Lactate Injection by Electric Currents for Bioremediation of Tetrachloroethylene in Clay

Abstract

Biological transformation of tetrachloroethylene (PCE) in silty clay samples by ionic injection of lactate under electric fields is evaluated. To prepare contaminated samples, a silty clay slurry was mixed with PCE, inoculated with KB-1^® dechlorinators and was consolidated in a 40 cm long cell. A current density between 5.3 and 13.3 A m⁻² was applied across treated soil samples while circulating electrolytes containing 10 mg L⁻¹ lactate concentration between the anode and cathode compartments to maintain neutral pH and chemically reducing boundary conditions. The total adsorbed and aqueous PCE was degraded in the soil to trichloroethylene (TCE), cis-1,2-dichloroethene (cis-DCE), vinyl chloride (VC) and ethene in 120 d, which is about double the time expected for transformation. Lactate was delivered into the soil by a reactive transport rate of 3.7 cm² d⁻¹ V⁻¹. PCE degradation in the clay samples followed zero order transformation rates ranging from 1.5 to 5 mg L⁻¹ d⁻¹ without any significant formation of TCE. cis-DCE transformation followed first order transformation rates of 0.06 to 0.10 per day. A control experiment conducted with KB-1 and lactate, but without electricity did not show any significant lactate buildup or cis-DCE transformation because the soil was practically impermeable (hydraulic conductivity of 2×10⁻⁷ cm s⁻¹). It is concluded that ionic migration will deliver organic additives and induce biological activity and complete PCE transformation in clay, even though the transformation occurs under slower rates compared to ideal conditions.

1. Introduction

Tetrachloroethylene (PCE) and its transformation byproducts are among the most frequently encountered contaminants in soil and groundwater. In situ reductive dechlorination by pumping of electron donors has been applied at sites with relatively permeable aquifers [1 – 3]. However, uniform injection of additives is a key for successful implementation and challenges exist when remediating fractured clay, heterogeneous aquifers and/or leaky aquitards. In low permeability soils, contaminants usually penetrate into preferential flow pathways and diffuse into the matrix [4]. Removing contamination sources from sites is not sufficient as contaminants may back-diffuse into the aquifer for decades, serving as long-term contamination sources [5 – 8].

Ineffective pumping of organic additives for remediation is also affected by degradation of the additives. In some cases, the additives (e.g. lactate [9]) are not detected in soil when delivered by hydraulic gradients because the additives’ degradation occurs at higher rates than their transport by hydraulic gradients. Pumping of additives and consequent transformation of PCE in heterogeneous soil and groundwater deposits is complex and is a function of the transformation kinetics relative to transport rates. Remediation of such deposits is difficult because of the low permeability, complexity of the processes, and limitation of mass transfer due to the slow diffusion process [10].

Electrokinetic injection uses direct electric currents (DC) to generate uniform electric fields and uniform transport by electroosmosis and ion migration [11], resulting in mixing of remediation additives in heterogeneous and low permeability deposits. The process of ion transport under DC in clays can occur at uniform rates of up to a few centimeters per day [12]. The concept of DC injection for enhanced bioremediation was introduced in the mid-1990’s. Several studies have reported efficient transport and injection rates for additives, such as sulfate transport of up to 20 cm d⁻¹ across 80 cm beds of fine sand [13], acetate and nitrate migration at 2.0 to 2.6 cm d⁻¹ across 40 cm silty loam and clay samples (hydraulic conductivity less than 3 cm per year) [14], and citrate migration through contaminated marine sediments over a distance of 4.6 meters at a rate of 30 cm d⁻¹ [15]. Lactate reactive transport rates, which account for sorption and transformation, on the order of 5 cm d⁻¹ in sand and 3.7 cm d⁻¹ in clay were reported [9] under electric currents. Nutrient amendment by electrokinetics has been reported [16] for the bioremediation of a chromium-contaminated soil.

Studies on biological transformation in consolidated clay deposits or leaky aquitards are rare. It is necessary to assess the feasibility and efficiency of delivery of organic additives in such media, which is further complicated by adsorption and biological transformation of the additives. The objective of this study is to investigate in situ bioremediation of PCE in clay by electrokinetic injection of lactate, a common electron donor for anaerobic biodegradation [3, 17]. The study includes assessment of sorption and transformation of PCE and lactate and evaluates effects of electrokinetic injection of lactate on biological transformation of PCE in clay.

2. Experimental

This section describes several difficult experimental tasks that are necessary because of the challenges of working with a volatile contaminant, anaerobic bacteria and clay. The soil description and properties are first presented, followed by a description of the procedure to evaluate PCE adsorption on the soil, a description of the biological culture and its maintenance, a description of the electrokinetic-enhanced bioremediation (EK-Bio) experimental setup, clay sample preparation, mixing with contaminant and biological culture, and finally a description of the EK-Bio experiments.

Silty clay soil, classified as low plasticity clay (CL) by the Unified Soil Classification System, was used. The soil had an appreciable amount (1.4% by dry weight) of organic matter and a pH of 6.8 which was measured using 20% soil slurry. The soil solids had a specific gravity of 2.69, a cation exchange capacity of 15 meq 100g⁻¹ and a zeta potential of −75 mV at neutral pH [18]. Other soil properties include a plasticity index of 12, and a maximum dry density of 1.67 g cm⁻³ at an optimum moisture content of 18.3%. Sodium lactate [(Sigma-Aldrich), 60% by weight liquid], and PCE [(Sigma-Aldrich), 99% pure with a density of 1.62 g cm⁻³] were used.

Contaminant retention and adsorption will impact transport and transformation of target contaminants. Therefore, duplicate PCE and trichloroethylene (TCE) soil adsorption tests were conducted in 60 mL glass serum bottles with Teflon-lined rubber stoppers. 50 mL PCE solutions were mixed with each 10g of soil for 2 days. The average concentration of triplicate blank samples was assumed to be the initial values when calculating partitioning factors. All adsorption samples had 200 mg L⁻¹ of sodium azide to prevent any biodegradation.

A PCE degrading source culture, known as KB-1® dechlorinator, was obtained from SiREM (Guelph, ON, Canada). The culture is described as a natural, non-pathogenic microbial culture that contains various species of Dehalococcoides to promote complete dechlorination of chlorinated ethene to non-toxic ethane (http://www.siremlab.com/). While the source KB-1® 90 culture was obtained from SiREM in serum bottles, it was developed and maintained at Northeastern University (NU) by feeding it PCE and lactate. At SiREM, the KB-1® culture was maintained with TCE with methanol as the electron donor. A similar procedure was followed at NU but was adjusted to use PCE with lactate as an electron donor source. Although the procedure is adequate to maintain cultures in a porous media, it was not clear if the bacteria would be active in a tight clay formation as described in this work.

EK-Bio experiments were conducted using laboratory cells which are made of rectangular boxes (Figure 1) that are 15 cm high, 5 cm wide and 40 cm long. The cells include three levels of sampling ports for pore water collection and voltage measurements. The ports are distributed evenly across the box, with vertical and horizontal spacing of 5 and 8 cm, respectively.

Figure 1.

Diagram of laboratory system for electrokinetic enhanced bioremediation

Preparing a silty clay sample with PCE contamination and an anaerobic degrading culture in the 40 cm EK-Bio cell is difficult because of the volatility of the contaminants and the need to maintain an anaerobic environment for the degrading culture. The soil, contaminant and culture were prepared as slurry and then consolidated to represent field conditions. To prepare the slurry, 4500 g of dry soil was mixed with 3000 mL of deionized water and 400 µL of pure PCE. The PCE was allowed to completely dissolve in the deionized water before mixing with the clay, thus PCE was added as a dissolved phase and not as a NAPL phase. The clay and PCE solution slurry was allowed to equilibrate for 2 days. After that, 100 mL of KB-1® culture was inoculated in the slurry. Culture volume was selected to achieve final dechlorinator cell concentration of about 10⁴ cells mL⁻¹ in pore water [2] with source cell concentration at 10⁶ cells mL⁻¹ (as reported by SiREM lab) and safety factor of 3. Immediately after inoculation, the slurry was consolidated in the EK-Bio rectangle setup. The consolidation was conducted with initial loading 4.2 psi (29 kPa) and increased every two days, up to final stress of 17.5 psi (120.7 kPa) after 11 days of consolidation. Mixing and consolidation were conducted in an anaerobic chamber.

Soil and pore water samples were collected after consolidation for measurements of initial concentrations and chemical conditions. It was necessary to do this step after consolidation, which lasted 14 days, to make sure that biological activity and reactivity (e.g. sorption) were at equilibrium. Samples were collected from the ports for measurement of the initial distribution of PCE. After consolidation, the soil hydraulic conductivity was 2 × 10⁻⁷ cm s⁻¹, measured by a falling head test.

After consolidation, the cell was connected to electrode reservoirs at both ends. Graphite electrodes were placed in both reservoirs, and the reservoirs were sealed airtight. Gas bags were connected at each electrode reservoir to collect the gases generated from electrolysis or bioreactivity. The anode and cathode electrolytes were circulated to neutralize the pH. An additional pH control system was connected to ensure that the pH range was maintained around 7.0 ± 0.2 with additives of dilute solutions of sodium hydroxide or sulfuric acid.

Four cells that include two experiments (EK-A and EK-B) and two non-electricity control experiments (Control-A and Control-B) were conducted. Experiments EK-A and EK-B were conducted under similar conditions but for different processing periods. Experiments Control-A and Control-B were conducted under similar conditions without electricity but for different periods of times. The samples had initial dry density of 1.41 to 1.58 g cm⁻³, porosity of 0.41 to 0.47 and initial PCE concentrations of 24 to 33 mg L⁻¹. It was not feasible to conduct more experiments because of the difficulties in preparing the samples and the duration of each test (one month for sample preparation and up to 4 months testing for each sample). A constant current density mode was applied in order to facilitate controlling electrolysis at the electrodes, which will allow the voltage to adjust depending on the temporal changes in electrical conductivity. For EK-A and EK-B, an initial current density of 5.3 A m⁻² was applied, which was later increased to 13.3 A m⁻², mainly because the voltage dropped as a result of increasing electrical conductivity of the system. The applied constant currents generated an electrical gradient between 0.5 and 1.0 V cm⁻¹ across the sample. An initial lactate concentration of 10 g L⁻¹ was maintained at the electrolytes. To make N and P not the limitation factors, 0.5 g of KH₂PO4, 1 g of K₂HPO4, and 5 g of (NH₄)₂SO4 were dissolved in the electrode reservoirs as well at the beginning. Control-A and Control-B were conducted at same initial conditions without applying electricity.

Lactate concentrations were analyzed using an ion chromatograph (Dionex 120). A Dionex AS15 column was used with 40 mM sodium hydroxide eluent and a flow rate of 1.2 mL min⁻¹. The standard curve was prepared with a concentration range from 0 to 200 mg L⁻¹, and updated every three months. PCE, TCE, and cis-1,2-dichloroethene (cis-DCE) concentrations in pore water were analyzed using the purge and trap method with a gas chromatograph (Model 8610C, SRI, CA) with helium carrier gas. The GC was equipped with MXT-VOL stationary column and a photo ionization detector (PID). Vinyl chloride (VC) and ethene were analyzed by GeoLab (Braintree, MA) using a purge and trap method and detected by GC-MS.

3. Results and Discussion

3.1 Culture Activity and Transformation Rates

Control experiments were conducted under a hydraulic gradient that is equal to one without electricity. Increased cis-DCE concentration appears in all sampling ports (Figure 2), indicating biological transformation of PCE. The increased cis-DCE started in ports 1 and 4 (which are 8 cm from the boundaries) followed by the inside ports (Port 2 and 3). Transport of lactate by hydraulic gradients was limited [9] because of the low hydraulic conductivity and lactate degradation. The increase in cis-DCE concentrations may be due to the presence of intrinsic pollution, and their utilization as electron donor. However, after this initial transformation, cis-DCE concentrations were persistent in all ports, showing final cis-DCE concentrations ranging from 20 to 60 mg L⁻¹ after more than 100 days. cis-DCE concentrations in pore water are higher than the initial PCE concentrations because of PCE desorption and because cis-DCE adsorption capacity is smaller than that of PCE. Overall, the results indicate some transformation occurred in the control experiments, but the transformation was slow, limited, and incomplete after more than 100 days.

Figure 2.

Tetrachloroethylene and cis-1,2-dichloroethene concentration in Control-B experiment.

For bioremediation experiments with electrokinetic injection of lactate, initial pore fluid PCE concentration was 16.6 mg L⁻¹ in EK-A and 24.7 mg L⁻¹ in EK-B. The reason for the difference is that after analyzing the initial conditions of EK-A, the procedure for EK-B was modified so that additional dissolved PCE was mixed with slurry to retain high initial concentration of PCE. PCE concentrations decreased to below detection limits in all ports in EK-A within the first 14 days (Figure 3). Although there was no significant TCE accumulation during the culture development process, small concentrations of TCE (about 2 to 10 mg L⁻¹) were detected in the ports after 10 days in EK-A and 15 days in EK-B. This may indicate that the transformation process is occurring at a slower rate than in batch water system, possibly due to the packed soil system. TCE concentrations decreased within 7 days and no further TCE accumulation was detected (data not shown).

Figure 3.

Tetrachloroethylene concentration in clay during EK enhanced bioremediation.

Increased cis-DCE concentrations were detected at all ports while PCE concentrations were decreasing (Figure 4). For EK-A, the concentration at Port 1 increased to 46.9 mg L⁻¹ after 14 days and then gradually decreased to 16.3 mg L⁻¹ after 40 days of processing. The same trend occurred at Ports 2, 3 and 4. cis-DCE concentrations increased to more than 70 mg L⁻¹ and then the concentrations decreased. cis-DCE degradation was relatively fast in the first 20 days after cis-DCE reached peak concentration and then the degradation slowed down. In EK-A Port 3, cis-DCE reached 80 mg L⁻¹ after 36 days, and decreased to 20 mg L⁻¹ after 54 days. However, the degradation rate was limited after that and an additional 30 days were required to consume the residual cis-DCE to a non-detectable level. Note that the cis-DCE concentration decreased first at Port 1 after 14 days, followed by Port 2 after 25 days, Port 3 after 32 days and finally Port 4 after 39 days. This trend indicates that the cis-DCE decrease followed a pattern starting from Port 1 to Port 4, or from the anode towards the cathode, reflecting the impact of electroosmosis.

Figure 4.

cis-1,2-dichloroethene concentration in clay during EK enhanced bioremediation.

The experiments showed that there is a time lag of 10 days (8 days for duplicate) for PCE transformation in Port 2 and 3, the middle ports of the clay cell. This time lag may be due to either a delay in the delivery of electron donors during this initial period or due to the effect of soil packing. After successful lactate injection in the clay, a decrease in PCE coupled with an increase in cis-DCE concentration was immediately detected. During the first 30 days, a zero order transformation rate was assumed for PCE because no PCE tail was detected at any of the ports. The transformation rates were calculated by fitting the data after the lag time and are summarized in Table 1. The overall PCE transformation rates were between 1.5 to 5.0 mg L⁻¹ d⁻¹.

Table 1.

Tetrachloroethylene (PCE) and cis-1,2-dichloroethene (cis-DCE) transformation rates in ports across the soil cell

Port No.	[Tetrachloroethylene]/ mg L−1 d−1		[1,2-dichloroethene]/ d−1
	EK–A	EK–B	EK–A	EK–B
#1	1.82	1.11	0.0577	0.0828
#2	3.48	2.15	0.0776	0.0748
#3	4.48	2.32	0.139	0.0735
#4	2.49	1.04	0.254	0.0964

cis-DCE transformation rates were analyzed after the PCE was completely transformed (no generation of cis-DCE). cis-DCE transformation in soil follows a first order transformation and the rates (summarized in Table 1) varied from 0.06 to 0.10 per day (except one port in EK-A which showed a high rate of 0.25 per day.) The rates are significantly higher than the natural attenuation rates as reported at 0.00044–0.00061 d⁻¹ [19]. The reason is that transformation in the lab study occurs under controlled and optimum conditions in the presence of a microbial culture that is designed for transformation of chlorinated ethenes. According to the first order transformation, the remediation of cis-DCE was controlled by the concentration of source substrate cis-DCE and the culture bioactivity. It is concluded that electrokinetic transport is effective for injection of lactate in clay, and the concentration of the electron donor is not a limiting factor for the reductive dechlorination under such conditions.

3.2 Final Dechlorination Product and Lactate Delivery

After cis-DCE concentrations decreased to below the detection limit, the bioremediation experiments EK-A and EK-B were continued for 14 and 35 days, respectively, to allow for further transformation of VC. After testing, the soil samples were sliced into 10 sections, centrifuged and the supernatants were analyzed. No PCE, TCE or cis-DCE were detected in any of the soil samples collected from the two experiments. Measurements showed presence of VC in the EK-A samples, but not in the EK-B samples. Ethene was the only major final product in samples collected from EK-B, with concentrations as high as 5 to 14.5 mg kg⁻¹ (Figure 5). An interesting observation in EK-B is that the final concentration of ethene decreased from the anode towards the cathode across the soil. The reason for this profile is not clear; there was enough lactate throughout the cell. However, this may be related to the final redox potential distribution which changed relatively linearly from −150 mV near the anode to less than −300 mV near the cathode. The results confirm that PCE can be completely transformed to ethene in 4 months across the 40 cm clay sample using lactate injection under electric fields.

Figure 5.

Final product of Tetrachloroethylene dechlorination with electrochemical injection of lactate (Duplicate A at 93 days and Duplicate B at 127 days).

A relatively uniform high concentration of lactate (more than 2000 mg L⁻¹) was measured across the soil samples (Figure 6). Higher lactate concentrations were detected near the anode side in EK-A, which again reflect the role of electroosmotic flow.

Figure 6.

Final lactate concentration across the clay with bioremediation (Duplicate A at 93 days and Duplicate B at 127 days).

3.3 Physicochemical Soil Properties

Circulation of the electrolytes through the cathode and anode prevented extreme changes in the electrolyte pH, which were kept at neutral conditions. Figure 7 shows the final pH and the Oxidation-Reduction Potential (ORP) across the soil samples. The pH in the soil was around neutral and varied between 7 to 7.5 and the ORP profile showed a chemically reducing environment across the soil, with values ranging from −150 mV near the anode to −300 mV near the cathode. Mixing of the electrolytes resulted in favorable pH and ORP conditions for reductive dechlorination.

Figure 7.

Final soil pH and ORP distribution by EK-enhanced bioremediation.

Initial water content (after consolidation) was 33.6% in EK-A and 26.3% in EK-B. The final soil water content increased to average respective values of 38.4% and 29.0% for EK-A and EK-B, indicating some swelling in both experiments. The increase in water content, though not significant compared to other electrokinetic applications, is due to the nonlinear electrical conductivity profiles during testing as reported by Alshawabkeh et al. [20].

The voltage gradient decreased from 0.9 to 0.2 V cm⁻¹ in EK-A after 35 days under 5.3 A m⁻². The current was then increased to 13.3 A m⁻², increasing the gradient to 1.1 V cm⁻¹, which dropped to 0.7 V cm⁻¹ at the end of testing. Similar behavior is noted in EK-B, but the current was adjusted to 13.3 A m⁻² after 15 days. Both experiments showed that the voltage tends to stabilize after 14 to 21 days of testing. The transport of lactate, as well as other ionic additives, resulted in an increase in electrical conductivity of the soil. Although this increase was not uniform across the sample, the total voltage decreased over time. It can be concluded that lactate delivery was achieved within first 21 to 28 days as the voltage gradient was stable in 14 to 21 days. Lactate transport is expected to occur at a rate of 3.7 cm d⁻¹ under 1V cm⁻¹ in this soil [9]. For a 40 cm sample, under a gradient of 0.5 V cm⁻¹, 28 days will be required to achieve lactate delivery. Although lactate delivery occurs in less than a month, biological transformation occurs at a relatively slow rate, as discussed earlier, and the samples were processed for longer periods to allow complete transformation.

3.4 Mass Balance

Mass balance calculations are difficult to calculate and to assess because of the complexity of reactive transport of additives and contaminants, the possible effects of sorption, transformation, gas formation, abiotic reduction at the cathodes or sorption on setup components. In this study, the total mass of chloroethenes was calculated by averaging the concentrations in pore water of the soil samples and assuming the final soil properties (porosity and dry density). Accounting for adsorption, the total mass (T) is calculated by:

$$T=K_d{c}_W{M}_S+c_W{V}_W$$ [1]

where K_d is the adsorption coefficient factor (L Kg⁻¹), c_W is the aqueous concentration in pore water (mg L⁻¹), V_W is the volume of pore water (L), and M_S is the mass of soil solid (Kg). Linear adsorption isotherm is assumed based on laboratory data.

Only PCE, TCE, and cis-DCE were regularly monitored during the experiments, and the total mass was calculated without considering VC and ethene. The K_d values for PCE and TCE were measured as 1.77 L kg⁻¹ and 0.81 L kg⁻¹, respectively. Assuming same effective organic carbon fraction and adsorption, average K_d value for cis-DCE was predicted as 0.23 L kg⁻¹.

Figure 8 shows the total mass of PCE, TCE and cis-DCE based on values measured at the ports over time. The initial total mass was calculated using the PCE/TCE/DCE concentrations immediately after consolidation. The final mass for EK-B was calculated based on the total ethene concentration in final soil samples. The total mass of chloroethenes was relatively maintained during the first month of treatment, indicating that transformation is occurring without losses due to other reasons (e.g. electroosmotic flow or evaporation at the boundaries). The loss of chloroethenes that is calculated based on PCE/TCE/ cis-DCE occurs when peak cis-DCE concentration is reached, which is then followed by continuous formation of VC. The recovery rate of initial chloroethenes as ethene at the end of testing in Ek-B was 67%, which further substantiates that biotransformation is the primary mechanism of remediation under these conditions.

Figure 8.

Amount of PCE, TCE and cis-DCE in clay during bioremediation with electrochemical injection. Amount is calculated in milli-mole (mmol). Final amount in EK-B refers to ethane.

The remediation process involves three primary stages: 1) lactate injection, 2) PCE desorption and 3) PCE transformation. Under conditions of this study, 21 to 28 days are sufficient for lactate delivery through the 40 cm sample, and more than 30 days are required for biological transformation of PCE to ethene in batch samples. Thus, complete transformation of PCE to ethene will require at least 60 days across the 40 cm clay sample (hydraulic conductivity is 2 × 10⁻⁷ cm d⁻¹). However, column testing shows that more than 120 days are required for complete transformation. The difference may be due to the less than ideal conditions for biological activity within the clay pores. It is possible that microbial activity decreased (by as much as 50%) due to the compacted clay condition and small pore space available for microbial growth.

4. Conclusions

Biotransformation of PCE in clay by hydraulic injection of additives is difficult to implement because of the low permeability of clay deposits. This study evaluates biotransformation of PCE in clay by electrokinetic injection of lactate, a common electron donor for anaerobic biodegradation. Transformation of PCE in clay occurs but at a slower rate, ranging from 1.5 to 5.0 mg L⁻¹ d⁻¹ which is about half the transformation rate expected in sand under similar conditions. No significant formation of TCE appears to occur while cis-DCE transformation occurs under first order transformation rates of 0.06 to 0.10 per day. It is concluded that ionic migration will deliver organic additives and induce biological activity and complete PCE transformation in clay, even though the transformation occurs under slower rates compared to ideal conditions.