2c

Pa

b

a

ARRAA

KCE2SM

1

mg[awttp

([seEp

(a2

j

0d

Journal of Hazardous Materials 177 (2010) 1126–1133

Contents lists available at ScienceDirect

Journal of Hazardous Materials

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

D crossed electric field for electrokinetic remediation of chromiumontaminated soil

eng Zhanga,b, Chunji Jina,b,∗, Zhenhuan Zhaoa,b, Guobin Tiana,b

Key Laboratory of Marine Environment and Ecology (Ocean University of China), Ministry of Education, Qingdao 266100, PR ChinaCollege of Environmental Science and Engineering, Ocean University of China, Qingdao 266100, PR China

r t i c l e i n f o

rticle history:eceived 21 May 2009eceived in revised form 7 January 2010ccepted 8 January 2010vailable online 15 January 2010

a b s t r a c t

Chromium contaminated soil can be remediated by electrokinetic techniques. However, in practical appli-cation, Cr(VI) may migrate with water deep into the soil, contaminating previously unpolluted layers.Both horizontal and vertical electric fields were applied simultaneously to improve traditional electroki-netic remediation. Contrasting experiments using four operation modes (none, solely horizontal, solelyvertical and 2D crossed electric field) were designed and tested at the bench-scale with the practical

5

eywords:hromiumlectrokineticD crossedoil treatmentigrate

sample of chromium contaminated soil (1.3 × 10 mg/kg) from a chemical plant to investigate Cr(VI)migration downward in each test and the effectiveness and feasible of the new design. During the tests,Cr(VI) could migrate deep into the soil in the solely horizontal mode. Cr(VI) migration downward couldbe prevented by vertical barrier in the solely vertical mode. However, using the 2D crossed mode, Cr(VI)was significantly prevented from migrating downward and the chromium contaminated soil was treatedeffectively. Thus, the 2D crossed electric field is a promising and practical method for the remediation ofcontaminated soils.

. Introduction

Chromium slag is a primary cause of soil pollution. In China,ore than 20% of the hyper-chromium contaminated soil and

roundwater arise from the long-term stacking of chromium slag1]. In regions with this chromium waste material, although morend more chromium slags have been treated by recycling, the soilhere chromium slags deposited has been polluted seriously. If

he chromium contaminated soil was not been remedied in time,he environment of the chromium contaminated soil would haveotential risk yet.

In recent years, in China and abroad, electrokinetic remediationEK) methods for in situ soil remediation have been widely studied2–5]. The principle is that an electric field applied to a soil-water

ystem results in a movement of metal contaminants, primarily bylectroosmosis and electromigration [6]. As a remediation method,K has many merits, including its minor impacts on soil hydraulicermeability, easy installation and high efficiency, among others.

∗ Corresponding author at: Key Laboratory of Marine Environment and EcologyOcean University of China), Ministry of Education, College of Environmental Sciencend Engineering, Ocean University of China, 238 Songling Road, Qingdao, Shandong66100, PR China. Tel.: +86 136 8764 7051/ 532 6678 1061; fax: +86 532 6678 2571.

E-mail addresses: phevos1983@yahoo.com.cn (P. Zhang),inhou@ouc.edu.cn (C. Jin).

304-3894/$ – see front matter © 2010 Elsevier B.V. All rights reserved.oi:10.1016/j.jhazmat.2010.01.038

© 2010 Elsevier B.V. All rights reserved.

At present, EK is widely applied in practical engineering in Americaand Europe.

In a typical procedure, a horizontal electric field is adopted,and water or electrolytes are added to moisten the soil. How-ever, contaminants often migrate downward along with the liquid,by dispersion and diffusion, as they migrate in the electric field.Since many areas with contamination have not adopted effectiveanti-seepage measures, this can result in serious pollution of deepsoils and groundwater. The EK reactor which was used in mostresearches was a horizontal column. And contamination migrationinto deep soil was not investigated. However, Reynolds et al. [7]found that permanganate could migrate downward during EK inwhich a deep and wide V shaped reactor was used.

The application of vertical electric fields has gained interest inrecent years. For example, Chen et al. [8] used a vertical electric fieldto study the migration of copper and zinc complexes, and showedthat metal complexes were effectively prevented from penetratingdeeper soils. Chen et al. [9] also found that transport of metal com-plexes could be influenced by a vertical electric field, and that theefficiency was related to the potential of the electric field. Wang etal. [10] applied an upwardly directed electrokinetic soil remedial

technology in a metal contaminated soil, and found that most ofthe metal ions moved from the inner regions to the surface of thesoil. By using the Lasagna technology [11,12], which included eitherhorizontal or vertical zones between electrodes, soluble organiccompounds may be removed from contaminated low-permeability

ous Materials 177 (2010) 1126–1133 1127

sdhfi

erctetactwEcs

2

2

ptoucpls

2

F

Table 1List of physical and chemical properties of the soil.

Physical and chemical properties Contaminated soil Uncontaminatedsoil

Density (g/cm3) 1.276 1.205

Particle size distribution (%)0.2–0.5 mm 55.47 51.810.5–1 mm 44.53 48.19

Carbonate (%) 1.2 1.4pH 9.08 9.13

P. Zhang et al. / Journal of Hazard

oils. A vertical electric field has also been used to treat deep andense soil [11,12]. However, to data, there have been no studies thatave used a combination of both horizontal and vertical electricelds for soil remediation.

In general, it was assumed that if both horizontal and verticallectric fields were used, heavy metal ions would move along theesultant force direction. Under this condition, contaminated soilould be remedied, while at the same time, the downward migra-ion of contaminants could be slowed or prevented. Contrastingxperiments using four operation modes (none, solely horizon-al, solely vertical and 2D crossed electric field) were arrangednd tested at the bench-scale with the practical aged sample ofhromium contaminated soil in present study. The Cr(VI) migra-ion into deep soil, using traditional horizontal field electric field,as investigated. And the feasible and efficiency of the 2D crossed

K system was assessed in bench-scale experiments. Chromiumontaminated soil from a slag region was used as a model soil toimulate practical application of this method.

. Experimental

.1. The soil sample

An actual chromium contaminated soil was used to simulateractical engineering. A range of soil samples from the surfaceo 30 cm depth were taken from a former chromium slag stockf a chemical plant in Qingdao, China. In addition, a range ofncontaminated soil from 10 cm to 100 cm depth, away from theontaminated region, was selected as reference sample. Each sam-le was air dried, mixed, and passed through a 2 mm sieve in the

aboratory. Physical and chemical properties of the soil are pre-ented in Table 1.

.2. Reactor design

This study used a self-designed reaction system as detailed inig. 1.

Fig. 1. Diagram of the experimental reactor (a: the contami

EC (�S/cm) 1062 112Total chromium (mg/kg) 23357.74 129.91Hexavalent chromium (mg/kg) 837.98 6.34

The reaction system was composed of a DC power, a mainreactor, two electrolyte reservoirs and two hydraulic head con-trol devices. The main reactor was an acrylic cube-shapedcolumn with dimensions of 280 mm × 105 mm × 560 mm, andcomposed of a soil cell (200 mm × 105 mm × 510 mm), two elec-trode chambers (35 mm × 105 mm × 140 mm) and a groundwatercontainer (280 mm × 105 mm × 40 mm) to simulate real contam-inated site. The soil cell and the water container were separatedby a porous acrylic plate (200 mm × 105 mm × 10 mm). In addi-tion, the soil cell was divided into two parts: an EK treatmentzone (200 mm × 105 mm × 100 mm) and a non-treatment zone(200 mm × 105 mm × 370 mm). All the dimensions above wereinner dimension. Three pipes (∅5 mm × 40 mm) were placed atdepths of 160 mm, 260 mm and 360 mm from the top of the mainreactor to investigate Cr migration during electrokinetic remedia-tion process.

Direction of electric field could be controlled according to exper-iments need. When horizontal potential was needed, two graphite

plates (100 mm × 100 mm × 10 mm) were installed in the electrodechambers and separated by a distance of 240 mm. Because most ofthe migrated chromium was hexavalent chromium (Cr(VI)) anionssuch as Cr2O7

2−, CrO42− and HCrO4

− [13], when vertical potential

nated soil zone and b: the uncontaminated soil zone).

1128 P. Zhang et al. / Journal of Hazardous Materials 177 (2010) 1126–1133

F rizone

wwpsp

2

e

eCt2342

bsitnasspofiawCaAteai

asba

TS

ig. 2. Electrodes arrangement in each test (a: no electric field in Test 1, b: solely holectric field in Test 4).

as applied, a graphite anode plate (200 mm × 100 mm × 10 mm)as placed at the top of soil surface as anode, and a porous copperlate (200 mm × 100 mm × 10 mm) was placed 90 mm below theoil surface as cathode. A constant potential was supplied by DCower, and potential gradient used in experiments was 1 V/cm.

.3. Experimental procedure

As shown in Fig. 2, four tests were arranged by changing thelectrode layout. And the testing program is summarized in Table 2.

When contaminated soil was saturated by water and had nolectric field, Cr migration could be investigated in the Test 1 (T1).r migration downward and remediation efficiency will be inves-igated when solely horizontal electric field was applied in the Test(T2). Vertical barrier effectiveness could be discussed in the Test(T3) where solely vertical electric field was applied. And the Test(T4) was arranged to discuss feasibility and effectiveness of the

D crossed electric field for electrokinetic remediation.Before all of the tests, filter paper was placed at each interface

etween soil and liquid. Next, a certain amount of uncontaminatedoil was filled from top of the soil cell. Then distilled water wasnjected into the soil cell from bottom, and the soils were tampedo force the air out at the same time. Layer and layer uncontami-ated soils were filled into the soil cell continuously as this methodnd up to 390 mm above the bottom of the soil cell. Contaminatedoils were then packed into the soil cell over the uncontaminatedoils and up to 460 mm above the bottom of the soil cell. During thisrocess, three 5 mm thickness silica sand layers were filled insteadf the soils at the position of the three pipes, respectively. And thelter paper was placed at each interface between the silica sandnd the pipe to prevent the silica sand from driving out. Soil porousater could through the silica layer and effluent from the pipe. Ther migration downward in this experiment could be investigatedccording to variation of Cr content of porous water in each layer.constant head level at the top of contaminated zone was main-

ained by adding distilled water to the electrolyte chambers andlectrolyte reservoirs with a head control device and pump duringll test. Then electrical connections were made and the test wasnitiated.

The horizontal electrodes could be installed in terms of testrrangement. The graphite anode plate was placed at the top ofoil surface as anode, and a porous copper plate was placed 90 mmelow the soil surface as cathode during soil filling process in T3nd T4. However, the interface between contaminated soil zone

able 2ummary of testing program.

Parameter Test1

Electrolyte solution Distilleda

Applied voltage gradient (V/m) 1.0Electric field style Noneb

Duration (h) 240

a Distilled: distilled water.b None: no electric field was used.c Horizontal: only horizontal field was used.d Vertical: only vertical electric field was used.e 2D crossed: both horizontal and vertical electric field were used, and anode up catho

tal electric field in Test 2, c: solely vertical electric field in Test 3, and d: 2D crossed

and uncontaminated soil zone was 20 mm above the cathode. The20 mm uncontaminated soil layer was arranged as buffer layerbecause of the Cr migration downward at the test beginning. Thisdesign could improve feasibility of the vertical barrier.

During each test, 4 ml porous water was sampled from the eachpipe. The electrolyte pH, the current intensity, and the concen-tration of Cr(VI) in the porous water were measured. After EKtreatment, uncontaminated soils were sampled from three layersof non-treatment soil zone (depths of 140–160 mm, 220–260 mmand 340–360 mm from the top of the main reactor), and each layerwas sliced into three segments from cathode to anode along thehorizontal direction, The EK treatment zone was sliced into six seg-ments from cathode to anode along the horizontal direction in T1and T2. However, the EK treatment zone was divided into threelayers (10–45 mm, 45–80 mm, 80–100 mm from the top of the EKtreatment zone), and then each layer was divided into six sectionsfrom cathode to anode along the horizontal direction in T3 and T4.The soil pH, electroconductivity (EC), oxidation–reduction poten-tial (ORP), moisture, total chromium, and Cr(VI) of these sectionswere measured.

2.4. Chemical analysis

The soil pH was determined using a soil to water ratio of 1:2.5(w/w) by a pH meter (pHS-25, LIDA instrument Co. Ltd). The ECvalue of soil sample was measured in a suspension of 2 g soil in10 ml distilled water using an electroconductivity meter (DDS-307A, Shanghai Precision & Scientific Instrument Co. Ltd). And thewater content was determined using standard methods [14]. Cr(VI)was determined using a diphenylcarbazide colorimetric method[15]. The soil porous solution sample was passed 0.45 �m film, andthe concentration of Cr(VI) in solution tested by visible spectropho-tometer (722N, Shanghai Precision & Scientific Instrument Co. Ltd)at 540 nm. The concentration of Cr(VI) in soil was extracted from 1 gsoil suspended in 15 ml 0.4 M KCl [16]. Then the suspension was alsopassed 0.45 �m film and analyzed with visible spectrophotometerat 540 nm. The total Cr concentration in the soil was determined bywet digestion analysis. Soil samples were digested in a mixture ofHNO3:HClO4:HF–HCl for analysis by atomic absorption and flame

emission spectrophotometer (AAS. M6 THERMO, Thermo FisherScientific).

The reagents which were used in analysis were AR grade. TheCr(VI) standard samples (National Standard Institute, China) wereused in Cr(VI) and total Cr analysis. In addition, all of the analyses

Test2 Test3 Test4

Distilled Distilled Distilled1.0 1.0 1.0Horizontalc Verticald 2D crossede

240 255 255

de down.

ous Materials 177 (2010) 1126–1133 1129

wv

3

3

fsdtvftncapcic2r1ter0te

3

cwmtt

spijmct

P. Zhang et al. / Journal of Hazard

ere performed in duplicate and the data represented the averagedalues.

. Results and discussion

.1. Changes in current intensity during electrokinetic treatment

Fig. 3 shows the variation in the current intensity with timeor each test. (T1 has no electric field, therefore no current inten-ity.) The current intensity immediately increased from 0, and thenecreased slowly, approaching 0 again. At the beginning of theest, the electrolyte was distilled water and the soil resistance wasery high. More and more exchangeable metal ions were desorbedrom the soil into the soil solution with the electrokinetic remedia-ion progressed. The anions desorption could be enhanced by OH−

ear the cathode and hindered by H+ near the anode. However, theations desorption was opposite to anions. Cr(VI) commonly existss soluble ion in alkali soil and precipitation with decrease of theH value. Cr(III) exists as soluble ion when pH below 4 and pre-ipitation or adsorption with increase of the pH [17,18]. Chromiumons and other ions could migrate in soil during electrokinetic pro-ess and current intensity rose to a maximal level. T2 reached.36 mA/cm2 after 35.5 h, T3 reached 0.64 mA/cm2 after 35 h, T4 (H)ose to 1.49 mA/cm2 after 13 h and T4 (V) rose to 0.91 mA/cm2 after7 h. The increase in T4 was the most rapid among all tests due tohe presence of both horizontal and vertical electric fields. Effect oflectrokinetic was enhanced. After reaching a peak value, the cur-ent intensity curves began to descend in all tests and approached. It is due to soluble ions migrating to the opposite electrode andhe other soluble ions adsorption and precipitation. In addition, thelectrode polarization is also the reason.

.2. Migration of Cr(VI) in soil during electrokinetic treatment

Due to Cr(VI) is more toxic and greater mobility than Cr(III),hanges in content of Cr(VI) of porous solution from each pipehich was below the contaminated soil zone could indicate theigration of Cr(VI). Comparing the migration of Cr(VI) in the tests,

he feasibility and effectiveness of 2D crossed electric filed for elec-rokinetic remediation could be discussed.

Fig. 4 shows the variation of Cr(VI) concentration in porousolution with the time progressed. The Cr(VI) concentration of theorous solution increased with time, and then decreased, maintain-

ng fluctuation in a range in T1. The maximum reached 127.78 mg/Lust after 75.8 h at S1 pipe in T1. This indicated that Cr(VI) could

igrate downward due to dispersion and diffusion when the soilontained enough water. If contaminated soil was not treated inime, deep uncontaminated soil and even groundwater would be

Fig. 3. Current measurement (T4 (H) is horizontal and T4 (V) is vertical).

Fig. 4. Cr(VI) concentration of the soil solution sampled from different layers ofuncontaminated soil zone changed with time in (a) T1, (b) T2, (c) T3 and (d) T4. Ineach test, S1, S2 and S3 were three pipes position that were 160 mm, 260 mm, and360 mm from the top of main reactor, respectively, in vertical direction.

1130 P. Zhang et al. / Journal of Hazardous Materials 177 (2010) 1126–1133

FTw

pccrftdmIbtcT[CSfavfi

nS7abotrdv

3e

1r8nec

tbOfi

Before the treatment was initiated, the initial soil moisture con-tent was measured at about 45.6%. The soil moisture content forall tests was smaller than the initial values after test. Evaporationof moisture is likely to happen with the rise in temperature dur-

ig. 5. Soil pH profiles distribution in the EK treatment zone after treatment in T4.op: soil with depth of 0–45 mm; mid: soil with depth of 45–80 mm; bottom: soilith depth 80–100 mm in EK treatment zone.

olluted over the long term due to rainfall. The variation of Cr(VI)oncentration in T2 is similar to T1. However, the maximum Cr(VI)oncentration in T2 was lower than T1 at the same pipe position,espectively. Many of the Cr(VI) ions could migrate horizontallyrom the soil to the anode chamber in T2 due to the horizon-al electric field applied. The amount of Cr(VI) ions transportedownward in T2 was smaller than T1. These indicated Cr(VI) couldigrate to deep layer during horizontal electrokinetic remediation.

n Fig. 4c, the Cr(VI) concentrations of porous solutions fluctuatedelow 1.6 mg/L at each pipe in T3. The maximum in T3 was lowerhan T1 and T2 remarkably. This indicated that vertical electric fieldould prevent Cr(VI) from migrating downward with porous water.he vertical electric field could be applied to plant uptake nutrients19] and decontamination technology [20]. In T4 the variation ofr(VI) concentrations also fluctuated and maximum was 2.1 mg/L at1 pipe. A large amount of the Cr(VI) ions could migrate horizontallyrom the soil to the anode chamber due to horizontal electric field,nd Cr(VI) migration downward was prevented synchronously byertical electric field in T4. This indicated that 2D crossed electriceld was feasible and effective arrangement.

Fig. 4 also shows that Cr(VI) ions migrate from the contami-ated zone to the uncontaminated zone first via S1 and finally via3. Taking Fig. 4a as an example, S1 reached 127.78 mg/L just after5.8 h, S2 reached 77.72 mg/L after 81 h, and S3 reached 21.67 mg/Lfter 166.7 h. The porous solution transportation could be blockedecause the channel in the soil is an irregular net, not a linearne. Therefore, Cr(VI) penetrated from contaminated soil zone tohe deep layer could take some times. In addition, Cr(VI) may beeduced to Cr(III) and adsorbed to soil during Cr(VI) migrationownward. So the deeper the pipe position was, the lower the peakalue of Cr(VI) was.

.3. Soil pH, oxidation–reduction potential andlectroconductivity after electrokinetic treatment

After electrokinetic treatment, the soil pH near the cathode was0.73, and this decreased gradually in a horizontal direction, toeach 8.06 near the anode in T2. The soil pH ranged from 7.47 to.09 near the top layer, while soil pH fluctuated from 11.23 to 11.5ear the bottom layer in T3. It is because H+ and OH− generated bylectrolysis migrated to opposite electrodes through the soil, andhanged the soil pH [21].

As shown in Fig. 5, the soil pH decreased from the cathode tohe anode in the horizontal direction, and increased from top toottom in the vertical direction in T4. This indicated that H+ andH− migration path was not a simple line due to the 2D crossedeld in use.

Fig. 6. The variations of ORP profiles along the cell after treatment in T4. Top: soilwith depth of 0–45 mm; mid: soil with depth of 45–80 mm; bottom: soil with depth80–100 mm in EK treatment zone.

Generally, as the potential is greater than 200 mV, the system isconsidered as slightly oxidative environment, and when it is below100 mV, it indicated highly reducing [22]. Seen from Fig. 6, slightlyoxidizing condition exists only in the soil very next to both horizon-tal and vertical anode, while highly reductive environment prevailsthroughout the soil specimen. It is expected that Cr(VI) reduced toCr(III) and affects the remediation efficiency of total Cr. This resultwas in agreement with previous studies [23].

EC increased from the cathode to the anode in T2, with a maxi-mum of 1470 �S/cm near the anode and a minimum of 730 �S/cmnear the cathode. The EC distribution was layered in contaminatedsoil in T3, with a maximum at 3060 �S/cm, and a minimum at256 �S/cm. This indicated that Cr(VI) ions had migrated from thebottom to the top of the EK treatment zone due to the verticalpotential in T3. And this also explained only a little Cr(VI) couldmigrate to deep soil layer in Fig. 4b. In T4, EC increased from thecathode to the anode in the horizontal direction, and decreasedfrom top to bottom in the vertical direction.

In addition, anions were affected by potential, gravity, thehydraulic gradient, and the concentration gradient. It is hypothe-sized that the track of the anions is an upward parabola, as shown inFig. 7. The changes in soil pH and EC in T4 could be used as evidenceto the hypothesis.

3.4. Soil moisture content after electrokinetic treatment

Fig. 7. Analysis of main effect of anion in 2D crossed potential. F1 and F2 are potentialeffects, F3 is the hydraulic gradient, F4 is the concentration gradient, and G is gravity.

ous Materials 177 (2010) 1126–1133 1131

i[teccafite

3t

atbtcTaCCIcmlCenottattpu

allndtideie

ieC(p53(tfotoi

P. Zhang et al. / Journal of Hazard

ng EK treatment in agreement with previously published findings24]. In addition, the soil moisture content increased from anodeo soil in T1. The current theory of the mechanisms responsible forlectroosmosis [25] holds that the electric field causes positivelyharged ions in this double layer to migrate toward the negativelyharged electrode. However, this variation in T4 was not displayeds T1. Gas which was generated near the cathode of vertical electriceld due to electrolysis may be changed the soil channel. The elec-roosmotic flow could be affected by gas generation and moisturevaporation in T4.

.5. Distribution of Cr(VI) and total Cr in soil after electrokineticreatment

Fig. 8 shows the distribution of Cr(VI) in EK treatment zonefter treatment. In T1 without an electric field, the Cr(VI) concen-ration in each section was lower than the initial concentrationecause some of the Cr(VI) ions had diffused to deeper layers andhe electrode chambers. The Cr(VI) concentration increased fromathode to anode, with a maximum value of 1101.36 mg/kg in T2.his indicated that Cr(VI) ions had migrated to the anode and hadssembled near the anode above the initial values. However, somer(VI) had migrated into deep soil, which has shown in Fig. 4. Andr(VI) could be adsorbed to soil near the anode due to low pH [26].

n addition some Cr(VI) had been reduced to Cr(III) in soil as dis-ussed in the soil ORP section, and a large number of Cr(VI) hadigrated into the anode chamber from soil. These resulted in a

ower than expected concentration of Cr(VI). A large number ofr(VI) had been transported from the bottom to the top by verticallectric field in T3. And in the top soil layer, Cr(VI) concentrationsear the electrode chambers were lower than that in the middlef the top layer due to Cr(VI) diffusion to the chambers. In addi-ion, Cr(VI) had been transported to the upper right quadrant ofhe EK treatment zone which was near both horizontal and verticalnodes in T4. Cr(VI) concentration of the top layer rose first andhen declined along the horizontal direction, with the peak leaningoward the anode. These also indicated that Cr(VI) infiltration wasrevented and that the track of Cr(VI) was a parabola that leanedpwards.

Fig. 9 shows the distribution of total Cr in non-treatment zonefter each test. Concentration of total Cr in the top layer wasarger than initial in T1. And the content of total Cr in the topayer increased significantly in T2. The maximum value reachedear 510 mg/kg. These indicated that chromium could migrate intoeep soil during electrokinetic remediation with solely horizon-al electric field again. And this also explained that the decreasen Cr(VI) in EK treatment zone was due to the Cr(VI) migrationownward beside the Cr(VI) migration into anode chamber. How-ver, concentration of total Cr in each layer was smaller thannitial in T3 and T4 because of the vertical electric field barrierffect.

Remediation efficiency of total Cr in EK treatment zone is shownn Fig. 10. Total Cr decreased about 20% due to diffusion withoutlectric potential in T1. However, the remediation efficiency of totalr was from 20.7% to 60.2% in T2 with a horizontal potential. In T3Fig. 10b), the total Cr in the top layer increased with the verticalotential, but the remediation efficiency of the middle layer was3.3–62.2%. Remediation efficiency of total Cr ranged from 7.3% to0.6% in the top layer, and 43% to 70.4% in the middle layer in T4Fig. 10c). It could be deduced that an amount of Cr(VI) was reducedo Cr(III) in terms of the ORP and residual Cr(VI)/total Cr. This was a

actor which affected the remediation efficiency of total Cr. Becausef the low soil pH near the anode regions, Cr(VI) was adsorbed ontohe soil surfaces [26]. This also affected the remediation efficiencyf total Cr. The total Cr in the bottom layer showed a greater increasen T4 (Fig. 10c). The bottom layer which acted as a buffer was uncon-

Fig. 8. Distribution of Cr(VI) in EK treatment zone after treatment in (a) T1 and T2, (b)T3 and (d) T4. Top: soil with depth of 0–45 mm; mid: soil with depth of 45–80 mm;bottom: soil with depth 80–100 mm in EK treatment zone.

taminated soil in each test. Some contaminated soils were mixedinto the bottom layer during soil filling process. In addition, Cr(III)from the upper layer migrated downward to the cathode and Cr(VI)infiltrated along with water. These factors resulted in pollution ofthis layer.

3.6. Chemical form analysis

A Tessier sequential extraction procedure was used to deter-mine the speciation of the metal forms in the original contaminatedsoil [27]. The soil had been extracted into five forms: exchangeable

form (Ex-), carbonates-bound (Car-), Fe–Mn oxides-bound (Fe–Mn-), organic-bound (Org-), and residual form (Re-).

As shown in Table 3, residual form played an important role. Indecreasing order, the forms were Re- > Fe–Mn- > Org- > EX- > Car- in

1132 P. Zhang et al. / Journal of Hazardous Materials 177 (2010) 1126–1133

Fig. 9. Distribution of total Cr in non-treatment zone after treatment in (a) T1, (b)T2, (c) T3 and (d) T4. Top, mid and bottom at depths of 140–160 mm, 220–260 mmand 340–360 mm from the top of the main reactor.

Fig. 10. Distribution of total Cr in EK treatment zone after treatment in (a) T1 and T2,

(b) T3 and (c) T4. C is the remaining total Cr concentration in the contaminated soil,C0 is the initial concentration of Cr in the contaminated soil in both top and middlelayers, and the initial concentration of Cr in uncontaminated soil in the bottom layer.

the original contaminated soil. The EX- and Car- are the forms mostlikely to be released to the environment due to their unstable bind-ing to soil particle. Fe–Mn- and Org- forms may release Cr due topH and ORP changes. Re- varies little because the metal ions forman interior crystal lattice resulting in one whole complex. Manyresearch reported that Cr contaminated soil could be remediatedand had better remediation efficiency by EK [10,23,28,29]. How-ever, the soils in these researches were artificial contaminated soil.From chemical form analysis, the aged chromium contaminatedsoil has a large percentage of Re- in this study. It is the important

factor that resulted in a low remediation efficiency of total Cr inthis experiment.

Table 3Percentage of chemical forms in original contaminated soil.

Step Form Percentage (%)

1 Exchangeable (Ex-) 0.35–3.332 Carbonates-bound (Car-) 2.5–4.53 Fe–Mn oxides-bound (Fe–Mn-) 23.13–274 Organic-bound (Org-) 10.32–13.55 Residual (Re-) 43.15–51.27

ous M

4

mdcictattc

oocwtet

A

C

R

[

[

[

[

[

[

[

[

[

[

[

[

[[

[

[

[

P. Zhang et al. / Journal of Hazard

. Conclusions

EK is an efficient treatment for remediation of soils with heavyetal contamination. However, heavy metal ions will infiltrate

ownward if EK is applied with a horizontal electric field. The 2Drossed electric field electrokinetic remediation technology wasntroduced to improve the EK. And the present method differs fromonventional horizontal EK in that both horizontal and vertical elec-ric fields were applied. When the 2D crossed electric field waspplied, Cr contaminated soil was remediated and the Cr(VI) migra-ion into deep soil was prevented simultaneously. It is expected thathe track of Cr(VI) may be parabolic and leaned upwards if the 2Drossed electric field is applied.

In practical engineering, the direction of the vertical electric fieldf the 2D crossed electric field could be easily modified in termsf the polarity of contamination. And it is necessary to establish ariterion for soil remediation to assess the remediation efficiency,hich depends on contamination species, contamination seques-

ration, bio-toxicity and bioavailability. Overall, the 2D crossedlectrical field appears to be a promising and practical method forhe remediation of soils contaminated with heavy metals.

cknowledgement

This work was supported by Natural Science Foundation ofhina (20307009)

eferences

[1] S. Lei, Chromium slag treatment and utilization, Chin. J. Resour. Compr. Util. 10(2004) 5–8 (in Chinese).

[2] Y.B. Acar, R.J. Gale, A.N. Alshawabkeh, Electrokinetic remediation: basics andtechnology status, J. Hazard. Mater. 40 (1995) 117–137.

[3] A.N. Alshawabkeh, A.T. Yeung, M.R. Bricka, Practical aspects of in-situ electroki-netic extraction, J. Environ. Eng. 125 (1999) 27–35.

[4] M.M. Page, C.L. Page, Electroremediation of contaminated soils, J. Environ. Eng.128 (2002) 208–219.

[5] J. Virkutyte, M. Sillanpaa, P. Latostenmaa, Electrokinetic soil remediation-

critical overview, Sci. Total Environ. 289 (2002) 97–121.

[6] J. Hamed, Y.B. Acar, R.J. Gale, Pb(II) removal from kaolinite by electrokinetics, J.Geotech. Eng. 117 (1991) 241–271.

[7] D.A. Reynolds, E.H. Jones, M. Gillen, et al., Electrokinetic migration of per-manganate through low-permeability media, Ground Water 46 (2008) 629–637.

[

[

[

aterials 177 (2010) 1126–1133 1133

[8] H. Chen, D. Zhou, L. Cang, Influence of vertical electric field on migration ofheavy metal complexes in soil column, Chin. J. Environ. Sci. 26 (2006) 78–82(in Chinese).

[9] H. Chen, D. Zhou, L. Cang, et al., Effects of vertical electric field and EDTA applica-tion on ryegrass copper and zinc uptake and their leaching risks, Acta Pedolog.Sin. 44 (2007) 174–178 (in Chinese).

10] J.Y. Wang, X.J. Huang, C.M. Kao, et al., Removal of heavy metals from kaolin usingan upward electrokinetic soil remedial (UESR) technology, J. Hazard. Mater. 136(2006) 532–541.

11] S.A. Ho, P.W. Sheridan, C.J. Athmer, et al., Integrated in situ remediation tech-nology: the Lasagna process, Environ. Sci. Technol. 29 (1995) 2528–2534.

12] S.A. Ho, P.W. Sheridan, B.M. Hughes, et al., The Lasagna technology for in situsoil remediation, Environ. Sci. Technol. 33 (1999) 1086–1091.

13] R. Zhang, H. Sun, Remediation of chromate contaminated soils by combinedtechnology of electrokinetic and iron PRB, Chin. J. Environ. Sci. 28 (2007)1131–1136 (in Chinese).

14] R.K. Lu, Analytic Methods for Soil Agricultural Chemistry, China AgricultureScience and Technology Press, Beijing, 2000 (in Chinese).

15] M.S. Fishman, L.C. Fridman (Eds.), Methods for Determination of Inorganic Sub-stances in Water and Fluvial Sediments, U.S. Geological Survey, TWRI Book 5,1989, p. 626 (Chapter A1).

16] S. Yu, G. Zhang, L. Qi, Determination of hexavalent chromium in some contam-inated soil, Chin. J. Arid. Environ. Monit. 10 (1996) 207–208 (in Chinese).

17] R.A. Griffin, A.K. Au, R.R. Frost, Effect of pH on adsorption of chromium fromlandfill leachate by clay minerals, J. Environ. Sci. Health A 12 (1977) 431–449.

18] Y.X. Chen, Z.Y. He, J.P. Wu, Form and transformation of chromium species andin soils, Environ. Sci. 15 (1994) 53–57 (in Chinese).

19] D.M. Zhou, H.F. Chen, L. Cang, et al., Ryegrass uptake of soil Cu/Zn inducedby EDTA/EDDS together with a vertical direct-current electrical field, Chemo-sphere 67 (2007) 1671–1676.

20] G.N. Kim, B.I. Yang, W.K. Choi, et al., Development of vertical electrokinetic-flushing decontamination technology to remove 60Co and 137Cs from a Koreannuclear facility site, Sep. Purif. Technol. 68 (2009) 222–226.

21] V.B. Acar, A. Alshawabkeh, Principles of electrokinetic remediation, Environ.Sci. Technol. 27 (1993) 2638–2647.

22] G. Sposito, Chemistry of Soils, Oxford University Press, New York, 1989.23] C. Weng, Y. Lin, T.Y. Lin, et al., Enhancement of electrokinetic remediation

of hyper-Cr(VI) contaminated clay by zero-valent iron, J. Hazard. Mater. 149(2007) 292–302.

24] J. Ma, H. Wang, Q. Luo, Movement-adsorption and its mechanism of Cd in soilunder combining effects of electrokinetics and a new type of bamboo charcoal,Environ. Sci. 28 (2007) 1829–1834 (in Chinese).

25] A.P. Shapio, R.F. Probstein, Removal of contaminants from saturated clay byelectroosmosis, Environ. Sci. Technol. 27 (1993) 283–291.

26] S. Chinthamreddy, K.R. Reddy, Oxidation and mobility of trivalent chromium inmanganese-enriched clays during electrokinetic remediation, J. Soil Contam. 8(1999) 97–216.

27] A. Tessier, P.B.C. Campbell, M. Bisson, Sequential extraction procedure for thespeciation of trace metals, Anal. Chem. 51 (1979) 844–850.

28] A. Sawadaa, K. Mori, S. Tanaka, et al., Removal of Cr(VI) from contaminated soilby electrokinetic remediation, Waste Manag. 24 (2004) 483–490.

29] K.R. Reddy, S. Chinthamreddy, Effects of initial form of chromium on electroki-netic remediation in clays, Adv. Environ. Res. 7 (2003) 353–365.