Journal of the Taiwan Institute of Chemical Engineers 70 (2017) 252–259

Contents lists available at ScienceDirect

Journal of the Taiwan Institute of Chemical Engineers

journal homepage: www.elsevier.com/locate/jtice

A combination of reducing and chelating agents for electrolyte

conditioning in electrokinetic remediation of As-contaminated soil
So-Ri Ryu a, Eun-Ki Jeon a, Kitae Baek a,b,∗
a
Department of Environmental Engineering and Soil Environment Research Center, Chonbuk National University, 567 Baekje-daero, Deokjin-gu, Jeonju,
Jeollabukdo, Republic of Korea
b
Department of Bioactive Material Sciences, Chonbuk National University, 567 Baekje-daero, Deokjin-gu, Jeonju, Jeollabukdo, Republic of Korea

a r t i c l e i n f o a b s t r a c t

Article history: Arsenic (As)-mobility in soils is dependent on redox chemistry. Generally, As(III) is more mobile than
Received 26 May 2016 As(V), and reductive dissolution of Fe hydroxides promotes the release of As in soil. In this study,
Revised 13 September 2016
electro-kinetic remediation (EKR) coupled with a redox reaction was investigated for remediation of As-
Accepted 30 October 2016
contaminated soil. To reduce As(V) and Fe hydroxides, reducing agents such as ascorbic acid, oxalic acid,
Available online 3 December 2016
and dithionite were used. In addition, chelating agents including oxalate, EDTA, and citrate were used
Keywords: to prevent the formation of Fe precipitates, well-known sinks for As(III) and As(V). After application of
Arsenic reducing and chelating agents, NaOH was used to enhance the electroosmotic flow (EOF) because As(III)
Reducing agent is non-charged over a broad pH range (under pH 9.2). The reducing agents enhanced release of As in
Chelating agent the form of As(III) or As(V). Additionally, NaOH application increased the removal of As(III) and As(V) by
Electrokinetic remediation increasing the current and EOF. Ascorbic and oxalic acids removed only a small amount of As due to their
Secondary mineral
mild reducing potential. Even dithionite, a powerful reducing agent, was not effective for the reduction
of As(V) and Fe in the EKR because it was unstable under acidic conditions. Chelating agents did not
prevent the formation of secondary Fe minerals in this system because of the low stability of complexes
under acidic conditions. The overall removal efficiency of As was not exceeded 25%, therefore, precise pH
control is required to apply the combination of dithionite and chelating agent to EKR.
© 2016 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction as adsorption/desorption, precipitation/dissolution, and complex-

Arsenic (As) is a naturally occurring element that is abundant
in the earth’s crust. However, As contamination of soil has been
aggravated by industrial activities such as mining, smelting, and
combustion of fossil fuels. As is toxic to organisms, reduces soil
fertility, and influences plant growth, moreover, it causes human
diseases such as hyperpigmentation, keratosis, anemia, neuropa-
thy, burning sensation of eyes, solid swelling of legs, chronic
lung diseases, liver fibrosis and skin cancer [1–4]. Therefore,
various technologies including soil washing, pyro-metallurgical
extraction, soil flushing, and electrokinetic remediation (EKR) have
been applied to treat As-contaminated soils [1,5,6]. Among these
methods, EKR is effective even in fine-grained soils because the
primary removal mechanisms are independent of soil pore size
[7,8]. In EKR, removal of contaminants is basically achieved by
electro-migration and electro-osmotic flow (EOF) of charged and
uncharged pollutants, respectively. Additionally, phenomena such
∗
Corresponding author. Fax: +82 63 270 2449.
E-mail address: kbaek@jbnu.ac.kr (K. Baek).
ation reactions could also influence contaminant transport [9],
and various enhancements including electrolyte conditioning have
been used to enhance contaminant transport [8].
Acid conditioning has been effective in the removal of cationic
heavy metals; however, it was not effective for the removal of
oxyanionic As [8]. Although acidic conditioning can enhance the
release of As by dissolution of Fe-hydroxides associated with As,
oxyanionic As is easily re-adsorbed onto the positively charged
acidified soil surface [1,9]. The fate and transport of As is highly
dependent on the speciation of As, i.e. As(III) or As(V). Under ox-
idizing conditions, As(V) is the predominant species existing as
H2 AsO4 − (pH 2.2–6.9) and HAsO4 2− (pH 6.9–11.5) [7]. In a re-
ducing environment, As(III) is the predominant species existing as
H3 AsO3 (below pH 9.2) and H2 AsO3 − (pH 9.2–12) [1]. A positively
charged soil surface has high affinity for oxyanionic As, and the
particular speciation influences adsorption. In the natural environ-
ment, As(III) has weaker binding affinity and more mobility than
As(V) because of its neutral charge; therefore, the reduction of
As(V) to As(III) could prevent re-adsorption of extracted As. There-
fore, reducing condition is required to enhance reductive dissolu-
tion of Fe to facilitate As extraction from soil to soil pore. Oxalate,

http://dx.doi.org/10.1016/j.jtice.2016.10.058
1876-1070/© 2016 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
 S.-R. Ryu et al. / Journal of the Taiwan Institute of Chemical Engineers 70 (2017) 252–259 253

ascorbate, and dithionite have been used to achieve the reductive Table 1
Initial soil properties.
dissolution and/or reduce As(V) to As(III). Dithionite is a relative
strong reducing agent compared to others, and the chemical has Chemical Properties
stand reduction potential (Eo ) of −1.13 V or −1.12 V even in alka- Soil pH 5.5
line conditions, therefore, it was expected to work in EKR [2,10]. Soil EC (μS/cm) 88.6
However, As mobility in soils is highly dependent on Fe miner- Soil texture (wt%)
als such as ferrihydrite, lepidocrocite, and goethite as well as their Sand (50 μm–2 mm) 41
Silt (2 μm–50 μm) 32
speciation [9,11]. Fe minerals are crucial adsorption sites for As due
Clay (<2 μm) 27
to their abundance and high affinity [12]. Fe minerals predomi-
nantly retain oxyanionic As under oxidizing conditions; however, Concentration of heavy metals (mg/kg)

As bound to such minerals is released under reducing conditions As 193.5 25a

via reductive dissolution of Fe [11,13]. Cu 284.3 150
Pb 591.1 200
H2 AsO4 − + 3H+ + 2e− ↔ HAsO2 + 2H2 O (13) Cd 3.9 4
Zn 210.3 300
Ni 12.7 100
Fe(OH )3(s) + 3H+ + e− ↔ Fe2+ + 3H2 O (11) Fe 30,270.7 –

Reduction of As and/or Fe could increase the mobility of As a

which is complex with Fe minerals. However, reductively dissolved
Fe2+ and Fe3+ could form secondary Fe minerals such as goethite,
magnetite, green rust, siderite, and ferric arsenate, which act as
new adsorption sites for extracted As [2,6,14,15]. Thus, it is neces-
sary to prevent the formation of secondary Fe minerals to enhance
As removal from soil. Chelating agents forming complexes with
dissolved Fe2+ and Fe3+ could reduce or prevent the formation
of secondary Fe minerals when the reductive dissolution of Fe
mineral occurs [2,15]. Kim et al. (2015) applied various combina-
tion of reducing and chelating agents including oxalate, ascorbate,
dithionate, EDTA and citrate to extract As, and reported that the
combination of reducing and chelating agents could extract more
than 90% of As from soil and mine tailing, but sole application
of reducing agent was not effective due to the re-adsorption of
As onto the secondary Fe mineral formed in the solution [15].
Although As contamination of soil is highly related to redox
reactions, few studies applying redox chemistry to EKR have been
reported for As-contaminated soil, and those have been mainly
the combination of biotic reduction and EKR [8]. However, the
combination of chemical reduction and chelating agents is worthy
of application to EKR because such a method could release As
without re-adsorption [2,15]. In this study, we investigated the
applicability of the combination of reducing and chelating agents
in EKR to remediate As-contaminated soil. To evaluate the combi-
nation of reducing and chelating agents in EKR, ascorbate, oxalate,
and dithionite, were used for reducing agents and oxalate, citrate,
and EDTA were used for chelating agents. The agents were used
for electrolytes to improve mobility and extractability of As. The
mass balance of As was prepared to verifying the mobility of As.
2. Materials and methods
2.1. Soil sampling and characterization
Soil samples were collected from rice fields nearby a former
smelter located in Janghang-eup, Chungcheongnam-do, Republic
of Korea As (193.5 mg/kg), Cu (284.3 mg/kg), and Pb (591.1 mg/kg)
concentration of soil samples exceeded the Korean regulation level
(As: 25 mg/kg, Cu: 150 mg/kg, Pb: 200 mg/kg). Among the metals,
we focused on As because of the relatively high concentration and
widespread contamination in the site. The soil texture was clay
loam consisting of 27% clay, 41% sand, and 32% silt. Samples were
air-dried and passed through a No. 10 sieve (2.0 mm) before exper-
iments. Other soil characteristics are summarized in Table 1.
2.2. Experimental setup
A schematic diagram of the experimental setup is shown in
Fig. 1. The apparatus consisted of one soil chamber (20 × 4 ×
Soil Environment Conservation Act Standards for Arsenic in Republic of Korea.
4 cm3 ), two electrolyte chambers (5 × 4 × 4 cm3 ), two electrolyte
reservoirs (1 L), two peristaltic pumps, two electrodes (3.5 × 5.5 ×
0.1 cm3 ) and a power supply. After adjusting the water content to
30%, the soil sample was placed in the soil chamber. One liter of
electrolyte, circulated in the electrolyte chamber and reservoir by a
peristaltic pump, was replaced every two days. Mesh type iridium-
coated titanium oxide electrodes (3.5 × 5.5 × 0.1 cm3 ) were placed
in each electrolyte chamber. Direct current was supplied with a
voltage gradient of 1 V/cm. Reducing agents (such as ascorbic, ox-
alic acids and dithionite) and chelating agents (such as EDTA, cit-
rate and oxalate) were used for electrolyte conditioning. After ap-
plication of reducing or chelating agents for 22 days (EKR1) or 16
days (EKR 2-6), the electrolyte was changed to NaOH to enhance
the electroosmotic transport of neutral As(III). Detailed conditions
are listed in Table 2.
2.3. Analytical method
The pH, electrical conductivity (EC), oxidation-reduction poten-
tial (ORP), and As concentration of the electrolytes were regularly
measured with a pH/EC meter (Istek, Republic of Korea), ORP me-
ter (Istek, Republic of Korea), and inductively coupled plasma opti-
cal emission spectroscopy (ICP-OES, Agilent, USA), respectively. Af-
ter the experiments, the soil compartments were divided into 5
equal sections from the anode to cathode and dried. Each section
was crushed and sieved with a No. 100 sieve (0.125 mm) before
analysis. The pH and EC of soils were measured with a pH/EC me-
ter (Istek, Republic of Korea) after extraction with deionized water
using a ratio of 1: 5 (soil: water). Heavy metals in the soils were
extracted with aqua regia as follows: hydrochloride: nitric acid =
21: 7 at 80 °C for 1 h, and then ICP-OES (Agilent, USA) was used
to analyze the metal concentration in the extract. As fractionation
in soil was analyzed by the modified Wenzel’s method [3], and As
fractionations were classified as non-specifically bound (F1), specif-
ically bound (F2), amorphous and poorly crystalline hydrous oxides
(F3), well-crystallized hydrous oxides (F4), and residual phase (F5).
We modified the extraction of residual phase (F5) using aqua regia
instead of HNO3 /H2 O2 .
3. Results and discussion
3.1. Electrical current
The change of current during operation is presented in Fig. 2.
The initial current of EKR 1–5 was higher than 10 mA because
the pre-treatment process provided ionic matter to the system
(Table 2). The current of EKR 1–5 decreased gradually with op-
eration time because ionic matter was removed from the system
254 S.-R. Ryu et al. / Journal of the Taiwan Institute of Chemical Engineers 70 (2017) 252–259

Fig. 1. Experimental setup for electrokinetic remediation.

Table 2
Electro-kinetic remediation(EKR) experimental conditions.

Exp. no. Pretreatment solution pHa ECb Electrolytesc Electrolytesd

Anolyte Catholyte Anolyte Catholyte

EKR 1 0.1 M Ascorbic acid + Oxalate 4.7 686.0 Tap water 0.1 M Ascorbic acid + Oxalate 0.1 M NaOH Tap water
EKR 2 0.1 M Dithionite 4.6 1383.0 0.1 M Dithionite
EKR3 0.1 M Dithionite + EDTA 5.8 1694.0 0.1 M Dithionite + EDTA
EKR4 0.1 M Dithionite + Citrate 6.6 1680.0 0.1 M Dithionite + Citrate
EKR5 0.1 M Dithionite + Oxalate 5.5 1703.0 0.1 M Dithionite + Oxalate
EKR6 Tap water 5.5 88.6 0.1 M Dithionite + Oxalate
a
soil pH after the pretreatment.
b
soil EC after the pretreatment.
c
Electrolyte circulation before the electrolyte change.
d
Electrolyte circulation after the electrolyte change

by the electric field. In EKR 6, the current increased continuously pH, ORP, and the ratio of iron to oxalate ion concentrations. When
because ionic matter was supplied by electrolyte circulation, which a high ratio of Fe:C2 O4 exists in an acidic and reducing environ-
contained more ionic matter compared to the pretreatment pro- ment, a stable precipitate of Fe(C2 O4 )(s) is formed [2,6,20,21].
cess. The electrolyte change to NaOH caused an increase in current
because Na+ and OH− were continuously provided to the system. Fe2+ + C2 O4 2− ↔ Fe(C2 O4 )(s )
It was observed that the current increased or stabilized after the
The pretreatment and electrolyte circulation of dithionite and
increase. However, the current in EKR 4 was sharply lowered
oxalate (EKR 5) maintained a high ratio of Fe:C2 O4 and reducing
because of clogging of soil pores by precipitation of cations such
environment before the change of electrolyte. Additionally, the soil
as CaCO3 , FeOH3 , and some metal hydroxide due to supplying
of EKR 5 would be acidified by electrolysis because the electrolyte
OH− ion [16]. After pretreatment, the soil pH in EKR 4 was 6.6
maintained an acidic condition before the change of electrolyte
(Table 2). Because the reducing power of dithionite is increased in
(Fig. S1). Thus, the precipitate, Fe(C2 O4 )(s) , led to soil clogging.
the pH range of 6 to 9 and decreases rapidly below 5.5, EKR 4 had
The distribution of soil pH after electrokinetic treatment is pre-
the advantage of being able to reductively dissolve large amounts
sented in Fig. 3. The soil pH in the cathodic region was higher than
of Fe compared to the other conditions [17]. Large amounts of
in the anodic region in all experiments except for EKR 5. Generally,
Fe2+ released from soil formed stable Fe(II)-citrate complexes in
electrolysis caused acidification of the anodic region and alkaliza-
the moderately acidic conditions that moved toward the anode
tion of the cathodic region [5]. However, the electrolyte change to
due to the net negative charge [18,19]. Then, these complexes were
NaOH caused a slight increase of soil pH in the anodic region. In
oxidized at the anodic chamber due to oxygen generated by the
EKR 5, the soil pH after the experiment was lowered to 4.7, and
anodic reaction to form Fe hydroxides at high pH of analyte (>12)
the soil inhibited the movement of OH− toward the anode after
after changing the electrolyte to NaOH [5,18,19].
the change of electrolyte.
2H2 O − 4e− → O2 + 4H+ NaOH (5)
− 3.2. As in the electrolyte
Fe(II ) − Cit + O2 → Fe(III ) − Cit + HO2 /O2 (19)
In EKR 5, a significant increase in the current was not ob- Figs. 4 and 5 show the accumulated mass of As removed to
served after the electrolyte change because precipitation clogged electrolytes from soil and the accumulated EOF. As was scarcely
the pores of the soil. Iron-oxalate complexation depends on the transported to electrolytes before the change of electrolyte to
 S.-R. Ryu et al. / Journal of the Taiwan Institute of Chemical Engineers 70 (2017) 252–259 255

Fig. 2. Electrical current during operation (a) EKR 1, (b) EKR 2 and 3, (c) EKR 4, (d) EKR 5 and 6.

Fig. 3. pH distribution in the soil after the experiment.

NaOH. EKR 1 removed relatively more As from the soil during the
first period of operation, and As was removed by electromigration
and electro-osmotic flow (Fig. 4a and c). Ascorbic and oxalic acids
could extract As bound to amorphous Fe oxides [9,15], and the
removal of As by the organic acids decreased with pH with the
maximum amount at pH 2–3 [2]. After pretreatment, the soil of
EKR 1 was acidified to pH 4.7, and the electrolytes maintained the
acidic conditions until the change of electrolyte (Table 2 and Fig.
S1); thus, ascorbic and oxalic acids could extract greater amounts
of As. The extracted As(V) was transported toward the anode by
electromigration, and reduced neutral As(III) was removed toward
the cathode by EOF. However, the overall removal of As was not ef-
ficient. As removal in EKR 2–6 was negligible even though dithion-
ite was expected to extract larger amounts of As from soil via its
strong reducing power. Dithionite could extract As bound to both
amorphous and crystalline iron phase [15]; however, As extraction
was lowered under acidic conditions because of rapid dithionite
decomposition [2]. Moreover, the reduction potential of dithionite
is highly dependent on the solution pH. The reduction potential
in alkaline condition (at pH 14) was −1.13 V and neutral condition
(at pH 7) was −0.527 V or −0.511 V, and under acidic condition (at
pH 0) was +0.07 V. Thus, it was not effective to reduce As(V) to
As(III) under acidic condition, and it is required to maintain the
pH as neutral or alkaline condition for the reduction reaction by
dithionite [10,22]. Similar to EKR 1, the soil was acidified after pre-
treatment, and the electrolytes maintained the acidic conditions
until the change of electrolyte (Table 2 and Fig. S1). Additionally,
As could precipitate together with the self-decomposition products
of dithionite [15].
The change of electrolyte to NaOH enhanced the current and
EOF toward the cathode (Figs. 2 and 5). As was removed by
both electro-migration and electro-osmotic flow, which means that
As(V) and As(III) coexisted in the soil. In EKR 4 and 5, small
amounts of As accumulated in the catholyte because pore clogging
by iron precipitates and Fe(C2 O4 )(s) reduced the current and inhib-
ited the electroosmotic flow (Figs. 2 and 5b).
256 S.-R. Ryu et al. / Journal of the Taiwan Institute of Chemical Engineers 70 (2017) 252–259

Fig. 4. Accumulated arsenic in the electrolytes (a) EKR 1: Accumulated arsenic in the anolyte, (b) EKR 2∼6: Accumulated arsenic in the anolyte, (c) EKR 1: Accumulated
arsenic in the catholyte, (d) EKR 2∼6: Accumulated arsenic in the catholyte.

Fig. 5. Accumulated EOF toward the cathode (a) Accumulated EOF in EKR 1, (b) Accumulated EOF in EKR 2 −6.

3.3. As removal 3.4. As fractionation

Fig. 6 shows the arsenic distribution in soil after the exper-
iments. As was removed toward both the anodic and cathodic
regions because the directions were directly related to electro-
migration and electro-osmosis. Although the movement of As was
generally insignificant, As in soil accumulated in the cathodic re-
gion in all experiments except EKR 2 and 6, which shows that the
release of As by reduction was valid since the main removal mech-
anism of As(III) is via electro-osmotic flow. In EKR 2 and 6, As in
the soil accumulated in the anodic region despite the circulation of
reducing agents. This result is related to the soil pH, as the soil pH
after the experiments was over 9.2 over a wide range of the soil
in EKR 2 and 6 (Fig. 3). Although As(III) is neutral under pH 9.2, it
changes from H3 AsO3 to H2 AsO3 − above pH 9.2. Since the soil pH
after the experiments was over 9.2, the negatively charged As(III)
was transported toward the anodic region.
Initially, most of the As was associated with Fe hydroxides,
i.e., amorphous and poorly crystalline hydrous oxides (F3) and
well-crystallized hydrous oxides (F4) were 65.6% and 15.9%, re-
spectively (Fig. 7). After electrokinetic treatment, As bound to
amorphous and poorly crystalline hydrous oxides (F3) decreased
while non-specifically bound As (F1) and specifically bound As
(F2) increased. In EKR 1, As bound to amorphous and poorly crys-
talline hydrous oxides (F3) decreased and specifically bound As
(F2) increased. Ascorbic and oxalic acids could extract Fe, mainly
amorphous and poorly crystalline hydrous oxides, by reductive
dissolution and chelation; therefore, As bound to amorphous and
poorly crystalline hydrous oxides (F3) could be released. However,
a significant change in the overall fractionation was not observed,
which indicated that ascorbic and oxalic acids were not effective
in extracting As in EKR. In EKR 2–6 using dithionite, with a strong
 S.-R. Ryu et al. / Journal of the Taiwan Institute of Chemical Engineers 70 (2017) 252–259 257

Fig. 6. Fractionation of arsenic in the soil after electrokinetic treatment.

Fig. 7. Distribution of arsenic in the soil after electrokinetic treatment.

reducing power, a similar trend was observed in specifically bound
As (F2) and As bound to amorphous and poorly crystalline hydrous
oxides (F3). In addition, As bound to well-crystallized hydrous ox-
ides (F4) and the residual phase (F5) increased in the anodic region
in these experiments because As formed precipitates with the self-
decomposition products of dithionite. Dithionite self-decomposed
in the following reactions, which are more active under acidic
conditions [2].
HS2 O4 − + H2 O → HSO2 − + HSO3 − + H+
3HSO2 − + H+ → H2 S + 2HSO3 −
2HSO2 − + H+ → S + HSO3 − + H2 O
The soil in the anodic region was acidified by electrolysis before
the change of electrolyte; thus, released As appeared to precipitate
as an arsenic sulfide phase [2]. This phenomenon was observed in
EKR 3–5 using chelating agents. It has been shown that As is re-
adsorbed at the positively charged soil surface under anoxic con-
ditions by the change of electrolyte [14]. EDTA, citrate and oxalate
form stable Fe-EDTA, Fe-citrate or Fe-oxalate complexes under
acidic conditions, and these complexes are transported toward the
anode due to the negative charge; however, they would be dis-
sociated in the strongly acidic anode region because the stability
of these complexes is decreased under acidic conditions [23,24].
Thus, Fe was released and released As(V) could precipitate as
258 S.-R. Ryu et al. / Journal of the Taiwan Institute of Chemical Engineers 70 (2017) 252–259

Table 3
Mass balance of Arsenic in the system.

Exp. no. Concentration of Asa (mg/kg) Initialb (mg) Anolytec (mg) Catholyted (mg) Soile (mg) Removal efficiencyf (%) Mass balanceg (%)

EKR 1 169.8 85.8 5.5 3.6 75.3 12 98

EKR 2 144.3 75.5 6.4 7.3 56.4 25 93
EKR 3 185.7 88.0 5.9 5.8 84.5 4 109
EKR 4 191.5 81.9 4.7 2.1 81.1 1 107
EKR 5 181.9 81.8 2.7 0.6 76.9 6 98
EKR 6 174.7 77.2 3.8 7.2 69.7 14 105

Previous Electrolyte (0–2 weeks) Changed electrolyte (2–4 weeks) Removal Reference

study Anolyte Catholyte Anolyte Catholyte efficiency (%)

Exp. 1 Tap water Oxalic acid 0.1 M NaOH 0.1 M Tap water 18.7
Exp. 2 Tap water Citric acid 0.1 M NaOH 0.1 M Tap water 25.3
a
Korean regulation level (As: 25 mg/kg).
b
Arsenic mass in the initial soil.
c
Removed mass of arsenic to the anolyte.
d
Removed mass of arsenic to the catholyte.
e
Residual mass of arsenic in the soil after the experiment.
f
1 − (d/a )
g
(b + c + d)/a.

ferrous arsenate or ferric arsenate because the soil at the chamber
is under anoxic conditions [14]. Thus, the combination of dithion-
ite with other chelating agents had an adverse effect by enhancing
precipitates. In the case of EKR 6, this tendency was less than EKR
5 due to the difference in pretreatment although they used the
same electrolyte.
Table 3 shows the removal efficiency of As and the residual
concentration after EKR. Generally, EKR could not meet the Ko-
rean regulation levels because of too low removal efficiency. In EKR
1, although ascorbic and oxalic acids could extract As bound to
amorphous iron oxides with their mild reducing power, they did
not have sufficient removal efficiency. Using dithionite, which is
a powerful reducing agent in EKR 2–6, it was shown that forma-
tion of precipitates decreased As transport. However, EKR 2 and 6,
in which relatively small amounts of precipitation were observed,
showed better efficiencies than EKR 1. However, the removal of As
used organic acids and NaOH was slightly higher or similar with
this study [25], which indicates that dithionite used in this study
was not effective to remove or extract As in the EKR.
4. Conclusions
The combination of reducing and chelating agents was applied
to the removal of As from soil by EKR. Reducing agents such as
ascorbic and oxalic acids and dithionite could release As in the
form of As(III) or As(V). The application of NaOH increased removal
of both As(III) and As(V) through the increase in current and EOF
after the change of electrolyte. However, ascorbic and oxalic acids,
mild reducing agents, were not able to achieve sufficient efficiency
even though acidic conditions, which are optimal for ascorbic and
oxalic acids, were maintained before the change of electrolyte. In
the case of dithionite with powerful reducing ability, As-removal
was strongly constrained due to precipitation. Under acidic condi-
tions, self-decomposition of dithionite was accelerated prior to the
change of electrolyte. The self-decomposition products of dithion-
ite appeared to form precipitates with As in the form of arsenic
sulfide. Chelating agents used to prevent formation of secondary Fe
minerals were also not effective in this system due to the low sta-
bility of complexes under acidic conditions, which enhanced rather
than prevented precipitation. The reducing power of the agent, rate
of self-decomposition, and formation of precipitates are all highly
related to the soil pH. However, the soil pH was changed by elec-
trolysis during the operation of EKR. Thus, precise pH controls in-
cluding alkaline pursing solution and Fe anode to prevent serious
acidification are required to apply the combination of dithionite
and chelating agents in EKR.
Acknowledgment
This work was supported by KEITI through GAIA project and
NRF (2015R1D1A1A09060537).
Supplementary materials
Supplementary material associated with this article can be
found, in the online version, at doi:10.1016/j.jtice.2016.10.058.
References
[1] Lee J-C, Kim EJ, Kim H-W, Baek K. Oxalate-based remediation of arsenic bound
to amorphous Fe and Al hydrous oxides in soil. Geoderma 2016;270:76–82.
[2] Kim EJ, Baek K. Enhanced reductive extraction of arsenic from contami-
nated soils by a combination of dithionite and oxalate. J Hazard Mater
2015;284:19–26.
[3] Choong TSY, Chuah TG, Robiah Y, Koay FLG, Azni I. Arsenic toxicity, health
hazards and removal techniques from water: an overview. Desalination
2007;217:139–66.
[4] Roychowdhury T, Tokunaga H, Uchino T, Ando M. Effect of arsenic-contami-
nated irrigation water on agricultural land soil and plants in West Bengal, In-
dia. Chemosphere 2005;58:799–810.
[5] Acar YB, Alshawabkeh AN. Principles of electrokinetic remediation. Environ Sci
Technol 1993;27:2638–47.
[6] Lee SO, Tran T, Jung BH, Kim SJ, Kim MJ. Dissolution of iron oxide using oxalic
acid. Hydrometallurgy 2007;87:91–9.
[7] Jeon E-K, Jung J-M, Kim W-S, Ko S-H, Baek K. In situ electrokinetic remedi-
ation of As-, Cu-, and Pb-contaminated paddy soil using hexagonal electrode
configuration: a full scale study. Environ Sci Pollut Res 2015;22:711–20.
[8] Jeon E-K, Ryu S-R, Baek K. Application of solar-cells in the electrokinetic reme-
diation of As-contaminated soil. Electrochim Acta 2015;181:160–6.
[9] Pedersen HD, Postma D, Jakobsen R. Release of arsenic associated with
the reduction and transformation of iron oxides. Geochim Cosmochim Acta
2006;70:4116–29.
[10] Mayhew SG. Redox potential of dithionite and SO2-from equilibrium reaci-
tons with flavodoxins, methyl viologen and hydrogen plus hydrogenase. Eur
J Biochem 1978;85:535–47.
[11] Bose P, Sharma A. Role of iron in controlling speciation and mobilization of
arsenic in subsurface environment. Water Res 2002;36:4916–26.
[12] Nowack B, Sigg L. Dissolution of Fe(III)(hydr)oxides by metal-EDTA complexes.
Geochim Cosmochim Acta 1997;61:951–63.
[13] Kim K, Choi W. Enhanced redox conversion of chromate and arsenite in ice.
Environ Sci Technol 2011;45:2202–8.
[14] Fan JX, Wang YJ, Liu C, Wang LH, Yang K, Zhou DM, et al. Effect of iron ox-
ide reductive dissolution on the transformation and immobilization of arsenic
in soils: new insights from X-ray photoelectron and X-ray absorption spec-
troscopy. J Hazard Mater 2014;279:212–19.
 S.-R. Ryu et al. / Journal of the Taiwan Institute of Chemical Engineers 70 (2017) 252–259
[15] Kim EJ, Lee JC, Baek K. Abiotic reductive extraction of arsenic from con-
taminated soils enhanced by complexation: Arsenic extraction by reducing
agents and combination of reducing and chelating agents. J Hazard Mater
2015;283:454–61.
[16] Ryu BG, Park GY, Yang JW, Baek K. Electrolyte conditioning for electroki-
netic remediation of As, Cu, and Pb-contaminated soil. Sep Purif Technol
2011;79:170–6.
[17] de Carvalho LM, Schwedt G. Polarographic determination of dithionite and its
decomposition products: kinetic aspects, stabilizers, and analytical application.
Anal Chim Acta 2001;436:293–300.
[18] Pham AN, Waite TD. Oxygenation of Fe(II) in the presence of citrate in aqueous
solutions at pH 6.0–8.0 and 25 degrees C: interpretation from an Fe(II)/citrate
speciation perspective. J Phys Chem A 2008;112:643–51.
[19] Jones AM, Griffin PJ, Waite TD. Ferrous iron oxidation by molecular oxygen
under acidic conditions: The effect of citrate, EDTA and fulvic acid. Geochim
Cosmochim Acta 2015;160:117–31.
259
[20] Pliego G, Zazo JA, Casas JA, Rodriguez JJ. Fate of iron oxalates in aqueous so-
lution: the role of temperature, iron species and dissolved oxygen. J Environ
Chem Eng 2014;2:2236–41.
[21] Lee SO, Tran T, Park YY, Kim SJ, Kim MJ. Study on the kinetics of iron oxide
leaching by oxalic acid. Int J Miner Process 2006;80:144–52.
[22] McMillan W Jr, Roberts JD, Coryell CD. The thermodynamic constants of the
dithionite (hydrosulfite) ion. J Am Chem Soc 1942;64:398–9.
[23] Panias D, Taxiarchou M, Paspaliaris I, Kontopoulos A. Mechanisms of dis-
solution of iron oxides in aqueous oxalic acid solutions. Hydrometallurgy
1996;42:257–65.
[24] Waite TD, Morel FMM. Photoreductive Dissolution of Colloidal Iron-Oxide - Ef-
fect of Citrate. J Colloid Interface Sci 1984;102:121–37.
[25] Shin SY, Park SM, Baek K. Electrokinetic removal of as from soil washing
residue. Water Air Soil Pollut 2016;227:223.