Nitrobenzene (NB) is one of the main organic contaminants in water and is extensively used in dyes, explosives, and insecticides. However, it is infamous for its great carcinogenicity, toxicity, and other opposing influences on environments and organisms. Furthermore, NB demonstrates poor activity towards oxidative degradation due to the strong electronwithdrawing features of
the nitrogroup. According to the U.S. Environmental Protection Agency, the level of NB in water and lake streams should not exceed 17 ppm (1). Therefore, environmental risks have urged the advance of effective methods for NB elimination.
Elimination of NB through reducing it to less poisonous aniline has been investigated and the results have shown that the aquatic environment was still at risk even when the level of aniline reached the environmental boundary (2). Hence, the progress of operative approaches for NB degradation remains an important challenge. Thanks to the strong oxidative and nonselective features of hydroxyl radical, different advanced oxidation processes (AOPs) are used to degrade organic pollutants in the water phase (3, 4).
Several techniques of AOPs, including Fenton (5), ozonation in alkaline media (6), electro Fenton (7), nano catalytic ozonation (8), and sonophotoFenton have been applied for the treatment of organic contaminants in an aqueous environment (9). Chemical treatment techniques, such as chlorination for NB degradation, can form chlorinated products which are nonbiodegradable and toxic. The physical approaches just transfer pollutants from one phase to another.
Therefore, the Fenton and photoFenton processes (10–13) have been employed for this purpose. They are good choices since they need lowcost chemicals and low energy
(14–15). Obviously, the Fenton method creates hydroxyl radicals (OH^{•}) which are based on the reaction between H_{2}O_{2} and Fe^{2+ }(Equations 1 and 2).
(Equation 1)
(Equation 2)
Furthermore, iron (III) can react with the residual amount of hydrogen peroxide and create iron (II) (Equation 2) (16). Nevertheless, it should be noted that the homogeneous
Fenton process has some important issues. Homogeneously catalyzed reactions require up to 50–80 mg/l of Fe ions in solution which is well above the Regulation. In addition, at the end of the wastewater treatment, the treatment of the sludge enriched with Fe ions is costly and needs substantial amounts of chemicals and the workforce. Some efforts have been made to solve these problems and investigate the conditions for the recovery of the catalyst. The Fentonlike processes have been employed to form highly reactive hydroxyl radicals with H_{2}O_{2 }in the presence of metal cations under appropriate reaction conditions (1718).
The comparison of electroFenton, Fenton, and photoFenton in terms of their efficiency in dye removal was investigated by Huang
et al. (19). They tried to transform organic wastewater, such as Remazol Black B dye into inorganic matter using the abovementioned methods. They found that the removal efficiency was 70% by the Fenton technique, whereas it reached about 93% by the electroFenton method. This conclusion was due to the reduction of iron ions in the cathode applied in the process. After the application of ultraviolet light, photoFenton had an efficacy of more than 98% that was based on the recovery of iron ions. Based on the results of the abovementioned research, organic contaminants, like formic acid, which are hard to remove by the Fenton process, could be quickly and completely degraded by the photoFenton technique. The oxalic acid, which can be remediated hardly by other approaches can be removed by the electroFenton method (19).
In the present study, the synthetic wastewater comprising NB was treated by the Fentonlike process. This research inspected the influence of various variables, including pH, the concentration of hydrogen peroxide, and ferric ion on NB removal using the BoxBehnken design (BBD) method. Materials & Methods
Materials
The tests were performed employing hydrogen peroxide solution (30% w/w), Ferric chloride (≥99.5%) as a source of iron (III) ions (manufactured in Merck, US), and MnO_{2} (≥99%) (manufactured in SigmaAldrich, Germany). Moreover, doubledistilled water was used in all the experiments.
Figure 1) Experimental setup in laboratory scale: 1) cooling water jacket, 2) magnetic stirrer bar, 3) UV lamp, 4) glass reactor, 5) magnetic stirrers, 6) input wastewater, 7) output wastewater, 8) cooling water input, 9) cooling water output, and 10) electric connection
Photoreactor and experiments
All runs were performed in a oneliter glass photo reactor. The system was prepared with a sampling system (Figure 1). A 15W (UVC) mercury lamp (manufactured in Philips, Netherlands) was used as the source of light and positioned vertically at the center of the reactor. The reactor was equipped with a water jacket with an external flow controlled by a thermostat to control the temperature at 25° C. The BW 20G water bath (manufactured in JeioTech, Korea) was employed to fix the temperature at 25° C. A stirrer was used to mix the solution and prevent the precipitation of iron ions. The PT10P pH meter (manufactured in Sartorius, Germany) was employed to adjust the initial pH of the solution. The samples were transferred to a
spectrophotometer (DR 5000, Hach, Jenway, USA) at 600 nm to determine the percentage of chemical oxygen demand (COD) removal using standard methods (20).
Approximately 1000 ml of synthetic wastewater was used during each experiment. Ferric and Hydrogen peroxide concentrations and pH were the chosen variables, and the response was the removal of COD in wastewater. It should be noted that the reaction time was 30 min in all tests. The percentage of NB removal was calculated using Equation (3):
(Equation 3)
Where and are NB values at the initial and times, respectively. The remaining amount of hydrogen peroxide in the samples can interfere with the COD test; hence, it was removed by MnO_{2} powder. The MnO_{2} powders were separated by filtration (21).
Statistical analysis and experimental design
The experimental design was employed to optimize the percentage of NB removal in a scientific way. Effects of Ferric concentration ( ), hydrogen peroxide concentration ( ), and acidity ( ) on NB removal were explored. The input variables, including , , and , as well as their values and dimensions are presented in Table 1.
The BoxBehnken experimental design needs a fewer number of experiments (15 experiments for three variables) (22). Based on the design of experiments, as a polynomial equation of independent variables, the subsequent model for the response variable (Y) was offered (Equation 4):
Table 1) Experimental levels and ranges of independent factors
Variable
Unit
Symbol
Range and levels
Low (1)
Middle (0)
High (+1)
Hydrogen peroxide concentration
mM
6
12
18
Ferric Concentration
mM
0.6
1.2
1.8
pH

3
7
11
(Equation 4)
Where is the remainder of the equation, is a constant number, is a linear interaction between the input variables of and ( and ), is the slope of the variable, and is the second order of the input variable ( ). Analysis of variance (ANOVA) was applied to test the significance of the operating variable in the polynomial equation (Equation 4) (23). In the ANOVA, a pvalue of 0.05 was considered statistically significant, and pvalues of less than 0.0500 indicated the significance of model terms. The values greater than 0.1000 indicate that the model terms are not significant.
The statistical significance of the secondorder models was defined by Fvalue. If the calculated Fvalue is higher than the Fvalue, the pvalue will be much smaller which indicates the significance of the statistical model. As can be seen in Equation 5, the calculated Fvalue was gained by dividing the mean squares of regression (including square, linear, and interaction) by the mean squares of residual as the following (24):
(Equation 5)
Residual degrees of freedom are the total degrees of freedom minus the regression degree of freedom, and the regression degree of freedom is the number of terms minus one (24). Design of the experiments included 15 tests (2 replicates at the central point and 13 tests). All the experiments were performed randomly to reduce experimental errors. Table 2 summarizes the percentages of NB removal.
Table 2) Design of the experiments with three independent parameters and their response
Responses (%)
Manipulated variables
Run no.
68.5
11
12
1.8
1
71.0
11
18
1.2
2
97.4
7
12
1.2
3
97.1
7
12
1.2
4
85.9
7
6
1.8
5
78.5
3
12
1.8
6
89.6
7
18
1.8
7
77.3
3
6
1.2
8
81.0
7
6
0.6
9
96.3
7
12
1.2
10
63.5
11
6
1.2
11
71.2
3
12
0.6
12
89.0
3
18
1.2
13
85.1
7
18
0.6
14
60.0
11
12
0.6
15
Results
Design of experiments and analysis of variance
Optimum conditions for maximizing the removal of NB were determined using the Fentonlike process. In this project, BBD was used to investigate the influence of three independent variables (H_{2}O_{2} and Fe^{3+} concentration, and pH) on the percentage of NB removal ( ) for the achievement of desired conditions. Table 2 tabulates the empirical results in the percentage of NB
removal. The quadratic relationship was gained using the least squares error method to determine the response in terms of the three independent variables as the following:
(Equation 6)
Predicted values of the quadratic model in relationship with the observed values in NB removal are showed in Figure 2 and the estimated values were calculated via Equation 6.
Table 3 shows the results of ANOVA for evaluation of the model. According to Table 3, the degree of freedom for the residual error and model were five and nine, respectively. Comparison of the tabulated Fvalue in Table 3 with the calculated Fvalue revealed that the calculated Fvalue for the model was greater than the tabulated Fvalue. Therefore, the pvalue was very low (<0.0001) and the model was extremely significant.
Terms of C_{HP}, C_{F}, pH, and their squares had a pvalue of less than 0.01; hence, they were highly significant (25). However, the term of the binary interaction between the variables had a pvalue of more than 0.05, meaning that the interaction of variables was nonsignificant. In addition, the pvalue of lack of fit was more than 0.05 which indicated the model did not fit the data.
Effects of operational parameters
As it was mentioned earlier, the effects of Fe^{3+} concentration (0.6, 1.2, and 1.8 mM),
Figure 2) The predicted versus experimental values in the removal of nitrobenzene
Table 3) ANOVA results for the existing model in the removal of nitrobenzene
pvalue
FValue
MS
DF
SS
Source
0.001
32.47
229.93
9
2069.41
Model
0.016
12.87
91.12
1
91.12
XC_{HP}
0.020
11.21
79.38
1
79.38
X_{C Fe}^{3+}
0.001
49.58
351.12
1
351.12
X_{CpH}
0.070
5.26
37.22
1
37.22
X^{2}_{CHP}
0.001
40.60
287.56
1
287.56
X^{2}_{CFe}^{3+}
0.000
188.71
1336.43
1
1336.43
X^{2}_{CpH}
0.943
0.01
0.04
1
0.04
X_{CHP}.X_{CFe}^{3+}
0.466
0.62
4.41
1
4.41
X_{CHP}.X_{CpH}
0.831
0.05
0.36
1
0.36
X_{CFe3+}.X_{CpH}
7.08
5
35.41
Error
11.80
3
35.41
Lack of fit
0.00
2
0.00
Pure error
14
2104.82
Total
Model Summary
S
R^{2}
2.66120
98.32%
9529%
73.08%
Figure 3) Contour plots of nitrobenzene removal in terms of Ferric ion and hydrogen peroxide concentration (°C, , and )
hydrogen peroxide concentration (6, 12, and 18 mM), and pH values (3, 7, and 11) on the removal of NB were investigated. The threedimension (3D) plots versus these variables in the removal of NB are presented in Figures 3 to 5. It should be noted that they are plotted based on Equation 6.
A 3D plot of the NB removal in terms of Ferric and hydrogen peroxide concentration is presented in Figure 3. It was found that the highest percentage of NB removal was gained at the mean values for both variables (Ferric ion and hydrogen peroxide concentrations). In the Fentonlike process, the removal efficiency of NB was decreased with an excessive increase in the Fe (III) concentration. The treatment proficiency was reduced with a decline in H_{2}O_{2 }dosage since the breakdown of H_{2}O_{2} was decreased in the presence of a catalyst to form hydroxyl radical. However, an extreme H_{2}O_{2} dosage was not suitable since first, it would raise the treatment cost dramatically. Moreover, an extreme H2O2 dosage results in an excessive scavenger effect of H_{2}O_{2} on hydroxyl radical (Equation 7) (2627).
→ + (Equation 7)
Therefore, the concentration of H_{2}O_{2} should be in the optimum amounts. For this reason, it was optimized in the present research.
Figure 4) Contour plots of nitrobenzene removal in terms of pH and hydrogen peroxide concentration ( mM, ° C, and )
As it was clear in figures 3 to 5, the increase in the catalyst dosage was helpful to some extent. However, the catalyst could not be added without any limitation. The undue loading of the catalyst has a negative effect on the removal of COD since it may lead to the scavenger effect (Equations 8–10).
→ +(Equation 8)
→ + (Equation 9) → +(Equation 10)
If the catalyst is excessive, the extra catalyst would spend the produced hydroxyl radical. Higher amounts of catalyst would clearly increase the treatment cost and result in the production of a large volume of sludge which, in turn, would raise the need for the sludge treatment. Figure 4 shows the percentage of NB removal in terms of hydrogen peroxide concentration and pH.
It was obvious that in mean values of pH, the percentage of NB removal was increased and the changes in the concentration of H_{2}O_{2} in similar pH values had a low influence on NB removal. In addition, the interaction between the concentration of H_{2}O_{2} and pH was nonsignificant. Nevertheless, the removal of NB was slightly higher in the mean values of hydrogen peroxide concentrations, compared to other values.
Discussion
Figure 5 displays the percentage of NB removal in terms of pH and Ferric concentration. It was clear that in the mean value of pH, the percentage of NB removal from wastewater was at the highest level. Effect of Fe^{3+} on NB removal was the same as the effect of hydrogen peroxide concentration. As presented in Figure 5, the best pH in this research was 6.23 based on maximum NB removal. This could be created from the coagulation of ferric species which can play a significant role under neutral pH. These results are in agreement with those of other investigators about the homogeneous Fentonlike catalyst (28).
At pH > 3, the main fraction of Fe (III) precipitates as Fe (OH)_{3}, dwindling the reaction between H_{2}O_{2 }and Fe^{3+} which leads to the manufacture of Fe^{2+}. However, at neutral pH, the coagulant properties of Ferric ion overcome this phenomenon. In the alkaline condition (pH at 9), the removal of NB was reduced since the H_{2}O_{2} degraded faster into H_{2}O and O_{2}.
As shown in figures 3 to 5, it is obvious that the percentage of NB removal can be high in some circumstances. For optimization of the
Figure 5) Contour plots of nitrobenzene removal in terms of Ferric concentration and acidity ( mM, ° C, and ) Table 4) Optimal conditions in the removal of NB (°C and )
Parameters
Unit
Value
Hydrogen peroxide concentration
mM
15.33
Ferric ion concentration
mM
1.30
pH

6.23
Predicted NB removal
%
99.23
Experimental NB removal
%
99.0
percentage of NB removal, the values of different variables were gained via DesignExpert software (version 7.0.0). These optimum values, along with the percentage of NB removal are summarized in Table 4. As it is obvious, the maximum rate of NB removal was gained under the subsequent conditions: Ferric concentration of 1.30 mM, hydrogen peroxide concentration of 15.33 mM, and a pH of 6.23. Under these conditions, the maximum rate of NB removal by the experiments and model (Equation 6) with three replications were 99.0% and 99.23%, respectively.
In this project, the efficacy of the Fentonlike method versus the initial concentration of the NB can be explored but it was fixed in all runs. the lifetime of hydroxyl radicals is only a few nanoseconds and they can only react where they are produced, therefore tChance of interaction between NB molecules and hydroxyl radicals can be improved by increasing the number of NB molecules per volume unit which results in the enhancement of the degradation efficiency (29).
Conclusion
In this study, the synthetic wastewater containing NB was treated by the Fentonlike process. The BBD method was used and the effect of H_{2}O_{2}, Fe^{3+}, and pH on the removal of NB was studied in synthetic wastewater. The ANOVA was used for the analysis of experimental results. The results showed that the operational factors were significant and effective.
A quadratic model was applied to analyze the influence of variables on the percentage of NB removal. At optimum conditions (hydrogen peroxide and Ferric with concentrations of 15.33 and 1.30 mM, respectively, and a pH of 6.23) and after 30 min of reaction, the removal efficiency for NB and COD were 99.0% and 56.7%, respectively. The Fentonlike process was powerful in the removal of NB; however, it could remove the COD to some extent.
Footnotes
Acknowledgements
The authors’ deepest appreciations go to the officials and managers of Pars Special Economic Energy Zone for their technical support. Conflicts of Interest
The authors declare that there was no conflict of interest in this research.
References
Sun L, Song H, Li Q, Li A. Fe/Cu bimetallic catalysis for reductive degradation of nitrobenzene under oxic conditions. Chem Eng J 2016;283:36674. Link
Mu Y, Yu HQ, Zheng JC, Zhang SJ, Sheng GP. Reductive degradation of nitrobenzene in aqueous solution by zerovalent iron. Chemosphere 2004;
54(7):78994. PMID: 14637335
Duan H, Liu Y, Yin X, Bai J, Qi J. Degradation of nitrobenzene by Fentonlike reaction in a H_{2}O_{2}/ schwertmannite system. Chem Eng J 2016;283:87379. Link
Xu S, Zhu H, Cao W, Wen Z, Wang J, FrancoisXavier CP, et al. CuAl_{2}O_{3}gC_{3}N_{4} and CuAl_{2}O_{3}Cdots with dualreaction centres for simultaneous enhancement of Fentonlike catalytic activity and selective H_{2}O_{2} conversion to hydroxyl radicals. Appl Catal B Environ 2018;234:22333. Link
Bae S, Kim D, Lee W. Degradation of diclofenac by pyrite catalyzed Fenton oxidation. Appl Catalysis B Environ 2013;134:93102. Link
Shokri A. The treatment of spent caustic in the wastewater of olefin units by ozonation followed by electrocoagulation process. Desal Water Treat 2018;
111:17382. Link
Shokri A. A kinetic study and application of electroFenton process for the remediation of aqueous environment containing toluene in a batch reactor. Russ J Appl Chem 2017;90(3):4527. Link
Shokri A. Degradation of 4–Nitrophenol from industerial wastewater by nano catalytic Ozonation. Int J Nano Dimension 2016;7(2):1607. Link
Shokri A. Application of Sono–photoFenton process for degradation of phenol derivatives in petrochemical wastewater using full factorial design of experiment. Int J Ind Chem 2018;9(4):295303. Link
Trovo AG, Melo SA, Nogueira RF. Photo degradation of the pharmaceuticals amoxicillin, bezafibrate and paracetamol by the photoFenton processapplication to sewage treatment plant effluent. J Photochem Photobiol A Chem 2008;198(23):21520. Link
Babuponnusami A, Muthukumar K. A review on Fenton and improvements to the Fenton process for wastewater treatment. J Environ Chem Eng 2014;
2(1):55772. Link
da Silva Leite L, de Souza Maselli B, de Aragao Umbuzeiro G, Nogueira RF. Monitoring ecotoxicity of disperse red 1 dye during photoFenton degradation. Chemosphere 2016;148:5117. PMID: 26841294
PeraltaHernandez JM, Vijay S, RodriguezNarvaez O, PachecoAlvarez MA. Photo and solar fenton processes for wastewater treatment. Electrochemical water and wastewater treatment. Oxford: ButterworthHeinemann; 2018. Link
GarciaSegura S, Bellotindos LM,YaoHui H, Brillas E, MingChun L. Fluidizedbed Fenton process as alternative wastewater treatment technologya review. J Taiwan Institute Chem Eng 2016;67:21125. Link
Clarizia L, Russo D, Di Somma I, Marotta R, Andreozzi R. Homogeneous photoFenton processes at near neutral pH: a review. Appl Catal B Environ 2017;209:35871. Link
Funai DH, Didier F, Gimenez J, Esplugas S, Marco P, Junior AM. PhotoFenton treatment of valproate under UVC, UVA and simulated solar radiation.
J Hazard Mater 2017;323(Pt A):53749. PMID: 27387276
Shokri A, Bayat A, Mahanpoor K. Employing Fentonlike process for the remediation of petrochemical wastewater through Box–Behnken design method. Desal Water Treat 2019;166:13543. Link
Huang YH, Huang YF, Chang PS, Chen CY. Comparative study of oxidation of dyeReactive Black B by different advanced oxidation processes: Fenton, electroFenton and photoFenton. J Hazard Mater 2008;154(13):65562. PMID: 18063472
American Public Health Association. Water Pollution control federation, & water environment federation. Standard methods for the examination of water and wastewater. Washington, D.C: American Public Health Association; 1965. Link
Elsousy K, Hussen A, Hartani K, El Aila H. Elimination of organic pollutants using supported catalysts with hydrogen peroxide. J Chem 2007;
2(1):97103. Link
Ahmadi M, Rahmani K, Rahmani A, Rahmani H. Removal of benzotriazole by PhotoFenton like process using nano zerovalent iron: response surface methodology with a BoxBehnken design. Pol J Chem Technol 2017;19(1):10412. Link
Rumky J, Ncibi MC, BurgosCastillo RC, Deb A, Sillanpa M. Optimization of integrated UltrasonicFenton system for metal removal and dewatering of anaerobically digested sludge by BoxBehnken design. Sci Total Environ 2018;645:57384. PMID: 30032079
Karimi S, Shokri A, Aghel B. Remediation of spent caustic in the wastewater of oil refinery by photoFenton process. Arch Hyg Sci 2020;9(3):17988. Link
Shokri A, Hosseini J. Treatment of synthetic wastewater containing diethyl phthalate through photoFenton method by boxBehnken design. Arch Hyg Sci 2020;9(2):12131. Link
Shokri A, Karimi S. Treatment of aqueous solution containing acid red 14 using an electro peroxone process and a boxBehnken experimental design. Arch Hyg Sci 2020;9(1):4857. Link
Xu HY, Liu WC, Qi SY, Li Y, Zhao Y, Li JW. Kinetics and optimization of the decoloration of dyeing wastewater by a schorlcatalyzed Fentonlike reaction. J Serb Chem Soc 2014;79(3):36177. Link
Yang XJ, Tian PF, Zhang XM, Yu X, Wu T, Xu J, et al. The generation of hydroxyl radicals by hydrogen peroxide decomposition on FeOCl/SBA15 catalysts for phenol degradation. AlChE J 2014;61(1):16676. Link
Daud NK, Hameed BH. Decolorization of acid red 1 by Fentonlike process using rice husk ashbased catalyst. J Hazard Mater 2010;176(13):93844. PMID: 20042285
Shokri A, Noshad G, Ali Hosseini A. Degradation of Nitrobenzene in an Aqueous Environment through Fentonlike Process Using Box–Behnken Design Method. Arch Hyg Sci. 2021; 10 (1) :5866 URL: http://jhygiene.muq.ac.ir/article1460en.html