Journal of Water Process Engineering 22 (2018) 73–79

Contents lists available at ScienceDirect

Journal of Water Process Engineering

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

Optimization of sequential chemical coagulation - electro-oxidation process T

for the treatment of an industrial textile wastewater
⁎
Edison GilPavasa,c, , Izabela Dobrosz-Gómezb, Miguel Ángel Gómez-Garcíac
a
GIPAB: Grupo de Investigación en Procesos Ambientales, Departamento de Ingeniería de Procesos, Universidad EAFIT, Cr 49 # 7 Sur 50, Medellín, Colombia
b
Grupo de Investigación en Procesos Reactivos Intensificados con Separación y Materiales Avanzados – PRISMA, Departamento de Física y Química, Facultad de Ciencias
Exactas y Naturales, Universidad Nacional de Colombia, Sede Manizales, Campus La Nubia, km 9 vía al Aeropuerto la Nubia, Apartado Aéreo 127, Manizales, Caldas,
Colombia
c
Grupo de Investigación en Procesos Reactivos Intensificados con Separación y Materiales Avanzados – PRISMA, Departamento de Ingeniería Química, Facultad de
Ingeniería y Arquitectura, Universidad Nacional de Colombia, Sede Manizales, Campus La Nubia, km 9 vía al Aeropuerto la Nubia, Apartado Aéreo 127, Manizales,
Caldas, Colombia

A R T I C L E I N F O A B S T R A C T

Keywords: In this study, the sequential Chemical Coagulation-Electro-Oxidation (CC-EO) process was proposed as an al-
Chemical coagulation ternative for the treatment of an industrial textile wastewater. Complete characterization of the effluent was
Electro-oxidation made in the terms of its organic load (Total Organic Carbon (TOC), Chemical Oxygen Demand (COD), Biological
Industrial textile wastewater Oxygen Demand (BOD5)), biodegradability (BOD5/COD ratio) and solids content (total solids and turbidity). For
BDD anode
CC, the jar test was used to determine both the most favorable dosage of coagulant and pH of the process (i.e.,
Optimization
600 mg/L of Al2(SO4)3 at pH of 9.3). CC let to remove ca. 93% of turbidity, 53% of COD and 24% of TOC. It also
increased BOD5/COD ratio of raw textile wastewater from 0.16 to 0.27. Next, CC effluent was treated by EO. Its
performance was optimized using Box-Behnken experimental Design and Response Surface Methodology. The
following EO optimal conditions were found: current density = 15 mA/cm2, conductivity = 4.7 mS/cm and
pH = 5.6. At these conditions, the sequential CC-EO process removed 100% of color, 93.5% of COD, and 75% of
TOC after 45 min of electrolysis with an estimated operating cost of 6.91 USD/m3. Moreover, the CC-EO process
yield a highly oxidized (Average Oxidation State, AOS = 2.3) and biocompatible (BOD5/COD > 0.4) effluent.

1. Introduction [6]. Nowadays, chemical oxidations by Advanced Oxidation Processes

(AOPs) are of great interest. Although these methods are effective for
Textile industry produces large amounts of effluents that are not color removal, they are also expensive and/or they can generate other
amenable to chemical or biological treatment. High content of dye- pollutants [7–9]. Recently, special attention is paid to electrochemical
stuffs, surfactants and additives, generally organic compounds of technologies due to their large number of advantages: versatility,
complex structures, causes non-biodegradability of industrial textile safety, selectivity, simplicity of automation, environmentally friendly
wastewater (ITWW). Indeed, textile dyes themselves present low or any and low investment costs [10]. However, in many cases, single treat-
biodegradability. ment process results inadequate due to the nature of the ITWW.
So far, open literature concerning ITWW treatment has been scarce. Lately, the application of combined processes has been proposed to
Holkar et al. [1] presented a complete survey about different methods overcome the disadvantages of individual ones [11–13]. Electro-
to treat textile wastewater along with cost per unit volume of treated chemical variations of AOṔs, called Electrochemical Advanced Oxida-
water. Conventional methods applied for their handling include che- tion Processes (EAOPs), such as anodic oxidation (electro-oxidation,
mical coagulation (CC), adsorption, and filtration [2]. Indeed, CC pro- EO), electro-Fenton [13], and photo-electro-Fenton [15], have received
cess exhausts a large amount of flocculent reagent and can generate an a lot of attention due to its environmental compatibility and high effi-
extra sludge [3]. On the other hand, biological methods result useless if ciency. EO consists of pollutant oxidation by direct electron transfer to
applied alone [4,5]. The recalcitrant nature of various dyes, together the anode (M), formed from water discharge on its surface at high
with their toxicity to microorganisms, makes aerobic treatment difficult current, as follows [10]:

⁎
Corresponding author at: GIPAB: Grupo de Investigación en Procesos Ambientales, Departamento de Ingeniería de Procesos, Universidad EAFIT, Cr 49 # 7 Sur 50, Medellín,
Colombia.
E-mail address: egil@eafit.edu.co (E. GilPavas).

https://doi.org/10.1016/j.jwpe.2018.01.005
Received 22 November 2017; Received in revised form 6 January 2018; Accepted 9 January 2018
Available online 03 February 2018
2214-7144/ © 2018 Elsevier Ltd. All rights reserved.
E. GilPavas et al. Journal of Water Process Engineering 22 (2018) 73–79

M + H2 O → M (HO∙) + H+ + e− (1) Aldrich; reagent grade ≥90%) was also used as explained. For iodo-
metric titration, KI (Merck, 99.5%) and Na2S2O3·5H2O (Panreac,
According to the literature [9], the most effective anode material for
99.7%) were applied.
EO is a non-active boron-doped diamond (BDD) thin-film electrode.
This is due to its inert surface at low adsorption load, corrosion stabi-
2.3. CC treatment
lity, and extremely high O2-evolution overvoltage. BDD electrodes have
been applied for the degradation of organic pollutants such as textile
Aluminum salts are the most commonly used inorganic coagulants
dyes [8], and herbicides [10]. The electrolysis of water to form HO%
due to their good performance, high efficiency, low cost, easy handling,
radicals, once the electric current passes through the BDD electrode, is
and availability. The CC efficiency and the quality of the treated ef-
presented in Eq. (2) [16]:
fluents strongly depend on the amount of coagulant and treatment’́ s pH.
∙
H2 O → HOAds + H+ + e− (2) Insufficient coagulant dosage or its overdosing can result in poor per-
formance of CC and its high cost. On the other hand, pH value affects
Different authors have recently addressed the combination of CC
the surface charge of coagulants and the stabilization of the suspension.
with an AOP. It was found to be highly efficient for the treatment of
Besides, the solubility of coagulant in aqueous solution is also influ-
wastewaters coming from various industries, among them: sugar cane
enced by pH level [4]. Following the D2035 ASTM standard method
distillery [12], landfills [6], textile [11,17,18], winery [3] and others
(v.g., the jar-test), the CC operational parameters were determined.
[19,20]. As far as we know, the coupled treatment of CC-EO for the
Thus, aluminum sulfate dosage and pH of effluent were varied in the
treatment of ITWW has not been reported.
ranges of 300–1200 mg/L and 3–9, respectively. After completing each
Here, the potential application of a sequential CC-EO process, at lab
CC treatment, supernatantś color, turbidity, TOC and COD were mea-
scale, for the treatment of textile effluents resulting from an industrial
sured. Their removal percentages were calculated as follows:
facility located in Medellín-Colombia, is presented. The ITWW was
characterized for organic matter content (Total Organic Carbon (TOC), (Absi − Abst )
Color removal (%) = %RC = x100
Chemical Oxygen Demand (COD), and Biological Oxygen Demand Absi (3)
(BOD5)), solids amount (Total solids (TS), Turbidity, Conductivity (κ))
where: Absi and Abst are the initial absorbance and the absorbance at
and biodegradability (BOD5/COD). The influence of different operating
time t, respectively.
conditions (for CC: pH and coagulant dose, and for EO: current density,
conductivity and pH) on the efficiency of CC-EO process (removal of: (Ti − Tt )
Turbidityremoval (%) = RT % = x100
color, COD, TOC and turbidity) was evaluated. The jar tests was used to Ti (4)
determine operational variable values with the most significant influ-
where: Ti and Tt are the initial turbidity value (NTU) and the turbidity
ence on CC. EÓs performance was optimized using a Box-Behnken ex-
at time t, respectively.
perimental Design (BBD) and Response Surface Methodology (RSM).
The optimal operational conditions were validated experimentally. Zi − Zt
COD or TOC degradation (%) = %DCOD or %DTOC = x100
Zi
2. Materials and methods (5)
where: Zi and Zt are the initial COD or TOC value (ppm) and the COD or
2.1. Samples handling and characterization TOC value at time t, respectively.
Finally, when CC was completed, the obtained supernatant was
ITWW samples were collected from an equalization tank in a textile directly treated by EO.
industrial plant located in Medellín (Colombia). They were dark blue
color due to the mixture of several dyes (reactive, direct, dispersive, 2.4. EO treatment
acid and cuba dyes) as well as other pollutants used during the textile
processing. ITWW samples were kept refrigerated, to avoid compounds A plexiglass, 0.12 L, batch continuously stirred jacketed reactor was
degradation during storage and transportation, following APHÁs stan- used to perform EO experiments. It consisted of two vertically placed
dard procedures [21]. Samples resulting from the laboratory tests were electrodes: (i) the anode, a boron-doped diamond (BDD) thin-film
analyzed by triplicate using a UV–vis double-beam spectrophotometer (0.5 mm thickness) on titanium substrate (supplied by NeoCoat,
(Spectronic Genesys 2PC) in the range of 200–700 nm, with a 1 cm path Switzerland), with a total area of 12.5 cm2 (2.5cm × 5 cm), and (ii) the
length quartz cell. Standard methods were followed for the quantitative cathode, an iron (99.7%) electrode, with the following dimensions:
analysis of COD (the closed reflux method with colorimetric determi- 1.8 cm × 0.55 cm × 5 cm. Basing on our preliminary results (not shown
nation, method 5220D), TOC (method 5310D), BOD5 (the respirometric here), the gap between electrodes was set at 1 cm. The voltage was
method, 5210B), TS (method 2130B), Apparent Color (AC) (method regulated with a BK-Precision source (0–30 V, 0–5 A). The reaction
2120B) and turbidity (method 2540D) [21]. Finally, H2O2 concentra- mixture of 80 mL was stirred at 240 rpm during 15 min. The test tem-
tion was measured by iodometric titration. To avoid the interference of perature was maintained at 25 °C using a Polyscience 712 thermostat
H2O2 during COD measurements, the residual H2O2 was quenched (Niles, Illinois) connected to the reactor jacket.
using MnO2.
2.5. Experimental design and statistical analysis
2.2. Reagents
Three factor’s BBD, including 15 tests, was used to optimize EO
All reagents were used as received from suppliers without any fur- performance (i.e., to maximize the organic matter degradation). The
ther purification. Their aqueous solutions were prepared using ultra- following independent variables (or operating factors) and their levels
pure water (Milli-Q system, Billerica, Massachusetts; were chosen in accordance with our previous work [22]: (i) current
conductivity < 1 μS/cm). Their pH values were adjusted with sulfuric density (j): in the range of 5–25 mA/cm2; (ii) solution conductivity (κ):
acid (H2SO4; Merck, purity = 98%) or sodium hydroxide (NaOH; Carlo ranging from 3.33 to 4.7 mS/cm; and (iii) pH, between 3 and 8.2. In all
Erba, purity = 97%). Aluminum sulfate (Al2(SO4)3.18H2O, Merck, cases, the initial pollutant concentration corresponded to the industrial
purity = 99%) was used for the chemical coagulation process. conditions of the ITWW. All experiments were made by triplicate and
Anhydrous sodium chloride (NaCl; Merck; purity ≥99%) was used as the average value of each response variable (%DCOD and operating
support electrolyte to adjust solutions conductivity. MnO2 (Sigma cost) were used for statistical analysis.

74
E. GilPavas et al. Journal of Water Process Engineering 22 (2018) 73–79

Individual and synergistic effects of the operating factors on each Table 2

response variables can be visualized in three-dimensional response %RC, %DCOD, and %RT for different coagulant dosage and pH́ s value.
surface plots. Details of the RSM can be consulted elsewhere
pH Al2(SO4)3·18H2O dose (mg/L) RC (%) DCOD (%) RT (%) pHfinal
[11,23,24]. Response variables were fitted to quadratic models of the
operating factors, as in Eq. (6), using Statgraphics Centurion XVI Soft- 9.3 300 16.75 23 17.68 6.4
ware. 9.3 400 40.70 32 32.93 6.2
9.3 600 97.27 53 93.29 5.6
3 3 3 3
9.3 900 97.89 58 93.90 4.48
Yi = β0 + ∑ βi x i + ∑ βii x ii2 + ∑ ∑ βij x i x j 9.3 1200 98.14 56 94.51 4.29
1 1 1 1 (6) 7 300 14.64 27 11.59 6.19
7 400 75.68 43 71.34 6.0
where: β0, βi, βii, and βij are the regression coefficients for the intercept, 7 600 97.02 56 92.07 4.66
linear, square, and interaction terms, respectively; and xi and xj are 7 900 97.15 56 93.29 4.27
independent variables. Regression coefficients and their effects were 7 1200 97.40 58 93.90 4.15
analyzed using analysis of variance (ANOVA), Pareto diagrams, and p- 5 300 17.12 22 9.15 3.9
5 400 24.32 26 14.63 3.78
values (with a confidence coefficient of 95%). It is important to remark
5 600 28.79 24 17.68 3.7
that RSM was not used to understand the mechanism for wastewater 5 900 39.83 28 29.27 3.63
degradation but to determine the optimal operational conditions using 5 1200 50.12 39 38.42 3.58
empirical models. 3 300 22.46 21 14.02 2.84
3 400 28.29 22 14.63 2.84
3 600 29.90 27 8.54 2.83
2.6. Operating cost 3 900 31.64 31 8.54 2.83
3 1200 39.95 34 13.41 2.82
Operating cost (OC) is one of the most significant variables for the
EO process. Indeed, it allows defining its economic feasibility. The OC
evaluation of EO process must include, at least, the energy consumption Indeed, the wastewater displays an intense blue color, due to the pre-
(EEO, kWh/m3) and reactive consumption (RC, kg/m3). In this study, sence of indigo dye, implying high content of organic compounds. In
OC (expressed in the terms of USD per m3 of treated effluent) was addition, due to the presence of salts, the sample grants high con-
calculated according to Eq. (7): ductivity. High COD value, more than twice of that of permissible limit,
infers the presence of large amounts of non-biodegradable organic
USD matter. In fact, the BOD5/COD ratio, equaled to 0.16 (< 0.4), indicates
OC ⎛ 3 ⎞ = aEEO + bRC + CC
⎝ m ⎠ (7) that the effluent is not biodegradable [25]. Although, the legally per-
where: a equals to 0.1867 USD/kWh; b varies as: 0.4 USD/kg for NaOH, missible limit for organic compounds is usually expressed in the terms
0.60 USD/kg for H2SO4 and 0.22 USD/kg for NaCl; and the cost for the of COD, in this study the TOC evolution was also monitored. This
CC process was estimated as 0.41 USD/m3 (including cost of solids parameter gives a direct measurement of carbon content without in-
disposal) [11]. The EEO was calculated as follows: terference from oxidizable inorganic substances also present in the
studied sample.
VIt
EEO =
VR (8)
3.2. The efficiency of CC for ITWW treatment
where: V is the cell voltage (V), I is the applied current (A), t is elec-
trolysis time (hour) and VR is the effluent volume (m3). The jar test was used to determine the effect of both coagulant
dosage and treatmentś pH on %RC, %DCOD and %RT. The obtained
3. Results and discussion results are presented in Table 2. Notice that the use of high coagulant
dosage and high pH value led to significant removal of color and COD.
3.1. ITWW sample characterization Under increasing dosage of coagulant from 300 to 600 mg/L, the total
COD removal increases from ca. 21–56%, while the corresponding color
A summary of raw textile wastewater characteristics together with elimination changes from ca. 14.6–97.3%, respectively. However, the
the governmental limits (according to the Colombian environmental addition of larger dosages of coagulant caused decrease in effluent pH,
legislation for ITWW) is presented in Table 1. Notice that pH, COD, and due to the hydrolysis of Al3+ ions, provoked by the release of H+ ions
BOD5 values do not fulfill Colombián environmental regulations. into the solution [7]. Contrariwise, it is well known that effective

Table 1
Main characteristics of raw ITWW, together with the corresponding legislation limits, and values obtained after CC and CC-EO processes. In addition, global treatment efficiency, at
optimized operational conditions, along with the respective process operating costs is summarized.

Parameter Raw ITWW sample Emission limita After CC process After CC-EO process Global treatment efficiency (%)

30 min 45 min 30 min 45 min

pH 9.3 6–9 5.7 4.0 4.7 – –

Abs (660 nm) 1.2 – 0.033 0 0 100 100
Conductivity (μS/cm) 4010 – 3300 4087 4100 – –
Turbidity (NTU) 161 – 11 0 0 100 100
COD (mg O2/L) 720 400 338 109.87 46.96 84.74 93.48
TOC (mg C/L) 164 – 124.4 51.12 41.53 68.83 74.68
BOD5 (mg O2/L) 115 200 90.5 60.56 59.88 47.34 47.93
BOD5/COD ratio 0.16 0.5 0.27 0.55 0.83 – –
Generated sludge (kg/m3) – – 0.456 0 0 – –
Global operating costs (USD/m3) – – 0.41 4.79 6.91 – –

a
Emission limit values for industrial wastewater discharges into the municipal sewer system according to Res. 0631, 17/03/2015, issued by the Ministry of Environment and
Sustainable Development, Colombia.

75
E. GilPavas et al. Journal of Water Process Engineering 22 (2018) 73–79

formation of flocculants is favored in a range of pH between 7 and 9, Table 4

using Al2(SO4)3·18H2O as coagulant agent [25,26]. Therefore, de- The ANOVA for %DCOD and OC, as a function of j (A), pH (B), and κ (C), according to the
BBD.
termining an optimum coagulant dosage became very important. Fi-
nally, it was possible to reach ca. 53% of COD, 97% of color and 93% of DCOD (%) Sum of squares Degree of Mean square F value P value
turbidity removals (see Table 2), using 600 mg/L of Al2(SO4)3·18H2O freedom
and pH of 9.3 (the most efficient conditions for CC treatment). These
A:j 0.2155 1 0.2155 320.04 0.0000
results are comparable with those reported by Aquino et al. [17] for an
B:pH 0.0002 1 0.0002 0.24 0.6445
industrial textile effluent characterized with initial COD and TOC of C: κ 0.0010 1 0.0010 1.41 0.2891
963 mg O2/L and 358.5 mg C/L, respectively. They suggested a coagu- AA 0.0023 1 0.0023 3.36 0.1263
lant dose (Al2(SO4)3·18H2O) of 1000 mg/L and a pH in the range of AB 0.0008 1 0.0008 1.21 0.3221
5–10 as the most suitable conditions for effective CC treatment. AC 0.0029 1 0.0029 4.33 0.0919
BB 0.0013 1 0.0013 1.88 0.2290
The supernatant obtained after CC process, realized at the most
BC 0.0003 1 0.0003 0.51 0.5078
proper conditions, was also characterized (Table 1). Although CC pre- CC 0.0028 1 0.0028 4.22 0.0950
treatment eliminates wastewater turbidity, it was inefficient in re- Error total 0.0034 5 0.0007
moving surfactants and other soluble organic pollutants existing in the Total (corr.) 0.2308 14
R2 = 98.54% R2adj = 95.92%
sample. Notice that CC was able to reduce efficiently organic load (e.g.,
COD decreases from 720 to 400 mg O2/L). However, the BOD5/COD OC (USD/m3)
ratio increased only from 0.16 to 0.27, indicating that this effluent still A:j 36.7691 1 36.7691 1590.63 0.0000
B:pH 0.0034 1 0.0034 0.14 0.7192
cannot be considered as biodegradable. This implies that an additional C: κ 0.3883 1 0.3883 16.80 0.0094
wastewater treatment is required before its final discharge into the AA 0.5928 1 0.5928 25.65 0.0039
surface waterbodies. Thus, EO was chosen to complete the ITWW AB 0.0384 1 0.0384 1.66 0.2541
handling and to reach the permissible legislation discharge limits. AC 0.0640 1 0.0640 2.77 0.1572
BB 0.0306 1 0.0306 1.32 0.3022
BC 0.0218 1 0.0218 0.94 0.3764
3.3. Electro-oxidation (EO) process CC 0.0447 1 0.0447 1.93 0.2231
Error total 0.1156 5 0.0231
Total (corr.) 38.0275 14
EO process was applied to reduce COD concentration and to in- R2=99.70% R2adj = 99.15%
crease the biodegradability of the effluent resulting from CC process.
The BBD and RSM analysis were used to determine the individual ef-
fects of EO operational parameters and their interactions on EO effi- (or factors) and their double-interactions in the variance of each re-
ciency. The experimental design together with the obtained results is sponse variable. The p-values were used to identify experimental
presented in Table 3. Notice that the additional COD removal percen- parameters that present statistical influence on particular response. If p-
tages achieved by EO varied between ca. 17 and 55%. The following value is lower than 0.05, it is considered that the specific variable
most efficient experimental conditions were found: j = 25 mA/cm2, shows statistical significance within the 95% confidence level [22].
pH = 5.6, and κ = 4.7 mS/cm. However, the OC was considerably high The ANOVA results are presented in Table 4. Notice that the p-value
at such settings. To optimize the EO operating conditions (v.g., both to of j was lower than 0.05 for both response variables (%DCOD and OC).
maximize %DCOD and to minimize OC), the data were adjusted to a In the case of OC, κ and j-j interactions also presented this character-
second order polynomial (Eqs. (9) and (10)). These empirical models istic. Therefore, they can be considered as statistically relevant in the
represent the %DCOD and OC as a function of three independent pro- performance of EO process. The factor with the major effect on %DCOD
cess variables (j, pH, and κ), respectively. Model predictions match was j. Indeed, the oxidation of pollutants and production of oxidizing
accurately to the experimental results, with determination coefficients species depend on the potential difference in the electrochemical cell.
of R2 = 98.54% and R2adj = 95.92% for %DCOD and of R2 = 99.70% The use of high j values maintains the anode activity and promotes the
and R2adj = 99.15% for OC (Table 4). simultaneous oxidation of pollutants and water. Conversely, at low j
The ANOVA was used to determine statistical significance of main values, O2 evolution is avoided causing the loss of anode activity due to
factors and their interactions. It consists of classifying and cross-clas- the adsorption on its surface of intermediates formed from direct anodic
sifying statistical results, decomposing the contribution of each variable oxidation, making this procedure not useful for wastewater remediation
in practice [10]. Such effect on the response variable will be discussed
Table 3 subsequently.
The experimental and predicted results of %DCOD and OC, according to the BBD.
Electrolysis time 15 min. %DCOD = 1.5456 + (0.0049*j ) − (0.058*pH ) − (0.579*κ) − (0.0003*(j )2)
Run j (mA/cm2) pH κ (mS/cm) DCOD (%) OC (USD/m3) + (0.0006*j *pH ) + (0.0039*j *κ) + (0.0027*(pH )2)
+ (0.0052*pH *κ) + (0.0591*(κ)2) (9)
yexp ypred yexp ypred

1 5 8.2 4.015 17 17 0.772 0.7942 USD

OC ⎛ 3 ⎞ = 3.1017 + (0.1894*j ) + (0.0799*pH ) − (1.6948*κ)
2 15 3 4.7 36 38 2.245 2.3620 ⎝ m ⎠
3 25 8.2 4.015 51 53 4.892 4.8860
4 15 5.6 4.015 36 35 2.229 2.3280 + (0.004*(j )2) − (0.0038*j *pH ) − (0.0185*j *κ)
5 5 3 4.015 19 19 0.552 0.5574 + (0.0135*(pH )2) − (0.0414*pH *κ) + (0.2344*(κ)2) (10)
6 15 8.2 4.7 38 40 2.155 2.2554
7 25 5.6 3.33 49 51 5.206 5.3293 Response surface plots were used to observe the simultaneous ef-
8 5 5.6 4.7 17 16 0.724 0.6010
9 15 5.6 4.015 33 35 2.333 2.3280
fects of two factors over each response variable. They were constructed,
10 25 3 4.015 49 49 5.064 5.0410 from Eqs. (9) and (10), for a fixed pH value of 5.7 (Figs. 1(a) and (b)).
11 15 5.6 4.015 38 35 2.422 2.3280 As it can be seen, the application of the highest j let to achieve the
12 15 8.2 3.33 42 41 2.961 2.8435 highest %DCOD, ca. 53%, if both of independent parameters (j and κ)
13 15 3 3.33 43 42 2.755 2.6550
were kept at their maximum values. Consequently, OC also increased.
14 25 5.6 4.7 55 54 4.730 4.6358
15 5 5.6 3.33 22 23 0.694 0.7887 The enhancement in COD degradation as well as in decolorization ef-
ficiency, obtained under increasing j, can be associated to the

76
E. GilPavas et al. Journal of Water Process Engineering 22 (2018) 73–79

Fig. 1. Response surface diagrams for the interactive

effect of j and κ on %DCOD (a) and OC (b) for EO
process (electrolysis time = 15 min, temperature =
25 °C, pH = 5.6).

acceleration of reaction (1) on the anodic surface, which regulates the different intermediates [9]. These compounds can be produced through
amount of produced oxidant (HO%). Thus, larger quantities of (HO%) a number of reaction pathways, including oxidation, addition and
can be generated (Eq. (2)) and the conjugated chromophore system of substitution [15]. Low efficiencies of EO process realized at high κ
organic matter (OM) can be destroyed [13]. In particular, it is known (Figs. 1a and 1b) can be related to the production of persistent oxy-
that (HO%) radicals can be completely mineralized, due to their high chloro species that delay or even inhibit complete mineralization [27].
oxidation potential (E0 = 2.80 V), contrary to active chlorine, leading On the other hand, low chloride concentrations can also contribute to
to contaminant transformation into CO2 and H2O [10]. Regarding to the organic matter mineralization due to the formation of chlorine radicals
effect of κ (e.g., NaCl), although the derived chloro oxidant species are (Cl% and Cl2%−) [21]. Finally, the pH value of 5.6 was slightly bene-
not capable of mineralizing textile wastewater, owing to their low ficial to COD removal. Thus, as it can be seen in Table 4 (e.g. runs 2, 6 y
oxidation potentials (E0 = 1.36 V for chlorine, 1.63 V for hypochlorous 9), at studied conditions, effluentś pH does not present influence on the
acid and 0.90 V for hypochlorite), they can result in the formation of rate of removal of the organic load. This probably because under acidic

77
E. GilPavas et al. Journal of Water Process Engineering 22 (2018) 73–79

conditions, the formation of chlorates (ClO3−) and perchlorates

(ClO4−), mediated by HO% radicals and the consequent depletion of
these radicals at the electrode surface, is reduced [14]. Panizza and
Cerisola [28] also obtained similar results, for the electrochemical de-
gradation of an azo dye, using BDD and PbO2 anodes. Fig. 1b shows
high increase in OC with an increasing current density, as expected by
the increase in applied cell voltage. Indeed, applied j is a key factor in
electrochemical treatments and it deserves special attention in view of
its economic as well as performance implications. Thus, the optimal
conditions that let to achieve simultaneously high %DCOD and low OC
(USD/m3) were obtained from RSM models and calculated by Minitab
15 software. The following conditions were selected to study the evo-
lution of the reaction with time: j = 15 mA/cm2, pH = 5.6 and κ = 4.7
mS/cm.

3.4. The evolution of color, %DCOD and%DTOC during the sequential CC-
EO process Fig. 3. Chemical and biological characteristics of wastewater during CC-EO process.
Experimental conditions: CC (Al2(SO4)3·18H2O) = 600 mg/L and pH (9.3). EO process
pH = 5.6, κ = 4.7 mS/cm, j = 15 mA/cm2, BDD anode and Fe cathode.
Validation of results obtained by BBD-RSM methodology was per-
formed experimentally at optimized conditions. The variations of color
removal, %DCOD and %DTOC, as a function of time, during the se- COD
AOS = 4 − 1.5
quential CC-EO process are summarized in Fig. 2. COD removal effi- TOC (11)
ciency felt within the 95% prediction intervals, confirming the model’s where: TOC and COD are expressed in mg of C/L and mg of O2/L,
prediction capability. During CC stage, color removal reaches ca. 97%. respectively. The AOS can take a value between −4 (for the most re-
Further EO treatment, after 30 min of electrolysis, let to reach ca. 100% duced state of carbon, CH4) and +4 (for the most oxidized state of
of discoloration. On the other hand, the combined CC-EO let to decrease carbon, CO2). The CC treatment produces only a small amount of oxi-
COD and TOC in ca. 94% and 75%, respectively (after 45 min of elec- dized organic intermediates (AOS value is equal to ca. −0.0756). The
trolysis). Fig. 2 also presents the OC estimated for the CC-EO process. AOS rises gradually up to ca. 2.3, after 45 min of EO (Fig. 3). This
OC highly increases with electrolysis time, with an associated operating implies that organic intermediates go through different oxidation-mi-
cost of ca. 7 USD/m3 after time 45 min of electrolysis. Notice that OC neralization reactions. Even if high AOS values are frequently asso-
can be reduced to 4.8 USD/m3 (with expected removal of ca. 100, 85, ciated to biocompatible solutions, this is not always the case for traces
and 69% of color, COD, and TOC, respectively, and BOD5/COD of highly toxic compounds. The BOD5/COD ratio is customarily used for
ratio = 0.55), if 30 min of electrolysis time is used during the EO pro- the assessment of physicochemical processes as a pretreatment before
cess. Under these conditions, the Colombian permissible limits for the the biological ones [30]. Fig. 3 shows its evolution during EO. The
industrial wastewater discharges are fulfilled. BOD5/COD of raw wastewater was 0.16 (Table 2), indicating that the
biodegradability of this wastewater was very low. This value increased
3.5. An improvement of wastewater biodegradability and sequential up to ca. 0.27 after the CC process and then progressively up to ca. 0.83
treatment efficiency after 45 min of electrolysis. This last value implies that the final effluent
is highly biocompatible (BOD5/COD ratio > 0.4).
Useful additional information, regarding on the reaction progress
during wastewater treatment, can be obtained analyzing the Average 4. Conclusions
Oxidation State (AOS) of the treated sample (Fig. 3). AOS was calcu-
lated as follows [29]: According to the findings of this research, sequential Chemical
Coagulation (CC, using Al2(SO4)3·18H2O as coagulant) – Electro-
Oxidation (EO, using BDD anode) process appears as an efficient al-
ternative for the treatment of textile wastewater characterized with low
biodegradability and high amount of recalcitrant compounds. The
ecological parameters such as color, COD and TOC decreased sig-
nificantly. The obtained BOD5/COD ratio implies further biological
processes to work more effectively.
CC stage removed ca. 97% of color, 53% of COD and 24% of TOC.
Nevertheless, its biodegradability improvement was quite low (BOD5/
COD ratio increased from ca. 0.16–0.27), excluding biological methods
as alternative for the following stage of its treatment. At these levels,
the CC presents an operating cost of 0.39 USD/m3.
The CC effluent was subsequently treated by EO. Its operational
conditions were optimized by BBD and RSM (j = 15 mA/cm2, κ = 4.7
mS/cm, and pH = 5.6). At such conditions, after 45 min of electrolysis,
removals of 100% of color, > 90% of COD and > 70% of TOC were
reached. Moreover, wastewater biodegradability increased up to BOD5/
COD = 0.83, with an operating cost of 6.91 USD/m3. Finally, during
Fig. 2. The evolution of color, COD and TOC removal and the OC of wastewater treated the EO process, the AOS raised from −0.0756 to 2.3.
using CC-EO process. Experimental conditions: CC (Al2(SO4)3·18H2O) = 600 mg/L and If 30 min of electrolysis is applied, at optimized conditions, high
pH (9.3). EO process pH = 5.6, κ = 4.7 mS/cm, j = 15 mA/cm2, BDD anode and Fe efficiencies can be also reached (v.g., reductions of 100% of color, 85%
cathode.
of COD, 69% of TOC and BOD5/COD of 0.55) with lower operating

78
E. GilPavas et al. Journal of Water Process Engineering 22 (2018) 73–79

costs (ca. 4.8 USD/m3), accomplishing Colombian permissible limits for [13] J. Zou, X. Peng, M. Li, Y. Xiong, B. Wang, F. Dong, B. Wang, Electrochemical oxi-
the textile industrial wastewater discharges. dation of COD from real textile wastewaters: kinetic study and energy consumption,
Chemosphere 171 (2017) 332–338.
[14] K. Cruz, O. Torres, A. García, E. Brillas, A. Hernández, J.M. Peralta, Optimization of
Acknowledgements electro-Fenton/BDD process for decolorization of a model azo dye wastewater by
means of response surface methodology, Desalination 286 (2012) 63–68.
[15] J. Ruiz, C. Arias, E. Brillas, A. Hernández-Ramírez, J.M. Peralta-Hernández,
The authors thank to the Dirección de Investigación de la Mineralization of Acid Yellow 36 azo dye by electro-Fenton and solar photoelectro-
Universidad EAFIT, Medellin–Colombia for financial support of this Fenton processes with a boron-doped diamond anode, Chemosphere 82 (2011)
research. The staff of the Laboratorio de Ingeniería de Procesos is also 495–501.
[16] Y. Yavuz, S. Koparal, U. Ogutveren, Treatment of petroleum refinery wastewater by
recognized for their participation. electrochemical methods, Desalination 258 (2010) 201–205.
[17] J. Aquino, G. Pereira, R. Rocha, N. Bocchi, S. Biaggio Combined Coagulation and,
References Electrochemical process to treat and detoxify a real textile effluent, Water, Air, Soil
Pollut. 227 (8) (2016) 266.
[18] D. Manenti, A. Módenes, P. Soares, F. Espinoza, R. Boaventura, R. Bergamasco,
[1] C.R. Holkar, A.J. Jadhav, D.V. Pinjari, N.M. Mahamuni, A.B. Pandit, A critical re- V.J. Vilar, Assessment of a multistage system based on electrocoagulation solar
view on textile wastewater treatments: possible approaches, J. Environ. Manag. 182 photo-Fenton and biological oxidation processes for real textile wastewater treat-
(2016) 351–366. ment, Chem. Eng. J. 252 (2014) 120–130.
[2] F. El-Gohary, A. Tawfik, Decolourisation and COD reduction of disperse and re- [19] L. Guerreiro, C.S.D. Rodrigues, R.M. Duda, R.A. Oliveira, R.A. Boaventura,
active dyes wastewater using chemical-coagulation followed by sequential batch L.M. Madeira, Treatment of sugarcane vinasse by combination of coagulation/
reactor (SBR) process, Desalination 249 (2009) 1159–1164. flocculation and Fenton’s oxidation, J. Environ. Manag. 181 (2016) 237–248.
[3] N. Amaral-Silva, R.C. Martins, C. Paiva, S. Castro-Silva, R.M. Quinta-Ferreira, A new [20] E. Hernández-Francisco, J. Peral, L.M. Blanco-Jerez, Removal of phenolic com-
winery wastewater treatment approach during vintage periods integrating ferric pounds from oil refinery wastewater by electrocoagulation and Fenton/photo-
coagulation Fenton reaction and activated sludge, J. Environ. Chem. Eng. 4 (2016) Fenton processes, J. Water Process Eng. 19 (2017) 96–100.
2207–2215. [21] Standard Methods for the Examination of Water and Wastewater, 22th edn., APHA,
[4] S. Bhatia, Z. Othman, A.L. Ahmad, Coagulation-floculation process for POME American Public Health Association, Washington DC, USA, 2012.
treatment using Moringa oleifera seeds extract: optimization studies, Chem. Eng. J. [22] E. GilPavas, I. Dobrosz-Gómez, M.A. Gómez-García, Electrochemical degradation of
133 (2007) 205–212. acid yellow 23 by anodic oxidation–optimization of operating parameters, J.
[5] P.A. Soares, R. Souza, J. Soler, T.F. Silva, S. Guelli, R.A.R. Boaventura, V.J.P. Vilar, Environ. Eng. 142 (2016) 5 (ASCE).
Remediation of a synthetic textile wastewater from polyester-cotton dyeing com- [23] M.S. Secula, C.S. Stan, C. Cojocaru, B. Cagnon, I. Cretescu, Multi-objective opti-
bining biological and photochemical oxidation processes, Sep. Purif. Technol. 172 mization of indigo carmine removal by an electrocoagulation/GAC coupling process
(2017) 450–462. in a batch reactor, Sep. Sci. Technol. 49 (6) (2014) 924–938.
[6] C. Amor, J.A.P. Torres-Socías, M.I. Maldonado, I. Oller, S. Malato, M.S. Lucas, [24] F. Ghanbari, M. Moradi, A comparative study of electrocoagulation, electro-
Mature landfill leachate treatment by coagulation-flocculation combined with chemical Fenton, electro-Fenton and peroxi-coagulation for decolorization of real
Fenton and solar photo-Fenton processes, J. Hazard. Mater. 286 (2015) 261–268. textile wastewater: electrical energy consumption and biodegradability improve-
[7] A.K. Verma, R.R. Dash, P. Bhunia, A review on chemical coagulation/flocculation ment, J. Environ. Chem. Eng. 3 (2015) 499–506.
technologies for removal of colour from textile wastewaters, J. Environ. Manag. 93 [25] M. Lapertot, S. Ebrahimi, S. Dazio, A. Rubinelli, C. Pulgarin, Photo-Fenton and
(2012) 154–168. biological integrated process for degradation of a mixture of pesticides, J.
[8] E. GilPavas, I. Dobrosz-Gómez, M.A. Gómez-García, Decolorization and miner- Photochem. Photobiol. A: Chem. 186 (2007) 34–40.
alization of Diarylide Yellow 12 (PY12) by photo-Fenton process: the Response [26] V. Golob, A. Vinder, M. Simonic, Efficiency of the coagulation/flocculation method
Surface Methodology as the optimization tool, Water Sci. Technol. 65 (10) (2012) for the treatment of dyebath effluents, Dyes Pigm. 67 (2005) 93–97.
1795–1800. [27] M. Sales, C.K. Costa de Araújo, J. Vieira de Melo, J.M. Peralta-Hernandez,
[9] F.C. Moreira, R.A.R. Boaventura, E. Brillas, V.J.P. Vilar, Electrochemical advanced D. Ribeiro da Silva, C.A. Martínez-Huitle, Decontamination of real textile industrial
oxidation processes: a review on their application to synthetic and real wastewaters, effluent by strong oxidant species electrogenerated on diamond electrode: viability
Appl. Catal. B: Environ. 202 (2017) 217–261. and disadvantages of this electrochemical technology, Appl. Catal. B: Environ.
[10] E. Brillas, C.A. Martínez-Huitle, Decontamination of wastewaters containing syn- 130–131 (2013) 112–120.
thetic organic dyes by electrochemical methods. An updated review, Appl. Catal. B: [28] M. Panizza, G. Cerisola, Electrochemical degradation of methyl red using BDD and
Environ. 166–167 (2015) 603–643. PbO2 anodes, Ind. Eng. Chem. Res. 47 (2008) 6816–6820.
[11] E. GilPavas, I. Dobrosz-Gómez, M.A. Gómez-García, Coagulation-flocculation se- [29] F. Vogel, J. Harf, A. Hug, P.R.V. Rohr, The mean oxidation number of carbon (MOC)
quential with Fenton or Photo-Fenton processes as an alternative for the industrial a useful concept for describing oxidation processes, Water Res. 34 (2000)
textile wastewater treatment, J. Environ. Manag. 191 (2017) 189–197. 2689–2702.
[12] C.S. Rodrigues, A. Neto, R. Duda, R. Oliveira, R. Boaventura, L.M. Madeira, [30] Y.C. Hsu, H.C. Yang, J.H. Chen, The enhancement of the biodegradability of phe-
Combination of chemical coagulation photo-Fenton oxidation and biodegradation nolic solution using preozonation based on high ozone utilization, Chemosphere 56
for the treatment of vinasse from sugar cane ethanol distillery, J. Clean. Prod. 142 (2) (2004) 149–158.
(2017) 3634–3644.

79