Treatment Of Dissolved Perchlorate Biology Essay

Published:

Waters containing perchlorate and dissolved nitrate, derived from detonated explosives and solid propellants, often also contain elevated concentrations of other dissolved constituents, including sulfate. Four column experiments, containing mixtures of silica sand, zero-valent iron (ZVI) and organic carbon (OC) were conducted to evaluate the potential for simultaneous removal of nitrate, sulfate and perchlorate. Initially, the flow rate was maintained at 0.5 pore volume (PV) day-1, and then decreased to 0.1 PV day-1 after 100 PV of flow. Nitrate concentrations decreased from 10.8 mg L-1 (NO3-N) to trace levels through NO3- reduction to NH4+ using ZVI alone, and through denitrification using OC. Observations from the mixture of ZVI and OC suggest a combination of nitrate reduction and denitrification. Up to 71% of input sulfate (24.5 ± 3.5 mg L-1) was removed in the column containing OC and more than 99.7% of the input perchlorate (857 ± 63 μg L-1) was removed by the OC- and ZVI+OC-containing columns as the flow rate was maintained at 0.1 PV day-1. Nitrate and perchlorate removal followed first-order and zero-order rates, respectively. Nitrate >2 mg L-1 (NO3-N) inhibited perchlorate removal in the OC-containing column but not in the ZVI+OC-containing column. Sulfate did not inhibit perchlorate degradation within any of the columns.

Lady using a tablet
Lady using a tablet

Professional

Essay Writers

Lady Using Tablet

Get your grade
or your money back

using our Essay Writing Service!

Essay Writing Service

Keywords: perchlorate, nitrate, sulfate, zero-valent iron, organic carbon

1. Introduction

Perchlorate (ClO4-) is a common contaminant in groundwater and surface water. Approximately 90% of perchlorate salts are manufactured as ammonium perchlorate, which is widely used in large volumes of solid propellants for rockets, missiles, explosives, and pyrotechnics. A variety of non-military sources, such as the use and manufacture of road flares, perchloric and chloric acids, fireworks displays, and blasting agents used in mining and construction, can also cause widespread, low concentration perchlorate contamination in water. In addition to these anthropogenic sources, naturally-occurring perchlorate generated via atmospheric processes and contained in nitrate fertilizers and some natural minerals also contributes low levels of perchlorate in some parts of the world .

Perchlorate is highly soluble, mobile, and recalcitrant in the environment. It is potentially toxic to various forms of life, with low concentrations inhibiting iodide uptake in human thyroid and animal thyroid glands . Due to these health impacts, the US EPA adopted the National Research Council (NRC) recommended reference dose of 0.0007 mg kg-1 per day for perchlorate as an interim health advisory in 2005, which translates to a drinking water equivalent level (DWEL) of 24.5 µg L-1.

Physical-chemical treatment technologies, such as ion exchange, carbon adsorption, and advanced oxidation are effective for treating a range of contaminants, but are less cost-efficient and effective for removing perchlorate from water . Ion-exchange, a widely accepted water treatment technology, can effectively remove perchlorate from water; however, the highly saline (7-12%) brine generated during the ion-exchange process requires costly management before disposal . Microbial reduction of perchlorate is an area of intense interest because this strategy is relatively cost-effective and environmentally compatible . The hazardous perchlorate is converted into two innocuous compounds-chloride and oxygen-catalyzed by at least two separate enzymes .

Perchlorate and nitrate are often found as co-contaminants in water, as a large number of propellants, blasting agents, and explosives contain perchlorate as well as nitrogen-containing compounds. Moreover, nitrate-containing compounds are widely used in agriculture and perchlorate-containing pyrotechnics. High concentrations of sulfate, derived from sulfide oxidation at some mining sites, also require management. Because perchlorate- and nitrate-containing blasting agents and explosives are used at mine sites, water containing co-mingled perchlorate, nitrate, and sulfate can develop . The purpose of this study was to identify an effective, economical, and feasible technology to remediate perchlorate-, nitrate-, and sulfate- contaminated waters associated with mining and blasting sites. A series of column experiments was conducted using mixtures of zero-valent iron (ZVI) and wood chips (organic carbon, OC) as reactive media to remove co-mingled perchlorate, nitrate, and sulfate in simulated groundwater.

2. Material and methods 

2.1 Column design and experimental setup

Four acrylic columns were used, each 30 cm in length with an internal diameter of 5 cm. Influent ports were located at the base of each column, and effluent ports were located at the top of each column for discharge of the effluent solution and for sample collection. In addition, 13 equally spaced sampling ports were installed at approximately 2.1 cm intervals along the length of the columns. Both the bottom and top layers of the four columns were packed with a 1.0 cm thick layer of silica sand as a non-reactive material to separate the reactive mixture from the influent and effluent end ports.

Lady using a tablet
Lady using a tablet

Comprehensive

Writing Services

Lady Using Tablet

Plagiarism-free
Always on Time

Marked to Standard

Order Now

Column 1 was packed with 100% silica sand (SS) as a control. Column packings were composed of 50 vol.% granular ZVI with the balance as SS (Column 2), 50 vol.% wood chips (OC) with the balance as SS (Column 3), and a mixture of 10 vol.% ZVI and 40 vol.% OC adjusted with 50 vol.% SS (Column 4). The granular ZVI (0.17-1.41 mm) was obtained from Connelly-GPM, Inc. (Chicago, USA). The silica sand (0.6-0.8 mm) was obtained from Silica Company (Ottawa, USA). The wood chips (~1-9.5 mm, deciduous in type) were obtained from a lumber producer in Waterloo, ON. Prior to initiating the experiments, all four columns were placed in an anaerobic glove box (COY, Ltd., Grass Lake, USA) that contained 5% H2 and 95% N2. The columns were flushed with CO2 (g), which is more soluble in water than N2 and O2, for 24 h to displace atmospheric gases from the void pore space of the column packing and enhance saturation of the packing material. The columns were then wet with simulated groundwater, composed of CaCO3 saturated water, over a 48 h period.

A conservative tracer test was performed to determine the flow and transport characteristics of each column. The computer code CXTFIT 2.0 , a series of analytical solutions to the one-dimensional advection-dispersion equation with non-linear, least-squares parameter optimization, was used to determine the average linear velocities and dispersion coefficients from the column effluent concentration data. The flow rates of the column experiments were approximately 0.5 pore volumes (PV) per day during the conservative tracer test.

The input solution was prepared by adding nitrate, sulfate, and perchlorate to the CaCO3 saturated water as sodium salts to obtain concentrations of 10.8 ± 0.3 mg L-1 (NO3-N) nitrate, 24.5 ± 3.5 mg L-1 sulfate, and 857 ± 63 µg L-1 perchlorate. The ratios of concentrations were based on measurements of effluent from test-scale waste-rock piles [8]. The purpose of using CaCO3 saturated water as the input solution is to simulate the presence of HCO3- and CO32-, which are common in natural surface water and groundwater. Prior to introducing the input solution, Columns 3 (OC) and 4 (ZVI+OC) were saturated with a solution containing 5% (vol.) sodium lactate for 3 d to promote the growth of microorganisms. The input solution initially was displaced through the four columns from the bottom to the top at a flow rate of 0.5 PV day-1 during the first stage of the experiment. During the second stage of the experiment, the flow rates were decreased to 0.1 PV day-1 (after 99 PV in Column 2, 112 PV in Column 3, and 110 PV in Column 4) to evaluate the effect of residence time on contaminant removal. Three profiles were collected along the columns during the course of the experiments: after 95.4, 115, and 129 PV of flow passed through Column 2 (ZVI); after 108, 132, and 151 PV of flow through Column 3 (OC); and after 106, 128, and 144 PV of flow through Column 4 (ZVI+OC).

2.2 Sample collection and analytical methods

Water samples were collected using 30 mL glass syringes attached directly to the ports along the length of the columns and to the effluent ports, such that the syringes were filled at the same rate as the input solution being introduced to the columns. Except where noted, all samples were passed through 0.45 µm cellulose acetate filters prior to measurement. The pH, Eh, and alkalinity were determined immediately after sampling. All other samples were kept chilled (<4°C) until analysis within one month of collection.

Values of pH and Eh were measured on unfiltered samples in sealed cells to minimize O2 exposure. The pH measurements were made using a Ross combination glass electrode (Orion 815600) calibrated using standard 4.0 and 7.0 buffers and then checked against a 10.0 buffer. The Eh measurements were made using a Pt-billeted Ag/AgCl combination electrode (Orion 9678BNWP). The performance of the electrode was checked using Zobell's and Light's solutions before and after each measurement. Measurements were corrected to the standard hydrogen electrode (SHE). Alkalinity was determined using a Hach digital titrator with bromcresol green/methyl red indicator and 0.16 N H2SO4.

The concentrations of major anions (Br-, NO3-, NO2-, Cl- and SO42-) were determined by ion chromatography (IC; DX600, Dionex, Sunnyvale, USA). Ammonia (NH3-N) concentrations were determined using a Hach spectrophotometer DR/8400 (SMEWW, 2005) (salicylate method). Dissolved hydrogen sulfide (H2S) was determined using the methylene blue spectrophotometric method .

Lady using a tablet
Lady using a tablet

This Essay is

a Student's Work

Lady Using Tablet

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Examples of our work

Filtered 20 mL samples were collected separately in 30 mL polyethylene bottles for ClO4- analysis. A headspace was maintained in these sample bottles to minimize the possibility of anaerobic conditions developing during storage. Perchlorate was analyzed following the method described by with the addition of equal aliquots of isotopically enriched sodium perchlorate (NaCl18O4, Cambridge Isotope Laboratories, Andover, USA) to all the samples and calibration standards as an internal standard (IS). All water samples were prepared by elution through one OnGuard-II Ba cartridge and one OnGuard-II H cartridge (Dionex, Sunnyvale, USA) in series to remove sulfate and carbonate, respectively. Perchlorate was analyzed by high performance liquid chromatography (HPLC; Agilent 1100, Agilent Technologies, Mississauga, Canada) followed by electrospray tandem mass spectrometry (MS/MS; 4000 Q TRAP, Applied Biosystems, Foster City, USA) and detected by negative electrospray ionization mass spectrometry using Multiple Reaction Monitoring (MRM) detection. The instrument and practical detection limits were 0.02 and 0.05 μg L-1, respectively. QA/QC results showed that relative method recovery over the entire standard curve (0.5-100 µg L-1) fell into the range of 98-110%, and the relative internal standard recovery for unknown samples was 71 to 118%.

Results and Discussion

Conservative tracer test and control column results

The transport model CXTFIT 2.0 was used to determine the velocity and dispersion coefficient for each column during the first stage of the experiments. The fitted velocities estimated by CXTFIT using the Br- breakthrough curve data ranged from 40.3 to 54.3 m yr-1 for the four columns. These values were in agreement with the average linear velocities calculated from porosity and flow volume measurements, which ranged from 49.1 to 59.6 m yr-1, with the exception of Column 2 (ZVI). The dispersion coefficients ranged from 0.1 to 1.2 m2 yr-1 for the four columns (Table 1). Measurements made on the control column indicated that transport of ClO4-, NO3- and SO42- was conservative (Fig. 1)

3.2 Geochemistry conditions of columns

The three pH profiles within Columns 2-4 measured at different times almost overlapped, exhibiting similar values with distance (Fig. 2). The influent water pH was 8.3 throughout the experiments. The average pH value within Columns 2 (ZVI) was 9.7. The increase in pH was due to the reduction of water by ZVI. The average pH in Column 4 (ZVI+OC) was slightly lower, at 8.6. The pH of Column 3 (OC), which did not contain ZVI, was much lower at pH 7.4. Similarly, the average pH of the column effluent was 9.7 in Column 2 and 7.2 in Column 3, remaining within 0.5 pH units throughout the entire experiment (Fig. 3). The effluent pH of Column 4 increased quickly from below 7 to about 9.5 during the first 10 PVs and then decreased to about 8.6 for the remainder of the first stage of the experiment.

The three alkalinity profiles for each of Columns 2, 3, and 4 indicated similar average values and changing trends, with slight increases as the input solution advanced through the columns (Fig. 2). Total alkalinities with an average value of 96 mg L-1 (as CaCO3) were generated in Column 2 (ZVI) effluent; in the effluent of the columns containing OC, slightly higher alkalinity values of 119 mg L-1 (as CaCO3) in Column 3 (OC) and 113 mg L-1 (as CaCO3) in Column 4 (ZVI+OC) were observed (Fig. 3).

The decrease in flow rate in the second stage of the experiments did not cause obvious shifts in pH and alkalinity profiles, but the prolonged residence time likely enhanced the reducing conditions within Columns 2, 3, and 4 (Fig. 2). A much higher average Eh value of 300 mV was observed within Column 3 (OC) than in the middle parts of Columns 2 (ZVI; -200 mV) and 4 (ZIV+OC; -400 mV). Addition of the lactate solution into Columns 3 and 4 likely resulted in relatively low Eh and high alkalinity values in column effluent at the beginning of the first stage of the experiment compared to Column 2 (Fig. 3).

3.3 Removal of nitrate, sulfate, and perchlorate in columns

Reduction of nitrate by ZVI has been observed to proceed rapidly through nitrite to ammonium. The proposed pathway for the overall reaction follows :

NO3- + 4Fe0 + 10H+ → 4Fe2+ + NH4+ + 3H2O. (1)

Nitrate was almost completely removed in the column containing ZVI (Column 2) from the average input concentration of 10.8 mg L-1 (NO3-N) to a trace concentration. Nitrite concentrations remained < 0.02 mg L-1 (NO2-N) and the average ammonium (NH4+) concentration was 10.7 mg L-1 (NH3-N) indicating that NH4+ was the principal byproduct throughout the course of the experiment (Fig. 1). The concentrations of total N (sum of NO3-N, NO2-N, and NH3-N) in Column 2 (ZVI) effluent were consistent with the input nitrate concentrations, indicating complete conversion of NO3- to NH4+.

Bacterially mediated denitrification by organic carbon (OC) such as wood waste as a carbon source is thermodynamically favored and leads to the step-wise reduction of nitrate to form nitrogen gas:

NO3-(aq) → NO2-(aq) → NO(enzyme complex) → N2O(gas) → N2(gas). (2)

The overall reaction can be expressed as :

5CH2O + 4NO3- → 2N2 + 5HCO3- + H+ + 2H2O. (3)

The removal of nitrate in Column 3 containing OC was variable (Fig. 1). In the first stage of the experiment (108 PV) at a flow rate of 0.5 PV day-1, up to 2.6 mg L-1 NO3-N nitrate remained in the effluent. In addition, up to 2.4 mg L-1 NO2-N nitrite, an intermediate reaction product, was observed in the effluent in the first stage of the experiment. However, nitrate was more completely removed in the second stage of the experiment when the flow rate decreased to 0.1 PV day-1, from an average input concentration of 10.8 mg L-1 to < 0.02 mg L-1 (NO3-N) without measurable nitrite observed. A trace of ammonia (average value 0.1 mg L-1 (NH3-N)) was observed in Column 3 effluent samples, perhaps due to the decomposition of OC. In the first stage of the experiment, the total N concentrations in the Column 3 (OC) effluent progressively rose from 0.1 to 5.2 mg L-1. The increase in total N concentrations is predominantly due to an increase in the concentrations of NO3- and NO2-, indicating the increase in total N was due to incomplete denitrification, possibly due to a depletion of labile organic carbon. In the second stage of the experiment the total N concentration decreased to as low as 0.7 mg L-1 (N), suggesting that the rate of organic carbon consumption was sufficient to result in complete denitrification during this stage.

The mixture of ZVI+OC in Column 4 resulted in extensive removal of nitrate, no detectable release of nitrite, and the release of 9.5 mg L-1 of ammonia (NH3-N) at the beginning of the experiment that gradually decreased to 6.6 mg L-1 (NH3-N) at 100 PV (Fig. 1). This decreasing trend suggests that nitrate reduction by ZVI predominated early in the experiment and that denitrification became established, resulting in partial conversion of NO3- to N2(g). In stage 2 the average total N concentration was 5.5 mg L-1, almost exclusively as NH4+, indicating an increase in the extent of denitrification as the residence time increased, resulting in approximately equal removal by nitrate reduction and denitrification. Less ammonium is observed when ZVI is used in conjunction with a microbial consortia to reduce nitrate .

Effluent sulfate concentrations did not decrease in the columns containing ZVI (Column 2) or the mixture of ZVI and OC (Column 4) over the duration of the experiment (Fig. 1). Sulfate was observed to break through Column 3 (OC) in the first stage of the experiment; however, effluent sulfate concentrations in Column 3 decreased from 24.7 mg L-1 after 108 PV in the first stage of the experiment to 7.1 mg L-1 after 132 PV in the second stage of the experiment (when the flow rate slowed), with approximately 71% of the input sulfate (24.5 mg L-1) removed. The effluent sulfate concentration then increased to 15.0 mg L-1 after 151 PV in the second stage of the experiment. This sulfate removal is attributed to the onset of biologically mediated sulfate reduction coupled to OC oxidation :

2CH2O + SO42- → 2HCO3-+ H2S. (4)

The decrease in the rate of sulfate removal over time was probably due to depletion of labile OC over the long-term operation of the experiment. The inhibition of sulfate reduction observed in Column 4 (ZVI+OC) might be attributed to the higher pH conditions developed in this column compared to Column 3 (OC) (Figs. 2 and 3).

Although reduction of perchlorate by ZVI is thermodynamically favored, with ∆G°=-596.27 kcal mol-1, perchlorate was not removed by ZVI in Column 2 throughout the experiment; this is likely due to the high activation energy required for perchlorate reduction . The effluent perchlorate concentration was 857 ± 63 μg L-1, similar to the input concentration (Fig. 1). The total effluent Cl concentrations (sum of ClO4- and Cl in μmol L-1) in Column 2 (ZVI) were much higher than the input perchlorate concentrations, this difference may be due to the release of Cl initially present on the ZVI (Fig 4).

The effluent concentrations of perchlorate in Column 3 decreased from 547 μg L-1 after 108 PV in the first stage of the experiment to 28 μg L-1 after 151 PV in the second stage of the experiment (Fig. 1). Similarly, in the second stage of the experiment when the flow rate was lowered, OC combined with ZVI exhibited more uniform removal of perchlorate in Column 4 compared to Column 3. The concentration of perchlorate in the effluent of Column 4 decreased from 679 μg L-1 after 106 PV in the first stage of the experiment to 1.37 μg L-1 after 144 PV in the second stage of the experiment. Previously studies suggest the following pathway for biological degradation of perchlorate :

. (5)

The first two reactions are catalyzed by (per)chlorate reductase. Chlorite dismutase catalyzes the disproportionation of chlorite into chloride and oxygen . The OC probably provided sufficient carbon and nitrogen for microbial degradation of perchlorate in Columns 3 (OC) and 4 (ZVI+OC); however, complete removal of perchlorate was not consistently observed until the second stage of the experiment (Fig. 1). Extensive removal of perchlorate initially was observed in Column 3, but the rate of perchlorate removal declined after 30 PVs. After the flow rate decreased, complete removal of perchlorate was observed. Initially, little perchlorate removal was observed in Column 4. More extensive removal was observed after 10 PVs, suggesting a period of acclimation was required before perchlorate reduction was established. After 60 PVs, the rate of perchlorate removal declined, suggesting depletion of labile OC. During stage 2 of the experiment, complete removal of perchlorate was observed, suggesting that the rate of OC fermentation was sufficient to provide labile OC for sustained perchlorate reduction.

Chloride, the final potential product of perchlorate biological degradation, should be released at an amount equivalent to the mass of perchlorate removed. The changes in chloride concentrations in both Columns 3 (OC) and 4 (ZVI+OC) effluent were inversely correlated to perchlorate concentrations. Very consistent concentrations of total effluent Cl (sum of ClO4- and Cl in μmol L-1) relative to the input perchlorate concentrations for these columns indicated that the expected mass of Cl was accounted for in both Columns 3 and 4 (Fig 4).

3.4 Removal rates of nitrate and perchlorate within columns

The two nitrate and perchlorate profiles collected for each of Columns 2, 3, and 4 in the second stage of the experiment were similar (Fig. 5). Therefore, only the profiles measured in the first stage of the experiment (95.4 PV for Column 2; 108 PV for Column 3, and 106 PV for Column 4) and one of the profiles measured in the second stage of the experiment (115 PV for Column 2; 132 PV for Column 3; and 128 PV for Column 4) were used for calculation of removal rate parameters.

Nitrate removal rates within Columns 2, 3, and 4 were consistent with a first-order rate model (Fig. 6) as reported in other studies :

C = C0 exp(-k1 ­t) (6)

RN = k1 C (7)

where C is the nitrate concentration (mg L-1), C0 is the initial nitrate concentration (mg L-1), k1 is a first-order rate constant (d-1), t is the residence time (d), and RN is the reaction rate (removal rate) of nitrate (mg L-1 d-1). The best fit nitrate degradation equation based on calculated residuals for each column (SigmaPlot, SPSS Inc.) was selected from the two first-order expressions derived from the two stages of the experiment.

Nitrate removal by ZVI within Column 2 (ZVI) followed first order removal model, with equations of y=10.55exp(-0.96x) (R2=0.9829) in the first stage of the experiment (Fig. 6 (1)) and y=10.96exp(-0.68x) (R2=0.9991) in the second stage of the experiment (Fig. 6 (2)); however, neither of these equations can describe the degradation curves in both experimental stages well. A modified first-order removal rate equation y=10.99exp(-0.80x), which fell between the first and second stage data sets, was selected to describe the nitrate removal within Column 2 (ZVI).

Denitrification within Column 3 (OC) can be described by the equation: y=10.91exp(-0.62x) (R2=0.9934) (Fig. 6 (3), (4)). Both zero-order and first-order rate expressions have been observed to provide reasonable descriptions of the rate of denitrification (R2>0.86) . Similar findings were observed in this study: nitrate removal by ZVI and OC in Column 4 in both first and second stages of the experiment were consistent with both the first-order and zero-order rate equations in terms of R2 (>0.94) (Fig. 6 (5), (6)). However, to maintain consistency between the nitrate removal equations for Column 4 (ZVI+OC) and those for Column 2 (ZVI) and Column 3 (OC), an exponential expression y=11.54exp(-1.56x) (R2=0.9468) was selected to describe the rates of nitrate removal in Column 4 throughout the experiment. The removal rate of nitrate within Columns 2 (ZVI), 3 (OC), and 4 (ZVI+OC) can be represented by RN,2 = 0.80C mg L-1 day-1, RN,3 = 0.62C mg L-1 day-1, and RN,4 = 1.56C mg L-1 day-1, respectively. Given the same initial concentration, nitrate was removed much more rapidly in Column 4 than in Columns 2 or 3 (Fig. 5).

Perchlorate removal in some bioreactors has been observed to follow first-order reaction rates with respect to perchlorate concentration . However, in this study, perchlorate removal in Columns 3 (OC) and 4 (ZVI+OC) followed zero-order rate equations (Fig. 7):

, (8)

where C is the perchlorate concentration (µg L-1), t is the residence time (day), and k0 is a zero-order rate constant for perchlorate removal (µg L-1 day-1). The best fit equation for perchlorate removal, based on least squares regression, was also obtained using a zero-order rate expression derived from the two stages of the experiment. The perchlorate removal within Columns 3 (OC) and 4 (ZVI+OC) was described by: y = 955.17-135.51x (R2 = 0.9495) and y = 889.99-113.07x (R2 = 0.9677) (Fig. 7). The perchlorate removal rates within Columns 3 (OC) and 4 (ZVI+OC) followed RP,3 = 136 µg L-1 day-1 and RP,4 = 113 µg L-1 day-1. Perchlorate was removed more rapidly in Column 3 than in Column 4 (Figs. 5 and 7).

3.5 Effect of nitrate and sulfate on perchlorate removal rate

The impact of nitrate on perchlorate reduction is important because nitrate is a common co-contaminant in perchlorate-contaminated water. Most perchlorate reducing bacteria (PRB) are also denitrifiers, and the simultaneous removal of nitrate and perchlorate from contaminated waters has been observed . Likewise, the nitrate and perchlorate within Columns 3 (OC) and 4 (ZVI+OC) were removed simultaneously. However, the overall nitrate removal within both Columns 3 and 4 occurred at a more rapid rate than perchlorate removal (Fig. 5), which may have been due to competition for common electron donors (organic matter) between the two removal processes or the inhibition of perchlorate reductases in the presence of nitrate . Moreover, complete nitrate removal within Columns 3 and 4 was observed prior to complete perchlorate removal (Fig. 5), similar to other studies .

Nitrate has different effects on perchlorate removal rates. Nitrate was found to inhibit the perchlorate reduction rate in some studies but not others . In this study, the inhibition of perchlorate removal by nitrate was observed within Column 3 (OC) but not Column 4 (ZVI+OC). The inhibiting effects of nitrate on perchlorate removal occurred at nitrate concentrations >2 mg L-1 (NO3-N) within Column 3 (OC) throughout the experiment. Moreover, rapid perchlorate degradation did not proceed until nitrate concentrations were reduced to relatively low levels (normally <2 mg L-1 (NO3-N)), which is consistent with previous findings . However, the presence of nitrate within Column 4 (ZVI+OC) did not inhibit perchlorate removal (Fig. 5), which followed similar linear rates in the presence and absence of nitrate. These results suggest that adding ZVI to wood chips can potentially reduce the inhibition of nitrate on perchlorate removal. The presence of sulfate did not inhibit perchlorate removal in Column 3 (OC) or 4 (ZVI+OC) (Fig. 5), as is consistent with previous reports .

Conclusions

Reactive media containing OC or a mixture of ZVI+OC were found effective for removing nitrate and perchlorate from water. The removal of nitrate and perchlorate followed first-order and zero-order rates in these column experiments, respectively. Nitrate and perchlorate were removed simultaneously within the columns; however, complete nitrate removal occurred prior to complete perchlorate removal. Addition of ZVI to wood chips reduced the inhibition of nitrate on perchlorate degradation. Decreasing the flow rate from 0.5 to 0.1 PV day-1 resulted in more complete removal of nitrate, sulfate, and perchlorate. These results suggest that bioreactors containing OC or mixtures of OC and ZVI may be suitable for treating nitrate, perchlorate and sulfate at mining and blasting sites.

Acknowledgements

Funding for this research was provided by the Ontario Ministry of Research and Innovation - Ontario Research Excellence Fund and the Natural Sciences and Engineering Research Council of Canada. Laura Groza and Joy Hu provided assistance with the perchlorate and anion analyses. Additional technical support was provided by Peng Liu.