Heterogeneous electro fenton process for textile wastewater treatment: application of fe3o4-Mn3o4 as a catalyst - Nguyen Duc Dat Duc

Vietnam Journal of Science and Technology 56 (2C) (2018) 194-200 HETEROGENEOUS ELECTRO FENTON PROCESS FOR TEXTILE WASTEWATER TREATMENT: APPLICATION OF Fe3O4-Mn3O4 AS A CATALYST Nguyen Duc Dat Duc1, 2, Nguyen Thi Chi Nhan1, Nguyen Hoang Ha2, Nguyen Tan Phong1, * 1Ho Chi Minh City University of Technology, VNU-HCM, 268 Ly Thuong Kiet, HCMC City 2Ho Chi Minh City University of Food Industry, 140 Le Trong Tan, Tan Phu District, HCMC *Email: ntphong@hcmut.edu.vn Received: 18 May 2018; Accepted for publication: 23 August 2018 ABSTRACT To produce quality textile products, many stable synthetic dyes are invented. However, textile wastewater discharged into the environment contains a large amount of dyes which are bio-recalcitrants and stable to heat and solar radiation. These toxic contaminants impact negatively the environment for a long time and poison plants, animals, and human. In this study, Heterogeneous Electro Fenton process with Fe3O4-Mn3O4 as a catalyst was used for textile wastewater treatment. Graphite electrodes were chosen for their stability and good conductivity. Three operating factors greatly affected the removal efficiency are current density, pH and catalyst dosage. Response surface methodology was applied to find empirical mathematical models for these factors, then find an accordant operating condition for treatment which provides high removal efficiency and low operating cost. Investigated ranges scaled in previous studies were: current density (10 – 20 mA/cm2), pH (3 – 5), and catalyst dosage (0.5 – 1.5 g/l). At current density of 17.03 mA/cm2, pH of 3.77, catalyst dosage of 1.17 g/l, the treatment reached its optimum condition, COD and Color removal efficiencies were 93.3 % and 99.2 %, respectively, where COD and Color values in the effluent are 56 mg/l and 16 Pt-Co. Keywords: Heterogeneous Electro Fenton; textile wastewater; Fe3O4-Mn3O4; Box Behnken. 1. INTRODUCTION Main pollution in textile wastewater often originate from dyeing and finishing processes. Textile wastewater contains high non-biodegradable organic loading and toxic compounds, the structures of dye molecules are generally complex and stable to heat and light. Furthermore, synthetic dyes usually contain aromatic rings which are harmful and transformable into toxic compounds could be found in chlorination process of wastewater treatment plan. Even if dye in the water resource remain at low concentration, the Turbidity and Color will be increased. So these wastewater sources are hard to discharge to the environment legally [1]. Heterogeneous Electro Fenton process for textile wastewater treatment: Application of 195 Biological treatment, a popular stage in wastewater treatment plan, can’t be used in this situation, when 53 % of 87 colors are stable with biological agent [2]. Other treatment processes such as adsorption, coagulation and filtration are very expensive and a large amount of sludge will occur [3]. Among advanced processes, Electro Fenton process is based on the continuous supply of H2O2. Fe(OH)2+ was reduced at the cathode, and Fe2+ ion will be created. This ion will work as catalyst for the Fenton reaction. Due to this process, a high amount of sludge is discharged and many chemicals must be used. Furthermore, sedimentation velocity of sludge is not high so the volume of sedimentation tank will increase. This disadvantage can be solved by Heterogeneous Electro Fenton (HEF) process with Fe3O4-Mn3O4 as a catalyst, in this process, Fe3O4-Mn3O4 was quickly separated by a magnet and recycled for the treatment and little chemicals was used. The Mn3O4 has redox cycle between Mn(II)/Mn(III) = 15.1 V, which could be used as a good catalyst [4]. Many evidences showed that Manganese ion play a similar role with Fe ion in Fenton-like processes [4]. Furthermore, Fe3O4-Mn3O4 was used in heterogeneous Fenton processes to treat some dyes and antibiotics which have total removal efficiency of over 95 % [5, 6]. Catalytic mechanism of Fe3O4-Mn3O4 in Fenton-like process was developed as Haber Weiss reaction [4]: Mn2+surf + H2O2
Mn3+ + OH• + OH- (1) Mn3+surf + H2O2
Mn2+surf + HOO• + H+ (2) OH• radical can attack an organic substrate such as methylene blue: RH + OH•
R• + H2O (3) Many previous reports showed that, pH, catalyst loading, current density (J) are significantly influent to the HEF process [7]. These parameters were investigated simultaneously by Response surface methodology. To lower the investment cost, carbonaceous materials such as graphite electrodes will be used instead of Platinum or Iridium. To the best of our knowledge, the application of Fe3O4-Mn3O4 in HEF process for real textile wastewater treatment has not been reported. This study was carried out to determine the optimum operating condition of three parameters, including catalyst dosage, pH, and current density to remove Color and COD from a real textile wastewater. 2. MATERIALS AND METHODS 2.1. Wastewater characteristic Wastewater was collected from the equalization tank of THANHCONG Textile Garment Investment Trading Joint Stock Company wastewater treatment plan. Then, coagulation with FeCl3 1 % w/w was performed to remove suspended solid [8]. The characteristics of the sample after pre-treatment are shown in Table 1. 2.2. Materials and Apparatus Fe3O4-Mn3O4 was synthesized by co-precipitation method [5] and characterized for solid identification by X-ray diffraction (XRD) analysis. K2CrO7 (Merck, Germany), FeSO4.7H2O (Merck, Germany), and Fe(NH4)2(SO4)2.6H2O (Merck, Germany) were used for COD analysis. Color of the sample was determined by 2120C, Spectrophotometric Method (EPA, 1999). Nguyen Duc Dat Duc, Nguyen Thi Chi Nhan, Nguyen Hoang Ha, Nguyen Tan Phong 196 pH meter (WTW Inolab 7110, Germany) was used to measure pH. Graphite plates with 40 cm2 areas was used as electrodes. Table 1. Textile wastewater characteristics. Figure 1. Schema of the reactor. Parameter unit Value (n=3) pH - 9.1 ± 1.2 TSS mg/l 45 ± 9.6 COD mgO2/l 673 ± 32 color Pt-Co 1025 ± 28 2.3. HEF Model operation A 500 ml glass beaker was used for the reaction. Cathode and anode were connected to the power supply (MATRIX MPS-3005S) to provide constant voltage. The distance between electrodes was fix at 4 cm. An air blower was pumped at 1 lair/lsample/min near the cathode and Magnetic stirring with 100 rpm was used to avoid separation (Figure 1). The HEF process was then performed under differential operation condition, which were adjusted for each set experiment. Next, the solution was separated by a centrifuge, the water was used for COD and color analyses. The electrodes were washed with HCl 1N solution and distilled water after the experiments. 2.4. Data handling Experimental designs were made by using Modde 5.0 (Umetrics, Umea, Sweden). Based on Box Behnken Design, 15 experiments were designed for 3 factors at three levels, including catalyst dosage (x1; 0.5, 1, 1.5 g/l), pH (x2; 3, 4, 5), and Current density (x3; 10, 15, 20 mA/cm2), 2 responses were the removal efficiencies of COD (Y1) and Color (Y2). The polynomial models for the removal efficiencies of COD and color with respect to the HEF variables were expressed as equation (4) [9]: εββββ ++++= ∑ ∑∑∑ = = == 3 1 3 1 3 1 3 1 2 0 i i j jiij i iiiii xxxxY (4) where Y is the predicted response by the model, β0, βi, βii, βij, are the constant regression coefficients of the model. xi, xi2 and xixj represent the linear, quadratic and interactive in terms of the uncoded independent variables, respectively; ε is the random error. The removal efficiencies of COD and Color were calculated as equation (5): 100,% in outin X XX Y − = (5) where Y, % is the removal efficiencies of COD or Color, Xin is the input of COD, mg/l or Color, Pt-Co, Xout is the output of COD, mg/l or Color, Pt-Co. 2.5. Parameter Analysis Sample parameters were analyzed based on EPA and national technical regulation: Color Heterogeneous Electro Fenton process for textile wastewater treatment: Application of 197 (2120C, Spectrophotometric Method (EPA, 1999)); COD (5220D. Closed Reflux, Titrimetric Method (EPA, 1999); pH (TCVN 6492:2011). 3. RESULTS AND DISCUSSION 3.1. Statistical Analysis and Second-order Polynomial Equation The ANOVA for the polynomial models of HEF process is shown in Table 2. The determination coefficient values (R2) of the model were 0.970 and 0.994 for Y1 and Y2, respectively, which were higher than 0.97 and considered as realizable model [10]. The F values calculated for the model regressions were 18.139 for Y1 and 85.002 for Y2. All these values were greater than 3.29 as theoretically calculated for 95 % confidence level as given by Neto and Scarminio [11], corroborating that the developed models are statistically significant. In addition, the F values calculated for the lack of fit of the models were 2.072 for Y1 and 1.831 for Y2 , which were both lower than 19.30 tabulated for 95 % confidence level as given by Neto and Scarminio [11]. Table 2. ANOVA results for the fitted models. Y1: R2 = 0.97; R2adjusted = 0.917 Y2: R2 = 0.994; R2adjusted = 0.982 DF SS MS F p DF SS MS F p Regression 9 1546.47 171.83 18.139 0.003 9 1421.8 157.978 85.002 0 Lack of Fit 3 35.8376 11.9459 2.0727 0.342 3 6.81258 2.270 1.831 0.372 Note: DF: Degree of freedom; SS: Sum of square; MS: Mean of square; F: Fisher value; p: p values. (a) (b) (c) (d) Figure 2. Statistical analysis graphs. The residuals for Y1 and Y2 are distributed without rule around the mean, so the agreement of the models Figure 2(a), (b) are good and systematic errors are eliminated [12]. Figure 2(c),(d) show that, the data points on these graphs lied close to a diagonal lines, so all variations in the second-order equations are significant factors. All these evidences were proved for the comfort of initial assumptions [11]. The ANOVA for the polynomial equations is shown in Table 3. Almost p values were lower than 0.05, implied that the model terms were significant at 95 % of probability level [11]. In the polynomial equation for Y1, three model terms x1x2, x1x3, x2x3 have p values higher than -3 -2 -1 0 1 2 3 60 70 80 90 Re sid u al s fo r Y1 Y1,% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 -3 -2 -1 0 1 2 3 70 80 90 Re sid u al s fo r Y2 Y2, % 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0.005 0.01 0.02 0.05 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.95 0.98 0.99 0.995 -4 -3 -2 -1 0 1 2 3 4 N- Pr ob ab ilit y fo r Y1 Residuals for Y1 3 8 9 14 6 1 11 13 10 4 7 15 12 5 2 0.005 0.01 0.02 0.05 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0.95 0.98 0.99 0.995 -4 -3 -2 -1 0 1 2 3 4 N- Pr ob ab ilit y fo r Y2 Residuals for Y2 1 8 6 15 10 12 3 14 2 9 11 13 7 5 4 Nguyen Duc Dat Duc, Nguyen Thi Chi Nhan, Nguyen Hoang Ha, Nguyen Tan Phong 198 0.05, indicated that there are no interaction between x1, x2, x3 in pairs. Model terms where p values higher than 0.05 were ignored from the equations. The final equations were displayed in equation (6), (7): Y1 = 91.033 +5.850x1 -3.763x2 + 5.937x3 – 9.417x12 – 10.042x22 – 9.042x32 (6) Y2 = 97.300 + 5.325x1 – 3.463x2 + 5.638x3 – 4.800x12 – 4.475x22 – 13.725x32 + 3.325x1x2 (7) Table 3. ANOVA for the second-order polynomial equation. Y1: R2 = 0.97; R2adjusted = 0.917 Y2: R2 = 0.994; R2adjusted = 0.982 Coefficient SD p-value Coefficient SD p-value Constant 91.0333 1.77697 5.36E-08 97.3 0.78709 6.57E-10 X1 5.85001 1.08817 0.003 5.32501 0.48199 0.00011 X2 -3.7625 1.08817 0.01809 -3.4625 0.48199 0.00081 X3 5.93749 1.08817 0.00281 5.6375 0.48199 8.03E-05 X1*X1 -9.4167 1.60174 0.00202 -4.8 0.70947 0.00107 X2*X2 -10.042 1.60174 0.00152 -4.475 0.70947 0.00147 X3*X3 -9.0417 1.60174 0.00242 -13.725 0.70947 6.81E-06 X1*X2 -0.525 1.5389 0.74685 3.32499 0.68164 0.00456 X1*X3 -0.725 1.5389 0.6574 1.625 0.68164 0.06286 X2*X3 -1.2 1.5389 0.4708 1.14999 0.68164 0.15239 3.2. Determination and Accreditation of the Optimum Condition (a) (b) Figure 3. Contour graphs of removal efficiencies of COD and Color. With Modde 5.0 software, the maximum points represented the optimum condition were shown in Figure 3. In view of that, the optimal condition was at pH = 3.8, catalyst dosage = 1.1 g/l, and current density = 17 mA/cm2. The treatment at this condition provide the removal efficiencies of 93.2 % for Y1 and 99.6 % for Y2. To evaluate the precise of the predicted optimum condition, some experiments were made 99.0 94.8 90.6 86.4 82.2 78.0 73.8 69.6 65.4 Heterogeneous Electro Fenton process for textile wastewater treatment: Application of 199 at this condition. Table 4 shows that, the removal efficiencies of COD and Color were 91.6 % (COD = 56 mg/l) and 98.4 (Color = 16 Pt-Co) observed that the experimental values and predicted values were similar. Furthermore, Color and COD in the effluent achieved the QCVN 13:2015/BTNMT (Column A). These results are slightly different from previous reports [4-6, 8] because of the different in wastewater, treatment process and experiment design method. Table 4. Experiment results under optimum condition. No. Parameters input (n = 3) (Average ± SD) output (n = 3) (Average ± SD) Removal Efficiencies QCVN 13:2015/BTNMT, Column A 1 COD, mg/l 673 ± 32 56 ± 12 91.6 % 100 2 Color, Pt-Co 1025 ± 28 16 ± 6 98.4 % 75 4. CONCLUSIONS The application of Fe3O4-Mn3O4 in the Heterogeneous Electro Fenton process for textile wastewater treatment was successfully optimized by Box Behnken design. The regression polynomial equations were formulated for the relations between variables and responses have high reliability (R2Y1 = 0.97 and R2Y2 = 0.994. At pH = 3.8, Fe3O4-Mn3O4 dosage of 1.1 g/l, and current density of 17.00 V, the treatment efficiencies are highest. In the validation test, the removal efficiencies of COD and Color were 91.6 % and 98.4 %, respectively, which are similar to their predicted response values. Lastly, the Heterogeneous Electro Fenton process with Fe3O4- Mn3O4 as a catalyst is a good solution for textile wastewater treatment since the Color and COD in the effluent achieved the national standard QCVN 13:2015/BTNMT, column A. Acknowledgment. This research is funded by Vietnam National University Ho Chi Minh City (VNU- HCM) under grant number B2018-20-03. REFERENCES 1. Nidheesh P. V., Gandhimathi R., and Sanjini N. S. - NaHCO3 enhanced Rhodamine B removal from aqueous solution by graphite–graphite electro Fenton system, Separation and Purification Technology 132 (2014) 568-576. 2. Azbar N. - Comparison of various advanced oxidation processes and chemical treatment methods for COD and color removal from a polyester and acetate fiber dyeing effluent, Chemosphere 55 (2004) 35-43. 3. Ahmad A. - Effect of preparation conditions of activated carbon from bamboo waste for real textile wastewater, Journal of Hazardous Materials 173 (2010) 487-493. 4. Rhadfi T., Piquemal J., Sicard L., Herbst F., Briot E., Benedetti M., and Atlamsani A. - Polyol-made Mn3O4 nanocrystals as efficient Fenton-like catalysts, Applied Catalysis A: General 386 (2010) 132-139. 5. Silva G. C., Ciminelli V. S. T., Ferreira A. M., Pissolati N. C., Paiva P. R.P., and López J. L. - A facile synthesis of Mn3O4/Fe3O4 superparamagnetic nanocomposites by chemical Nguyen Duc Dat Duc, Nguyen Thi Chi Nhan, Nguyen Hoang Ha, Nguyen Tan Phong 200 precipitation: Characterization and application in dye degradation, Materials Research Bulletin 49 (2014) 544-551. 6. Wan Z. and Wang J. - Degradation of sulfamethazine antibiotics using Fe3O4-Mn3O4 nanocomposite as a Fenton-like catalyst, Journal of Chemical Technology & Biotechnology (2016) n/a-n/a. 7. Nidheesh P. V. and Gandhimathi R. - Trends in electro-Fenton process for water and wastewater treatment: An overview, Desalination 299 (2012) 1-15. 8. Costa R. C. C., Lelis M. F. F., Oliveira L. C. A., Fabris J. D., Ardisson J. D., Rios R. V. A., Silva C. N., and Lago R. M. - Remarkable effect of Co and Mn on the activity of Fe3 − xMxO4 promoted oxidation of organic contaminants in aqueous medium with H2O2, Catalysis Communications 4 (2003) 525-529. 9. Box G. E. P and Draper N. R. - Empirical Model-Building and Response Surfaces, New York: Wiley, 1987. 10. Maran P. J., Sivakumar V., Thirugnanasambandham K., and Sridhar R. - Response surface modeling and analysis of barrier and optical properties of maize starch based edible films, International Journal of Biological Macromolecules 60 (2013) 412-421. 11. Neto B. B., Scarminio I. S., and Bruns R. E., Statistical Design - Chemometrics. Data Handling in Science and Technology, ed. I. S. Scarminio R.E. Bruns and B. de Barros Neto, Elsevier 25 (2005) 1-7. 12. El-Ghenymy A., Garcia-Segura S., Rodríguez R. M., Brillas E., Begrani M. S. E., and Abdelouahid B. A. - Optimization of the electro-Fenton and solar photoelectro-Fenton treatments of sulfanilic acid solutions using a pre-pilot flow plant by response surface methodology, Journal of Hazardous Materials 221 (2012) 288– 297.