A long-term performance test on an autotrophic denitrification column for application as a permeable reactive barrier

Chemosphere 73 (2008) 723–728

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

A long-term performance test on an autotrophic denitrification column for application as a permeable reactive barrier Hee Sun Moon a, Do Yun Shin b, Kyoungphile Nam b, Jae Young Kim b,* a b

Department of Civil and Environmental Engineering, Princeton University, NJ 08544, USA Department of Civil and Environmental Engineering, Seoul National University, 599 Gwanangno, Gwanakgu, Seoul 151-744, Republic of Korea

a r t i c l e

i n f o

Article history: Received 24 April 2008 Received in revised form 19 June 2008 Accepted 23 June 2008 Available online 22 August 2008 Keywords: Permeable reactive barrier Long-term performance test Autotrophic denitrification Elemental sulfur Thiobacillus denitrificans

a b s t r a c t The long-term performance of a sulfur-based reactive barrier system was evaluated using autotrophic denitrification in a large-scale column. A bacterial consortium, containing autotrophic denitrifiers attached on sulfur particles, serving as an electron donor, was able to transform 60 mg N L1 of nitrate into dinitrogen. In the absence of phosphate, the consortium was unable to remove nitrate, but after the addition of phosphate, nitrate removal was readily evident. Once the column operation had stabilized, seepage velocities of 1.0  103 and 0.5  103 cm s1, corresponding to hydraulic residence times of 24 and 48 h, respectively, did not affect the nitrate removal efficiency, as determined by the nitrate concentration in the effluent. However, data on the nitrate, nitrite and sulfate distribution along the column indicated differential transformation patterns with column depths. Based on the dinitrogen concentration in the total gas collected, the denitrification efficiency of the tested column was estimated to be more than 95%. After 500 d operation, the hydrodynamic characteristics of the column slightly changed, but these changes did not inhibit the nitrate removal efficiency. Data from a bacterial community analysis obtained from four parts of the column demonstrated the selective a spatial distribution of predominant species depending on available electron acceptors or donors. Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction Many technologies are available for treating nitrate from groundwater, such as ion exchange, reverse osmosis, electrodialysis and distillation. However, such technologies inevitably involve the removal of groundwater for further treatment. As an alternative, bacterial denitrification, either heterotrophic or autotrophic, has been proposed and especially, autotrophic process combined with a permeable reactive barrier system is likely to be feasible as an in situ groundwater treatment. Autotrophic denitrification seems to be competitive at implementing the process into a permeable reactive barrier system, as it requires no external carbon source and exhibits lower cell yield compared to heterotrophic denitrification (Hiscock et al., 1991; Soares, 2002; Rocca et al., 2007; Sierra-Alvarez et al., 2007). Since the 1970s, autotrophic denitrification systems have found applications in the treatment of waste effluent (Sikora and Keeney, 1976), groundwater (Kruithof et al., 1988; Gayle et al., 1989; Hiscock et al., 1991) and landfill leachate (Koenig and Liu, 1996). Our group has also reported on the feasibility of an in situ biological reactive barrier system consisting of sulfur/limestone and autotrophic denitrifiers for the treatment of nitrate in bank filtrate * Corresponding author. Tel.: +82 2 880 8364; fax: +82 2 889 0032. E-mail address: [email protected] (J.Y. Kim). 0045-6535/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2008.06.065

(Moon et al., 2004). The effect of the composition of reactive materials and co-contaminant in autotrophic denitrification (Moon et al., 2006a) and different alkalinity sources on sulfur/limestone autotrophic denitrification (Moon et al., 2006b) were also explored to find the applicability of the autotrophic denitrification process to a permeable reactive barrier system in groundwater. A variety of reactions can occur as groundwater flows through a permeable reactive barrier, which can generate mineral precipitates and/or gases, the accumulation of which can diminish the effective porosity of the barrier (Mackenzie et al., 1999). This may limit the access of pollutants to reactive materials, and also reduce the hydrodynamic characteristics of the reactive barrier toward the creation of constricted flow paths and preferential paths of water flow through the porous media (Phillips et al., 2000; Vikesland et al., 2000, 2003; Sorel et al., 2001). Therefore, the long-term success of a permeable reactive barrier is critically dependent on its ability to maintain the reactivity of the reactive materials, with a sufficient hydraulic residence time for contact with the pollutants (Fogler, 1999; Farrell et al., 2000). As part of a project for developing a biological permeable reactive barrier system for nitrate-contaminated groundwater, an autotrophic denitrification, using elemental sulfur as an electron donor, was chosen, with it long-term performance tested using a large-scale column. The objectives of this study were: (1) to determine the autotrophic nitrate removal efficiency according to

724

H.S. Moon et al. / Chemosphere 73 (2008) 723–728

different hydraulic residence times, (2) to investigate the nitrogen mass balance for evaluation of the complete denitrification efficiency and (3) to compare the hydrodynamic characteristics of the column before and after long-term operation. The other objectives were to obtain an idea of the appropriate barrier thickness to attain acceptable denitrification efficiency and to understand a microbial community in a permeable reactive barrier.

2. Materials and methods 2.1. Experimental setup and performance test A large-scale column reactor was made of acrylic, with an inside diameter and height of 30 and 120 cm, respectively. The column was packed with sulfur granules, 5–10 mm in diameter, as an electron donor, and limestone, 5–10 mm in diameter, as an alkalinity source, at a 3:1 volume ratio (Moon et al., 2004; Moon et al., 2006a). A schematic diagram of the column reactor is shown in Fig. 1. A bacterial consortium, capable of transforming nitrate into dinitrogen under anoxic, autotrophic denitrifying conditions, was enriched from anaerobic digested sludge and maintained at 30 °C in a liquid medium, containing (g L1) 2 KNO3, 5 Na2S2O3  5H2O, 2 K2HPO4, 1 NaHCO3, 0.5 NH4Cl, 0.5 MgCl2  6H2O and 0.01 FeSO4  7H2O in sterile distilled water (Koenig and Liu, 1996). The consortium was found to contain Thiobacillus denitrificans as a key member, which utilizes sulfur as an electron donor (Moon et al., 2006a; Moon et al., 2006b). The consortium was introduced into the column, and then recycled for 7 d for the microorganisms to attach onto the surface of sulfur particles. The column reactor was continuously fed with a synthetic groundwater in an upflow mode, with a seepage velocity of either 1.0  103 (flow rate: 20 mL min1) or 0.5  103 cm s1 (flow rate: 10 mL min1). The synthetic groundwater contained 0.18 CaCl2  2H2O, 0.4 MgCl2  6H2O, 0.2 KH2PO4, 0.03 K2SO4 and 0.434 KNO3 g L1 and pH was 5.23 (KH2PO4 was not added during phase I operation). The concentrations of nitrate, nitrite and sulfate in the influent and effluent solutions were determined at appropriate intervals, and the evolved gas from the column was collected into a 1 L of Tedlar bag for an analysis of the gas composition. During the column operation, the effluent pH was maintained between 6.7 and 6.9.

The large-scale column was operated at 20 °C for 510 d. Experiments were performed in four phases, as follows; (i) phase I: no phosphate addition, with a flow rate of 20 mL min1 (1.0  103 cm s1 of seepage velocity), (ii) phase II: phosphate addition (as 0.2 g L1 KH2PO4), with a flow rate of 20 mL min1, (iii) phase III: synthetic groundwater containing phosphate, with a flow rate of 10 mL min1 (0.5  103 cm s1 of seepage velocity) and (iv) phase IV: synthetic groundwater containing phosphate, with a flow rate of 20 mL min1. During phases II to IV, phosphate concentrations were the same as 0.2 g L1 KH2PO4 with different seepage velocities, which yielded to the N:P ratios of 1.32:1 for phase II, III, and IV, respectively. To investigate whether the hydrodynamic characteristics of the column changed, tracer tests were conducted before and after operation, using a 200 mg L1 bromide solution. A bacterial community analysis along the column was also performed. After 510 d operation, the column was sacrificed, and the reactive materials in the column (i.e., sulfur granules and limestone) destructively were sampled along the column from the bottom (i.e., 0– 30, 30–50, 50–80 and 80–100 cm). The samples were subjected to extraction of the bacterial bulk DNA, and then analyzed by PCR-DGGE (polymerase chain reaction-denaturing gradient gel electrophoresis). 2.2. Analytical methods A Dionex 500 IC (ion chromatography), equipped with an AS14A anion column and a CD20 conductivity detector, was used to analyze for the nitrite, nitrate, sulfate and bromide concentrations in the influent and effluent solutions. The volume of gas evolved from the autotrophic denitrification column was measured everyday during phases III and IV. The composition of the collected gas was analyzed using gas chromatography equipped with a thermal conductivity detector (Younglin 600, Younglin, Korea). For obtaining 16S rRNA genes for the bacterial community structure analysis, the extracted genomic DNA was first amplified with 27F and 1522R primers, using the GeneAmp PCR System 9700 (Applied Biosystems, USA). Another round of PCR was performed with the amplified 16S rRNA genes for DGGE analysis. Firstly, 1 lL each of GC-338F (GC clamp-CTCCTACGGGAGGCAGCAGT) and 518R (GTATTACCGCGGCTGCTG) primer solutions (10 pmol lL1 each), 20 ng of the amplified, purified 16S rRNA genes and Accupower HotStart PCR Premix (Bioneer, Korea) were

Gas

Effluent

Gas sampling port

Large-scale column reactor

Gas collection system Pump Reservoir Fig. 1. A schematic diagram of the large-scale column reactor and gas collection system used in this study.

725

H.S. Moon et al. / Chemosphere 73 (2008) 723–728

When the sulfur-utilizing, autotrophic bacterial consortium had successfully attached to the sulfur particles, a phosphate-free synthetic groundwater, containing 60 mg N L1 of nitrate, was flowed into the column at a seepage velocity of 1.0  103 cm s1. This was conducted to verify the effect of phosphate on the bacterial denitrification activity. During the initial 60 d operation, no nitrate removal was evident. However, the autotrophic denitrification activity seemed to decrease with time (Fig. 2a, Phase I), probably due to the lack of phosphate in the influent. Inorganic phosphate tends to form precipitates with Na+, Ca2+, Mg2+ and Fe3+ in soil and; thus, the phosphate mobility is much slower than that of nitrate (Gonzales-Pradas et al., 1993). Therefore, low levels of phosphate may limit denitrification in deep soils and underlying aquifers, even when the overlying soils are contaminated with high levels of both phosphate and nitrate. In these situations, a phosphate supplement may be required to obtain proper denitrification activity (Hunter, 2003). After 0.2 g L1 KH2PO4 had been supplemented to the system, the nitrate concentration in the effluent sharply decreased, indicating the active expression of autotrophic denitrification activity (Fig. 2a, Phase II). However, the nitrate was unlikely to be completely transformed to dinitrogen; and more than 40 mg N L1 nitrite was observed in the effluent at this stage (Fig. 2b). Since a significant amount of nitrite was observed during Phase II of the operation, which was thought to be due to insufficient contact time between nitrate and/or its metabolites and attached microorganisms, the flow rate was reduced to 10 mL min1, giving a seepage velocity of 0.5  103 cm s1, providing sufficient hydraulic residence time (i.e., 48 h) on the 150th day of column operation. As shown during the Phase III period (Fig. 2), almost all the nitrate was transformed and; moreover, no nitrite was detected. Accordingly, the sulfate concentration in the effluent dramatically increased during this period (Fig. 2c), which resulted from sulfur oxidation. On the 250th day of operation, the seepage velocity was increased to 1.0  103 cm s1, again providing a hydraulic residence time of 24 h (Fig. 2a, Phase IV). However, complete nitrate removal was still attained. Interestingly, no accumulation of nitrite was observed, unlike during Phase II that had also had a hydraulic residence time of 24 h; sulfate production was also evident, as during Phase III. These data present an interesting aspect of the transformation of nitrate during the operation of the autotrophic denitrification column when using elemental sulfur as an electron donor. Initially, a hydraulic residence time of 24 h was not sufficient for complete nitrate transformation, as shown in Phase II, which was supported

Phase IV : Seepage velocity :1.0x10-3 cm s-1

a

90

Nitrate (mg-N L-1)

80 70 60 50 40 30 Influent Effluent

20 10 0 60

b

50 -1

Nitrite (mg-N L )

3.1. Nitrate removal

Phase III : Phase II : Phosphate Seepage velocity : addtion 0.5x10-3 cm s-1

40 30 20 10 0 400

c

350

Influent Effluent

300

-1

3. Results and discussion

Phase I : Phosphate free 100

Sulfate (mg-S L )

combined to give a final reaction volume of 20 lL. The next steps involved: (i) an initial denaturing step at 94 °C for 10 min, 30 cycles of denaturation (at 94 °C for 45 s), (ii) an annealing step at 55 °C for 45 s and (iii) an extension step at 72 °C for 45 s. Using these PCR products, a DGGE analysis was performed using the D-Code system (Bio-Rad, USA). Gels consisted of 1.0-mm-thick 9% polyacrylamide, with a denaturant gradient of 35–55% urea-formamide solution (100% concentration of the denaturant was 7 M urea-40% deionized formamide). Electrophoresis was performed in 1  TAE (40 mM Tris, 20 mM acetic acid and 1 mM EDTA; pH 8.3) at 60 °C and 100 V for 7 h. The DGGE gels were stained with 1  GreenStar staining dye solution (Bioneer, Korea) and bands were visualized using a UV illuminator (302 nm, Vilber Lourmat, France). The DGGE band intensities were quantified by using a Bio-Rad Quantity One (v.5.2).

250 200 150 100 50 0 0

30 60 90 120 150 180 210 240 270 300 330 360 390 420 450 480 510 540

Elapsed time (d) Fig. 2. Long-term operation of the autotrophic denitrification column, and the concentrations of (a) nitrate, (b) nitrite and (c) sulfate.

by the absence of sulfate in the effluents during the same period. However, the same hydraulic residence time was sufficient for complete nitrate transformation to dinitrogen gas (Phase IV) once the column had stabilized (in other words, after the column exhibited its full denitrification activity), which was achieved by increasing the hydraulic residence time to 48 h (Phase III). 3.2. Dinitrogen production The volume and composition of the gas produced by autotrophic denitrification were determined. During Phases I and II, the production of gas was insignificant, with increased dinitrogen production, indicating the autotrophic bacterial activity had not stabilized. During Phase III, the averaged production of gas was approximately 750 mL d1, with more than 95% dinitrogen gas in the total gas produced. It is known that full induction of the denitrification pathway (i.e., nitrate ? nitrite ? nitrous oxide ? dinitrogen) occurs with complete exhaustion of molecular oxygen. Otherwise, denitrifiers are able to reduce nitrate to nitrite, but avoid synthesis of the full array of denitrification enzymes unless anaerobiosis is achieved (Tate, 2000). Therefore, the conversion of nitrate to nitrous oxide is often considered the rate-limiting step for denitrification, where nitrate removal or nitrite production does not necessarily imply complete denitrification in the system. In this respect, the observation of more than 95% dinitrogen gas

H.S. Moon et al. / Chemosphere 73 (2008) 723–728

production in this study was meaningful. Less than 2% oxygen gas and carbon dioxide, 2–5 ppmv nitrous oxide and 10 ppmv hydrogen sulfide were also detected as trace gases (data not shown). The amount of dinitrogen in the total gas produced remained almost constantly during Phase IV. However, the gas volumes generated for the same period ranged from 1400 to 1800 mL d1. The volume of nitrogen gas generated by autotrophic denitrification based on following stoichiometric equation was calculated from the total gas volume measured and the gas composition.

55S þ 20CO2 þ 50NO3 þ 38H2 O þ 4NHþ4 þ ! 4C5 H7 O2 N þ 25N2 þ 55SO2 4 þ 64H

This observed gas volume was compared to the theoretical value obtained using the following equation, with the results presented in Table 1:

Sulfate (mg-S L-1) 110

0

30

60

Sulfate (mg-S L1)

90 120 150 180 210 110

a

100

Nitrate Nitrite Sulfate

90

Column height (cm)

726

0

30

60

90 120 150 180 210

b

100

Nitrate Nitrite Sulfate

90

80

80

70

70

60

60

50

50

40

40

30

30

20

20

10

10

1

Theoretical nitrogen gas volumeðmL d Þ 22:4 mL Temp ðKÞ  ¼ Removed nitrate conc ðmg  NL Þ  28 mg 273:15 ðKÞ 1

1

0

 Flow rate ðL d Þ

0

0

10

20

30

40

50

60

70

-1

During phase III, where the seepage velocity was 0.5  103 cm s1, the averaged dinitrogen gas volume observed was approximately 602 mL d1, while the theoretical dinitrogen gas volume was approximately 745 mL d1. This corresponded to about 81% dinitrogen gas production in the column. However, during Phase IV, where the seepage velocity was 1.0  103 cm s1, the observed and the theoretical volumes were 1050 and 1500 mL d1, respectively. Such a difference can be attributed to the entrapment of the gas produced inside the pores. This view was supported by the sharp increase in gas production immediately after the seepage velocity was increased. If this is accepted, the real production of dinitrogen gas during Phase III would be comparable to that during Phase IV. 3.3. Spatial ion distribution The spatial ion distribution throughout the column was determined during Phases III and IV of the operation (Fig. 3). When the seepage velocity was 0.5  103 cm s1, nitrate removal occurred rapidly, with almost complete removal within 30 cm from the bottom of the column (Fig. 3a). Although a concentration of less than 5 mg N L1 nitrite production occurred up to this point, it quickly disappeared within 50 cm from the bottom. A similar spatial ion distribution pattern was observed in a previous study performed with small columns (Moon et al., 2004). At a seepage velocity of 1.0  103 cm s1, the nitrate removal was slower, with transformation complete at 50 cm from the bottom (Fig. 3b). More nitrite accumulated and then disappeared within this area. The patterns of sulfate production in these two cases were a good match to the nitrate and nitrite transformation behavior.

Table 1 Average volumes of dinitrogen gas evolved during the column operation Operation stage (Seepage velocitya)

Phase III (0.5) Phase IV (1.0) a

Unit: 103 cm s1.

Average dinitrogen gas volume (mL d1)

Nitrate, Nitrite (mg-N L )

Theoretical

602 (±167) 1409 (±208)

745

81

1496

95

10

20

30

40

50

60

70

Nitrate, Nitrite (mg-N L-1)

Fig. 3. Spatial ion distribution along the column at different seepage velocities: (a) 0.5  103 cm s1 and (b) 1.0  103 cm s1.

3.4. Hydrodynamic characteristics Tracer tests were performed to investigate the change in the hydrodynamic characteristics in the test column before and after operation. The experimentally obtained breakthrough curves were compared to the predicted data by curve-fitting using a mathematical mass transport model consisting of advection, dispersion. The least square method was used for curve-fitting. (Moon et al., 2004). Before the column operation, a hydrodynamic dispersion coefficient of 4.5  103 cm2 s1 was measured in the column and seepage velocity was 1.07  103 cm s1. An effective porosity of 0.42 was also obtained (Table 2). After 500 d operation, the seepage velocity increased by 10% compared to the initial value, which in turn resulted in an increased hydrodynamic dispersion coefficient. The column Peclet number at the beginning was 23.8, which would indicate that advection was dominant during mass transport (Shackelfold, 1994). However, this was reduced to 12.7 after 500 d column operation, indicating that dispersion became to play a role in solutes transport as the hydrodynamic characteristics changed during the column operation. The effective porosity decreased to 0.38, probably due to the decreased sulfur granules size as a result of the autotrophic sulfuroxidizing bacterial activity and the dissolution of the limestone added as an alkalinity source. Also, the microbial growth, as well as the biotic and abiotic reactions, which caused precipitation, may also have slightly contributed to the decrease in the effective

Table 2 Hydrodynamic parameters before and after the column operations Parameters

Dinitrogen gas recovery (%)

Observed

0

Determined values Before operation

Hydrodynamic dispersion coefficient (Dh) (cm2 s1) Seepage velocity (v) (cm s1) Effective porosity (ne) a Column Peclet number ðvLD1 h Þ Dispersivity (a) a

L: column length.

After operation

4.5  103

9.4  103

1.07  103 0.42 23.8 4.21

1.19  103 0.38 12.7 7.90

H.S. Moon et al. / Chemosphere 73 (2008) 723–728

727

effective porosity and seepage velocity, slightly changed, but these changes did not result in a decline in the denitrification performance of the column. The information presented is to be widely consulted with in constructing permeable reactive barrier system for nitrate-contaminated groundwater treatment in the field. Especially, the spatial ions distribution data indicate an appropriate thickness of the reactive barrier. At least a barrier thickness of 50 cm seemed to be required to insure complete nitrate removal without the accumulation of nitrite under the seepage velocities tested. Acknowledgements The financial support for this research was provided by the Brain Korea 21 Project, with additional support from the KOSEF (Korea Science and Engineering Foundation) through the AEBRC (Advanced Environmental Biotechnology Research Center) at POSTECH. The authors thank the Research Institute of Engineering Science at Seoul National University for their technical assistance. Fig. 4. Comparative DGGE analysis of the bacterial consortium attached to the sulfur particles after 500 d column operation: Lane 1, Sample from 80 to 100 cm; Lane 2, Sample from 50 to 80 cm; Lane 3, Sample from 30 to 50 cm and Lane 4, Sample from 0 to 30 cm.

porosity (Kamolpornwijit et al., 2003). Examination of the precipitates on the reactive materials, using energy dispersive spectroscopy and X-ray diffraction, revealed calcium carbonate as being dominant, with calcium phosphate and calcium sulfate as trace minerals (data not shown). 3.5. Microbial community analysis After the column had been operated for over 500 d, the reactive materials were collected at different depths, with bulk genomic DNA extracted and subjected to PCR-DGGE analysis. As seen in Fig. 4, all samples from the four parts of the column showed bands corresponding to T. denitrificans-like bacterium, indicating the dominant presence of this bacterium throughout the column. The distribution and intensity of the bands from the samples taken at depths of 30–50 and 0–30 cm (lanes 3 and 4, Fig. 4) seemed to be almost identical, which can be explained by the spatial ion distribution data, which was consistent with patterns of nitrate and nitrite reduction in these areas. However, the DGGE fingerprint of the sample from the 80 to 100 cm region (lane 1, Fig. 4) was different from that of the other regions. At least three detectable bands disappeared in this region, with a marked intensity of the band for C. limicola-like bacterium. C. limicola is able to use sulfide and carbon dioxide as electron donor and acceptor, respectively, and produce oxidized species, such as sulfur and sulfate (Cork et al., 1985). A similar spatial distribution of predominant bacterial species in denitrification columns has also been reported (Koenig et al., 2005). 4. Conclusions A large-scale autotrophic denitrification column, using elemental sulfur as an electron donor, was successfully operated for over 500 d. Nitrate (60 mg N L1) was almost completely transformed to dinitrogen gas, accounting for more than 95% of theoretical evolution. In addition, the bacterial species attached to the sulfur particles seemed to dynamically adapt to the microenvironments, especially the electron acceptors and donors, along the column. During the operation, the hydrodynamic characteristics, such as

References Cork, A., Mathers, J., Maka, A., Srnak, A., 1985. Control of oxidative sulfur metabolism of Chlorobium limicola forma thiosulfatophilum. Appl. Environ. Microbiol. 49, 269–272. Farrell, J., Kason, M., Melitas, N., Li, T., 2000. Investigation of the long-term performance of zero-valent iron for reductive dechlorination of trichloroethylene. Environ. Sci. Technol. 34, 514–521. Fogler, H.S., 1999. Elements of Chemical Reaction Engineering. Prentice Hall PTR, Upper Saddle River, NJ. Gayle, B.P., Boardman, G.D., Sherrard, J.H., Benoit, R.E., 1989. Biological denitrification of water. J. Environ. Eng. – ASCE 115, 930–943. Gonzales-Pradas, E.M., Villafranca-Sanches, M., Socias-Viciana, M., 1993. Phosphate and nitrate sorption on calcareous soils from Spain. Arid Soil Res. Rehab. 7, 181– 190. Hiscock, K.M., Lloyd, J.W., Lerner, D.N., 1991. Review of natural and artificial denitrification of groundwater. Water Res. 25, 1099–1111. Hunter, W.J., 2003. Accumulation of nitrite in denitrifying barriers when phosphate is limiting. J. Contam. Hydrol. 66, 79–91. Kamolpornwijit, W., Liang, L., West, O.R., Moline, G.R., Sullivan, A.B., 2003. Preferential flow path development and its influence on long-term PRB performance: Column study. J. Contam. Hydrol. 66, 161–178. Koenig, A., Liu, L.H., 1996. Autotrophic denitrification of landfill leachate using elemental sulfur. Water Sci. Technol. 34 (5–6), 469–476. Koenig, A., Zhang, T., Liu, L.H., Fang, H.H.P., 2005. Microbial community and biochemistry process in autosulfurotrophic denitrifying biofilm. Chemosphere 58, 1041–1047. Kruithof, J.C., van Bennekom, C.A., Dierx, H.A., Hijnen, W.A.M., van Paassen, J.A.M., Schooners, J.C., 1988. Nitrate removal from groundwater by sulphur/limestone filtration. Water Supply 6, 207–217. Mackenzie, P.D., Horney, D.P., Sivavec, T.M., 1999. Mineral precipitation and porosity losses in granular iron columns. J. Hazard. Mater. 68, 1–17. Moon, H.S., Ahn, K.-H., Lee, S., Nam, K., Kim, J.Y., 2004. Use of autotrophic sulfuroxidizers to remove nitrate from bank filtrate in a permeable reactive barrier system. Environ. Pollut. 129, 499–507. Moon, H.S., Chang, S.W., Nam, K., Choe, J., Kim, J.Y., 2006a. Effect of reactive media composition and co-contaminants on sulfur-based autotrophic denitrification. Environ. Pollut. 144, 802–807. Moon, H.S., Nam, K., Kim, J.Y., 2006b. Initial alkalinity requirement and effect of alkalinity sources in sulfur-based autotrophic denitrification barrier system. J. Environ. Eng. – ASCE 132, 971–975. Phillips, D.H., Gu, B., Watson, D.B., Roh, Y., Liang, L., Lee, S.Y., 2000. Performance evaluation of a zero valent iron reactive barrier: mineralogical characteristics. Environ. Sci. Technol. 34, 4169–4176. Rocca, C.D., Belgiorno, V., Meric, S., 2007. Hetertrophic/autotrophic denitrfication (HAD) of drinking water: prospective use for permeable reactive barrier. Desalination 210, 194–204. Shackelfold, C.D., 1994. Critical concepts for column testing. J. Geotechnol. Eng. – ASCE 120, 1804–1828. Sierra-Alvarez, R., Beristain-Cardoso, R., Salazar, M., Gomez, J., Razo-Flores, E., Field, J.A., 2007. Chemolithotrophic denitrification with elemental sulfur for groundwater treatment. Water Res. 41, 1253–1262. Sikora, L.J., Keeney, D.R., 1976. Evaluation of sulfur-Thiobacillus denitrificans nitrate removal system. J. Environ. Qual. 5, 298–303. Soares, M.I.M., 2002. Denitrification of groundwater with elemental sulfur. Water Res. 36, 1392–1395. Sorel, D., Warner, S.D., Longino, B.L., Honniball, J.H., Hamilton, L.A., 2001. Dissolved hydrogen measurements at a permeable zero-valent iron reactive barrier. Nat. Meet. Am. Chem. Soc., Div. Environ. Chem.. San Diego, CA. pp. 1175–1180.

728

H.S. Moon et al. / Chemosphere 73 (2008) 723–728

Tate III, R.L., 2000. Soil Microbiology. John Wiley & Sons, New York NY. Vikesland, P.J., Klausen, J., Zimmermann, H., Ball, W.P., Roberts, A.L., 2000. Contaminant reduction by Fe(0): Evaluation of transport properties and chemical reactivity over time. Nat. Meet. Am. Chem. Soc., Div. Environ. Chem. Washington, DC, pp. 674–678.

Vikesland, P.J., Klausen, J., Zimmermann, H., Roberts, A.L., Ball, W.P., 2003. Longevity of granular iron in groundwater treatment processes: Changes in solute transport properties over time. J. Contam. Hydrol. 64, 3–33.