Fate of Microbial Metabolites of Hydrocarbons in a Coastal Plain

Jennifer J. Crawford, Gerald K. Sims, F. William Simmons, Loyd M. Wax, and David L. Freedman. Journal of Agricultural and Food Chemistry 2002 50 (6), ...
0 downloads 0 Views 2MB Size
Environ. Sci. Techno/. 1995, 29, 458-469

Fate of Miembid Mdddites of "r0C s in a Coastal Plain Aquifer: The Role of Electron Acceptors I S A B E L L E M . COZZARELLI,*jt JANET S . H E R M A N , * A N D M A R Y J O BAEDECKER' US.Geological Survey, Reston, Virginia 22092, and Department of Environmental Sciences, University of Virginia, Charlottesville, Virginia 22903

A combined field and laboratory study was undertaken to understand the distribution and geochemical conditions that influence the prevalence of low molecular weight organic acids in groundwater of a shallow aquifer contaminated with gasoline. Aromatic hydrocarbons from gasoline were degraded by microbially mediated oxidation-reduction reactions, including reduction of nitrate, sulfate, and Fe(lll). The biogeochemical reactions changed overtime in response to changes in the hydrogeochemical conditions in the aquifer. Aliphatic and aromatic organic acids were associated with hydrocarbon degradation in anoxic zones of the aquifer. Laboratory microcosms demonstrated that the biogeochemical fate of specific organic acids observed in groundwater varied with the structure of the acid and the availability of electron acceptors. Benzoic and phenylacetic acid were degraded by indigenous aquifer microorganisms when nitrate was supplied as an electron acceptor. Aromatic acids with t w o or more methyl substituents on the benzene ring persisted under nitrate-reducing conditions. Although iron reduction and sulfate reduction were important processes in situand occurred in the microcosms, these reactions were not coupled to the biological oxidation of aromatic organic acids that were added to the microcosms as electron donors.

Introduction The fate of aromatic hydrocarbons in surficial aquifers has been the subject of many studies because of their toxicity and high aqueous solubility. Benzene and the lower molecular weight alkylbenzenes are the hydrocarbons most frequently reported as being present in groundwater (e.g., refs 1-6). Aromatic hydrocarbons are subject to degradation by aerobic and anaerobic microbial processes in the environment (e.g., refs 7-9). In groundwater, anoxic conditions frequently develop as oxygen is consumed during the aerobic degradation of organic contaminants (e.g., refs 2-4 and 10-12). The degradation of monoaromatic hydrocarbons has been demonstrated to occur in the absence of oxygen by a variety of microbially mediated reaction pathways, includingnitrate reduction (13-13, iron reduction (16, 13, sulfate reduction (18, 19), and methanogenesis (20, 21). As a result of these anaerobic degradation processes, compounds such as low molecular weight (LMW) organic acids that were not present initially in the hydrocarbon source material are formed. Avariety of metabolic products, including aromatic, aliphatic, and alicyclic organic acids, phenols, and aldehydes, have been identified in laboratory experiments during the anaerobic microbial oxidation of petroleum hydrocarbons (13,21). Reports of aliphatic and aromatic organic acids in shallow aquifers contaminated with petroleum products suggest that these compounds are associated with the in situ anaerobic degradation of aromatic hydrocarbons (22-25). The metabolic intermediates produced during hydrocarbon degradation should be further degraded in the aquifer. Anaerobic metabolism of aromatic acids, such as benzoic acid, by microbial oxidation to COz is well documented in laboratory experiments (16, 26-28). The rate of degradation of these intermediates may control the rate of degradation of hydrocarbon precursors in some anaerobic environments in which a consortium of microorganisms works in syntrophic association to degrade organic carbon completely to carbon dioxide (9, 29). Although organic acids may contribute significantly to the total organic carbon pool, little is known about the distribution and fate of organic acids in groundwater environments. Organic acids are geochemicallyreactive and participate in processes such as metal complexation, sorption, and mineral dissolution. Thus, understanding the biogeochemical fate of organic acids is essential to predicting the geochemical evolution of shallow aquifers containing degradable organic compounds. The presence of these compounds in subsurface environments may have significant implications to the modeling of the geochemical effects of organic solutes in groundwater and the implementation of bioremediation technologies. This paper describes a combined field and laboratory study of LMW organic acids in groundwater. The purpose of our study was to determine how the distribution and biodegradability of organic acids in a shallow Coastal Plain * Corresponding author; FAX: 703-648-5274; e-mail address: [email protected]. + U.S. Geological Survey. University of Virginia.

*

458 1 ENVIRONMENTAL SCIENCE &TECHNOLOGY / VOL. 29, NO. 2, 1995

0013-936x/95/0929-0458$09.00/0

0 1995 American Chemical Society

METERS

FEET

NW

18

SE

FEET

60 8'

17

55 16

50

15 14

45 13

12

40

11

35 10

9

30

8

25 FIGURE 1. Cross-section of the study area illustrating the generalized hydrogeology within 25 ft of the land surface. The locations of the perching unit, e sand layer with dense stringers of clay, and the former gasoline tank are shown. The position of the sampling intewals at site VW9 and the perched and regional water tables in May 1990 are indicated.

aquifer contaminated with gasoline is affected by the availability of electron acceptors. Laboratory microcosm experiments, constructed with aquifer sediment and groundwater, were used to test the biodegradability under controlled conditions of specific organic acids identified in the anoxic aquifer. The importance of the acid structure as well as the availability of electron acceptors was investigated to determine the biogeochemicalfate of specific LMW organic acids. Additional discussions of the geochemical and hydrologic processes occurring in the aquifer are in Cozzarelli (30).

Methods Field Site. The field site is in Galloway Township, New Jersey, 10 mi west ofAtlantic City. An underground storage tank on a farmer's property leaked, resulting in the accumulation of pockets of a separate gasoline phase in the saturated and unsaturated zones of the shallow aquifer (31). The upper 20 ft of sediment at the site is predominately fine- to medium-grained sand with iron oxide coatings, and discontinuous clay lenses, and coarse-grained sand lenses (Figure 1). Below a depth of 20 ft, coarse-grained sand predominates. At approximately11- 15ft below land surface, a clayey-sand layer contains 1-3 in. thick, dense, clay lenses that have a low hydraulic conductivity (31).Due to the presence of the clayeylayer, perched water is present above the regional water table aquifer beginning at about 10 ft below the land surface. The regional water table fluctuated between approximately 14 and 17 ft below land surface during the period June 1989-April 1991. During

periods with large amounts of recharge, the unsaturated zone between the two hydrologic regimes disappeared in some locations. Groundwater Sampling. For this study, groundwater samples were collected from probes consisting of 0.25 in. diameter stainlesssteel tubes perforated over a 6-in. interval (31). This paper focuses on water sampled in February 1990 and April 1991, from probes installed at 10.7, 12.5, and 15.5ft below land surface (siteVW9) in both the perched water and regional groundwater (Figure 1). At this site, the two shallowest probes (10.7 and 12.5ft belowland surface) intersect the perched water. The probe at 15.5feet is at the interface between the perched water and regional aquifer, near the bottom of the low permeability clay that forms the base of the perched water zone. Water samples were collected from the probes with a peristaltic pump and analyzed for geochemical constituents that are reactants, intermediates, or end products involved in the degradation of the dissolved hydrocarbons. Groundwater samples for analyses of temperature, pH, 02, CHI, hydrocarbons, and LMW organic acids were collected without filtration. Temperature was measured, using a mercury thermometer, in a cup that collected water as it flowed from the end of the pump tubing. The pH of the water was determined at the well site using an Orion portable pH meter (Model SA250, Orion Research Inc., Boston, MA) and a Fischer combination glass-body electrode (Accu-pHast, Fisher Scientific, Pittsburgh, PA). Dissolved oxygen was determined in the field using a modified version of the Winkler titration (32,331on samples collected VOL. 29, NO. 2, 1995 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

469

in 60-300-mL BOD bottles. Water samples for the measurement of CH4 concentrations were collected in Glasspak syringes connected directlyto the sample-pump outlet. The water sample was transferred from the syringe into 25-mL serum bottles containing mercuric chloride (HgClZ, at a concentration of 0.2 mM Hg). Groundwater for determination of volatile hydrocarbons was collected in 40-mL amber glass vials preamended with HgC12. A recovery surrogate solution (o-xylene-dlo)was added to the samples immediately after theywere collected (34,35). Groundwater samples for analyses of LMW organic acids were collected in 14-mL glass vials with Teflon-lined caps. The samples were preserved with HgC12 and adjusted to pH 10 by the addition of 1 N potassium hydroxide. The samples were frozen within 8 h of collection. Water taken for NVDOC (nonvolatile dissolved organic carbon), VDOC (volatile dissolved organic carbon), "IS, HzS,NO3-, S042-,and HC03- analyses was filtered in-line through 0.2-pm Nuclepore filters. Samples taken for NVDOC and VDOC were preserved with HgClZ. Samples taken for NH4+were collected in glass vials and acidified to pH 2 with ultrapure nitric acid. For HPSmeasurements, the water was sampled with a syringe already containing methylene blue complexing reagents (36). Alkalinity (as HC03-) was determined in the field by potentiometric titration (32). Samples taken for Fez+analyseswere filtered through 0.1-pm filters and fixed with 2,2'-bipyridine method reagents in the field (32). All samples were stored at 4 "C during transport to the laboratory except for the organic acid samples which were kept frozen and shipped on dry ice. Analytical Procedures. The concentrations of Fez+and H2S were determined by modifications of the bipyridine method (33)and the methylene blue colorimetric method (36),respectively. Dissolved methane concentrations were measured by headspace analysis and gas chromatography (32). Ammonia concentrations were determined with an Orion Model 95-12 ion-selective electrode. Nitrate and Soh2-were analyzed by ion-exchange chromatography. Concentrations of NVDOC andVDOC were determined by the persulfate wet oxidation technique using a carbon analyzer (12). Water samples for hydrocarbon analyses were analyzed using a liquid-liquid extraction technique as described by Phinney and Cozzarelli (35). The hydrocarbons were analyzed by gas chromatography with ion-trap mass spectrometric detection using a DB-5 bonded-phase fused silica capillary column. Samples for analysis of LMW organic acids were analyzed on freeze-dried samples by extraction with diethyl ether as described by Cozzarelli et al. (23). The ether extracts were analyzed by high-resolution gas chromatography with flame ionization detection on a 30-m DB-Wax or DBFFAP coated fused silica capillary column (0.032 mm i.d., 0.25 mm film; J&W Scientific). Positive identification of organic acid structures was made on selected samples by co-injection with authentic standards and by gas chromatography/mass spectrometry using a Finnigan ion-trap detector. Microcosm Experiments. Microcosm experimentswere conducted with NO3-, S042-, or iron(II1) oxyhydroxide added as electron acceptors and specific organic acids added as potential substrates. A control experiment with no electron acceptors added was also conducted. Sediment was collected for the microcosm experiments with a 2 ft long split-spoon corer (3 in. diameter) from the anoxic 460

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 29, NO. 2, 1995

perched water zone at a location 5 ft west of site VW9 in April 1991 (Figure 1). Cores were collected from depths between 9.5 and 12.5 feet. The cores were opened in a nitrogen (N2)-filledglovebagat the field site and transferred to cleaned and sterilized glass jars. The jars were topped off with anoxic groundwater collected at site VW9 at 12.5 ft belowland surface (VW9-12.5)to minimize any headspace and kept on ice for transport to the laboratory. Twelve jars of sediment and an additional 2.5 L of water from VW912.5 were collected for use in the microcosms. The microcosms were constructed using 60-mL BOD bottles that were tested for exclusion of atmospheric oxygen (30).All manipulations of the microcosms were performed in a Nz-filledglovebag in the laboratory. Approximately 40 g of sediment and 40 mL of organic acid solution made with groundwater from site VW9-12.5 were added to each BOD bottle. Five organic acids were added to the groundwater for all the experiments to yield a concentration of approximately0.05-0.20 mMof each compound. The acids added were benzoic acid (added as sodium benzoate), o-toluic acid, phenylacetic acid, 3,4-dimethylbenzoicacid, and 2,4,6-trimethylbenzoic acid. The pH of the organic acid solutions was adjusted (with 1 N KOH) to 6.0, the value measured at VW9-12.5 in the field when groundwater samples were collected. Electron acceptors were added to separate aliquots of the organic acid solution for the different experiments in sufficient quantities such that their concentrations would not limit the degradation reactions. For the nitrate and sulfate reduction experiments, 30 mM NO3- or SO4*-was added. For the iron reduction experiment, 5 mL (0.47g dry weight) of an amorphous iron oxyhydroxide was added directly to each microcosm in sufficient quantity to result in 20-30 mM Fez+in solution if all of the iron were reduced. The iron(II1) oxyhydroxide was prepared as described by Lovley and Phillips (37).The microcosmswith no additional electron acceptor (Nom) contained the organic acid solution without any additionalFe(III),S042-,or NO3-. Four microcosms for each experiment contained autoclaved sediment and poisoned water (using HgC12) to serve as abiological controls. The microcosms were incubated in the dark at 14 "C. The degradation intermediates and end products produced during the incubation period were monitored at time intervals of 1, 2, 4, 8, 16, and 32 days. The microcosms were sacrificed in duplicate. The following analyses were done for each microcosm experiment: NO3- experiment: NO3-, NOz-, S042-,N&+, alkalinity,pH, organic acids; SO4'experiment: Nos-, NOz-, SO.?, H A alkalinity,pH, organic acids; Fe experiment: total dissolved Fe and Mn, alkalinity, pH, organic acids; N O W experiment: NO3-, NOZ-,SO4"-, NH4+,HzS, total dissolved Fe and Mn, alkalinity, pH, CHI, organic acids. Samples for analyses of NO3-, NOz-, S042-, NH4+, total dissolved Fe and Mn, and alkalinity were filtered through a 0.2-pm Nuclepore filter using a syringe. Samples for organic acids, pH, CH4, and H2S were collected without filtration. Transfer of the CHI and H2S samples by syringe into the collection bottles was done inside the glovebag. Iron and manganese were measured as total dissolved ions by direct current plasma emission spectrometry on an ARL Spectraspan V. Analytical methods used for the analyses of the other dissolved constituents in the microcosms were those previously described for the groundwater samples.

I

I I'

TABLE 1

Concentrations of Aromatic Hydrocarbons (in mgl.) Identified in Groundwater at Site W 9 . at 15.5 It below Land Surface in February 1990 (mSnl

pH

DO

fez'

NH;

H$

TDOC HCO> SO:

NO,

FIGURE 2. Concentrations (in mgA) d 00, W+,NH,+, HIS. TOOC, HCO,-.SOf-,and NO,-,andpHvaluesincontaminatedgmundwmer ham the perched water zone in January 1990 compared to the concentrationsat an uncontaminated location in the perched water.

In cases in which the concentrations of some of the constituents were too low to be illustrated with a bar. the numerical values are indicated on the figure.

Results and Discussion Groundwater Chemistry. The geochemical patterns observed in the groundwater of the contaminated aquifer, compared to uncontaminated water (Figure 21, can be explained by microbiaJly mediated oxidation-reduction reactions. Compared to theuncontaminatedperchedwater at a site located 20 ft to the south WlO at 12.5 ft depth), thecontaminatedwater atVW9 (at 10.7ftdepth) contained high concentrations of total dissolved organic carbon (TDOC) from the dissolutionof gasoline hydrocarbons and the production of organic acid metabolites from hydrocarbon degradation. The uncontaminated groundwater was acidic (pH = 4.731, whereas the contaminated water was less acidic (pH = 5.99). The HC03- concentration of 150 mglL in the contaminated water was nearly 2 orders of magnitude higher than the concentration found in the area unaffected by hydrocarbon degradation (2.4 mglL). Alkalinity and pH increased in the area affected by the gasoline as hydrocarbons were oxidized to bicarbonate by aerobic and anaerobic degradation reactions. The background water had dissolved 02,N03-,and SO4*- concentrations of 6.1, 39.7, and 23.7 mglL, respectively. The contaminated water was anoxic, contained no NOS-, and had a S042- concentration of 13.2 mg1L indicating that nitrate, iron, and sulfate reduction were important anaerobic respiratoryprocesses at this site. The possible reduced end products of these processes including N&+, FG+, and H2S accumulated in the anoxic groundwater reaching concentrations of 2.6, 19.9, and 1.2 mgIL, respectively. Concentrationsofthese constituents were below detection limits in the background water. The low concentrations of CHI (c0.05 mg1L) in contaminated groundwater indicate that methanogenesis was not an important reaction. Atotal of41 aromatichydrocarbons, includingbenzene, toluene, ethylbenzene,andxylenes (togetherreferred to as BTFXI as well as C3-and G-alkylbenzenes, naphthalene, and methylnaphthalenes were identified in the perched water (Table 1). Aliphatic hydrocarbons were not found in the groundwater. Several classes of LMW organic acids were identitied in the perched water at VWS (15.5 ft below land surface), including branched and straight-chain aliphatic acids, aromatic acids, and alicyclic acids (Table

benzene toluene ethylbenzene pxylene mxylene isopropylbenzene o-xylene-dto (recovery surrogate) c-xylene mpropyi benzene 1-ethyl-4-methylbenzene 1-ethyl-3-methylbenzene 1.3.5-trimethylbenzene 1-ethyl-2-methylbenzene 1.2.4-trimethyl benzene n-butylbenzene lf.3-trimethyi benzene 1.2.4.5-tetramethylbenzene 1.2.3.5-tetramethylbenzene 1.2.3.4-tetramethylbenzene naphthalene 2-methylnaphthalene 1-methylnaphthalene

0.193 1.50 1.11 1.97 4.25 0.084 4.49 2.08

0.223 0.396 0.890 0.517 0.485 1.73 0.015

0.139 0.081 0.258 0.079 0.377 0.070 0.043

TABLE 2

Concentrations of Organic Acids (in pgll) Identified in Groundwater at Site W 9 . at 15.5 H below Land Surface in February 1990 bsnl acetic acid propanoic acid isobutanoic acid butanoic acid isopentanoic acid pentanoic acid heptanoic acid cyclohexanoic acid c-toluic acid ptoiuic acid phenylacetic acid 2.4- and 2.5-dimethylbenzoic acid 2.6-dimethylbenzoic acid pmethylphenylacetic acid 3.4-dimethyibenzoic acid 2.4.6-trimethylbenzoic acid

43 5 2 2 7 2 4 8

70 77 5 365 16 91 32 71

2). The most abundant aliphatic acids were acetic acid

and 3,3-dimethylbutanoic acid, which was semi-quantified by GClMS (30). The aromatic acids were mostly methyl and dimethyl benzoic acids as well as phenylacetic acids. The most abundant alkylbenzoic acids were 0-toluic acid, p-toluic acid, 2,4-andlor 2,5-dimethylbenzoic acid, p methylphenylacetic acid, 3,4-dimethylbenzoic acid, and 2,4,6-trimethylbenzoic acid. The presence of LMW organic acids (intermediates in the degradation of hydrocarbons) is huther evidence that microbial degradation of the hydrocarbons occurred. The highest concentrations of LMW organic acids were found in the anoxic perched water where anaerobic microbial degradation processes were dominant (38). Although the metabolic pathways resulting in production of the alkylbenzoic acids in sikd cannot be ascertained from the field data alone, the geochemical evidence indicates that the alkylbenzenes were oxidized in the anoxic groundwater. VOL. 29, NO. 2.1995 I ENVIRONMENTAL SCIENCE &TECHNOLOGY.

481

For example, the aromatic hydrocarbon 1,2,4-trimethylbenzene (1,2,4-TMB) was one of the more abundant alkylbenzenes in the anoxic perched water zone in 1990 (Table 1). The predominant organic acid found, a dimethylbenzoicacidwith a2,4- or 2.5-dimethyl substitution, could result from the oxidation of one of the methyl groups of 1,2,4-TMB. LaboratoryexperimentsofMillsandRandall (39) identified anaerobic bacteria in groundwater at this site that could grow on gasoline hydrocarbons with NOsand SOP2- as electron acceptors. Funhermore, the same types of organic acids found in this study were reported in anoxic groundwater downgradient from a crude oil spill near Bemidji, M N (221, in which anaerobic degradation of monoaromatic hydrocarbons by iron reduction and fermentative processes occurred in situ. In subsequent anaerobic microcosm experiments with Bemidji aquifer sediment, Cozzarelli et al. (23)verified that the production of alkylbenzoicacids occurred during the anaerobic oxidation of alkylbenzenes. In another study, Barbaro et al. (40) identified 2-methylbenzoic acid (0-toluic acid) during a natural-gradient injection experiment in which a solution containingBTEXandNO3-wasinjectedintoanuncontined, sandy aquifer at Canadian Forces Base (CFB) in Borden, Ontario. The o-toluic acidwasfoundinanoxicgroundwater and was presumed to result from degradation of BTEX compounds by denitrifying bacteria. Changes in the Biogeochemical Reactions over Time. Changes in the geochemical reactions that control the observed groundwater composition result, in pan. from changes in the hydrologic conditions at this site. Oxygen, nitrate, and sulfate are supplied to the shallow perched water (10.7-ft depth) by vertical recharge fromprecipitation events. After water levels had reached a low point at the end of 1990, recharge in the winter-early spring of 1991 (30)resulted in increased concentrations of nitrate and sulfate in the groundwater. The higher nitrate and sulfate concentrations in the shallow perched water in April 1991 (Figure 3.4) indicate that nitrate and sulfate reduction by anaerobic respirationwerenot limited by lack of availability of electron acceptors. These processes likely contributed to the depletion of hydrocarbons observed at this depth in April 1991 relative to February 1990 (Figure 3B). Hydrocarbon concentrations were c0.30 mg/L in April 1991, a decrease of nearly 90% from those observed in 1990. The high concentrations of total LMW organic acids in April 1991 (greater than 1.5 mglL), compared to the residual concentrations of hydrocarbons at this location (Figure 3B), indicate that some of the acids produced during the metabolism of the hydrocarbons are resistant to further anaerobic degradation and thus accumulate in the gmundwater. The changes in concentrations of specific classes and structures of LMW organic acids over time in the groundwater at site VW9 provide insight into the persistence of certain organic acids in situ. Although the concentrations of total LMW organic acids did not change significantly between 1990 and 1991 in the perched water (Figure 3B), the increase in the aromatic acid fraction and simultaneous decrease in the aliphatic acid fraction indicate that the relative persistence of these two classes of compounds differs. Although the fate of the aliphatic acids cannot be concluded from the field data alone, in the nitrate-rich groundwater the degradation of aliphatic acids by nitratereducing bacteria in the aquifer may prevent their accumulation in the groundwater. This is in contrast to the

A

30.0 I

E v

8

*

a

2 5

0 so,' h9 NO,

25.0 20.0 15.0

L

10.0 5.0

QJ

I3

I

February 1990

0.0

2 2.0

-

3

-

1.5

0 Aromatic Acids

3

5 a

1.0 -

e

s

$

0

OS -

6 0.0

-

February 1990

April 1991

W Benzoic Acid L l Phenylacetic Acid 0 0-Toluic Acid

0 3,4-Dimethylbenwic Acid 2,4,6-TrirnethylbenzoicAcid r-

February 1990

April 1991

FIGURE3. Concentrations(in m a ) of (A) SO& and NO,: (E)total ammatic hydrocarbons, LMW aromatic acids, and LMW aliphatic acids: and (C) individual aromatic organic acids: benMic acid, phenylacetic acid, etoluic acid.3.4-dimethylbenzoic acid. and 2.4.6trimethylbenzoic acid in the shallow perched water (at 10.7 ft) at site wy9 in February 1990 and April 1991. The data on graphs B and C represent the average of field duplicates, and the error bars (C) represent the magnitude of the range of values for the individual organic acids.

results of a study of groundwater downgradient of a crude oil body in which Cozzarelli et al. (23) found that the concentrations of aliphatic organic acids increased by 2 orders of magnitude as microbial degradation reactions shifted from iron reduction to methanogenesis. The depletion of electron acceptors by microbial processes in the aquifer led to the accumulation of aliphatic organic acids in anoxic groundwater. In the current study, the decrease in the concentrations of hydrocarbons as substrates may also have affected the observed low concentrations of aliphatic acids: in the absence of appreciable

concentrations of readily degradable aromatic hydrocarbons, the rate of utilization of the aliphatic acids may have increased. Under these conditions, some of the aromatic acids were persistent. A study of the persistence of individual aromatic acids in the perched water where nitrate reduction was occurring indicates that benzoic acid and phenylacetic acid concentrations decreased over time whereas o-toluicacid and2,4,6trimethylbenzoicacid increased by factors of 2.5-7 (Figure 3 0 . The concentrations of 3,4-dimethylbenzoic acid remained relatively constant over the time period shown. Because the metabolic intermediates of hydrocarbon degradation are both produced and consumed by microbiological processes, rates of degradation of the LMW organic acids cannot be inferred from the concentration changes observed in groundwater over time. However, the relative persistence of the different compounds indicates that the simple aromatic acids, benzoic acid and phenylacetic acid, are either degraded more quickly when nitrate is available as an electron acceptor or that they were not produced as extracellular intermediates during hydrocarbon degradation under nitrate-reducingconditions. Laboratory studies of anaerobic microbial degradation pathways have indicated that benzoic acid is a transient intermediate in the degradation of toluene under nitrate-reducing (13)as well as sulfate-reducing (41) conditions. The turnover of benzoic acid in these culture studies was rapid, resulting in the detection of very low concentrations in solution. In the current study, the organic acids with more substituents (o-toluic acid, 3,4-dimethylbenzoic acid, and 2,4,6-trimethylbenzoic acid) remained constant or increased in concentration under the same conditions in which decreases in the concentrations of benzoic acid and phenylacetic acid were observed (Figure 3C). Other studies have also shown that slight changes in the number or type of substituents on the aromatic ring results in significant differences in the degradability of the compounds (see, for example, refs 7 and 42). Biological Oxidation-Reduction Reactions. Results from the microcosm studies (Figures4-7) provide insight into the fate of some of the aromatic organic acids, in the presence of different electron acceptors, that could not be achieved by analysis of the field data alone. Microcosm experiments in which NO3- was added as a potential electron acceptor clearly establish that the degradation of the simplest aromatic acids, benzoic acid and phenylacetic acid, is coupled to NO3- reduction and support the conclusions of the field study (Figure 5A-F). Complete mineralization occurs, as evidenced by the loss of these organic acids (Figure5D,E) coincident with the production of bicarbonate (Figure 5A) in the microcosms over time. No significant increase in HC03- or decrease in organic acid concentrations was observed in the sterile control. No decreases in concentrations were observed for the other alkybenzoic acids in either the viable or sterile microcosms as illustrated by o-toluic acid (Figure 5F). In the viable nitrate-amended microcosms,benzoic acid degradation began first and phenylacetic acid degradation began before benzoic acid was depleted. Between days 16 and 32 (after the benzoic acid was depleted), 0.093 mM phenylacetic acid was degraded (Figure5E). For complete oxidation ofthis amount of organic carbon, 0.74 mMHCO3should have been produced. Over this time period, an increase of 0.50 mM HCO3- was measured, indicating that 33% of the COawent into biomass assuming there were no

losses of aqueous C 0 2 by precipitation of carbonates. Alternatively, some carbon could be present as metabolic intermediates, although no LMW organic acids were detected other than those added as substrates. If the oxidation of phenylacetic acid over this time period (0.093 mM) was coupled with nitrate reduction to NH4+,0.42 mM of NO3- would need to be reduced to balance the electrons released. If, on the other hand, denitrification was occurring, 0.67 mM of NO3- would be reduced. The actual amount of NO3- reduced was 1.06 mM NO3-, indicating that microbiallymediated nitrate reduction may be coupled to the oxidation of other organic compounds in addition to the added organic acids. Although there was no degradation of organic acids in the sulfate-amended microcosms, the production of 0.45 mM HC03- in the microbially active microcosms (Figure SA) indicatesthat organic material was degraded. Between days 16 and 35 there was an apparent decrease in sulfate concentration (loss of 2.5 mM, Figure 6B) and a slight increase in the concentration of H2S (Figure 6C) in the microbially active microcosms, indicating that the bicarbonate produced during this same time interval may have been the product of microbially mediated oxidation of DOC coupled to the reduction of sulfate. However, because the concentrations of S042- were large and the duplicates showed significant variation early in the experiments, it is difficult to assess the significance of a change in concentration of this magnitude (approximately 10%;Figure 6B). In addition to these changes in concentrations of S042and HzS in the viable microcosms, the difference between the viable and sterile microcosms on day 32 supports the contention that microbially mediated sulfate reduction occurred. In the microcosms without any additional electron acceptor, no comparable increase in HzS concentrations was observed (Figure 4D), indicating that the reactions observed in the sulfate-amended microcosms were stimulated by the addition of sulfate to the system. The most likely biodegradable carbon sources (other than the added organic acids) in the nitrate- and sulfateamended microcosms were the aromatic hydrocarbons present in the aquifer sediment and groundwater from site VW9 used in the construction of the microcosms. Mills and Randall (39)verified the presence of nitrate- and sulfatereducing microorganisms capable of utilizing benzene, toluene, and xylenes at the location from which the sediment and water were collected for the current microcosm study. The concentrations of total monoaromatic hydrocarbons measured in the groundwater at site VW9 were more than sufficient to produce the 0.45 mM HC03produced in the microcosms (FigureSA). The large number of organic compounds present in the aquifer results in numerous possible reactions, resulting in the observed reduction of nitrate and sulfate in situ. The fact that no losses of organic acids occurred over time in the sulfateamended microcosms (Figure 6) or the iron-amended microcosms (Figure 7) indicates that the sulfate and iron reduction observed in the field study were likely coupled to the oxidation of other organic compounds present in the aquifer that are more readily degradable by sulfate and iron reducers. Although other investigators have established sulfate-reducing enrichment cultures in laboratory incubationsthat utilized benzoic acid and phenylaceticacid (28),and the microbially mediated oxidation of benzoic acids in pure culturesbythe Fe(II1)-reducingbacterium GS15 is well documented (16,17), the conditions in the VOL. 29, NO. 2, 1995 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 1483

A

D

Microcosms without Added Electron Acceptors

0.25 0.20

-

0.15

- $

viable

: .

‘8

sterile

0.05 0.00

5

0

10

B

8

15

20

25

30

0

35

1 0

,

1

5

,

,?

10

15

Days

0’05 0.04

v

o,5

\

0.03

2 0’02

25

5 - sterile

1’ t

0.12

viable

i

viable

LJ -,

i

0.06 0 0.00

5

10

15

Days

C

t

1

8

sterile

t

1

0.30 0.24

A‘

-

/

1 -

viable

-

t

01 0

0.18

10

15

20

25

30

I

20

25

30

! 35

f

I

I

35

0.00

3

I

viable

I‘

1

sterile

‘8

j

it* * -

0.12 0.06

5

35

Days

0.5

0.4

30

i

sterile

0.24

t 0.01 c I I

0.2

20

, sterily

Days

-

0.3

,

1

Days FIGURE 4. Concentration (in mM. except H2S in pM) of (A) bicarbonate; (B) nitrate; (C) sulfate; (D) sulfide; (E)benzoic acid; and (F) phenylaceticacid in the microcosmswith no additionalelectron acceptor added as a function of time (in days). The closed diamonds indicate the viable microcosms whereas the open triangles are the sterile controls. The microcosms were sacrificed in duplicate at all times except for the sterile microcosms on day 1. Where only one data point is visible, the measured values of the duplicate microcosms were identical.

microcosms used in the present study apparently did not favor these reactions. The time scale of the microcosm experiments is short relative to the time available for decomposition of the LMW organic acids in the aquifer. It is possible that the degradation of aromatic organic acids observed in groundwater at the Galloway site is coupled to sulfate reduction and Fe(II1) reduction at a rate too slow to observe in these laboratory experiments. The persistence of the higher alkylated benzoic acids in the nitrate-amended microcosms indicates that they were not biodegraded under the conditions employed in the 464 1 ENVIRONMENTAL SCIENCE &TECHNOLOGY / VOL. 29. NO. 2, 1995

experiments (e.g.,o-toluicacid, Figure 5F). The observation that biodegradation of LMW organic acids is structurespecific is consistent with reports of selective degradation of other aromatic compounds (e.g., refs 6, 19, 43, and 44). In particular, aromatic compounds with substitutions at the ortho position are more stable than aromatic compounds with substitutions at other positions. This may explain the observed persistence of o-toluic acid and 2,4,6trimethylbenzoic acid in the shallow groundwater (Figure 3C). In a study of the degradation of benzoic acid and phenylacetic acid by a sulfate reducer, Desulfosarcina strain

A

D

Nitrate Amended Microcosms

I

3

Y

.-u a 8 m 0.5 l.OL*J

0.0

I

,

,

,

20

25

30

h

,

i:c/ 0.00

5

0

10

15

0

35

,

5

,

,~

,

10

15

20

25

1

\~

viable

30

35

Days

Days

B

,

E 21 26

3

sterile

25

\

24

3

;i:

1

t

5

0.25

-

2 -

23 22 21

viable

20

I

4

5

10

15

20

25

30

Days

C 0.20

,

!

!

1

sterile

I'

1

0.00 1 0

'.

* 5

10

15

I

0

5

10

15

I

0.25

1

0.20

I 35

0.05 0 0.00

viable

20

25

30

Days

20

25

1- 1 35

30

Days

F

-1

0.05

0.00

35

j

0.15 0.10

5 5

10

15

20

25

30

35

Days

FIGURE 5. Concentration (in mM) of (A) bicarbonate; (B) nitrate; (C) ammonium; (D) benzoic acid; (E) Phenylacetic acid; and (F) @toluic acid as a function of time (in days) i n the microcosm experiments in which nitrate was added as an electron acceptor. The closed diamonds indicate the viable microcosms whereas the open triangles are the sterile controls. The microcosms were sacrificed i n duplicate et ell times except for the sterile microcosms on day 1. Where only one data point is visible, the measured values of the duplicate microcosms were identical.

DSUS,Sembiring and Winter (28) observed that benzoic acid degraded first when in the presence of phenylacetic acid-an observation consistent with the results of the nitrate-amended microcosms presented here. However, Sembiring and Winter (28)found sigmficant accumulation of acetic acid during the degradation of benzoic acid, suggesting that the potential for acetate accumulation may be greater in pure cultures than in heterogeneous systems where mixed cultures and multiple substrates are present.

Abiological Oxidation-Reduction Reactions. Although the reduction of nitrate and sulfate in the microcosm experiments was biologically mediated, dissolution of amorphous ferric iron in the Fe(II1)-amendedmicrocosms was not biologically mediated, as evidenced by the dissolution of iron in the sterile controls (Figure7B). Analysis of the geochemical data from the microcosms with the computerized geochemical reaction path model PHREEQE (45) revealed that the microcosm solutions were underVOL. 29. NO. 2. 1995 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

485

0.60

v

0.40

:

0.30

-

,

0.25

I

I

0.50 -

Q

D

Sulfate Amended Microcosms

A

viable

\

a2 c)

4.s rn

V

a

.I

0.20 -

0.10 -

5

t

m

:

0.10 -

-

0.05

sterile 0.00

1

5

0

10

15

,

,

I

I

,

1

'*

i

30

35

0.25

i

sterile

25

Days

E 28

20

I

-

viable

':

,t

t

21

0.00

1 0

10

5

20

15

25

j i

30

35

Days

F 3.00 2.50

0.25

-

g

viable,

2

2.00

Lz 4

1.50

3

1.00

A v

21 1

9

o'20

;]I

0.15 -

i

I

'I 1 '

sterile'

4

:

viable

4

.-V 3

-

0.00 1 0

1

-A

1

I

I

5

10

1 15

I

sterile \

, 20

I

25

30

...

\A

35

I 40

0.10 0.05

-

0.00 I

Days FIGURE 6. Concentration (in mM,except H2S in pM)of (A) bicarbonate; (b)sulfate; (C) sulfide; (D) benzoic acid (E) phenylacetic acid and (F) 0-toluic acid as a function of time (in days) in the microcosm experiments in which sulfate was added as an electron acceptor. The closed diamonds indicate the viable microcosms whereas the open triangles are the sterile controls. The microcosms were sacrificed in duplicate at all times. Where only one data point is visible, the measured values of the duplicate microcosms were identical.

saturated with respect to amorphous iron oxyhydroxides, indicating the potential for chemical dissolution of Fe(II1). The mechanism of the dissolution is unknown; however, it is assumed that the dissolution was a reductive process and, due to the low solubility of Fe3+at near-neutral pH, the total iron measured in solution (1.1mM) was in the Fe2+ oxidation state (46). The most likely chemical mechanism for the reductive dissolution of Fe(II1)in these experiments is reduction by sulfide and, possibly,by organic solutes (common chemical reductants of iron(II1) oxyhydroxides; 47). The dissolution of Fe(II1) by reaction with 466 m ENVIRONMENTAL SCIENCE &TECHNOLOGY / VOL. 29, NO. 2,1995

sulfides has been documented in anoxic marine sediments (48). Although the concentrations of dissolved HzS in the microcosms were low ( < 3 pM in the microcosms with no added electron acceptors; Figure 4D), the geochemical profiles from the field site show that the aquifer supports sulfate reduction. In addition, the presence of solid FeS in the aquifer sediment as authigenic coatings on mineral grains (381,supports the conclusion that sulfate reduction is occurring in the aquifer. It has been demonstrated that pyrite oxidation is coupled to Fe(II1) reduction even at circumneutral pH where the solubility of Fe3+is virtually

A

C

Iron Amended Microcosms 0.05 0'25

1

t

0.20

1i viable

I'

sterile

\

0.00I

0

5

4

10

15

25

20

30

0.00

35

1

0

1

5

I

15

10

Days

t

20

1

25

30

1

35

Days

D B 0.25

1.4

E -E

-B-

g

1.2

0.20

c t

,

1

,

4

1

viable

*'

1

E

E

8

0.8

I

0.6

/

sterile 0.4

0.2

t 0

t

1

5

1

10

1

15

I

25

20

1

30

35

Days

0.00

1 0

1 5

10

15

20

25

30

1

35

Days

FIGURE 7. Concentration (in mM) of (A) bicarbonate; (B) total dissolved iron; (C) benzoic acid; and (D) phenylacetic acid as a function of time (in days) in the microcosm experiments in which Fe(lll) was added as an electron acceptor. The closed diamonds indicate the viable microcosms whereas the open triangles are the sterile controls. The microcosms were sacrificed in duplicate at all times except for the sterile microcosms for the organic acid analyses on day 1. Where only one data point is visible, the measured values of the duplicate microcosms were identical.

zero (49).The following equation, from Moses et al. (491, describes the reaction:

FeS,(s)

+ 14Fe3++ 8H,O

-

+

15Fe2+ 2SO:-

+ 16H'

(1)

If all of the Fe2+in the sediment was generated by the above reaction, it would require the oxidation of only 0.24 mg of FeS2,which is less than 0.001% of the sediment by weight. The lack of significant disappearance of organic acids (added as substrates) over time in these microcosms indicates that the reduction of iron was not directly coupled to the abiological oxidation of these species but possibly to other organic substances present in the sediment and groundwater used to construct the microcosms. Reactions with oxygenated compounds, such as indenones and benzeneacetaldehydes, tentatively identified in groundwater at site VW9 (30)could account for some of the reduction of Fe(II1) observed in the Fe(II1)-amended microcosms. However,the organic compounds must have been only partially oxidized since no COz production was observed over time in the microcosms (Figure 7A). Lovley et al. (50) observed the nonenzymatic dissolution of amorphous iron(II1) oxides by benzaldehyde, benzyl alcohol, and phenylacetate, although the rates of dissolution

in the absence of microbial enzymes were much slower than those in the presence of enzymes. The amount of Fe2+produced in their experiments, 0.4-1.4 mM after 21 days, is comparable to the amount of dissolution (0.6 mM) observed in the present experiments. Lovley et al. (50)and Lakind and Stone (51) also found that the nonenzymatic reduction of Fe(II1) by aromatic compounds was typically only a two-electron transfer and did not result in any COZ production. Phenolic compounds have also been demonstrated to nonenzymatically reduce Fe(III), although Lakind and Stone (51)found the reaction rate above pH 6 was too slow to be detected. Although the exact nature of the dissolution of the iron(II1) oxyhydroxide added to the microcosms is not known, the most likely mechanism appears to be the chemical reduction of iron by sulfide minerals or aromatic compounds present in the sediments at the time the microcosms were constructed. This mechanism is consistent with the presence of both dissolved sulfide and iron in the groundwater and sulfide minerals in the aquifer sediment at site VW9. Although microbially mediated Fe(II1) reduction was not observed in the microcosms, the possibilityof direct microbial reduction of iron in situ should not be discounted. The groundwater contained high VOL. 29. NO. 2, 1995 / ENVIRONMENTAL SCIENCE &TECHNOLOGY 467

concentrations ot many organic compounds, including aromatic hydrocarbons. The enzymatic oxidation of these compounds by iron-reducing bacteria has been well documented (16, 17) and has resulted in high dissolved iron concentrations in aquifers contaminated with crude oil (12).

Conclusions The fate of LMW organic acids in a shallow aquifer contaminated with gasoline is controlled by a complex interplay between microbial degradation processes and geochemical and hydrologic changes in the aquifer. At the field site, LMW organic acids were found to be associated with hydrocarbon degradation in oxygen-depleted zones of the aquifer. The organic acid pool in the groundwater changed in composition and concentration over time as the availability of electron acceptors changed. The availability of nitrate from infiltrating water appeared to be important in regulating the concentrations of simple aromatic acids in groundwater. Laboratory microcosm experiments supported the conclusion of the field study that the fate of specific LMW organic acids observed in the groundwater depended on the structure of the acid and the availabilityof electron acceptors. The addition of nitrate enhanced microbial degradation of benzoic acid and phenylacetic acid. Sulfate, and some of the nitrate, reduced in the microcosms was apparently coupled to the biologically mediated oxidation of other organic compounds that are important substrates for the aquifer bacteria, whereas the dissolutionof Fe(II1)in the Fe(II1)-amended microcosms was due to abiological mechanisms. Understanding the biogeochemical fate of organic acids is essential to predicting the geochemical evolution of shallow aquifers containing degradable organic compounds. In shallow aquifers contaminated with petroleum products, the fate of microbial metabolites of hydrocarbons will depend on the availability of electron acceptors. Geochemical studies of natural and contaminated subsurface environments have shown that as organic matter degradation proceeds in systems in which electron acceptors become limited, aliphatic organic acids reach high concentrations (23,24,52)and their presence in groundwater can be used as an indicator of redox processes (53). The results presented here, combined with these previous studies, may provide a basis for understanding the hydrogeochemical conditions under which organic compounds persist in the subsurface and illustrate the challenges we face in studying geochemical processes in heterogeneous systems.

Acknowledgments This project was conducted as part of the U S . Geological SurveyToxic Substances Hydrology Program. The authors benefitted from discussion with Roseanne Ford, George Hornberger, and Aaron Mills and from review comments by Eric Roden, Ronald Baker, and three anonymous reviewers. The technical assistance of and helpful discussionswith US. GeologicalSurveyscientistsin West Trenton, NJ, Art Baehr, Jeff Fischer, and Nick Smith, are greatfully acknowledged. We thank Curtis Phinney for his assistance with the GUMS analyses of hydrocarbons and organic acids and Jessica Hopple and Carol Wicks for providing help with field work. Elizabeth Phillips generously provided the amorphous iron(II1)oxyhydroxide for the microcosms. MyChau Tran provided analytical support and assistance with 468

1

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 29, NO. 2, 1995

the preparation of the data tables. We thank Melissa Schomody for typing the tables and Leslie Robinson for preparing the illustrations.

literature Cited (1) Dell’Acqua, R.; Bush, B.; Egan 7.1. Chromatogr. 1976,128,271280. (2) Schwarzenbach, R. P.; Giger, W.; Hoehn, E.; Schneider, J. K. Environ. Sci. Technol. 1983, 17, 472-479. (3) Reinhard, M.; Goodman, N. L.; Barker, J. F. Environ. Sci. Technol. 1984, 18, 953-961. (4) Barker, J. F.; Tessman, J. S.; Plotz, P. E.; Reinhard, M.J. Contam. Hydrol. 1986, 1, 171-189. (5) Wilson, J. T.; Miller, G. D.; Ghiorse, W. C.; Leach, F. R.J. Contam. Hydrol. 1986, 1, 163-170. (6) Eganhouse, R. P.; Baedecker, M. J.; Cozzarelli, I. M.; Aiken, G. R.; Thorn, K. A.; Dorsey, T. F. Appl. Geochem. 1993,8551-567. (71 Berry, D. F.; Francis, A. J.; Bollag, J. M. Microbiol. Rev. 1987,51, 43-59. (8) Evans, W. C.; Fuchs, G.Annu. Rev. Microbiol. 1988,42,289-317. (9) GrbiC-Galit, D. Geomicrobiol. J. 1990, 8, 167-200. (10) Golwer, A.; Matthess, G.; Schneider, W. Groundwater PollutionSymposium: Proceedingsof theMoscowSymposium,Augustl971; IAHS Publication 103; IAHS Press: Wallingford, 1975; p p 159166. (1 1) Baedecker, M. J.; Apgar, M. A. Studies in Geophysics: Groundwater Contamination;National Academy Press: Washington, DC, 1984; pp 127-138. (12) Baedecker, M. J.; Cozzarelli, I. M.; Siegel, D. I.; Bennett, P. C.; Eganhouse, R. P. Appl. Geochem. 1993, 8, 569-586. (13) Kuhn, E. P.; Zeyer, J.; Eicher, P.; Schwarzenbach, R. P. Appl. Environ. Microbiol. 1988, 54, 490-496. (14) Major, D. W.; Mayfield, C. I.; Barker, J. F. Ground Water 1988, 26, 8-14. (15) Hutchins, S. R.; Sewell, G. W.; Kovacs, D.A.; Smith, G.A. Environ. Sci. Technol. 1991, 25, 68-76. (16) Lovley, D. R.; Baedecker, M. J.; Lonergan, D. J.; Cozzarelli, I. M.; Phillips, E. J. P.; Siegel, D. I. Nature 1989, 339, 297-299. (17) Lovley, D. R.; Lonergan, D. J. Appl. Environ. Microbiol. 1990,56, 1858-1864. (18) Edwards, E. A.; GrbiC-GaliC, D. Appl. Environ. Microbiol. 1992, 58, 2663-2666. (19) Haag, F. M.; Reinhard, M.; McCarty, P. L. Environ. Toxicol. Chem. 1991, 10, 1379-1390. (20) Wilson, B. H.; Smith, G. B.; Rees, J. F. Environ. Sci. Technol. 1986,20, 997-1002. (21) GrbiC-GaliC, D.; Vogel, T. M. Appl. Environ. Microbiol. 1987,53, 254-260. (22) Cozzarelli, I. M.; Eganhouse, R. P.; Baedecker, M. J. Environ. Geol. Water Sci. 1990, 16, 135-141. (23) Cozzarelli, I. M; Baedecker, M. J.; Eganhouse, R. P.; Goerlitz, D. F. Geochim. Cosmochim. Acta 1994, 58, 863-877. (24) Godsy, E. M.; Goerlitz, D. F.; GrbiC-Galit, D. Ground Water 1992, 30, 232-242. (25) Wilson, B. H.; Wilson, J. T.; Kampbell, D. H.; Bledsoe, B. E.; Armstrong, J. M. Geomicrobiol. J. 1990, 8, 225-240. (26) Evans, W. C. Nature 1977, 270, 17-22. (27) Bak, F.; Widdel, F. Arch. Microbiol. 1986, 146, 177-180. (28) Sembiring, T.; Winter, J. Appl. Microbiol. Biotechnol. 1989, 31, 84-88. (29) Ferry, J. G.; Wolfe, R. S. Arch. Microbiol. 1976, 107, 33-40. (30) Cozzarelli, I. M. Ph.D. Thesis, University of Virginia, May 1983. (31) Fischer, J. M.; Baehr, A. L.; Smith, N. P. US.Geological Survey Toxic Substances Hydrology Program-Proceedings of the technical meeting, Monterey, California, March 11-15, 1991; Mallard, G. E., Aronson, D. A., Eds.; Water Resources Investigations Report 91-4034; U.S. Geological Survey: Reston, VA, 1991; pp 243249. (32) Baedecker, M. J.; Cozzarelli, I. M. In Ground-Water Contamination and Analysis at Hazardous Waste Sites; Lesage, S., Jackson, R. E., Eds.; Marcel Dekker, Inc.: New York, 1992; pp 425-462. (33) Brown, E.; Skougstad, M. W.; Fishman, M. J. Methods for collection and analysis of water samples for dissolved minerals and gases; Techniques of Water-ResourcesInvestigations of the US.Geological Survey; U.S. Government Printing Office: Washington, DC, 1970. (34) Baedecker, M. J.; Cozzarelli, I. M.; Phinney, C. S. US.Geological Survey Toxic Substances Hydrology Program-Proceedings of the technical meeting, Monterey, California, March 11 -15, 1991; Mallard, G. E., Aronson, D. A., Eds.; Water Resources Investigations Report 91-4034;U.S. Geological Survey: Reston, VA, 1991; pp 287-293.

(35) Phinney, C. S.;Cozzarelli, I. M. Proceedings, 40th ASMS Conference on Mass Spectrometry and Allied Topics: Washington, DC, 1992. (36) Lindsay, S. S.;Baedecker, M. J. In Ground-Water Contamination: Field Methods; Collins, A. G., Johnson, A. I., Eds.; ASTM Special Technical Publication 963;American Society for Testing and Materials: Philadelphia, PA, 1988; pp 349-357. (37) Lovley, D. R.; Phillips, E. J. P. Appl. Environ. Microbiol. 1987,53, 2636-2641. (38) Cozzarelli, I. M.; Baedecker, M. J. In Water-Rock Interaction; Kharaka. Y. K.. Maest. A. S... Eds... A. A. Balkema: Rotterdam, 1992;pp 275-278. (39) Mills, A. L.: Randall, S. E. In U S . Geoloaical Survev Toxic Substances Hydrology Program-Proceedings of the technical meeting, Monterey, California, March 11-15,1991;Mallard, G. E.,Aronson, D. A,, Eds.; Water Resources Investigations Report 91-4034,US.Geological Survey, Reston, VA, 1991;pp 263-267. (40) Barbaro, J.R.; Barker, J. F.; Lemon, L.A.; Mayfield, C. I.]. Contam. Hydrol. 1992,11, 245-272. (41)Beller, H. R.; Reinhard, M.; Grbic-Galic, D. Appl. Environ. Microbiol. 1992,58,3192-3195. (42) Horowitz, A.; Shelton, D. R.; Cornell, C. P.; Tiedje, J. M. Dew. Ind. Microbiol. 1982,23,435-444. (43) Kuhn, E. P.; Suflita,J. M.; Rivera, M. D.; Young, L. Y.App1.Environ. Microbiol. 1989,55, 590-598. (44) Evans, P. J.; Mang, D. T.; Young, L. Y. Appl. Environ. Microbiol. 1991,57,450-454. (45)Parkhurst, D. L.; Thorstenson, D. C.; Plummer, L. N. PHREEQE-A computer program for geochemical calculations; Revised and

reprinted January 1985;U.S.Geological SurveyWater-Resources Investigation 80-96; U.S. Geological Survey: Reston, VA, 1985. (46)Hem, J. D. Study and interpretation of thechemicalcharacteristics of natural water, 3rd ed.; US.Geological Survey Water-Supply Paper 2254;US.Government Printing Office: Washington, DC, 1989. (47) Sulzberger, B.; Suter, D.; Siffert, C.; Banwart, S.; Stumm, W. Mar. Chem. 1989,28,127-144. (48)Canfield, D. E. Geochim. Cosmochim. Acta 1989,53,619-632. (49) Moses, C. 0.;Nordstrom, D. IC;Herman, J. S.; Mills,AL. Geochim. Cosmochim. Acta 1987,51, 1561-1571. (50)Lovley, D. R.; Phillips, E. J. P.; Lonergan, D. J. Environ. Sci. Technol. 1991,25,1062-1067. (51) Lakind, J. S.;Stone, A. Geochim. Cosmochim.Acta 1989,53,961971. (52) McMahon, P. B.; Chapelle, F. H. Nature 1991,349,233-235. (53) Barcelona, M. J.; Tomczak, D.; Lu, I.; Wrkhaus, C. Fractionation and identification of organic matter in natural and fossil-fuel contaminated aquifer systems. Presented at API-NGWA Petroleum Hydrocarbons in the Subsurface, Houston, TX,Nov 1012, 1993.

Received for review May 16, 1994. Revised manuscript received October 20, 1994. Accepted October 26, 1994.* ES9403017 @

Abstract published in Advance ACSAbstracts, December 1,1994.

VOL. 29, NO. 2, 1995 I ENVIRONMENTAL SCIENCE &TECHNOLOGY

469