Selective removal of nitrate ion using a novel composite carbon electrode in capacitive deionization

w a t e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 0 3 3 e6 0 3 9

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/watres

Selective removal of nitrate ion using a novel composite carbon electrode in capacitive deionization Yu-Jin Kim, Jae-Hwan Choi* Department of Chemical Engineering, Kongju National University, 34 Gongupdae-gil, Seobuk-gu, Cheonan-si, Chungnam 331-717, South Korea

article info

abstract

Article history:

We fabricated nitrate-selective composite carbon electrodes (NSCCEs) for use in capacitive

Received 11 April 2012

deionization to remove nitrate ions selectively from a solution containing a mixture of

Received in revised form

anions. The NSCCE was fabricated by coating the surface of a carbon electrode with the

6 July 2012

anion exchange resin, BHP55, after grinding the resin into fine powder. BHP55 is known to

Accepted 17 August 2012

be selective for nitrate ions. We performed desalination experiments on a solution con-

Available online 31 August 2012

taining 5.0 mM NaCl and 2.0 mM NaNO3 using the NSCCE system constructed with the fabricated electrode. The selective removal of nitrate in the NSCCE system was compared

Keywords:

to a membrane capacitive deionization (MCDI) system constructed with ion exchange

Capacitive deionization

membranes and carbon electrodes. The total quantity of chloride and nitrate ions adsorbed

Nitrate-selective composite carbon

onto the unit area of the electrode in the MCDI system was 25 mmol/m2 at a cell potential

electrode

of 1.0 V. The adsorption of nitrate ions was 8.3 mmol/m2, accounting for 33% of the total. In

Anion-exchange resin

contrast, the total anion adsorption in the NSCCE system was 34 mmol/m2, 36% greater

Selectivity

than the total anion adsorption of the MCDI system. The adsorption of nitrate ions was

Desalination

19 mmol/m2, 2.3-times greater than the adsorption in the MCDI system. These results showed that the ions were initially adsorbed by an electrostatic force, and the ion exchange reactions then occurred between the resin powder in the coated layer and the solution containing mixed anions. ª 2012 Elsevier Ltd. All rights reserved.

1.

Introduction

The concentration of nitrate ions is gradually increasing in groundwater and surface water because of contamination of soil and rivers through the excessive use of nitrogen fertilizers and various organic pollutants. Nitrate ions in drinking water are known to cause methemoglobinemia in infants (Fewtrell, 2004). The WHO therefore advises that nitrate-nitrogen concentrations in drinking water should be maintained under 10 mg/L (WHO, 1996). Currently known nitrate removal methods include ion exchange (Bae et al., 2002), biological treatment (Shrimali and

Singh, 2001), reverse osmosis (RO) (Schoeman and Steyn, 2003), and electrodialysis (ED) (Midaoui et al., 2002; Van der Bruggen et al., 2004). Biological treatment is relatively inexpensive but requires large bioreactors. Additionally, the removal efficiency of bioreactors during the winter is reduced because of decreased microbial activity at low water temperatures (Koparal and Ogutveren, 2002). Ion exchange has the advantage of removing nitrate ions easily. However, the disadvantage of ion exchange is that a large amount of secondary pollutants is discharged during the process of regenerating the used resins (Bae et al., 2002). Reverse osmosis and electrodialysis can remove nitrate ions effectively.

* Corresponding author. Tel.: þ82 41 521 9362; fax: þ82 41 554 2640. E-mail address: [email protected] (J.-H. Choi). 0043-1354/$ e see front matter ª 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.watres.2012.08.031

6034

w a t e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 0 3 3 e6 0 3 9

However, reverse osmosis and electrodialysis require regular membrane replacement and high pressure and voltage during operation. Reverse osmosis and electrodialysis are less economical in the case of brackish water, which contains 2000e10,000 ppm of salt (Koparal and Ogutveren, 2002). Recently, the capacitive deionization (CDI) technique has attracted a great deal of attention. A characteristic of CDI technology that differentiates it from existing desalination techniques such as ED and RO is that CDI significantly reduces energy costs. In addition, CDI is recognized as an environmental friendly desalination process because it does not produce pollutants during operation (Gregory and Lovley, 2005; Zou et al., 2008; Xu et al., 2008; Leonard et al., 2009; Biesheuvel et al., 2009; Anderson et al., 2010; Tsouris et al., 2011; Huang et al., 2012). A CDI process removes ions by adsorbing the oppositely charged ions onto a porous carbon electrode when an electrical potential is applied to the electrode. Increasing the adsorption capacity of the carbon electrode to increase its desalination efficiency is very important. For this purpose, carbon electrodes for CDI have been developed using numerous carbon materials (Zou et al., 2008; Xu et al., 2008; Leonard et al., 2009; Tsouris et al., 2011; Cohen et al., 2011; Jia and Zou, 2012; Nie et al., 2012). In the early 2000s, Andelman (2002) was able to improve the desalination efficiency of the CDI process dramatically by using membrane capacitive deionization (MCDI), in which an ion exchange membrane is combined with carbon electrodes. Additionally, Kim and Choi (2010) have fabricated a new carbon electrode that performs similarly by coating carbon electrodes with ion exchange polymers. Recently, Porada et al. (2012) have developed a novel capacitive wire-based desalination technology. The ion exchange electrodes developed thus far adsorb either cations or anions; i.e., anion exchange electrodes adsorb all types of anions, regardless of the types of anions present, and prevent cations from being adsorbed (Biesheuvel et al., 2011; Lee et al., 2011; Kwak et al., 2012; Biesheuvel and Van der Wal, 2010). However, it is sometimes not necessary to remove all the ions in the feed solution. The most desirable electrode in a CDI application is one that can selectively adsorb specific anions from a solution containing several anions (Noked et al., 2009). When specific ion-selective carbon electrodes are used, operating costs can be reduced significantly because only the specific target ions are removed. Additionally, CDI has the advantage of increasing throughput per unit size of the carbon electrodes. Nevertheless, a study on the preparation of specific ion-selective electrodes for CDI applications has not yet been reported. In this study, we fabricated a novel nitrate-selective composite carbon electrode (NSCCE) to increase the selective adsorption of nitrate from a solution of mixed anions. To make an ion-selective composite carbon electrode, a method that involves coating the surface of carbon electrodes with a material that can provide high selectivity for a specific ion can be considered. The NSCCE was fabricated by coating a carbon electrode with a finely ground anion exchange resin that is known to exhibit high selectivity for the nitrate ion. We performed a desalination experiment using the unit cell made with the prepared NSCCE with a solution of mixed nitrate and chloride ions. To verify the performance of the NSCCE, the

desalination results were compared with those of the MCDI system that was constructed with a carbon electrode and an ion-exchange membrane.

2.

Materials and methods

2.1.

Preparation of a carbon electrode

To coat nitrate-selective resin powder, a carbon electrode was fabricated by mixing activated carbon powder with the polymer. Polyvinylidene fluoride (PVdF, M.W. ¼ 530,000, Aldrich) was dissolved in the organic solvent dimethylacetamide (DMAc, Aldrich), then mixed with activated carbon powder (CEP-21K, PCT Co., Korea, surface area ¼ 1320 m2/g). A uniform carbon slurry was produced by stirring the mixture for 12 h using a magnetic stirrer. The slurry was then cast onto the current collector graphite sheet (F02511, Dongbang Carbon) using a doctor blade. The slurry was then dried in a drying oven at 50  C for 6 h. To remove the remaining DMAc, the carbon electrode was further dried in a vacuum oven at 50  C for 2 h. The content of the polymer binder (PVdF) in the carbon electrode was 10 wt%. After drying, the thickness of the carbon layer on the graphite sheet was approximately 150 mm.

2.2. Fabrication of the nitrate-selective composite carbon electrode To make the NSCCE, anion exchange resin (Bonlite Co., BHP55, ion exchange capacity ¼ 1.05 meq/mL) was used. The BHP55 resin is expected to demonstrate selective adsorption because the resin is known to be nitrate-selective. The purchased resin was dried in a drying oven at 80  C for 24 h and pulverized using a grinder. The resin powder was sieved through a 90 mm mesh and used for coating. To coat the powdered resin onto the surface of the carbon electrode, an anion exchange polymer was obtained from Sion Tech Co., Korea and used as a binder. The polymer was prepared by introducing amine groups in the form of a tertiary amine after the chloromethylation of polystyrene. The ion exchange capacity of the anion exchange polymer was measured to be 1.20 meq/g. The coating solution was produced by mixing the anion exchange polymer dissolved in DMAc with the BHP55 resin powder. Through the preliminary experiments, we determined the optimum amount of anion exchange polymer needed to bind the resin powders tightly. The weight ratio of the binder to the resin powder was 1:1. The previously fabricated carbon electrode was coated with the uniformly mixed coating solution using a doctor blade. The fabrication of the NSCCE was completed by drying the coated electrode in a drying oven at 50  C for 12 h. The surface structure of the carbon electrode was examined using a scanning electron microscope (MIRA LMH, TESCAN Ltd.).

2.3.

Capacitive deionization experiment

To confirm that the fabricated composite carbon electrode is nitrate-selective, a desalination experiment was performed on two types of capacitive deionization unit systems. The “MCDI system” consisted of ion exchange membranes and

w a t e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 0 3 3 e6 0 3 9

carbon electrodes, and the “NSCCE system” included the NSCCE that had been fabricated. As shown in Fig. 1, the MCDI system consisted of a cation exchange membrane (Neosepta CMX, Astom Co., Japan), a spacer, and an anion exchange membrane (Neosepta AMX, Astom Co., Japan), arranged in order between two carbon electrodes and fixed by bolting Plexiglass plates onto both ends. The NSCCE system was constructed by replacing the anion exchange membrane and carbon electrode (þ) in the MCDI system with the NSCCE. All of the electrodes were cut into 10  10 cm2 sections before using in the cell construction. The apparent area of each electrode was 100 cm2. The influent was supplied via a peristaltic pump through the rubber gasket placed on the outside of the carbon electrode. To form a flow path, a hole (1 cm in diameter) was created in the center of the cation exchange membrane and the carbon electrode (). The influent was designed to enter from the direction of the electrode border, pass the spacer, and then flow out through the hole. The influent entered the cells at a constant rate (50 mL/min), and the effluent was sent to the feed tank to be circulated. The initial feed solution was prepared by mixing NaCl and NaNO3 so that the concentrations of chloride and nitrate were 5.0 and 2.0 mM, respectively. We used 300 mL of mixed solution for the desalination experiment, and the solution was stirred with magnetic stirrer during the experiment. The electrode assembly was powered using a potentiostat (WPG100, WonA Tech). Pre-treatment of the CDI system was conducted by running a few cycles of adsorption and desorption. After confirming that the system reached the dynamic steady state when the conductivity change of each cycle was the same as the previous cycle, we carried out actual experiments. An adsorption experiment was performed by applying a potential of 1.0 V to the cells for 15 min. The adsorbed ions were subsequently desorbed by setting the potential to 0.0 V for 5 min. These adsorption and desorption processes were successively repeated three times. During the adsorption and desorption processes, the electrical conductivity of the feed tank was measured. A data

6035

logger was connected to a conductivity meter, and the conductivity was measured every 3 s using a computer. At the end of the adsorption and desorption process for the first and second cycles, samples were collected from the feed tank. In addition, samples were collected at certain time intervals during the third cycle. The quantities of ions adsorbed and desorbed were calculated by measuring the ion concentrations of the samples. The ion concentrations were analyzed using ion chromatography (Metrohm Compact IC). An SI-90 4E column from Shodex was used for ion chromatography. The eluent was a mixture of 1.8 mM Na2CO3 and 1.7 mM NaHCO3, and the flow rate was 1.0 mL/min.

3.

Results and discussions

3.1.

SEM analysis of the composite carbon electrode

SEM images of the top and side views of the fabricated composite carbon electrode are shown in Fig. 2. The coated resin powder particles occurred in a particle size distribution that ranged from several micrometers to 60 mm. Because the sizes of the resin powder particles were large compared with the sizes of the activated carbon particles (approximately 4 mm), the composite carbon electrode appeared to have a relatively rough surface. However, the resin was combined well with the binder and formed a uniform coating layer on the surface of the carbon electrode. The thickness of the coating layer was approximately 70 mm. When a potential is applied to the carbon electrode, cations and anions are adsorbed onto the cathode and anode, respectively, by electrostatic attraction. However, because of the ion exchange resin coating on the surface of the composite carbon electrode, the ions are adsorbed onto the resin before migrating to the electrode surface. Therefore, if the ion exchange resin layer exhibits selectivity for nitrate ions, more nitrate ions are expected to be adsorbed onto the resin. In the CDI process, adsorption is induced by supplying a solution between the carbon electrodes. Therefore, a large

Fig. 1 e Schematic diagram of the MCDI system and the capacitive deionization experiment.

6036

w a t e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 0 3 3 e6 0 3 9

1.5

800 MCDI (solid) NSCCE (dotted)

0.5 600

MCDI (solid) NSCCE (dotted) 0.0

current (A)

conductivity ( S/cm)

1.0 700

500 -0.5

1st cycle

2nd cycle

3rd cycle

400

-1.0 0

10

20

30

40

50

60

time (min)

Fig. 3 e Conductivity and current changes during the operation of the MCDI and NSCCE systems.

Fig. 2 e SEM image of a composite carbon electrode coated with BHP55 resin powder, (a) top view, (b) side view.

surface area for the electrode is important to achieve adsorption equilibrium between the ions in the solution and the electrode surface. As shown in Fig. 2, the fabricated composite carbon electrode exhibits a large surface area that allows ions in the solution to contact the resin. However, the size distribution of the resin powder had a wide range because the resin powder was sieved through a 90-mm mesh. The particle size of the resin powder is an important factor to enhance selective adsorption. The small and uniform resin powder is expected to improve the selective adsorption of a specific ion. Thus, further research on the size and size distribution of the resin powder will be carried out.

3.2. Desalination performance in the capacitive deionization cell Adsorption (1.0 V for 15 min) and desorption (0.0 V for 5 min) processes in the MCDI and NSCCE systems were repeated three times, and the changes in the electrical conductivity of the solution were measured. The results are shown in Fig. 3. The conductivity changes show reproducible results for the three cycles of adsorption and desorption. The conductivity of the solution began to decrease drastically immediately after

the adsorption potential was applied. After 5 min, adsorption was completed, and the conductivity remained at a constant value. After 15 min, the cell potential was set to 0.0 V, and the conductivity rapidly returned to the initial value because of desorption of the adsorbed ions. Adsorption and desorption occurred rapidly and reproducibly in both types of systems, as indicated by the changes in conductivity. The current showed changes similar to those of the conductivity. The current decreased rapidly and then converged to zero after 5 min. The conductivity in the experiment using the NSCCE electrode decreased more than the conductivity of the MCDI system when the cell potential was applied. The conductivity of the MCDI system decreased to 670 mS/cm, whereas the conductivity of the NSCCE system decreased to 640 mS/cm at the completion of adsorption. The conductivity measurements showed that the NSCCE electrode adsorbed more ions than the MCDI system when an identical potential was applied.

3.3. Adsorption of nitrate and chloride ions in the MCDIsystem After the adsorption and desorption processes were completed during the desalination experiment, the ion concentrations in the solution were analyzed. The amounts of adsorbed and desorbed ions calculated from the ion concentration in each cycle are listed in Table 1. The amounts of adsorbed nitrate and chloride were approximately the same

Table 1 e Total amount of anions adsorbed and desorbed during each cycle for MCDI system operation. Cycle

1st 2nd 3rd Average

Amount of desorbed ions (mmol/m2)

Amount of adsorbed ions (mmol/m2) NO 3

Cl

% of NO 3

NO 3

Cl

% of NO 3

8.7 8.3 7.8 8.3

16 16 17 16

35 34 31 33

8.9 7.8 9.0 8.6

17 17 18 17

34 31 33 33

6037

w a t e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 0 3 3 e6 0 3 9

3.4. Adsorption of nitrate and chloride ions in the NSCCE system Desalination experiments were performed on the cell constructed with an electrode coated with the BHP55 resin under the same conditions used for the MCDI system. The adsorption and desorption of ions were measured from the ion concentrations during the adsorption and desorption process, which was repeated three times. The results shown in Table 2

1.05

30 2

amount of anions adsorbed (mmol/m )

during each cycle. The average total amount of the adsorbed ions from three adsorption experiments was 25 mmol/m2, and the ratio of nitrate ions to the total adsorbed ions was approximately 33%. In contrast, the total amount of the desorbed ions was 26 mmol/m2, indicating that all of the adsorbed ions were desorbed. During the third cycle of the adsorption process, the adsorption trend for each ion was determined by analyzing the concentrations of nitrate and chloride ions over time. The concentrations of nitrate and chloride ions as a function of time are presented in Fig. 4. The adsorption of nitrate and chloride began while the cell potential was being applied, and the concentrations of the ions decreased. The concentrations stabilized after 5 min as the adsorption capacity of the electrodes was saturated. Similar to the results of the conductivity measurements, the concentration results indicate that adsorption was complete approximately 5 min after the cell potential was applied. Ion adsorption as a function of time is shown in Fig. 5. The amounts of adsorbed nitrate and chloride were 7.8 mmol/m2 and 17 mmol/m2, respectively, with a total of 25 mmol/m2 at 15 min. Nitrate ions accounted for 31% of the total ions. Initially, nitrate and chloride were prepared at concentrations of 2.0 and 5.0 mM, respectively, and nitrate accounted for 29% of the total ions. Chloride ions are known to migrate better than nitrate ions under an applied electric field because the limiting molar conductivities of nitrate and chloride ions are 71.5 and 76.4 cm2 U1 eq1, respectively (Strathmann, 2004). Considering the ratio of the ions in the initial solution and their molar conductivity, the adsorbed nitrate was expected to be 27% of the total.

25

20

15

Nitrate Chloride total

10

5

0 0

2

4

6

8

10

12

14

16

adsorption time (min)

Fig. 5 e The amount of ions adsorbed as a function of adsorption time during MCDI system operation.

indicate that adsorption and desorption were consistent throughout the cycles. The total quantity of adsorbed ions was 34 mmol/m2. Of this total, the amount of nitrate ions was 19 mmol/m2 (56% of the total). Compared with the results obtained using the MCDI system, the total adsorption increased by approximately 36%. More importantly, the amount of adsorbed nitrate ions increased 2.3 times, from 8.3 mmol/m2 to 19 mmol/m2. These results show that the fabricated NSCCE is highly effective for the selective adsorption of nitrate ions. Fig. 6 shows the changes in the ion concentrations during the application of the cell potential (1.0 V). Compared with the results obtained using the MCDI system (Fig. 4), a difference was observed in the concentration profiles. In the case of the MCDI system, the concentrations did not change after 5 min because adsorption was complete. However, the nitrate concentration decreases continuously during 15 min, while the chloride concentration decreases during the first 5 min and then increases continuously. Based on the concentration changes shown in Fig. 6, the mechanism of ion adsorption in the NSCCE system can be examined. When a potential is applied to the electrode, the ions adsorbed onto the resin powder migrate to the carbon electrode via electrostatic attraction. Additional ions would then be adsorbed onto the ion exchange group of the resin

1.00 Nitrate Chloride

C/Co

0.95

Table 2 e Total amount of anions adsorbed and desorbed during each cycle for NSCCE system operation.

0.90

Cycle 0.85

0.80 0

2

4

6

8

10

12

adsorption time (min)

Fig. 4 e Changes in ion concentrations during the adsorption period in MCDI system operation.

14

16

1st 2nd 3rd Average

Amount of desorbed ions (mmol/m2)

Amount of adsorbed ions (mmol/m2) NO 3

Cl

% of NO 3

NO 3

Cl

% of NO 3

19 20 19 19

15 14 14 14

56 59 58 58

19 19 20 19

15 15 14 15

56 56 59 56

6038

w a t e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 0 3 3 e6 0 3 9

1.1

mole fraction of nitrate adsorbed (%)

70

nitrate chloride

1.0

C/Co

0.9

0.8

0.7

0.6

0.5

MCDI system NSCCE system

60

50

40

30

20

10

0 0

2

4

6

8

10

12

14

16

0

2

4

6

8

10

12

14

16

adsorption time (min)

adsorption time (min)

Fig. 6 e Changes in ion concentrations during the adsorption period in NSCCE system operation.

Fig. 8 e Mole fraction of nitrate adsorbed as a function of adsorption time for MCDI and NSCCE systems.

from the solution. Because chloride ions are more prevalent in the solution at this time, more chloride ions are expected to be adsorbed onto the resin. After 5 min of potential application, the carbon electrode becomes saturated, and additional adsorption by the electrode is not possible. From this point forward, ion exchange between the BHP55 resin in the coating layer and the solution is thought to occur to maintain the adsorption equilibrium. Because the chloride ions adsorbed onto the resin exceed the equilibrium amount, the chloride ions on the resin are thought to be exchanged with nitrate ions in the solution. To summarize the above adsorption mechanism: (1) after the adsorption potential is applied, adsorption is achieved by electrostatic attraction for the first 5 min; and (2) afterward, ion exchange occurs between the resin coating layer and the solution. This mechanism can explain the concentration changes shown in Fig. 5. Fig. 7 shows the quantity of ions adsorbed during the adsorption time. Five minutes after the cell potential was applied, the amount of adsorbed chloride was 17 mmol/m2. However, the amount of adsorbed chloride decreased to 14 mmol/m2 after 15 min. The amount of adsorbed nitrate ions increased from 15 mmol/m2 to 19 mmol/m2 after 15 min.

Although the amount of adsorbed chloride and nitrate ions changed after the 5 min mark, the sum of the two remained constant. This result supports the fact that adsorption is complete 5 min after the application of the cell potential, followed by ion exchange between the resin and the solution. The molar ratios of the adsorbed nitrate ions over time in the MCDI and NSCCE systems are presented in Fig. 8. In the case of the MCDI system, the adsorption ratio of nitrate did not change dramatically but did increase slightly over time. In contrast, the adsorption ratio of nitrate increased from the initial 30% to 58% after 15 min in the NSCCE system. The molar ratio of the adsorbed nitrate ions after 5 min of adsorption was 47%, suggesting that the BHP55 resin in the coating layer demonstrates selectivity for nitrate ions.

2

amount of anions adsorbed (mmol/m )

40 35 30 Nitrate Chloride Total

25 20 15 10 5 0 0

2

4

6

8

10

12

14

adsorption time (min)

Fig. 7 e The amount of ions adsorbed as a function of adsorption time during NSCCE system operation.

16

4.

Conclusion

We fabricated a nitrate-selective composite carbon electrode for use in capacitive deionization to selectively desalinate nitrate ions from a solution containing a mixture of anions. The ion exchange resin, BHP55, known to be selective for nitrate ions, was pulverized into fine powder, and the composite carbon electrode was fabricated by coating the resin powder onto the surface of a carbon electrode. To test the selective removal of nitrate, desalination experiments were performed on the MCDI and NSCCE systems on a solution that contained nitrate and chloride ions in concentrations of 2.0 and 5.0 mM, respectively. During the application of a cell potential of 1.0 V for 15 min, a total of 25 mmol/m2 of anions were adsorbed in the MCDI system. In addition, the adsorption of nitrate was 8.3 mmol/ m2, accounting for 33% of the total adsorption. A second desalination experiment was performed on the cell with the NSCCE under the same conditions. The results showed that the quantity of total adsorbed ions increased to 34 mmol/m2, representing a 36% increase compared with the results for the MCDI system. Additionally, the adsorption of nitrate ions was 19 mmol/m2, a 2.3-fold increase over the adsorption achieved using the MCDI system.

w a t e r r e s e a r c h 4 6 ( 2 0 1 2 ) 6 0 3 3 e6 0 3 9

In this study, we fabricated a novel NSCCE that can improve the CDI technology. We verified the enhancement of nitrate removal from a solution of mixed nitrate and chloride ions. In addition to chloride ions, however, water contains many other anions such as sulfate and fluoride ions. Studies of the effect of other anions on the selective removal of nitrate are needed in the future. In addition, a selective composite carbon electrode is not limited to nitrate ions, and similar methods in which a carbon electrode is coated with resins that are selective for other specific ions may be possible. If a composite carbon electrode is fabricated using a chelating resin, the resulting electrode is expected to be effective in the selective removal of calcium ions or heavy metal ions.

Acknowledgment This work was supported by a grant (223-111-002) from the Converging Technology Project of Korea Institute of Environmental Science and Technology (KIEST) funded by Korea Ministry of Environment.

references

Andelman, M.D., 2002. Charge barrier flow-through capacitor. Can. Patent CA 2444390. Anderson, M.A., Cudero, A.L., Palma, J., 2010. Capacitive deionization as an electrochemical means of saving energy and delivering clean water. Comparison to present desalination practices: will it compete? Electrochimica Acta 55, 3845e3856. Bae, B.U., Jung, Y.H., Han, W.W., Shin, H.S., 2002. Improved brine recycling during nitrate removal using ion exchange. Water Research 36, 3330e3340. Biesheuvel, P.M., Van der Wal, A., 2010. Membrane capacitive deionization. Journal of Membrane Science 346, 256e262. Biesheuvel, P.M., Van Limpt, B., Van der Wal, A., 2009. Dynamic adsorption/desorption process model for capacitive deionization. Journal of Physical and Chemistry C 113, 5636e5640. Biesheuvel, P.M., Zhao, R., Porada, S., Van der Wal, A., 2011. Theory of membrane capacitive deionization including the effect of the electrode pore space. Journal of Colloid and Interface Science 360, 239e248. Cohen, I., Avraham, E., Noked, M., Soffer, A., Aurbach, D., 2011. Enhanced charge efficiency in capacitive deionization achieved by surface-treated electrodes and by means of a third electrode. Journal of Physical Chemistry C 115, 19856e19863. Fewtrell, L., 2004. Drinking-water nitrate, methemoglobinemia, and global burden of disease: a discussion. Environmental Health Perspectives 112, 1371e1374. Gregory, K.B., Lovley, D.R., 2005. Remediation and recovery of uranium from contaminated subsurface environments with electrodes. Environmental Science and Technology 39, 8943e8947. Huang, Z.H., Wang, M., Wang, L., Kang, F., 2012. Relation between the charge efficiency of activated carbon fiber and its desalination performance. Langmuir 28, 5079e5084.

6039

Jia, B., Zou, L., 2012. Graphene nanosheets reduced by a multistep process as high-performance electrode material for capacitive deionization. Carbon 50, 2321e2345. Kim, Y.J., Choi, J.H., 2010. Improvement of desalination efficiency in capacitive deionization using a carbon electrode coated with an ion-exchange polymer. Water Research 44, 990e996. Koparal, A.S., Ogutveren, U.B., 2002. Removal of nitrate from water by electroreduction and electrocoagulation. Journal of Hazardous Materials 89, 83e94. Kwak, N.S., Koo, J.S., Hwang, T.S., Choi, E.M., 2012. Synthesis and electrical properties of NaSS-MAA-MIMA cation exchange membranes for membrane capacitive deionization (MCDI). Desalination 285, 138e146. Lee, J.Y., Seo, S.J., Yun, S.H., Moon, S.H., 2011. Preparation of ion exchanger layered electrodes for advanced membrane capacitive deionization (MCDI). Water Research 45, 5375e5380. Leonard, K.C., Genthe, J.R., Sanfilippo, J.L., Zeltner, W.A., Anderson, M.A., 2009. Synthesis and characterization of asymmetric electrochemical capacitive deionization materials using nanoporous silicon dioxide and magnesium doped aluminum oxide. Electrochimica Acta 54, 5286e5291. Midaoui, A.E., Elhannouni, F., Taky, M., Chay, L., Sahli, M.A.M., Echihabi, L., Hafsi, M., 2002. Optimization of nitrate removal operation from ground water by electrodialysis. Sep. Purif. Technol 29, 235e244. Nie, C., Pan, L., Liu, Y., Li, H., Chen, T., Lu, T., Sun, Z., 2012. Electrophoretic deposition of carbon nanotubes-polyacrylic acid composite film electrode for capacitive deionization. Electrocimica Acta 66, 106e109. Noked, M., Avraham, E., Bohadana, Y., Soffer, A., Aurbach, D., 2009. Development of anion steroselective, activated carbon molecular sieve electrodes prepared by chemical vapor deposition. Journal of Physical Chemistry C 113, 7316e7321. Porada, S., Sales, B.B., Hamelers, H.V.M., Biesheuvel, P.M., 2012. Water desalination with wires. Journal of Physical Chemistry Letters 3, 1613e1618. Schoeman, J.J., Steyn, A., 2003. Nitrate removal with reverse osmosis in a rural area in South Africa. Desalination 155, 15e26. Shrimali, M., Singh, K.P., 2001. New methods of nitrate removal from water. Environmental Pollution 112, 351e359. Strathmann, H., 2004. Ion-exchange Membrane Separation Processes. Elsevier, Amsterdam. Tsouris, C., Mayes, R., Kiggans, J., Yiacoumi, S., DePaoli, D., Dai, S., 2011. Mesoporous carbon for capacitive deionization of saline water. Environmental Science and Technology 45, 10243e10249. Van der Bruggen, B., Koninckx, A., Vandecasteele, C., 2004. Separation of monovalent and divalent ions from aqueous solution by electrodialysis and nanofiltration. Water Research 38, 1347e1353. WHO, 1996. Guidelines for Drinking-water Quality. In: Health Criteria and Other Supporting Information, second ed., vol 2. WHO, Geneva. Xu, P., Drewes, J.E., Heil, D., Wang, G., 2008. Treatment of brackish produced water using carbon aerogel-gased capacitive deionization technology. Water Research 42, 2605e2617. Zou, L., Li, L., Song, H., Morris, G., 2008. Using mesoporous carbon electrodes for brackish water desalination. Water. Research 42, 2340e2348.