Accepted Manuscript

Title: Capacitive desalination of ZnO/activated carbonasymmetric capacitor and mechanism analysis

Author: Jianyun Liu Miao Lu Jianmao Yang Jian ChengWenshu Cai

PII: S0013-4686(14)02200-2DOI: http://dx.doi.org/doi:10.1016/j.electacta.2014.11.023Reference: EA 23695

To appear in: Electrochimica Acta

Received date: 9-9-2014Revised date: 31-10-2014Accepted date: 5-11-2014

Please cite this article as: Jianyun Liu, Miao Lu, Jianmao Yang, Jian Cheng, Wenshu Cai,Capacitive desalination of ZnO/activated carbon asymmetric capacitor and mechanismanalysis, Electrochimica Acta http://dx.doi.org/10.1016/j.electacta.2014.11.023

This is a PDF file of an unedited manuscript that has been accepted for publication.As a service to our customers we are providing this early version of the manuscript.The manuscript will undergo copyediting, typesetting, and review of the resulting proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

1

Capacitive desalination of ZnO/activated carbon

asymmetric capacitor and mechanism analysis Jianyun Liu a,*, Miao Lua, Jianmao Yangb, Jian Cheng a, Wenshu Caia

a College of Environmental Science and Engineering, Donghua University, Shanghai 201620, China

b Research Center for Analysis & Measurement, Donghua University, Shanghai 201620, China

* Corresponding author. Tel.: 86-21-67792381; Fax: 86-21-67792522. E-mail:

jianyun.liu@dhu.edu.cn (J.-Y. Liu)

ABSTRACT

Zinc oxide/activated carbon composite electrode (ZnO/AC) was prepared by simply

mixing ZnO nanoparticles with AC granules in the presence of Teflon emulsion.

Scanning electron microscopy shows an even and seamless surface with effective filling

of ZnO nanoparticles in between AC granules. Cyclic voltammetry and impedance

analysis demonstrate the ideal double-layer capacitor behavior. The desalination behavior

of the asymmetric capacitor with ZnO/AC as positive electrode and AC as negative

electrode (+ZnO/AC||AC), or ZnO/AC as negative electrode and AC as positive electrode

(ZnO/AC||AC ) was studied, respectively. As compared with pure AC||AC capacitor,

ZnO/AC||AC capacitor showed a very stable desalination behavior with high

desalination amount of 9.4 mg/g and charge efficiency of 80.5%; while +ZnO/AC||AC

capacitor showed no obvious difference after several desalination cycles due to poor

ACCEPTED MANUSCRIP

T

2

stability. The mechanism was analyzed based on zeta potential of ZnO particles and pH

variation near the electrode surface during charging process. The different desalination

properties on positive and negative electrodes due to zeta potential variation of ZnO with

pH change at electrode surface were further confirmed by using other metal oxides like

CuO, MnO2 and WO3. This study provides a particularly important guidance for

screening electrode materials and optimizing operation parameters for capacitive

desalination.

Keywords: Zinc oxide, Asymmetrical capacitive desalination, surface pH of electrode,

zeta potential, polarization of electrode

1. Introduction

Capacitive deionization (CDI), as one of the most promising techniques to produce the

fresh water from saline water [1], has attracted much interest recently, because of its

merits such as low energy consumption, high water recovery rate, the renewability of the

electrode and easy operation with no secondary pollution due to low applied voltage [2,

3]. By developing porous carbon-based electrode materials such as carbon aerogel [4, 5],

carbon nanofiber [6-8], carbon nanotubes [9, 10], graphene [11, 12], the desalination

performance has been improved due to the large surface area, ideal pore distribution and

good conductivity. Meanwhile, activated carbon (AC), as the most commonly used

commercially available and cost efficient carbon material, is still attractive as the typical

ACCEPTED MANUSCRIP

T

3

CDI electrode material because of its high specific surface area, low cost, easy

manufacturing on a large scale [2, 13-15]. Especially, by mixing with some additives

such as conductive carbon black [16], ionic exchange particles [17, 18], functionalized

carbon nanotube [19] and graphene [20-22], the desalination performance of the

composite activated carbon electrode has been further improved due to enhanced

conductivity, ion transportation and charge efficiency.

Charge efficiency (CE), defined as the ratio of the equivalent charge of adsorbed salt

to the total charge input during charging process, is one of the most important parameters

in CDI cell. The charge efficiency can be increased by membrane-based CDI cell [23, 24]

or by chemical modification of electrode surface [21, 22] to diminish co-ions

migration/repulsion. The modification of carbon surface with some ionic selective groups

can produce an extra potential field in terms of altering the potential of zero charge (PZC)

on the electrode surface [25, 26], which drives the working electrode into a high charge

efficiency. Recently, metal oxides (MOs) have attracted much attention as additives of

electrode materials in electrochemical capacitive deionization due to low toxicity, high

stability, and capability to store more electronic charge. For example, TiO2 [15, 27-29],

ZnO [30-32], MnO2 [33, 34], Fe3O4 [35] and Al2O3[36, 37] have been used successfully

to improve the desalination performance. Usually, the good water-wettability of MOs and

the facile ion transportation are generally considered to be the main reason for improving

the desalination efficacy. Recently, Han [36] and Leonard [37] demonstrated the

improvement of charge efficiency on the Al2O3||SiO2 asymmetric capacitor due to the

ACCEPTED MANUSCRIP

T

4

inherent charge on the MOs surface.

It is well-known that the polarization of electrode occurs when applying a high

potential, which results in the carbon corrosion accompanying with pH variation [38].

The pH variation on positive and negative electrode may change the zeta potential of

MOs particle surface, thus influencing the PZC of electrode surface and charge efficiency

[27, 39]. The effect of polarization on the deionization performance and stability has not

been mentioned in the Al2O3||SiO2 asymmetric capacitor [36, 37]. On the symmetric

capacitor by assembling the two identical MOs-containing electrodes [30, 31, 33, 35], it

is difficult to differentiate their contribution in positive electrode and negative electrode,

respectively. Moreover, it has been reported that the asymmetric capacitor with MOs

cathode and AC anode could give an extended operating voltage [40]. Therefore, it is

necessary to look into the influence of transition metal oxide on desalination performance

by assembling asymmetric capacitor.

In this report, using ZnO as a candidate of metal oxides, we prepared zinc

oxide/activated carbon composite electrode by simply mixing ZnO nanoparticles with AC

granules in the presence of the Teflon emulsion. The asymmetric capacitor was

assembled to investigate the effect of ZnO in positive and negative electrode on

desalination performance, respectively. By monitoring surface pH of the electrode during

charging step, we clarified the polarization issue of the electrodes and its influence on the

desalination behavior of the asymmetric capacitor. Combining with the zeta potential

analysis of various metal oxides, we also revealed the rules about how to choose the

ACCEPTED MANUSCRIP

T

5

metal oxides additives as the positive or negative electrode material. Finally WO3, MnO2

and CuO mixed carbon composite electrodes were tested to confirm the relation of zeta

potential with desalination performance. This work will provide a very important

guidance to electrode material selection in CDI technology.

2. Materials and methods

2.1. Materials

Activated carbon (AC) was supplied by Yihuan Carbon Co., Ltd. ZnO (~200 nm),

MnO2 (~50 nm), CuO (~50 nm) and WO3 (~50 nm) were bought from (Suzhou Yougao

Nanomaterial Co., Ltd). Polytetrafluoroethylene (PTFE, 60 wt %) were obtained from Du

Pont Co., Ltd. All other chemicals were from Sinopharm Chemical Reagent Co., Ltd.

2.2. Fabrication of ZnO/AC composite electrode

The ZnO-AC composite electrode was prepared as follows: AC powder, ZnO

nanoparticles and PTFE binder were mixed at a ratio of 10:1:0.8. After agitating for 2 h,

the carbon slurry was dried in an oven at 100 ℃ for 12 h. Finally, moistened by a certain

amount of ethanol, the mixture was rolled into a flat sheet and pressed on the titanium

mesh (current collector) with the electrode size of 60 mm×30 mm. Other metal oxides-

containing AC electrode (MOs/AC) was prepared by the same method with MnO2, CuO

and WO3 particles instead of ZnO, and the electrodes were designated by MnO2/AC,

CuO/AC and WO3/AC, respectively. AC electrode was fabricated with the above similar

ACCEPTED MANUSCRIP

T

6

method but without metal oxides addition.

2.3. Assembly of asymmetric CDI cell and capacitive desalination experiments

The asymmetric CDI cell was assembled by using MO/AC electrode and AC counter

electrode sandwiched together, separated by a Nylon mesh which provides the flow

channel to prevent the short-circuit of the two electrodes. The desalination experiment

was carried out by a batch –mode with a 50 mL of 500 ppm NaCl aqueous solution,

which was continuously pumped through the unit cell by a peristaltic pump with a

volumetric flow rate of 10 mL/min.

The capacitive desalination test was conducted on a computer-controlled battery test

system (LANHE, CT2001A, Wuhan, China) using a constant current (16 mA) charging

and short circuit self-discharging mode. The charging terminal voltage was set at 1.2 V

unless otherwise stated. The concentration variation of NaCl solution was measured with

conductivity meter (Mettler Toledo, S230). Concentration was calculated based on the

calibration curve between conductivity and concentration made prior to the experiment.

The salt removal amount was represented by the average adsorbed NaCl mass in one

charging process divided by electrode mass (mg/g).

When the positive terminal of the test system was connected with ZnO/AC electrode,

the cell was denoted as +ZnO/AC||AC, and conversely the cell was denoted as

ZnO/AC||AC.

ACCEPTED MANUSCRIP

T

7

2.4. Characterization

The surface morphology of the electrodes was examined using scanning electron

microscopy (SEM, Hitachi, S-4700, Japan).

Cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS)

measurement were carried out in 1 M NaCl solution using a Princeton 2273

electrochemical workstation with a two-electrode testing cell similar to the desalination

cell. For EIS measurement, the frequency range was from 100 kHz to 10 mHz with 5 mV

amplitude voltage.

The release amount of ZnO during charging process was measured by atomic

absorption spectrometry (AAS). Samples were taken from the effluent of a continuous

desalination cell at regular intervals and filtrated by membranes (0.22 m). The

calculation curve was first obtained by using a series of zinc standard solution.

The zeta potential of MO particles in different pH solutions was analyzed by a zeta

meter (Zeta-Meter, PG314, USA). The pH of the electrode surface was measured by

precise pH test paper. For surface pH measurement of the electrode, the electrode was

immediately removed out from the cell after charging to a terminal voltage from 0.4 V to

1.5 V.

3. Results and discussion

3.1. Morphology of ZnO/AC composite electrodes

By simple rolling, the obtained ZnO/AC composite electrode keeps good flexibility

ACCEPTED MANUSCRIP

T

8

and high mechanical strength. Fig.1 (A) and (B) show the surface morphology of the

ZnO/AC and AC electrodes, respectively. There is loose contact between AC granules

with deep ditches on the AC electrode. With ZnO doping, the ZnO nanoparticles have a

good distribution near AC granules, and the surface of the obtained ZnO/AC electrode is

flat with ZnO particles well filling the gap between AC particles. This may facilitate the

ionic diffusion and mass transfer along the electrode in the electrochemical cell, due to

hydrophilicity and semi-conductivity properties of ZnO particles.

3.2. Electrochemical characterization

Cyclic voltammetry (CV) measurements are widely used to evaluate the

electrosorption performance and determine the specific capacitance of the fabricated

electrode. Fig. 2(A) displays the CV curves of AC and ZnO/AC composite electrodes in

the presence of 1 M NaCl aqueous solution at a scan rate of 2 mV s-1. These curves

exhibit ideal rectangular shape, indicating a typical electric double layer capacitor

behavior. Compared to that of AC electrode, CV curve of ZnO/AC composite electrode

has a large integrated CV area and exhibits a very rapid current response on voltage

reversal at each end potential and possesses high reversibility, indicating an excellent

capacitive behavior of ZnO/AC composite electrode. The specific capacitances of AC and

ZnO/AC are calculated from CV curves to be 49.4 F/g and 66 F/g, respectively. The

remarkable enhancement of the specific capacitance is mainly due to the improvement of

conductivity and wettability of ZnO/AC composite electrode. The mechanism for the

10 m

ACCEPTED MANUSCRIP

T

9

good charge storage property on metal oxides is also interpreted as the hydroxylation of

ZnO surface [41]. Therefore, the diffusion and adsorption of ions become facile on the

composite electrode. The possible mechanisms of ions adsorption are as follows:

ZnO (H2O)n = ZnO(OH¯)n+n H+ (1)

ZnO (OH¯)n + nNa+ = ZnO(OH) n Nan (2)

The EIS experiment was carried out in 1 M NaCl solution to examine the dynamic

behavior of ions transferring from the electrolyte to electrode. Fig. 2(B) presents the

corresponding Nyquist plots of ZnO/AC and AC electrode. They exhibit a frequency-

independent slope near to 45°at the middle frequency range, suggesting a porous

electrode behavior [42]. From the intercept of the plots on real axis at high frequency, the

ZnO/AC possesses a low equivalent series resistance (esr) of 0.1 , while AC electrode

possesses a higher esr (0.3 ), suggesting ZnO/AC electrode has a slightly lower internal

resistance since the same electrolyte solution was used in the cell in both cases. At lower

frequency of ZnO/AC, the straight line, which represents the diffusion resistance

(Warburg impedance), has larger slope than 45, with the line inclining steeply to the

imaginary axis, proving that ion diffusion and migration on the surface of the composite

electrode are faster, implying an excellent electrochemical double-layer capacitance

behavior. However, the spectrum of AC electrode at lower frequency region is far away

from imaginary axis, indicating a slow ion diffusion velocity. It concludes that the

addition of ZnO facilitates the ion transportation and diffusion inside the electrode, being

consistent with the CV test.

ACCEPTED MANUSCRIP

T

10

3.3. Desalination performance of ZnO/AC-based asymmetric capacitor

With AC as counter electrode, two asymmetric cells were assembled with ZnO/AC

composite as positive/negative electrode (denoted as +ZnO/AC||AC and ZnO/AC||AC,

respectively). The cell was cycled with 500 ppm NaCl aqueous solution overnight to

arrive at the physical adsorption balance. The cycling stability of the electrode was

investigated by recycling the CDI cell with charging and discharging alternate mode. The

multi-cycle voltage and current curves (A, C), and the corresponding conductivity

variation (B, D) of ZnO/AC||AC and +ZnO/AC||AC capacitor were shown in Fig.3.

ZnO/AC||AC capacitor shows a steady conductivity variation (panel B) during charge-

discharge process, indicating an excellent stability and regeneration ability of electrode

materials for deionization; while there is a gradual ascending on conductivity with

continuous charge-discharge cycling on +ZnO/AC||AC capacitor, indicating an

decreasing desalination behavior. After 12 charge-discharge cycles, the obtained variation

curves of conductivity and voltage with time on ZnO/AC||AC, +ZnO/AC||AC and

AC||AC capacitor were collected in Fig.4. The AC||AC capacitor shows a slow

adsorption-desorption process with the desalination amount of 4.5 mg/g per cycle. The

ZnO/AC||AC and +ZnO/AC||AC obviously demonstrate a fast desalination rate. The

enhancement of adsorption on ZnO/AC||AC was observed with the desalination amount

of 9.4 mg/g. While, the desalination amount on +ZnO/AC||AC was only 4.8 mg/g, similar

to AC||AC capacitor. It turns out that ZnO additive is effective to enhance the desalination

amount in negative electrode with good stability, but ZnO in positive electrode loses its

ACCEPTED MANUSCRIP

T

11

effectiveness with several cycles due to its instability.

3.4. Mechanism analysis

In order to understand the difference of desalination performance of ZnO on positive

and negative electrode, electrode stability and water splitting due to polarization of

electrode were investigated in detail according to the chemical property of ZnO.

3.4.1 ZnO release amount test

In the charge-discharge process, the samples were taken from the circulating solution to

measure the concentration of zinc ions by atom adsorption spectroscopy in order to

monitor the release of ZnO out of electrode. Fig.5 presents the concentration of zinc ions

in 500 mL NaCl solution at different desalination periods. The release amount of Zn from

negative electrode is neglectable with the concentration of zinc ions in circulation

solution is pretty low in seven days circulating, indicating that ZnO is stable when used

as negative additive. However, the release amount of zinc ions from positive electrode

increased gradually with time, indicating the dissolution of ZnO out of the electrode. It

confirms that the continuous release of zinc ions results in the reduced desalination

performance as described previously. Dutta group [30] ever observed the poor

regeneration of the symmetric capacitor assembled with two ZnO-nanowire modified

carbon fibers during desalination. We speculate that the cause is similarly from the

instability of ZnO positive electrode.

ACCEPTED MANUSCRIP

T

12

3.4.2 Surface pH changes of the electrodes during charging process

As we know, ZnO is dissoluble in low pH solution. Release of ZnO may result from

pH change due to water splitting during charge process. However, there is only a little pH

variation found (from pH 6.8 to 6.3) in the circulating water in continuous desalination

cycling. Cohen [38] measured the solution pH value by controlling the potential of a

single electrode to demonstrate the decomposition of water. The AC particles possess a

rich porous structure, and there is an overlapped diffusion layer between inner particles,

which may result in the aggravation of polarization at the particles interface inside the

electrode. Therefore, our experiment was designed to measure the surface pH value of the

electrode to explore the effect of electrode polarization. Under a constant-current

charging mode, when the voltage arrived at a certain value, charging was stopped, and

then the surface pH value was measured immediately by precise pH test paper. Fig.6

presents the surface pH variation of AC and ZnO/AC electrodes with terminal voltage. It

can be seen clearly that with the increase of terminal voltage and thus the quantity of

electric charge, pHs on the positive and negative electrode surface become more acid and

basic, respectively, independent of electrode materials. It confirms that the polarization of

electrodes becomes more and more severe with increasing the quantity of electric charge.

At the terminal voltage of 1.2 V, pH value on the positive electrode close to 3.5. At this

pH value, ZnO particles in the electrode obviously tend to dissolve. These can well

explain the cause of the deterioration of the desalination performance on +ZnO/AC||AC

ACCEPTED MANUSCRIP

T

13

capacitor. It was reported that MnO2 could extend the potential window to inhibit water

electrolysis in MnO2/graphite electrode with MnO2 as dominant material (70%) [40, 43].

However, the polarization is similar on both AC and ZnO/AC electrodes, probably due to

the existence of the porous AC (mass ratio,85%) in the ZnO/AC composite electrodes,

which dominates the properties of the electrode.

It is difficult to explain the cause of the pH change at low terminal voltage due to the

complexity of the surface of the electrode. In the AC-dominated MOs/AC composite

electrode material, owing to existence of the unique porous structure of AC, the pH

changes from the faradic charge involved in the decomposition of H2O could not be

neglectable. Therefore, apart from the absorption-desorption of the salt ions, the

additional redox reaction such as water splitting may occur on carbon surface [44],

Negative electrode: C+H2O + e¯ = C-H2(1/2)+ OH¯ (3)

Positive electrode: C+H2O - 2e¯ = C-O2 (1/2)+ 2H+ (4)

At negative electrode, OH¯ was accumulated near the surface of porous carbon,

companying with the adsorption of hydrogen. At positive electrode, H+ was accumulated

with the adsorption of oxygen and formation of oxygen-containing groups. Therefore,

water splitting results in pH change at the electrode surface when charging up to a certain

voltage.

Meanwhile, with AC||AC capacitor, the effect of salt concentration in the circulating

solution on the pH of electrode surface was studied. The cells filled with NaCl solution at

different concentrations were applied by the same coulomb charge in order to eliminate

ACCEPTED MANUSCRIP

T

14

the effect of the IR drop. Thus, there are different terminal voltages of charging for

different NaCl concentration. Fig. 7 gives the pH changes on the electrode surface after

charging 8 mAh at various concentrations. The pH variation trends on electrode surface

are similar in the concentration range of 0.5 ~ 35 g L-1 NaCl solution. In 200 g L-1 NaCl

solution, the pH variation is unobvious due to very low terminal voltage (0.56 V).

Nevertheless, the surface pH at positive and negative electrode still became acidic (pH 4)

and basic (pH 10.5), respectively, when applying higher coulomb charge input of 10 mAh

(terminal voltage, 0.80 V). It indicates that the positive and negative electrode will

definitely become acid and basic, respectively, if only a certain charge quantity was

imposed on the porous electrode. Therefore, the polarization processes is unavoidable

during the charging process, even if the cell voltage is lower than the theoretical

electrolysis voltage of 1.2 V. This conclusion is very important for practical desalination

application, such as the selection of electrode material and evaluation of scaling at

electrode.

3.4.3 The zeta potential of metal oxides and the desalination performance in the

asymmetric capacitor

As described previously, ZnO/AC electrode enhanced the desalination performance. The

hydrophilicity of ZnO and its polarization are deemed to be the main reason for

increasing ion transportation rate and desalination efficacy [32]. Actually, the occurrence

of the polarization is not necessarily related to materials (vide supra). On the other hand,

ACCEPTED MANUSCRIP

T

15

the zeta potential of ZnO may contribute to the electric field, resulting in fast ion

adsorption, in the light of pH variation on electrode surface upon charging. Fig.8 (a)

shows the zeta potential of ZnO nanoparticles dispersed in dilute NaCl aqueous solution

at different pHs. When pH value is lower than 5, the movement of ZnO particles could

not be observed due to the dissolution of ZnO. The isoelectric point of ZnO is close to 8.

It indicates that ZnO surface is positively charged at lower pH, and negatively charged at

higher pH value. The further the pH deviates from the isoelectric point, the larger the

absolute value of zeta potential becomes. Therefore, with the pH change on positive and

negative electrode during charging process, the charges of ZnO particles in the electrode

form an additional electric field, promoting the migration of ions from solution to the

electrode. It is very reasonable to explain the desalination properties on positive and

negative electrodes (Fig. 3).

As is known, the migration of co-ions in CDI cell consumes charging energy, resulting

in poor charge efficiency [32]. In the presence of ZnO nanoparticles at negative electrode,

the negative charges of the ZnO surface diminish the migration of counter ions.

Therefore, the high charge efficiency on ZnO/AC||AC (CE=80.5%) was obtained

compared with AC||AC (40%), as shown in Table 1; meanwhile, +ZnO/AC||AC also has

an enhanced charge efficiency (CE = 59%, calculated from 5th cycle due to positive zeta

potential at positive electrode with low surface pH environment. But the desalination

amount and charge efficiency decrease with cycling, and there is no difference from

ACCEPTED MANUSCRIP

T

16

blank AC||AC at 25th cycles due to the dissolution of ZnO under low pH condition.

As studied above, because of severe polarization of the porous carbon, the positive

electrode and negative electrode in CDI cell are very acidic and basic, respectively.

Therefore, the relation between zeta potential of ZnO and the desalination performance of

its corresponding composite electrode is applicable to other additives. Zeta potential of

various metal oxides (MO) nanoparticles (CuO, MnO2 and WO3) at different pH

solutions were measured as shown in Fig. 8 (b-d). And the desalination results together

with those of ZnO/AC composite electrodes were collected in Table 1. Obviously, there is

different desalination performance on the positive and negative electrodes in all cases.

Similar to ZnO/AC, there is an obvious increase in desalination amount and charge

efficiency on both CuO/AC positive and CuO/AC negative electrodes (data from 5th

cycle), due to the reverse zeta potential (charge property) in acidic and basic conditions

(curve a and b of Fig. 8). But the cycling stability is poor. The desalination amount at the

25th cycle became less than 4 mg/g for both positive and negative electrode, lower than

that on pure AC||AC. The acidic surface on positive electrode resulted in the dissolution

of CuO. Some blue deposits were observed on negative electrode surface, probably

forming copper hydroxide, which further blocks the active site of AC. Therefore, CuO is

unsuitable for use as negative and positive additives. Because the isoelectric points of

MnO2 and WO3 are rather low (Fig. 8 c&d), they remain the negative charge state either

on negative electrode or on positive electrode. When they were in positive electrode, a

ACCEPTED MANUSCRIP

T

17

retarded charge-discharge behavior was observed with lower charge efficiency than that

on pure AC electrode. It indicates that with the addition of MnO2 and WO3, the negative

charges on the positive electrode produce a reverse electric field, inhibiting the

adsorption-desorption of the anions. As negative additives of CDI, MnO2 is better than

WO3, while WO3 showed low desalination efficiency with much slow discharge process,

probably due to high dielectric characteristic.

It reveals that zeta potential analysis of MO is much helpful for the selection of

suitable electrode materials for practical application, especially for the materials whose

charge state is sensitive to pH.

4. Conclusions

The ZnO/AC composite electrode was successfully fabricated by simple mechanical

mixing, with excellent electrochemical behaviors as compared with the AC electrode. The

desalination performance was studied by assembling the asymmetric capacitor with

ZnO/AC composite as either negative electrode or positive electrode. And ZnO/AC||AC

has better desalination performance with excellent cycling stability and reproducibility

during charge-discharge process. While the dissolution of ZnO results in poor

desalination behavior on +ZnO/AC||AC. The polarization of electrode, independent of

electrode materials and electrolytes, was analyzed by measuring the surface pH of

electrode. The enhancement of desalination amount and charge efficiency on negative

electrode was explained according to zeta potential of ZnO. The extended studies on

ACCEPTED MANUSCRIP

T

18

other MOs/AC composites demonstrate that the charge efficiency and desalination

amount can be enhanced by properly choosing the positive/negative electrode materials

in the light of zeta potential of the additives. This research provides a particularly

important base and guidance for the selection of electrode materials and optimizing

operation process in capacitive desalination. More detailed studies such as MO size and

ratio and charge-discharge mode effects are still under way.

Acknowledgement:

This work was financially supported by the National Natural Science Foundation (No.

21105009, 21476047), the Fundamental Research Funds for the Central Universities

(2232012A3-05,) and foundation of State Key Laboratory of Electroanalytical Chemistry

(SKLEAC201205).

References:

[1] Y. Oren, Capacitive delonization (CDI) for desalination and water treatment - past,

present and future (a review), Desalination, 228 (2008) 10-29.

[2] S. Porada, R. Zhao, A. van der Wal, V. Presser, P.M. Biesheuvel, Review on the

science and technology of water desalination by capacitive deionization, Prog.

Mater. Sci., 58 (2013) 1388-1442.

[3] M.A. Anderson, A.L. Cudero, J. Palma, Capacitive deionization as an electrochemical

means of saving energy and delivering clean water. Comparison to present

desalination practices: Will it compete?, Electrochim. Acta, 55 (2010) 3845-3856.

ACCEPTED MANUSCRIP

T

19

[4] C.J. Gabelich, T.D. Tran, I.H. Suffet, Electrosorption of inorganic salts from aqueous

solution using carbon aerogels, Environ. Sci. Technol., 36 (2002) 3010-3019.

[5] D.C. Wu, X. Chen, S.H. Lu, Y.R. Liang, F. Xu, R.W. Fu, Study on synergistic effect

of ordered mesoporous carbon and carbon aerogel during electrochemical charge-

discharge process, Micropor. Mesopor. Mat., 131 (2010) 261-264.

[6] Z.H. Huang, M. Wang, L. Wang, F.Y. Kang, Relation between the charge efficiency

of activated carbon fiber and its desalination performance, Langmuir, 28 (2012)

5079-5084.

[7] L. Pan, X. Wang, Y. Gao, Y. Zhang, Y. Chen, Z. Sun, Electrosorption of anions with

carbon nanotube and nanofibre composite film electrodes, Desalination, 244 (2009)

139-143.

[8] G. Wang, Q. Dong, Z. Ling, C. Pan, C. Yu, J. Qiu, Hierarchical activated carbon

nanofiber webs with tuned structure fabricated by electrospinning for capacitive

deionization, J. Mater. Chem., 22 (2012) 21819-21823.

[9] K. Dai, L.Y. Shi, D.S. Zhang, J.H. Fang, NaCl adsorption in multi-walled carbon

nanotube/active carbon combination electrode, Chem. Eng. Sci., 61 (2006) 428-433.

[10] Y. Wimalasiri, L.D. Zou, Carbon nanotube/graphene composite for enhanced

capacitive deionization performance, Carbon, 59 (2013) 464-471.

[11] Z.-Y. Yang, L.-J. Jin, G.-Q. Lu, Q.-Q. Xiao, Y.-X. Zhang, L. Jing, X.-X. Zhang, Y.-

M. Yan, K.-N. Sun, Sponge-templated preparation of high surface area graphene

with ultrahigh capacitive deionization performance, Adv. Funct. Mater., (2014) n/a-

ACCEPTED MANUSCRIP

T

20

n/a.

[12] B.P. Jia, L.D. Zou, Graphene nanosheets reduced by a multi-step process as high-

performance electrode material for capacitive deionisation, Carbon, 50 (2012) 2315-

2321.

[13] H. Marsh, Activated carbon., 1st ed, Elsevier, Boston (MA), 2006.

[14] G. Wang, B. Qian, Q. Dong, J. Yang, Z. Zhao, J. Qiu, Highly mesoporous activated

carbon electrode for capacitive deionization, Sep. Purif. Technol., 103 (2013) 216-

221.

[15] L. Zou, G. Morris, D. Qi, Using activated carbon electrode in electrosorptive

deionisation of brackish water, Desalination, 225 (2008) 329-340.

[16] S. Nadakatti, M. Tendulkar, M. Kadam, Use of mesoporous conductive carbon black

to enhance performance of activated carbon electrodes in capacitive deionization

technology, Desalination, 268 (2011) 182-188.

[17] J.B. Lee, K.K. Park, S.W. Yoon, P.Y. Park, K.I. Park, C.W. Lee, Desalination

performance of a carbon-based composite electrode, Desalination, 237 (2009) 155-

161.

[18] J. Liu, S. Lu, H. Yang, W. Cai, Supercapacitor and method for making the same, US

patent, 2009, US200900991289.

[19] J. Yang, L.D. Zou, N.R. Choudhury, Ion-selective carbon nanotube electrodes in

capacitive deionisation, Electrochim. Acta, 91 (2013) 11-19.

[20] B. Jia, L. Zou, Wettability and its influence on graphene nansoheets as electrode

ACCEPTED MANUSCRIP

T

21

material for capacitive deionization, Chem. Phys. Lett., 548 (2012) 23-28.

[21] H.B. Li, F. Zaviska, S. Liang, J. Li, L.J. He, H.Y. Yang, A high charge efficiency

electrode by self-assembling sulphonated reduced graphene oxide onto carbon fibre:

towards enhanced capacitive deionization, J. Mater. Chem. A, 2 (2014) 3484-3491.

[22] M. Lu, J. Liu, S. Wang, J. Cheng, Preparation of sulfonated graphene/activated

carbon composite electrode and asymmetric capacitor for deionization, Chem. J.

Chinese Uni. 35 (2014) 1546-1552.

[23] R. Zhao, P.M. Biesheuvel, A. van der Wal, Energy consumption and constant

current operation in membrane capacitive deionization, Energy Environ. Sci., 5

(2012) 9520.

[24] Y.J. Kim, J.H. Choi, Improvement of desalination efficiency in capacitive

deionization using a carbon electrode coated with an ion-exchange polymer, Water

research, 44 (2010) 990-996.

[25] X. Gao, J. Landon, J.K. Neathery, K.L. Liu, Modification of carbon xerogel

electrodes for more efficient asymmetric capacitive deionization, J. Electrochem.

Soc., 160 (2013) E106-E112.

[26] E. Avraham, M. Noked, I. Cohen, A. Soffer, D. Aurbach, The dependence of the

desalination performance in capacitive deionization processes on the electrodes

PZC, J. Electrochem. Soc., 158 (2011) P168-P173.

[27] C. Kim, J. Lee, S. Kim, J. Yoon, TiO2 sol–gel spray method for carbon electrode

fabrication to enhance desalination efficiency of capacitive deionization,

ACCEPTED MANUSCRIP

T

22

Desalination, 342 (2014) 70-74.

[28] P.I. Liu, L.C. Chung, H. Shao, T.M. Liang, R.Y. Horng, C.C.M. Ma, M.C. Chang,

Microwave-assisted ionothermal synthesis of nanostructured anatase titanium

dioxide/activated carbon composite as electrode material for capacitive deionization,

Electrochimica. Acta, 96 (2013) 173-179.

[29] H.J. Yin, S.L. Zhao, J.W. Wan, H.J. Tang, L. Chang, L.C. He, H.J. Zhao, Y. Gao,

Z.Y. Tang, Three-dimensional graphene/metal oxide nanoparticle hybrids for high-

performance capacitive deionization of saline water, Adv. Mater., 25 (2013) 6270-

6276.

[30] M.T.Z. Myint, J. Dutta, Fabrication of zinc oxide nanorods modified activated

carbon cloth electrode for desalination of brackish water using capacitive

deionization approach, Desalination, 305 (2012) 24-30.

[31] M.T.Z. Myint, S.H. Al-Harthi, J. Dutta, Brackish water desalination by capacitive

deionization using zinc oxide micro/nanostructures grafted on activated carbon cloth

electrodes, Desalination, 344 (2014) 236-242.

[32] K. Laxman, M.T.Z. Myint, H. Bourdoucen, J. Dutta, Enhancement in ion adsorption

rate and desalination efficiency in a capacitive deionization cell through improved

electric field distribution using electrodes composed of activated carbon cloth coated

with zinc oxide nanorods, ACS Appl. Mater. Interfaces, 6 (2014) 10113−10120.

[33] A.G. El-Deen, N.A.M. Barakat, H.Y. Kim, Graphene wrapped MnO2-nanostructures

as effective and stable electrode materials for capacitive deionization desalination

ACCEPTED MANUSCRIP

T

23

technology, Desalination, 344 (2014) 289-298.

[34] J. Yang, L. Zou, H. Song, Z. Hao, Development of novel MnO2/nanoporous carbon

composite electrodes in capacitive deionization technology, Desalination, 276

(2011) 199-206.

[35] M.C. Zafra, P. Lavela, G. Rasines, C. Macías, J.L. Tirado, C.O. Ania, A novel

method for metal oxide deposition on carbon aerogels with potential application in

capacitive deionization of saline water, Electrochim. Acta, 135 (2014) 208-216.

[36] L.C. Han, K.G. Karthikeyan, M.A. Anderson, J.J. Wouters, K.B. Gregory,

Mechanistic insights into the use of oxide nanoparticles coated asymmetric

electrodes for capacitive deionization, Electrochim. Acta, 90 (2013) 573-581.

[37] K.C. Leonard, J.R. Genthe, J.L. Sanfilippo, W.A. Zeltner, M.A. Anderson, Synthesis

and characterization of asymmetric electrochemical capacitive deionization

materials using nanoporous silicon dioxide and magnesium doped aluminum oxide,

Electrochim. Acta, 54 (2009) 5286-5291.

[38] I. Cohen, E. Avraham, Y. Bouhadana, A. Soffer, D. Aurbach, Long term stability of

capacitive de-ionization processes for water desalination: The challenge of positive

electrodes corrosion, Electrochim. Acta, 106 (2013) 91-100.

[39] X. Gao, A. Omosebi, J. Landon, K.L. Liu, Enhancement of charge efficiency for a

capacitive deionization cell using carbon xerogel with modified potential of zero

charge, Electrochem. Commun., 39 (2014) 22-25.

[40] M.S. Hong, S.H. Lee, S.W. Kim, Use of KCl Aqueous Electrolyte for 2 V

ACCEPTED MANUSCRIP

T

24

manganese oxide/activated carbon hybrid capacitor, Electrochem. Solid State Lett.,

5 (2002) A227-A230.

[41] H. Noei, H.S. Qiu, Y.M. Wang, E. Loffler, C. Woll, M. Muhler, The identification

of hydroxyl groups on ZnO nanoparticles by infrared spectroscopy, Phys. Chem.

Chem. Phys., 10 (2008) 7092-7097.

[42] B.E. Conway;, Electrochemical supercapacitors: Scientific fundamentals and

techanological applications, Kluwer Academic/Plenum Publishers, New York, 2002.

[43] T. Brousse, M. Toupin, D. Bélanger, A Hybrid activated carbon-manganese dioxide

capacitor using a mild aqueous electrolyte, J. Electrochem. Soc., 151 (2004) A614.

[44] A. Soffer, M. Folman, The electrical double layer of high surface porous carbon

electrode, J. Electroanal. Chem. Interf. Electrochem., 38 (1972) 25-43.

Fig. 1

10 m10 m

A B

ACCEPTED MANUSCRIP

T

25

Fig.1. SEM images of ZnO/AC (A) and AC (B) electrodes

ACCEPTED MANUSCRIP

T

26

Fig. 2

Fig.2. (A) CVs of the ZnO/AC (solid line) and AC (dotted line) electrodes in 1 M NaCl

solution at a scan rate of 2 mV s-1. (B) EIS of the ZnO/AC (square) and AC (circle)

electrode in 1 M NaCl solution. The inset of B is the magnified part in high

frequency.

-0.2 -0.1 0.0 0.1 0.2 0.3

-60

-30

0

30

60

i /(m

A g-1

)

V / V

ZnO/AC AC A

3 6 9 12 150

3

6

9

12

15

-Z''/

Ohm

Z'/Ohm

ZnO/AC AC

B

1 20

1

2

-Z''/

Ohm

Z'/Ohm

ACCEPTED MANUSCRIP

T

27

Fig. 3

Fig.3. Cell voltage/current curves (A, C) and conductivity variation (B, D) with time on

ZnO/AC||AC (A, B) and +ZnO/AC||AC (C, D) in the continuous charge-discharge

process.

A

-1

-0.5

0

0.5

1

1.5

0 150 300 450t /min

Δ V

/V

-10

5

20

35

50

65

I /m

A

C

-1

-0.5

0

0.5

1

1.5

0 150 300 450t /min

Δ V

/V

-10

5

20

35

50

65

I /m

A

B

750

850

950

1050

0 150 300 450t /min

Con

duct

ivity

/ μS

cm

-1

D

750

850

950

1050

0 150 300 450t /min

Con

duct

ivity

/ μS

cm

-1

ACCEPTED MANUSCRIP

T

28

Fig. 4

Fig.4. Conductivity variations of CDI cells in 500 ppm NaCl solution. Solid, dotted and

dashed lines correspond to ZnO/AC||AC, +ZnO/AC||AC and AC||AC, respectively.

370

410

450

490

0 20 40 60t / min

cNaC

l / m

g L-1

-ZnO/AC II AC+ZnO/AC II AC AC II AC

ACCEPTED MANUSCRIP

T

29

Fig. 5

Fig.5. Concentration of zinc ions in the desalination solution with long-term circulating,

(square) ZnO/AC||AC, (triangle) +ZnO/AC||AC

0

2

4

6

8

0 50 100 150t / h

cZn /

mg

L-1

+ZnO/AC II AC

-ZnO/AC II AC

ACCEPTED MANUSCRIP

T

30

Fig. 6

Fig.6. Surface pH variation of negative (circle: AC, square: ZnO/AC) and positive

(triangle: ZnO/AC, diamond: AC) electrodes as changing terminal voltage of the

capacitors during charging process

0

3

6

9

12

0.3 0.6 0.9 1.2 1.5ΔV / V

pHZnO- AC+ZnO+ AC-

ACCEPTED MANUSCRIP

T

31

Fig. 7

Fig.7. Surface pH of positive and negative electrodes in AC|| AC capacitor by charging 8

mAh in different concentrations of NaCl solution

0

3

6

9

12

15

0.5 1 8.5 35 200

cNaCl / g L-1

Surfa

ce p

Hpositive electrodenegative electrode

ACCEPTED MANUSCRIP

T

32

Fig. 8

Fig.8. Zeta potential of (a) ZnO, (b) CuO, (c) MnO2 and (d) WO3 nanoparticles at different pH

solutions.

Table 1. Comparison of desalination amount and charge efficiency (CE) on different capacitors

Metal oxides Isoelectric

point Capacitor Desalination

amount(mg/g) CE (%) Comments

- - AC‖AC 4.5 40 stable

ZnO 8 +ZnO/AC‖AC 7.2(a), 4.2(b) 59(a), 39(b) unstable, Zn release

–ZnO/AC‖AC 9.4 80.5 stable

CuO 9 +CuO/AC‖AC 8.7(a), 3.8(b) 74(a), 40(b) unstable, Cu release

–CuO/AC‖AC 7.9(a), 3.6(b) 77(a), 60(b) unstable, blue deposits

MnO2 <2 +MnO2/AC‖AC 5.9 38 retarded charge-discharge time

-55

-35

-15

5

25

45

2 4 6 8 10 12pH

Zeta

pot

entia

l /m

V

(d)

(c)

(b)(a)

ACCEPTED MANUSCRIP

T

33

–MnO2/AC‖AC 8.2 71.4 stable

WO3 <2 +WO3/AC‖AC 2.7 12.7 retarded charge-discharge time

–WO3/AC‖AC 6.4 44.7 retarded discharge time (a) calculated from 5th charging cycle; (b) calculated from 25th charging cycle.

ACCEPTED MANUSCRIP

T