Environ. Sci. Technol. 2010, 44, 3474–3480
Impact of Nanoscale Zero Valent Iron on Geochemistry and Microbial Populations in Trichloroethylene Contaminated Aquifer Materials T E R E S A L . K I R S C H L I N G , †,‡ K E L V I N B . G R E G O R Y , ‡,§ E D W I N G . M I N K L E Y , J R . , ‡,| G R E G O R Y V . L O W R Y , * ,‡,§ A N D R O B E R T D . T I L T O N * ,†,‡,⊥ Department of Chemical Engineering, Center for the Environmental Implications of NanoTechnology (CEINT), Department of Civil and Environmental Engineering, Department of Biological Sciences, and Department of Biomedical Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania 15213
Received December 10, 2009. Revised manuscript received March 18, 2010. Accepted March 22, 2010.
Nanoscale zerovalent iron (NZVI) particles are a promising technology for reducing trichloroethylene (TCE) contamination in the subsurface. Prior to injecting large quantities of nanoparticles into the groundwater it is important to understand what impact the particles will have on the geochemistry and indigenous microbial communities. Microbial populations are important not only for nutrient cycling, but also for contaminant remediation andheavymetalimmobilization.Microcosmswereusedtodetermine the effects of NZVI addition on three different aquifer materials from TCE contaminated sites in Alameda Point, CA, Mancelona, MI, and Parris Island, SC. The oxidation and reduction potential of the microcosms consistently decreased by more than 400 mV when NZVI was added at 1.5 g/L concentrations. Sulfate concentrations decreased in the two coastal aquifer materials, and methane was observed in the presence of NZVI in Alameda Point microcosms, but not in the other two materials. Denaturing gradient gel electrophoresis (DGGE) showed significant shifts in Eubacterial diversity just after the Fe0 was exhausted, and quantitative polymerase chain reaction (qPCR) analyses showed increases of the dissimilatory sulfite reductase gene (dsrA) and Archaeal 16s rRNA genes, indicating that reducing conditions and hydrogen created by NZVI stimulate both sulfate reducer and methanogen populations. Adding NZVI had no deleterious effect on total bacterial abundance in the microcosms. NZVI with a biodegradable polyaspartate coating increased bacterial populations by an order of magnitude relative to controls. The lack of broad bactericidal effect, combined with the stimulatoryeffectofpolyaspartatecoatings,haspositiveimplications for NZVI field applications. * Address correspondence to either author. Phone: 412-268-2948 (G. V. L.); 412-268-1159 (R. D. T.). Fax: 412-268-7813 (G. V. L.); 412268-7139 (R. D. T.). E-mail:
[email protected] (G. V. L.);
[email protected] (R. D. T.). † Department of Chemical Engineering. ‡ Center for the Environmental Implications of NanoTechnology. § Department of Civil and Environmental Engineering. | Department of Biological Sciences. ⊥ Department of Biomedical Engineering. 3474
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 9, 2010
Introduction Nanoscale zerovalent iron (NZVI) particles have shown promise for treating TCE contaminated groundwater, without the accumulation of chlorinated intermediates seen with other in situ treatments (1, 2). NZVI has an Fe0 core and a predominantly magnetite shell (3). Typically NZVI is stabilized by anionic polyelectrolytes or other macromolecules to reduce aggregation and improve transport in groundwater (4-9). This allows the particles to be injected into an aquifer for in situ remediation. NZVI treatments have shown promise in the laboratory and at pilot scale field sites, but there are considerable concerns about the potential hazards of nanomaterials in the environment (10). Of particular concern in this case is the bactericidal effect of NZVI in culture studies (11, 12). Regardless of how well NZVI reduces TCE in situ, most contaminated sites will rely on bioremediation as a concurrent or terminal process to meet remediation goals. Therefore, it is important to understand how NZVI will affect site geochemistry, microbial abundance and diversity, and their interrelationships. In order to effectively treat TCE source zones, large concentrations of NZVI (averaging between 1 and 5 g per liter) (2, 13) will be injected into the affected area. NZVI is 20-30 times more reactive than micrometer sized iron filings on a mass basis and therefore has the ability to rapidly reduce the local oxidation reduction potential (ORP) (3). While every site will be impacted differently by the addition of NZVI, assessing what types of changes to expect and what changes are consistent between different aquifer materials will allow effective decisions to be made at field sites. Geochemical changes in various NZVI amended aquifers have yet to be studied and while most current research focuses on how geochemistry will impact the lifetime and reactivity of NZVI, changes in the site biogeochemistry induced by NZVI are equally important. Microbial communities are of utmost importance not only to nutrient cycling (C, N, S, Fe, etc.) in groundwater systems, but also to contaminant and heavy metal reduction and immobilization (14). Reductions in bacterial abundance in the aquifer system could be detrimental to aquifer remediation, particularly if natural attenuation is occurring. NZVI is bactericidal in pure cultures of Escherichia coli, under both aerobic and anaerobic conditions, likely because of a combination of membrane disruption, the reduced state of the particle, and Fe2+ induced generation of reactive oxygen species (11, 12, 15). Particle coatings, whether engineered or natural, tend to reduce this toxicity (15). While NZVI interactions with bacteria have been studied in pure culture, it is not yet known how or whether NZVI addition to the subsurface will affect microbial abundance. Impacts of micrometer sized Fe0 on subsurface microbial communities have been studied in permeable reactive barriers (PRB), but NZVI injection is fundamentally different from Fe0 deployment in a PRB. Even though both systems react with water to generate cathodic hydrogen, a preferred electron donor for many metabolic groups of bacteria including sulfate reducers, methanogens, and reductive dehalogenators (1, 16-21), NZVI injection requires the addition of comparably lower concentrations (1-5 g/L, 0.5 wt %) of iron, distributed over a larger area and in source zones where high TCE concentrations may be present, whereas a PRB generally contains an Fe0-sand mixture between 25 and 50 wt % Fe0, confined in a limited region down gradient of a source zone (22-24). Another notable difference is the polymer coating on NZVI, which is required 10.1021/es903744f
2010 American Chemical Society
Published on Web 03/30/2010
for successful injection. If an appropriate biodegradable polymer modifier is chosen, it is possible to biostimulate while increasing particle transport. Fe0 has been used successfully as an electron donor for biological dechlorination (19-21, 25), however, a better understanding of the impact of NZVI is needed to determine whether NZVI treatment will be compatible with natural attenuation or bioaugmentation remediation treatments. Microcosms with aquifer material from three different TCE contaminated sites were used to examine the impact of field relevant concentrations of NZVI on both the biogeochemistry and microbial diversity of the material. Samples from Alameda Point, CA, Mancelona, MI, and Parris Island, SC were chosen for their wide range of geochemistry and previous TCE exposure. The change in pH, oxidation and reduction potential (ORP), sulfate concentration, and methane partial pressure were compared between the three sites upon addition of TCE and/or NZVI. The Alameda Point microcosms were also exposed to polyaspartate modified NZVI to determine the effect of biodegradable polymer coatings, and to magnetite nanoparticles to determine the effect of Fe0 content on the microbial community. The microbial community was examined by quantitative polymerase chain reaction (qPCR) and denaturing gradient gel electrophoresis (DGGE). The objectives of this study were to (1) compare NZVI-induced geochemical changes between different aquifer materials and determine which effects translate between sites, (2) determine if there is any decrease in microbial abundance indicating a toxic effect of NZVI, (3) determine how NZVI affects microbial communities with particular regard to the inhibition or stimulation of particular metabolic groups, and (4) determine whether a biodegradable polyaspartate coating alters the effect of NZVI on microbial abundance and diversity.
Materials and Methods Aquifer Materials and Nanoparticles. Microcosms contained aquifer material from three different TCE contaminated sites: Alameda, CA, Mancelona, MI, and Parris Island, SC. These samples were taken during 2006 and 2007 and were shipped in coolers to the laboratory where they were subsequently stored at 4 °C. Alameda Point material is a clayey sand with high sulfate and chloride concentrations because of its proximity to the San Francisco Bay. Mancelona material is a glacial sandy soil with particularly low organic carbon. Parris Island, also a coastal site, has a low groundwater pH (3.5). Additional site information can be found in the Supporting Information (SI). For the NZVI, reactive nanoiron particles (RNIP) and polyaspartate modified reactive nanoiron particles (mRNIP) were provided by Toda Kogyo Corp (Japan) and were stored in an anaerobic glovebox. Detailed physical and chemical properties of RNIP have been reported previously (3, 17). Magnetite particles with approximately the same primary particle size (27.5 nm) and surface characteristics as RNIP were obtained from NanoAmor (Houston, TX) and also stored in an anaerobic glovebox. Microcosm Design. Microcosms were assembled in autoclaved 160 mL serum bottles with 25 mL of wet aquifer material and 75 mL of N2 purged groundwater for each site. For all three materials, unamended (control) and NZVIamended reactors were made. A subset of each was spiked with TCE (ACS grade, Fischer Scientific, Fairlawn, NJ) to an initial concentration of 100 mg/L and the other contained no TCE. All reactors were created in duplicate for a total of eight reactors for each site evaluated at each time point. Additionally, two sets of duplicate microcosms with Alameda material were also amended with modified NZVI, modified NZVI plus TCE, or magnetite. NZVI and modified NZVI were added under a nitrogen atmosphere to a total concentration of 1.5 ( 0.1 g/L. Magnetite nanoparticles were added to a
final concentration of 2.5 g/L also under nitrogen. Serum bottles were capped with Teflon-lined butyl rubber septa. The Fe0 content of the NZVI was determined by acid digestion of the particles followed by hydrogen measurement by GCTCD and iron measurement by atomic absorption spectroscopy following the procedures of Liu and co-workers (17). The Fe0 content of the bare particles at the time of microcosm assembly was 32% for Mancelona, 28% for Parris Island, and 24% for Alameda. The modified NZVI had an Fe0 content of 33%. Microcosms were incubated on an orbital laboratory shaker (100 rpm) in the dark at 23 ( 2 °C. Microcosm Sacrifice. Initial groundwater pH and Fe0 content of the NZVI was used to estimate the reactive lifetime of the NZVI in the microcosms (defined as the time for Fe0 content to fall to 5%) based on data reported previously (17). Microcosms were sacrificed at two time points. The first set was sacrificed after approximately the end of the NZVI reactive lifetime for Alameda and Mancelona soils, (110 days and 140 days, respectively). The predicted reactive lifetime in Parris Island soil with pH 3.5 is less than 2 weeks (17), therefore the first sample point for Parris Island was delayed to 68 days to allow time for the microbial community to react as this amount of time had been previously determined to be adequate to observe changes in the microbial community. The second set of microcosms from all three sites was sacrificed at 250 days (predicted to be well after all of the Fe0 had been oxidized) to determine the potential for rebound of the geochemistry or microbial communities. Prior to sacrifice, headspace samples were analyzed for TCE and dechlorination products by gas chromatography with FID detection on an HP6890 GC with a GSQ PLOT capillary column as previously described (3). Hydrogen and methane were measured by gas chromatography using an HP6850 GC with TCD detection and Hayesep D packed column as previously described (3). Microcosms were sacrificed in an anaerobic glovebox under a nitrogen atmosphere, where aquifer material samples were removed for DNA extraction. ORP was measured with an Accumet probe (rinsed with 70% ethanol) by placing the tip of the probe into the aquifer solids slurry immediately after opening the reactors. The groundwater from each microcosm was filtered through a 0.45 µm nylon filter for pH measurements and ion chromatography. The pH of the filtered groundwater was measured using an Accumet pH probe. Chloride and sulfate concentrations were measured from the microcosm extracts using a Dionex high performance liquid chromatograph (HPLC) with an IonPac AS9HC column (Dionex, Sunnyvale, CA) according to U.S. Environmental Protection Agency Method 300.1. Molecular Biology Methods. Duplicate DNA extractions were performed on material from each of the microcosms using the Power Soil DNA Extraction Kit (MoBio, Carlsbad, CA) according to the manufacturer’s directions, with the exception that 0.5 g of aquifer material were used for each extraction. Polymerase chain reaction (PCR) was performed for denaturing gradient gel electrophoresis (DGGE) using primers BAC341F-GC and BAC534R (26), (Operon, Huntsville, AL) targeting the 16s rRNA gene. The thermocycler program was as follows: 95 °C for 1 min followed by 35 cycles of 94 °C for 1 min, an annealing temperature of 50 °C for 30 s and an elongation time of 40 s at 72 °C. PCR products from duplicate extracts from the same microcosm were combined and 50 µL was loaded into an 8% polyacrylamide DGGE gel with a denaturant (urea and formamide) concentration ranging from 45 to 65%. The gel was run using the BioRad DCode System and buffers (BioRad, Hercules, CA) for 15 h at 70 V. The gel was removed and stained with SYBR gold (Invitrogen, Carlsbad, CA) stain for 20 min. Cluster analysis on the DGGE bands was performed on the same sections of each gel using GelCompar software (Applied Maths, Austin, TX) (27). A Jaccard (band based) correlation was chosen to VOL. 44, NO. 9, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3475
TABLE 1. Microcosm Carbon Mass Balances after 250 Daysa Alameda NZVI + TCE (%) trichloroethylene acetylene ethylene ethane 1,1-dichloroethylene cis-1,2-dichloroethylene trans-1,2-dichloroethylene vinyl chloride total a
1.9 28 50 12 1.4 2.3 0.053 96
Alameda Modified NZVI + TCE (%)
Mancelona NZVI + TCE (%)
1.9 46 11 0.065 0.32 3.9 0.017 63
2.2 0.03 75 20 0.19 0.56 0.02 98
Parris Island NZVI + TCE (%) 67 18 8.3 2.7 0.086 0.019 96
Concentrations are stated in terms of the percent TCE reacted.
FIGURE 1. Change in the oxidation and reduction potential with respect to the standard hydrogen electrode in microcosms with aquifer material from Alameda, Mancelona, and Parris Island with indicated amendments. Error bars represent one standard error in the measurement. generate a similarity matrix. This comparison measures the similarity between common and different bands, and is not sensitive to band intensity. The unweighted pair group method arithmetic algorithm (UPGMA) was used to create trees (28). Quantitative polymerase chain reaction (qPCR) was performed on DNA extracts using an Applied Biosystems 7500 real time PCR thermocycler. qPCR for total Bacteria and Archea was performed using TaqMan Universal PCR Master Mix (Applied Biosystems, Foster City, CA) according to the procedures of DaSilva and Alvarez (29). Total Eubacteria quantification used forward primer BAC1364F (30), universal reverse primer PROK1492R (30), and probe TM1389F (30) with DNA extracted from an overnight culture of Escherichia coli DH5R as a standard. Total Archaea was quantified using forward primer ARCH1369F 3476
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 9, 2010
(30), reverse primer PROK1541R (30), and probe TM1389F (30) with standard DNA from Methanococcus maripaludis (ATCC 43000). The reaction mixtures consisted of 300 nM forward and reverse primers and 250 nM probe concentration in a 20 µL volume. The dsrA gene was quantified using SYBR green PCR master mix (Applied Biosystems, Foster City, CA) with forward primer dsr1F (31), and reverse primer dsr500R (31) with DNA from Desulfovibrio vulgaris (ATCC 29579) as a standard according to the procedures of Wilms and co-workers (31).
Results and Discussion Dechlorination. A carbon mass balance for TCE loss was calculated from measured headspace concentrations of TCE, ethane, ethene, and acetylene (Table 1). Concentrations of chlorinated byproducts were insignificant in TCE control
microcosms (less than 0.5 mol % of initial TCE) indicating a lack of natural attenuation. Mass balances are good (g96%) for Alameda with bare NZVI, Mancelona, and Parris Island. The mass balance for Alameda with polyaspartate modified NZVI was lower (63%). This may be caused by the increased microbial activity in those microcosms, and possibly the degradation of acetylene (32). Geochemistry. The addition of NZVI to all three aquifer materials created conditions that were significantly more reduced than the control microcosms (Figure 1) despite significant differences in the site geochemistry (SI Tables S4-S6). ORP decreased by over 400 mV by the end of the NZVI reactive lifetime in all three aquifer materials when NZVI was added. The ORP decrease for microcosms amended by both NZVI and TCE was slightly smaller than their NZVI counterparts without TCE. This result is attributed to the consumption of a fraction of the NZVI reducing equivalents to dechlorinate TCE, leaving the system in a more oxidized state. TCE had little effect on microcosm ORP by itself. After 250 days and after the Fe0 in NZVI had been oxidized, the NZVI-amended microcosms remained significantly more reduced than the control microcosms, although a partial rebound was observed in the Alameda microcosms. The polyaspartate-modified NZVI microcosms for Alameda material followed the same trend as NZVI. Alameda control microcosms remained stable while ORP in Mancelona and Parris Island microcosms decreased as a result of microbial activity in the sealed system. Also consistent through all three aquifer materials is that the groundwater pH increased when NZVI (modified or unmodified) was added (data in SI, Figure S1). Magnetite and TCE alone had little impact on pH. Both Alameda and Mancelona microcosms had an initial pH of 7.8 and reached a pH above 8.5 at the end of the NZVI reactive lifetime. Parris Island groundwater has a very low pH of 3.5 that increases to 5 with NZVI addition after 68 days. This would not be an ideal site for NZVI treatment because the reactive lifetime of Fe0 at this pH would be less than two weeks (17). The pH increase with NZVI is caused by the anaerobic corrosion of iron (Fe0 + 2H2O f Fe2+ + 2OH- + H2) (3). The increase was smaller when TCE was present as some reducing equivalents from Fe0 were required to reduce TCE, thereby decreasing OH- production. In all three aquifer materials, 1.5 g/L NZVI failed to exceed the buffering capacity of the material, as the equilibrium pH of NZVI in an unbuffered system is greater than 10.6 (17). It is noteworthy that all of these pH increases are smaller than those occurring in permeable reactive barriers, which typically reach a pH greater than 9, even with groundwater flow (16, 33). This is because the NZVI dosing (0.15 wt %) yields a much lower Fe0 to aquifer solids ratio than in PRBs where iron is present at 25-50 wt %. The comparatively low local Fe0 concentration after NZVI injection would lead to less extreme conditions, placing less selective pressure on the microbial community. Figure 2 shows that adding NZVI to the two coastal aquifer materials with high sulfate concentrations (Alameda and Parris Island) decreased sulfate concentration relative to both the groundwater (indicated by the dashed line in Figure 2) and the control microcosms. The decrease in sulfate occurred during the reactive lifetime of the NZVI, but did not decrease further after the Fe0 was depleted. In most systems, field and laboratory scale, with micrometer or larger sized Fe0, abiotic reduction of sulfate (to a reduced form) is not observed (34) and any sulfate reduction is attributed to an increase in sulfate reducing bacteria (35). A microbial role is consistent with the microcosm results, as NZVI corrosion produces a significant amount of hydrogen (data in SI Figures S3 and S5), and Fe0 has been shown to stimulate sulfate reducers (18, 22). Concentrations of hydrogen are below detection at
FIGURE 2. Sulfate reduction in microcosms with aquifer material from Alameda, Mancelona, and Parris Island with indicated amendments. Error bars represent the standard error. *, ** as in Figure 1. both microcosm sample points indicating the hydrogen produced by NZVI is consumed by the time that the first sacrifice occurs. The fact that polyaspartate-modified NZVI microcosms produced a larger decrease in sulfate is also consistent with microbial sulfate reduction, as modified NZVI caused an increase in total bacteria (discussed below). Nevertheless, the current data do not rule out abiotic sulfate attrition by interaction with NZVI. Based on the available surface area, sulfate adsorption to NZVI could produce similar decreases in free sulfate concentration. Sulfate interactions with NZVI in sterile conditions are currently under investigation. Mancelona microcosms had low initial sulfate concentrations, and therefore large increases in sulfate reducer populations are not likely. Methane was only detected in certain Alameda microcosms. After 250 days, detectable levels of methane were seen in microcosms with NZVI (pCH4 ) 0.010 ( 0.001 atm), polyaspartate-modified NZVI (pCH4 ) 0.0020 ( 0.0001 atm), and magnetite (pCH4 ) 0.0041 ( 0.0003 atm). Methane was at or below the detection limit (1 × 10-4 atm) in the control microcosms and all the microcosms with TCE added. Methane was not detected in any of the Parris Island or Mancelona microcosms. In Alameda microcosms the increase in methanogenesis with NZVI addition is consistent with the hydrogen production and more reducing conditions, but the lack of methane production in microcosms containing TCE suggests that either TCE or its degradation products (primarily acetylene and ethane) are inhibiting methanogens. TCE toxicity to methanogens has been observed at less than 100 ppm (36). Also notable is that significant acetylene production occurred in the Alameda material amended with NZVI and TCE (Table 1). Acetylene is also a known inhibitor of methanogenesis VOL. 44, NO. 9, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3477
FIGURE 3. qPCR analysis of DNA extracted from Alameda microcosms after 250 days of incubation. NZVI addition enhances sulfate reducers and Archaea, but does not decrease the overall bacterial abundance. Error bars are 95% confidence intervals. The trends in Eubacterial DNA concentrations are the same at 110 days (see SI Figure S3).
FIGURE 4. Cluster diagram of Eubacterial DGGE profiles of Alameda microcosms at the end of the NZVI reactive lifetime (110 days). One and 2 indicate duplicate microcosms. The scale bar in the upper left represents the percent similarity between profiles based on the branch length. The numbers at nodes are cophenetic correlations representing the faithfulness of each cluster by comparison to the similarity matrix. Bands included in the comparison are marked in red. (37). Acetylene was not detected after 250 days in the polyaspartate-modified NZVI microcosms, which also experienced inhibition of methanogenesis, indicating that it cannot be the sole inhibiting factor. Methanogenesis enhancement by magnetite may be attributed to the high iron requirements of methanogens (38). Dissolved iron was below detection in the Alameda point groundwater (SI). Increasing iron concentrations by NZVI or magnetite addition not only would help meet the iron requirements of methanogens, but can also remove dissolved sulfide which is toxic to methanogens. Microbial Diversity. NZVI did not decrease bacterial abundance in Alameda microcosms (Figure 3), indicating it is not universally bactericidal in aquifer systems. NZVI may only be toxic to certain groups of microorganisms, and may select for others. Additionally, in aquifer systems, toxicity may be reduced because, unlike previous culture studies which demonstrated toxicity, aquifer material contains surfaces, clays, and other components to which NZVI may adsorb, limiting contact with bacterial surfaces. A marked increase in bacterial DNA was observed when polyaspartatemodified NZVI was added to the microcosm. Polyaspartate is a biodegradable polymer that certain bacteria are able to hydrolyze to aspartate (39) increasing the available carbon and nitrogen for the microbial community. The increase in 3478
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 9, 2010
bacterial abundance suggests that the system is carbon or nitrogen limited and that the microbial community can use polyasparate as a nutrient source. NZVI stimulated sulfate reducer and methanogen populations. The Archaeal primers chosen for this study strongly target methanogenic Archaea. Archaeal DNA (Figure 3) was only detected in the Alameda microcosms amended with NZVI, polyaspartate modified NZVI and magnetite, consistent with the observed methane concentrations reported above. Fe0 filings can support the growth of methanogens through cathodic hydrogen generation (21, 40), consistent with the results for the NZVI. Archaeal DNA is below the detection limit when TCE is present likely because TCE and acetylene can be toxic for methanogenic Archaea. Copies of the dsrA gene also increased with NZVI addition, indicating that at least a portion of the observed decrease in sulfate concentration was caused by selection for sulfate reducers. The increase in dsrA was smaller when TCE was present at the beginning of the experiment, suggesting an inhibition of sulfate reducers by either TCE or one of its degradation products. These results are consistent with those obtained with Fe0 filings despite the significantly smaller weight fraction of Fe0 in the NZVI treatment. Enhancement of sulfate reducers and the precipitation of iron sulfides has been previously
FIGURE 5. Cluster diagram of Eubacterial DGGE profiles of Alameda microcosms after 250 days. The scale bar in the upper left represents the percent similarity between profiles based on the branch length. The numbers at nodes are cophenetic correlations representing the faithfulness of each cluster by comparison to the similarity matrix. Bands included in the comparison are marked in red. observed in zerovalent iron barriers, columns, and batch studies (18, 23). One notable difference between the Alameda material and some observations of permeable reactive barriers is the enhancement of methanogens (23, 34). Low methanogen populations in PRBs are attributed to high pH, or competition with other metabolic groups for hydrogen. The lower relative pH and the shorter lifetime (short-term hydrogen source) means NZVI addition may be more conducive to methanogen growth in certain soils than permeable reactive barriers. Denaturing gradient gel electrophoresis (DGGE) reveals distinct Eubacterial communities in all of the Alameda microcosms (Figure 4). Duplicate microcosms produced DGGE banding patterns that cluster together, indicating reproducibility between microcosms. After one NZVI reactive lifetime, cluster analysis sorts the microcosms into two distinct groups: those with reactive Fe0 amendment, and those without Fe0 amendment. Microcosms amended with magnetite nanoparticles have a DGGE profile that is more similar to the control, suggesting that the initial changes in diversity are caused primarily by the NZVIinduced changes in geochemical conditions, particularly ORP, hydrogen evolution, and available iron, and not by direct microbial interactions with nanoparticle surfaces. A second cluster separates the NZVI-amended microcosms. Microcosms with bare NZVI were distinct from those amended with polyaspartate-modified NZVI (modified NZVI). TCE has less of an impact on DGGE profiles in the short term than NZVI. After more than one NZVI lifetime (250 days) (Figure 5), the microbial communities have changed and no longer cluster into the distinct groups with and without Fe0 observed after 110 days. The DGGE banding patterns remain reproducible between the treatments. The presence of two to three darker (relative to the control) bands in the NZVI and modified NZVI banding patterns indicates the community changes are likely the result of decreased abundance of several populations at the expense of the specific enrichment of another, especially since qPCR does not reveal a significant decrease of Eubacterial sequences recovered. Additionally, the DGGE profiles among the different amendments are more dissimilar after 250 days suggesting the communities have continued to diverge. Implications. The addition of NZVI to an aquifer can be expected to create reducing conditions and produce hydrogen which will enhance sulfate reduction, if sulfate is present, and methanogenesis. Geochemical evolution in NZVI amended aquifer material causes changes in the Eubacterial diversity in the short term at NZVI concentra-
tions expected to be applied in the field. The fact that NZVI does not decrease microbial abundance, and that polyaspartate coatings can stimulate microbial growth, has positive implications for field implementation with little detrimental ecological side effects. The stimulatory effect of polyaspartate coatings could be used to combine NZVI treatment with bioremediation.
Acknowledgments We thank Dr. Pedro Alvarez, Rice University, for assistance with qPCR methodology. This material is based on work supported by the National Science Foundation (BES-068646), the U.S. Environmental Protection Agency (R833326 and the NSF Cooperative Agreement EF-0830093, Center for the Environmental Implications of NanoTechnology (CEINT). Any opinions, findings, conclusions, or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NSF or the EPA. This work has not been subjected to EPA review and no official endorsement should be inferred.
Supporting Information Available Additional data for site geochemistry, TCE and product concentrations, pH changes, and total bacteria after one iron lifetime. This material is available free of charge via the Internet at http://pubs.acs.org.
Literature Cited (1) Reardon, E. J.; Fagan, R.; Vogan, J. L.; Przepiora, A. Anaerobic corrosion reaction kinetics of nanosized iron. Environ. Sci. Technol. 2008, 42 (7), 2420–2425. (2) Zhang, W.; Elliott, D. W. Applications of iron nanoparticles for groundwater remediation. Remediation 2006, 16 (2), 7–21. (3) Liu, Y.; Majetich, S. A.; Tilton, R. D.; Sholl, D. S.; Lowry, G. V. Tce dechlorination rates, pathways, and efficiency of nanoscale iron particles with different properties. Environ. Sci. Technol. 2005, 39 (5), 1338–1345. (4) Kim, H.-J.; Phenrat, T.; Tilton, R. D.; Lowry, G. V. Fe0 nanoparticles remain mobile in porous media after aging due to slow desorption of polymeric surface modifiers. Environ. Sci. Technol. 2009, 43 (10), 3824–3830. (5) Phenrat, T.; Liu, Y.; Tilton, R. D.; Lowry, G. V. Adsorbed polyelectrolyte coatings decrease Fe0 nanoparticle reactivity with tce in water: Conceptual model and mechanisms. Environ. Sci. Technol. 2009, 43 (5), 1507–1514. (6) Saleh, N.; Phenrat, T.; Sirk, K.; Dufour, B.; Ok, J.; Sarbu, T.; Matyjaszewski, K.; Tilton, R. D.; Lowry, G. V. Adsorbed triblock copolymers deliver reactive iron nanoparticles to the oil/water interface. Nano Lett. 2005, 5 (12), 2489–2494. (7) Sirk, K. M.; Saleh, N. B.; Phenrat, T.; Kim, H.-J.; Dufour, B.; Ok, J.; Golas, P. L.; Matyjaszewski, K.; Lowry, G. V.; Tilton, R. D. VOL. 44, NO. 9, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
3479
(8)
(9)
(10)
(11)
(12)
(13) (14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
(23)
(24)
3480
Effect of adsorbed polyelectrolytes on nanoscale zero valent iron particle attachment to soil surface models. Environ. Sci. Technol. 2009, 43 (10), 3803–3808. He, F.; Zhao, D. Preparation and characterization of a new class of starch-stabilized bimetallic nanoparticles for degradation of chlorinated hydrocarbons in water. Environ. Sci. Technol. 2005, 39 (9), 3314–3320. Tiraferri, A.; Sethi, R. Enhanced transport of zerovalent iron nanoparticles in saturated porous media by guar gum. J. Nanopart. Res. 2009, 11 (3), 635–645. Wiesner, M. R.; Lowry, G. V.; Alvarez, P.; Dionysiou, D.; Biswas, P. Assessing the risks of manufactured nanomaterials. Environ. Sci. Technol. 2006, 40 (14), 4336–4345. Auffan, M. l.; Achouak, W.; Rose, J. r. m.; Roncato, M.-A.; Chanea´c, C.; Waite, D. T.; Masion, A.; Woicik, J. C.; Wiesner, M. R.; Bottero, J.-Y. Relation between the redox state of iron-based nanoparticles and their cytotoxicity toward Escherichia coli. Environ. Sci. Technol. 2008, 42 (17), 6730–6735. Lee, C.; Kim, J. Y.; Lee, W. I.; Nelson, K. L.; Yoon, J.; Sedlak, D. L. Bactericidal effect of zero-valent iron nanoparticles on Escherichia coli. Environ. Sci. Technol. 2008, 42 (13), 4927–4933. Otto, M.; Floyd, M.; Bajpai, S. Nanotechnology for site remediation. Remediation 2008, 19 (1), 99–108. Valls, M.; Lorenzo, V. Exploiting the genetic and biochemical capacities of bacteria for the remediation of heavy metal pollution. FEMS Microbiol. Rev. 2002, 26 (4), 327–338. Li, Z.; Greden, K.; Alvarez, P. J. J.; Gregory, K. B.; Lowry, G. V., Adsorbed polymer and NOM limits adhesion and toxicity of nano scale zero-valent iron (NZVI) to Escherichia coli. Environ. Sci. Technol. 2010 in press. Liang, L.; Korte, N.; Gu, B.; Puls, R.; Reeter, C. Geochemical and microbial reactions affecting the long term performance of in situ ‘iron barriers’. Adv. Environ. Res. 2000, 4, 273–386. Liu, Y.; Lowry, G. V. Effect of particle age (Fe0 content) and solution pH on NZVI reactivity: H2 evolution and TCE dechlorination. Environ. Sci. Technol. 2006, 40 (19), 6085–6090. Dinh, H. T.; Kuever, J.; Muszmann, M.; Hassel, A. W.; Stratmann, M.; Widdel, F. Iron corrosion by novel anaerobic microorganisms. Nature 2004, 427 (6977), 829–832. Gregory, K. B.; Mason, M. G.; Picken, H. D.; Weathers, L. J.; Parkin, G. F. Bioaugmentation of Fe(0) for the remediation of chlorinated aliphatic hydrocarbons. Environ. Eng. Sci. 2000, 17 (3), 169–181. Lampron, K. J.; Chiu, P. C.; Cha, D. K. Reductive dehalogenation of chlorinated ethenes with elemental iron: The role of microorganisms. Water Res. 2001, 35 (13), 3077–3084. Weathers, L. J.; Parkin, G. F.; Alvarez, P. J. Utilization of cathodic hydrogen as electron donor for chloroform cometabolism by a mixed, methanogenic culture. Environ. Sci. Technol. 1997, 31 (3), 880–885. Gu, B.; Phelps, T. J.; Liang, L.; Dickey, M. J.; Roh, Y.; Kinsall, B. L.; Palummbo, A. V.; Jacobs, G. K. Biogeochemical dynamics in zero valent iron columns: Implications for permeable reactive barriers. Environ. Sci. Technol. 1999, 33, 2107–2177. Gu, B.; Watson, D. B.; Wu, L.; Phillips, D. H.; White, D. C.; Zhou, J. Microbiological characteristics in a zero-valent iron reactive barrier. Environ. Monit. Assess. 2001, 77, 293–309. Scherer, M. M.; Richter, S.; Valentine, R. L.; Alvarez, P. J. J. Chemistry and microbiology of permeable reactive barriers for in situ groundater clean up. Crit. Rev Microbiol. 2000, 26 (4), 221–264.
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 9, 2010
(25) Novak, P. J.; Daniels, L.; Parkin, G. F. Enhanced dechlorination of carbon tetrachloride and chloroform in the presence of elemental iron and Methanosarcina barkeri, Methanosarcina thermophila, or Methanosaeta concillii. Environ. Sci. Technol. 1998, 32 (10), 1438–1443. (26) Muyzer, G.; de Waal, E. C.; Uitterlinden, A. G. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction amplified genes coding for the 16s rDNA. Appl. Environ. Microbiol. 1993, 59 (3), 695– 700. (27) el Fantroussi, S.; Verschuere, L.; Verstraete, W.; Top, E. M. Effect of phenylurea herbicides on soil microbial communities estimated by analysis of 16s rRNA gene fingerprints and community-level physiological profiles. Appl. Environ. Microbiol. 1999, 65 (3), 982–988. (28) Smalla, K.; Oros-Sichler, M.; Milling, A.; Heuer, H.; Baumgarte, S.; Becker, R.; Neuber, G.; Kropf, S.; Ulrich, A.; Tebbe, C. C. Bacterial diversity of soils assessed by DGGE, t-RFLP and SSCP fingerprints of PCR-amplified 16s rRNA gene fragments: Do the different methods provide similar results. J. Microbiol. Methods 2007, 69 (3), 470–479. (29) Da Silva, M. L. B.; Alvarez, P. J. J. Assessment of anaerobic benzene degradation potential using 16s rRNA gene-targeted real-time pcr. Environ. Microbiol. 2007, 9, 72–80. (30) Suzuki, M. T.; Taylor, L. T.; DeLong, E. F. Quantitative analysis of small-subunit rRNA genes in mixed microbial populations via 5′-nuclease assays. Appl. Environ. Microbiol. 2000, 66, 4605– 4614. (31) Wilms, R.; Sass, H.; Kopke, B.; Cypionka, H.; Engelen, B. Methane and sulfate profiles within the subsurface of a tidal flat are reflected by the distribution of sulfate reducing bacteria and methanogenic archaea. FEMS Microbiol. Ecol. 2007, 59 (3), 611– 621. (32) Schink, B. Fermentation of acetylene by an obligate anaerobe Pelobacter acetylenicus sp. Nov. Arch. Microbiol. 1985, 142 (3), 295–301. (33) Wilkin, R. T.; Puls, R., W.; Sewell, G. W. Long-term performance of permeable reactive barriers using zero-valent iron: Geochemical and microbiological effects. Ground Water 2003, 41 (4), 493–503. (34) Da Silva, M. L. B.; Johnson, R. L.; Alvarez, P. J. J. Microbial characterization of groundwater undergoing treatment with a permeable reactive iron barrier. Environ. Eng. Sci. 2007, 24 (8), 1122–1127. (35) Kober, R.; Schlicker, O.; Ebert, M.; Dahmke, A. Degradation of chlorinated ethylenes by Fe0: Inhibition processes and mineral precipitation. Environ. Geol. 2002, 41, 644–652. (36) Sponza, D. T. Toxicity and treatability of carbontetrachloride and tetrachloroethylene in anaerobic batch cultures. Int. Biodeterior. Biodegrad. 2003, 51, 119–127. (37) Oremland, R. S.; Taylor, B. F. Inhibition of methanogenesis in marine sediments: Validity of the acetylene reduction assay for anaerobic microcosms. Appl. Microbiol. 1975, 30 (4), 707–709. (38) Jarrell, K. F.; Kalmakoff, M. L. Nutritional requirements of the methanogenic archaebacteria. Can. J. Microbiol. 1988, 34 (5), 557–576. (39) Obst, M.; Steinbuchel, A. Microbial degradation of poly(amino acid)s. Biomacromolecules 2004, 5 (4), 1166–1176. (40) Daniels, L.; Belay, N.; Rajogopal, B. S.; Weimer, P. J. Bacterial methanogenesis and growth from CO2 with elemental iron as the sole source of electrons. Science 1987, 237 (3), 509.
ES903744F
