Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=geac20
International Journal of Environmental Analytical
Chemistry
ISSN: 0306-7319 (Print) 1029-0397 (Online) Journal homepage: https://www.tandfonline.com/loi/geac20
Treatment of CNC industry wastewater by
electrocoagulation technology: an application
through response surface methodology
Muhammed Kamil Oden
To cite this article: Muhammed Kamil Oden (2020) Treatment of CNC industry wastewater by electrocoagulation technology: an application through response surface methodology, International Journal of Environmental Analytical Chemistry, 100:1, 1-19, DOI: 10.1080/03067319.2019.1628955
To link to this article: https://doi.org/10.1080/03067319.2019.1628955
Published online: 19 Jun 2019.
Submit your article to this journal
Article views: 227
View related articles
View Crossmark data
ARTICLE
Treatment of CNC industry wastewater by electrocoagulation
technology: an application through response surface
methodology
Muhammed Kamil Oden
Department of Environmental Protection Technology, Selcuk University, Sarayönü Vocational High School, Konya, Turkey
ABSTRACT
Removal of COD, and several toxic heavy metals (Cu2+ and Ni2+) from CNC (metalworkingfluid) wastewater was investigated using electrocoagulation method (EC) with Fe and Al electrodes. The interaction effects of the current density, reaction time and initial pH were analyzed and were correlated to assess the removal efficiencies for COD, copper, and nickel. Coefficient of determina-tion (R2) and adjusted R2was found to be higher than 96.81% and 92.77; 99.01% and 89.94 for all responses at Fe and Al electrodes, respectively. Removal efficiencies were determined to be 95.72%, 96.03%, 95.22% and 97.11%, 98.51%, 92.49% for COD, copper and nickel at iron and aluminum electrodes, respectively under opti-mum operating conditions. The operational cost of the EC process for COD, copper, and nickel removal, were found to be 2.54, 3.36, 2.50€/m3for iron electrode and 7.16, 8.95, 8.50€/m3for aluminum electrode at optimum conditions, respectively. The results provide that The EC process seems to be an effective treatment method for removing COD and several trace heavy metals from the CNC machine (metalworkingfluid) wastewater.
ARTICLE HISTORY Received 10 April 2019 Accepted 30 May 2019 KEYWORDS CNC machine wastewater; metalworkingfluid; electrocoagulation; heavy metal removal; response surface methodology
1. Introduction
Some industrial and technological changes and reforms have been experienced with the advent of industrial transformation in our country and all over the world. These
advances accelerated innovations and mechanization in allfields in order to meet the
basic and other needs of the increasing number of the world population. These new-technology machines facilitated production but they also increase the amount of waste-water and diversity of pollutants. When industrial wastewaste-waters are discharged into receiving environments without treatment, this may result in permanent environmental problems.
Given that CNC (Computer Numerical Control) automated a considerable amount of
the tasks made by machinists, it was considered as one of the most significant
innova-tions of the 20th century in the manufacturing industry. CNC improved freedom for the
CONTACTMuhammed Kamil Oden muhammedkoden@gmail.com
This article has been republished with minor changes. These changes do not impact the academic content of the article.
2020, VOL. 100, NO. 1, 1–19
https://doi.org/10.1080/03067319.2019.1628955
designs and process repeatability and shortened delivery times in production. CNC technology, apart from mass production, facilitated the production of almost any
geometrical shape [1]. In our country, CNC machine tools have been used since 1984
[2]. Figure 1 shows the steps from the production design to the production stage in CNCmachine. Characteristics of industrial wastewater are totally dependent on the kind
of production and diversity of the product [3]. It is very challenging to treat industrial
wastewaters effectively because of the large variations in their composition and their
amount [4]. Dry production in metalworking process can shorten the life of the machine
used, and may give harm to final characteristics of the product [5]. Therefore, cutting
fluid is commonly utilized for surface cleaning, cooling and corrosion prevention in the
course of the machining [6,7] and oily wastewater occurs widely in mechanic industries
[8]. Cooling liquid extends tool life and helps to obtain a good surface quality. Several
chemicals are used to this end, and they are called generally called cooling oil or boron
oil [9]. Among the environmental impacts of CNC machining are cutting fluids (waste
management, cleaning and work environmental issues), noise, lubricants and various
other process waste (metal chips and cutting tools etc.) [1,10]. There are large amounts
of organic and inorganic materials with oil, grease and heavy metals in the wastewaters
resulting from the metalworking industry [11]. Although the liquid wastes discharged
from these industries are not voluminous, they are highly dangerous due to their toxic
contents [12,13].
A look at the related studies shows that several processes or treatment techniques
are used in the removal of pollutants in different characteristics from industrial
waste-water. This includes examples such as treatment of wastewater by Fenton process [14]
and Sono-Fenton process [15], oxidation process [16], ozonation [17], electrocoagulation
[18,19], coagulation [20], coagulation-UV/H2O2process [21], adsorption process [22], ion
exchange process [23], membrane filtration process [24], biological treatment [25].
Recent research shows that electrocoagulation (EC) process is preferred more than
chemi-cal coagulation (CC) process thanks to some advantages it offers [26]. Electrocoagulation is
a process which includes the generation of coagulants from an electrode through the effect of
electric current carried out to these electrodes. The ions are drawn by the colloidal particles,
their charge is neutralized and allow for theirflocculation. The hydrogen gas released from the
cathode gets into interaction with the particles resulting inflocculation, causing the
unsoli-cited material to rise and be removed removed [27,28]. EC process involves a large number of
advantages in comparison to chemical coagulation process. These include no requirements for chemicals, simple equipment, more stable and less amount of sludge, and easy operation
in addition to low operating costs [27,29]. In recent years, studies on the treatment of various
wastewaters show that the interest in electrochemical treatment methods such as
electro-oxidation (EO), electrocoagulation (EC) and electroflotation (EF) has increased [30]. A variety of
metals such as stainless steel, aluminum, iron, and platinum have been used in during
operation as electrodes [28,31]. Iron and aluminum electrodes are widely used for the
electrocoagulation process [32]. Dissolved metal ions make up several charged hydroxylated
species during the electrocoagulation process. Being adsorbed on Al(OH)3(s)and Fe(OH)3(s),
these species function as coagulants and make the removal of pollutants in wastewater easier. The main reactions taking place in the EC process for iron and aluminum electrodes are shown
in Equtions (1–7) [7,18,33]. Anode : Al! Al3þþ3e (1) Fe! Fe2þþ2e1 (2) Cathode : 3H2Oþ 3e! 3=2H2þ3OH (3) 2H2Oþ 2e! H2þ2OH (4) Overall : Feþ 2H2O$ Fe OHð Þ2þH2 (5)
4Fe2þþO2þ10H2O! 4Fe OHð Þ3þ4H2 (6)
Cationic monomeric species are produced during the electrocoagulation of the aluminum
anode such as Al3+and Al(OH)2at low pH, At appropriate pH values, it is initially converted
to Al(OH)3andfinally polymerized to Aln(OH)3naccording to the following reactions:
Overall : Al3þðaqÞþ3H2O$ Al OHð Þ3þ3HþðaqÞ (7)
All of the hydroxide flocs generated assist as coagulants, colloidal and dissolved
contaminant material in the wastewater become destabilized and come in sight as
precipitate in the reactor [34].
The current study investigated the removal efficiency of the electrocoagulation on
the wastewater of the metalworking process. The process efficiency is evaluated by
analyzing its effect on different operating factors (initial pH, current density and
electro-lysis time) with the purpose offinding the most sufficient operating conditions for iron
and aluminum electrodes. Optimum operating parameters for COD, copper and nickel
removal in the EC process were analyzed with the removal efficiencies obtained.
Response surface methodology (RSM) drawing on central composite design (CCD) was
utilized, by developing a mathematical model to study the interactive effects of the
2. Experimental
2.1. Characterization of CNC process wastewater
Wastewater samples used in this study were taken from the storage tank at the CNC
machine in the metal cutting workshop in Konya City (Turkey) at different times. Being
a composite one, the wastewater used in the experiments was mixed and utilized. The
characteristics of CNC (metalworking fluid) wastewater are shown in Table 1. The
analyses of heavy metals and other parameters were performed using a PerkinElmer ICP Mass spectrometer (ICP-MS) NexION 350X and Hach Lange Dr 2800. In the course of
the experimental study, the process scheme inFigure 2was used.
2.2. Experimental details and procedure
In the course of the experimental study, the process scheme inFigure 2was used. As can be
seen inFigure 2, two electrodes (an anode and a cathode) spaced by 3 cm were placed in
the reactor in a vertical manner. A laboratory-scale EC reactor consists of Plexiglas being 9 cm in diameter cm 13 cm in height. The iron and aluminum electrodes which function as one anode and one cathode had the sizes of 6 cm width, 11.5 cm height and 0.1 cm
thickness. The total area of the effective electrode approximately was 85 ± 5 cm2. Mervesan
305D II was used in the EC experiments as the power supply. The volume of the
metalwork-ingfluid wastewater used in the experiments was 500 mL. The wastewater was stirred at
200 rpm speed by using a Velp brand magnetic stirrer. In advance of each of the experi-mental batch tests, the surfaces of the electrodes were cleaned with acetone.
Afterward, they were kept in a cleaning solution (100 mL of 35% HCl and 200 mL of 2.8%
C6H12N4) for at least 5 min, followed by the rinsing up with tap water. In the EC experiments,
optimization was performed under the conditions of 16–48 mA/cm2current density, a
reac-tion time of 10–30 min and pH of 4–10. In order to increase conductivity from 1.756 over
10.0 mS/cm, NaCl was added so that salinity was kept at the same level.
By taking samples from supernatant 120 min after each experimental run, analyses were conducted. The supernatant samples were preserved under favorable conditions for further analyses. Chemicals used in the experimental studies had an analytical-reagent grade and
the analyses were performed in accordance with standard methods [35].
2.3. Experimental design and statistical analysis
In wastewater treatment research, a few multivariate statistical models have been utilized in order to optimize the processes recently. In order to understand and optimize the structure of the EC reactor, the powerful experimental design tool whose name is response surface methodology (RSM), particularly central composite design (CCD) has been used. According to the studies, CCD is a statistical technique using quantitative data obtained from experi-ments in order to determine a regression model equation among operating parameters. Among these statistical models, the response surface methodology (RSM) is a widely used
methodology in variousfields for improvement and optimization processes [32,36–40]. By
using the experimental design matrix which is the most popular design under RSM, Optimization process was carried out with the help of Statgraphics Centurion XVI.I software.
Table 1. Characterization of metalworking fluid wastewater. Parameters pH Conductivity, µS/cm Temp., °C COD, mg/L TSS, mg/L Nitrite, mg/L Chloride, mg/L Cr +6 , mg/L Nickel, mg/L Zinc, mg/L Iron, mg/L Copper, mg/L Mean <9 1756 25,1 32,350 25 0.7 950 23.1 54.8 17.1 154 77,5 Value ±0.3 ±80 ±1 ±500 ±1,5 ±0.1 ±20 ±0.9 ±1.3 ±0.5 ±1,4 ±1
RSM can be utilized for resolving the main (X1, X2and X3), interactive (X1X2, X1X3and X2
X3) and curvature effects (X12, X22and X32). 15 experiments were carried out with three
central points and three parameters which were the current density (X1: 16–48 mA/cm2),
electrolysis time (X2: 10–30 min) and initial pH (X3: 4–10). These parameters are considered
relevant independent factors which affect the removals of COD, copper, and nickel as
responses. The coded, symbol and variables for the EC process are shown inTable 2.
Statistical analysis of the data and the analysis of the variance: In order to estimate
the statistical significance of the data, Statgraphics Centurion XVI.I software which
utilizes the Fisher’s F test was used. The conformity of the model can be verified in
relation to the coefficient of determination (R2) by using the respective probability
values the same software. Depending on the effects of independent factors,
three-dimensional graphics and related shape parcels were obtained.
3. Results and discussion
3.1. Statistical analysis and experimental modeling
In the present study. the experimental three variables and factors are employed to optimize the EC process by response surface methodology (RSM). RSM uses mathematical and
Figure 2. Schematic representation of EC in the laboratory (1: Reactor; 2:Electrode; 3: Magnetic Stirrer Bars; 4: Magnetic Stirrer; 5: DC Power Supply).
Table 2.Coded levels and independent variables for EC process.
Coded factors
Symbol and variables
X1: Current density (mA/cm2) X2: Electrolysis time (min) X3: initial pH
−1 16 10 4
0 32 20 7
statistical techniques for optimizing processes. In order tofit the least squares technique into a model, RSM utilizes an experimental model such as Central Composite Design (CCD)
[37,41]. The RSM used three variables (initial pH. current density and operating time) in order
to determine process responses such as copper, COD and nickel removal efficiencies,
amounts of the electrode (ELC) and energy (ENC) consumptions. Level of these three
variables is shown in Table 2. A second-order polynomial response surface model was
used tofit the experimental data obtained by CCD (Equation 4) to identify. An estimate of
the electrocoagulation test results is obtained from this model.
Y¼ β0þ β1X1þ β2X2þ β3X3þ β11X12þ β22X22þ β33X32þ β12X1X2þ β13X1X3
þ β23X2X3 (4)
X1, X2and X3symbolize the coded levels of design variables (X1: current density; X2: time and
X3: initial pH). Interaction is shown withβ12,β13,β23,quadratic coefficients with β11,β22,β33,
linear withβ1,β2,β3,intercept withβ0and set of regression coefficients with β. Y represents
the predicted response (% removal efficiency) (Y1: COD removal efficiency. %; Y2: copper
removal; and Y3: Nickel removal). The wastewater treatment data is defined in the program
with meaning equations. The regression equations of the model test for Fe and Al electrodes
are shown inTable 3.Table 4shows the effects of 15 experiments on COD, copper and nickel
removal efficiencies. Removal efficiencies obtained from those three parameters on the
variables are shown inTable 4for Fe electrode and for Al electrode. Positive and negative
figures in front of the values of equations represent the synergistic and antagonistic effects
[39]. In Table 5, variance analysis (ANOVA) was calculated for the quadratic polynomial
model of the response surface model. The significance of regression was confirmed by
F-test. Correlation coefficients (R2and adjusted R2) were used to verify the concordance of
quadratic equations. The obtained p-values give information about whether the F values are
statistically significant. The quality of fitness (Table 7) are as follows: p–value of Models <
0.05, p–value of Lack of Fit (f-value)>0.05.Values of Prob>F were between 0.003–0,04 or less
than 0.04 showed that the model was statistically significant. As it can be observed from
Table 7for Fe and Al electrodes. R2and adjusted R2values were found to be higher than
92.77% and 79.76; 89.94% and 71.84 respectively. R2 values are listed from high to low
copper>nickel> COD for Fe electrode and COD>nickel>copper for Al electrode. The model F-values of 7.13, 16.90, 8.00 for iron electrode and 55.90, 4.96, 10.13 for Al electrode implied
significant models for COD, copper, and nickel removal efficiencies for the EC process,
respectively.
3.2. Process optimization and effect of operating variables on COD, copper, and nickel removals
For investigating the COD, copper and nickel metals removal with an iron and aluminum
as the electrode, current density, pH and time variations were used. Table 6 shows
ANOVA results of the response surface quadratic model of the COD, copper and nickel
removals at Fe electrode. As seen fromTable 6, linear coefficients of the current density
(X1) and pH (X3) on COD and nickel removal efficiency, and electrolysis time (X2) and pH
(X3) on copper removal efficiency have significant effects for an iron electrode.
The second order terms of the current density (X1X1) have also significant effects on
Table 3. Regression equations by EC process using Fe and Al electrodes. Electrode Parameters Equations Fe COD removal. % Y1 = 94.6555 –0.671331* X1 – 0.253863* X2 + 1.90225* X3 + 0.0108501* X1 2 – 0.0107227* X1 *X 2 + 0.0197391* X1 *X 3 + 0.0248048* X2 2 -0.0480808* X2 *X 3 – 0.140325* X3 2 Copper removal (%) Y2 = 47.0486 + 1.46476*X 1 + 2.93071*X 2 – 3.75744*X 3 – 0.0180185* X1 2 – 0.0300403*X 1 *X 2 + 0.032426*X 1 *X 3 – 0.0312886*X 2 2 – 0.0642475*X 2 * X3 + 0.355754*X 3 2 Nickel removal (%) Y3 = 11.1991 + 1.48279*X 1 + 2.74293*X 2 + 6.45267*X 3 – 0.0267551*X 1 2 + 0.0104072*X 1 *X 2 – 0.00475208*X 1 *X 3 – 0.0497703*X 2 2 – 0.144404*X 2 * X3 – 0.114537*X 3 2 Al COD removal. % Y1 = 158.716 –3.52182*X 1 – 1.0666*X 2 – 4.4196*X 3 + 0.0570054*X 1 2 – 0.00183056*X 1 *X 2 – 0.0100464*X 1 *X 3 + 0.0285149*X 2 2 + 0.0180835*X 2 * X3 + 0.256125*X 3 2 Copper removal (%) Y2 = 73.1876 + 0.501176*X 1 – 0.555376*X 2 + 5.00747*X 3 – 0.0178196*X 1 2 – 0.00322581*X 1 *X 2 + 0.128696*X 1 *X 3 + 0.00518817*X 2 2 + 0.101075*X 2 * X3 – 0.82945*X 3 2 Nickel removal (%) Y3 = 87.7568 + 1.4848*X 1 – 0.656174*X 2 – 5.22143*X 3 – 0.0334283* X1 2 + 0.00342153* X1 *X 2 + 0.127832* X1 *X 3 + 0.000874392* X2 2 + 0.11253* X2 * X3 – 0.160094* X3 2
Table 4. Experimental and predicted response of removal effi ciency in the EC process. Fe electrode Al electrode X1 :j X2 :t X3 : pH Y1 : COD removal (%) Y2 : Copper removal (%) Y3 : Nickel Removal (%) Y1 : COD removal (%) Y2 : Copper removal (%) Y3 : Nickel Removal (%) Run mA/cm 2 min -Exp. Pred. Exp. Pred. Exp. Pred. Exp. Pred. Exp. Pred. Exp. Pred. 1 16 10 7 90.38 90.20 76.25 77.50 81.56 81.10 90.94 90.60 89.74 86.99 76.27 74.82 2 48 10 7 92.55 91.93 84.25 85.12 75.27 76.02 92.87 91.81 94.58 94.33 83.57 83.61 3 16 30 7 94.18 94.80 93.35 92.48 80.01 79.26 92.97 94.03 92.90 93.15 79.28 79.25 4 48 30 7 89.49 89.67 82.12 80.87 80.38 80.84 93.73 94.07 95.67 98.42 88.77 90.22 5 16 20 4 90.22 90.88 89.41 89.95 75.71 78.47 93.35 93.67 90.58 89.66 86.40 85.01 6 48 20 4 86.18 87.28 83.93 84.85 75.63 77.18 94.22 95.26 87.03 83.61 85.49 82.61 7 16 20 10 87.73 86.63 93.61 92.69 91.33 89.78 90.90 89.87 71.09 74.51 63.13 66.01 8 48 20 10 87.49 86.82 94.35 93.81 90.34 87.58 89.85 89.53 92.25 93.17 86.77 88.16 9 32 10 4 87.24 86.76 86.06 84.27 76.92 74.62 81.82 81.84 88.51 92.18 90.23 93.07 10 32 30 4 92.09 90.81 93.16 93.49 86.78 84.77 84.97 83.60 90.58 91.25 90.41 91.84 11 32 10 10 86.01 87.29 94.32 93.98 92.13 94.14 74.63 76.00 84.0 83.33 81.02 79.59 12 32 30 10 85.09 85.58 93.70 95.49 84.66 86.96 79.94 79.92 98.19 94.52 94.70 91.86 13 32 20 7 86.38 86.39 91.74 91.74 91.14 91.13 75.03 75.18 97.41 97.26 90.69 90.45 14 32 20 7 86.39 86.39 91.73 91.74 91.11 91.13 75.36 75.18 97.09 97.26 90.23 90.45 15 32 20 7 86.41 86.39 91.74 91.74 91.10 91.13 75.16 75.18 97.29 97.26 90.41 90.45
Table 5.The results of ANOVA for response surface of the quadratic model for the used electrodes.
Electrodes Model R2 Adj.R2 Sum of squares Mean square F-value p-value (Prob.>F)
Fe COD 92.77 79.76 108.771 11.21 7.13 0.02175 Copper 96.81 91.09 424.249 45.63 16.90 0.00310 Nickel 93.50 81.82 619.653 64.38 8.00 0.01695 Al COD 99.01 97.24 874.958 96.26 55.90 0.00017 Copper 89.94 71.84 678.326 67.79 4.96 0.04610 Nickel 94.80 85.45 863.955 91.00 10.13 0.01004
Table 6.The results of ANOVA for Fe and Al electrodes.
Sum of squares Mean Square F-Ratio P-Value
Source Df Fe Al Fe Al Fe Al Fe Al COD X1 1 5.79701 0.78560 5.79701 0.78560 3.69 0.46 0.1129 0.5293 X2 1 2.75069 16.1316 2.75069 16.1316 1.75 9.37 0.2432 0.0281 X3 1 11.0784 45.3529 11.0784 45.3529 7.05 26.34 0.0452 0.0037 X1X1 1 28.487 786.339 28.487 786.339 18.12 456.72 0.0080 0.0000 X1X2 1 11.7735 0.34313 11.7735 0.34313 7.49 0.20 0.0410 0.6740 X1X3 1 3.59084 0.93016 3.59084 0.93016 2.28 0.54 0.1911 0.4953 X2X2 1 22.7179 30.0222 22.7179 30.0222 14.45 17.44 0.0126 0.0087 X2X3 1 8.32236 1.17724 8.32236 1.17724 5.29 0.68 0.0697 0.4459 X3X3 1 5.88912 19.6195 5.88,912 19.6195 3.75 11.40 0.1107 0.0198 Total error 5 7.86,231 8.60855 1.57246 1.72171 Total (corr.) 14 108.771 874.958 Copper X1 1 7.93573 79.5427 7.93573 79.5427 2.94 5.83 0.1471 0.0605 X2 1 57.5209 52.6139 57.5209 52.6139 21.31 3.86 0.0058 0.1067 X3 1 68.558 15.5718 68.558 15.5718 25.40 1.14 0.0040 0.3342 X1X1 1 78.5625 76.8379 78.5625 76.8379 29.10 5.63 0.0030 0.0637 X1X2 1 92.4078 1.06556 92.4078 1.06556 34.23 0.08 0.0021 0.7911 X1X3 1 9.69015 152.642 9.69015 152.642 3.59 11.19 0.1167 0.0204 X2X2 1 36.1469 0.99386 36.1469 0.99386 13.39 0.07 0.0146 0.7980 X2X3 1 14.8599 36.7784 14.8599 36.7784 5.50 2.70 0.0659 0.1615 X3X3 1 37.8515 205.761 37.8515 205.761 14.02 15.08 0.0134 0.0116 Total error 5 13.4981 68.2015 2.69963 13.6403 Total (corr.) 14 424.249 678.326 Nickel X1 1 6.11835 195.103 6.11835 195.103 0.76 21.73 0.4231 0.0055 X2 1 4.41832 60.9424 4.41832 60.9424 0.55 6.79 0.4920 0.0479 X3 1 235.62 90.5593 235.62 90.5593 29.29 10.09 0.0029 0.0247 X1X1 1 173.217 270.399 173.217 270.399 21.53 30.12 0.0056 0.0027 X1X2 1 11.0909 1.19879 11.0909 1.19879 1.38 0.13 0.2932 0.7298 X1X3 1 0.20811 150.599 0.20811 150.599 0.03 16.77 0.8785 0.0094 X2X2 1 91.4615 0.02822 91.4615 0.02822 11.37 0.00 0.0199 0.9575 X2X3 1 75.0692 45.5871 75.0692 45.5871 9.33 5.08 0.0283 0.0740 X3X3 1 3.92348 7.66536 3.92348 7.66536 0.49 0.85 0.5161 0.3979 Total error 5 40.2241 44.8913 40.2241 8.97826 Total (corr.) 14 619.653 863.955
Table 7.Optimum operating values for Fe and Al electrodes.
COD Copper Nickel
Fe Al Fe Al Fe Al
Goal of model Maximize
Current density (mA/cm2) 16 48 32 40 30 38
Reaction time (min) 30 30 20 30 16 30
Initial pH 4 4 10 8 10 10
Model prediction results (%) 97.10 98.69 97.91 100 96.28 94.56
time (X1X2) have also significant effects on copper removal at the iron electrode. In terms
of copper removal, pH (X3) and the coefficients having two factors of X1X3and X2X3have
effects on EC process (p > 0.05), whereas the current density (X1) and the quadratic
terms of current density (X1X1) have highly significant effects on copper removal
(p < 0.003). For nickel removal, initial pH (X3), the quadratic terms of current density
(X1X1), reaction time (X2X2) and initial pH (X3X3) have significant effects on nickel
removal.
Table 6 shows ANOVA results of the response surface quadratic model of the COD,
copper and nickel removals at Al electrode. As can be seen from Table 6, linear
coefficients of the current density (X1) and pH (X3) on nickel removal efficiency, linear
coefficients of the current density (X1) and electrolysis time (X2) on copper removal
efficiency, and electrolysis time (X2) and pH (X3) on COD removal efficiency have
significant effects for aluminum electrode. The second order terms of the current density
(X1X1) have also significant effects on COD, copper and nickel removal at aluminum
electrode. In terms of copper removal, pH (X3) and the coefficients having two factors of
X1X3and X3X3have effects on EC process (p < 0.05). For nickel removal, current density
(X1) electrolysis time (X2) and initial pH (X3), the quadratic terms of current density (X1X1),
the second order terms of the current density and pH (X1X3) have significant effects on
nickel removal at the aluminum electrode.
Three-dimensional response surface graphics that belong to the experimental study
are shown inFigure 3for iron electrode and inFigure 4for the aluminum electrode. The
experimental data were prepared in colors three dimensionally with current density against time, current density against pH and time against pH graphics. Generally speak-ing, during the experimental study, it is seen that current density, electrolysis time and
pH variables affected removal efficiency at the central level. It is impossible to claim that
no variable is influential. In the COD removal study conducted with the iron electrode, as
seen inFigure 3.a1, removal efficiency is quite high in 16 mA/cm current density. In case
of the current increase, about a 7% decrease was observed in removal efficiency. The
response and contour plots of the quadratic model for copper and nickel removal are
given in Figure 3. a4–c9 for the iron electrode and Figure 4. a4–c9 for the aluminum
electrode. In copper removal, it seems that the link between pH increase and time did
not affect removal efficiency that much. It is seen that current density had
a considerable effect on removal. As seen inFigure 3.b5-6and 3.b8-9, the biggest removal
was obtained in copper and nickel removal in pH 10.
The ANOVA result of experimental optimum data for iron and aluminum electrode
show inTable 7. The most suitable condition values for COD removal in iron electrode
are 16 mA/cm2for current density, 30 minutes for electrolysis time and 4 for pH. In line
with these conditions, the estimated COD removal efficiency of the model was
deter-mined as 97.10%. Experimental COD removal efficiency was found as 95.72% as a result
of the experiment performed at the optimum conditions. This finding shows that the
model prediction for COD removal was within the confidence interval. In optimum
conditions, the COD value was calculated as 1384 mg/L after treatment of CNC machine wastewater. This value is considerably high in terms of discharge limits. According to the results, the wastewater generated during metalworking on the CNCmachine does not meet the direct discharge limit values determined by Water Pollution Control Regulation of Turkey (COD < 100 mg/L) (Ministry of Environment and Urbanization). However, this
experimental research was obtained a high COD removal efficiency and this wastewater was became less harmful. It was found that the optimum conditions for copper removal
were 32 mA/cm2for current density, 20 minutes for operation time and 10 for pH by
using the iron electrode. Experimental removal efficiency and model prediction removal
efficiency in copper removal were found as 96.03% and 97.91% respectively. The
optimum conditions for nickel removal were 30 mA/cm2for current density, 16 minutes
for time and 10 for pH by the aid of iron electrode. The experimental removal efficiency
(a1) (b2) (c3)
(a4) (b5) (c
6)
(a7) (b
8) (c9)
Estimated Response Surface pH=7,0 16 32 48 Current Density 10 20 30 Time 82 84 86 88 90 92 94 96 98 CO D r e m o va l COD removal 82,0-84,0 84,0-86,0 86,0-88,0 88,0-90,0 90,0-92,0 92,0-94,0 94,0-96,0 96,0-98,0
Estimated Response Surface Time=20,0 16 32 48 Current Density 4 7 10 pH 82 84 86 88 90 92 94 96 98 CO D r e m o va l COD removal 82,0-84,0 84,0-86,0 86,0-88,0 88,0-90,0 90,0-92,0 92,0-94,0 94,0-96,0 96,0-98,0
Estimated Response Surface Current Density=32,0 10 20 30 Time 4 7 10 pH 82 84 86 88 90 92 94 96 98 CO D r e m o v a l COD removal 82,0-84,0 84,0-86,0 86,0-88,0 88,0-90,0 90,0-92,0 92,0-94,0 94,0-96,0 96,0-98,0
Estimated Response Surface pH=7,0 16 32 48 Current Density 10 20 30 Time 75 80 85 90 95 100 C u p p er r em o val Cupper removal 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Time=20,0 16 32 48 Current Density 4 7 10 pH 75 80 85 90 95 100 C u p p er r em o val Cupper removal 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Current Density=32,0 10 20 30 Time 4 7 10 pH 75 80 85 90 95 100 Cu p p e r r e m o v a l Cupper removal 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface pH=7,0 16 32 48 Current Density 10 20 30 Time 65 70 75 80 85 90 95 100 N ick e l r em o va l Nickel removal 65,0-70,0 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Time=20,0 16 32 48 Current Density 4 7 10 pH 65 70 75 80 85 90 95 100 N ick el r e m o v al Nickel removal 65,0-70,0 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Current Density=32,0 10 20 30 Time 4 7 10 pH 65 70 75 80 85 90 95 100 Ni c k e l r e m o v a l Nickel removal 65,0-70,0 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Figure 3.Three-dimensional graphs for COD, copper, and nickel at Fe electrode (a) current density vs. electrolysis time, (b) current density vs. pH, (c) electrolysis time vs. pH.
and model prediction removal efficiency obtained in the nickel removal study were 95.22% and 96.28%, respectively. It was observed that experimental values were con-sistent with predicted values for copper and nickel removal. The optimum conditions of COD in experiments with aluminum electrode were found to be 30 min for reaction
time, 48 mA/cm2 for the current density and 4 for pH. The COD removal efficiency of
optimum model prediction and laboratory experiment were obtained 98.69% and 97.11%, respectively. The COD concentration of the wastewater of the treated CNC machine was 934 mg/L at optimum operating conditions at the aluminum electrode.
(a1) (b2) (c3)
(a4) (b5) (c6)
(a7) (b8) (c9)
Estimated Response Surface pH=7,0 16 32 48 Current Density 10 20 30 Time 75 80 85 90 95 100 CO D r e m o v a l COD removal 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Time=20,0 16 32 48 Current Density 4 7 10 pH 75 80 85 90 95 100 CO D r e m o v a l COD removal 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Current Density=32,0 10 20 30 Time 4 7 10 pH 75 80 85 90 95 100 CO D r e m o v a l COD removal 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface pH=7,0 16 32 48 Current Density 10 20 30 Time 70 75 80 85 90 95 100 Cu p p e r r e m o v a l Cupper removal 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Time=20,0 16 32 48 Current Density 4 7 10 pH 70 75 80 85 90 95 100 Cu p p e r r e m o v a l Cupper removal 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Current Density=32,0 10 20 30 Time 4 7 10 pH 70 75 80 85 90 95 100 C u p p er r em o val Cupper removal 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface pH=7,0 16 32 48 Current Density 10 20 30 Time 60 65 70 75 80 85 90 95 100 N ick el r em o val Nickel removal 60,0-65,0 65,0-70,0 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Time=20,0 16 32 48 Current Density 4 7 10 pH 60 65 70 75 80 85 90 95 100 Ni c k e l r e m o v a l Nickel removal 60,0-65,0 65,0-70,0 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Estimated Response Surface Current Density=32,0 10 20 30 Time 4 7 10 pH 60 65 70 75 80 85 90 95 100 N ick el r em o val Nickel removal 60,0-65,0 65,0-70,0 70,0-75,0 75,0-80,0 80,0-85,0 85,0-90,0 90,0-95,0 95,0-100,0
Figure 4.Three-dimensional response surface graphs for COD, copper, and nickel at Al electrode (a) current density vs. electrolysis time, (b) current density vs. pH, (c) electrolysis time vs. pH.
Aluminum electrode provided more removal efficiency than the iron electrode. Unfortunately, The COD concentration value is much higher than the discharge limits.
It was found that optimum conditions for copper removal were 30 min for time,
40 mA/cm2for the current density and 8 for pH by the aid of aluminum electrode. The
optimum removal efficiencies of copper were calculated as 100% and 98.51%
respec-tively as a result of the model prediction and laboratory tests. Optimum conditions for
nickel removal were determined as 30 min for electrolysis time, 38 mA/cm2 for the
current density, and 10 for pH by using the aluminum electrode. In optimum conditions
of nickel, the estimated removal efficiency of the model and the efficiency of the
experimental study were found as 94.56% and 92.49%, respectively. In Turkey’s Water
Pollution Control Regulation, there are values pertaining to the discharge limits of metal preparation and processing industry waste. As a result of the study, 96.03% and 98.51% copper removal were obtained for iron and aluminum electrodes, respectively. The copper concentration after the treatment was calculated as 3.07 and 1.15 mg/L. The discharge limit in the regulation is provided for the copper parameter. However, nickel concentration was calculated as 2.61 mg/L for iron electrode and 4.11 mg/L for the aluminum electrode. Since the nickel discharge limit is 3 mg/L, the remaining nickel concentration does not provide the discharge limit. It was observed that the experi-mental values of the study performed to remove all pollutants from the wastewater were consistent with the predicted values. A very low margin of error in the model prediction value showed a good correlation between experimental study data and model results. In the experimental study for COD removal, which was performed
under optimum conditions, the removal efficiency was achieved, over 90% for all
pollutants. On iron and aluminum electrodes, COD, copper and nickel removal e
fficien-cies were obtained as 95.72%, 96.03%, 95.22% and 97.11%, 98.51%, 92.49%, respectively.
When the optimum application results inTable 7are examined, it can be said that as
the current density increases, the suitability and treatment efficiency increase, because
only the iron electrode with the COD removal of the current length is found to be
16 mA/cm2. When the reaction times are examined; In the removal of COD with iron and
aluminum electrode, and also used in aluminum electrode, copper and nickel removal were also determined as the most suitable contact time for 30 minutes. The pH value was found to be 4 for COD removal during both electrodes. In the removal of copper
and nickel, the pH of the medium increased the efficiency in more alkaline conditions.
Nickel was found to have a pH of 10 in both electrodes.
3.3. Economic analysis
In addition to labor, disposal, sludge treatment and maintenance, and fixed costs,
equipment, mainly electrodes and electricity consumption costs constituted the
operat-ing cost (OC) of the EC process [41,42]. The cost of the treatment (OC) in the appropriate
conditions is calculated by [39] Equation (5);
OC cost
m3
¼ aENC þ bELC (5)
where an aENC is the electrical energy consumption (kWh/m3) and bELC is the electrode
ENC¼U:i:tEC
V (6)
where i is current (A), tEC is the operating time (min), U is cell voltage (V). Electrode
consumption (ELC) in an electrocoagulation process can be calculated using Faraday’s
law (Equation (7)) [38,43,44], in which i is the current (A), M is the atomic mass of the
electrode (MFe= 55.84 g mol−1, MAl= 26.98 g mol−1), t is the running time (s), z is the
valence (zFe= 2, zAl= 3), F is Faraday’s constant (96.485 C mol−1), and v is the volume of
the treated wastewater (m3).
ELC¼M:i:t
z:F:v (7)
Faraday’s laws give an ideal estimate of the electro dissolution in an ideal case. In a real
application, many factors affect the amount and quality of electro dissolution [45]. The
operational cost of the EC process, which was applied to metalworking process
waste-water for COD, copper, and nickel removal, wasfound to be 2.54, 3.36, 2.50€/m3for iron
electrode and 7.16, 8.95, 8.50 €/m3 for aluminum electrode at optimum conditions,
respectively.
3.4. Comparison of optimized experiment results with previous literature studies Table 8shows thefindings of previous COD removal studies and some heavy metal irons
from various wastewaters by the EC method. Briefly, pollutants can be removed from
wastewater effectively by the EC method. For the treatment of various real or synthetic
wastewater such as wastewater of motor vehicle factories [3], metalworkingfluid
waste-water [7], metal plating wastewater [18], carwash wastewater [27], lithographic
waste-water [32], rice mill wastewater [36], color wastewater [40], and oily wastewater [46],
electrocoagulation was successfully applied.
High-efficiency results of electrocoagulation process wereproved as a result of the
various research similar to this study for the removal of heavy metal ions from waste-water. Operating time, pH, direct current density and electrical conductivity on the
pollutants removal have vital importance for the success level of the EC process [34].
Table 8 shows the evaluation of the electrode, the removal efficiencies and waste-water types of this research under the optimum conditions. According to the results of this study, it can be said that the removal of heavy metal (copper and nickel) and COD
arepossible and the EC method is a cost-effective treatment. In the light of results of the
current and previous studies, it can be said that EC is an appropriate, easy and
cost-effective method for the removal of pollutants from wastewater of CNC machines.
4. Conclusions
The present study investigated the removal of COD and heavy metal ions from wastewater ofCNC industry wastewater, by a batch electrocoagulation process with Fe and Al
electro-des. Three response parameters applied to as COD, copper, and nickel removal efficiency
was determined. The influence of electrode type, electrolysis time, wastewater pH, current
density, and conductivity on the removal of pollutants was investigated in this study. In
Table 8. Comparing the results of some previous studies with the results of this study. Operating parameters Electrode Water types Responses C0 (mg/L) Conductivity (mS/cm) j (mA/cm 2 )p Hi t (min) Re (%) References Al & Fe real COD, TOC 17312, 3155 6.1 80 5– 7 25 93, 80 [ 7 ] Fe real COD, color, total Cr, Ni, Zn 475.8, 5983, 358, 8.1, 149.25 17.14 30 5 30 76.2, 99.8, 98.5, 96, 99.7 [ 18 ] Fe & Al real COD. oil-grease, chloride 560, 125, 150 0.98 3 8 30 88, 90, 50 [ 27 ] Fe real TOC 18400 16.08 125 8.23 30 >58 [ 32 ] Steel real Cr, Ni, Zn, Cu 93.2,57.6, 20.4, 33.3 8.9 9.56 4 45 97 [ 34 ] Steel real COD, TSS 2200.768 -15 7 -89 [ 36 ] Fe synthetic Color, COD 1750, 2080 12.37 20 –30 7.2 15 89.2, 76.1 [ 40 ] Al real oil and grease 35 6.19 20 –80 3.6 –8.7 20 98 [ 46 ] Fe & Al real Cu, Cr, Ni 45, 44.5, 394 2.0 10 3 20 100 [ 47 ] Fe & Al real COD*, Cu** and Ni** 32350, 77.5, 23.1, 54.8 >10.0 16* and 32** 4* and 10** 30 >90 This Study C0 : initial concentration. pH i : initial pH. . j:current density. t: reaction time. Re : removal effi ciency
interaction between the parameters in the EC process, optimization with RSM (with the help of central composite design) is applied. ANOVA results show that the high correlation
coefficient (R2) and adjusted R2value verify a satisfactory adjustment of the second-order
regression model with the experimental data. The quality offitness (Table 5) are as follows:
p–value of Models < 0.05, the p–value of Lack of Fit (f-value)>0.05, the difference between
R2pred and R2adj are less than 20% . The proposed model complied with the experimental
data very well. A positive effect was identified on the removal efficiency as a result of
increasing the operation time for the pollutant removal in the EC reactor by using Fe and Al electrodes. As a result of the experiments, the following results were obtained;
● The removal efficiencies of COD, copper and nickel were determined as 95.72%, 96.03%
and 95.22% at optimum operating conditions for all responses at the iron electrode,
respectively. The COD, copper and nickel removal efficiencies were found as 97.11%,
98.51% and 92.49% at optimum operating conditions for all responses at the aluminum electrode, respectively.
● The operational cost of the EC process, which was applied to CNC industry
waste-water for COD, copper, and nickel removal, was found to be 2.54, 3.36, 2.50€/m3
for iron electrode and 7.16, 8.95, 8.50 €/m3 for aluminum electrode at optimum
conditions, respectively.
● Fe-Fe electrode was a successful application of CNC industry in the treatment of
wastewater Al-Al electrode application of nickel removal found to be more success-ful, but in the removal of COD and copper Al-Al was found to be successful. For copper and nickel removals, the highest rate was found at pH 10.
● It is important to study both thefirst usage of wastewater from the Konya industry
and the implementation of the RSM model. It is also valuable that the data obtained have been acquired as a result of the treatment of real wastewater. These data will shed light on the treatment of CNC industry wastewaters and help design industrial type treatment plants.
Authors’ contributions
MKO is the only author of the manuscript. Electrocoagulation studies, all optimization, model application and characterization analysis, etc. mentioned in the article were performed by the author. The author read and approved thefinal manuscript.
Disclosure statement
No potential conflict of interest was reported by the author.
ORCID
References
[1] S. Anderberg, Ph. D. thesis, Chalmers Universıty of Technology, (2012). ISBN 978-91-7385-678-2.
[2] Ministry of Economy, Export General Directorate, Automotive, Machinery, 3 vols. (Electrical and Electronic Products Head of Department, 2018). https://trade.gov.tr/data/5b8fd58313b 8761f041fee92/02da75c2b78fb8ff53c22420922323ce.pdf
[3] S. Ahmad Mirbagheri and M. Salehi Moayed, Iran. J. Environ. Health. Sci. Eng. 3, 289 (2006). [4] C. Barrera-Díaz, B. Frontana-Uribe and B. Bilyeu, Chemosphere 105, 160 (2014). doi:10.1016/
j.chemosphere.2014.01.026.
[5] G. Rotella, P. Priarone, S. Rizzuti and L. Settineri, Glocalized Solutions for Sustainability in Manufacturing, 1 vols. (Springer, Berlin,2011, p. 365). doi:10.1007/978-3-642-19692-8_63 [6] C. Cheng, D. Phipps and R.M. Alkhaddar, Water Environ. J. 20, 227 (2006). doi:
10.1111/j.1747-6593.2005.00010.x.
[7] E. Demirbas and M. Kobya, Process Saf. Environ. Prot. 105, 79 (2017). doi:10.1016/j. psep.2016.10.013.
[8] F. Ozyonar, Int. J. Electrochem. Sci 11, 1456 (2016).
[9] Bacak, S, Karabıyık, Ö., J Tech. Sci, 9, 26 (2019).https://dergipark.org.tr/tbed/issue/43343/ 514035
[10] A. Zein, W. Li, C. Herrmann and S. Kara, Glocalized Solutions for Sustainability in Manufacturing, 1 vols. (Springer, Berlin,2011), p. 274. doi:10.1007/978-3-642-19692-8_48 [11] T. Asano, Water Sci. Technol. 45, 24 (2002). doi:10.2166/wst.2002.0137.
[12] F. Ansari, Y.K. Pandey, P. Kumar and P. Pandey, Int. J. Energy Env. 4, 1079 (2013).
[13] F. El-Gohary, R. Abdel Wahaab, F. Nasr and H.I. Ali, Presented at 2nd Specialized Conference on Pretreatment of Industrial Wastewaters, Athens, Greece, 16, 148,1996.
[14] C. Özdemir, M.K. Öden, S. Şahinkaya and D. Güçlü, Color. Technol. 127, 268 (2011). doi:10.1111/j.1478-4408.2011.00310.x.
[15] C. Özdemir, M.K. Öden, S. Şahinkaya and E. Kalipçi, Clean–Soil, Air, Water 39, 60 (2011). doi:10.1002/clen.201000263.
[16] Y. Zhao and W. Qiu, Fuel Process. Technol. 167, 355 (2017). doi:10.1016/j.fuproc.2017.07.021. [17] O.A. Alsager, M.N. Alnajrani, H.A. Abuelizz and I.A. Aldaghmani, Ecotoxicol. Environ. Saf. 158,
114 (2018). doi:10.1016/j.ecoenv.2018.04.024.
[18] M.K. Oden and H. Sari-Erkan, Process Saf. Environ. Prot. 119, 207 (2018). doi:10.1016/j. psep.2018.08.001.
[19] H. Singh and B.K. Mishra, Environ. Eng. Res. 22, 141 (2017). doi:10.4491/eer.2016.029. [20] N. Li, G.P. Sheng, Y.Z. Lu, R.J. Zeng and H.Q. Yu, Water Res. 111, 204 (2017). doi:10.1016/j.
watres.2017.01.010.
[21] F. Qian, M. He, J. Wu, H. Yu and L. Duan, J. Environ. Sci. 76, 329 (2018). doi:10.1016/j. jes.2018.05.025.
[22] M.K. Oden and S. Kucukcongar, J. Global Nest. 20, 234 (2018). doi:10.30955/gnj.002500 [23] P. Finkbeiner, J. Redman, V. Patriarca, G. Moore, B. Jefferson and P. Jarvis, Water Res. 146,
256 (2018). doi:10.1016/j.watres.2018.09.042.
[24] T. Li, W. Zhang, S. Zhai, G. Gao, J. Ding, W. Zhang, Y. Liu, X. Zhao, B. Pan and L. Lv, Water Res. 143, 87 (2018). doi:10.1016/j.watres.2018.06.031.
[25] A. Sarı and M. Tuzen, J. Hazard. Mater 164, 1372 (2009). doi:10.1016/j.jhazmat.2008.09.047. [26] F. Özyonar and B. Karagözoğlu, J. Fac. Eng. Archit. Gazi. Univ. 27, 81 (2012).
[27] Z.B. Gönder, G. Balcıoğlu, I. Vergili and Y. Kaya, J. Environ. Manage. 200, 380 (2017). doi:10.1016/j.jenvman.2017.06.005.
[28] A.M.H. Elnenay, E. Nassef, G.F. Malash and M.H.A. Magid, Egypt. J. Petr. 26, 203 (2017). doi:10.1016/j.ejpe.2016.03.005.
[29] G. Chen, Sep. Purif. Technol. 38, 11 (2004). doi:10.1016/j.seppur.2003.10.006.
[30] F. Ozyonar and B. Karagozoglu, Sep. Purif. Technol. 150, 268 (2015). doi:10.1016/j. seppur.2015.07.011.
[32] A. Suárez-Escobar, A. Pataquiva-Mateus and A. López-Vasquez, Catal. Today 266, 120 (2016). doi:10.1016/j.cattod.2015.09.016.
[33] M.Y.A. Mollah, R. Schennach, J.R. Parga and D.L. Cocke, J. Hazard. Mater. 84, 29 (2001). doi:10.1016/S0304-3894(01)00176-5.
[34] M. Al-Shannag, Z. Al-Qodah, K. Bani-Melhem, M.R. Qtaishat and M. Alkasrawi, Chem. Eng. J. 260, 749 (2015). doi:10.1016/j.cej.2014.09.035.
[35] Federation, W. E., and APHA, Standard Methods for the Examination of Water and Wastewater, 21sted. (Washington, DC, USA,2005).
[36] T. Karichappan, S. Venkatachalam, J.P. Maran and K. Sengodan, J. Korean Chem. Soc. 57, 761 (2013). doi:10.5012/jkcs.2013.57.6.761.
[37] U.T. Un, A. Kandemir, N. Erginel and S.E. Ocal, J. Environ. Manage. 146, 245 (2014). doi:10.1016/j.jenvman.2014.08.006.
[38] S. Ahmadzadeh, A. Asadipour, M. Pournamdari, B. Behnam, H.R. Rahimi and M. Dolatabadi, Process Saf. Environ. Prot. 109, 538 (2017). doi:10.1016/j.psep.2017.04.026.
[39] E. Sık, M. Kobya, E. Demirbas, E. Gengec and M.S. Oncel, J. Environ. Chem. Eng. 5, 3792 (2017). doi:10.1016/j.jece.2017.07.004.
[40] K. Hendaoui, F. Ayari, I.B. Rayana, R.B. Amar, F. Darragi and M. Trabelsi-Ayadi, Process Saf. Environ. Prot. 116, 578 (2018). doi:10.1016/j.psep.2018.03.007.
[41] E. Gengec, M. Kobya, E. Demirbas, A. Akyol and K. Oktor, Desalination 286, 200 (2012). doi:10.1016/j.desal.2011.11.023.
[42] Y. Shi, H. Liu, X. Zhou, A. Xie and C. Hu, Chin. Sci. Bull. 54, 2124 (2009). doi: 10.1007/s11434-009-0177-4.
[43] X.Y. Zheng, H.N. Kong, D.Y. Wu, C. Wang, Y. Li and H.R. Ye, Water Sci. Technol. 60, 2929 (2009). doi:10.2166/wst.2009.309.
[44] E. Nariyan, M. Sillanpää and C. Wolkersdorfer, Sep. Purif. Technol. 193, 386 (2018). doi:10.1016/j.seppur.2017.10.020.
[45] İ.A. Şengil, M. Özacar, J. Hazard. Mater. 137, 1197 (2006). doi:10.1016/j.jhazmat.2006.04.009. [46] M. Changmai, M. Pasawan and M.K. Purkait, Sep. Purif. Technol. 210, 463 (2019).
doi:10.1016/j.seppur.2018.08.007.
