Chemosphere 64 (2006) 415–422

www.elsevier.com/locate/chemosphere

In situ bioelectrokinetic remediation of phenol-contaminated soil

by use of an electrode matrix and a rotational operation mode
a,b,*
Qishi Luo , Hui Wang a, Xihui Zhang a,c
, Xiangyu Fan a, Yi Qian a

a
Department of Environmental Science and Engineering, Tsinghua University, Beijing 100084, PR China
b
Shanghai Academy of Environmental Sciences, Shanghai 200233, PR China
c
Research Center for Environmental Engineering and Management, Shenzhen Graduate School of Tsinghua University,
Shenzhen 518055, PR China

Received 13 October 2005; received in revised form 22 November 2005; accepted 23 November 2005
Available online 10 January 2006

Abstract

In situ bioremediation is a safe and cost-effective technology for the cleanup of contaminated sites, but its remediation rate is usually
very slow. This study attempted to accelerate the process of bioremediation by employing non-uniform electrokinetic transport processes
to mix organic pollutants and degrading bacteria in soils under in situ conditions (namely, in situ bioelectrokinetic remediation) by use of
an electrode matrix and a rotational operation mode. A bench-scale non-uniform electrokinetic system with periodic polarity-reversal
was developed for this purpose, and tested by using a sandy loam spiked with phenol as a model organic pollutant. The results demon-
strated that non-uniform electrokinetic processes could enhance the in situ biodegradation of phenol in the soil, the efficiency of which
depended upon the operational mode of the electric field. Compared with the unidirectional operation and the bidirectional operation,
the rotational operation could effectively stimulate the biodegradation of phenol in the soil if adopting appropriate time intervals of
polarity-reversal and electrode matrixes. A reversal interval of 3.0 h and a square-shaped electrode matrix with four electrode couples
appeared appropriate for the in situ biodegradation of phenol, at which a maximum phenol removal of 58% was achieved in 10 d
and the bioremediation rate was increased about five times as compared to that with no electric field applied. The results also showed
that adopting a small polarity-reversal interval and an appropriate electrode array could produce a high and uniform removal of phenol
from the soil. It is believed that in situ bioelectrokinetic remediation holds the potential for field application.
 2005 Elsevier Ltd. All rights reserved.

Keywords: Bioelectrokinetic remediation; In situ bioremediation; Non-uniform electrokinetics; Electrode matrix; Electrokinetic transport; In situ bio-
degradation; Phenol

1. Introduction 1998). However, extended treatment time under field con-

In situ bioremediation holds the highest potential for the
cleanup of contaminated sites with organic pollutants due
to its safety and cost-effectiveness. It neither alters the
intrinsic soil properties nor does it produce secondary pol-
lution in the process of remediation (Yeom and Ghosh,
*
Corresponding author. Address: Shanghai Academy of Environmental
Sciences, 508 Qinzhou Road, Shanghai 200233, PR China. Tel.: +86 21
64085119x2317; fax: +86 21 64840680.
E-mail addresses: qsluo99@yahoo.com.cn, luoqishi@tsinghua.org.cn
(Q. Luo).
ditions is a major disadvantage. It often takes many
months or even several years to complete the remediation.
The slow remediation rate may be attributed to many
factors such as low soil temperature, limited nutrient sup-
ply, and low bacterial activity. However, it is commonly
believed that the slow rate results primarily from the lim-
ited opportunities of interactions among contaminants,
bacteria, and bacterial nutrients (Zhang et al., 1995;
Mohammed, 1996; Simoni et al., 2001). When organic pol-
lutants, especially hydrophobic compounds, are released
into subsurface environments, their mobility is greatly
reduced, typically through absorption on solid surfaces,

0045-6535/$ - see front matter  2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.chemosphere.2005.11.064
416 Q. Luo et al. / Chemosphere 64 (2006) 415–422
partitioning into soil organic matter, sequestration within
tortuous micropores, and/or formation into non-aqueous-
phase liquids (NAPL) (Pignatello and Xing, 1996; Juhasz
et al., 2000). On the other hand, the degrading bacteria
tend to adhere to soil particle surfaces forming continuous
or/and discontinuous biofilms; only a very small propor-
tion of the total bacteria exists in soil liquids (Rittmann
and McCarty, 2001). Consequently, the mass transfer and
the opportunities of contact and interaction between the
pollutants and the degrading cells are very limited due to
the slow movement between them. Previous studies have
supported this concept by demonstrating that the degrada-
tion rate for the well-mixed ex situ systems are often orders
of magnitude faster than those of undisturbed systems
(Dragun, 1998).
The slow movement of organics and bacteria in soils is
traditionally overcome by applying surfactants to dissolve
NAPL and desorb organics into soil liquids, or by employ-
ing hydraulic gradients to mobilize organics and bacterial
cells through the soil matrix (Saba et al., 2001). However,
these techniques are not effective in most cases due to the
heterogeneity of the soil matrix and the shortcomings of
hydraulically driven transport processes. In addition, the
energy consumption and costs associated with these tech-
niques are considerable.
Electrokinetically induced transport processes may pro-
vide an alternative solution. When a direct current electric
field is imposed upon the soil matrix, various transport
processes, including electroosmosis, electromigration,
electrophoresis and dielectrophoresis, can be induced (Als-
hawabkeh and Bricka, 2000). Electroosmosis is the mobili-
zation of pore fluids in an electric field, usually from the
anode toward the cathode; electroosmotic flow is able to
drive the free-phase dissolved and even sorbed organics
toward cathode regions (Acar et al., 1992). Electromigra-
tion is the movement of ionic species such as heavy metal
ions and NO 3 in an electric field; cations move towards
the cathode while anions move towards the anode. Electro-
phoresis is the transport of charged particles, including clay
particles and bacterial cells, toward the electrode of oppo-
site polarity, while dielectrophoresis is the translational
motion of neutral matter and bacterial cells toward stron-
gest field regions due to polarization effects in a non-uni-
form field regardless of its direction (Pohl, 1978). These
transport mechanisms can cause the movement of organic
molecules, bacterial cells and nutrients, and pore fluids
through the soil matrix (Luo, 2005), and thus may have
the potential for accelerating the mass transfer and interac-
tions among organic pollutants, bacterial cells and nutri-
ents during in situ bioremediation.
An integrated approach, here termed in situ bioelec-
trokinetic remediation, is generated by incorporating elec-
trokinetics into bioremediation to enhance the in situ
biodegradation of contaminants in soils. In earlier studies,
it was carried out mainly for the purpose of employing elec-
trokinetic transport processes to deliver bacterial cells and
nutrients, electron acceptors/donors, and other additives
like chelating agents and surfactants to stimulate in situ
bioremediation. These additives can be transported with
a precise directionality through soils by electrokinetics.
Lee and Lee (2001) and Jackman et al. (2001) studied the
transport of bacterial cells by applying electrokinetics to
degrade the diesel in sandy loam and the 2,4-dichlorephen-
oxyacetic acid in silt soil, respectively. In their studies, elec-
trophoresis was considered as the main mechanism of cell
transport. It has been demonstrated that, NO 2
3 , SO4 and
þ
NH4 as nutrients or electron acceptors can be injected into
soils through electromigration (Acar et al., 1997; Luo et al.,
2004a). Organic pollutants including phenol, 2,4-dichloro-
phenol and phenanthrene can be desorbed and moved
through the soil matrix by use of electrokinetic processes,
the effect of which depends upon the soil properties, the
type and operational mode of the electric field, and the
application of surfactants (Saichek and Reddy, 2003;
Luo, 2005).
The control of soil pH is of significance for in situ bio-
electrokinetic remediation. When an electric field is applied
to wet soils through inert electrodes, acid fronts (H+) and
base fronts (OH) will be generated at the anode and cath-
ode, respectively, due to electrolysis of pore fluids. These
acids and bases will advance through the soil matrix and
change soil pH. The change of soil pH will affect the elec-
trokinetic transport processes, the activity and viability of
bacterial cells, and the sorption and biodegradation of
organics in soil (Hamed and Bhadra, 1997; Luo et al.,
2004c). In order to maintain a proper pH value for bio-
electrokinetic remediation, measures must be taken to
avoid the adverse effects of electrolysis reactions. Applying
pH conditioning agents, constructing modified reactors,
and circulating electrolyte solutions in electrode compart-
ments are the common strategies (Lee and Yang, 2000).
However, reversing the polarity of the electric field seems
a simple solution since the H+ and OH ions generated
at the electrodes may be automatically neutralized in this
way. Reversing the polarity at appropriate intervals could
keep the soil pH and moisture levels (Luo et al., 2005a).
In addition, adopting an appropriate electric field is also
helpful to avoid extreme changes of soil pH. It has been
demonstrated that non-uniform electric field has advanta-
ges in maintaining soil pH as compared to uniform electric
field (Luo et al., 2004b).
The present studies in this research team use non-uni-
form electric fields as an in situ mixing tool. They are car-
ried out only for the purpose of using non-uniform
electrokinetic transport processes induced by the applied
non-uniform electric field (namely, non-uniform electroki-
netics) to mobilize and mix the organics and bacteria in
soils under in situ conditions, rather than injecting any
additives. Therefore, it is a new strategy as compared to
the earlier studies in the field of bioelectrokinetic remedia-
tion. The feasibility of this strategy has been reported pre-
viously by Luo et al. (2005a). In their study, the effects of
four operational modes (including unidirectional, bidirec-
tional, continuous, and intermittent operation) and two
 Q. Luo et al. / Chemosphere 64 (2006) 415–422 417

bacterial inoculation methods (including local and whole DC power supply

inoculation) on the biodegradation of phenol in sandy Electrode control system

loam were tested by using one pair of electrodes and rect- Monitoring system

angle-shaped electrokinetic reactors. In order to examine

Side +B
the applicability of this technique in field conditions, this 3cm 3cm 3cm 3cm 3cm 3cm

4cm
study further investigates the effects of rotational operation
Middle

12 cm
on the in situ biodegradation of phenol in the soil under the
condition of whole inoculation by using electrode matrixes

4cm
Columnar electrode
with couples of electrodes, and square-shaped electroki-
Side -B
netic cells.
(a) 24 cm

2. Materials and methods 4cm 4cm 4cm 4cm 4cm 4cm

4cm
2.1. Experimental soil, contaminant and bacteria a
4cm

A natural sandy loam was used as the experimental soil b

4cm
with various characteristics as described by Luo et al.
c
(2004a). The soil was collected from the topsoil layer (0–
4cm
30 cm) of a woodland, passed through a 2-mm sieve, ster- d
Columnar electrode

ilized three times by alternately using an autoclave 4cm

(121 C for 45 min) and a drier (105 C for 30 min), and e
stored in a desiccator for later tests. 4cm A B C D E
(b)
Phenol was selected as the tested organic pollutant. The
choice of phenol was based on its prevalence in the envi- 4cm 4cm 4cm 4cm 4cm 4cm
ronment and its biodegradability in a wide range of con-
4cm
centrations (Farooq et al., 1996). a
A mixed culture of phenol-degrading bacteria was used 4cm
as the experimental bacteria. The bacteria were isolated b
from a contaminated soil with petrochemicals by using 4cm

basic mineral medium with phenol as the sole carbon source c

4cm
with components as described by Luo (2005). The phenol- Columnar electrode
d
degrading ability was confirmed through degradation tests
4cm
in a phenol solution. The bacterial cells were cultivated in e
the medium on a shaker (30 C, 150 rpm), harvested in 4cm A B C D E
exponential growth phase by centrifugation, and resus- (c)
pended in sterilized deionized water to obtain a highly con- Fig. 1. Schematic of experimental setup and sampling positions. Testing
centrated bacterial suspension (1.2 · 1010 CFU ml1; CFU: system 1 with one pair of electrodes (a); testing system 2 with electrode
colony forming units) for the tests. array of two pairs (b) and four pairs (c) of electrodes. The dashed lines and
the interjunctions between them denoted sampling lines and positions,
respectively.
2.2. Experimental systems

Two sets of testing systems were used in this study. The
first was used to optimize the polarity-reversing interval
for the tests. It consisted of a soil cell, a pair of electrodes,
an electrode control system, an electric current and voltage
real-time monitoring system, and a power supply (Fig. 1a).
The soil cell was made of rectangle-shaped Perspex with an
inner size of length 24 cm, width 12 cm and height 10 cm.
One pair of column-shaped graphite electrodes, length
12 cm and diameter 0.5 cm, were used to generate a non-
uniform electric field. The electrode control apparatus
was capable of reversing the polarity of the electric field,
thus allowing changes in the operational mode for the tests.
The monitoring system could monitor electric current and
voltage on-line and stored them into a personal computer
for later analysis. The power supply could provide a con-
stant direct current electric voltage in a range from 0 to
60 V for the electrokinetic tests.
The second system was used to determine the optimal
electrode configuration for the in situ biodegradation of
phenol. It was the same as the first system, except for the
soil cell and the electrodes. A square-shaped soil cell with
a size of length 24 cm, width 24 cm and height 10 cm was
used, two or four pairs of electrodes were installed symmet-
rically into the soil bed to form the electrode matrixes as
shown Fig. 1b and c, respectively.
2.3. Experimental procedures
The contaminated soil was artificially spiked with phe-
nol by blending a mixed solution of phenol and degrading
418 Q. Luo et al. / Chemosphere 64 (2006) 415–422
bacteria into the sterilized soil and adjusting the moisture
level. The moist soil was then placed into the soil cell by
layers. Each layer was pressed down using a Perspex pestle
and vibrated so that the amount of void space was mini-
mized. And the soil specimen was compacted for 12 h at
a pressure of 0.1 kg cm2. The extruded pore fluids were
removed from the surface layer using bibulous paper.
Before the reactor assembly was conducted, a small frac-
tion of the soil specimen was obtained to determine the
initial content of phenol and degrading bacteria, and mois-
ture, since a portion of phenol and water could volatilize or
be consumed by the bacteria, and bacteria may die or grow
during the preparation. Two types of soil beds, with vol-
umes of length 24 cm, width 12 cm, height 2.5 cm, and
length 24 cm, width 24 cm, height 2.5 cm, were prepared
for testing system 1 and 2, respectively. The characteristics
of the moist soil beds are listed in Table 1.
Pairs of electrodes were symmetrically inserted into the
soil beds to form the electrode matrixes as shown in
Fig. 1, with a distance of 2 cm from the soil cell wall. Power
supply, electrode control apparatus, and monitoring equip-
ment were then connected to the electrodes. Once the assem-
bly was completed, the electrokinetic cell was enclosed with
a Perspex cover to prevent the soil bed from excessive evap-
oration of water and phenol.
Only a specific range of electric current or field strength
can be used to manipulate the movement of bacteria
through soils. A previous study demonstrated that a cur-
rent of no more than 20 mA has little effect on the cell sur-
face properties, degrading activity and viability of the
mixed bacterial cells tested in this study (Luo et al.,
2005b). If this current strength is imposed on the tested
soil, the electric gradient should be less than 2.0 V cm1,
or the current density less than 0.67 mA cm2 (Luo,
2005). Therefore, a constant electric field gradient of
1.0 V cm1 was applied through electrodes in this study.
For the tests in system 1, the electric field was applied
unidirectionally or bidirectionally, while for the tests in sys-
tem 2, the electric field was applied through an electrode
array in a rotational way at an optimal polarity-reversing
interval, which was determined by using the system 1. Dur-
ing rotational operation, only a couple of electrodes are
electrified at each moment. The rotational operation mode
when using the electrode array of four couples and an
interval of 3 h is shown schematically in Fig. 2. Control
Table 1
The main characteristics of the moist soil beds
Parameters Value
Total volume (cm3) 720 (for system 1)
1440 (for system 2)
Bulk density (g cm3) 2.29
Water saturation (%) 53.0
Effective porosity (%) 42.0
Water content (%) 17.5
Soil (pH) 7.7
Phenol content (mg kg1) 180.2
Bacterial number (log CFU g1) 9.18
tests with no electric field applied or no bacteria inoculated
were run in parallel. During the tests, the electric voltage
and current through the system were recorded every
15 min by using the on-line monitoring system.
At the end of each test, the soil beds were destructively
sampled using a U-shaped sampler and a spatula to deter-
mine the final phenol concentrations, bacterial numbers
and moisture levels. For the tests in system 1, three sam-
pling lines were arranged along the middle line in which
the two electrodes were located, and along the two sidelines
(side +B and B) with a distance of 4 cm from the middle.
Seven sampling points with distances of 1, 4, 7, 10, 13, 16
and 19 cm from the (initial) anode were taken on each line,
as shown in Fig. 1a. For the tests in system 2, five sampling
lines, with a distance of 4 cm from each other and from the
cell wall, were arrayed lengthways and widthways, respec-
tively, and the intersections (25 positions in total) between
two lines were sampling positions, as shown in Fig. 1b
and c.
2.4. Analytical methods
Methylene chloride was used to extract phenol from the
soil samples. Approximately 1.0–1.5 g of soil sample was
mixed with 5 ml of methylene chloride in a Teflon-sealed
glass vial, extracted by sonication, and filtered through
0.45 lm Teflon membranes. Each soil sample was extracted
three times, and the filtrate was combined. The phenol con-
centration in the filtrate was analyzed by high performance
liquid chromatography (HPLC) (Hewlett Packard, 1050) at
275 nm using a C18 reverse phase column (Agilent, Zorbax
extend-C18) and a mobile phase containing methanol and
2% (v/v) acetic acid (60:40, v/v). The HPLC was calibrated
using four external standards prior to performing analyses.
About 0.5 g of soil sample was taken to examine the bac-
terial distribution across the soil bed. CFU were counted by
preparing serial dilutions of soil suspensions in a sterilized
solution of 0.1% sodium pyrophosphate (w/v), plating onto
an agar plate, and incubating at 30 C for 2 d. In addition,
the water content was determined using the methods
described by Lu (2000).
All the analyses were performed in triplicate, and the
results were calculated as the average. The phenol content
and the amount of bacteria in the soil were expressed in
units of mg-phenol and log CFU per kg-dried-soil, respec-
tively. In order to directly reflect the relative change of phe-
nol in the soil, all the results are presented as the ratio of
the final phenol content (c) to the initial content (c0).
2.5. Quality control and mass balance
The Perspex soil cells and the graphite electrodes were
all new and specifically constructed for the tests. All the
equipment and testing systems were calibrated prior to
use. All the glassware including vials and sample bottles
were verified to be clean. The chemicals used as solvents
or in analyses were fresh and high grade purity.
 Q. Luo et al. / Chemosphere 64 (2006) 415–422
E 2 (+)
3 hours later
(-) E2
At the beginning
E1 (+) (-) E 1
(-) E 4
21 hours later
E 4 (+)
Fig. 2. Schematic of rotational operation at an interval of 3 h through an electrode array with four couples for one operation cycle. The electrified anode
(d) and cathode ( ); the un-electrified electrodes (s).
Phenol was used as an external standard for calibration.
Sample blanks and standard samples were injected regu-
larly into HPLC to ensure that the system remained uncon-
taminated and to certify a uniform response. After the tests
were completed, a mass balance was conducted for the phe-
nol in the soil cells. Over the test period of 10 d, 87–96% of
the initial phenol was recovered from the soil, indicating
there was an overall mass recovery of phenol before and
after the test. The discrepancies in mass balance of phenol
might be caused by volatilization throughout the test, and
by adsorption to the experimental equipment, such as the
Plexiglas cells, electrodes and sample vials.
Bacterial cells were analyzed by using standard method
(e.g., counting the CFU after incubating). All the glass-
ware, inoculating compartment, diluting solution, and
incubator were thoroughly sterilized and verified to be ster-
ile. All the analyses were performed in triplicate to mini-
mize testing errors.
3. Results and discussion
3.1. Biodegradation of phenol under unidirectional
operation
A saturated sorption amount of approximately 240 mg-
phenol kg-dried-soil1 was previously reported for the soil
(Luo et al., 2005c). In this study, the actual initial content
of phenol was about 180 mg kg1, much less than the satu-
rated sorption amount. The movement of phenol at this
content requires desorption of phenol from the soil particle
surface since most of the phenol is absorbed at this concen-
419
E3 (+) (+) E4
6 hours later 9 hours later
E3 (-) E 4 (-)
E1 (-) (+) E1
12 hours later
E3 (-) E 2 (-)
18 hours later 15 hours later
E3 (+) (+) E 2
tration (Acar et al., 1992). At the beginning of test, the phe-
nol in the spiked soil was uniformly distributed. However,
after applying 1.0 V cm1 unidirectionally for 10 d (not
adding bacterial cells), the phenol was concentrated to
the region of 4–10 cm from the cathode (Fig. 3a). This sug-
gested that the applied electric field could desorb and drive
the phenol through the soil towards the cathode.
When the phenol and the degrading bacteria were both
homogeneously added into the soil, a maximum opportu-
nity of contact and interaction between them might exist.
With no application of the electric field (control test), only
about 9% of the phenol was reduced after 10 d, suggesting
the mass-transfer between the phenol and the bacteria was
still very slow even if they were both evenly distributed
throughout the soil. After applying 1.0 V cm1 in one
direction for 10 d, however, the phenol was reduced on
average by 46%, about four times greater than the control
test (Fig. 3a). This indicated that the electric field acceler-
ated the in situ biodegradation of phenol in the soil.
The phenol removal from the soil beds might result from
electrolysis, volatilization, and biodegradation. Over the
test period of 10 d, a maximum current of 12 mA (current
density of 0.042 mA cm2) was observed. Under such a
field, the electrolyzed phenol would be negligible, as dem-
onstrated by Luo (2005). In addition, the electric current
(far less than 5.0 mA cm2) and the experiment duration
(10 d) were low enough to neglect the fluctuations of soil
temperature (Baraud et al., 1999), and hence the volatilized
phenol could be considered equal for different tests due to
the covering. Therefore, the discrepancy in removal effi-
ciency might primarily result from the biodegradation
420 Q. Luo et al. / Chemosphere 64 (2006) 415–422

1.6 the cathode after 10 d. This implied that the phenol in the
1.4 soil beds could not be uniformly removed by using unidi-
1.2 rectional operation.
Phenol / c c0-1

1.0
0.8 3.2. Biodegradation of phenol under bidirectional operation
0.6
0.4
When reversing the polarity every 1.5, 3.0 and 12.0 h
and running for 10 d, the phenol in the middle region
0.2
was reduced on average by 68%, 60%, and 49%, respec-
(a) 0.0
tively. The phenol was relatively uniformly removed at
1.2 intervals of 1.5 and 3.0 h, while at an interval of 12.0 h,
the phenol remained higher near both electrodes and lower
1.0
in the middle cell (Fig. 3b). These revealed that polarity-
reversal enhanced significantly the biodegradation of phe-
Phenol / c c0-1

0.8

0.6
nol in the soil, and a smaller interval induced a higher
and more uniform phenol removal from the soil. According
0.4 to the results, an interval of no more than 3.0 h seemed
0.2 appropriate for the tests.
(b) 0.0
Reversing the polarity of the electric field could cause
the phenol and the bacteria to move back and forth
1.2 through the soil. Luo (2005) demonstrated that the phenol
could be moved through the tested soil towards the cathode
1.0
at a rate of 1.5 cm d1 V1, while the degrading bacteria
0.8 towards the anode at 6.9 cm d1. When reversing the polar-
Phenol / c c0-1

ity at intervals of 1.5, 3.0 and 12.0 h, the movement direc-

0.6
tion of phenol would be artificially changed after moving
0.4 distances of 0.09, 0.19 and 0.75 cm, respectively, while
0.2 the bacteria would be changed after moving 0.43, 0.86
and 3.45 cm, respectively. The larger the time interval,
(c) 0.0
the further the movement distance before turning around,
10.3 suggesting that a larger interval might cause the phenol
or bacteria to concentrate in specific regions in the soil.
Bacteria / logCFU g-1

10.1 However, repeatedly reversing the polarity at a smaller

10.0
9.9
9.8
9.7
0 4 8 12 16
(d) Distance from the anode / cm
Fig. 3. The change of phenol in the middle of the soil cell under (a)
unidirectional and (b) bidirectional operation, and the spatial distribution
of (c) phenol and (d) bacteria in the soil bed at an interval of 1.5 h (10 d,
system 1). Adding bacteria but not applying electric field (s); applying
electric field but not adding bacteria (d); adding bacteria and applying
electric field (m); adding bacteria and applying electric field at intervals of

12.0 h (h), 3.0 h (j), and 1.5 h ( ); sampling lines of side +B (), side B
(), and middle (n).
process. The bacteria could be moved by the applied electric
field toward the anode (Luo, 2005), in a direction opposite
to the movement of phenol. During this movement, more
opportunities for interaction could be generated.
However, under unidirectional operation, the phenol
remained very high in local regions (Fig. 3a); about 86%
of the initial phenol remained in the region of 4–7 cm from
interval could cause the phenol and bacteria to move back
and forth in initial locations.
Compared with the control test (not applying electric
field), the phenol in the peripheral regions also dropped sig-
nificantly (Fig. 3c). In the two side areas of +B and B,
about 52% of the initial phenol was reduced after operation
20 for 10 d at interval of 1.5 h, indicating that the two periph-
eral regions were still covering by the electric current
despite a distance of 4 cm from the electrode. This finding
suggests that a larger site could be treated at the cost of
fewer electrode materials by applying non-uniform electric
field through tubular electrodes.
However, the phenol remained higher in the peripheral
regions than in the middle (Fig. 3c). After operation for
10 d at an interval of 1.5 h, the remaining phenol in the
two peripheral regions was about 1.5 times that in the mid-
dle. This indicated that applying electric field bidirection-
ally through one pair of electrodes could not evenly
remove the phenol from the whole soil beds, as was the case
under unidirectional operation (Fig. 3a).
In contract to the phenol, the degrading bacteria were
mainly distributed in the middle of each soil cell, especially
in the two electrodes regions (Fig. 3d). After operation for
10 d at an interval of 1.5 h, the bacterial numbers in the
 Q. Luo et al. / Chemosphere 64 (2006) 415–422 421

middle region were 1.6 times those in the peripheral Range of c c0-1
regions. This might be attributed to the dielectrophoretic
0.80-1.00
movement of bacteria in a non-uniform electric field.
0.60-0.80
Reversing the polarity of non-uniform electric field could 1.0
disperse the phenol throughout the soil beds (Luo et al., 0.40-0.60
0.8
2005c), but might cause the bacteria to move through

Phenol / c c0-1
0.20-0.40
dielectrophoresis towards the two electrodes regions with 0.6
0.00-0.20
higher electric density, regardless of the direction of electric 0.4
a
field applied (Betts, 1995). 0.2 b e
li n
c
m ple
0.0 d Sa
3.3. Biodegradation of phenol under rotational operation (a) A B C D E e

In order to evenly remove phenol from whole soil beds Range of c c0-1
using bioelectrokinetic remediation, the electrode array 0.80-1.00
and rotational operation mode were tested by using system 0.60-0.80
1.0
2. Two and four pairs of electrodes were arrayed in the soil
0.40-0.60
beds as shown in Fig. 1b and c, respectively, and electric 0.8

Phenol / c c0-1
gradient of 1.0 V cm1 was applied alternately through 0.6
0.20-0.40

electrode couples at a reversal interval of 3 h in a clockwise 0.00-0.20

0.4
direction as shown in Fig. 2. a
When no electric field was applied (control test), the phe- 0.2 b
in e
c lel
nol in the soil beds reduced approximately 10% after 10 d mp
0.0 d Sa
(Fig. 4a). However, when using an electrode array of two (b) A B C D E e
couples, the phenol was reduced on average by 51%, about
Range of c c0-1
four times higher than the control test. Nevertheless, there
existed some dead zones, in which up to 75–83% of the initial 0.80-1.00

phenol remained, and the difference in phenol removal 1.0 0.60-0.80

between sampling positions was up to 50% (Fig. 4b), suggest- 0.40-0.60
0.8
Phenol / c c0-1

ing that this electrode array was not effective. When using the 0.20-0.40
electrode array of four couples, the phenol in the whole soil 0.6
0.00-0.20
beds was reduced on average by 58%, about five times higher 0.4
than control test. Importantly, no dead zones existed in this a
0.2 b
case; the difference in phenol removal efficiency was no more c line
le
than 14% (Fig. 4c). These findings indicated that the elec- 0.0 d mp
(c) A B C D E Sa
e
trode array of four couples might be efficient in evenly
removing the phenol from the whole soil beds. Fig. 4. Spatial distribution of phenol in the soil beds after rotational
When applying an electric field through the electrode operation for 10 d at interval of 3 h (in system 2). Not applying electric
matrix in a rotational way at a smaller time interval, the field (a); applying electric field through electrode arrays with two (b) and
four (c) pairs of electrodes.
electric field was capable of mixing the phenol and the
degrading bacteria in the soil under in situ condition.
Quickly rotating the polarity of electric field applied could study (Fig. 4). However, an optimal electrode matrix is
exert a rotating force on bacterial cells through electrorota- determined by many factors such as soil properties, space
tion (Pethig and Huang, 1995). This force might partly between electrodes, electric field strength applied, opera-
contribute to the faster biodegradation of phenol in the soil tional mode, and so on (Alshawabkeh et al., 1999). There-
beds under rotational operation. fore, an appropriate electrode array for a specific site
The polarity-reversing interval would be a critical factor should be obtained according to the actual conditions.
for effective removal of phenol. The smaller the interval, Theoretically, an optimal electrode matrix could be easily
the more uniform and more efficient the phenol removal, scaled up to a field process by repeating the matrix in a cer-
as described by Fig. 3b. However, reversing the polarity tain way. From this perspective, the bioelectrokinetic reme-
could increase the electricity consumption for the bioelec- diation holds significant potential for field application.
trokinetic system; the smaller the interval of polarity-rever-
sal, the greater the energy consumption (Luo et al., 2005a). 4. Conclusions
Therefore, an optimal interval should be selected for the
bioelectrokinetic remediation. (1) Non-uniform electrokinetic processes could accelerate
Electrode matrix with a specific array of electrodes the in situ bioremediation of phenol-contaminated
would be another critical factor. The electrode array of soil, the efficiency of which depends upon the opera-
four couples appeared appropriate for the soil tested in this tional mode. Compared with the unidirectional and
422 Q. Luo et al. / Chemosphere 64 (2006) 415–422
bidirectional operation, the rotational operation could
effectively stimulate the in situ biodegradation of phe-
nol in the soil if adopting an appropriate electrode
matrix and a proper time interval of polarity-reversal.
(2) A reversal interval of 3.0 h and a square-shaped elec-
trode matrix with four electrode couples appeared
desirable appropriate for the biodegradation of phe-
nol in the soil, at which a maximum phenol removal
of 58% was achieved in 10 d and the bioremediation
rate was increased about five times. Adopting a small
polarity-reversal interval and an appropriate elec-
trode array could produce a high and uniform
removal of phenol from the soil.
(3) As an in situ remediation technique, bioelectrokinetic
remediation holds high potential for field application.
Acknowledgements
The main work was completed in Department of Envi-
ronmental Science and Engineering at Tsinghua Univer-
sity, and supported partly by the National 973 program
of China (2004CB418506) and partly by the National 863
programs of China (2002AA601120, 2004AA649220).
References
Acar, Y.B., Li, H., Gale, R.J., 1992. Phenol removal from kaolin by
electrokinetics. J. Geotech. Eng. 118, 1837–1852.
Acar, Y.B., Rabbi, M.F., Ozsu, E.E., 1997. Electrokinetic injection of
ammonium and sulfate ions into sand and kaolinite beds. J. Geotech.
Geoenviron. Eng. 123, 239–249.
Alshawabkeh, A.N., Bricka, M., 2000. Basics and applications of
electrokinetic remediation. In: Wise, D.L., Trantolo, D.J., Cichon,
E.J., Inyang, H.I., Stottmeister, U. (Eds.), Remediation Engineering of
Contaminated Soils. Marcel Dekker Inc., New York, pp. 95–111.
Alshawabkeh, A.N., Gale, R.J., Ozsu-Acar, E., Bricka, R.M., 1999.
Optimization of 2-D electrode configuration for electrokinetic reme-
diation. J. Soil Contam. 8, 617–635.
Baraud, F., Tellier, S., Astruc, M., 1999. Temperature effect on ionic
transport during soil electrokinetic treatment at constant pH. J.
Hazard. Mater. 64, 263–281.
Betts, W.B., 1995. The potential of dielectrophoresis for the real-time
detection of microorganisms in foods. Trend Food Sci. Tech. 6, 51–58.
Dragun, J., 1998. Organic chemical biodegradation in soil. In: Calabrese,
E.J., Kostecki, P.T. (Eds.), The Soil Chemistry of Hazardous
Materials. The Hazardous Materials Control Research Institute,
Maryland, pp. 325–445.
Farooq, S., Nakhla, G.F., Essa, M.H., 1996. Simulation of in-situ
bioremediation of phenol contaminated sandy aquifers, 2. Effect of
phenol concentration. Water Air Soil Poll. 87, 283–295.
Hamed, J.T., Bhadra, A., 1997. Influence of current and pH on
electrokinetics. J. Hazard. Mater. 55, 279–294.
Jackman, S.A., Maini, G., Sharman, A.K., Sunderland, G., Knowles,
C.J., 2001. Electrokinetic movement and biodegradation of 2,4-
dichlorephenoxyacetic acid in silt soil. Biotechnol. Bioeng. 74,
40–48.
Juhasz, A.L., Megharaj, M., Naidu, R., 2000. Bioavailability: the major
challenge (constraint) to bioremediation of organically contaminated
soils. In: Wise, D.L., Trantolo, D.J., Cichon, E.J., Inyang, H.I.,
Stottmeister, U. (Eds.), Bioremediatiom of Contaminated Soils.
Marcel Dekker Inc., New York, pp. 217–241.
Lee, K.S., Lee, K., 2001. Bioremediation of diesel-contaminated soil by
bacterial cells transported by electrokinetics. J. Microbiol. Biotechnol.
11, 1038–1045.
Lee, H.H., Yang, J.W., 2000. A new method to control electrolyte pH by a
circulation system in electrokinetic soil remediation. J. Hazard. Mater.
B 77, 227–240.
Lu, R.K., 2000. Analytic Methods for Soil Agricultural Chemistry. China
Agricultural Science and Technology Press, Beijing, pp. 57–59.
Luo, Q.S., 2005. Study on in situ bio-electrokinetic remediation of soils
and its mechanism. Ph.D. thesis, Tsinghua University, Beijing, China,
pp. 31–102.
Luo, Q.S., Wang, H., Zhang, X.H., Qian, Y., 2004a. Movement and
transformation of nitrate in soil by non-uniform electrokinetics.
Environ. Sci. 25, 107–113.
Luo, Q.S., Zhang, X.H., Wang, H., Qian, Y., 2004b. Influence of non-
uniform electrokinetic remediation technology on soil properties.
Tech. Eq. Environ. Poll. Control 5, 40–45.
Luo, Q.S., Zhang, X.H., Wang, H., Qian, Y., 2004c. The movement and
its mechanism of genetic engineering bacteria in soils under non-
uniform electric field. China Environ. Sci. 24, 34–39.
Luo, Q.S., Zhang, X.H., Wang, H., Qian, Y., 2005a. The use of non-
uniform electrokinetics to enhance in-situ bioremediation of phenol-
contaminated soil. J. Hazard. Mater. 121, 187–194.
Luo, Q.S., Wang, H., Zhang, X.H., Qian, Y., 2005b. Effect of direct
electric current on the cell surface properties of phenol-degrading
bacteria. Appl. Environ. Microb. 71, 423–427.
Luo, Q.S., Zhang, X.H., Wang, H., Qian, Y., 2005c. Mobilization of
phenol and chlorophenol in unsaturated soils by non-uniform
electrokinetics. Chemosphere 59, 1289–1298.
Mohammed, N., 1996. State-of-the-art review of bioremediation studies.
J. Environ. Sci. Heal. A 31, 1547–1574.
Pethig, R., Huang, Y., 1995. Dielectrophoretic and electrorotation
behavior of cells: theory and experiment. In: Walz, D., Berg, H.,
Milazzo, G. (Eds.), Bioelectrochemistry of Cells and Tissues. Birkha-
user, Basel, pp. 265–300.
Pignatello, J.J., Xing, B., 1996. Mechanisms of slow sorption of organic
chemicals to natural particles. Environ. Sci. Technol. 30, 1–11.
Pohl, H.A., 1978. Dielectrophoresis—The Behavior of Neutral Matter in
Non-uniform Electric Fields. Cambridge University Press, Cambridge,
pp. 23–45.
Rittmann, B.E., McCarty, P.L., 2001. Environmental Biotechnology:
Principles and Applications. McGraw-Hill, New York, pp. 695–730.
Saba, T., Illangasekare, T.H., Ewing, J., 2001. Investigation of surfactant-
enhanced dissolution of entrapped non-aqueous phase liquid chemicals
in a two-dimensional groundwater flow field. J. Contam. Hydrol. 51,
63–82.
Saichek, R.E., Reddy, K.R., 2003. Effect of pH control at the anode for
the electrokinetic removal of phenanthrene from kaolin soil. Chemo-
sphere 51, 273–287.
Simoni, S.F., Schafer, A., Harms, H., Zehnder, A.J.B., 2001. Factors
affecting mass transfer limited biodegradation in saturated porous
media. J. Contam. Hydrol. 50, 99–120.
Yeom, I.T., Ghosh, M.M., 1998. Mass transfer limitation in PAH-
contaminated soil remediation. Water Sci. Technol. 37, 111–118.
Zhang, W., Bouwer, E., Wilson, L., Durant, N., 1995. Biotransformation
of aromatic hydrocarbons in subsurface biofilms. Water Sci. Technol.
31, 1–14.