Environ. Sci. Technol. 2001, 35, 4783-4788
Hydrogen Concentrations in Sulfate-Reducing Estuarine Sediments during PCE Dehalogenation CHRISTOPHER S. MAZUR AND W. JACK JONES* Ecosystems Research Division, National Exposure Research Laboratory, U.S. Environmental Protection Agency, Athens, Georgia 30605
Despite recent progress made evaluating the role of hydrogen (H2) as a key electron donor in the anaerobic remediation of chloroethenes, few studies have focused on the evaluation of hydrogen thresholds relative to reductive dehalogenation in sulfidogenic environments. Competition for hydrogen exists among microbial populations in anaerobic sediments, and direct evidence indicates that lower hydrogen thresholds are observed with more energetically favorable electron-accepting processes. This study examined aqueous hydrogen concentrations associated with sulfate reduction and perchloroethylene (PCE) dehalogenation in anoxic estuarine sediment slurry microcosms and evaluated the competition for H2-reducing equivalents within these systems. After an initial lag period of 13 days, PCE was reductively transformed to trichloroethylene (TCE). During the time of continuous PCE dehalogenation, a significantly (P < 0.05) lower hydrogen concentration (0.5 nM) was observed in the sediment slurries amended with PCE as compared to slurries without PCE (0.8 nM). Sulfate reduction to sulfide was observed in all sediment slurries, but in microcosms actively dechlorinating PCE, the amount of reducing equivalents directed to sulfate reduction was approximately half the amount in sediment slurries without PCE. These findings provide evidence that a lower hydrogen threshold exists in anoxic estuarine sediment slurries with PCE as a terminal electron acceptor as compared to sediment slurries in which sulfate reduction was the predominant electronaccepting process. Furthermore, our results utilizing the inhibitor molybdate indicated that H2-utilizing methanogens may have the potential to effectively compete with dechlorinators for hydrogen when sulfate reduction is initially inhibited.
Introduction Perchloroethylene (PCE) is a widely used chlorinated solvent and a common contaminant in U.S. groundwater. Aquifers containing PCE-contaminated water may be used as a source for municipal water supplies (1). Since PCE is a suspected carcinogen, the fate of PCE in natural environments has become a primary concern (2). Subsurface environments are commonly anoxic, and the primary PCE transformation * Corresponding author e-mail:
[email protected]; phone: (706)355-8228; fax: (706)355-8202. 10.1021/es0110372 Not subject to U.S. Copyright. Publ. 2001 Am. Chem. Soc. Published on Web 11/16/2001
pathway under such conditions is via sequential reductive dehalogenation (3). Reductive dehalogenation is defined as the addition of electrons to a molecule with the concomitant removal of a halogen substituent. Although many studies have been conducted to elucidate the biological fate of PCE under anoxic conditions (4-11), information remains limited concerning the fate of chlorinated organic compounds in environments containing sulfate as a prevalent electron acceptor (12, 13). Hydrogen (H2) is an important intermediate in the anaerobic degradation of organic matter (14). Electrons liberated during the oxidation of organic matter by primary fermenting organisms may be transferred to protons during metabolism, forming H2. Fermentative and other microorganisms that oxidize organic compounds in the absence of oxygen often require low hydrogen partial pressures for energy conservation and therefore may benefit from the “syntrophic” association with hydrogen-consuming organisms (15). Lovley and Goodwin (16) provided evidence that the steady-state hydrogen concentration observed in aquatic sediments is a result of microbial oxidation of hydrogen coupled to specific terminal electron-accepting processes. On the basis of the “hydrogen threshold” concept, microorganisms that couple oxidative metabolism to an energetically more favorable electron acceptor will achieve a minimum hydrogen concentration (i.e., threshold) that is lower than that attained by an organism using an electron acceptor of lower energy yield (Table 1) (17, 18). The hydrogen threshold model thus provides an explanation for the redox zonation observed among iron-reducing, sulfate-reducing, methanogenic, and other hydrogen-consuming organisms in natural environments in which hydrogen production is rate limiting (19-22). Microbial competition for hydrogen plays an important role in the natural attenuation of chloroethylenes, and recent studies have focused on the role of hydrogen as a key electron donor for the reductive transformation of these compounds. Several of these studies, primarily using freshwater sediments and mixed-culture systems, compared the competition for hydrogen between dehalogenators and other hydrogenconsuming organisms such as methanogens (23, 24). Using a PCE-dehalogenating enrichment culture, Smatlak et al. (25) reported nearly a 10-fold lower H2 half-velocity constant for the process of dehalogenation than for methanogenesis. Additional studies have been conducted to evaluate the selective enhancement of PCE dechlorination through the use of fermentable substrates that degrade slowly and provide a steady release of low levels of hydrogen (26, 27). Yang and McCarty (28) recently observed a lower hydrogen concentration during the reductive transformation of cis-1,2 dichloroethene (cis-DCE) as compared to methanogenic conditions in batch reactors. According to the hydrogen threshold concept, one would predict a lower hydrogen concentration during the reductive dehalogenation of a chloroethylene than either sulfate reduction or methanogenesis (Table 1). However, the H2 values reported by Yang and McCarty during cis-DCE transformation (2.2 ( 0.9 nM) were higher than H2 values measured for sulfate reduction (1.0-1.5 nM) in other studies (16, 20). Additional research is needed to better understand hydrogen thresholds associated with the fate of chloroethylenes in a sulfate-reducing environment. Investigations of reductive dehalogenation of chloroethylenes under sulfidogenic conditions are almost exclusively from pure-culture studies (29-31). A recent study conducted using pure and enriched microbial cultures clearly VOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4783
TABLE 1. Energetics of Various Environmentally Relevant Hydrogen-Utilizing Reactions process
reaction
energetics (kJ/mol of H2)
Mn(IV) reduction Fe(III) reduction PCE reduction TCE reduction cis-DCE reduction sulfate reduction methanogenesis
2Mn4+ + 2H2 f 2Mn2+ + 4H+ 2Fe3+ + H2 f 2Fe2+ +2H+ PCE + H2 + H+ f TCE + Cl- + 2H+ TCE + H2 + H+ f cis-DCE + Cl- + 2H+ cis-DCE + H2 + H+ f VC + Cl- + 2H+ SO42- + 4H2 f S2- + 4H2O CO2 + 4H2 f CH4 + 2H2O
-334a -228a -173b -169b -139b -38a -33a
a Ref 18. b Calculated according to ref 17: organic substrate + H f organic product + H+ + Cl-. Aqueous ∆G ° values were used for organic 2 f substrates at 25 °C, pH 7. ∆Gf°(H+) ) -39.9 kJ/mol; ∆Gf°(Cl-) ) -131.3 kJ/mol.
demonstrated a lower hydrogen concentration during PCE dechlorination as compared to either sulfidogenic or methanogenic cultures (32). However, such studies conducted with microbial isolates do not replicate the competitive nature for hydrogen observed in natural environments. The transformation of PCE to cis-1,2 DCE has also been reported in a sulfate-reducing culture inoculated with anaerobic sludge from a wastewater treatment facility, but no hydrogen concentration data were presented (33). The objectives of the present study were to compare hydrogen concentrations in sulfate-reducing estuarine sediments in the presence and absence of PCE and to examine the effects of the sulfatereducing inhibitor molybdate on terminal electron-accepting processes.
Materials and Methods Chemicals. Neat solutions of perchloroethylene (PCE), trichloroethylene (TCE) (Aldrich Chemical Co, Milwaukee, WI), and cis-1,2-dichloroethylene (cis-1,2-DCE) (Supelco Park, Bellefonte, PA) were used to prepare analytical standards. Scott Speciality Gases (99%, Aldrich Chemical Co.) were used to prepare methane standards. Hydrogen standards were prepared using Grade 5.5 (99.999%) hydrogen (BOC Group Inc., Murray Hill, NJ) and were calibrated using a 500 ppb (v/v) hydrogen standard purchased from Scott-Marrin, Inc. (Riverside CA). Zinc acetate, diamine reagent (n,n-dimethylp-phenylenediamine sulfate), and molybdate (sodium salt, 99%) were purchased from Sigma Chemical Co. (St. Louis, MO). Analytical-grade sodium sulfide (Na2S‚9H2O) crystals were purchased from J. T. Baker Chemical Co. (Phillipsburgh, NJ). Sediment Collection and Microcosm Incubation. Sediment samples were collected as cores at depths ranging from 5 to 17 cm in the intertidal zone of a salt-marsh creek on Sapelo Island, GA. Sediments were placed in a sealed container and stored at 4 °C under a nitrogen atmosphere until use. All slurries were prepared in an anaerobic chamber containing a 1% hydrogen atmosphere. The sediment material was composited, passed through a 1-mm sieve, and thoroughly mixed with anoxic (N2 sparged) site-water (conductivity equivalent to 50% seawater) to achieve an approximate 10 gL-1 slurry. Forty-five milliliters of the slurry was dispensed into 60-mL amber serum bottles that were subsequently sealed with Teflon-lined, butyl-rubber septa and aluminum crimp caps. PCE was added to bottles at a final concentration of approximately 50 µM from a saturated aqueous stock solution. The serum bottles were subsequently transferred to a 25 °C incubator and were continuously mixed on a shaker table at a rate of 100 rpm. Sterile controls were prepared by autoclaving sediment slurry microcosms at 120 °C for 30 min for 3 consecutive days. All experimental microcosms were prepared in triplicate. Molybdate was added at a final concentration of 13.5 mM to some sediment microcosms to examine the effects of inhibition of sulfate reduction on PCE dehalogenation. The composited sediment 4784
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001
sample used in all studies had a total organic carbon content of 4.2% (w/w). Analytical Techniques. Gas chromatography techniques were used to quantify PCE, TCE, cis-1,2 DCE, methane, and hydrogen. Subsamples (50 µL) of the sediment slurry were extracted into hexane (1 mL) and analyzed for chlorinated aliphatics by gas chromatography. PCE and TCE were analyzed by direct injection (1 µL) using a 5890 HewlettPackard gas chromatograph equipped with an electron capture detector (ECD). Separation was achieved with a 30 m × 0.32 mm i.d., 0.25 mm film thickness DB-5 capillary column (J&W Scientific, Folsom, CA). The detector and injection-port temperatures were 300 and 150 °C, respectively. The temperature program was as follows: 60 °C, hold 5 min, 10 °C min-1 to 90 °C, no hold, 40 °C min-1 to 220 °C. Less chlorinated dehalogenation products, including cis-DCE, were quantified by gas chromatography using a split injection (8:1 ratio) and a 30 m DB-624 capillary column (0.25 mm i.d., 1.4 µm film thickness; J&W Scientific), and an ECD detector. The detector and injection-port temperatures were 250 and 150 °C, respectively. The temperature program for separation of the lower chlorinated aliphatics was as follows: 65 °C, hold 8 min, 40 °C min-1 to 200 °C, hold for 2 min. Headspace samples (200 µL) were analyzed for methane by gas chromatography using a Hewlett-Packard 5890 gas chromatograph equipped with a flame ionization detector. Methane was separated on a Poropak N column (6 ft × 1/8 in. o.d., 80/100 mesh; Alltech, Deerfield, IL) at 50 °C with a N2 carrier flow of 15 mL min-1. Sulfate analysis was conducted by ion chromatography using a Dionex instrument equipped with a CD20 conductivity detector and a Dionex Ionpac AS-16 column. A degassed 0.1 M sodium hydroxide eluent at a flow rate of 1.5 mL min-1 was used to separate anions. Microcosm subsamples were first centrifuged and then passed through a Dionex-Onguard silver filter before analysis to remove Cl- anion. Ferric iron (Fe3+) and ferrous iron (Fe2+) species were analyzed in both the pore water and the solid-phase using the ferrozine assay described in Viollier et al. (34). A 1-mL sample of sediment slurry was directly filtered (0.22 µm syringe filter) into the ferrozine solution to avoid exposure to the air. A 0.5 M HCl extraction procedure described in Kostka and Luther (35) was implemented to determine the availability of reactive amorphous iron(III) oxides from the solid phase. Hydrogen concentration was measured by direct injection of headspace samples into a Trace Analytical model RGA3 gas chromatograph (Menlo Park, CA) equipped with 1 mL loop and a reduction gas detector (RGD2). Hydrogen was sampled every 5-7 days to ensure complete equilibrium during the continuous production and consumption of aqueous hydrogen in the sediment material (36). The concentration of dissolved hydrogen H2(dissolved) in the aqueous solution was calculated by assuming equilibrium as follows: H2(dissolved) ) (LP)/(RT), where H2(dissolved)
TABLE 2. Calculation of Electron Equivalents Transferred via Dehalogenation of PCE to TCE and Reduction of Sulfate to Sulfide As Estimated from Product (Sulfide, TCE) Formation in Estuarine Sediment Microcosms with and without PCE PCE + H+ + 2(e-) f TCE + Cl SO42- + 9H+ + 8(e-) f HS- + 4H2O day 27 data HS-
e-
(µmol equivalents) TCE (µmol e- equivalents) total e- transferred (µmol)
FIGURE 1. (A) Reaction kinetics of PCE dehalogenation and TCE formation in estuarine sediment microcosms not previously exposed to chloroethylenes. (B) Comparison of dissolved hydrogen concentrations in estuarine sediment microcosms in the absence and presence of PCE dehalogenation. (C) Dissolved sulfide production in estuarine sediment slurries in the presence and absence of PCE dehalogenation. All results are the mean of triplicate determinations. is the concentration of dissolved H2 in moles per liter; L is the Ostwald coefficient for H2 (37); R is the universal gas constant; P is pressure (atm); and T is the temperature (K). All hydrogen concentration data are reported as the aqueousphase concentration. Dissolved sulfide concentrations were measured from aqueous-phase samples using a modified Cline assay (38). A 1-mL sample of sediment slurry was directly filtered (0.22 µm syringe filter) into an acidified 2% (w/v) zinc acetate trapping solution to avoid exposure to the air. The sample was then amended with the diamine reagent, vortexed, and incubated in the dark for 30 min before measuring the absorbance at 670 nm. All sulfide standards were prepared by thoroughly washing and drying sodium sulfide crystals.
Results and Discussion Reduction of sulfate was identified to be the predominant electron-accepting process in the natural estuarine sediment slurries. Sulfate concentrations remained in excess of 10 mM throughout the course of the experiment. Ferric iron (Fe3+) was not detected in any of the samples in either the pore water or the solid-phase material when extracted for amorphous iron(III) oxides. With the exception of the molybdate inhibition studies, only trace levels of methane were produced during the course of the experiment. Dissolved hydrogen concentrations measured in PCEamended slurries (PCE present) were compared to hydrogen levels observed in natural estuarine sediment slurries (PCE absent). Results presented in Figure 1A indicate a lag time
2(e-) transferred/TCE formed 8(e-) transferred/HS- formed PCE absent
PCE present
4.1 0.0 4.1
2.2 6.6 8.8
of approximately 13 days before the detection of PCE dehalogenation and TCE product formation. PCE transformation was nearly complete after 32 days. Mass balance calculations indicated that the principal product of the initial transformation of PCE was TCE. Neither cis-1,2-DCE nor other less-chlorinated products were detected at any time during the initial 32 days of PCE transformation. Autoclaved controls showed no transformation of PCE. During the initial 5 days of incubation, dissolved hydrogen levels were initially high (>1 nM) in all microcosms due to carryover of hydrogen gas from the anaerobic chamber during initial microcosm setup. However, during the period of appreciable PCE transformation (days 21-32), a significantly (P < 0.05) lower hydrogen concentration (0.4 nM) was observed in the PCE-amended slurries in comparison to natural slurries without PCE (0.6 nM) (Figure 1B). At the time of complete disappearance of the added PCE (day 39), the concentration of hydrogen increased to the level observed in the sediment slurries without PCE. The variability observed in the hydrogen concentration data during the initial 20 days of incubation may be attributed to either the variation among replicate bottles with regard to the onset and completion of PCE dehalogenation or the lack of development of steady-state conditions in the microcosms. Products of redox reactions may be informative indicators of the predominant terminal electron-accepting processes within a sedimentary environment (15). The concentration of dissolved sulfide (Figure 1C) and TCE product formation (Figure 1A) were used to assess the net consumption of reducing equivalents within slurry microcosms. On the basis of the maximum values measured for ferrous iron (8mM) and dissolved sulfide and using the solubility constants of Davison (39), these sediment microcosms were determined to be undersaturated with respect to amorphous ferrous sulfide. The mean pH for all sediment slurries during the course of the experiment was 7.9 ( 0.1 (mean ( one standard deviation) and did not vary by more than 0.3 pH unit throughout the course of the experiment, thus allowing direct comparison of dissolved sulfide species in PCE-present versus PCE-absent slurries. The total electron equivalents necessary for the processes of sulfide production and TCE formation were calculated (Table 2) assuming (i) 8 electrons were transferred during the reduction of sulfate to sulfide and (ii) 2 electrons were transferred during reduction of PCE to TCE. At day 27, approximately 75% of the initially added PCE (50 µM) was transformed to TCE in the PCE-present microcosms. The concentration of sulfide measured in PCE-absent slurries equated to approximately 4.1 µmol of electron equivalents transferred via sulfate reduction and was nearly twice the amount of electron equivalents calculated for sulfate reduction in the PCE-present sediment slurries (approximately 2.2 µmol of e- equiv). However, the net amount of reduction products detected in the PCE-present slurries (8.8 µmol of e- equiv, HS- + TCE) indicated that a greater number of reducing equivalents had been transferred than in the PCEVOL. 35, NO. 24, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4785
FIGURE 2. Steady-state aqueous hydrogen concentrations in estuarine sediment microcosms in the absence of PCE and with continual addition of PCE. Results are the mean of duplicate determinations. absent slurries (4.1 µmol of e- equiv, HS-). These results further suggest that PCE dehalogenation was the energetically preferred terminal electron-accepting process. Hydrogen serves as a key intermediate in the microbial metabolism of organic matter and steady-state hydrogen kinetics in anaerobic sediments are reached when the concentration of hydrogen consumption equals the rate of hydrogen production. In natural ecosystems, competition exists for the limited concentration of hydrogen produced through fermentative processes, and the production of hydrogen in microbial systems is generally believed to be the rate-limiting step. Thus, the hydrogen threshold observed is a result of a predominant hydrogen-consuming process and may be used as a redox indicator (12). However, a recent study conducted by Jakobsen and Postma (40) indicates that microbial competition for hydrogen-reducing equivalents may not always result in complete exclusion of one process over another and concomitant redox processes may occur. Under these circumstances, the reliability of hydrogen as a redox indicator can be questionable, and a partial equilibrium approach may be more appropriate when assessing simultaneous geochemical processes often found at the iron/sulfate reduction interface (41). To determine if the PCE-amended slurries in our study could maintain a lower steady-state hydrogen concentration than PCE-absent slurries, the dehalogenating slurries were refed to the original PCE concentration every 3-5 days to ensure that these microcosms were not limited by the availability of the chlorinated electron acceptor. The result was that a significantly (P < 0.02) lower hydrogen concentration (mean of 0.5 nM) was observed and maintained in the PCE-amended microcosms (Figure 2) than in the PCEabsent slurries (mean of 0.8 nM). The continuous addition of PCE to the sediment slurries resulted in accumulation of cis-1,2 DCE as the major transformation product during this incubation period. The rate of hydrogen production, microbial population size, and availability of electron acceptors are all factors that contribute to the observed H2 concentrations in anaerobic sediments (23). On the basis of free energy, the sequential reduction of PCE via hydrogen oxidation should theoretically result in a lower hydrogen concentration than sulfate reduction (Table 1). Our aqueous hydrogen values for sulfidogenesis (0.8 nM) were comparable to the values observed by Lovley and Goodwin for sulfate reduction (1.01.5 nM) (16), and reductive dehalogenation of PCE within 4786
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 24, 2001
our experimental microcosms resulted in a lower aqueous hydrogen value (0.5 nM). These results indicate that the observed hydrogen values proved to be a useful indicator of the predominant terminal redox process when hydrogen production is rate limiting. To understand the possible role of sulfate-reducing organisms with regard to dehalogenation of PCE in our system, we conducted a comparative study using the specific sulfate-reducing inhibitor molybdate. In one set of experiments, molybdate was added to estuarine sediment slurries actively dehalogenating PCE to cis-1,2 DCE. Another set of experiments was performed in which molybdate was added with the initial amendment of PCE. Slurries that were actively dehalogenating PCE and then dosed with molybdate continued upon re-feed to transform PCE to cis-1,2 DCE at the same rate while in the presence of the inhibitor (data not shown). In these microcosms, sulfate reduction was inhibited, only trace levels of methane were detected, and dissolved hydrogen concentrations remained at low levels (
