S

U

SE

a

ARRAA

KSSEH

1

dtbltriaccwaVattbg

I

sh

0d

Journal of Hazardous Materials 162 (2009) 1583–1587

Contents lists available at ScienceDirect

Journal of Hazardous Materials

journa l homepage: www.e lsev ier .com/ locate / jhazmat

hort communication

se of solar cell in electrokinetic remediation of cadmium-contaminated soil

onghu Yuan, Zhonghua Zheng, Jing Chen, Xiaohua Lu ∗

nvironmental Science Research Institute, Huazhong University of Science and Technology, Wuhan 430074, PR China

r t i c l e i n f o

rticle history:eceived 29 November 2007eceived in revised form 11 June 2008ccepted 11 June 2008vailable online 20 June 2008

a b s t r a c t

This preliminary study used a solar cell, instead of direct current (DC) power supply, to generate electricfield for electrokinetic (EK) remediation of cadmium-contaminated soil. Three EK tests were conductedand compared; one was conducted on a cloudy and rainy day with solar cell, one was conducted on a sunnyday with solar cell and another was conducted periodically with DC power supply. It was found that theoutput potential of solar cell depended on daytime and was influenced by weather conditions; the applied

eywords:olar celloil remediationlectrokineticeavy metals

potential in soil was affected by the output potential and weather conditions, and the current achievedby solar cell was comparable with that achieved by DC power supply. Solar cell could be used to drivethe electromigration of cadmium in contaminated soil, and removal efficiency achieved by solar cell wascomparable with that achieved by DC power supply. Compared with traditional DC power supply, usingsolar cell as power supply for EK remediation can greatly reduce energy expenditure. This study provided

the E

patdsrctraor

gaaarT

an alternative to improveremediation.

. Introduction

Electrokinetic (EK) technology has been developed for the reme-iation of low permeable contaminated soil [1]. The great potentialo remove pollutants (organic and inorganic) has been proved inatch and field tests [2–4]. However, the field application is very

imited till now. High cost, including electrode-conditioning solu-ion and energy expenditure, is one of the most important factorsestricting the application. Most previous work was focused on themprovement of electrode conditioning, such as using appropriategents [5–7], ionic exchange membrane [8,9], electrode solutionirculation [10] and so on [11], to increase the remediation effi-iency. Researchers have proposed the integration of EK processith other technologies to propel its application [12,13]. Little

ttention has been paid to reduce the electric energy expenditure.ery recently, we have developed a galvanic cell, made from ironnd carbon, to generate electric potential to drive the electromigra-ion of cadmium in contaminated soil [14]. Nevertheless, the elec-romigration rate was very low due to the low potential generated

y the galvanic cell (less than 1 V). How to achieve high potentialradient across soil is the bottleneck for its use in EK remediation.

Solar cell has attracted more and more interest in the past years.t has been deployed for a wide range of applications including sup-

∗ Corresponding author. Tel.: +86 27 87792159; fax: +86 27 87792159.E-mail addresses: yuansonghu622@hotmail.com (S. Yuan),

corpionzzh@163.com (Z. Zheng), chjing@mail.tsinghua.edu.cn (J. Chen),ust-esri@hotmail.com (X. Lu).

ro

2

2

wa

304-3894/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.jhazmat.2008.06.038

K soil remediation and expanded the use of solar cell in environmental

© 2008 Elsevier B.V. All rights reserved.

lying power for consumer products such as electronic calculatorsnd garden lights, and for supplying power in developing coun-ries for water pumping and street lighting [15]. What is more,irect current (DC), not alternating current (AC), is generated byolar cell. This is beneficial for EK remediation because direct cur-ent is normally required. As known, the power generation of solarell depends on sunlight, which suggests that the power genera-ion was influenced by daytime and weather conditions. Since EKemediation is a long time course, the intermittent running can becceptable. In fact, Reddy and Saichek have addressed that peri-dic running led to higher remediation efficiency than continuousunning [16].

In the present study, we used solar cell, instead of DC power, toenerate electric field for EK soil remediation. Cadmium was used asrepresentative heavy metal pollutant because of its high toxicitynd extensive existence. The remediation efficiency on a cloudynd rainy day was compared with that on a sunny day. Also, theesults were compared with those achieved by DC power supply.he objectives are to evaluate the feasibility of using solar cell for EKemediation and to investigate the influence of weather conditionn the remediation efficiency.

. Experimental

.1. Chemicals

CdCl2·2.5H2O (Tingxin Chemical Reagent Factory, China, >99.0%)as used as the source of cadmium. Kaolin was often used

s the model clayed soil in EK laboratory experiments [5,7]

1584 S. Yuan et al. / Journal of Hazardous M

bekcvct(

2

cc[uio(enttew

LlncelTcW

NTb

a((fts(o8TtHvw

2

ttfmsmftbAla

3

3

cvratdwdtap

f6r44crtS2, it was dark in the first 12.5 h (6:00 p.m. to 6:30 a.m.), sunny

Fig. 1. Sketch of experimental setup.

ecause of its low buffer capacity, low organic content, low cationxchange capacity and inertia. The main characteristics of theaolin (Shanghai Fengxian Fengcheng Chemical Reagent Factory;hemical purity) used in this study have been provided in our pre-ious work [17]. It is clayed, slightly acidic (pH 5.75), negativelyharged (zero point of charge (ZPC) 3.36) and contains low con-ent of organic matters (0.28%) and low cation exchange capacity1.71 cmol/100 g).

.2. Procedures

The preparation of cadmium-contaminated soil (140 mgadmium/kg dry soil) was similar to our previous work [14]. Theoncentration was far less than the maximal adsorption capacity17]. The water content of 50% was made by controlling the vol-me of water added. The sketch of experimental setup is shown

n Fig. 1. The soil cell was made of perspex with an inner sizef length 11.1 cm × width 5.4 cm × height 5.0 cm. Graphite sheetslength 4.9 cm × width 4.9 cm × thickness 1.0 cm) were placed atach end and were used as anode and cathode. The contami-ated soil was gradually added to soil cell and was compactedo avoid space. About 300 g moist soil was added to the soil cello attain the density of 1.31 g/cm3. The top of soil cell was cov-red with a piece of weighing paper to prevent the evaporation ofater.

Solar cell panel (SL005-12, Zhangjiagang Yongneng Photovoltaictd., Inc., China) was used as the power supply. The dimension isength 30.6 cm × width 21.5 cm. The nominal peak power, nomi-al voltage, nominal current, open circuit voltage, and short circuiturrent are 5 W, 16.8 V, 0.3 A, 21 V and 0.34 A, respectively. In thexperiment, the solar cell panel was horizontally placed in a table

ocated in the top of a five-floor building (about 25 m height).he side with crystalline was faced to sunlight. Note that the soilell was placed inside the building. The longitude and latitude of

uhan city, where experiments were carried out, are E114◦21′ and

f(6c

aterials 162 (2009) 1583–1587

30◦37′, respectively, indicating the relatively sufficient sunlight.he experiments were performed in late autumn (around Novem-er 1, 2007).

Two experiments were conducted using the solar cell panels power supply. One was conducted on a cloudy and rainy daydenoted as S1) and the other was conducted on a sunny daydenoted as S2). The surrounding temperature in S1 and S2 rangedrom 11 ◦C to 18 ◦C and from 11 ◦C to 27 ◦C, respectively. One addi-ional experiment was conducted as a comparison with DC powerupply (GPC-H, 30V/5A, Taiwan Guwei Electronic Ltd., Inc., Taiwan)denoted as D1). In order to get comparable result, the maximalutput potential was used D1 and the running periodically lastedh per day to simulate the alternation of daytime and darkness.he duration of all the experiments was for 48 h. The output poten-ial and electric current were measured by a multimeter (Hangzhouuasheng Instrument Factory, China) connected in the circuit. Theoltage between anode and cathode, the applied potential in soil,as measured by a multimeter.

.3. Analysis of samples

At the end of the experiments, soil in the cell was equally slicedo five pieces from anode to cathode. pH and cadmium concen-ration in each piece were analyzed. The samples were air-driedor over 48 h, then ground and sieved by 0.25 mm screen. pH was

easured with a pH meter (Hanna Company, Italy) by mixing 5 goil sample with 10 mL deionized water. For the analysis of cad-ium, 1 g soil was digested with 5 mL of concentrated nitric acid,

ollowed by centrifugation for 10 min (5000 rpm). The concentra-ion of cadmium in the supernatant was diluted and determinedy atomic absorbance spectrophotometer (WFX-110, Beijing Ruilinalytical Instrument Co. Ltd.). Each sample was prepared in trip-

icate. The recovery of cadmium in the processes was verified to bebove 90%.

. Results and discussion

.1. Variation of electric potential and current

The variation of output potential, applied potential in soil andurrent in the EK process are shown in Fig. 2. The discontinuousariation of potential and current in D1 was due to the intermittentun, while the discontinuous variation in S1 and S2 was due to thelternation of daytime and darkness. As the maximal output poten-ial of solar cell was determined to be 18.8 V in the first 24 h andeclined in the second 24 h, the output potential of DC power (D1)as set at 18.8 V in the first 24 h and 16.8 V in the second 24 h. Theuration at high output potential was about 8 h per day (Fig. 2a), sohe running time D1 was set at 8 h per day. Fig. 2a shows that thepplied potential in soil in D1 was a litter lower than the outputotential.

The variation of output potential of solar cell was quite differentrom that in D1. In S1, it was cloudy in the first 5 h (1:00 p.m. to:00 p.m.), dark from 5 h to 17.5 h (6:00 p.m. to 6:30 a.m.), slightlyainy from 17.5 h to 29 h (6:30 a.m. to 6:00 p.m.), dark from 29 h to1.5 h (6:00 p.m. to 6:30 a.m.) and cloudy with weak sunlight from1.5 h to 48 h (6:30 a.m. to 1:00 p.m.). The output potential of solarell was zero in darkness. It is encouraging that the output potentialeached around 15 V on rainy day. The output potential greatly fluc-uated, particularly during the time of sunrise and sunset. While in

rom 12.5 h to 24 h (6:30 a.m. to 6:00 p.m.), dark from 24 h to 36.5 h6:00 p.m. to 6:30 a.m.) and sunny from 36.5 h to 48 h (6:30 a.m. to:00 p.m.). Thus, the output potential in S2 was maintained almostonstant at relatively high level in daytime compared with that in

S. Yuan et al. / Journal of Hazardous Materials 162 (2009) 1583–1587 1585

Fe

Stt

STtebTopCtsooc

twptDi[w

li

tsbbap

3

IcdwawoFDtlnZema

3

saaogWwd

ig. 2. Variation of (a) output potential and real applied potential to soil and (b)lectric current.

1. A little higher potential was obtained in S2 than in S1, suggestinghat sunny day is more beneficial to generate high output potentialhan cloudy and rainy day.

It should be noted that the applied potential in soil in S1 and2 were much lower than the output potential, particularly in S1.his suggested that cloudy and rainy day had negative impact onhe applied potential in soil. Fig. 2a also shows that the differ-nce between the output potential and applied potential in soilecame significant when the output potential was less than 15 V.his implied that the applied potential in soil was dependent on theutput potential. The difference of output potential and appliedotential in soil was not great when DC power supply was used.ompared with DC power supply, solar cell has larger inner resis-ance [15]. The potential partitioned in the inner circuit becameignificant when the output potential was less than the nominalutput potential. As a consequence, it is necessary to maintain theutput potential above 15 V to achieve a high EK remediation effi-iency.

When the electric current is concerned (Fig. 2b), it was foundhat the current decreased with time in D1. The maximal currentas often obtained at the start of test when the quantity of ions inore solution is greatest [5,14]. The lower initial current in S1 and S2

han in D1 was due to the lower applied potential in soil achieved. In1, the current declined with duration because of the migration of

ons and the precipitation of cations with OH− produced at cathode5]. While in S1 and S2, the variation of current was in agreementith that of the applied potential in soil, which conformed to Ohm’s

psaww

Fig. 3. Distribution of soil pH at the end of test.

aw. The current obtained in S1 and S2 was comparable with thatn D1, so the effective electromigration can be expected.

With respect to the application of solar cell, it can be concludedhat solar cell can be used to generate potential gradient acrossoil, the output potential depended on daytime and was influencedy weather conditions, the applied potential in soil was affectedy weather conditions and the output potential, and the currentchieved by solar cell was comparable with that achieved by DCower supply.

.2. Variation of soil pH

The variation of pH from anode to cathode is presented in Fig. 3.t is clear that the soil pH near anode dropped to about 2 and nearathode rose to above 9. It has been well known that anode pro-uces H+ and cathode produces OH− [1]. The variation of soil pHas attributed to the electromigration of H+ to cathode and OH− to

node. Soil pH in most regions was below the initial value, whichas ascribed to the faster electromigration rate of H+ than thatf OH− [1]. This result was in agreement with the literature [5,7].ig. 3 shows that the soil pH in S1 was slightly higher than that in1 and S2. The current and applied potential in soil in S1 was lower

han that in D1 and S2 during most operating time, resulting in theess electrode reactions. Fig. 3 also demonstrates that the soil pHear anode dropped to below ZPC. When soil pH became less thanPC, the direction of electroosmosis was reversed [1]. Although thelectroosmotic flow was not collected in the process, water accu-ulation around graphite cathode was observed in the initial 30 h

nd became minute in the later stage.

.3. Electromigration of cadmium

Generally, the electromigration of heavy metal in contaminatedoil involved two steps, the desorption from soil to pore solutionnd the subsequent electromigration [14]. It is impossible to movedsorbed heavy metals in soil. The variation of cadmium at the endf test is presented in Fig. 4. In all the tests, cadmium electromi-rated from anode to cathode and was accumulated near cathode.e previously found that the desorption of cadmium from kaolinas highly dependent on soil pH [17]. The desorption efficiencyecreased sharply with the increase of pH and became minute at

H above 4.8 [17]. The variation of soil pH in Fig. 3 reveals that theoil pH from anode to the section of 0.7 (normalized distance fromnode) was lower than 4.8 and from the section of 0.7 to cathodeas higher than 4.8. Therefore, cadmium in the region where pHas below 4.8 was greatly removed. Cadmium in low pH region

1586 S. Yuan et al. / Journal of Hazardous M

wtclh[

otmrwTgpetEwo

3a

sstsccTbbb

(ea1rledatc¥

ihtDt

cPpcncirr

4

cpbbapcete

A

eFSn

R

Fig. 4. Distribution of cadmium at the end of test.

as moved to the next region where pH was above 4.8. Due tohe increase of adsorption capacity with the increase of pH [17],admium was re-adsorbed in this region, which led to the accumu-ation of cadmium near cathode. Our previous work also found theeavy dependence of cadmium removal with soil pH in EK process14].

Fig. 4 shows that the variation of cadmium from anode to cath-de at the end of test was similar in the three experiments. Fromhe mass balance calculation, the total removal efficiency of cad-

ium in D1, S1 and S2 were obtained as 14.9%, 17.1% and 18.3%,espectively. Although the potential and current in the three testsere different, the removal efficiency of cadmium was comparable.

his result may be due to the influence of soil pH on the electromi-ration. Soil pH determines the desorption of cadmium from soilarticles to pore solution, and the desorption is prerequisite for thelectromigration. Different potential led to similar electromigra-ion suggesting that desorption was the rate-limited step for theK removal. Similar results were also found in our previous work,here pH showed much more significant effect on the EK removal

f cadmium than potential gradient [14].

.4. Discussion on the energy expenditure and full-scalepplication

When the power is supplied by solar cell, the electric energy con-umption can be absolutely spared. It has been proved above thatolar cell panel alone can be used as power supply for EK remedia-ion. As a result, the other equipment that is necessary in solar cellystem like storage battery and inverter will not be used. The costoncerning electric energy consumption is solely resulted from theapital of solar cell. The price of solar cell is about ¥35/W in China.he lifetime of solar cell is about 20 years, so the running cost cane calculated by the division of capital by lifetime. The running costy using traditional DC power supply will be compared with thaty using solar cell in the following paragraph.

Ground-Water Remediation Technologies Analysis CenterGWRTAC) concluded from pilot scale EK remediation that thenergy consumption in extracting heavy metals from soil waspproximately 500 kW h/m3 or more at an electrode spacing of.0–1.5 m [18]. In pilot-scale EK tests, Acar and Alshawabkeheported the energy expenditure of 700 kW h/m3 to remove 55%ead from soils for 122 days [19], and Zhou et al. showed the energyxpenditure of 224 kW h/t to remove 76% copper from soils for 140

3

ays [6]. Assuming the average energy expenditure is 500 kW h/mnd the remediation duration is 100 days [18], we may calculatedhe power as 208 W/m3 (500/(100 × 24)). Thus, the capital of solarell is ¥7280/m3 (35 × 208), and the running cost is calculated to be100/m3 (7280/(20 × 365/100)). The electric energy cost in China

[

[

aterials 162 (2009) 1583–1587

s ¥0.5/kW h, so the energy expenditure is ¥250/m3, which is muchigher than the running cost of using solar cell. From this aspect,he use of solar cell is more economical than the use of traditionalC power supply. Furthermore, the capital of solar cell is expected

o decrease in future.In the full-scale application of EK remediation driven by solar

ell, the cell panel can be installed above the contaminated site.resently, 1 m2 panel area can generate the power of 100 W, so theanel area is about 2 m2/m3 soil. Since direct current from solarell can be used in EK remediation, the wires from positive andegative electrode can be directly connected with the anode andathode of EK system, respectively. The installation of other affil-ated equipment was similar to that of traditional EK system. EKemediation driven by solar cell is particularly effective in the areasich in sunlight.

. Conclusions

This study investigated the EK remediation of cadmium-ontaminated soil driven by a solar cell. It was found that the outputotential of solar cell depended on daytime and was influencedy weather conditions, the applied potential in soil was affectedy the output potential and weather conditions, and the currentchieved by solar cell was comparable with that achieved by DCower supply. The electromigration of cadmium in soils by solarell was comparable with that by traditional DC power supply. How-ver, the running cost of EK remediation by solar cell is much lowerhan that by DC power supply. Solar cell could be used to drive thelectromigration of cadmium in contaminated soil.

cknowledgements

This work was partly supported by the National Natural Sci-nce Foundation (No. 20777024), the key project of Natural Scienceoundation of Hubei Province (No. 2006ABD005) and the Nationalcience Foundation of Huazhong University of Science and Tech-ology.

eferences

[1] R.F. Probstein, R.E. Hicks, Removal of contaminants from soils by electric field,Science 260 (1993) 498–503.

[2] R. Lageman, R.L. Clarke, W. Pool, Electro-reclamation, a versatile soil remedia-tion solution, Eng. Geol. 77 (2005) 191–201.

[3] G.M. Nystroem, L.M. Ottosen, A. Villumsen, Electrodialytic removal of Cu, Zn, Pb,and Cd from harbor sediment: influence of changing experimental conditions,Environ. Sci. Technol. 39 (2005) 2906–2911.

[4] S. Ho, C. Athmer, P.W. Sheridan, B.M. Hughes, B. Orth, P.H. Mckenzie, A.M.Shapiro, T.M. Stivavec, J. Salvo, D. Schultz, R. Landis, R. Griffith, S. Shoemaker,The Lasagna technology for in situ soil remediation. 2. Large field test, Environ.Sci. Technol. 33 (1999) 1092–1099.

[5] R.E. Saichek, K.R. Reddy, Effect of pH control at the anode for the electroki-netic removal of phenanthrene from kaolin soil, Chemosphere 51 (2003) 273–287.

[6] D.M. Zhou, L. Cang, A.N. Alshawabkeh, Y.J. Wang, X.Z. Hao, Pilot-scale elec-trokinetic treatment of a Cu contaminated red soil, Chemosphere 63 (2006)964–971.

[7] S.H. Yuan, M. Tian, X.H. Lu, Electrokinetic movement of hexachlorobenzene inclayed soils enhanced by Tween 80 and �-cyclodextrin, J. Hazard. Mater. B137(2006) 1218–1225.

[8] Z.M. Li, J.W. Yu, I. Neretnieks, Electroremediation: removal of heavy metalsfrom soils by using cation selective membrane, Environ. Sci. Technol. 32 (1998)394–397.

[9] N. Sabba, D.E. Akretche, Selective leaching of a copper ore by an elec-tromembrane process using ammonia solutions, Miner. Eng. 19 (2006) 123–

129.

10] H.H. Lee, J.W. Yang, A new method to control electrolytes pH by circulationsystem in electrokinetic soil remediation, J. Hazard. Mater. B77 (2000) 227–240.

11] M. Pazos, M.A. Sanromán, C. Cameselle, Improvement in electrokinetic reme-diation of heavy metal spiked kaolin with the polarity exchange technique,Chemosphere 62 (2006) 817–822.

ous M

[

[

[

[

[

[copper and cadmium from kaolin enhanced by organic acids, Chemosphere 68

S. Yuan et al. / Journal of Hazard

12] S. Pamukcu, A. Weeks, J.K. Wittle, Enhanced reduction of Cr(VI) by direct electriccurrent in a contaminated clay, Environ. Sci. Technol. 38 (2006) 1236–1241.

13] P. Thepsithar, E.L. Roberts, Removal of phenol from contaminated kaolin usingelectrokinetically enhanced in situ chemical oxidation, Environ. Sci. Technol.40 (2006) 6098–6103.

14] S.H. Yuan, C. Wu, J.Z. Wan, X.H. Lu, Electromigration of cadmium in contami-nated soils driven by single and multiple primary cells, J. Hazard. Mater. B151(2008) 594–602.

15] R.W. Miles, K.M. Hynes, I. Forbes, Photovoltaic solar cells: an overview of state-of-the-art cell development and environmental issues, Prog. Cryst. Growth Ch.51 (2005) 1–42.

[

[

aterials 162 (2009) 1583–1587 1587

16] K.R. Reddy, R.E. Saichek, Enhanced electrokinetic removal of phenanthrene fromclay soil by periodic electric potential application, J. Environ. Sci. Heal. A 39(2004) 1189–1212.

17] S.H. Yuan, Z.M. Xi, Y. Jiang, J.Z. Wan, C. Wu, Z.H. Zheng, X.H. Lu, Desorption of

(2007) 1289–1297.18] Ground-Water Remediation Technologies Analysis Center, Technology

overview report, Electrokinetics, July 1997.19] Y.B. Acar, A.N. Alshawabkeh, Electrokinetic remediation. I. Pilot-scale tests with

lead-spiked kaolinite, J. Geotech. Eng. ASCE 122 (1996) 173–185.