Electrokinectic soil remediation

Uncorrected Proof
ARTICLE IN PRESS
1
2
3
Ž.
The Science of the Total Environment xxx 2001 xxx᎐xxx
4
5
6
7
8
910
Electrokinetic soil remediation ᎏ critical overview
11
12
Jurate Virkutyte
a,
U
, Mika Sillanpaa
a
, Petri Latostenmaa
b
¨¨
13
14
a
Uni¨ersity of Oulu, Water Resources and En¨ironmental Engineering Laboratory, Tutkijantie 1 F 2, 90570 Oulu, Finland15
b
¨
Finnish Chemicals Oy, P.O. Box 7, FIN-32741 Aetsa, Finland

¨
16
17
Received 28 May 2001; accepted 31 August 200118
19
20
21
Abstract22
23
In recent years, there has been increasing interest in finding new and innovative solutions for the efficient removal24
of contaminants from soils to solve groundwater, as well as soil, pollution. The objective of this review is to examine25
several alternative soil-remediating technologies, with respect to heavy metal remediation, pointing out their26
strengths and drawbacks and placing an emphasis on electrokinetic soil remediation technology. In addition, the27
review presents detailed theoretical aspects, design and operational considerations of electrokinetic soil-remediation28
variables, which are most important in efficient process application, as well as the advantages over other technologies29
and obstacles to overcome. The review discusses possibilities of removing selected heavy metal contaminants from30
clay and sandy soils, both saturated and unsaturated. It also gives selected efficiency rates for heavy metal removal,31
the dependence of these rates on soil variables, and operational conditions, as well as a cost᎐benefit analysis. Finally,32
several emerging in situ electrokinetic soil remediation technologies, such as Lasagna
TM
, Elektro-Klean
TM
, elec-33
trobioremediation, etc., are reviewed, and their advantages, disadvantages and possibilities in full-scale commercial34
applications are examined. ᮊ 2001 Published by Elsevier Science B.V.
35
36
Keywords: Electrokinetic soil remediation; Heavy metals37
38
39

1. Introduction
40
41
Every year, millions of tonnes of hazardous
42
waste are generated in the world. Due to ineffi-43
cient waste handling techniques and hazardous
44
waste leakage in the past, thousands of sites were
45
46
47
4849
U
Corresponding author.50
.
51
contaminated by heavy metals, organic com-
52
pounds and other hazardous materials, which 53
made an enormous impact on the quality of
54
groundwater, soil and associated ecosystems. Dur- 55
ing the past decades, several new and innovative
56
solutions for efficient contaminant removal from 57
soils have been investigated and it is strongly
58
believed that they will help to solve groundwater 59
and soil pollution. Despite numerous promising 60

laboratory experiments, there are not many suc-
61
0048-9697r01r$ - see front matter ᮊ 2001 Published by Elsevier Science B.V.62
Ž.
PII: S 0 0 4 8 - 9 6 9 7 0 1 01027-0
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. rThe Science of the Total En
¨ironment xxx 2001 xxx᎐xxx2
63
cessfully implemented in situ soil-treatment tech-64
niques yet. Because of uncertainty, lack of ap-65
propriate methodology and proven results, many66
in situ projects are currently under way. It is67
likely that there will not be a single universal in68
situ soil-treatment technology. Instead, quite a69
large variety of technologies and their combina-70
tions suitable for different soil remediation situa-71
tions will be developed and implemented.72
Although the successful and environmentally73
friendly soil treatment technologies have not been74
completely investigated and implemented, there75
are several techniques which have attracted in-76
creased interest among scientists and industry77
officials. These are:78
7980
ⅷ
Bioremediation ᎏ despite a demonstrated81
ability to remove halogenated and non-82

halogenated volatiles and semi-volatiles, as
83
well as pesticides, this technique has failed to84
show efficient results in removing heavy met-85
als from contaminated soils.8687
ⅷ
Thermal desorption ᎏ this treats halogenated88
and non-halogenated volatiles and semi-vola-89
tiles, as well as fuel hydrocarbons and pesti-90
cides. It has failed to demonstrate an ability to91
remove heavy metals from contaminated soils.9293
ⅷ
Soil vapour extraction ᎏ there are several94
promising results in reducing the volume of95
treated heavy metals. Nevertheless, this tech-96
nique cannot reduce their toxicity.9798
ⅷ
Soil washing ᎏ this technique has demon-99
strated potential effectiveness in treating100
heavy metals in the soil matrix.101102
ⅷ
Soil flushing ᎏ according to laboratory-scale103
experiments, this is efficient in removing heavy104
metals from soils, despite the fact that it can-105
not reduce their toxicity.
106107
ⅷ
Electrokinetic soil remediation.108
109
As none of the other in situ soil remediation

110
techniques has demonstrated the efficient re-111
moval of heavy metals, there was a necessity to
112
develop other methods to remediate soil contami-113
nated by heavy metals.
114
Electrokinetic soil remediation is an emerging115
technology that has attracted increased interest116
among scientists and governmental officials in the
117
last decade, due to several promising laboratory 118
and pilot-scale studies and experiments. This 119
method aims to remove heavy metal contami- 120
nants from low permeability contaminated soils 121
under the influence of an applied direct current. 122
However, regardless of promising results, this 123
method has its own drawbacks. First of all, the 124
whole electrokinetic remediation process is highly 125
dependant on acidic conditions during the appli- 126
cation, which favours the release of the heavy 127
metal contaminants into the solution phase. How- 128
ever, achieving these acidic conditions might be 129
difficult when the soil buffering capacity is high. 130
In addition, acidification of soils may not be an 131
environmentally acceptable method. Second, the 132
remediation process is a very time-consuming ap- 133
plication; the overall application time may vary 134
from several days to even a few years. There are 135
some other limitations of the proposed technique 136

that need to be overcome: i.e. the solubility of the 137
contaminant and its desorption from the soil ma- 138
trix; low target ion concentration and high non- 139
target ion concentration; requirement of a con- 140
ducting pore fluid to mobilise contaminants; and 141
heterogeneity or anomalies found at sites, such as 142
large quantities of iron or iron oxides, large rocks 143
Ž.
or gravel, etc. Sogorka et al., 1998 . 144
According to the experiments and pilot-scale 145
studies conducted, metals such as lead, chromium, 146
cadmium, copper, uranium, mercury and zinc, as 147
well as polychlorinated biphenyls, phenols, 148
chlorophenols, toluene, trichlorethane and acetic 149
acid, are suitable for electrokinetic remediation 150
and recovery. 151
152
2. Theoretical, design and operational
153
considerations 154
155
2.1. Theoretical aspects 156
157
The first electrokinetic phenomenon was 158
observed at the beginning of the 19th Century,
159
when Reuss applied a direct current to a 160
Ž
clay᎐water mixture Acar and Alshawabkeh,
161

.
1993 . However, Helmholtz and Smoluchowski 162
were the first scientists to propose a theory deal- 163
ing with the electroosmotic velocity of a fluid and
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 3
164
the zeta potential under an imposed electric gra-165
Ž.Ž .
dient ␨ Acar and Alshawabkeh, 1993 . Sibel166
Pamukcu and her research group have derived167
the following Helmholtz᎐Smoluchowski equation:
168
␧␨ Ѩ␾
Ž.
u s 1
EO
␮Ѩx
169
where u is the electroosmotic velocity, ␧ is the
EO170
dielectric constant of the pore fluid, ␮ is the171
viscosity of the fluid and Ѩ␾rѨx is the electric172
gradient.173
When DC electric fields are applied to con-174
taminated soil via electrodes placed into the175
ground, migration of charged ions occurs. Positive176

ions are attracted to the negatively charged cath-177
ode, and negative ions move to the positively178
charged anode. It has been experimentally proved179
that non-ionic species are transported along with180
the electroosmosis-induced water flow. The direc-181
tion and quantity of contaminant movement is182
influenced by the contaminant concentration, soil
183
type and structure, and the mobility of contami-184
nant ions, as well as the interfacial chemistry and185
the conductivity of the soil pore water. Electroki-186
netic remediation is possible in both saturated187
and unsaturated soils.188
Electrokinetic soil treatment relies on several189
interacting mechanisms, including advection,190
which is generated by electroosmotic flow and191
externally applied hydraulic gradients, diffusion192
of the acid front to the cathode, and the migra-193
tion of cations and anions towards the respective194
Ž.
electrode Zelina and Rusling, 1999 . The domi-
195
nant and most important electron transfer reac-196
tions that occur at electrodes during the elec-197
trokinetic process is the electrolysis of water:
198
q
Ž.
y
HOª 2H q1r2Ogq2e

22199
200
yy
Ž. Ž.
2H O q2e ª 2OH qHg 2
22
201
The acid front is carried towards the cathode202
by electrical migration, diffusion and advection.
203
The hydrogen ions produced decrease the pH204
near the anode. At the same time, an increase in205
the hydroxide ion concentration causes an in-206
crease in the pH near the cathode. In order to
207
solubilise the metal hydroxides and carbonates 208
formed, or different species adsorbed onto soils 209
particles, as well as protonate organic functional 210
groups, there is a necessity to introduce acid into 211
the soil. However, this acid addition has some 212
major drawbacks, which greatly influence the ef- 213
ficiency of the treatment process. The addition of 214
acid leads to heavy acidification of the contami- 215
nated soil, and there is no well-established method 216
for determining the time required for the system 217
to regain equilibrium. 218
The main goal of electrokinetic remediation is 219
to effect the migration of subsurface contami- 220
nants in an imposed electric field via electro- 221
osmosis, electromigration and electrophoresis. 222

These three phenomena can be summarised as 223
follows: 224
225226
ⅷ
Electroosmosis is the movement of soil mois- 227
ture or groundwater from the anode to the
228
cathode of an electrolytic cell. 229230
ⅷ
Electromigration is the transport of ions and 231
ion complexes to the electrode of opposite
232
charge. 233234
ⅷ
Electrophoresis is the transport of charged 235
particles or colloids under the influence of an
236
electric field; contaminants bound to mobile
237
particulate matter can be transported in this 238
manner. 239
240
The phenomena occur when the soil is charged 241
with low-voltage direct current. The process might 242
be enhanced through the use of surfactants or 243
reagents to increase the contaminant removal 244
rates at the electrodes. Upon their migration to 245
the electrodes, the contaminants may be removed 246
by electroplating, precipitationrco-precipitation, 247
pumping near the electrode, or complexing with 248

ion exchange resins. 249
Electromigration takes place when highly solu-
250
ble ionised inorganic species, including metal 251
cations, chlorides, nitrates and phosphates, are 252
present in moist soil environments. Electrokinetic 253
remediation of soils is a unique method, because 254
it can remediate even low-permeability soils.
255
Other mechanisms that greatly affect the elec- 256
trochemical remediation process are electroosmo- 257
sis, coupled with sorption, precipitation and disso-
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx4
258
Ž.
lution reactions van Cauwenberghe, 1997 . This259
is the reason why all the appropriate processes260
should be taken into consideration and investi-261
gated before implementation of the technique262
can take place.263
Once the remediation process is over, extrac-264
tion and removal of heavy metal contaminants265
are accomplished by electroplating at the elec-266
trode, precipitation or co-precipitation at the267
electrode, pumping water near the electrode, or268
complexing with ion exchange resins. Adsorption269

onto the electrode may also be feasible, as some270
ionic species will change their valency near the271
Ž.
electrode depending on the soil pH , making
272
Ž
them more likely to adsorb van Cauwenberghe,273
.
1997 .
274
Prediction of THE decontamination time is of275
great importance in order to estimate possible276
power consumption and to avoid the occurrence277
of reverse electroosmotic flow, i.e. from the cath-278
Ž
ode to the anode, during the process Baraud et279
.
al., 1997, 1998 . The phenomenon of reverse elec-
280
troosmotic flow is not well understood and should281
be further investigated.282
Decontamination velocity depends on two283
Ž.
parameters Baraud et al., 1997, 1998 :
284
285286
ⅷ
Contaminant concentration in the soil solu-287
tion, which is related to the various possible288
Ž

solidrliquid interactions adsorptionrdesorp-
289
tion, complexation, precipitation, dissolution,290
.
etc. and to the speciation of the target species.291292
ⅷ
Velocity in the pore solution when species are293
in the soil solution and not engaged in any294
reactions or interactions. The velocity depends295
Ž
on different driving forces electric potential296
gradient, hydraulic head differences and con-
297
.
centration gradient and is not closely related298
to soil properties, except for the electroosmo-299
sis phenomenon.300
301
The success of electrochemical remediation de-302
pends on the specific conditions encountered in
303
the field, including the types and amount of con-304
taminant present, soil type, pH and organic con-
305
Ž.
tent Acar and Alshawabkeh, 1993 .306
For in situ conditions, the contaminated site307
itself and the immersed electrodes form a type of
308
electrolytic cell. Usually, the electrokinetic cell 309

design in laboratory experiments consists of an 310
open-flow arrangement at the electrodes, which 311
permits injection of the processing fluid into the 312
porous medium, with later removal of the con- 313
Ž
taminated fluid Sogorka et al., 1998; Reddy and 314
Chinthamreddy, 1999; Reddy et al., 1997, 1999; 315
.
Zelina and Rusling, 1999 . 316
It seems that there is a controversy as to where 317
electrodes should be placed to obtain the most 318
reliable and efficient results. It is obvious that 319
imposition of an electrical gradient by having 320
inert electrodes results in electroosmotic flow to 321
the cathode. Many authors propose that position- 322
ing of the electrodes directly into the wet soil 323
Ž
mass produces the most desirable effect Sims, 324
1990; Acar and Alshawabkeh, 1993; Reddy et al., 325
.
1999; Sogorka et al., 1998 . Through seeking im- 326
provements in experiments, some researchers tend 327
to place the electrodes not directly into the wet 328
soil mass, but into an electrolyte solution, at-
329
tached to the contaminated soil, or else to use 330
Ž
different membranes and other materials van 331
Cauwenberghe, 1997; Baraud et al., 1998; Bena-
332

.
zon, 1999 . In order to maintain appropriate
333
process conditions, a cleaning agent or clean wa- 334
ter may be injected continuously at the anode. 335
Thus, contaminated water can be removed at the 336
cathode. Contaminants at the cathode may be 337
removed by electrodeposition, precipitation or ion 338
exchange. 339
Electrodes that are inert to anodic dissolution 340
should be used during the remediation process. 341
The most suitable electrodes used for research 342
purposes include graphite, platinum, gold and sil- 343
ver. However, for pilot studies, it is more ap- 344
propriate to use much cheaper, although reliable, 345
titanium, stainless steel, or even plastic elec-
346
trodes. Using inert electrodes, the electrode reac- 347
tions will produce H
q
ions and oxygen gas at the 348
anode and OH
y
ions and hydrogen gas at the 349
cathode, which means that if pH is not controlled,
350
an acid front will be propagated into the soil 351
pores from the anode and a base front will move
352
out from the cathode. 353

It has been proved by experiments that when 354
heavy metals enter into basic conditions, they 355
adsorb to soil particles or precipitate as hydrox-
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 5
356
ides, oxyhydroxides, etc., and in acidic conditions,357
those ions desorb, solubilise and migrate.
358
Another important parameter in the electroki-
359
netic soil-remediation technique is the conductiv-
360
ity, since this, together with soil and pore fluid,361
affects the electroosmotic flow rate.
362
The conductivity of soil depends on the concen-
363
tration and the mobility of the ions present, i.e.364
contaminant removal efficiencies decrease with a
365
Ž
reduction in contaminant concentration Reddy
366
et al., 1997, 1999; Reddy and Chinthamreddy,
367
.

1999; Zelina and Rusling, 1999 . This is due to
368
hydrogen ion exchange with cationic contami-
369
nants on the soil surface, with release of the
370
contaminants. As the contaminant is removed,371
the hydrogen ion concentration in the pore fluid
372
increases, resulting in an increasing fraction of
373
the current being carried by the hydrogen ions374
rather than by the cationic contaminants.
375
It is possible to conclude that the variables
376
which have impact on the efficiency of removing377
contaminants from soils are:
378
379
380
ⅷ
Chemical processes at the electrodes;
381382
ⅷ
Water content of the soil;
383384
ⅷ
Soil type and structure;
385386

ⅷ
Saturation of the soil;
387388
ⅷ
pH and pH gradients;
389
390
ⅷ
Type and concentration of chemicals in the 391
soil;
392393
ⅷ
Applied current density; and
394395
ⅷ
Sample conditioning. 396
397
In addition, insoluble organics, such as heavy 398
hydrocarbons, are essentially not ionised, and the 399
soils in contact with them are not charged. The 400
removal of insoluble organics by electric field is 401
limited to their movement out of the soil by 402
electroosmotic purging of the liquid, either with 403
water and surfactant to solubilise the compounds, 404
or by pushing the compounds ahead of a water 405
Ž.
front Probstein and Hicks, 1993 . 406
Ionic migration is the movement of ions sub- 407
jected to an applied DC electric field. Electromi- 408
Ž

gration rates in the subsurface depend upon van 409
.
Cauwenberghe, 1997 : 410
411412
ⅷ
Soil porewater current density;
413414
ⅷ
Grain size;
415416
ⅷ
Ionic mobility;
417418
ⅷ
Contaminant concentration; and
419420
ⅷ
Total ionic concentration. 421
422
The process efficiency is not as dependent on
423
the fluid permeability of soil as it is on the pore-
424
water electrical conductivity and path length
425
Ž.
Fig. 1. Electroosmosis and electromigration of ions adapted from Acar et al., 1994, 1996; Acar and Alshawabkeh, 1996 .
Uncorrected Proof
ARTICLE IN PRESS
()

J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx6
426
through the soil, both of which are a function of
427
the soil moisture content. As electromigration
428
does not depend on the pore size, it is equally
429
Ž
applicable to coarse and fine-grained soils van430
.
Cauwenberghe, 1997 .
431
Electroosmosis in water-saturated soil is the
432
movement of water relative to the soil under the
433
influence of an imposed electric gradient. When
434
there is direct current applied across the porous
435
media filled with liquid, the liquid moves relative
436
to the stationary charged solid surface. When the
437
surface is negatively charged, liquid flows to the
438
Ž.
cathode. Acar et al. 1994, 1996 have conducted

439
numerous experiments and found that this process
440
Ž.
works well in wet i.e. water-saturated fine-
441
grained soils and can be used to remove soluble
442
pollutants, even if they are not ionic. The dis-
443
solved neutral molecules simply go with the flow.
444
Fig. 1 shows a schematic representation of this445
process.
446
An excess negative surface charge exists in all
447
kinds of soil. For example, many clays are anionic,
448
colloidal poly-electrolytes. The surface charge
449
density increases in the following order: sand-
450
silt - kaolinite - illite - montmorillonite. Injec-
451
tion of clean fluid, or simply clean water, at the
452
anode can improve the efficiency of pollutant
453
removal. For example, such a flushing technique

454
using electroosmosis has been developed for the
455
removal of benzene, toluene, trichlorethane and456
m-xylene from saturated clay.
457
According to that stated above, the main fac-
458
tors affecting the electroosmotic transport of con-459
taminants in the soil system are as follows:460
461
462
ⅷ
Mobility and hydration of the ions and charged
463
particles within the soil moisture;
464465
ⅷ
Ion concentration;
466
467
ⅷ
Dielectric constant, depending on the amount
468
of organic and inorganic particles in the pore469
solution; and
470471
ⅷ
Temperature.472
473

Most soil particle surfaces are negatively
474
charged as a result of isomorphous substitution
475
Ž
and the presence of broken bonds Yeung et al.,476
.
1997 .
477
Experiments have determined the dependence
478
of the zeta potential of most charged particles on
479
solution pH, ionic strength, types of ionic species,
480
Ž
temperature and type of clay minerals Vane and
481
.
Zang, 1997 . For water-saturated silts and clays,
482
the zeta potential is typically negative, with values 483
measured in the 10᎐100-mV range.
484
However, if ions produced in the electrolysis of
485
water are not removed or neutralised, they lower
486
the pH at the anode and increase it at the cath-
487

ode, accompanied by the propagation of an acid
488
front into the soil pores from the anode and a
489
base front from the cathode. This process can
490
Ž
significantly effect the soil zeta potential drop in
491
.
zeta potential , as well as the solubility, ionic state
492
and charge, level of adsorption of the contami- 493
Ž.
nant, etc. Yeung et al., 1997 .
494
In addition, different initial metal concentra-
495
tions and sorption capacity of the soil may pro-
496
duce soil surfaces that are less negative, which at
497
the same time may become positive at a pH of
498
approximately the original zero-point charge
499
Ž.
Yeung et al., 1997 . Similarly, chemisorption of
500
anions makes the surface more negative.

501
Electroosmotic flow from the anode to the
502
cathode promotes the development of a low-pH
503
environment in the soil. This low-pH environment
504
inhibits most metallic contaminants from being
505
sorbed onto soil particle surfaces and favours the 506
formation of soluble compounds. Thus, electro-
507
osmotic flow from the anode to cathode, resulting
508
from the existence of a negative zeta potential,
509
enables the removal of heavy metal contaminants 510
by the electrokinetic remediation process.
511
The pH of the soil should be maintained low
512
enough to keep all contaminants in the dissolved
513
phase. Nevertheless, when the pH becomes too
514
low, the polarity of the zeta potential changes and
515
Ž
reversed electroosmotic flow i.e. from the cath-
516

.
ode to the anode may occur. In order to achieve
517
efficient results in removing contaminants from
518
soils, it is necessary to maintain a pH low enough
519
pH to keep metal contaminants in the dissolved
520
phase and high enough to maintain a negative
521
Ž.
zeta potential Yeung et al., 1997 . Despite this
522
apparently easily implemented theory, simultane-
523
ous maintenance of a negative zeta potential and
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 7
524
dissolved metal contaminants remains the great-525
est obstacle in the successful implementation of526
the electrokinetic soil remediation process.527
528
2.2. Design considerations529
530
In order to obtain efficient and reliable results,531

electrokinetic remediation of soil should be im-532
plemented under steady-state conditions. It is533
obvious that during the remediation process, other534
reactions, such as transport and sorption, and535
precipitation and dissolution reactions, occur and536
affect the remediation process.537
There have been numerous indications of the538
importance of heat and gas generation at elec-539
trodes, the sorption of contaminants onto soil540
particle surfaces and the precipitation of contami-541
nants in the electrokinetic remediation process542
Ž
Acar and Alshawabkeh, 1993; Lageman, 1993;
543
.
Zelina and Rusling, 1999 . These processes should544
be further investigated, because it is believed that545
they may weaken the removal efficiency for heavy546
metal contaminants. It is reported that different547
physicochemical properties of the soil may influ-548
ence the removal rates of heavy metal contami-549
nants, due to changed pH values, hydrolysis, and550
oxidation and reduction reaction patterns.551
In order to enhance the electrokinetic remedia-552
tion process, several authors recommend the use553
of a multiple anode system, which is shown in Fig.554
2.555
556
2.3. Operational considerations557
558

As there are several experimental techniques559
to remediate coarse-grained soils, in situ elec-560
trokinetic treatment has been developed for con-561
taminants in low-permeability soils. Electrokinet-
562
ics is applicable in zones of low hydraulic conduc-563
tivity, particularly with a high clay content.
564
Contaminants affected by electrokinetic565
processes include:566
567568
ⅷ
Heavy metals;569570
Ž
ⅷ
Radioactive species Cs , Sr , Co , ura-
137 90 60571
.
nium ;
572573
Ž.
ⅷ
Toxic anions nitrates and sulfates ;574575
Ž.
ⅷ
Dense, non-aqueous-phase liquids DNAPLs ;
576
Fig. 2. Multiple anodes system US EPA, 1998.
577
578

ⅷ
Cyanides; 579580
Ž
ⅷ
Petroleum hydrocarbons diesel fuel, gasoline, 581
.
kerosene and lubricating oils ; 582583
ⅷ
Explosives; 584585
ⅷ
Mixed organicrionic contaminants; 586587
ⅷ
Halogenated hydrocarbons; 588589
ⅷ
Non-halogenated pollutants; and 590591
ⅷ
Polynuclear aromatic hydrocarbons. 592
593
Heavy metal interactions in the soil solution 594
Ž
are governed by several processes, such as Sims, 595
.
1990 :
596
597598
ⅷ
Inorganicrorganic complexation; 599600
ⅷ
Acid᎐base reactions; 601602
ⅷ

Redox reactions; 603604
ⅷ
Precipitationrdissolution reactions; and 605606
ⅷ
Interfacial reactions. 607
608
The choice of appropriate soil for electroki-
609
netic remediation process should be made with 610
extreme caution and possible soil pre-treatment
611
experiments should be carried out. 612
Soils that may be used for the electrokinetic
613
Ž.
remediation process should have Sims, 1990 : 614
615616
ⅷ
Low hydraulic conductivity;
617618
Ž
ⅷ
Water-soluble contaminants if there are any 619
poorly soluble contaminants, it may be essen- 620
.
tial to add solubility-enhancing reagents ; and
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En

¨ironment xxx 2001 xxx᎐xxx8
621
622
ⅷ
Relatively low concentrations of ionic materi-623
als in the water.624
625
It is reported that with applied electric fields,626
the most suitable soils for heavy metal remedia-627
Ž.
tion are kaolinite, clay and sand Sims, 1990 . As628
recommended, clay has low hydraulic conductiv-629
ity, reducing redox potential, slightly alkaline pH630
Ž
which is suitable for the remediation of several631
.
heavy metal contaminants , high cation exchange632
capacity and high plasticity. Under normal condi-633
tions, migration of ions is very slow, but is en-634
hanced by electrical fields or hydraulic pressure.635
The highest degree of removal of heavy metals636
Ž.
over 90% of the initial contaminant has been637
achieved for clayey, low-permeability soils,638
whereas for porous, high-permeability soils, such639
as peat, the degree of removal was only 65%640
Ž.
Chilingar et al., 1997 . Laboratory results showed
641
that electrokinetic purging of acetate and phenol642

from saturated kaoline clay resulted in greater643
than 94% removal of the initial contaminants.644
However, this methodology needs to be further645
investigated, because phenol has been reported to646
be toxic to humans and the environment.647
648
3. Removal of metals649
650
If heavy metal contaminants in the soil are in651
ionic forms, they are attracted by the static elec-652
trical force of negatively charged soil colloids.653
The attraction of metal ions to the soil colloids654
primarily depends on the soil electronegativity655
Ž
and the dissociation energy of ions Sah and656
.
Chen, 1998 . If there are appropriate pH condi-657
tions, heavy metals are likely to be adsorbed onto
658
the negatively charged soil particles. The main659
sorption mechanisms include adsorption andror
660
ion exchange. Desorption of cationic species from661
clay surfaces is essential in extraction of species662
from fine-grained deposits with high cation-
663
exchange capacity.664
As Acar and his research group have indicated
665
Ž

Acar and Alshawabkeh, 1993, 1996; Acar et al.,666
.
1994, 1996 , the sorption mechanisms depend on667
the surface charge density of the clay mineral, the668
characteristics and concentration of the cationic
669
species, and the presence of organic matter and 670
carbonates in the soil. The mechanism is also 671
significantly dependent on the pore fluid pH. The 672
higher the content of carbonates and organic 673
material in soils, the lower the heavy metal re- 674
moval efficiency, which is why the former should 675
be further investigated and taken into the con- 676
sideration. 677
During numerous experiments, a decrease in 678
Ž
current density was observed Acar and Al- 679
shawabkeh, 1993, 1996; Acar et al., 1994, 1996; 680
.
Sah and Chen, 1998 . The possible reasons might 681
be as follows: 682
683684
Activation polarisation: during the electroki- 685
Ž
netic remediation process, gaseous bubbles O
2 686
.
and H cover the electrodes. These bubbles
2 687
are good insulators and reduce the electrical 688

conductivity, subsequently reducing the cur- 689
rent. 690691
Resistance polarisation: after the electrokinetic 692
remediation process, a white layer was observed 693
on the cathode surface. This layer may be the 694
insoluble salt and other impurities that were 695
not only attracted to the cathode, but also 696
inhibited the conductivity, with a subsequent 697
decrease in current. 698699
Concentration polarisation: the H
q
ions gener- 700
ated at the anode are attracted to the cathode 701
and the OH
y
ions generated at the cathode 702
are attracted to the anode. If acid and alkaline 703
conditions are not neutralised, the current also 704
drops. 705
706
It is possible to conclude that soil containing 707
heavy metal contaminants influences the conduc- 708
tivity. 709
Interaction of the pollutants with the soil also
710
affects the remediation process. In order to in- 711
crease the solubility of complexes formed, or to
712
improve electromigration characteristics of speci- 713
fic heavy metal contaminants, an enhancement

714
solution may be added to the soil matrix. 715
Sometimes electroosmotic flow rates are too
716
low, and it may be necessary to flush the elec- 717
trodes with a cleaning agent, or simply clean tap 718
Ž.
water Probstein and Hicks, 1993 . In addition, 719
the electrode may be surrounded by ion-exchange
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 9
720
material to trap the contaminant and prevent its721
precipitation. It is essential to know the buffering722
capacity of the soil in order to alter the pH with723
suitable solutions or clean water. Many ground-724
waters contain high concentrations of bicarbon-725
ates, which consume added hydrogen ions to form726
carbonic acid, or hydroxyl ions to form carbonate727
ions. It is vital to draw attention to the limited728
solubility of metal carbonates, as well as the need729
for evaluation of sulfide, sulfate, chlor-730
ide and ammonia effects, which may occur when731
these compounds are introduced into the soil732
Ž
system during the remediation process Probstein733
.

and Hicks, 1993 .734
New alternatives have been suggested for the735
remediation of heavy metals from soils without
736
Ž
having low pH conditions Probstein and Hicks,
737
.
1993 . When the metal enters the region of high738
pH near the cathode, it may adsorb onto the soil,739
precipitate, or form hydroxy complexes. At higher740
pH values, the solubility increases because of the741
increasing stability of soluble hydroxy complexes.742
Despite favourable soluble complexes, the disso-743
lution process may be time-consuming and too744
slow to be successfully implemented.745
Concerning the process of transport of con-746
taminants and their derivatives, two major pheno-747
Ž.
mena were indicated Chilingar et al., 1997 :748
749750
1. The flow of contaminant solution through a751
solid matrix due to Darcy’s law and electroki-752
netics; and753754
2. Spatial redistribution of dissolved substances755
with respect to the moving liquid due to the756
diffusion and migration of charged particles.757
758
The total movement of the matter of the con-759
taminant solution in the DC electric field can be

760
expressed as the sum of four components761
Ž.
Chilingar et al., 1997 :
762
763764
ⅷ
The hydrodynamic flow of liquids driven by765
the pressure gradient;766767
ⅷ
The electrokinetic flow of fluids due to inter-768
action of the double layer with the DC field;769770
ⅷ
The diffusion of components dissolved in the771
flowing solution; and772773
ⅷ
The migration of ions inside moving fluids due
774
to the attraction of charged particles to the 775
electrodes. 776
777
The very questionable concept that removal of 778
heavy metals in the direct current field is effective 779
was also expressed, because electromigration of 780
ions is rapid and does not depend on the zeta 781
potential. In order to prove or disapprove this, 782
further investigations of this concept should be 783
carried out. Despite some disagreements, it was 784
agreed that in order to obtain efficient and reli- 785
able results and control the remediation process, 786

there is a need to provide continuous control of 787
Ž
the pH in the vicinity of the electrodes Acar and
788
Alshawabkeh, 1993, 1996; Acar et al., 1994, 1996; 789
.
Chilingar et al., 1997 . One possible way to achieve
790
this is periodic rinsing of the cathode with fresh 791
water. 792
Experiments have proved that electrical field
793
application in situ leads to an increase in temper- 794
ature, which in turn reduces the viscosity of hy- 795
Ž
drocarbon-containing fluids Chilingar et al.,
796
.
1997 . The reduction in fluid viscosity leads to an 797
increase in the total flow rate. 798
Ž.
It is reported Chilingar et al., 1997 that in 799
order to accelerate the fluid transport in situ, 800
electrical properties of soils, such as electrical 801
resistivity and the ionisation rate of the flowing 802
fluids that can affect the total rate flow, should 803
consider. In an applied DC field, some soil types 804
showed an increase in their hydraulic permeabil- 805
ity, which allows us to conclude that direct cur- 806
rent may accelerate fluid transport. However, this 807

method is not applicable to some clays, because 808
under the DC field, those clays become amor- 809
phous. It is possible to avoid such a transforma- 810
tion if interlayer clay water is trapped and is not 811
able to leave the system.
812
From the numerous laboratory and field experi- 813
ments and studies conducted, it is possible to
814
conclude that migration rates of heavy metal ions 815
Ž.
i.e. removal efficiencies are highly dependent on
816
soil moisture content, soil grain size, ionic mobil- 817
ity, pore water amount, current density and con- 818
Ž
taminant concentration Acar and Alshawabkeh,
819
1993, 1996; Acar et al., 1994, 1996; Chilingar et 820
.
al., 1997; Sah and Chen, 1998 . Also, in order to
821
assure the efficient and successful heavy metal
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx10
822
removal from soils, one of the main drawbacks of823

this process must be solved, which is premature824
precipitation of metal species close to the cathode825
compartment.
826
827
3.1. Limitations of the technique
828
829
The removal of heavy metals from soils using830
electrokinetic remediation has some limitations,831
which have been widely discussed among many832
scientists and researchers. For example, the sur-833
face of the electrode attracts the gas generated834
from the electrolytic dissociation process and in-835
creases the resistance, which significantly slows836
Ž
down the remediation process Sah and Chen,837
.
1998 . It is obvious that soil resistance is lower in838
the earlier stages of the electrokinetic process,839
and therefore a lower input voltage is required.840
When the electrokinetic process continues, gas841
bubbles from electrolytic dissociation cover the842
whole cathode surface and the resistance in-843
creases. To continue the soil remediation process,844
the input voltage must be increased to maintain
845
the same current, which also increases the voltage846
gradient. OH
y

ion that are formed react with847
cations and form a sediment, which plugs the848
spacing between soil particles, subsequently hin-849
dering the electrical current and decreasing the850
diffusive flow over time when the voltage is ap-851
Ž.
plied Sah and Chen, 1998 .
852
853
3.2. Enhancement and conditioning
854
855
To overcome the premature precipitation of856
ionic species, Acar and his research group have
857
recommended using different enhancement tech-858
niques to remove or to avoid these precipitates in859
the cathode compartment. Efficient techniques
860
should have the following characteristics:861
862863
ⅷ
The precipitate should be solubilised andror864
precipitation should be avoided.865866
ⅷ
Ionic conductivity across the specimen should
867
not increase excessively in a short period of868
time to avoid a premature decrease in the869
electroosmotic transport.

870
871
ⅷ
The cathode reaction should possibly be de-
872
polarised to avoid the generation of hydroxide 873
and its transport into the specimen.
874
875
ⅷ
Depolarisation will decrease the electrical po-
876
tential difference across the electrodes, which 877
would result in lower energy consumption.
878
879
ⅷ
If any chemical is used, the precipitate of the
880
metal with the new chemical should be per- 881
fectly soluble within the pH range attained.
882
883
ⅷ
Any special chemicals introduced should not
884
result in any increase in toxic residue in the 885
soil mass.
886
887

ⅷ
The cost efficiency of the process should be
888
maintained when the cost of enhancement is 889
included. 890
891
It is obvious that an enhancement fluid in-
892
creases the efficiency of contaminated soil treat-
893
ment; however, there is a lack of data which
894
would clarify further soil and contaminant inter- 895
actions in the presence of this fluid.
896
Ž.
As a depolariser i.e. enhancement fluid in the
897
cathode compartment, it is possible to use a low
898
Ž
concentration of hydrochloric or acetic acid Acar
899
and Alshawabkeh, 1993, 1996; Acar et al., 1994,
900
.
1996 . The main concern with hydrochloric acid
901
as the depolariser is that due to electrolysis, the
902

chlorine gas formed may reach the anode, as well
903
as groundwater, and increase its contamination.
904
Acetic acid is environmentally safe and it does
905
not fully dissociate. In addition, most acetate salts
906
are soluble, and therefore acetic acid is preferred 907
in the process.
908
The anode reaction should also be depolarised,
909
because of the dissolution and release of silica,
910
alumina and heavy metals associated with the clay
911
mineral sheets over long exposure to protons
912
Ž
Acar and Alshawabkeh, 1993, 1996; Acar et al.,
913
.
1994, 1996 .
914
In order to accomplish both tasks successfully,
915
it is better to use calcium hydroxide as the en-
916
hancement fluid to depolarise the anode reaction,

917
and hydrochloric acid as the enhancement fluid to
918
depolarise the cathode reaction.
919
The use of an enhancement fluid should be
920
Ž
examined with extreme care to prevent Yeung et 921
.
al., 1997 :
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 11
922
923
924
ⅷ
The introduction of a secondary contaminant925
into the subsurface;926927
ⅷ
The generation of waste products or by-prod-928
ucts as a result of electrochemical reactions;929
and930931
ⅷ
The injection of an inappropriate enhance-932
ment fluid that will aggravate the existing con-933
tamination problem.934

935
936
4. Electrokinetic soil remediation processes937
938
4.1. Remo¨al of hea¨y metals using cation-selecti¨e939
membrane940
941
In alkaline medium, heavy metals are likely to942
be adsorbed onto the soil particles and form943
insoluble precipitates. The high pH region in clos-944
est proximity to the cathode is the main obstacle945
Ž
to heavy metal removal Acar and Alshawabkeh,946
1993, 1996; Acar et al., 1994, 1996; Li et al., 1997;947
Li and Neretnieks, 1998; Li and Li, 2000; Yeung948
.
et al., 1997 . However, the latest experimental
949
studies show that it is possible to deal with the 950
Ž
pH impact Li et al., 1997; Li and Neretnieks, 951
.
1998; Li and Li, 2000 . A conductive solution, 952
which simulates the groundwater conditions, was 953
placed between the cathode and the soil to be 954
treated. However, the length of conductive solu- 955
tion must be at least twice the length of the 956
treated soil, which may be impossible to imple- 957
ment at a site. In addition, the solution has to be 958
placed in a special container, which would sig- 959

nificantly increase the costs of the overall remedi- 960
ation process. The pH buffer capacity, cation 961
exchange capacity of the medium, and interac- 962
tions of the solution with the soil may affect the 963
speed of the advancement of the acidic and the 964
Ž
basic fronts and the location of the pH jump Li 965
.
and Li, 2000 . In order to overcome these obsta- 966
cles, a new method was proposed which should 967
significantly improve the overall remediation 968
process. To reduce the relative length or volume 969
of the water in the system, a cation-selective 970
Ž
membrane is placed in front of the cathode Li et
971
al., 1997; Li and Neretnieks, 1998; Li and Li, 972
.Ž .
2000 Fig. 3 .
973
Ž.
Fig. 3. Electrokinetic cell with cation-selective membrane adapted from Li and Neretnieks, 1998; Li et al., 1997; Li and Li, 2000 .
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx12
974
Due to an applied electric current, ions move975
to the electrodes, according to their charges. The976

cation-selective membrane, placed between the977
soil and cathode, allows cations and very few978
anions to pass through it. This is why almost all979
the hydroxyl ions produced at the cathode remain980
on the cathodic side of the membrane. The hy-981
drogen ions generated at the anode move through982
the soil and into the membrane. The basic front983
cannot pass through the membrane, where it984
meets the acidic front. The main pH changes985
Ž
occur near the membrane Li et al., 1997; Li and986
.
Neretnieks, 1998; Li and Li, 2000 . It is possible987
that the membrane determines the pH jump and988
may control the cathode solution volume. A989
cation-selective membrane maintains the low soil
990
pH during the remediation process and signifi-991
cantly reduces the length of the conductive solu-992
tion required. Hence, the proposed electrokinetic993
cell consist of the treated soil, a conductive solu-994
tion, which is placed between the soil and the995
membrane, and the cathode compartment with996
electrolyte solution, which is between the mem-997
brane and cathode. After numerous experiments,998
it has been observed that the smaller the volume999
of conductive solution, the higher the pH will be1000
and the larger will be the leakage of the anions1001
Ž
through it Li et al., 1997; Li and Neretnieks,1002

.
1998; Li and Li, 2000 .
1003
However, a small amount of anions passing1004
through the membrane may be favourable for the1005
remediation process. Precipitation decreases the1006
remediation time, because this reduces the con-1007
Ž
centration of heavy metals in the liquid phase Li1008
.
and Li, 2000 . At the same time, back-diffusion of
1009
heavy metals is greatly reduced, since the concen-
1010
tration of heavy metals near the membrane does1011
not exceed the solubility of the metals. It has
1012
been proved by experiments that precipitation1013
decreases the electrical energy consumption, be-1014
cause the potential drop between the electrodes1015
and the remediation time are proportional to the
1016
Ž
distance between the electrodes Li et al., 1997;1017
.
Li and Neretnieks, 1998; Li and Li, 2000 .
1018
1019
4.2. Remo¨al of hea¨y metals using surfactant-coated1020
ceramic casings1021

1022
For many years, the main emphasis of elec-
1023
trokinetic soil remediation was on saturated, 1024
fine-grained soils and clays, which led to the mis- 1025
conception that electrokinetics was not suitable 1026
for unsaturated, sandy soils. Laboratory experi- 1027
ments proved that with appropriate technology 1028
and well-designed methods, it is possible to reme- 1029
diate heavy metals from unsaturated and sandy 1030
Ž.
soils Mattson and Lindgren, 1995 . The treat- 1031
ment of unsaturated soils has several limitations. 1032
The electrical conductivity of soil depends on the 1033
Ž.
moisture content Mattson and Lindgren, 1995 . 1034
During electroosmotic migration through the soil, 1035
the water content near the anode is reduced. As 1036
the moisture content decreases, the soil conduc- 1037
tivity becomes too low for the electrokinetic re- 1038
mediation application. In order to control the 1039
hydraulic flux of water in the treated soil, the use 1040
of porous ceramic castings has been proposed. 1041
During the application, it should be remembered 1042
that the direction of electroosmotic flow in porous
1043
ceramic media has a strong influence on the 1044
amount of water being added to the soil from the 1045
ceramic castings. Anode ceramic casting would be
1046

suitable for long-term electrokinetic remediation 1047
processes if it was ensured that electroosmotic 1048
flow occurred from the surrounding soil towards 1049
Ž
the interior of the anode casting Mattson and
1050
.
Lindgren, 1995 . As efficient electrokinetic reme- 1051
diation in unsaturated soils depends on the water 1052
amount at the anode, there is a necessity to 1053
continuously inject water during the whole reme- 1054
diation process. Despite the addition of water, it 1055
is important to maintain unsaturated conditions 1056
in the soil, because excess water may cause satu- 1057
rated conditions and contaminants will be able to 1058
migrate into the deeper layers of the soil. 1059
A number of experiments with an anode cer-
1060
amic casting were conducted and it was proved 1061
that it is possible to remove heavy metal contami-
1062
nants from unsaturated, sandy soils using the 1063
Ž
electrokinetic remediation technique Mattson
1064
.
and Lindgren, 1995 . 1065
First of all a laboratory cell was designed and 1066
constructed, which consisted of a plastic con-
1067

tainer filled with buffering solution. The polyvinyl 1068
chloride plate glued to the bottom of the con- 1069
tainer, the porous ceramic castings, woven wire 1070
cathode and graphite anode are shown in Fig. 4. 1071
The most suitable buffering solution for this ex-
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 13
1072
periment is a phosphate solution with a pH of 61073
Ž.
Mattson and Lindgren, 1995 . To overcome the1074
hydraulic counterflow, the experiment should only1075
be conducted until the fluid level difference1076
between the inner and outer reservoirs becomes1077
Ž.
) 1 cm Mattson and Lindgren, 1995 .1078
After laboratory experiments, a number of field1079
studies were conducted and the initial results1080
obtained are very promising. It is possible to state1081
that the use of anode ceramic casting may sig-1082
nificantly improve the application of electroki-1083
netic remediation in unsaturated soil media.
1084
1085
4.3. Lasagna
TM
process

1086
1087
In 1995, a novel integrated method for in situ1088
electrokinetic remediation of soils, called1089
Lasagna
TM
, was developed and implemented at1090
the Paducah site, in Kentucky, USA. This tech-1091
nology is useful for removing heavy metal con-1092
taminants from heterogeneous, low-permeability1093
Ž.
soils Ho et al., 1997, 1999 .
1094
In brief, the Lasagna
TM
process contains the1095
following concepts:1096
10971098
ⅷ
The creation of several permeable ‘treatment’1099
zones in close proximity through the whole1100
soil matrix by adding sorbents, catalytic1101
reagents, buffering solutions, oxidising agents,1102
etc.
1103
1104
ⅷ
Application of an electric current in order to
1105
transport contaminants into the ‘treatment’ 1106

zones created. 1107
1108
The Lasagna
TM
process has several advantages
1109
in comparison to other techniques. First, it is
1110
possible to recycle the cathode effluent by aiming
1111
it back to the anode compartment, which would
1112
favour neutralising of the pH and simplify water
1113
management. In addition, the fluid flow may be
1114
Ž
reversed by simply switching the polarity Ho et
1115
.
al., 1999 . The switching of polarity promotes
1116
multiple contaminant passes through the ‘treat-
1117
ment’ zones and helps to diminish the possibility
1118
of non-uniform potential and pH jumps in the soil 1119
system.
1120
Two schematic Lasagna

TM
model configura-
1121
Ž.
tions were suggested: horizontal Fig. 5 and verti- 1122
Ž.
cal Ho et al., 1999 .
1123
The process was called ‘Lasagna’ due to the
1124
layering of treatment zones between the elec-
1125
trodes. The formation of horizontal fractures in
1126
over-consolidating clays due to the horizontal
1127
electrodes and vertical pressuring system make
1128
this method especially effective in removing con-
1129
Ž
taminants from deeper layers of the soil Ho et
1130
.
al., 1999 . In addition, for shallow contamination
1131
which does not exceed 15 m and in not over-con-
1132
solidated soils, the vertical treatment configura- 1133
Ž.Ž.

tion is more appropriate Ho et al., 1997 Fig. 6 .
1134
Ž.
Fig. 4. Electrokinetic cell with ceramic castings Mattson and Lindgren, 1995 .
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx14
1135
TM
Ž.
Fig. 5. Horizontal Lasagna configuration adapted from Ho et al., 1999 .
1136
Accordingtolaboratoryexperimentsand
1137
promising pilot-scale studies at the Paducah site
1138
in Kentucky, Lasagna
TM
technology may become
1139
one of the most widely used electrokinetic reme-
1140
diation technologies for removing heavy metal
1141
contaminants from various soils. Nevertheless,
1142
there are several technological and other limita-
1143

tions, which should be improved for future stud-
1144
ies. It is obvious that Lasagna
TM
technology is
1145
potentially capable of treating multiple contami-
1146
nants in clay and laden soils, but additional exper-
1147
iments and studies should be conducted in order
1148
to assure that the treatment process is compatible
1149
TM
Ž
Fig. 6. Vertical Lasagna configuration adapted from Ho et1150
.
al., 1997 .
1151
for individual contaminants. In addition, one of 1152
the biggest technology drawbacks is the entrap- 1153
ment of gases formed by electrolysis and the 1154
assurance of good electrical contact to the elec- 1155
trodes. To increase the Lasagna
TM
process effi- 1156
ciency, there were attempts to implement biore-
1157
mediation in ‘treatment’ zones. It is believed that

1158
bioremediation together with electrokinetic reme- 1159
diation may significantly increase the overall re- 1160
moval of heavy metals, as well as other contami- 1161
nants, from clays and other soils.
1162
1163
4.4. Electro-Klean
TM
electrical separation
1164
1165
Electro-Klean
TM
technology is applied in situ, 1166
as well as ex situ, in Louisiana, USA. This is a
1167
new methodology, which is used to remove heavy 1168
metals, radionuclides and specific volatile organic
1169
contaminants from saturated and unsaturated 1170
sands, silts, fine-grained clays and sediments. This 1171
technology uses two electrodes to apply DC di-
1172
Ž
rectly into the contaminated soil mass van 1173
.
Cauwenberghe, 1997 . In order to improve the
1174
remediation efficiency, enhancement fluids, 1175

mostly acids, are added into the soil. The main
1176
limitation of this technique is the high buffering 1177
capacity of the soils and different coexisting 1178
chemicals and their concentrations.
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 15
1179
1180
4.5. Electrokinetic bioremediation
1181
1182
Electrokinetic bioremediation technology is de-1183
signed to activate microbes and other micro-1184
organisms present in soils by the use of selected1185
nutrients to promote the growth, reproduction1186
and metabolism of micro-organisms capable of1187
Ž
transforming organic contaminants in soil van1188
.
Cauwenberghe, 1997 . Nutrients reach the or-1189
ganic contaminants by specially applied bioelec-1190
tric technology. It is believed that this technology1191
may be very successful in the future, because it1192
does not require an external microbial population1193
to be added into the soil system. In addition,1194
nutrients may be uniformly dispersed over the

1195
contaminated soil or directed to a specific loca-1196
Ž.
tion van Cauwenberghe, 1997 and the method1197
avoids the problems associated with transport of1198
Ž
micro-organisms through fine-grained soils Fig.1199
.
7.1200
Despite promising results, this technology has1201
some major limitations. Sometimes the concen-1202
tration of organic pollutant exceeds the toxic limit1203
for the microbial population and micro-organisms1204
die. Simultaneous bioremediation of various or-1205
ganic contaminants may produce by-products,1206
which are highly toxic to micro-organisms. Those1207
by-products may significantly inhibit the bioreme-1208
diation rates.
1209
1210
4.6. Electrochemical geooxidation
1211
1212
Electrochemical geooxidation is used in Ger- 1213
many to remediate soil and water contaminated 1214
Ž
with organic and inorganic compounds van 1215
.
Cauwenberghe, 1997 . The in situ process in- 1216
volves the application of an electrical current to 1217

probes driven into the ground. The applied cur- 1218
rent creates favourable conditions for oxidation᎐ 1219
reduction reactions, which lead to the immobilisa- 1220
tion of inorganic contaminants in the soil or 1221
groundwater matrix between the electrode loca- 1222
tions. The main advantage of this technology is 1223
that there is no need to use catalysts for the 1224
oxidation᎐reduction reactions, because in almost 1225
all soils, natural catalysts, such as iron, magne- 1226
sium, titanium and elemental carbon, are present. 1227
The limitations of this technology are the very 1228
long remediation time and the lack of proven 1229
results.
1230
1231
4.7. Electrochemical ion exchange
1232
1233
This technology employs a series of electrodes, 1234
placed in porous castings, which are supplied with 1235
circulating electrolytes. During the remediation
1236
process, ion contaminants are captured in these
1237
electrolytes and pumped to the surface, where 1238
they are passed through an electrochemical ion
1239
Ž.
Fig. 7. Electrokinetic bioremediation according to Thevanayagam and Rishindran, 1998 .
Uncorrected Proof

ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx16
1240
Ž.
exchanger van Cauwenberghe, 1997 . This1241
method is used to remove heavy metals, halides1242
and specific organic species from different types1243
of soils. The most important limitation of this1244
technology is that it is a very expensive procedure1245
for cleaning effluents containing low levels of1246
contaminants.1247
1248
4.8. Electrosorb
TM1249
1250
Electrosorb
TM
technology is mostly used in1251
Louisiana, USA, and uses cylindrical electrodes1252
coated with a specially designed polymer mate-1253
rial. This polymer is impregnated with pH-regu-1254
lating chemicals in order to prevent pH jumps1255
Ž.
Reddy and Chinthamreddy, 1999 . During the1256
remediation process, electrodes are placed in1257
boreholes in the soil and direct current is applied.1258
Ions move through the pore water to the elec-1259
trode, where they are trapped in the electrode1260

polymer matrix. Although there are no indica-1261
tions of the limitations of the technique proposed,1262
it is believed that in order to be commercially1263
available, it should be further investigated.1264
1265
5. Remediation of specific heavy metal1266
contamination1267
1268
As the heavy metal contaminants in a soil and1269
solution primarily exist in the form of salts and1270
ions, the potential of an electrokinetic remedia-1271
tion technique depends on the quantity of those1272
compounds.1273
1274
5.1. Remo¨al of cadmium and lead1275
1276
Under alkaline conditions, cadmium and lead
1277
in the soil may become sediments of hydroxides1278
w Ž. Ž.x Ž
Cd OH , Pb OH and carbonates CdO ,
22 31279
.
PbCO . Soil pH determines the concentrations
31280
of hydroxide and carbonate in the soil solution,
1281
which play a crucial role in the formation of1282
Ž
heavy metal complexes in soil Sah and Chen,

1283
.
1998 .1284
In order to understand the migration of Pb and1285
Cd between electrified vs. non-electrified soil1286
samples under different times, locations and solu-
1287
tion types, it is important to use heavy metal 1288
Ž.
formal analysis Sah and Chen, 1998 . Also, due 1289
to varying stability of different heavy metals in 1290
the soil, there is a necessity to determine ap- 1291
propriate application times for electrokinetic re- 1292
mediation and the pH of the soil. 1293
Experiments conducted show that Pb-con- 1294
taminated soil is usually quite difficult to remedi- 1295
ate. However, high removal rates for Pb, as well 1296
as Cd, were obtained in experiments where HCl 1297
Ž
solution was used Acar and Alshawabkeh, 1993; 1298
.
Sah and Chen, 1998 . 1299
If the environment near the cathode is basic, it 1300
may favour the formation of the insoluble hydrox- 1301
Ž.
ide Cd OH . However, this Cd species may not
2 1302
Ž
be mobile under advective flow Acar and Al-
1303

.
shawabkeh, 1993, 1996; Acar et al., 1994, 1996 .
1304
In order to improve the removal rates of 1305
cadmium and lead from soils, the following pro- 1306
Ž.
posals should be considered Sah and Chen, 1998 : 1307
13081309
ⅷ
Experiments showed that soil could absorb 1310
more Pb than Cd, which should be taken into 1311
consideration in further laboratory experi- 1312
ments, as well as pilot-scale studies.
13131314
ⅷ
Cd-spiked samples have revealed a higher cur- 1315
rent density than Pb-spiked samples during 1316
the remediation process. A thin, white oxidant 1317
film was found on the cathode, which reduced 1318
the conductivity and removal efficiency of 1319
metals. Thus, an enhancement fluid should be 1320
added at the electrodes, or the electrodes 1321
must be cleaned regularly during the applica- 1322
tion. 13231324
ⅷ
The use of HCl acid increased the removal 1325
rates of lead and cadmium. In order to achieve 1326
optimal removal results, acid solution has to 1327
be added to the soil solution.
1328

1329
1330
5.1.1. Lead migration in soils
1331
Cationic heavy metals, such as Pb, are most 1332
soluble at a low pH. As the H
q
produced at the 1333
anode moves across the soil sample, cationic met- 1334
als which were sorbed or precipitated onto the
1335
soil particles are, in many cases, solubilised and 1336
may be able to undergo transport by diffusion, as 1337
well as via electrokinetic remediation processes,
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 17
1338
such as advection by electroosmotic flow and1339
electrolytic migration. Diffusion and electrolytic
1340
migration of OH
y
ions produced at the cathode1341
increase the pH of the system near the cathode
1342
Ž
and may precipitate desorbed ions Viadero et al.,1343

.
1998 . This is shown schematically in Fig. 8.
1344
Experiments showed that at a pH above 4᎐4.5,1345
lead was either adsorbed onto the soil andror
1346
Ž.Ž.
precipitated as Pb OH s , which reduced the
21347
conductivity of the soil by removing cations from
1348
Ž.
the liquid Viadero et al., 1998 . At high pH, most1349
of the lead is retained in hydroxide and carbonate1350
phases.
1351
1352
5.1.2. Cadmium migration in soils1353
When the initial pH is low, the conductivity of1354
the medium is high, and very low electrical poten-
1355
tial gradients are initially generated across the1356
Ž
specimen Acar and Alshawabkeh, 1993, 1996;
1357
Acar et al., 1994, 1996; Probstein and Hicks,1358
1993; Mattson and Lindgren, 1995; Sah and Chen,1359
.
1998; Viadero et al., 1998 .
1360

Numerous experiments have been conducted to1361
remove cadmium from kaolin. In kaolin, without1362
the addition of a reducing agent and in the pres-
1363
ence of humic acid and ferrous iron, low pH1364
conditions exists throughout most of the soil, ex-
1365
cept near the cathode. As low pH conditions1366
favours the dissolution of Cd species, cadmium is
1367
transported to the cathode compartment1368
Ž.
Pamukcu, 1997 . Low-concentration Cd speci-1369
mens exhibit a larger influx of water than high Cd
1370
concentration specimens for the same level of1371
Ž.
electricity Pamukcu, 1997 .
1372
qy
Ž.
H q e ª 1r2H
2 1373
1374
Cd
2q
q2e
y
ª Cd
0 1375

1376
Ž.Ž.
y 0 y
Ž.
Cd OH s q2e ª Cd q2OH 3
2
1377
When the current density is greater than 5 1378
mArcm
2
, secondary temperature effects are re- 1379
ported to decrease the efficiency of electro- 1380
Ž.
osmotic flow Hansen et al., 1997 .
1381
1382
5.2. Remo¨al of arsenic and chromium
1383
1384
The main substance used for desorbing cationic 1385
species is hydronium ions H O
q
produced at the
3 1386
anode during the electrolysis process. However, 1387
there are several major drawbacks of this process: 1388
it induces a dissolution of major soil components, 1389
Ž.
such as carbonates, as well as oxides Fe, Mg 1390
Ž.

when strongly acidified Hecho et al., 1998 .
1391
Anionic species are removed by the hydroxide 1392
ions generated at the cathode. It is necessary to 1393
add an anionic oxidising agent, which would mi- 1394
Ž
grate to the anode through the soil matrix Hecho
1395
.Ž.
et al., 1998 . Chromium III can be oxidised into
1396
Ž.
Cr VI as anionic species, which can be desorbed
1397
in alkaline medium. This method is useless with
1398
arsenic, because all soluble arsenic species are 1399
Ž.
anionic above pH 9 and arsenic V is more 1400
Ž.
strongly sorbed that arsenic III .
1401
In order to remove chromium from soils, it is 1402
Ž. Ž.
necessary to oxidise Cr III first to chromium VI , 1403
which is anionic. The removal of arsenic is not as 1404
complicated as that of chromium. The literature 1405
Ž.
indicates that arsenic III is more soluble than
1406

Ž.
Fig. 8. Lead removal from soils according to 29 .
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx18
1407
Ž.
arsenic V , so the use of an oxidising agent does1408
not seem useful.1409
Two alkaline reagents, i.e. sodium carbonate1410
and sodium hydroxide, are used to enhance the1411
Ž
remediation process Reddy and Chinthamreddy,
1412
.
1999 . Earlier, two alternatives, i.e. hydrogen per-1413
oxide and sodium hypochlorite, were used as oxi-1414
dising agents. However, experiments proved that1415
hydrogen peroxide tends to reduce very rapidly in1416
the soil, and only hypochlorite was used for fur-1417
Ž
ther laboratory and pilot studies Hansen et al.,
1418
.
1997; Hecho et al., 1998 .1419
1420
5.2.1. Chromium migration1421
Chromium can exist in valence states ranging1422

from y2toq6; however, q3 and q6 are the1423
only two valence states that prevail under subsur-1424
Ž
face conditions Reddy et al., 1997, 1999; Reddy1425
.
and Chinthamreddy, 1999 . Hexavalent
1426
Ž.
chromium VI is highly mobile and toxic in com-1427
Ž. Ž.
parison to Cr III . Cr VI exists as anions, speci-1428
Ž
y
.
fically hydrochromate HCrO , dichromate
41429
Ž
2y
.Ž
2y
.
Cr O and chromate CrO , and will mi-
27 41430
grate towards the anode during the electrokinetic1431
Ž.
remediation process. On the other hand, Cr III
1432
exists as a cation Cr
3q
and may form cationic,1433

neutral and anionic hydroxy complexes, specifi-1434
Ž.
2q
Ž.
q
Ž. Ž.
y
cally Cr OH , Cr OH , Cr OH , Cr OH
2341435
Ž.
2y
Ž.
and Cr OH . Cr III may also exist as other
51436
cationic, neutral and anionic inorganic and or-1437
ganic complexes, depending on the ligands pre-1438
Ž.
sent Reddy and Chinthamreddy, 1999 .1439
In acidic regions and at relatively low redox1440
Ž.
3q
potentials, Cr III exists as Cr and forms1441
Ž.
2q
cationic complexes Cr OH . Being positively1442
Ž.
q
charged, Cr OH will migrate towards the cath-
21443
ode during the electrokinetic remediation process.

1444
Ž. w Ž.x
Cr III precipitates as its hydroxide Cr OH
31445
between pH 6.8 and 11.3, while at higher pH
1446
Ž.
values, Cr III may form anionic hydroxy com-1447
w Ž.
y
Ž.
2y
x Ž
plexes Cr OH and Cr OH Reddy et al.,
451448
.
1997, 1999; Reddy and Chinthamreddy, 1999 .1449
The removal of chromium from soils by elec-
1450
trokinetic remediation is highly efficient if the1451
Ž.Ž
chromium exists as Cr VI Acar and Al-
1452
shawabkeh, 1993, 1996; Acar et al., 1994, 1996;1453
Reddy et al., 1997, 1999; Reddy and Chintham-1454
.
reddy, 1999; Sah and Chen, 1998 . If reducing
1455
agents, such as organic matter, sulfides or ferrous 1456
Ž.

iron, are present in natural soils, Cr VI is likely 1457
Ž.
to be reduced to Cr III , which may significantly 1458
affect the electrokinetic migration of chromium, 1459
as well as the migration of co-existing metals 1460
Ž. Ž.Ž
such, as Ni II and Cd II Reddy et al., 1997, 1461
.
1999; Reddy and Chinthamreddy, 1999 . 1462
As chromium species favour alkaline conditions 1463
in soils, an alkaline reagent must be injected into 1464
the soil system in order to neutralise H O
q
ions.
3 1465
In order to enhance the electrokinetic remedia- 1466
tion application, an oxidising agent ᎏ sodium 1467
hypochlorite ᎏ needs to be injected at the cath- 1468
Ž.
ode compartment Reddy et al., 1999 . Hypochlo- 1469
rite ions can migrate towards the anode and oxi- 1470
dise trivalent chromium to hexavalent chromium, 1471
which in turn migrates towards the anode. 1472
After close investigation of the effects of reduc- 1473
ing agents on chromium species migration, it was 1474
observed that when the chromate front meets the 1475
anodic reaction product Fe
2q
in a region adjacent 1476
to the anode, it reacts to form Cr

3q
and Fe
3q 1477
species:
1478
2q 6q 3q 3q
Ž.
Fe qCr m Fe q Cr 4
1479
Thus, further migration of chromate is inhib- 1480
ited due to redox reactions with ferrous ions 1481
Ž.Ž.
Haran et al., 1996 . Cr III is immobilised in
1482
sand due to the formation of complex sulfates 1483
Ž.
and hydroxides. When the pH is increased, Cr III 1484
is likely to be precipitated as chromic hydroxide:
1485
3qy
Ž. Ž.
Cr q3OH ª Cr OH 5
3
1486
The reduction reaction is controlled by two im- 1487
Ž.
portant factors, the amount of Fe II in the sand 1488
Ž.
and the soil pH Haran et al., 1996 :
1489

2qy
Ž.
Fem Fe q2e 6
1490
Ž.
y
Cr VI exists predominantly as HCrO at low pH
4 1491
2y
Ž
and as CrO at high pH in solution Reddy et
4 1492
.
al., 1997 :
1493
yq
Ž.
SᎏOHsSᎏO qH7.11494
1495
qq
Ž.
SᎏOHqH s SᎏOH 7.2
2
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 19
1496
1497

yq
Ž.
SᎏO qM s SᎏOM 7.31498
1499
qy
Ž.
SᎏOH qL s SᎏOH L 7.4
22
1500
where S᎐OH represents a typical surface functio-1501
nal group, and M
q
and L
y
represent a cation and1502
anion, respectively.1503
These complexation reactions are highly pH-1504
dependent, because the extent of surface depro-1505
Ž.
tonation Sogorka et al., 1998 and protonation1506
Ž.
reactions Acar and Alshawabkeh, 1993 is con-1507
Ž.
trolled by the solution pH Reddy et al., 1999 .1508
1509
5.2.2. Chromium remo¨al from different soils1510
Different experiments were conducted to ob-1511
tain results for chromium removal efficiency from1512
several types of naturally occurring soils, such as1513
Ž

kaolin and glacial till Acar and Alshawabkeh,1514
1993; Mattson and Lindgren, 1995; Reddy et al.,1515
1997, 1999; Reddy and Chinthamreddy, 1999; Sah1516
.
and Chen, 1998 .1517
The presence of reducing agent in soils, such as
1518
humic acid, did not retard the chromium migra-1519
tion, either in kaolin or in glacial till; actually, it1520
enhanced chromium migration towards the anode1521
Ž.
Reddy and Chinthamreddy, 1999 . On the other
1522
hand, ferrous iron, another reducing agent natu-
1523
rally present in soils, showed moderate retarda-1524
tion of chromium migration. Finally, the presence1525
of sulfides showed the highest rate of retardation1526
of chromium species migration towards the anode.1527
It is possible to conclude that when a reducing1528
Ž.
agent was present, higher Cr III concentrations1529
were observed near the anode. On the other1530
Ž.
hand, the reduced Cr III tends to migrate to-1531
Ž.
wards the cathode, resulting in high Cr III con-1532
Ž.
centrations in the section near the anode. Cr VI
1533

adsorption onto soil decreases with an increase in1534
Ž.
soil pH Reddy et al., 1997 .
1535
1536
5.2.2.1. Glacial till. Glacial till has high buffering1537
capacity because of the presence of carbonates in
1538
this soil. It is reported that there are no traces of1539
Ž
acid front formation in glacial till Reddy and
1540
.
Chinthamreddy, 1999 . Carbonates have the abil-1541
ity to neutralise H
q
ions generated, and block1542
development of an acidic pH environment near1543
the anode. The adsorption of HCrO
y
onto the
4
1544
soils is significant, but the adsorption of CrO
2y
is
4 1545
Ž.
negligible Reddy et al., 1999 . It is obvious that 1546
Ž.

high pH in glacial till causes all Cr VI to exist as 1547
CrO
2y
, which therefore results in low adsorption
4 1548
of species onto the soil. Soluble CrO
2y
ions are
4 1549
transported to the anode by electromigration. 1550
Ž. Ž.
The possibility of Cr VI conversion to Cr III
1551
Ž.
was evaluated Reddy et al., 1997, 1999 . It was 1552
proved that without reducing agents in the soil, 1553
Ž. Ž.
significant Cr VI reduction to Cr III would not
1554
occur. 1555
Iron deposits of hematite, pyrite and goethite 1556
occur in abundance in natural soils. When there 1557
are slightly alkaline conditions in glacial till, 1558
Ž.
2y
Cr VI exists predominantly in the form of CrO ,
4 1559
and it is reported in the literature that CrO
2y
4

1560
adsorption onto Fe O is significant. In addition,
23 1561
hematite may react with constituents of glacial 1562
Ž.
till, which may favour further removal of Cr VI
1563
Ž.
in the pore water Reddy et al., 1997 . 1564
1565
5.2.2.2. Kaolin. A distinct pH gradient developed 1566
Ž.
2y
in kaolin causes Cr VI to exist as both CrO
4 1567
y
Ž
and HCrO species Reddy and Chinthamreddy,
4 1568
.
1999 . In addition, alkaline conditions near the
1569
Ž.
cathode favour the existence of Cr VI in the
1570
form of CrO
2y
, which does not adsorb to the soil,
4 1571
Ž.

and therefore most Cr VI exists in solution and 1572
migrates toward the anode. On the other hand, 1573
CrO
2y
ions enter an acidic region near the anode,
4 1574
which favours the formation of HCrO
y
ions. As
4 1575
mentioned earlier, HCrO
y
adsorbs significantly
4 1576
Ž.
to the soil, which retards Cr VI migration.
1577
1578
5.2.3. Arsenic migration and remo
¨al 1579
In alkaline conditions, arsenic species do not
1580
demonstrate well-expressed adsorption, although 1581
Ž.
As V is usually more strongly adsorbed than
1582
Ž.
As III . It is indicated that alkaline conditions 1583
favour arsenic electromigration, although it is very
1584

Ž
slow and time-consuming Acar and Al- 1585
shawabkeh, 1993; Acar et al., 1996; Mattson and 1586
Lindgren, 1995; Haran et al., 1996; Sah and Chen,
1587
.
1998; Viadero et al., 1998 . In order to enhance 1588
the electromigration process, sodium hypochlorite
1589
is introduced into the process. To achieve the 1590
process efficiency desired and improve the system 1591
performance, it is necessary to inject an enhance-
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx20
1592
ment solution directly into the cathodic compart-1593
Ž
ment Reddy et al., 1997, 1999; Reddy and1594
.
Chinthamreddy, 1999 .1595
1596
5.3. Remo¨al of mercury1597
1598
Electrokinetic remediation of Hg-contaminated1599
soils is very difficult because of the low solubility1600
of Hg in most natural soils. The predominant1601
Ž.

species of insoluble Hg in the soils are HgS, Hg I1602
Ž.
and Hg Cl Cox et al., 1996 . Several years ago,
221603
a new method for Hg removal from soils was1604
introduced. It uses an I rI
y
lixiviant solution to
21605
solubilise Hg from contaminated solids. Oxidation1606
Ž.
of reduced insoluble Hg by I releases Hg II ,
2
1607
which is complexed as soluble HgI
2y
and Hg ions
41608
are ready to migrate through the soil towards the1609
Ž.
anode and be removed Cox et al., 1996 :
1610
HgSqI q2I
y
m HgI
2y
qS
24Žoxidised.1611
1612
Ž.

y 2y
Hg I qI q2I m HgI
241613
1614
y 2y 2y
Ž.
HgOq 4I m HgI q O8
4
1615
Once solubilised, Hg is able to migrate through
1616
the soil and be removed.1617
It should be mentioned that iodide solution1618
and I crystals introduced near the cathode react
21619
to form I
y
complex. Reduced forms of insoluble
31620
Hg can be oxidised by either I or I
y
; however,
231621
transport of oxidant through the soil is dependent1622
on the electromigration of the I
y
anion. The
31623
HgI
2y

complex formed via reactions with lixiviant
41624
solution is removed from the soil by electromigra-
1625
tion towards the anode.1626
A pH jump was observed during the electroki-
1627
netic remediation process. It is believed that this1628
pH increase may be caused by the following reac-
1629
tion, if an excess of Cl
y
is present under aerobic1630
Ž.
conditions Cox et al., 1996 :
1631
O q 2Hgq8Cl
y
q2H Oª 2HgCl
2y
q4OH
y
224
Ž.
9
1632
Mercury removal may be more efficient if chlo-
1633
ride or another suitable component is added to 1634
Ž.

the soil system Hansen et al., 1997 . Additional 1635
chloride ions are able to mobilise the mercury, 1636
forming complex ions which are easily transported 1637
out from the soil by electromigration. For in- 1638
stance, hypochlorite may be a suitable compound, 1639
which oxidises metallic mercury, forming HgCl
2y
:
4
1640
y 2yy
Ž.
HOClq Hgq3Cl ª HgCl qOH 10
4
1641
Although some promising results have been 1642
demonstrated, this method has several major 1643
drawbacks. First, in the presence of any organic 1644
matter, hypochlorite may form toxic, halogenated 1645
organic compounds, which are dangerous for hu- 1646
mans and may severely harm the environment. In 1647
addition, if not removed before the electrokinetic 1648
remediation process begins, metallic mercury 1649
would inhibit the overall remediation process due 1650
to its electric conductivity. 1651
1652
5.4. Remo¨al of zinc and copper 1653
1654
All calcium and magnesium should be removed
1655

before removal of zinc is initiated. The use of 1656
enhancing solutions, such as sodium acetate, in- 1657
creases the removal efficiency for metal ions, as 1658
well as reduces the process time. It is obvious that 1659
the cations with lower interaction energy will be 1660
removed first and will be followed by cations with 1661
higher interaction energy. 1662
After the number of experiments, the sequence 1663
of heavy metal removal from soils using sodium 1664
Ž
acetate as enhancement fluid was proposed Cox
1665
.
2q 2q 2q 2q
et al., 1996 : Ni f Cd ) Ca ) Cr) Zn ) 1666
K
q
fMg
2q
) Cu
2q
) Pb
2q
. 1667
Also, several experiments were conducted with 1668
distilled water as the enhancement fluid and the 1669
Ž.
following results were observed Cox et al., 1996 .
1670
Ca

2q
,Mg
2q
,Zn
2q
,K
q
and Pb
2q
percentage 1671
removal efficiencies were low and sometimes close 1672
to zero. Only Ni
2q
and Cr had removal efficien- 1673
cies quantified as medium᎐high.
1674
A schematic electrolytic cell for the removal of 1675
copper from contaminated soil was proposed by
1676
Ž.Ž .
Cox et al. 1996 Fig. 9 . 1677
The electrolytic cell is divided into three parts 1678
and the contaminated soil and electrodes are
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 21
1679
Ž.

Fig. 9. Electrokinetic cell for copper removal from soils adapted from Cox et al., 1996 .
1680
separated by anionic and cationic exchange mem-1681
branes. The anode and cathode compartments1682
contain electrolyte solution at constant pH 3.1683
A low pH value was maintained to keep copper1684
dissolved in the soil, thus making migration to-1685
wards the cathode and subsequent removal from1686
the soil feasible. Despite the fact that almost all1687
of the copper was found in the cathode compart-1688
ment, a certain amount was found in the anion1689
membrane. It was also suggested that copper
1690
found in the anionic membrane may be due to its1691
capability of forming complexes with different1692
Ž.
ligands present in soils Ribeiro et al., 1997 .
1693
1694
5.5. Other metals
1695
1696
Strontium remains as a divalent ion over a1697
large pH range. The cathode should not affect1698
strontium, since it will remain a divalent ion, even1699
Ž.
at high pH Pamukcu, 1997 .1700
In alkaline solution, the predominant species of1701
Co
2q

are either positively charged ions or hydrox-
1702
Ž.
ide Co OH salts. It is apparent that at high pH,
2 1703
cobalt tends not to precipitate onto soil particles, 1704
and may therefore be removed. 1705
According to the experiments, if Ca
2q
ions are 1706
removed first, then Zn
2q
follow, and finally Cu
2q 1707
2q
Ž
and Pb ions are removed Hansen et al., 1997;
1708
.
Hecho et al., 1998 . In order to mobilise contami- 1709
nants, energy may be wasted in dissolving lime 1710
and carrying harmless Ca
2q
ions out of the soil. It 1711
is obvious that further research concerning other 1712
suitable soil pre-treatment methods to mobilise 1713
contaminants need to be investigated and carried 1714
Ž
out Hansen et al., 1997; Hecho et al., 1998; 1715
.

Viadero et al., 1998 .
1716
1717
6. Heavy metal removal efficiency from 1718
contaminated soils
1719
1720
Electrokinetic remediation techniques have de- 1721
monstrated 85᎐95% efficiency in removing ar- 1722
senic, cadmium, chromium, cobalt, mercury,
1723
Table 1
Ž
Heavy metal removal efficiency from selective contaminated soils using the electrokinetic remediation technique adapted from
.
Lageman, 1993; Sengupta, 1995
1724
Ž.
Soil Metal removal efficiency %
Cd Cr Ni Pb Hg Cu Zn As Co Sr
Agrillaceous sand
River mud 50 64 91 54 60 71 94 66 ᎐᎐
Kaolinite 94.6 93.1 88.4 69 26.5 ᎐ 54.6 54.7 92.2 97.8
Kaolinite and 92.7 97.6 93.9 66.9 42.5 ᎐ 36.3 27.2 95.9 96
humic substances
Montmorillonite 86.6 93.5 93.6 ᎐᎐᎐64.4 64.3 89.4 92.3
Clayey sand 98 96.8 95.9 83 78.3 ᎐ 54.5 54.7 97.5 99
Uncorrected Proof
ARTICLE IN PRESS
()

J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx22
1725
nickel, manganese, molybdenum, zinc, antimony1726
and lead from low-permeability soils, i.e. clay,1727
peat, kaolinite, high-purity fine quartz, Na and1728
sand᎐montmorillonite mixtures, as well as from1729
Ž.
agrillaceous sand Yeung et al., 1997 . In addi-
1730
tion, highest removal efficiencies, i.e. more than1731
90% of heavy metals, were obtained in kaolinite1732
Ž.
Pamukcu and Wittle, 1992 . However, for porous,
1733
high-permeability soils, such as peat and river1734
sediment, the removal efficiency was approxi-1735
Ž.
mately 65% Chilingar et al., 1997 .
1736
A low pH profile in fine-grained soils may1737
contribute to higher efficiency for metallic con-1738
taminant removal. In addition, the low acidrbase1739
buffering capacity of kaolinite also contributes to1740
the higher heavy-metal removal efficiency for this1741
Ž
type of soil Hamed et al., 1991; Hicks and Ton-
1742
.
dorf, 1994 . Soils with a high content of humic1743

substances have higher cation exchange and buf-1744
fering capacity, which is why electrokinetic reme-1745
Ž.
diation efficiencies may decrease Table 1 .1746
It is very important to improve the removal1747
efficiency of heavy metals from high sorption-1748
capacity clays, such as illitic mixture, i.e. synclays.1749
Despite all the earlier accomplishments, elec-1750
trokinetic remediation of such soils still requires1751
higher current density, remediation time, energy1752
expenditure and costs in comparison to kaolinite1753
Ž.
Puppala et al., 1997 .
1754
1755
7. Cost–benefit analysis1756
1757
There are several factors that influence the
1758
cost of the electrokinetic remediation process.1759
Ž.
These are as follows van Cauwenberghe, 1997 :1760
17611762
ⅷ
Soil characteristics and moisture content;
17631764
ⅷ
Contaminant concentrations;17651766
ⅷ
Concentrations of non-target ions and con-

1767
ductivity of the pore water;17681769
ⅷ
Depth of the remediated soil;
17701771
ⅷ
Site preparation requirements; and17721773
ⅷ
Electricity and labour costs.1774
1775
During numerous laboratory experiments, it was1776
determined that if the distance between elec-1777
trodes was 1᎐1.5 m, the total removal of heavy
1778
metals from contaminated soil would require ap- 1779
proximately 500 kW hrm
3
of energy. Energy ex- 1780
penditure is directly proportional to the complete 1781
removal of contaminants from soil, i.e. remedia- 1782
Ž.
tion time van Cauwenberghe, 1997 . The total 1783
energy consumption can be lowered if appropri- 1784
ate cathodic polarisation techniques are used 1785
Ž.
Acar and Alshawabkeh, 1997; Li and Li, 2000 . 1786
The migration rate of contaminants through the 1787
soil matrix is approximately 2᎐3cmrday. If the 1788
distance between the electrodes is 2᎐3 m, the 1789
time frame for successful remediation would be 1790

Ž.
more than 100 days van Cauwenberghe, 1997 . 1791
However, the use of a cation-selective membrane 1792
reduces the remediation period to 10᎐20 days. 1793
The situation with in situ experiments is slightly 1794
different. The main parameters that influence the 1795
overall process cost are as follows: 1796
17971798
ⅷ
Soil properties; 17991800
ⅷ
Depth of contamination; 18011802
ⅷ
Cost of accommodating electrodes and placing
1803
treatment zones; 18041805
ⅷ
Clean-up time; and 18061807
ⅷ
Cost of labour and electrical power. 1808
1809
In order to avoid soil overheating and shorten the 1810
required time frame, the cost-optimised distance 1811
between electrodes needs to be maintained at 1812
Ž
3᎐6 m for most soils Lageman, 1993; Ho et al., 1813
.
1997, 1999 . Electrode construction costs account 1814
for up to 40% of the overall remediation costs. 1815
Ž.

Other expenses are Ho et al., 1997 : 1816
18171818
ⅷ
10᎐15% for electricity; 18191820
ⅷ
17% for labour; 18211822
ⅷ
17% for materials; and 18231824
ⅷ
Up to 16% for licenses and other fixed costs.
1825
1826
The first in situ electrokinetic remediation
1827
technique implemented, the Lasagna
TM
process, 1828
has reduced the clean-up time and power input
1829
required, as well as the total costs, by inserting 1830
treatment zones between the electrodes. Treat-
1831
ment zones diminish the need for above-ground 1832
waste handling and are cheaper to implement 1833
Ž.Ž.
than electrodes Ho et al., 1997 . Ho et al. 1997 1834
have presented a comparison of the cost᎐benefit
Uncorrected Proof
ARTICLE IN PRESS
()

J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 23
1835
Table 2
Cost᎐benefit analysis of selected techniques
1836
Technique Costs Remarks
3
TM
Lasagna 50᎐120 $USDrm Mandrelrtremie-tube method of
approximately Emplacement will be used instead
Ž.
over 3 years of earlier proposed steel plate
Electrodes with wick drains
and carbon-filled treatment zone
3
Soil heatingrvapour extraction 65᎐123 US$ryd
technology
3
Chemical oxidation 130᎐200 US$rm Technique was mostly used
Ž
with potassium permanganate to remove DNAPLs in situ
.
or hydrogen peroxide
1837
analysis for selected techniques, which is shown1838
in Table 2.
1839
1840
8. Conclusions

1841
1842
Electrokinetic soil remediation is an emerging1843
in situ technology with demonstrated efficiency to1844
remediate fine-grained soils, and especially to re-
1845
Ž.
move heavy metals from the soil matrix Table 3 .
1846
According to that stated in the articles re-1847
viewed, it is possible to draw the following conclu-1848
sions on the main advantages of this technique:1849
18501851
ⅷ
Electrokinetics is very targetable to any speci-1852
fic location, because treatment of the soil oc-1853
curs only between two electrodes.18541855
ⅷ
Electrokinetics is able to treat contaminated1856
soil without excavation being necessary.
1857
1858
ⅷ
Electrokinetics is most effective in clay, be- 1859
cause it has a negative surface charge, and in 1860
soils with low hydraulic conductivity.
18611862
ⅷ
Electrokinetics is potentially effective in both 1863
saturated and unsaturated soils.

18641865
ⅷ
Electrokinetics is able to treat both organic 1866
and inorganic contaminants, such as heavy
1867
metals, nitrates, etc.
18681869
ⅷ
Electrokinetics demonstrated the ability to re- 1870
move contaminants from heterogeneous natu- 1871
ral deposits.
18721873
ⅷ
Good cost effectiveness. 1874
1875
Despite all the advantages, this technique has 1876
some limitations, which are: 1877
18781879
ⅷ
The solubility of the contaminant is highly 1880
dependent on the soil pH conditions.
18811882
ⅷ
The necessity to apply enhancing solution.
1883
Table 3
Conclusions on heavy metal removal from contaminated soils1884
Metal Remarks
Lead and Successful removal is obtained only under acidic conditions
Cadmium High removal rates were achieved with the use of HCl solution

Ž. Ž.
Chromium Significant part of Cr VI is reduced to Cr III if there are sulfides or other
reducing agents present in the soil
Low chromium migration was observed in the soil in the presence of sulfides
and no retardation in the soil with humic acid
Arsenic Sufficient arsenic removal is achieved only in alkaline conditions
Migration of arsenic is accelerated by an oxidising agent
y
Mercury Efficient mercury removal is achieved using I rI lixiviant solution
2
Higher removal efficiency is obtained using chloride or other suitable
component added to the soil
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx24
1885
1886
ⅷ
When higher voltage is applied to the soil, the1887
efficiency of the process decreases due to the1888
increased temperature.18891890
ⅷ
Removal efficiency is significantly reduced if1891
soil contains carbonates and hematite, as well1892
as large rocks or gravel.1893
1894
In order to guarantee efficient electrokinetic1895
remediation of soil, among other variables, it is1896

important to investigate physicochemical con-1897
taminant᎐soil interactions and the impact of en-1898
hancing agents on these interactions, the occur-1899
rence of reverse electroosmotic flow and the in-1900
fluence of organic substances present in the re-1901
mediated soil.
1902
1903
References
1904
19051906
Acar YB, Alshawabkeh AN. Principles of electrokinetic reme-1907
Ž.
diation. Environ Sci Technol 1993;27 13 :2638᎐2647.19081909
Acar YB, Alshawabkeh AN. Electrokinetic remediation. I.1910
Pilot-scale tests with lead-spiked kaolinite. J Geotech Eng1911
1996;March:173᎐185.19121913
Acar YB, Alshawabkeh AN. Principles of electrokinetic reme-1914
Ž.
diation. Environ Sci Technol 1997;27 13 .
19151916
Ž.
Acar YB, Hamed JT et al. Removal of cadmium II from
1917
saturated kaolinite by the application of electrical current’.1918
Ž.
Geotechniques 1994;44 2 :239᎐254.
19191920
Acar YB et al. Enhance soil bioremediation with electric1921
Ž.

fields. Chemtech 1996;26 4 :40᎐44.19221923
Baraud F, Tellier S, Astruc M. Ion velocity in soil solution1924
during electrokinetic remediation. J Hazard Mater 1997;
1925
56:315᎐332.19261927
Baraud F, Fourcade MC et al. Modelling of decontamination1928
rate in an electrokinetic soil processing. Int J Environ Anal1929
Chem 1998;68:105᎐121.19301931
Benazon N. Soil remediation. Hazard Mater Manage 1999;Oc-1932
toberrNovember.19331934
Chilingar GV et al. Electrobioremediation of soils contami-
1935
nated with hydrocarbons and metals: progress report. En-
1936
ergy Sources 1997;19:129᎐146.19371938
Cox CD et al. Electrokinetic remediation of mercury-con-
1939
taminated soils using iodineriodide lixiviant. Environ Sci
1940
Technol 1996;30:1933᎐1938.
19411942
Ž.
Hamed J, Acar YB, Gale RJ. Pb II removal from kaolinite
1943
Ž.
using electrokinetics. J Geotech Eng ASCE 1991:241᎐271.
19441945
Hansen HK, Ottosen LM et al. Electrodialytic remediation of
1946
soils polluted with Cu, Cr, Hg, Pb and Zn. J Chem Technol

1947
Biotechnol 1997;70:67᎐73.
19481949
Haran BS, Popov BN et al. Development of a new electroki-1950
netic technique for decontamination of hexavalent1951
chromium from low surface charged soils. Environ Prog1952
Ž.
1996;15 3 :166᎐172.
1953
1954
Hecho LI, Tellier S, Astruc M. Industrial site soils contami- 1955
nated with arsenic or chromium: evaluation of the elec- 1956
trokinetic method. Environ Technol 1998;19:1095᎐1102.
19571958
Hicks RE, Tondorf S. Electrorestoration of metal-con- 1959
Ž.
taminated soils. Environ Sci Technol 1994;28 12 :
1960
2203᎐2210.
19611962
Ho SV et al. Scale-up aspects of the Lasagna
TM
process for in 1963
situ soil decontamination. J Hazard Mater 1997;55:39᎐60.
19641965
Ho SV et al. The Lasagna technology for in situ soil remedia- 1966
tion. 1. Small field test. Environ Sci Technol 1999;33: 1967
1086᎐1091.
19681969
Lageman R. Electroreclamation: application in the Nether- 1970

Ž.
lands. Environ Sci Technol 1993;27 13 :2648᎐2650.
19711972
Li RS, Li LY. Enhancement of electrokinetic extraction from 1973
Ž.
lead-spiked soils. J Environ Eng 2000;126 9 :849᎐857.
19741975
Li ZH, Neretnieks I. Electroremediation: removal of heavy 1976
metals from soils by using cation-selective membrane. Env- 1977
iron Sci Technol 1998;32:394᎐397.
19781979
Ž. Ž.
Li ZH, Yu JW, Neretnieks I. Removal of Pb II , Cd II and
1980
Ž.
Cr III from sand by electromigration. J Hazard Mater 1981
1997;55:295᎐304.
19821983
Mattson ED, Lindgren ER. Electrokinetic Extraction of Chro- 1984
mate from Unsaturated Soils. American Chemical Society, 1985
1995:11᎐20.
19861987
Pamukcu S. Electrochemical technologies for in situ restora- 1988
tion of contaminated subsurface soils. http:rrgeotech. 1989
civen.okstate.edurejgerppr9703rindex.htm.
19901991
Pamukcu S, Wittle JK. Electrokinetic removal of selected 1992
Ž.
heavy metals from soil. Environ Prog 1992;11 3 :241᎐250.
19931994

Probstein RF, Hicks RE. Removal of contaminants from soils
1995
by electric fields. Science 1993;260:498᎐503.
19961997
Puppala SK, Alshawabkeh AN, Acar YB, Gale RJ, Bricka M. 1998
Enhanced electrokinetic remediation of high sorption ca- 1999
pacity soils. J Hazard Mater 1997;55:203᎐220.
20002001
Reddy K, Chinthamreddy S. Electrokinetic remediation of
2002
heavy metal-contaminated soils under reducing environ- 2003
ments. Waste Manage 1999;19:269᎐282.
20042005
Reddy K, Parupudi US et al. Effects of soil composition on 2006
the removal of chromium by electrokinetics. J Hazard 2007
Mater 1997;55:135᎐158.
20082009
Reddy K, Donahue M, Sasaoka R. Preliminary assessment of 2010
electrokinetic remediation of soil and sludge contaminated 2011
with mixed waste. Air Waste Manage Assoc 2012
1999;49:823᎐830.
20132014
Ribeiro AB et al. A dynamic model for the electrokinetic 2015
removal of copper from polluted soil. J Hazard Mater 2016
1997;56:257᎐271.
20172018
Sah JG, Chen JY. Study of the electrokinetic process on Cd- 2019
and Pb-spiked soils. J Hazard Mater 1998;58:301᎐315.
20202021
Sengupta AK. Hazardous and industrial wastes. Proceedings

2022
of the 27th Mid-Atlantic Industrial Waste Conference,
2023
1995:824᎐834.
20242025
Sims RC. Soil Remediation Techniques at Uncontrolled Haz- 2026
ardous Waste Sites. Utah State University, 1990:704᎐732.
20272028
Sogorka DB et al. Emerging technologies for soils contami- 2029
nated with metals ᎏ electrokinetic remediation. Hazard 2030
Ind Wastes 1998;30:673᎐685.
Uncorrected Proof
ARTICLE IN PRESS
()
J. Virkutyte et al. r The Science of the Total En
¨ironment xxx 2001 xxx᎐xxx 25
2031
2032
Thevanayagam S, Rishindran T. Injection of nutrients and2033
TEAs in clayey soils using electrokinetics. J Geotech2034
Geoenviron Eng 1998;April:330᎐338.20352036
US EPA. Revised guidance document for the remediation of2037
contaminated soils. US Environmental Protection Agency2038
Annual Review, 1998.20392040
van Cauwenberghe L. Electrokinetics: Technology Overview2041
Report. Groundwater Remediation Technologies Analysis2042
Centre, 1997:1᎐17.20432044
Vane LM, Zang GM. Effect of aqueous phase properties on2045
clay particle zeta potential and electro-osmotic permeabil-
2046

ity: implications for electrokinetic soil remediation 2047
processes. J Hazard Mater 1997;55:1᎐22.
2048
2049
Viadero RCJr et al. A laboratory-scale study of applied volt-
2050
age on the electrokinetic separation of lead from soils. Sep 2051
Ž.
Sci Technol 1998;33 12 :1833᎐1859.
2052
2053
Yeung A et al. Physicochemical soil ᎏ contaminant interac- 2054
tions during electrokinetic extraction. J Hazard Mater 2055
1997;55:221᎐237.
2056
2057
Zelina JP, Rusling JF. Electrochemical remediation of soils, 2058
1999:532᎐539.
2059