Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater

G Model

ARTICLE IN PRESS

HAZMAT-18157; No. of Pages 8

Journal of Hazardous Materials xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat

Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater Lior Avishai, Hagar Siebner, Ofer Dahan ∗ , Zeev Ronen ∗ Department of Environmental Hydrology & Microbiology, Zuckerberg Institute for Water Research, Blaustein Institutes for Desert Research, Ben-Gurion University of the Negev, Sede Boker Campus, 8499000, Israel

h i g h l i g h t s • • • •

Integrated in-situ remediation treatment for soil, vadose zone and groundwater. Turning the topsoil into an efficient bioreactor for perchlorate degradation. Treating perchlorate leachate from the deep vadose zone in the topsoil. Zero effluents discharge from the remediation process.

a r t i c l e

i n f o

Article history: Received 16 July 2016 Received in revised form 31 October 2016 Accepted 2 November 2016 Available online xxx Keywords: Soil flushing In-situ remediation Perchlorate Vadose-zone monitoring

a b s t r a c t In this study, we examined the ability of top soil to degrade perchlorate from infiltrating polluted groundwater under unsaturated conditions. Column experiments designed to simulate typical remediation operation of daily wetting and draining cycles of contaminated water amended with an electron donor. Covering the infiltration area with bentonite ensured anaerobic conditions. The soil remained unsaturated, and redox potential dropped to less than −200 mV. Perchlorate was reduced continuously from ∼1150 mg/L at the inlet to ∼300 mg/L at the outlet in daily cycles. Removal efficiency was between 60 and 84%. No signs of bioclogging were observed during three operation months although occasional iron reduction observed due to excess electron donor. Changes in perchlorate reducing bacteria numbers were inferred from an increased in pcrA gene abundances from ∼105 to 107 copied per gram at the end of the experiment indicating the growth of perchlorate-reducing bacteria. We proposed that the topsoil may serve as a bioreactor to treat high concentrations of perchlorate from the contaminated groundwater. The treated water that infiltrates from the topsoil through the vadose zone could be used to flush perchlorate from the deep vadose zone into the groundwater where it is retrieved again for treatment in the topsoil. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Perchlorate is an environmental pollutant that is often attributed to ammonium perchlorate production in the explosives industry [1–3]. Its high solubility (220 g/l) and limited sorption to soil and sediments make it very mobile in the vadose zone and groundwater [4,5]. Perchlorate biodegradation in soils is possible under reducing conditions, in the presence of an electron donor and bacteria capable of perchlorate reduction [6–8]. The bacteria can use perchlorate as an alternative electron acceptor for metabolism in the absence of oxygen. Electron donor sources can be organic, such as acetate and ethanol, or inorganic, such

∗ Corresponding authors. E-mail addresses: [email protected] (O. Dahan), [email protected] (Z. Ronen).

as hydrogen gas [8,9]. The final byproducts of perchlorate reduction through the following general path—chloride and oxygen—are nontoxic: ClO4 − (perchlorate) → ClO3 − (chlorate) → ClO2 − (chlorite) → Cl− (chloride) + O2 (oxygen) [9]. Although biodegradation of perchlorate in deep soils is feasible [6], from a technical standpoint, efficient performance in the natural deep unsaturated zone is a challenge. It requires: (a) maintaining reducing conditions and complete absence of oxygen, (b) a substantial supply of efficient electron donor to the deep part of the unsaturated zone, (c) maintaining sufficient moisture to sustain microbial community development, and (d) presence of indigenous perchlorate-degrading bacteria in the target horizons that are being subjected to the cleanup operation. Frankel and Owsianiak [10] succeeded to remediate a shallow vadose zone (3–13 m) contaminated with perchlorate at concentrations of 1–13 mg/kg using a high-pressure injection of corn syrup and ethanol. Evans

http://dx.doi.org/10.1016/j.jhazmat.2016.11.003 0304-3894/© 2016 Elsevier B.V. All rights reserved.

Please cite this article in press as: L. Avishai, et al., Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater, J. Hazard. Mater. (2016), http://dx.doi.org/10.1016/j.jhazmat.2016.11.003

G Model HAZMAT-18157; No. of Pages 8 2

ARTICLE IN PRESS L. Avishai et al. / Journal of Hazardous Materials xxx (2016) xxx–xxx

et al. [8] removed perchlorate at a concentration of ∼75 mg/L from a contaminated vadose zone (3–12 m) using gaseous electron donor-injection technology. However, both cases demonstrated perchlorate-degradation capacity at relatively low perchlorate and salinity concentrations. Leakage from waste water lagoons of an ammoniumperchlorate manufacturing plant by the area of Ramat Hasharon, Israel, has been resulted in heavy contamination of the vadose zone and groundwater [11,12]. In groundwater underlying the site the contaminated plum was found to spread over an area of several square kilometers with peak concentrations exceeding 1000 [mg/L]. The thick vadose zone underlying the former waste ponds (∼40 m) was also found to contain extreme concentrations of perchlorate, exceeding 2200 mg/kg at depth of 20 m (equivalent to ∼23,000 mg/L in the sediment pore-water) [1]. Attempts to enhance biodegradation of perchlorate in the vadose zone of that site included application of ethanol enriched water solution through a subsurface drip irrigation under a covered infiltration gallery [13]. A vadose zone-monitoring system (VMS) that was installed at the site enabled continuous monitoring of water infiltration, ethanol (electron donor) propagation, perchlorate degradation and solute transport across the entire vadose zone during the infiltration experiments (Supporting information Fig. S1 and S2) [14–16]. The results showed that while enhanced degradation of perchlorate was demonstrated at the upper soil layers, the efficiency of delivering high concentration of electron donor as ethanol, to deep sections of the vadose zone was limited, because it was consumed rapidly in the top soil. In addition, temporal variation of perchlorate concentrations across the vadose zone indicated significant displacement of perchlorate from the vadose zone toward the groundwater with the percolating water [13]. Later investigation of the microbial potential for perchlorate degradation in sediments from the vadose zone of the site suggested high reduction activity in the shallow soils and only limited and bio reduction potential at the deep highly contaminated layers (Supporting information Fig. S3, [12]). A possibility to overcome the above limitations on in-situ bio degradation of perchlorate in deep vadose zone is through hydraulic flushing and leaching [17]. Nevertheless, this method requires recovery of the contaminated extraction fluids from the underlying aquifer for treatment and recycle, when possible. The recovered solution is usually treated above ground in engineered industrial facilities and then discarded into the local sewer system. For example, a pilot soil flushing from perchlorate was proposed for unsaturated soil at Tronox’s Henderson facility, located in Southern Nevada [18,19]. Here we propose an alternative approach for remediation of the deep vadose zone and decontaminate polluted groundwater using the bio reactive part of the soil. In this approach contaminated groundwater is pumped through a series of shallow wells tapping the upper groundwater (Fig. 1). The contaminated groundwater is than enriched with electron donor and reintroduced top soil for treatment. The treated water percolated gravitationally through the vadose zone and leach pollutants to the groundwater where it is retrieved back for treatment in the top soil through a cyclic process. The main advantages this remediation approach are: (a) it exploits the high degradation potential and accessibility of the top soils and saves substantially on expensive engineered infrastructure, (b) it treats large volumes of ambient contaminated groundwater without introducing additional chemicals which may affect the aquifer’s hydraulic properties [20,21], (c) it does not require large capital investment for the construction and operation of external treatment facilities, making this remediation process cost-efficient, and (d) it does not involve further contamination with polluted effluents. Upon combination of the remediation approach with application of vadose zone monitoring technologies the method

may be optimized by a real-time indication of the actual transport and degradation activity of the unsaturated zone microbial communities. While the leaching and pumping phases of the proposed treatment approach are well established in the literature [17], including results from previous stages for our study site [13], in the current study we focus on the first phase of the suggested remediation treatment: degradation of perchlorate in the upper unsaturated soil layers of the contaminated site. Degradation capacity was investigated through long-term column experiments, where synthetic contaminated groundwater was applied continuously to unsaturated columns of local soil for bio treatment.

2. Materials and methods The potential capacity of shallow soils to serve as a bio reactor for continuous treatment of perchlorate was tested through long-term column-infiltration experiments. Two main practical aspects guided the design of the column experiment: (a) maintaining unsaturated flow conditions in the entire soil column, as expected in the unsaturated zone [22,23], and (b) continuous treatment of percolating water that imitates the local groundwater, heavily contaminated with perchlorate, with the addition of excess electron donor.

2.1. Column design and treatment application We believe that the most suitable way to apply the suggested treatment in the field is by subsurface drip irrigation. Such irrigation systems allow accurate control of the water-application cycles as well as the chemical composition of the implemented water. Accordingly, the columns were designed to represent the coverage area of a typical commercial dripper with a discharge rate of 2.5 L/h. The dripper was placed in the center of the column with radius of influence of 21 cm to represent a typical dripper’s distribution of 30 × 30 cm in the field. The use of subsurface irrigation under impervious cover was expected to encourage the generation of reducing conditions, which are an essential prerequisite for successful biodegradation of perchlorate. Three large polyethylene columns (42 cm diameter, 55 cm height) were packed with soil from an area that was used in the past as a waste pond [6,12] (Fig. 2). The soil was collected from the upper 50 cm of the profile at three different locations, to reflect the local heterogeneity. Prior to packing, the soil was mixed and sieved with a coarse mesh (1 mm) to eliminate roots and stones. The soil texture (95% sand, 2.5% silt and 2.5% clay) was determined by hydrometer method [24]. During packing, the soil was physically compacted to prevent the formation of large unnatural voids. An infiltration gallery was constructed at the top of the column with a 2-cm layer of coarse tuff (1–10 mm). Initially the infiltration gallery was covered with a polyethylene sheet which was found inefficient at creating reducing conditions. As a result, later in the experiment, the cover was replaced with 2-cm thick layer of bentonite, which was found suitable for isolating the column from atmospheric oxygen (Supporting information Fig. S4). To maintain the unsaturated conditions that prevail in natural unsaturated zone, the drainage system was designed to create low tension at the outlet of the column; a 2-cm thick rockwool layer was placed at the bottom of each column [25]. The outlet was extended with a 60-cm long pipe filled with compressed rockwool, which created hydraulic continuity and resulted in light tension that resembled the conditions in an unsaturated profile [25]. The bottom of the drainage extension was connected to a syphon made of a 6-mm diameter tube to prevent air penetration into the column.

Please cite this article in press as: L. Avishai, et al., Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater, J. Hazard. Mater. (2016), http://dx.doi.org/10.1016/j.jhazmat.2016.11.003

G Model HAZMAT-18157; No. of Pages 8

ARTICLE IN PRESS L. Avishai et al. / Journal of Hazardous Materials xxx (2016) xxx–xxx

3

Fig. 1. Schematic illustration of the combined vadose zone–groundwater-remediation approach.

All columns were equipped with a suction cup for frequent collection of pore water, an oxidation-reduction potential (ORP) electrode (Van London-pHoenix Company, USA) and a time-domain transmissometry (TDT; Acclima, USA) for continuous water-content and temperature measurements. Probes were installed at a depth of 25 cm (middle of the soil profile). The protecting shield of the OPR electrode was removed to ensure good contact of the sensitive membrane with the soil. Special attention

was given while soil packing to ensure proper contact of the probes with the soil (Fig. S4). A synthetic groundwater solution was prepared with an ionic composition similar to that of the groundwater at the site (see Table 1 for details). Shortly before infiltration, each water pulse was amended with ethanol to a final concentration of 2740 mg/L. The high ethanol concentration was designed to ensure amounts in excess of that required for perchlorate degradation, so that electron donor and carbon sources would not be limiting factors for

Fig. 2. (a) Perchlorate (ClO4 − ), chlorate (ClO3 − ) and chloride (Cl− ) concentrations in the column outlet. (b) Redox potential (ORP) in the middle of the soil column and pH in the column outlet. (c) Concentrations of Fe and Mn in the column outlet. (d) Sediment volumetric water content (␪v ) in the middle of the soil column during the infiltration experiments. GW, groundwater.

Please cite this article in press as: L. Avishai, et al., Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater, J. Hazard. Mater. (2016), http://dx.doi.org/10.1016/j.jhazmat.2016.11.003

G Model

ARTICLE IN PRESS

HAZMAT-18157; No. of Pages 8

L. Avishai et al. / Journal of Hazardous Materials xxx (2016) xxx–xxx

4

Table 1 Chemical composition of the synthetic groundwater (GW) that was used for the experiment and local GW from the site. Chemical parameter

Synthetic GW

GW

EC (ms/cm) ClO4 − (mg/L) Cl− (mg/L) SO4 −2 (mg/L) NO3 − (mg/L) Na+ (mg/L) K + (mg/L) Ca+2 (mg/L) Mg+2 (mg/L) NH4+ (mg/L) C2 H6 O (mg/L)

5.8 1150 1250 49 0 837 76 152 35 76 2740

5.4 1050 1230 37 224.5 763 117.6 439 142 7

the process, considering additional consumption by soil bacteria. To further promote reducing conditions, in two out of the three columns (columns 1 and 2), oxygen was purged from the solution in the first stages by bubbling nitrogen gas before infiltration. Solution was added to the columns in 4-L pulses through drippers in the infiltration gallery at a discharge rate of 2.5 L/h. Treatments were applied at intervals of one or two days cycles of infiltration and drainage steps. This procedure had two main goals: (a) to ensure sufficient retention time of the contaminated water in the soil column for biodegradation to occur, and (b) to allow efficient drainage and maintaining of unsaturated conditions. Note that prolonged application of nutrient-enriched water to soils often leads to ponding and bioclogging of the surface; the application of wetting and drainage cycles, as practiced in agriculture, helps prevent this and encourages preservation of the soil column’s high permeability. Infiltration cycles and oxygen-stripping conditions were changed throughout the experiment on the basis of system performance and measurements of ORP, water content, and perchlorate concentrations in the column outlet. For example upon stabilization of perchlorate reduction at minimum levels during infiltration cycles of two days the infiltration capacity was increased to daily infiltration cycles in order to check system performance on a double load of both water and perchlorate mass. In the later stages the system performance was also tested without oxygen stripping.

(GC)–flame ionization detector (FID) (CP-3800, Varian). Water samples (1.5 ␮L) were injected by auto sampler. The FID and injector temperatures were set to 270 and 250 ◦ C, respectively. The GC oven temperature was first held at 50 ◦ C for 1 min, increased to 220 ◦ C at a rate of 25 ◦ C/min, and then held for 4 min. The separation was performed by Stabilwax® capillary column (60 m, 0.32 mm, 0.25 ␮m, Restek Corporation, USA); helium was used as the carrier gas (1 mL/min). For quantification, five external standards were used. At the end of the experiment, after 72, 58 and 45 days for column one, two and three respectively; soil samples were taken for DNA extraction from the surface (0 cm), middle (25 cm depth), and bottom (50 cm depth) of the profile. The samples were taken from two separate locations in each column: one at the center under the dripper and the other 9 cm to the side of the center. In addition, control samples were taken from each column before the beginning of the experiment and were stored in sealed jars in the dark at room temperature (similar to the column temperature conditions). DNA extractions from all soil samples were performed from 0.5 g using PowerSoilTM DNA Isolation Kit according to the manufacture instructions (MO BIO Lab. Inc., USA). Quantitative PCR conditions and primers [26] are presented in the Supplementary data (Supporting information Table S1). 3. Results and discussion 3.1. Experimental phases During the first period, when all columns were covered with polyethylene liner, no perchlorate reduction was observed, at this stage redox conditions (ORP) in the columns was still positive (∼ + 200 [mv]). Obviously, such conditions are not favorable for perchlorate reduction. It was assumed that the polyethylene cover failed to maintain reducing conditions and it was replaced with a 2-cm thick layer of bentonite clay. As soon as the column cover was changed to bentonite, anaerobic conditions were created in the columns. To avoid overloading the figures with data from the early stage when no perchlorate degradation was achieved, the day on which the surface cover was changed to bentonite is set as day 0 in Fig. 2.

2.2. Sampling and analysis 3.2. Perchlorate degradation Drainage from each column was collected continuously and weighed for mass-balance control on a daily basis. Following every infiltration pulse, a 50-mL sample was collected from the cumulative drainage and kept for chemical analysis. In addition, water samples were collected once a week from the suction cups installed in the middle of the soil column. All samples were filtered with a 0.45-␮m filter (Milex) and stored at 4 ◦ C until analysis. Perchlorate at the outlet was analyzed on a daily basis against a calibration curve using a perchlorate electrode (Laboratory Perchlorate Ion Electrode, Cole-Parmer, USA). All samples measured with ISE were adjusted by dilution to a concentration range of 10–100 mg/L. For conformation of the electrode results, duplicates were frequently analyzed by injecting 25 ␮L sample on an Thermo ScientificTM DionexTM ion chromatography systems (ICS 5000) equipped with Ion Pac AS19 column (detection limit of ±0.01 mg/L) and AG19 guard column Anion Self-Regenerating Suppressor (ASRS 300) at a flow rate of 1 mL/min. 60 mM KOH eluent was used for perchlorate analysis and 20 mM KOH for nitrate, chlorate, chloride and sulfate separation. The samples were separated over 20 min at a flow rate of 1 mL/min. Metal concentration was determined using inductively coupled plasma optical emission spectrometry (ICP–OES, Agilent ICP Varian, USA), with a detection limit for this analysis of 0.002 mg/L. Ethanol was analyzed by gas chromatograph

A significant decrease in perchlorate concentration from inlet to outlet was attained within 10–15 days after placing the bentonite cover on top of each column (Fig. 2). Perchlorate was rapidly declined from ∼1150 mg/L at the column inlet to 200–400, 400–600 and 0–200 mg/L in columns 1, 2 and 3, respectively, following each daily application of perchlorate-contaminated groundwater. This strong reduction of perchlorate was found in all treatments throughout the entire period. The decrease of perchlorate concentration in all columns was accompanied by an increase in chloride concentration, showing that perchlorate is indeed reduced to chloride. However, during the first period, an increase in chlorate concentration in the outlet (from 0 to 250 mg/L in column 1 and 0–50 mg/L in column 2) was observed, which then decreased to zero as development of the reducing conditions proceeded and chloride reached maximum values. In column 3, no chlorate was found in the outlet samples. Chlorate accumulation during high perchlorate reduction has been reported [27]. Competitive inhibition of perchlorate reduction by chlorate was suggested in the past, because perchlorate reductase reduces both anions [28]. Thus, the presence of chlorate might be an indication of transitional reducing conditions. Nevertheless, chlorate was observed only in part of the columns and only in the early stage before stabiliza-

Please cite this article in press as: L. Avishai, et al., Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater, J. Hazard. Mater. (2016), http://dx.doi.org/10.1016/j.jhazmat.2016.11.003

G Model HAZMAT-18157; No. of Pages 8

ARTICLE IN PRESS L. Avishai et al. / Journal of Hazardous Materials xxx (2016) xxx–xxx

5

Table 2 Experimental conditions and perchlorate reduction. Column

Time to stabilization at average minimum concentration (days) Average daily perchlorate reduction Total mass (g/day) From initial mass (%) [mass reduced/mass in] During 1-day cycle with oxygen-stripped water (g/day) During 2-day cycle with oxygen-stripped water (g/day) During 1-day cycle without oxygen stripping (g/day) During 2-day cycle without oxygen stripping (g/day)

1

2

3

12

17

12

2.9 76 3.623

1.8 60

2.7 84

1.875

1.414

3.121

2.5

3.128 2.19

ertheless, this unexpected positive effect was also documented by Shrout and Parkin [30], who examined the effect of oxygen on diverse perchlorate-degrading bacteria and found that when the electron donor is not a limiting factor, oxygen actually encourages perchlorate biodegradation. It was suggested that this effect of oxygen might be related to an increase in bacterial abundance under these conditions [31].

Fig. 3. Relative concentrations (Con.) of perchlorate (ClO4 − ), chlorate (ClO3 − ) and chloride (Cl− ) between the inlet and outlet of the column during the experiment. ON FIG: Y AXIS UNITS: [mmol/L].

tion of degradation on maximum capacity. Accordingly it may be concluded that in our case chlorate inhibition to perchlorate complete reduction was insignificant. During the experiment, perchlorate degradation and chloride and chlorate production were fitted stoichiometrically (Fig. 3). Measurements taken occasionally at the outlet showed ethanol consumption in the range of 60%–95% of the amount applied to the columns. On a weight basis, 0.31 g ethanol is needed to reduce one gram of perchlorate (2C2 H6 O + 3ClO4 − → 6H2 O + 4CO2 + 3Cl− ). However, more ethanol was consumed than the amount needed to reduce inlet perchlorate. This suggests that other bacteria competed with perchlorate reducing bacteria on ethanol energy. Increasing the treatment load from 4 L (equivalent loading of 2400 mg/day) every 2 days to 4 L every day (equivalent loading of 4800 mg/day) did not reduce the perchlorate-degradation capacity. Perchlorate concentration in pore water from the middle of the columns taken by the suction cups from a depth of 25 cm was generally lower than the inlet concentration and higher than the outlet concentration. This indicates that perchlorate degradation was active along the entire soil profile and not limited to shallow or deep sections. Dissolved oxygen was stripped from the solution used in columns 1 and 2 prior to infiltration into the soil column until days 55 and 41, respectively. In contrast, column 3 experiment was conducted with a solution that was not stripped of dissolved oxygen. Surprisingly, perchlorate-degradation capacity in column 3 was higher than that in the other two columns, reaching complete degradation during the first period and later stabilizing at 84% reduction (∼200 mg/L) (Fig. 2, Table 2). Moreover, even when columns 1 and 2 received solution that was not stripped of dissolved oxygen, i.e. after day 55 and 41, respectively, there was no decrease in perchlorate-reduction rate. Being a preferential electron acceptor, oxygen is known to prevent perchlorate degradation. Furthermore, it inhibits the expression of chlorite dismutase [29], and perchlorate reductase is an oxygen-sensitive enzyme [29]. Nev-

3.3. Development of reducing conditions The redox potential is a good indicator of favorable conditions for perchlorate metabolism [32]. Reports have suggested that an ORP of below −200 mV is needed; nevertheless, perchlorate is still reduced at more positive ORPs, albeit at a slower rate. In our study, before day zero (when the polyethylene cover was being used), ORP values fluctuated between +200 and 0 mV (Fig. 2). Decreases in redox potential were attributed to infiltration cycles of oxygen-stripped solution, increases to penetration of air during the drainage phase. Nevertheless, there was no degradation of perchlorate during this entire stage. However, within 10 days after the columns had been covered with a thin layer of bentonite (day 0), the redox potential declined to negative values ranging from −100 to −300 mV in all columns (Fig. 2). The decrease in redox potential was followed by immediate and significant degradation of perchlorate. Attaway and Smith [29] found that perchlorate degradation occurs only at redox potentials lower than −110 mV. In contrast, Shrout and Parkin [30] found that perchlorate can be reduced under redox potentials ranging between +180 and −220 mV, but that complete degradation could only be observed at redox potentials below −220 mV. However, there are microniches in soils that can have different redox potentials in very close proximity [30], creating different bacterial oxidation and reduction processes simultaneously within the soil profile [33]. The pH measurements of the outlet solution exhibited a significant drop from 8 to 7.5 to 5.5–6.5 along with a reduction in redox potential and degradation of perchlorate (Fig. 2). In flooded soil, pH drops due to the accumulation of carbonic acid formed from trapped CO2 produced during respiration [34]. The fact that perchlorate was always present in the effluent suggests that fermentation is not involved in the pH decline. In this study we concentrated on the main parameter that influence perchlorate reduction in unsaturated soil, oxygen. However, although nitrate is present in the groundwater (Table 1), its inhibitory effect is less significant in our case. In contrast to most situations, groundwater at Ramat Hasharon area contains on molar basis, three times more perchlorate than nitrate. Moreover, earlier studies testing the effect of equimolar concentration of nitrate on perchlorate reduction in this surface soil used in our study had shown that perchlorate reduction rates are not influenced by

Please cite this article in press as: L. Avishai, et al., Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater, J. Hazard. Mater. (2016), http://dx.doi.org/10.1016/j.jhazmat.2016.11.003

G Model HAZMAT-18157; No. of Pages 8

ARTICLE IN PRESS L. Avishai et al. / Journal of Hazardous Materials xxx (2016) xxx–xxx

6

nitrate and both reduce concurrently [35]. Nozawa-Inoue et al. [36] demonstrated that nitrate accelerate perchlorate reduction in unsaturated soil that was amended with acetate. Furthermore, concurrent perchlorate and nitrate reduction was described in an enrichment culture derived from the soil and use to treat contaminated water in Ion Exchange Membrane Bioreactor [37]. The appearance of both manganese and iron in the effluents could be a result of either washing of metal colloids or biochemical metal reduction. Because the increase in metals in the effluents corroborated with the decline in ORP, it is suggested that the metals in the effluents are associated with the microbial reduction of soil metal oxides. As perchlorate reduction provides more free energy than metal reduction, this observation is surprising [38]. Competition for the electron donor between metal reducers and perchlorate reducers might explain the decline in perchlorate reduction rate with the increase in soluble metals in the effluent. However, reduced iron is not inhibitory for perchlorate reducers [39]. Depending on the composition of the soil, enhanced dissolution might affect the porosity and reactivity of the matrix. 3.4. Abundance of perchlorate-reducing bacteria Soil samples were taken from the column profile immediately after the experiment for analysis of microbial gene copies. The analysis showed a two order of magnitude increase in gene copies of 16 s rRNA as well pcrA per gram of soil (Fig. 4). Both genes increased in the center cores and side cores, indicating that perchlorate degradation occurred throughout the entire column. The relative abundance of perchlorate reducers, as expressed by the ratio of pcrA to 16 s rRNA, was 0.008, 0.04 and 0.048 in columns 1, 2 and 3, respectively. A low abundance of perchlorate reducers was also observed by Nozawa-Inoue et al. [36], where the pcrA-to–16 s rRNA ratio in soil amended with perchlorate and acetate was about 0.1 ± 0.2%. The higher ratios in our experiment are likely due the larger amount of donors and acceptors consumed over time. The difference in the numbers of 16 s rRNA and pcrA gene copies among the columns indicated that in soil samples taken from the field in relatively close proximity (a few meters), there is variability in total bacterial abundance and in perchlorate-reducing bacteria. 3.5. Percolation conditions One of the critical properties characterizing the vadose zone is unsaturated conditions. Under natural conditions, it is practically impossible to create saturated conditions in the vadose zone by gravitational infiltration, and saturation can only be achieved by uprise of the water table or ponding [22,23]. Therefore, maintaining unsaturated conditions in the column during the entire infiltration and degradation experiment was essential for extrapolation of the results to full field setups. Indeed, continuous measurements of the volumetric water content of the sediment in the column during the experiment showed values ranging between 12 and 18 m3 /m3 in all columns, with the exception of days 55–68 in column 1 (Fig. 2). During that period, water content in column 1 rose from 15 to 25 m3 /m3 due to technical failure of the drainage system. Once the problem in the drainage was resolved, the water content decreased back to its original values. Note that water content over the course of the experiment in all treatments was far below saturation (∼40 m3 /m3 ), and all columns maintained the unsaturated conditions characteristic of natural unsaturated sediments in the field. The daily wetting and drainage cycles in the columns were expressed as a sharp increase in water content upon arrival of the wetting front to the middle of the column, where the moisture sensor was placed. Each wetting cycle was followed by a drainage phase which was expressed as a sequential decrease in

water content. Obviously, development of biomass in the column is essential for degradation. However, biomass development to a level that could significantly reduce hydraulic conductivity would have resulted in increased water content. Furthermore, significant biofilm formation would significantly increase the water content of the soil [40]. In unsaturated soils, perchlorate degradation is affected by the degree of water content [41]. Under low water content, oxygen availability is usually higher, which may result in lower perchlorate reduction. Nevertheless, in the current study, perchlorate degradation was achieved even at the relatively low water content of ∼14 m3 /m3 , much lower than reported in other studies [42,43]. Water content measurements in the soil columns throughout the entire experiment show that the sediment water content fluctuate in a relatively constant range between wetting and drainage cycles. There were no signs of water content build up or change in the drainage pattern. These are indications that there was no significant reduction in soil hydraulic conductivity or a bio clogging process that impacted the hydrological process in the column. It is likely that the use of pulse infiltration through drippers in large columns that resemble natural soil helped prevent bioclogging. In natural unsaturated soils, degradation “hot spots” are likely to occur. The theory of soil hot spots suggests that under unsaturated conditions, some pores are more exposed to a higher flux of water and nutrients, as well as electron donors and acceptors, where microbial activity and degradation may take place. Eventually, these hot spots may clog due to biomass generation. Nevertheless, once a hot spot is clogged, the water flow route will migrate, creating a new hot spot. The clogged pore will then eventually become unclogged due to lack of nutrients [44]. This is even more prominent when infiltration and drainage cycles are performed. Accordingly, managing a cleanup operation of the unsaturated zone may be beneficial if operated under frequent wetting cycles that prevent large-scale bioclogging.

4. Concluding remarks The difficulties in promoting in-situ remediation of the deep vadose zone have been attributed to the following: (a) limitations in delivering electron donor through injection of enriched water solution from the land surface to deep sections of the vadose zone (>15 m) [13]; (2) low perchlorate-degradation potential in sediment containing a high concentration (>10,000 mg/L) of perchlorate in the deep vadose zone at the site [12,13], and (c) enhanced infiltration of water from land surface, resulting in substantial down leaching of perchlorate to the groundwater [13]. The innovative remediation approach proposed herein involves three stages in a cyclical process (Fig. 1): (a) soil treatment – infiltration of contaminated groundwater enriched with a carbon source and electron donor for treatment in the bioactive topsoil; (b) downleaching – treated water from the topsoil infiltrates down from the bioactive zone to the deep sections of the vadose zone, downleaching perchlorate contamination toward the water table, and (c) retrieval – pumping back the contaminated leachates from the groundwater near the water table for treatment in the topsoil. Results from column experiments showed that under unsaturation condition it possible to maintain high rate of perchlorate degradation. These unsaturation conditions also prevented bio clogging during the three months of experiment as observed by the water content measurements that indicated the relatively small range and constant fluctuations in water content. Furthermore, we concluded that it possible, under unsaturated operation regimes, to sustain active perchlorate degrading microbial population as evidence from the qPCR results. Our observation that degradation was not limited to a particular section of the col-

Please cite this article in press as: L. Avishai, et al., Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater, J. Hazard. Mater. (2016), http://dx.doi.org/10.1016/j.jhazmat.2016.11.003

G Model HAZMAT-18157; No. of Pages 8

ARTICLE IN PRESS L. Avishai et al. / Journal of Hazardous Materials xxx (2016) xxx–xxx

Fig. 4. Gene copies of 16 s rRNA and pcrA per gram soil at different depths of the center and side profiles, compared to control, untreated soil.

Please cite this article in press as: L. Avishai, et al., Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater, J. Hazard. Mater. (2016), http://dx.doi.org/10.1016/j.jhazmat.2016.11.003

7

G Model HAZMAT-18157; No. of Pages 8

ARTICLE IN PRESS L. Avishai et al. / Journal of Hazardous Materials xxx (2016) xxx–xxx

8

umn and the consistent abundance of perchlorate reducing bacteria along the column imply that the deeper soil profile (below 50 cm) could potentially bring about a larger degradation of perchlorate. Based on the degradation rate of 51 g/m2 per day at 50 cm soil profile we conclude that by very simple means, over 51 tons of perchlorate per year can be degraded in the study site. The results of the research indicate that the alternative remediation approach, where the topsoil serves as a bioreactor for treating contaminated groundwater through a cleanup operation by flushing of the deep vadose zone, is a promising remediation solution with high feasibility for success in the field. Acknowledgement The authors wish to express their appreciation to the Israeli Military Industry for their support and to the Israeli Water Authority for funding the project. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jhazmat.2016.11. 003. References [1] H. Gal, N. Weisbrod, O. Dahan, Z. Ronen, R. Nativ, Perchlorate accumulation and migration in the deep vadose zone in a semiarid region, J. Hydrol. 378 (2009) 142–149. [2] C.W. Trumpolt, M. Crain, G.D. Cullison, S.J.P. Flanagan, L. Siegel, S. Lathrop, Perchlorate: sources, uses, and occurrences in the environment, Remediat. J. 16 (2005) 65–89. [3] D.S. Roote, Technology Status Report: Perchlorate Treatment Technologies, First Edition, Ground-Water Remediation Technologies Analysis Center, Pittsburgh, PA, 2001. [4] W. Motzer, Perchlorate: problems, detection, and solutions, Environ. Forensics 2 (2001) 301–311. [5] E.T. Urbansky, S.K. Brown, Perchlorate retention and mobility in soils, J. Environ. Monit. 5 (2003) 455–462. [6] H. Gal, Z. Ronen, N. Weisbrod, O. Dahan, R. Nativ, Perchlorate biodegradation in contaminated soils and the deep unsaturated zone, Soil Biol. Biochem. 40 (2008) 1751–1757. [7] J.D. Coates, L.A. Achenbach, Microbial perchlorate reduction: rocket-fueled metabolism, Nat. Rev. Microbiol. 2 (2004) 569–580. [8] P.J. Evans, R.A. Fricke, K. Hopfensperger, T. Titus, In Situ destruction of perchlorate and nitrate using gaseous electron donor injection technology, Groundwater Monit Remediat. 10 (2011) 3–11 (2). [9] J. Xu, Y. Song, B. Min, L. Steinberg, B.E. Logan, Microbial degradation of perchlorate: principles and applications, Environ. Eng. Sci. 20 (2003) 405–422. [10] A.J. Frankel, L.M. Owsianiak, B.J. Wuerl, J.F. Horst, In-situ Anaerobic Remediation of Perchlorate-impacted Soils, Arcadis, USA, 2005. [11] R. Nativ, E. Adar, Soil and Groundwater Contamination in the Ramat Hasharon Area, Annual Scientific Report Submitted to the Water Authority, Ministry of Infra Structure, Tel Aviv, 2005. [12] I. Sikron, Perchlorate Degradation in the Vadose Zone at the Israel Military Industry Site in Ramat Hasharon, Ben-Gurion University of the Negev, 2013 (M.Sc Thesis, Available online at: http://in.bgu.ac.il/en/aranne/Pages/default.aspx). [13] E. Katz, Transport and Degradation of Perchlorate in a Vadose Zone Undergoing Enhanced in Situ Bioremediation Treatment, Ben-Gurion University of the Negev, 2011 (M.Sc Thesis, Available online at: http://in.bgu.ac.il/en/aranne/Pages/default.aspx). [14] O. Dahan, R. Talby, Y. Yechieli, E. Adar, N. Lazarovitch, Y. Enzel, In-situ monitoring of water percolation in layered soils using a vadose-zone monitoring system, Vadose Zone J. 8 (2009) 916–925. [15] O. Dahan, A. Babad, N. Lazarovitch, E.E. Russak, D. Kurtzman, Nitrate leaching from intensive organic farms to groundwater, Hydrol. Earth Syst. Sci. 18 (2014) 333–341. [16] Y. Rimon, O. Dahan, R. Nativ, S. Geyer, Water percolation through the deep vadose zone and groundwater recharge: preliminary results based on a new vadose zone monitoring system, Water Resour. Res. 43 (2007). [17] O. Atteia, E.D. Estrada, H. Bertin, Soil flushing: a review of the origin of efficiency variability, Rev. Environ. Sci. Bio Technol. 12 (2013) 379–389.

[18] T.F. Battey, A.J. Shepard, R.J. Tait, Soil flushing through a thick vadose zone: perchlorate removal documented at Edwards AFB, California, in: Am. Geophys. Union, Fall Meeting 2007, Abstr. Publ., AGU, 2007. [19] Nevada Division of Environmental Protection, Treatability Study Work Plan: in Situ Soil Flushing Pilot Nevada Environmental Response Trust Site, Technology Innovation and Field Services Division, Henderson, Nevada,US EPA, 2014. [20] P. Baveye, P. Vandevivere, B.L. Hoyle, P.C. DeLeo, D.S. de Lozada, Environmental impact and mechanisms of the biological clogging of saturated soils and aquifer materials, Crit. Rev. Environ. Sci. Technol. 28 (1998) 123–191. [21] K. Seki, M. Thullner, J. Hanada, T. Miyazaki, Moderate bioclogging leading to preferential flow paths in biobarriers, Ground Water Monit. Remediat. 26 (2006) 68–76. [22] O. Dahan, R. Talby, Y. Yechieli, E. Adar, N. Lazarovitch, Y. Enzel, In-situ monitoring of water percolation in layered soils using a vadose-zone monitoring system, Vadose Zone J. 8 (2009) 916–925. [23] Y. Amiaz, S. Sorek, Y. Enzel, O. Dahan, Solute transport in the vadose zone and groundwater during flash floods, Water Resour. Res. 47 (2011). [24] G.W. Gee, J.W. Bauder, Particle-size analysis. P. 383–411, in: A.L. Page (Ed.), Methods of Soil Analysis, Part1, Physical and Mineralogical Methods, Second Edition, Agronomy Monograph 9, American Society of Agronomy, Madison, WI, 1986. [25] A. Ben-gal, U. Shani, A highly conductive drainage extension to control the lower boundary condition of lysimeters, Plant Soil 239 (2002) 9–17. [26] M. Nozawa-Inoue, M. Jien, N.S. Hamilton, V. Stewart, K.M. Scow, K.R. Hristova, Quantitative detection of perchlorate-reducing bacteria by real-time PCR targeting the perchlorate reductase gene, Appl. Environ. Microbiol. 74 (2008) 1941–1944. [27] M. Dudley, A. Salamone, R. Nerenberg, Kinetics of a chlorate-accumulating, perchlorate-reducing bacterium, Water Res. 42 (10) (2008) 2403–2410. [28] S.K. Chaudhuri, S.M.O. Connor, R.L. Gustavson, L.A. Achenbach, J.D. Coates, Environmental factors that control microbial perchlorate reduction, Appl. Environ. Microbiol. 68 (2002) 4425–4430. [29] S.W. Kengen, G.B. Rikken, W.R. Hagen, C.G. van Ginkel, A.J. Stams, Purification and characterization of (per)chlorate reductase from the chlorate-respiring strain GR-1, J. Bacteriol. 181 (1999) 6706–6711. [30] J.D. Shrout, G.F. Parkin, Influence of electron donor, oxygen, and redox potential on bacterial perchlorate degradation, Water Res. 40 (2006) 1191–1199. [31] D.E. Nozawa-Inoue, Perchlorate reduction using fine media fluidized bed bioreactor with oxidation-reduction potential-based feed control, Proceedings of the Annual International Conference on Soils, Sediments, Water and Energy, Vol. 15, Article 3 (2010). [32] H. Attaway, M. Smith, Reduction of perchlorate by an anaerobic enrichment culture, J. Ind. Microbiol. 12 (1993) 408–412. [33] S. Baram, S. Arnon, Z. Ronen, D. Kurtzman, O. Dahan, Infiltration mechanism controls nitrification and denitrification processes under dairy waste lagoon, J. Environ. Qual. 41 (2012) 1623–1632. [34] E.A. Paul, Soil Microbiology Ecology and Biochemistry, Academic Press Oxford, UK, 2014. [35] H. Gal, Perchlorate Fate in the Unsaturated Zone. PhD Dissertation, The Hebrew University of Jerusalem, 2010 (available at, http://harmanlib.huji.ac. il/index.html) 87 pp. [36] M. Nozawa-Inoue, M. Jien, K. Yang, D.E. Rolston, K.R. Hristova, K.M. Scow, Effect of nitrate acetate, and hydrogen on native perchlorate-reducing microbial communities and their activity in vadose soil, FEMS Microbiol. Ecol. 76 (2011) 278–288. [37] S. Fox, T. Bruner, Y. Oren, J. Gilron, Z. Ronen, Concurrent microbial reduction of high concentrations of nitrate and perchlorate in an ion exchange membrane bioreactor, Biotechnol. Bioeng. 113 (2016) 1881–1891. [38] K.A. Weber, L.A. Achenbach, J.D. Coates, Microorganisms pumping iron: anaerobic microbial iron oxidation and reduction, Nat. Rev. Microbiol. 4 (2006) 752–764. [39] A. Son, J. Lee, P.C. Chiu, B.J. Kim, D.K. Cha, Microbial reduction of perchlorate with zero-valent iron, Water Res. 40 (2006) 2027–2032. [40] X. Zhang, P.J. Van Geel, Development of a vertical TDR probe to evaluate the vertical moisture profile in peat columns to assess biological clogging, J. Environ. Eng. Sci. 6 (2007) 629–642. [41] H. Cai, A.G. Eramo, P.J. Evans, R. Fricke, R.A. Brennan, In situ bioremediation of perchlorate in vadose zone soil using gaseous electron donors: microcosm treatability study, Water Environ. Res. 82 (2010) 409–417. [42] P.J. Evans, I. Lo, A.E. Moore, W.J. Weaver, W.F. Grove, H. Amini, Rapid full-scale bioremediation of perchlorate in soil at a large brownfields site, Remediat. J. 18 (2008) 9–25. [43] M. Nozawa-Inoue, K.M. Scow, D.E. Rolston, Reduction of perchlorate and nitrate by microbial communities in vadose soil, Appl. Environ. Microbiol. 71 (2005) 3928–3934. [44] M. Bundt, F. Widmer, M. Pesaro, J. Zeyer, P. Blaser, Preferential flow paths: biological ‘hot spots’ in soils, Soil Biol. Biochem. 33 (2001) 729–738.

Please cite this article in press as: L. Avishai, et al., Using the natural biodegradation potential of shallow soils for in-situ remediation of deep vadose zone and groundwater, J. Hazard. Mater. (2016), http://dx.doi.org/10.1016/j.jhazmat.2016.11.003