense-Strategic Environmental Research and Development Program (DoD-SERDP) with the University of Michigan National Center for Integrated Bioremediation Research and Development (UM-NCIBRD). The authors thank Charles L. Major, Tony Brown, Dave Jaglowski, and Michael Gerdenich for assistance in field sampling and Briana Sye, Sandy Homola, Tim Baker, Nitin Barad, and Hongming Chen for laboratory assistance. This paper benefited greatly from the reviews of Dr. Peter Adriaens (University of Michigan) and Dr. Paul Bradley (United States Geological Survey) and editorial handling of Dr. David Lerner (University of Sheffield). Scholarship support from Engineering and Science Education Project-Department of Science and Technology (DOST-ESEP), Philippines through Dr. Estrella Alabastro and Dr. Ida Dalmacio of the Philippine Council for Advance Science and Technology Research and Development (PCASTRD) are gratefully acknowledged.

References

Albrechtsen, H.-J., Bjerg, P.L., Ludvigsen, L., Ru¨gge, K., Christensen, T.H., 1999. An anaerobic field injection experiment in a landfill leachate plume, Grinsted, Denmark-2. Deduction of anaerobic (methanogenic, sulfate-, and Fe(III)-reducing) redox conditions. Water Resour. Res. 35, 1247–1256. Altenschmidt, U., Fuchs, G., 1992. Anaerobic toluene oxidation to benzyl alcohol and benzaldehyde in a denitrifying Pseudomonas strain. J. Bacteriol. 174, 4860–4862. Anderson, R.T., Lovley, D.R., 2000. Anaerobic bioremediation of benzene under sulfate-reducing conditions in a petroleum-contaminated aquifer. Environ. Sci. Technol. 34, 2261–2266. 192 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194

Anderson, R.T., Rooney-Varga, J.N., Gaw, C.V., Lovley, D.R., 1998. Anaerobic oxidation in the Fe(III) reduction zone of petroleum-contaminated aquifers. Environ. Sci. Technol. 32, 122–129. Baedecker, M.J., Cozzarelli, I.M., Eganhouse, R.P., 1993. Crude oil in a shallow sand and gravel aquifer: III. Biogeochemical reactions and mass balance modeling in anoxic groundwater. Appl. Geochem. 8, 569–586. Barcelona, M.J., Wehrmann, H., Varljen, M., 1995. Reproducible well-purging procedures and stabilization criteria. Ground Water 32, 12–22. Bayly, R.C., Barbour, M.G., 1984. The degradation of aromatic hydrocarbons by the meta and gentisate pathways. In: Gibson, D.T. (Ed.), Microbial Degradation of Aromatic Compounds. Marcel Dekker, New York, NY, pp. 253–294. Beller, H.R., 2000. Metabolic indicators for detecting in situ anaerobic alkylbenzene degradation. Biodegradation 11, 125–139. Bennett, P., Siegel, D.E., Baedecker, M.J., Hult, M.F., 1993. Crude oil in a shallow sand and gravel aquifer: I. Hydrology and inorganic geochemistry. Appl. Geochem. 8, 529–549. Berry, D.F., Francis, A.J., Bollag, J.M., 1987. Microbial metabolism of homocyclic and heterocyclic aromatic compounds under anaerobic conditions. Microbiol. Rev. 51, 43–59. Borden, R.C., Gomez, C.A., Becker, M.T., 1994. Geochemical indicators of intrinsic bioremediation. Ground Water 33, 180–189. Brill, J.H., Narayanan, B.A., McCormick, J.P., 1991. Selective determination of pentafluorobenzyl ester derivatives of carboxylic acids by GC using microwave plasma and mass selective detection. Appl. Spectrosc. 45, 1617–1620. Canfield, D.E., Des Marais, D.J., 1991. Aerobic sulfate reduction in microbial mats. Science 251, 1471–1473. Champ, D.R., Gulens, J., Jackson, R.E., 1979. Oxidation–reduction sequences in groundwater flow systems. Can. J. Earth Sci. 16, 12–23. Chen, Y.-M., Abriola, L.M., Barcelona, M.J., 1997. Simulation of intrinsic bioremediation processes at Wurt- smith Air Force Base, Michigan. In: Alleman, B.C., Leeson, A. (Chairs), In situ and On-Site Bioremediation, vol. 1. Battelle Press, Columbus, OH, pp. 55–60. Cline, R.E., Todd, G.D., Ashley, D.L., Grainger, J., McGraw, J.M., Alley, C.C., Hill Jr., R.H., 1990. Gas chromatographic and spectral properties of pentafluorobenzyl derivatives of 2,4-dichlorophenoxy acetic acid and phenolic pesticides and metabolites. J. Chromatogr. Sci. 28, 167–172. Cozzarelli, I.M., Eganhouse, R.P., Baedecker, M.J., 1990. Transformation of monoaromatic hydrocarbons to organic acids in anoxic groundwater. Environ. Geol. Water Sci. 16, 135–141. Cozzarelli, I.M., Baedecker, M.J., Eganhouse, R.P., Goerlitz, D.F., 1994. The geochemical evolution of low- molecular weight organic acids derived from the degradation of petroleum contaminants in groundwater. Geochim. Cosmochim. Acta 58, 863–877. Cozzarelli, I.M., Herman, J.S., Baedecker, M.J., 1995. Fate of microbial metabolites of hydrocarbons in a coastal plain aquifer: the role of electron acceptors. Environ. Sci. Technol. 29, 458–469. Curtis, F., Lammey, J., 1998. Intrinsic remediation of a diesel fuel plume in Goose Bay, Labrador, Canada. Environ. Pollut. 103, 203–210. Dowling, N.J., Widdel, F., White, D.C., 1986. Analysis of phospholipid ester-linked fatty acid biomarkers of acetate-oxidizing sulfate reducers and other sulfide-forming bacteria. J. Gen. Microbiol. 132, 1815–1825. Eganhouse, R.P., Baedecker, M.J., Cozzarelli, I.M., Aiken, G.R., Thorn, K.A., Dorsey, T.F., 1993. Crude oil in a shallow sand and gravel aquifer: II. Organic geochemistry. Appl. Geochem. 8, 551–567. Eganhouse, R.P., Dorsey, T.F., Phinney, C.S., Westcott, A.M., 1996. Processes affecting the fate of monoaromatic hydrocarbons in an aquifer contaminated by crude oil. Environ. Sci. Technol. 30, 3304–3312. Fang, J., Barcelona, M.J., 1998. Biogeochemical evidence for microbial community change in a jet fuel hydrocarbons-contaminated aquifer. J. Org. Geochem. 29, 899–907. Fang, J., Barcelona, M.J., 2000. Evolution of aromatic hydrocarbon and metabolic intermediate plumes in a shallow sand aquifer contaminated with jet fuel. Proceedings of the 2000 Petroleum Hydrocarbons and Organic Chemicals in Ground Water: Prevention, Detection, and Remediation, November 17–19, 1999, Houston, TX, pp. 51–56. Fang, J., Barcelona, M.J., West, C., 1997. The use of aromatic acids and phospholipid ester-linked fatty acids for delineation of processes affecting an aquifer contaminated with JP-4 fuel. In: Eganhouse, R.P. (Ed.), Molec- ular Markers in Environmental Geochemistry. American Chemical Society, Washington, D.C., pp. 899–907. J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194 193

Findlay, R.H., Dobbs, F.C., 1993. Quantitative description of microbial communities using lipid analysis. In: Kemp, P.F., Sherr, B.F., Sherr, E.B., Cole, J.J. (Eds.), Aquatic Microbial Ecology. Lewis Publishers, Boca Raton, FL, pp. 271–284. Fru¨nd, C., Cohen, Y., 1991. Diurnal cycles of sulfate reduction under oxic conditions in cyanobacterial mats. Appl. Environ. Microbiol. 58, 70–77. Gibson, D.T., Subramanian, V., 1984. Microbial degradation of aromatic hydrocarbons. In: Gibson, D.T. (Ed.), Microbial Degradation of Aromatic Compounds. Marcel Dekker, New York, NY, pp. 181–252. Gieg, L.M., Kolhatkar, R.V., Mcinerney, M.J., Tanner, R.S., Harris Jr., S.H., Sublette, K.L., Suflita, J.M., 1999. Intrinsic bioremediation of petroleum hydrocarbons in a gas condensate-contaminated aquifer. Environ. Sci. Technol. 33, 2550–2560. Grbic´-Galic´, D., Vogel, T.M., 1987. Transformation of toluene and benzene by mixed methanogenic cultures. Appl. Environ. Microbiol. 53, 254–260. Haack, S.K., Bekins, B.A., 1988. Microbial populations in contaminant plumes. Hydrogeol. J. 8, 63–76. Janik-Spiechowicz, E., Wyszynska, K., 1999. Genotoxicity evaluation of tetramethylbenzenes. Mutat. Res. 439, 69–75. Kelly, W.R., Herman, J.S., Mills, A.L., 1997. The geochemical effects of benzene, toluene, and xylene (BTX) degradation. Appl. Geochem. 12, 291–303. Kuhn, E.P., Zeyer, J., Schwarzenbach, R.P., 1988. Anaerobic degradation of alkylated benzenes in denitrifying laboratory columns. Appl. Environ. Microbiol. 54, 490–496. Kukor, J.J., Olsen, R.H., 1996. Catechol 2,3-dioxygenases functional in oxygen-limited (hypoxic) environments. Appl. Environ. Microbiol. 62, 1728–1740. Ligocka, D., Sapota, A., Rydzynski, K., 2001. The disposition and metabolism of durene (1,2,4,5-tetramethylbenzene) in rats. Arch. Toxicol. 74, 673–679. Long, S.C., Aelion, C.M., 1999. Metabolite formation and toxicity measurements in evaluating bioremediation of a jet-fuel contaminated aquifer. Appl. Biochem. Biotechnol. 76, 79–97. Lovley, D.R., 1991. Dissimilatory Fe(III) and Mn(IV) reduction. Microbiol. Rev. 55, 259–287. Lovley, D.R., 1997. Potential for anaerobic bioremediation of BTEX in petroleum-contaminated aquifers. J. Ind. Microbiol. Biotechnol. 18, 75–81. Lovley, D.R., Chapelle, F.H., 1995. Deep subsurface microbial processes. Rev. Geophys. 8, 77–88. Lovley, D.R., Baedecker, M.J., Lonergan, D.J., Cozzarelli, I.M., Phillips, E.J., Siegel, D.I., 1989. Oxidation of aromatic contaminants coupled to microbial iron reduction. Nature 339, 297–300. Ludvigsen, L., Albrechtsen, H.-J., Heron, G., Bjerg, P.L., Christensen, T.H., 1998. Anaerobic microbial redox processes in a landfill leachate contaminated aquifer (Grindsted, Denmark). J. Contam. Hydrol. 33, 273–291. McMahon, P.B., Chapelle, F.H., 1991. Microbial production of organic acids in aquitard sediments and its role in aquifer geochemistry. Nature 349, 233–235. McNab Jr., W.W., 1999. Comparisons of geochemical signatures of biotransformation of fuel hydrocarbons in groundwater. Environ. Monit. Assess. 59, 257–274. National Research Council, 1993. In Situ Bioremediation: When Does it Work? Natural Academy Press, Wash- ington, DC. Prommer, H., Davis, G.B., Barry, D.A., 1999. Geochemical changes during biodegradation of petroleum hydrocarbons: field investigations and biogeochemical modeling. J. Org. Geochem. 30, 423–435. Reinhard, M., Shang, S., Kitanidis, P.K., Orwin, E., Hopkins, G.D., Lebron, C.A., 1997. In situ BTEX biotransformation under enhanced nitrate- and sulfate-reducing conditions. Environ. Sci. Technol. 31, 28–36. Schocher, R.J., Seyfried, B., Vazquez, F., Zeyer, J., 1991. Anaerobic degradation of toluene by pure cultures of denitrifying bacteria. Arch. Microbiol. 157, 7–12. Seyfried, B., Glod, G., Schocher, R., Tschech, A., Zeyer, J., 1994. Initial reactions in the anaerobic oxidation of toluene and m-xylene by denitrifying bacteria. Appl. Environ. Microbiol. 60, 4047–4052. Smith, M.R., 1990. The biodegradation of aromatic hydrocarbons by bacteria. Biodegradation 1, 191–206. United States Geological Survey, 1990. Installation restoration program phase II—confirmation/quantification stage 2, Wurtsmith Air Force Base, Michigan: Hydrology near Wurtsmith Air Force Base, Michigan. Prepared for Air Force Environmental Quality Branch Headquarters, Strategic Air Command, Offutt Air Force Base, Omaha, NE. U.S. Geological Survey, Lansing, MI. Vestal, J.R., White, D.C., 1989. Lipid analysis in microbial ecology. Bioscience 39, 535–541. 194 J.A. Namocatcat et al. / Journal of Contaminant Hydrology 67 (2003) 177–194

Vogel, T.M., Grbic´-Galic´, D., 1986. Incorporation of oxygen from water into toluene and benzene during anaerobic fermentative transformation. Appl. Environ. Microbiol. 52, 200–202. Vroblesky, D.A., Chapelle, F.H., 1994. Temporal and spatial changes of terminal electron-accepting processes in a petroleum hydrocarbon-contaminated aquifer and the significance for contaminant biodegradation. Water Resour. Res. 30, 1561–1570. Vroblesky, D.A., Bradley, P.M., Chapelle, F.H., 1996. Influence of electron donor on the minimum sulfate concentration required for sulfate reduction in a petroleum hydrocarbon contaminated-aquifer. Environ. Sci. Technol. 30, 1337–1381.