Journal of Hazardous Materials B137 (2006) 998–1007

Electrochemical treatment of textile dyes and dyehouse effluents

Efthalia Chatzisymeon, Nikolaos P. Xekoukoulotakis, Alberto Coz,
Nicolas Kalogerakis, Dionissios Mantzavinos ∗
Department of Environmental Engineering, Technical University of Crete, Polytechneioupolis, GR-73100 Chania, Greece
Received 24 November 2005; received in revised form 13 March 2006; accepted 14 March 2006
Available online 19 May 2006

Abstract
The electrochemical oxidation of textile effluents over a titanium–tantalum–platinum–iridium anode was investigated. Batch experiments were
conducted in a flow-through electrolytic cell with internal recirculation at current intensities of 5, 10, 14 and 20 A, NaCl concentrations of 0.5, 1,
2 and 4% and recirculation rates of 0.81 and 0.65 L/s using a highly colored, synthetic effluent containing 16 textile dyes at a total concentration
of 361 mg/L and chemical oxygen demand (COD) of 281 mg/L. Moreover, an actual dyehouse effluent containing residual dyes as well as various
inorganic and organic compounds with a COD of 404 mg/L was tested. In most cases, quantitative effluent decolorization was achieved after
10–15 min of treatment and this required low energy consumption; conversely, the extent of mineralization varied between 30 and 90% after
180 min depending on the operating conditions and the type of effluent. In general, treatment performance improved with increasing current
intensity and salinity and decreasing solution pH. However, the use of electrolytes not containing chloride (e.g. FeSO4 or Na2 SO4 ) suppressed
degradation. Although the acute toxicity of the actual effluent to marine bacteria Vibrio fischeri was weak, it increased sharply following treatment,
thus suggesting the formation of persistent toxic by-products.
© 2006 Elsevier B.V. All rights reserved.

Keywords: Decolorization; Dyes; Electrolysis; Textile effluent; Toxicity; Treatment

1. Introduction Dyes concentration in effluents is usually lower than any other

chemical found in these wastewaters, but due to their strong color
Color is one of the most obvious indicators of water pollution they are visible even at very low concentrations, thus causing
and the discharge of highly colored effluents containing dyes serious aesthetic problems in wastewater disposal [4]. There-
can be damaging to the receiving bodies [1]. Of these, textile fore, methods for decolorization of textile effluents have received
effluents typically have strong color due to unfixed dyes, as well considerable attention in recent years. Chemical precipitation,
as they are biorecalcitrant due to the presence of various auxiliary adsorption on activated carbon and natural adsorbents, as well
chemicals such as surfactants, fixation agents, bleaching agents, as several advanced oxidation processes have been employed
etc. [2]. The degree of dye fixation to fabrics depends on the for the treatment of textile effluents. Of the latter, ozonation,
fiber, depth of shade and mode of application and, depending on photocatalytic oxidation, Fenton and photo-Fenton oxidation,
the dye, 2–50% of unfixed dye can enter the waste stream [2]. ultraviolet (UV) irradiation and electrochemical oxidation have
Reactive dyes are usually found at relatively high concentrations been reported in the literature as effective means for the treat-
in wastewaters due to their low fixation especially to fibers such ment of synthetic and actual textile effluents [5,6].
as cotton and viscose. Electrochemical technologies such as electrooxidation, elec-
Dye molecules often receive the largest attention due to their trocoagulation and electroflotation have been widely used in
color, as well as the toxicity of some of the raw materials used water and wastewater treatment and several applications have
to synthesize dyes (e.g. certain aromatic amines), although dyes been recently reviewed elsewhere [7]. Electrooxidation over
are often not the largest contributor to the textile wastewater [3]. anodes made of graphite, Pt, TiO2 , IrO2 , PbO2 , several Ti-based
alloys and, more recently, boron-doped diamond electrodes in
the presence of a supporting electrolyte (typically NaCl) has
∗ Corresponding author. Tel.: +30 28210 37797; fax: +30 28210 37852. been employed for the decontamination of various industrial
E-mail address: [email protected] (D. Mantzavinos). effluents.

0304-3894/$ – see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jhazmat.2006.03.032
 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007 999

Several recent studies report the use of electrooxidation Table 1

to treat model aqueous solutions containing various dyes. Composition of effluents used in this study
Rajkumar et al. [8] studied the electrochemical degradation of Component Synthetic effluent Actual effluent
Reactive Blue 19 over a titanium-based dimensionally stable Remazol Black B 159 (44) nd
anode regarding the effect of operating conditions (current Remazol Red RB 37.3 (10.3) nd
density, salinity, reaction temperature and initial dye concen- Remazol Golden Yellow RNL 20.3 (5.6) nd
tration) on treatment performance, while they also identified Cibacron Black WNN 94.2 (26.1) nd
major reaction intermediates. The effect of various operating Cibacron Red FN-R 0.1 (<0.1) nd
Cibacron Blue FN-G 0.6 (<0.1) nd
conditions on Acid Blue and Basic Brown degradation over a Drimaren Red K-8B 6.7 (1.9) nd
lead/lead oxide anode and on Acid Orange 7 degradation over Drimaren Scarlet K-2G 9.4 (2.6) nd
a boron-doped diamond anode was studied by Awad and Abo Drimaren Yellow K-2R 12.1 (3.3) nd
Galwa [9] and Fernandes et al. [10], respectively. Basic Yellow Drimaren Navy K-BNN 5.6 (1.6) nd
28 and Reactive Black 5 were used as test substances to compare Drimaren Yellow K-4G 12 (3.3) nd
Drimaren Orange X-3LG 0.1 (<0.1) nd
the efficiency of a diamond electrode to that of conventional Drimaren Blue X-3LR 0.05 (<0.1) nd
metallic electrodes (iron, aluminium and copper) [11], while Drimaren Violet K-2RL 0.3 (<0.1) nd
an activated carbon fiber electrode was used to assess the Drimaren Red K-4BL 0.4 (0.1) nd
electrochemical degradability of 29 different textile dyes [12]. Drimaren Blue K-2RL 2.6 (0.7) nd
In further studies [13], several advanced oxidation processes, Total dye content 360.8 (100) nd
Organic auxiliary chemicals 0 nd
namely wet oxidation, TiO2 photocatalysis, electro-Fenton and Na2 SO4 0 5500
UV-assisted electro-Fenton were compared concerning their Na2 CO3 0 440
efficiency in treating Reactive Black 5. NaOH 0 110
Despite the relatively large number of papers dealing with COD 281 404
the electrochemical degradation of model aqueous solutions of Total solids 0 75
pH 7.5 9.5
dyes, appreciably fewer reports regarding the treatment of actual Absorbance (au) 5.5 0.25
effluents are available. Lin and Peng [14] developed a continuous EC50 (%) 4.1 75
process comprising coagulation, electrochemical oxidation and
Concentrations are quoted in mg/L. Numbers in brackets show percent dye
activated sludge to treat textile effluents. The effect of changing composition in the synthetic effluent; nd, not determined.
operating conditions such as coagulant concentration, solution
pH, current density, number of electrodes, residence times in
electrochemical and biological reactors on treatment efficiency of each dye to the total dye content is shown in Table 1 and this
was thoroughly investigated and optimal conditions were estab- composition was chosen to match exactly the percent compo-
lished. In further studies, Vlyssides et al. [15] investigated the sition of the actual dyestuff. All dyes were kindly provided by
electrooxidation of textile effluents over a Ti–Pt electrode at EPILEKTOS SA, a textile manufacturing industry located in the
different chloride concentrations and reported that electrochem- region of Sterea, Central Greece. The effluent is near-neutral,
ical treatment improved the biotreatability (as assessed by the strongly colored and highly ecotoxic with a COD content of
BOD/COD ratio) of the original effluent. Naumczyk et al. [16] 281 mg/L.
compared the decolorization and mineralization rates of textile
effluent electrochemical degradation over three Ti-based elec- 2.2. Actual effluent
trodes coated with different metals and they also attempted to
identify major reaction by-products. In a recent work, Sakalis et The actual effluent (AE) used in this study was provided
al. [17] demonstrated a continuous, pilot-scale cascade electro- by EPILEKTOS SA and it was used as received without any
chemical reactor capable of achieving 90% decolorization of a pretreatment. The dyeing process through which the effluent
textile effluent at a residence time of 40 min. was generated is shown in Fig. 1. The cotton fiber is first mer-
The aim of this work was to investigate the electrochemical cerized and bleached using NaOH, H2 O2 and other bleaching
treatability of both complex synthetic and actual textile effluents agents and then rinsed with water prior to dyeing. The dyestuff
over a Ti–Ta–Pt–Ir anode regarding the effect of varying oper- consists of about 200 kg of 16 dyes (whose percent compo-
ating conditions such as current, type and initial concentration sition is shown in Table 1) in 30 m3 water to dye 3 tonnes
of electrolyte and solution pH on decolorization, reduction of of cotton fibers and is also added large amounts of inorganic
COD and energy consumption. The effect of treatment on efflu- salts. Following dyeing, the fiber is washed several times with
ent acute ecotoxicity to marine bacteria Vibrio fischeri was also water and detergents in consecutive baths and finally under-
investigated. goes fixation and softening. Waste streams from the various
baths are collected in an equalization tank where the efflu-
2. Materials and methods ent used in this study was taken from. It consists of residual
2.1. Synthetic effluent dyes, NaOH, inorganic salts (Na2 SO4 and Na2 CO3 ) and various
organic components such as detergents, softening, dispersing
The synthetic effluent (SE) used in this study is a mixture of and fixing agents (collectively referred to as organic auxiliary
16 dyes with a total concentration of 361 mg/L. The contribution chemicals in Table 1). The effluent is highly alkaline, with a dark
1000 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007

Fig. 1. Flowsheet of the process generating the actual effluent used in this study.

green-blue color and partially ecotoxic with a COD content of of four major dyes were employed, absorbance was measured at
404 mg/L. each dye’s λmax ; this was 411, 542, 593 and 597 nm for Remazol
Golden Yellow RNL, Remazol Red RB, Cibacron Black WNN
2.3. Electrochemical degradation experiments and Remazol Black B, respectively.
COD was determined by the dichromate method. The appro-
Experiments were conducted in an electrolytic cell compris- priate amount of sample was introduced into commercially avail-
ing a stainless steel 316 L cathode, a titanium (grade II/VII) able digestion solution containing potassium dichromate, sulfu-
anode covered by a thin film of tantalum, platinum and iridium ric acid and mercuric sulfate (Hach Europe, Belgium) and the
alloy and the power supply unit. The cathode was a 5 cm in diam- mixture was then incubated for 120 min at 150 ◦ C in a COD reac-
eter and 12 cm long hollow cylinder in the center of which was tor (Model 45600-Hach Company, USA). COD concentration
housed the anode which, in turn, was 11 cm long with a diameter was measured colorimetrically using a DR/2010 spectropho-
of 3 cm. In a typical run, the effluent was mixed with the appro- tometer (Hach Company, USA).
priate amount of NaCl (or other electrolyte), batch loaded in a
vessel and continuously pumped in the cell through a pump at a 2.5. Acute toxicity
recirculation flowrate of 0.81 L/s (unless otherwise stated), thus
maintaining near-perfect mixing; the rate could be adjusted to The luminescent marine bacteria V. fischeri was used to assess
the desired value through a series of valves. The current intensity the acute ecotoxicity of textile wastewater samples prior to and
was then set to the desired value and the voltage was auto-
matically regulated to match the current value. In all cases, the
working volume was 8 L. A spiral coil immersed in the liquid
and connected to tap water supply was used to remove the heat
liberated from the reaction. All runs were conducted at ambi-
ent temperature which remained practically constant throughout
the experiment at 26 ± 1 ◦ C. Most of the runs were performed
at effluent’s ambient pH which was monitored throughout the
reaction with a Toledo 225 pH meter. For those experiments
where the starting solution pH was adjusted to acidic conditions,
the appropriate volume of H2 SO4 was added in the effluent. A
schematic of the experimental configuration is shown in Fig. 2.

2.4. Analytical measurements

The extent of decolorization that had occurred during

electrochemical treatment was assessed measuring sample
absorbance on a Shimadzu UV 1240 spectrophotometer. For
the synthetic and actual effluents, absorbance was measured at
λmax = 595 ± 5 nm. Prior to the measurement, the actual efflu- Fig. 2. Schematic of the experimental configuration. (1) Cathode, (2) power
ent was centrifuged for 15 min to remove any particles present supply, (3) anode, (4) OMW feed tank, (5) peristaltic pump, (6) coil for cooling
in the sample. In those cases where model aqueous solutions water and (7) valves.
 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007 1001

after treatment. The inhibition of bioluminescense of V. fischeri

exposed to untreated and treated dyehouse wastewater samples
for 15 min at 15 ◦ C was measured using a LUMIStox analyzer
(Dr. Lange, Germany) and the results were compared to an aque-
ous control. Toxicity is expressed as EC50 , which is the effective
concentration of a toxicant causing 50% reduction of light output
during the designated time intervals at 15 ◦ C. For each sample,
its EC50 value was determined by applying several dilutions.

3. Results and discussion

3.1. Effect of NaCl concentration on COD measurement

The standard COD measurement is known to be affected by a

number of inorganic substances which are outlined in the Stan- Fig. 3. Effect of salinity on color conversion for synthetic (SE) and actual (AE)
dard Methods for the Examination of Water and Wastewater effluents at 5 A and ambient solution pH. (X) shows run with SE at reduced
[18]. Of these, chloride may have a significant positive effect on salinity varying from 0.05 to 0.3%, with arrows showing periods of NaCl addi-
the test which is due to its reaction with potassium dichromate tion.
as follows:
resulting in 1 A at 25 V, about 15% decolorization occurred at
6Cl− + Cr2 O7 2− + 14H+ → 2Cr 3+ + 3Cl2 + 7H2 O (1) the end of this period. The effluent was then added an extra
The commercially available COD digestion solutions used 0.05% salt and left to react for another 30 min (2 A at 25 V)
in this study counterbalance chloride interference at chloride leading to a total of 21% decolorization after 60 min. This
concentrations up to 0.2% due to the presence of mercuric sul- procedure was repeated six times clearly showing the benefi-
fate. However, this concentration is usually lower than that used cial effect of adding NaCl on effluent decolorization. However
in electrochemical degradation experiments; therefore, it was and with the exception of the synthetic effluent at salinity val-
decided to estimate the effect of salinity on the COD test. This ues equal to or lower than 0.5% and the actual effluent at 0%
was done by measuring the COD content of effluents at various salinity, in all other cases color removal became quantitative
NaCl concentrations up to 4% as well as without NaCl. It was (e.g. >95%) within 10–15 min of reaction. For the run with the
found that the discrepancy between the measured COD values actual effluent in the absence of NaCl, decolorization proceeded
with and without salt (COD1 and COD2 , respectively) increased slowly reaching a final value of about 60% after 180 min of
linearly with percent salinity (S) as follows: reaction. Na2 SO4 , Na2 CO3 and NaOH, all found in the actual
 
effluent at considerable concentrations, serve as the supporting


COD1 − COD2 0 for S ≤ Sref electrolytes to induce electrochemical degradation. However,
= (2) addition of sodium chloride drastically increased decolorization.
COD1 0.18S + 0.01 for S > Sref
Several previous studies have shown the superiority of NaCl over
where Sref corresponds to the maximum chloride concentration Na2 SO4 and other electrolytes (e.g. NaOH and H2 SO4 ) for the
that is compensated due to the presence of mercury sulfate, i.e. treatment of dye-containing solutions [8–10,17,19].
0.2%. Eq. (2) was used to account for the effect of chloride The effect of changing current on color removal and COD
interference on COD measurement. The corrected values were conversion during the electrochemical treatment of synthetic and
then used to compute process efficiency in terms of specific actual effluents at 0.5 and 4% salinity, respectively, is shown in
energy consumption, anode efficiency and current efficiency. Fig. 4. Increasing current from 5 to 10–14 A improved decol-
orization from 39 to 68–87%, respectively, after 5 min with the
3.2. Effect of operating conditions on treatment synthetic effluent. Other than the early stages of the reaction,
the applied current, at the conditions employed in this study,
Fig. 3 shows the effect of varying salinity on effluent decol- does not seem to be critical to decolorization since quantitative
orization as a function of treatment time at 5 A current and ambi- removal could be achieved within 10–15 min of reaction; for
ent solution pH. Increasing salinity from 0.5 to 1–4% increased instance, over 98% color removal occurred after 15 min with
decolorization from 39 to 50–70%, respectively, after 5 min of the actual effluent regardless the applied current. Conversely,
reaction with the synthetic effluent. An additional experiment COD conversion appreciably increased with increasing current
was carried out with the synthetic effluent at 0% salinity; the and this was more pronounced for the run with the synthetic
salt-free solution exhibited no conductivity at the maximum effluent. For instance, the final COD reduction for the synthetic
operating output of the power supply (i.e. 25 V). The experi- effluent was 55, 65 and 86% at 5, 10 and 14 A, respectively;
ment was then repeated implementing a stepwise addition of these values for the actual effluent became 29, 34 and 39% at
0.05% NaCl every 30 min and the results are also shown in 5, 10 and 20 A, respectively. The fact that decolorization occurs
Fig. 3. For instance, the effluent was first added 0.05% salt at substantially greater rates than COD conversion implies that
and left to react for 30 min; although conductivity was very low electrochemical degradation by-products are more resistant to
1002 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007

Fig. 4. Effect of current on: (a) color and (b) COD conversion for synthetic
(SE) and actual (AE) effluents at 0.5 and 4% salinity, respectively, and ambient Fig. 5. Effect of pH on COD conversion for: (a) synthetic (5 A and 1% salinity)
solution pH. and (b) actual (20 A and 4% salinity) effluents.

electrooxidation than the original dyes; this is readily deducted in Fig. 5. As seen, acidic conditions appear to favor COD conver-
from the runs with the synthetic effluent whose starting organic sion for both actual and synthetic effluents. On the other hand,
content consists exclusively of dyes. Similar results regarding color removal was not affected by varying solution pH as, in all
the relative rates of electrochemical decolorization and miner- cases, complete decolorization was achieved within 10–15 min
alization have also been reported by several other investigators (data not shown). It should be mentioned here that, since the
[9,10,12,16]. solutions were not buffered, pH progressively recovered from
It is also interesting to note that the actual effluent is strongly acidic to alkaline conditions presumably due to the formation of
resistant to mineralization and this is possibly due to the presence hydroxyl anions in the solution, thus partly masking the effect
of persistent organic auxiliary chemicals rather than unoxidized of pH on degradation. However, it is well-documented [20,21]
dyes and their degradation by-products. Although the concen- that pH does not have a significant effect on the electrochemical
tration of residual dyes in the actual effluent is not known, a degradation of organics over titanium anodes in the range 3–10.
rough estimate can be made based on the absorbance values of Ceron-Rivera et al. [11] reported that the rate of electrochemical
the actual and synthetic effluents and the known dye content of degradation of Reactive Black 5 over an iron anode at pH 5.5 was
the latter. As seen from Table 1, the dye content of the actual as much as four times faster than that at pH 7.5, while Awad and
effluent should not exceed 15–20 mg/L; consequently, most of Abo Galwa [9] found that the electrochemical oxidation of Acid
the organic content originally present in the actual effluent is due Blue and Basic Brown over a lead/lead oxide anode was favored
to compounds other than the dyes. The discrepancy between the at pH 2–3 but strongly suppressed at pH 12. Lin and Peng [14]
recorded levels of mineralization becomes even greater bear- reported that treatment performance (in terms of COD decrease)
ing in mind that electrochemical oxidation experiments with the of an actual textile effluent over an iron anode was maximized
actual effluent were conducted at 4% salinity, while those with at neutral conditions but it deteriorated at alkaline conditions.
the synthetic one at 0.5%. However, it should be pointed out that In further experiments, the effect of two iron-containing
these discrepancies may, to some degree, be due to the different electrolytes, namely ferric chloride (FeCl3 ·6H2 O) and ferrous
water matrices of the two effluents. sulfate (FeSO4 ·7H2 O) on treatment efficiency was investigated.
To test the effect of solution pH on treatment, experiments The rationale behind the use of these electrolytes is that iron can
were conducted at pH values of 3 and 6 and the results are shown serve as an effective homogeneous catalyst to degrade organic
 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007 1003

Table 2
Percent color conversion during the electrochemical treatment of model solutions of four dyes at 5 A and 0.5% salinity
Treatment time (min) Remazol Black B Remazol Red RB Remazol Golden Yellow RNL Cibacron Black WNN

0 0 (4.5) 0 (9.5) 0 (2) 0 (6)

5 84.7 53 55.9 81.2
10 96.7 96.8 75.5 92.6
15 98.8 99.4 84.5 96
30 99.7 (0.6) 99.8 (0.4) 90.2 (1) 99.7 (1)

Initial dye concentration: 160 mg/L. Numbers in brackets show EC50 values (%) for each dye.

compounds as well as reduce hydrogen peroxide to hydroxyl electrochemical oxidation of 450 mg/L Reactive Black 5 over
radicals, thus simulating Fenton-type reactions. The synthetic a Ti–Pt anode at about 0.1% salinity and reported quantitative
effluent was treated at 5 A in the presence of 0.77% FeCl3 ; this decolorization after about 120 min of reaction. Kusvuran et al.
corresponds to chloride concentration in the reaction mixture [13] reported that the rate of Reactive Black 5 degradation over
equal to that of 0.5% NaCl. Treatment performance with ferric a Pt anode in the presence of iron ions was strongly dependent
chloride, in terms of color and COD removal, was similar to on the initial dye concentration; at 100 mg/L dye concentration
that with NaCl at a common chloride concentration. Complete and 20 mg/L iron concentration, only about 40% decolorization
decolorization occurred within 15–30 min and this was accom- occurred after 30 min of reaction.
panied by a final (i.e. after 180 min) COD reduction of about
50–60%. An additional experiment was performed with 0.79% 3.3. Electrochemical degradation mechanisms
FeSO4 ; this corresponds to iron concentration equal to that of
0.77% FeCl3 . Degradation was slow yielding only about 26 and Two mechanisms are thought to be responsible for organic
85% color removal after 30 and 180 min of reaction with the final matter (R) electrochemical degradation, namely: (a) direct
COD removal being only 10%, thus highlighting the beneficial anodic oxidation where the pollutants are adsorbed on the anode
role of chloride as an oxidizing agent. It should be mentioned surface (M) and destroyed by the anodic electron transfer reac-
that addition of FeCl3 or FeSO4 in the effluent resulted in par- tion and (b) indirect oxidation in the liquid bulk which is
tial iron precipitation and this was accompanied by a substantial mediated by the oxidants that are formed electrochemically;
pH decrease from 7.5 to 2–3; this value remained practically such oxidants include chlorine, hypochlorite, hydroxyl radicals,
constant throughout the experiment. ozone and hydrogen peroxide. Anodic water discharge results
To assess the effect of flowrate on treatment, experiments with in the formation of hydroxyl radicals that are adsorbed on the
the synthetic and actual effluents were performed at 0.5% NaCl anode surface and can then oxidize the organic matter [22,24]:
and 10 A and 4% NaCl and 20 A, respectively, and a reduced
recirculation rate of 0.65 L/s. Decreasing recirculation had prac- H2 O + M → M[OH• ] + H+ + e− (3)
tically no impact on color removal which was quantitative within R + M[OH• ] → M + RO + H+ + e− (4)
10–15 min in either case; however, the final extent of COD con-
version mildly decreased by about 10%. This can be explained In the presence of NaCl, chlorohydroxyl radicals are also
by the combined effect of: (a) decreased production of oxidants formed on the anode surface and then oxidize the organic matter:
and (b) the fact that a smaller number of organics flow through H2 O + M + Cl− → M[ClOH• ] + H+ + 2e− (5)
the cell in a single pass, both of which occur at reduced flow-
+ −
through velocities [22,23]. R + M[ClOH• ] → M + RO + H + Cl (6)
In a final set of experiments, the four dominant, in terms of
Reactions between water and radicals near the anode can
concentration, dyes found in the effluent were separately oxi-
yield molecular oxygen, free chlorine and hydrogen peroxide:
dized electrochemically at an initial concentration of 160 mg/L.
As seen in Table 2, all but Remazol Golden Yellow RNL were H2 O + M[OH• ] → M + O2 + 3H+ + 3e− (7)
completely decolorized within 10–15 min. The chemical formu-
lae of the major dyes are shown in Fig. 6; Remazol Black B (also H2 O + M[ClOH• ] + Cl− → M + O2 + Cl2 + 3H+ + 4e−
referred to as Reactive Black 5) is a diazo dye with two vinyl sul-
fone reactive groups, while Golden Yellow RNL (also referred (8)
to as Reactive Orange 107) is a vinyl sulfone-based monoazo
H2 O + M[OH• ] → M + H2 O2 + H+ + e− (9)
dye. On the other hand, Remazol Red RB (also referred to as
Reactive Red 198) is a heterobifunctional monoazo dye with Furthermore, hypochlorite can be formed as follows:
both a vinyl sulfone and a monochlorotriazine reactive group,
H2 O + Cl− → HOCl + H+ + 2e− (10)
while Black WNN (not shown in Fig. 6) is a proprietary mix-
ture of several azo dyes. The initial rate of Remazol Black 5 Therefore, direct anodic oxidation through reactions (4) and
decolorization seems to be greater than that of Remazol Red (6) results in reduced COD as well as the formation of primary
RB and Golden Yellow RNL possibly due to the existence of oxidants such as oxygen, chlorine, hypochlorite and hydrogen
two chromophore N N bonds. Sakalis et al. [17] studied the peroxide. Free chlorine and oxygen can further react on the
1004 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007

Fig. 6. Chemical formulae of the major dyes used in this study.

anode yielding secondary oxidants such as chlorine dioxide and 3.4. Energy consumption
ozone, respectively:
Electrochemical treatment is undoubtedly an energy-intense
H2 O + M[ClOH• ] + Cl2 → M + ClO2 + 3H+ + 2Cl− + e− process and its efficiency is usually assessed in terms of spe-
(11) cific energy consumption (SEC). This is defined as the amount
of energy consumed per unit mass of organic load (e.g. dye or
COD) removed. Alternatively, treatment performance may be
O2 + M[OH• ] → M + O3 + H+ + e− (12) expressed in terms of anode efficiency, i.e. the mass of organic
load removed divided by the electrode area, the applied cur-
The presence of Na2 SO4 and other sulfate-containing elec-
rent and the treatment time. Representative temporal profiles
trolytes in the reaction mixture leads to the formation of SO2
of cumulative SEC and anode efficiency with respect to dye
(especially in acidic media) which is a moderate reductant [17].
removal during the electrochemical treatment of synthetic and
Primary and secondary oxidants are quite stable and migrate
actual effluents are shown in Fig. 7. Anode efficiency reached
in the solution bulk where they indirectly oxidize the effluent.
a maximum within 10–15 min of reaction and then sharply
The efficiency of direct oxidation depends on the anode activ-
decreased presumably due to the rapid decolorization of both
ity, the diffusion rate of organics on the anode surface and the
effluents.
applied current density. On the other hand, the efficiency of indi-
Current efficiency (CE) is defined as the percentage of applied
rect oxidation depends on the diffusion rate of oxidants in the
current utilized to reduce COD:
solution and the pH value [24]. At acidic conditions, free chlo-


rine is the dominant oxidizing agent, while at slightly alkaline CODo − CODt
conditions hypochlorite, chloride ions and hydroxyl radicals are CE = FV (13)
8It
all important. Most of the experiments in this study were per-
formed at ambient pH which was alkaline for both effluents; where CODo and CODt , respectively, refer to the starting and
given that pH typically varied throughout the course of the final COD values, I the applied current, F the Faraday constant,
reaction between 7.5 and 8.5 for the runs with the synthetic V the liquid volume and t is the treatment time.
effluent and 9 and 9.5 for the runs with the actual one implies Table 3 summarizes SEC values with respect to dye and COD
that indirect oxidation might have proceeded through various removal as well as CE values for various runs with synthetic and
oxidants. actual effluents. Values of SEC with respect to dye degradation
 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007 1005

Fig. 7. Specific energy consumption (a and c) and anode efficiency (b and d) during the electrochemical treatment of synthetic (SE) and actual (AE) effluents at 0.5
and 4% salinity, respectively, and ambient solution pH.

were computed after 15 min of reaction given the fast effluent low energy requirement for decolorization; for instance, electro-
decolorization at the conditions employed in this study; con- chemical oxidation at 5 A would cost no more than D 0.5/kg of
versely, SEC values for COD removal were computed after dye removed for the complete decolorization of the synthetic
180 min given the relatively low mineralization rates recorded. effluent (based on D 0.1/kWh). SEC for the decolorization of
Likewise, CE values were computed after relatively long treat- the actual effluent is far greater than that for the synthetic one
ment times, i.e. 60 and 180 min. With the exception of run 10, dye (e.g. compare runs 2–4 with runs 11–13) since the latter is far
degradation occurred readily and this is reflected to the relatively more colored than the former. Nonetheless, the absolute energy

Table 3
Specific energy consumption (SEC), current efficiency (CE) and final toxicity (EC50 ) during the electrochemical treatment of synthetic (runs 1–9) and actual (runs
10–16) effluents at various conditions
Run Current Voltage Salinity pH SEC15 min SEC180 min CE60 min CE180 min EC50–180 min
(A) (V) (%) (kWh/kg dyerem ) (kWh/kg CODrem ) (%) (%) (%)

1 5 12 0.5 7.5 5.6 128 36.2 31.4 3.6

2 5 8 1 7.5 3.5 nd nd nd nd
3 5 6 2 7.5 2.6 nd nd nd nd
4 5 5 4 7.5 2.2 nd nd nd nd
5 10 19 0.5 7.5 15.7 367.6 30.2 17.4 1
6 14 25 0.5 7.5 32.3 529.4 31.1 15.8 1.1
7 5 8 1 6 3.4 90.5 41.2 29.6 0.5
8 5 8 1 3 3 77.3 36.9 30.3 0.5
9 10 19 0.5 7.5 15.5 468 26.6 13 5.2
10 5 15 0 9.5 2651 nd nd nd nd
11 5 6 1 9.5 61.4 nd nd nd nd
12 5 5 2 9.5 48.1 nd nd nd 3
13 5 4 4 9.5 39.9 94.6 24.1 13.1 1
14 10 6 4 9.5 138.3 119 nd 16.9 1
15 20 8 4 9.5 363.7 488.5 11.9 5.5 0.3
16 20 8 4 9.5 434 594.5 8.3 4.5 0.2

Runs 9 and 16 were performed at a recirculation rate of 0.65 L/s, while the rest at 0.81 L/s; nd, not determined.
1006 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007

consumption for the actual effluent is lower than that for the toxicity of the original solution which remained quite high even
synthetic one by about 25% since the former contains extra after 45 min of reaction.
electrolytes in addition to NaCl. Increasing current intensity at
constant salinity increased SEC for both dye and COD removal 4. Conclusions
and for both effluents. For instance, increasing current from 5 to
14 A at 0.5% salinity (runs 1 and 6) nearly quadrupled SEC for The main conclusions drawn from this study can be summa-
COD removal, thus showing that an increase in COD conversion rized as follows:
with current intensity (as shown in Fig. 4b) is accompanied by
a proportionately greater increase in energy consumption. The (1) Electrochemical oxidation is capable of destroying the chro-
cumulative energy consumption after 180 min of treatment was mophore groups of dyes found in textile effluents at short
0.18, 0.57 and 1.05 kWh at 5, 10 and 14, respectively, while treatment times and low energy consumption. However,
the respective amount of COD removed was 1.4, 1.55 and 2 g. this is accompanied by a moderate degree of mineraliza-
This is due to the fact that degradation by-products appear to tion. Treatment efficiency, in terms of both conversion and
be persistent to further electrochemical oxidation and possibly specific energy consumption, is affected by the operating
to the fact that some energy is wasted for the competing water conditions employed, i.e. applied current, type and con-
electrolysis reaction that becomes dominant at higher current centration of electrolyte and effluent pH. Moreover, the
densities and relatively low organic concentrations [25]. Cur- water matrix may also affect treatment since actual efflu-
rent efficiency was found to decrease with increasing treatment ents usually contain large concentrations of several organic
time, thus indicating energy waste to side reactions other than and inorganic species whose interference is not taken into
those contributing to COD decrease. account when synthetic model solutions of dyes are used
instead of actual effluents.
3.5. Effluent ecotoxicity (2) Although the ecotoxicity of the untreated actual effluent to
V. fischeri is weak, it sharply increases after electrochemical
Several samples collected at the end of electrochemical treat- oxidation and this is ascribed to the formation of organochlo-
ment at various conditions were tested with respect to their acute rinated and other toxic by-products that are persistent to
ecotoxicity to V. fischeri and representative results are given in further oxidation.
Tables 2 and 3. The original synthetic effluent is strongly toxic
and this is obviously due to its high dye content; as seen in Acknowledgments
Table 2, Remazol Black B at a concentration equal to that found
in the synthetic effluent has an EC50 value as low as 4.5% and The authors wish to thank G. Balagouras of EPILEKTOS
this is also the case with the other three major dyes. The toxicity SA for kindly providing the effluent and dyes used in this study.
of Reactive Black 5 to V. fischeri has also been demonstrated Financial support for this work was provided by the Hellenic
elsewhere [26]. On the other hand, the original actual effluent Ministry of National Education & Religious Affairs under the
is only weakly toxic although it contains several organic and “PYTHAGORAS” program. AC wishes to acknowledge the
inorganic species that are not present in the synthetic effluent. Marcelino Botin Foundation (Spain) for granting him a research
Although NaCl mediated electrochemical treatment was fellowship.
capable of degrading completely the dyes as well as reducing
the total organic load, the final effluent was always very toxic
References
and this is believed to be due to the formation of organochlo-
rinated compounds as well as secondary residual oxidants that [1] P. Nigam, I.M. Banat, D. Singh, R. Marchant, Microbial process for the
remained in the reaction mixture even after prolonged electro- decolorization of textile effluent containing azo, diazo and reactive dyes,
chemical oxidation. Toxicity remained high even after 180 min Process Biochem. 31 (1996) 435–442.
of synthetic effluent treatment at 14 A and 0.5% salinity, condi- [2] A.B.C. Alvares, C. Diaper, S.A. Parsons, Partial oxidation by ozone to
tions that yielded nearly 90% COD removal (i.e. run 6 in Table 3 remove recalcitrance from wastewaters—a review, Environ. Technol. 22
(2001) 409–427.
and Fig. 4b). Naumczyk et al. [16] reported that 80% of the 22 [3] A. Reife, H.S. Freeman, Environmental Chemistry of Dyes and Pigments,
identified (by means of GC/MS analysis) organic compounds John Wiley and Sons Inc., Canada, 1996.
originally present in a textile effluent could be removed electro- [4] H. Zollinger, Color Chemistry, Wiley-VCH, Zurich, Switzerland, 2003.
chemically over various Ti-based anodes (including a Ti–Pt–Ir [5] D. Mantzavinos, E. Psillakis, Enhancement of biodegradability of indus-
one); however, this gave rise to the formation of 20 organochlo- trial wastewaters by chemical oxidation pre-treatment, J. Chem. Technol.
Biotechnol. 79 (2004) 431–454.
rinated compounds most of which remained in the final effluent. [6] S. Meric, D. Kaptan, T. Olmez, Color and COD removal from wastewater
Increased toxicity was also recorded (data not shown) follow- containing Reactive Black 5 using Fenton’s oxidation process, Chemo-
ing treatment with FeCl3 or FeSO4 as the conducting electrolyte, sphere 54 (2004) 435–441.
implying that the presence of iron and/or other non-chlorinated [7] G. Chen, Electrochemical technologies in wastewater treatment, Sep. Purif.
Technol. 38 (2004) 11–41.
organic by-products may also contribute to effluent toxicity.
[8] D. Rajkumar, B.S. Song, J.G. Kim, Electrochemical degradation of Reac-
Kusvuran et al. [13] who studied the effect of electro-Fenton tive Blue 19 in chloride medium for the treatment of textile dyeing wastew-
treatment of Reactive Black 5 on the growth of Pseudomonas ater with identification of intermediate compounds, Dyes Pigments 72
putida reported that oxidation for 15 min increased sharply the (2006) 1–7.
 E. Chatzisymeon et al. / Journal of Hazardous Materials B137 (2006) 998–1007 1007

[9] H.S. Awad, N. Abo Galwa, Electrochemical degradation of Acid Blue and removal from textile wastewater, Chem. Eng. J. 111 (2005) 63–70.
Basic Brown dyes on Pb/PbO2 electrode in the presence of different con- [18] L.S. Clesceri, A.E. Greenberg, A.D. Eaton, Standard Methods for the
ductive electrolyte and effect of various operating factors, Chemosphere Examination of Water and Wastewater, 20th ed., APHA, AWWA, WEF,
61 (2005) 1327–1335. 1998.
[10] A. Fernandes, A. Morao, M. Magrinho, A. Lopes, I. Goncalves, Electro- [19] X.M. Zhan, J.L. Wang, X.H. Wen, Y. Qian, Indirect electrochemical
chemical degradation of C.I. Acid Orange 7, Dyes Pigments 61 (2004) treatment of saline dyestuff wastewater, Environ. Technol. 22 (2001)
287–296. 1105–1111.
[11] M. Ceron-Rivera, M.M. Davila-Jimenez, M.P. Elizalde-Gonzalez, Degra- [20] D. Rajkumar, K. Palanivelu, Electrochemical treatment of industrial
dation of the textile dyes Basic Yellow 28 and Reactive Black 5 wastewater, J. Hazard. Mater. B113 (2004) 123–129.
using diamond and metal alloys electrodes, Chemosphere 55 (2004) [21] D. Rajkumar, K. Palanivelu, Electrochemical degradation of cresols for
1–10. wastewater treatment, Ind. Eng. Chem. Res. 42 (2003) 1833–1839.
[12] Z. Shen, W. Wang, J. Jia, J. Ye, X. Feng, A. Peng, Degradation of dye solu- [22] M. Gotsi, N. Kalogerakis, E. Psillakis, P. Samaras, D. Mantzavinos, Elec-
tion by an activated carbon fiber electrode electrolysis system, J. Hazard. trochemical oxidation of olive oil mill wastewaters, Water Res. 39 (2005)
Mater. B84 (2001) 107–116. 4177–4187.
[13] E. Kusvuran, S. Irmak, H. Ibrahim Yavuz, A. Samil, O. Erbatur, Comparison [23] A. Kraft, M. Blaschke, D. Kreysig, B. Sandt, F. Schröder, J. Ren-
of the treatment methods efficiency for the decolorization and mineraliza- nau, Electrochemical water disinfection. Part II. Hypochlorite production
tion of Reactive Black 5 azo dye, J. Hazard. Mater. B119 (2005) 109–116. from potable water, chlorine consumption and the problem of calcareous
[14] S.H. Lin, C.F. Peng, Continuous treatment of textile wastewater by com- deposits, J. Appl. Electrochem. 29 (1999) 895–902.
bined coagulation, electrochemical oxidation and activated sludge, Water [24] C.J. Israilides, A.G. Vlyssides, V.N. Mourafeti, G. Karvouni, Olive oil
Res. 30 (1996) 587–592. wastewater treatment with the use of an electrolysis system, Bioresour.
[15] A.G. Vlyssides, M. Loizidou, P.K. Karlis, A.A. Zorpas, D. Papaioannou, Technol. 61 (1997) 163–170.
Electrochemical oxidation of a textile dye wastewater using a Pt/Ti elec- [25] N.N. Rao, K.M. Somasekhar, S.N. Kaul, L. Szpyrkowicz, Electrochemical
trode, J. Hazard. Mater. B70 (1999) 41–52. oxidation of tannery wastewater, J. Chem. Technol. Biotechnol. 76 (2001)
[16] J. Naumczyk, L. Szpyrkowicz, F. Zilio-Grandi, Electrochemical treatment 1124–1131.
of textile wastewater, Water Sci. Technol. 34 (1996) 17–24. [26] A. Gottlieb, C. Shaw, A. Smith, A. Wheatley, S. Forsythe, The toxicity of
[17] A. Sakalis, K. Mpoulmpasakos, U. Nickel, K. Fytianos, A. Voulgaropou- textile reactive azo dyes after hydrolysis and decolourisation, J. Biotechnol.
los, Evaluation of a novel electrochemical pilot plant process for azodyes 101 (2003) 49–56.