Environ Sci Pollut Res DOI 10.1007/s11356-015-4406-x
RESEARCH ARTICLE
Selective oxidative degradation of toluene for the recovery of surfactant by an electro/Fe2+/persulfate process Anhua Long 1,2 & Hui Zhang 1
Received: 29 October 2014 / Accepted: 18 March 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract An electro/Fe2+/persulfate process has been conducted for toluene removal from surfactant (SDS) flushing solution, and the pseudo-second-order reaction rate constant (k2 value) of toluene removal has been optimized by a response surface methodology (RSM). The results indicated that in this process, the reaction between persulfate and externally added Fe2+ generates sulfate-free radicals, and at the same time, Fe2+ is electro-regenerated at the cathode by the reduction of Fe3+. RSM based on Box–Behnken design (BBD) has been applied to analyze the experimental variables, of which the concentrations of persulfate and Fe2+ showed a positive effect on the rate constant of toluene removal, whereas the concentration of SDS showed a negative effect. The interactions between pairs of variables proved to be significant, such as between SDS, persulfate, and Fe 2+ concentrations. ANOVA results confirmed that the proposed models were accurate and reliable for analysis of the variables of the electro/Fe2+/persulfate process. The shapes of the 3D response surfaces and contour plots showed that the SDS, persulfate, and Fe2+ concentrations substantially affected the k2 value of toluene removal. The results indicated that increasing persulfate or Fe2+ concentration increased the k2 value, whereas Responsible editor: Angeles Blanco Electronic supplementary material The online version of this article (doi:10.1007/s11356-015-4406-x) contains supplementary material, which is available to authorized users. * Hui Zhang [email protected] 1
Department of Environmental Engineering, Wuhan University, Wuhan 430079, China
2
Jiangxi Science and Technology Normal University, Jiangxi 330013, China
increasing SDS concentration decreased the k2 value. The reaction intermediates have been identified by GC-MS, and a plausible degradation pathway for toluene degradation is proposed.
Keywords Surfactant recovery . Electrochemical treatment . Persulfate . Selective oxidation . Sulfate radical
Introduction Benzene, toluene, ethylbenzene, and xylenes (BTEX) is an abbreviation used for four related compounds found in coal tar, crude petroleum, and a wide range of petroleum products. Among them, toluene is one of the most abundantly produced chemicals, with worldwide annual production of 5–10 million tons (ATSDR 2000). The largest source of toluene release is during the production, transport, and use of petroleum, which contains about 5–8 % toluene. Once released into the environment, toluene usually evaporates into the air. Toluene can also dissolve in water, attach to soil particles, and it may be found in surface and groundwater at contaminated sites or in the close vicinity of natural oil, coal, and gas deposits (Li et al. 2013). The main effect of toluene is on the brain and nervous system, with fatigue and drowsiness being the most obvious symptoms (ATSDR 2000), and it can cause neurological damage (ATSDR 2001). In this context, activated persulfate oxidation has emerged as a novel effective remediation technology for the removal of toluene (Do et al. 2011; Huang et al. 2005; Liang et al. 2008, 2009). This system generates sulfate : radical (SO4 −), a strong oxidant (E0 =2.60 V) that can be thermally or chemically activated by initiators (Shukla et al. 2010; Tan et al. 2013). The activation methods include UV, heat, transition metal ions, alkaline conditions, and persulfate combined with other oxidants. As a natural substrate, ferrous
Environ Sci Pollut Res
ion (Fe2+) has the advantages of being inexpensive and nontoxic, and it has been widely applied as a catalyst to effectively activate persulfate (Jiang et al. 2013; Kusic et al. 2011; Long et al. 2014a; Xu et al. 2014; Zhao et al. 2014). However, some drawbacks are associated with the Fe2+/persulfate process. For instance, high Fe2+ dosage is required to activate persulfate because Fe2+ is difficult to regenerate after conversion to Fe3+ (Vicente et al. 2011). This results in a large amount of iron : sludge. Moreover, the excess Fe2+ could react with SO4 − (Eq. (1)) as a radical scavenger, resulting in reduced efficiency of the process. 3þ þ SO2− Fe2þ þ SO•− 4 →Fe 4
ð1Þ
In order to solve these problems, an electrochemical method has been applied to the reaction system so that Fe2+ could be regenerated according to Eq. (2) (Wu et al. 2012b): þ
Fe3 þ e− → Fe2
þ
Dionysiou 2004; Anipsitakis et al. 2005; Long et al. 2014b; : Zhou et al. 2014). Thus, in the flushing effluents, SO4 − may oxidize toluene at a faster rate than it oxidizes straight-chain aliphatic hydrocarbon surfactants. Indeed, selective oxidation of toluene in SDS flushing effluents was observed in our previous studies (Long et al. 2013, 2014a). Therefore, the objective of the present study was to optimize toluene removal for SDS recovery in flushing effluents from the electro/Fe2+/persulfate process using single-factor tests and response surface methodology (RSM). Three reaction parameters (SDS concentration, persulfate concentration, and Fe2+ concentration) were selected as single factors, and the optimal conditions were obtained by the RSM optimization approach. Furthermore, gas chromatography−mass spectrometry (GCMS) was also applied to determine the intermediates, on the basis of which a degradation pathway of the flushing effluents is proposed.
ð2Þ
Materials and methods In this electro/Fe2+/persulfate process, the cathodic reduction reaction enhances the regeneration of Fe2+, which further : activates persulfate to generate SO4 − (Wang and Chu 2011; 2+ Wu et al. 2012b). As a result, Fe can be continuously regenerated, and the degradation efficiency is improved when an electrochemical potential is applied to the Fe2+/persulfate process. In our previous studies, this process has been successfully applied to the removal of organic pollutants such as Acid Orange 7 (Wu et al. 2012b), bisphenol A (Lin et al. 2013), or clofibric acid (Lin et al. 2014) from water, and even to real wastewater, specifically a landfill leachate (Zhang et al. 2014). It is generally accepted that surfactant-based treatment provides an effective technology for the remediation of toluenecontaminated soil and groundwater (He et al. 2008; Laha et al. 2009; Lee et al. 2002b; Long et al. 2013). However, most surfactants are too expensive for flushing remediation of hydrophobic organic contaminants, and their residues may be biotoxic in the environment (Ying 2006). Usually, sodium dodecyl sulfate (SDS) is used as a surfactant because it is a food grade material and is easily biodegradable by soil and/or aquatic microorganisms (Lee et al. 2002a). Furthermore, the flushing effluents, containing surfactants and toluene, need to be treated before discharge or reuse. Such pollution necessitates investigations of suitable technologies to control flushing effluents and regenerate the surfactant. In a recent review, : SO4 − was proposed as a suitable electrophilic reagent, and when electron-donating groups are present on an aromatic : molecule, the rate of the reaction with SO4 − will increase : (Long et al. 2014b; Tsitonaki et al. 2010). Moreover, SO4 − is relatively stable in aqueous solution, and thus may be able to disperse more freely (Romero et al. 2010; Vicente et al. : 2011). Furthermore, it is generally accepted that SO4 − tends to react selectively via electron transfer (Anipsitakis and
Chemicals and reagents All materials were analytical grade reagents and were used without further purification. Toluene (C6H5CH3) was purchased from Shanghai No. 4 Reagent Factory (China). Sodium persulfate (Na 2 S 2 O 8 ), ferrous sulfate (FeSO 4 · 7H2O), and sodium dodecyl sulfate (SDS, CH 3 (CH 2 ) 11 SO 4 Na) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Apparatus and procedure Batch experiments were performed in an electrolytic reactor (glass beaker) containing 100 mL of solution. A 5 cm× 11.9 cm plate anode (Ti/RuO2-IrO2) and a plate cathode (stainless steel) of the same dimensions were arranged parallel to one another at a distance of 3 cm. The reactor was immersed in a water bath to keep the temperature constant at 20 °C. Electrolysis was conducted under constant current conditions using a direct current (DC) power supply (Model RYI-3010) from Shenzhen Yizhan Instrument Co., Ltd. (China). A magnetic stirrer (Model 78-1, Hangzhou Instrument Motors Factory, China) ensured mixing of the solution in the reactor. Before each run, a fresh stock solution of toluene was prepared with SDS. The initial concentration (C0) was fixed at 1 mM, and the initial pH (background value) was around 7.0. When the DC power supply was initiated, persulfate solution was added as a supporting electrolyte, and the current value was kept at 0.5 A to minimize hydrogen and oxygen evolution, and then, Fe2+ solution was introduced into the electrolytic cell. Aliquots of the mixture were sampled and analyzed at preselected time intervals.
Environ Sci Pollut Res
Analytical methods Direct aqueous injection–gas chromatography/flame ionization detection (DAI-GC/FID) was applied to investigate the concentration of toluene in the surfactant solutions (Kubinec et al. 2005; Long et al. 2014a). Samples were quantified using a gas chromatograph (GC-14C, Shimadzu) with the following settings. GC conditions: oven programmed to a constant temperature of 150 °C (12 min), N2 (99.999 % purity) as carrier gas, and SE-54 (30 m×0.32 mm×0.50 μm) column. FID conditions: temperature set at 250 °C, H2 flow of 60 mL/min, air flow of 400 mL/min, N2 make-up gas, and make-up flow of 20 mL/min. Injector conditions: injection mode with split inlets (split ratio=20:1), injection volume of 1 μL, and injector temperature 250 °C. The intermediate products during the electro/Fe2+/persulfate reaction were detected by headspace–gas chromatography–mass spectrometry (HS-GC-MS) (Esteve-Turrillas et al. 2007). Samples for GC-MS analysis (Shimadzu GCMSQP2010 Plus, Japan) were prepared as follows. A static headspace auto sampler (Thermo Finnigan model HS 2000, Waltham, MA, USA) equipped with standard glass vials of internal volume 10 mL was employed. Once equilibrium between the matrix and the gaseous phase was reached, a 1-mL vapor sample was injected into the system. An HP-5 MS capillary column (30 m length×0.25 mm i.d.×0.25 μm film thickness) was employed for GC separation. The GC equipment was operated in a temperature-programmed mode with an initial temperature of 40 °C held for 2 min, then ramping first to 100 °C at 10 °C min−1 and then further to 200 °C at 20 °C min−1, and finally holding at this temperature for 5 min. The injector and transfer-line temperature was 200 °C. The injector was operated in splitless mode. Helium was used as carrier gas at a flow rate of 5.5 mL/min. The MS-detected scan field was 2–500m/z, and the ion source temperature was 200 °C. The concentration of SDS was determined by the methylene blue method (Kenshi 1975). Methylene blue was first added to the supernatant liquid, then the anionic surfactant– cationic dye complex was extracted with chloroform, and the absorption of the chloroform phase was measured at λmax = 655 nm using a Rayleigh UV-9100 spectrophotometer (Rayleigh Co., China). From this figure, the concentration of SDS could be obtained. Experiment design and data analysis Box–Behnken design (BBD) was applied to investigate the effects of three independent variables, namely, SDS concentration (X1), persulfate concentration (X2), and Fe2+ concentration (X3), on the response function. The low, middle, and high levels of each variable were coded as −1, 0, and +1, respectively, as illustrated in Table 1. The corresponding
Table 1
Experimental range and level of independent variables
Variables
Symbol
SDS concentration (w/v), A Persulfate concentration (mM), B Fe2+ concentration (mM), C
X1 X2 X3
Range and level −1
0
1
0.5 10 5
1 20 10
1.5 30 15
actual values were selected based on available data and preliminary experiments. The dependent variable or objective function was the pseudo-second-order removal rate constant (k2 value). Each independent variable xi was coded as Xi according to the following equation (Singh et al. 2012): X i ¼ ðxi − x0 Þ=Δxi
ð3Þ
where x0 is the value of xi at the center point and Δxi represents the step change value. Thus, a total of 15 experiments based on three levels and three variables was required, including three replicates at the center point in order to estimate the pure error sum of squares. All 15 experiments were conducted in random order. Multiple regression analysis is expressed by the second-order polynomial model as follows (Nair and Ahammed 2014): Y ¼ β0 þ
3 X
βi X i þ
i¼1
3 X i¼1
βii X 2i
þ
2 3 X X
βi j X i X j
ð4Þ
i¼1 j¼iþ1
where Y is the predicted response by the model (k), Xi and Xj are the independent variables, and β0, βi, βii, and βij are the regression coefficients of the model. The validity of the predicted model was verified by analysis of variance (ANOVA), and the second-order model quality was assessed by the determination coefficient (R2). The results were analyzed by a Fischer trial and the probability value (with 95 % confidence level). Finally, the optimal values of the tested variables were obtained by analyzing the surface curves.
Results and discussion Single-factor tests The toluene removal efficiency depends on several main factors, such as the concentrations of SDS, persulfate, and Fe2+ in the electro/Fe2+/persulfate process. Some preliminary runs were performed to identify suitable single factors. Kinetic analysis illustrated that pseudo-second-order kinetics fitted the experimental data better than pseudo-first-order kinetics. As shown in Tables SM-1 and SM-2, the correlation coefficients (R2) for pseudo-second-order kinetics ranged from
Environ Sci Pollut Res
0.9683 to 0.9924, whereas those for pseudo-first-order kinetics were in the range 0.8035–0.9564. When the SDS concentration was increased from 0.5 to 2 %, the toluene removal rate decreased, as indicated in Fig. 1a. The values of rate constant k2 were 11.54×10−3, 8.28×10−3, and 3.02×10−3 L mg−1 min−1 at SDS concentrations of 0.5, 1, and 2 %, respectively. The results indicated that : there were competing reactions of SO4 − with toluene and SDS. A high concentration of SDS probably prevented the activated radicals from reaching the toluene. Moreover, a higher available Fe2+ concentration improved the decomposi: tion of persulfate to generate SO4 −. As can be seen in Fig. 1b, the toluene removal rate increased when the Fe2+ concentration was increased from 2 to 10 mM. The results indicated that the Fe3+ produced in the reaction could be effectively converted to Fe2+ by electro-reduction, further activating persulfate : for SO4 − generation.
Fig. 1 The influence of different parameters on the rate constants under electro/Fe2+/persulfate process: a the effect of Fe2+ concentration (C0 = 1 mM, [persulfate]=30 mM); b the effect of SDS concentration (C0 = 1 mM, [persulfate]=30 mM, [Fe2+]=10 mM)
In order to further investigate the effect of persulfate on the removal of toluene, a persulfate concentration of 20 mM was used. Under the same reaction conditions, the effect of Fe2+ concentration on the removal of toluene was also investigated by conducting experiments at 0, 2, 5, 10, and 20 mM when the SDS concentration was 2 %. As shown in Fig. 2a, b, the k2 value increased from 2.33×10−4 to 23.2×10−4 L mg−1 min−1 as the Fe2+ concentration was increased from 0 to 20 mM. That is to say, increasing the Fe2+ dosage corresponds to a higher available Fe2+ concentration under the electrochemical conditions. Moreover, according to the above results, the k2 value increased as the persulfate concentration was increased from 20 to 30 mM when the Fe2+ concentration was 5 or 10 mM. : This was because persulfate is a source of SO4 − in the system, and so, more of these reactive radicals were generated to degrade toluene at higher persulfate concentrations. The k2 value was 2.33×10−4 L mg−1 min−1 in the absence of Fe2+. : This result indicated that persulfate could generate SO4 − by
Fig. 2 The influence of Fe2+ concentration on the rate constants under electro/Fe2+/persulfate process (C0 =1 mM, [SDS]=2 %, [persulfate]= 20 mM)
Environ Sci Pollut Res
an electron-transfer reaction (Eq. (5)) (Lin et al. 2013; Wu et al. 2012b). − •− 2− S2 O2− 8 þ e →SO4 þ SO4
ð5Þ
Response surface evaluation The full factorial analysis with three factors at three levels is presented in Table 2, along with the k2 values for the respective runs. Based on the data in Table 2, the main effects plot and the interaction plots for k2 values at 30 min were developed. The main effects plot indicated that the effect of the Fe2+ or persulfate concentration on the k2 value was positive, whereas that of the SDS concentration was negative (Fig. SM-1). Thus, a high level (+1) of Fe2+ or persulfate concentration, but a low level (−1) of SDS concentration would lead to a higher k2 value (Wu et al. 2012a; Zhang et al. 2009, 2010). The pretreatment conditions were optimized by employing a Box–Behnken design (BBD) and a polynomial equation describing the k2 value over 30 min as a simultaneous function of SDS concentration (X1), persulfate concentration (X2), and Fe2+ concentration (X3), as shown in Eq. (6): −3
Y ¼ 10
ð1:238−1:850X 1 þ 1:088X 2 þ 1:863X 3
þ
þ
Run order
Full factorial BBD matrix for the rate constants X1
X2
X3
ð7Þ
−0:475X 1 X 2 −0:625X 1 X 3 þ 0:7X 2 X 3
ð6Þ
0:121X 23 Þ
The effects of two-factor interactions were investigated with one factor fixed at a high (+1) or low (−1) level while Table 2
Y ¼ 10−3 ð1:238−1:850X 1 þ 1:088X 2 þ 1:863X 3 þ 0:163X 21 þ 0:121X 23 Þ
−0:475X 1 X 2 −0:625X 1 X 3 þ0:7X 2 X 3 þ 0:163X 21 0:008X 22
the other was varied (Gong et al. 2010; Rastegar et al. 2011). Figure SM-2 shows interaction plots, which indicate whether or not there was an interaction between the factors. Here, the interaction between each pair of variables, namely, SDS, persulfate, and Fe2+ concentrations, was significant, as indicated by the curves crossing. That is to say, a change in one variable would affect the other. This was also confirmed by the high probability value ((Prob>F)>0.1) obtained through analysis of variance (ANOVA). The ANOVA results in Table 3 were used to assess the significance and adequacy of the model. The value of (Prob>F) was used to determine the significance of each model term. Corresponding model terms are more significant with decreasing value of (Prob>F), whereas a value over 0.1 implies insignificance (Zhang et al. 2011a). According to the selected confidence level ((Prob > F) > 0.1), the insignificant model term (X 2 2 ) was removed from the RSM models; Eq. (6) could then be simplified to Eq. (7), and the ultimate RSM models, in terms of the coded factors, were determined to calculate the rate constants.
Second-order rate constant (k2 value)×10−3 Observed
Predicted
1 2 3 4 5 6 7
0 −1 0 0 1 1 1
−1 0 0 1 0 1 −1
1 −1 0 1 −1 0 0
2.3 3.3 1.2 5.9 1.0 1.7 0.5
2.5 3.4 1.2 6.1 0.99 1.6 0.4
8 9 10 11 12 13 14 15
1 0 0 −1 0 −1 −1 0
0 0 0 −1 −1 0 1 1
1 0 0 0 −1 1 0 −1
3.6 1.3 1.2 3.1 0.4 8.7 6.2 1.2
3.5 1.2 1.2 3.2 0.2 8.4 6.3 1.0
The very low (Prob>F) values (F)
Model X1 X2 X3 X1X2
8.3×10−5 2.7×10−5 9.5×10−6 2.8×10−5 9.0×10−7
9 1 1 1 1
9.3×10−6 2.7×10−5 9.5×10−6 2.8×10−5 9.0×10−7
111.82 330.54 114.22 335.03 10.90
RESEARCH ARTICLE
Selective oxidative degradation of toluene for the recovery of surfactant by an electro/Fe2+/persulfate process Anhua Long 1,2 & Hui Zhang 1
Received: 29 October 2014 / Accepted: 18 March 2015 # Springer-Verlag Berlin Heidelberg 2015
Abstract An electro/Fe2+/persulfate process has been conducted for toluene removal from surfactant (SDS) flushing solution, and the pseudo-second-order reaction rate constant (k2 value) of toluene removal has been optimized by a response surface methodology (RSM). The results indicated that in this process, the reaction between persulfate and externally added Fe2+ generates sulfate-free radicals, and at the same time, Fe2+ is electro-regenerated at the cathode by the reduction of Fe3+. RSM based on Box–Behnken design (BBD) has been applied to analyze the experimental variables, of which the concentrations of persulfate and Fe2+ showed a positive effect on the rate constant of toluene removal, whereas the concentration of SDS showed a negative effect. The interactions between pairs of variables proved to be significant, such as between SDS, persulfate, and Fe 2+ concentrations. ANOVA results confirmed that the proposed models were accurate and reliable for analysis of the variables of the electro/Fe2+/persulfate process. The shapes of the 3D response surfaces and contour plots showed that the SDS, persulfate, and Fe2+ concentrations substantially affected the k2 value of toluene removal. The results indicated that increasing persulfate or Fe2+ concentration increased the k2 value, whereas Responsible editor: Angeles Blanco Electronic supplementary material The online version of this article (doi:10.1007/s11356-015-4406-x) contains supplementary material, which is available to authorized users. * Hui Zhang [email protected] 1
Department of Environmental Engineering, Wuhan University, Wuhan 430079, China
2
Jiangxi Science and Technology Normal University, Jiangxi 330013, China
increasing SDS concentration decreased the k2 value. The reaction intermediates have been identified by GC-MS, and a plausible degradation pathway for toluene degradation is proposed.
Keywords Surfactant recovery . Electrochemical treatment . Persulfate . Selective oxidation . Sulfate radical
Introduction Benzene, toluene, ethylbenzene, and xylenes (BTEX) is an abbreviation used for four related compounds found in coal tar, crude petroleum, and a wide range of petroleum products. Among them, toluene is one of the most abundantly produced chemicals, with worldwide annual production of 5–10 million tons (ATSDR 2000). The largest source of toluene release is during the production, transport, and use of petroleum, which contains about 5–8 % toluene. Once released into the environment, toluene usually evaporates into the air. Toluene can also dissolve in water, attach to soil particles, and it may be found in surface and groundwater at contaminated sites or in the close vicinity of natural oil, coal, and gas deposits (Li et al. 2013). The main effect of toluene is on the brain and nervous system, with fatigue and drowsiness being the most obvious symptoms (ATSDR 2000), and it can cause neurological damage (ATSDR 2001). In this context, activated persulfate oxidation has emerged as a novel effective remediation technology for the removal of toluene (Do et al. 2011; Huang et al. 2005; Liang et al. 2008, 2009). This system generates sulfate : radical (SO4 −), a strong oxidant (E0 =2.60 V) that can be thermally or chemically activated by initiators (Shukla et al. 2010; Tan et al. 2013). The activation methods include UV, heat, transition metal ions, alkaline conditions, and persulfate combined with other oxidants. As a natural substrate, ferrous
Environ Sci Pollut Res
ion (Fe2+) has the advantages of being inexpensive and nontoxic, and it has been widely applied as a catalyst to effectively activate persulfate (Jiang et al. 2013; Kusic et al. 2011; Long et al. 2014a; Xu et al. 2014; Zhao et al. 2014). However, some drawbacks are associated with the Fe2+/persulfate process. For instance, high Fe2+ dosage is required to activate persulfate because Fe2+ is difficult to regenerate after conversion to Fe3+ (Vicente et al. 2011). This results in a large amount of iron : sludge. Moreover, the excess Fe2+ could react with SO4 − (Eq. (1)) as a radical scavenger, resulting in reduced efficiency of the process. 3þ þ SO2− Fe2þ þ SO•− 4 →Fe 4
ð1Þ
In order to solve these problems, an electrochemical method has been applied to the reaction system so that Fe2+ could be regenerated according to Eq. (2) (Wu et al. 2012b): þ
Fe3 þ e− → Fe2
þ
Dionysiou 2004; Anipsitakis et al. 2005; Long et al. 2014b; : Zhou et al. 2014). Thus, in the flushing effluents, SO4 − may oxidize toluene at a faster rate than it oxidizes straight-chain aliphatic hydrocarbon surfactants. Indeed, selective oxidation of toluene in SDS flushing effluents was observed in our previous studies (Long et al. 2013, 2014a). Therefore, the objective of the present study was to optimize toluene removal for SDS recovery in flushing effluents from the electro/Fe2+/persulfate process using single-factor tests and response surface methodology (RSM). Three reaction parameters (SDS concentration, persulfate concentration, and Fe2+ concentration) were selected as single factors, and the optimal conditions were obtained by the RSM optimization approach. Furthermore, gas chromatography−mass spectrometry (GCMS) was also applied to determine the intermediates, on the basis of which a degradation pathway of the flushing effluents is proposed.
ð2Þ
Materials and methods In this electro/Fe2+/persulfate process, the cathodic reduction reaction enhances the regeneration of Fe2+, which further : activates persulfate to generate SO4 − (Wang and Chu 2011; 2+ Wu et al. 2012b). As a result, Fe can be continuously regenerated, and the degradation efficiency is improved when an electrochemical potential is applied to the Fe2+/persulfate process. In our previous studies, this process has been successfully applied to the removal of organic pollutants such as Acid Orange 7 (Wu et al. 2012b), bisphenol A (Lin et al. 2013), or clofibric acid (Lin et al. 2014) from water, and even to real wastewater, specifically a landfill leachate (Zhang et al. 2014). It is generally accepted that surfactant-based treatment provides an effective technology for the remediation of toluenecontaminated soil and groundwater (He et al. 2008; Laha et al. 2009; Lee et al. 2002b; Long et al. 2013). However, most surfactants are too expensive for flushing remediation of hydrophobic organic contaminants, and their residues may be biotoxic in the environment (Ying 2006). Usually, sodium dodecyl sulfate (SDS) is used as a surfactant because it is a food grade material and is easily biodegradable by soil and/or aquatic microorganisms (Lee et al. 2002a). Furthermore, the flushing effluents, containing surfactants and toluene, need to be treated before discharge or reuse. Such pollution necessitates investigations of suitable technologies to control flushing effluents and regenerate the surfactant. In a recent review, : SO4 − was proposed as a suitable electrophilic reagent, and when electron-donating groups are present on an aromatic : molecule, the rate of the reaction with SO4 − will increase : (Long et al. 2014b; Tsitonaki et al. 2010). Moreover, SO4 − is relatively stable in aqueous solution, and thus may be able to disperse more freely (Romero et al. 2010; Vicente et al. : 2011). Furthermore, it is generally accepted that SO4 − tends to react selectively via electron transfer (Anipsitakis and
Chemicals and reagents All materials were analytical grade reagents and were used without further purification. Toluene (C6H5CH3) was purchased from Shanghai No. 4 Reagent Factory (China). Sodium persulfate (Na 2 S 2 O 8 ), ferrous sulfate (FeSO 4 · 7H2O), and sodium dodecyl sulfate (SDS, CH 3 (CH 2 ) 11 SO 4 Na) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Apparatus and procedure Batch experiments were performed in an electrolytic reactor (glass beaker) containing 100 mL of solution. A 5 cm× 11.9 cm plate anode (Ti/RuO2-IrO2) and a plate cathode (stainless steel) of the same dimensions were arranged parallel to one another at a distance of 3 cm. The reactor was immersed in a water bath to keep the temperature constant at 20 °C. Electrolysis was conducted under constant current conditions using a direct current (DC) power supply (Model RYI-3010) from Shenzhen Yizhan Instrument Co., Ltd. (China). A magnetic stirrer (Model 78-1, Hangzhou Instrument Motors Factory, China) ensured mixing of the solution in the reactor. Before each run, a fresh stock solution of toluene was prepared with SDS. The initial concentration (C0) was fixed at 1 mM, and the initial pH (background value) was around 7.0. When the DC power supply was initiated, persulfate solution was added as a supporting electrolyte, and the current value was kept at 0.5 A to minimize hydrogen and oxygen evolution, and then, Fe2+ solution was introduced into the electrolytic cell. Aliquots of the mixture were sampled and analyzed at preselected time intervals.
Environ Sci Pollut Res
Analytical methods Direct aqueous injection–gas chromatography/flame ionization detection (DAI-GC/FID) was applied to investigate the concentration of toluene in the surfactant solutions (Kubinec et al. 2005; Long et al. 2014a). Samples were quantified using a gas chromatograph (GC-14C, Shimadzu) with the following settings. GC conditions: oven programmed to a constant temperature of 150 °C (12 min), N2 (99.999 % purity) as carrier gas, and SE-54 (30 m×0.32 mm×0.50 μm) column. FID conditions: temperature set at 250 °C, H2 flow of 60 mL/min, air flow of 400 mL/min, N2 make-up gas, and make-up flow of 20 mL/min. Injector conditions: injection mode with split inlets (split ratio=20:1), injection volume of 1 μL, and injector temperature 250 °C. The intermediate products during the electro/Fe2+/persulfate reaction were detected by headspace–gas chromatography–mass spectrometry (HS-GC-MS) (Esteve-Turrillas et al. 2007). Samples for GC-MS analysis (Shimadzu GCMSQP2010 Plus, Japan) were prepared as follows. A static headspace auto sampler (Thermo Finnigan model HS 2000, Waltham, MA, USA) equipped with standard glass vials of internal volume 10 mL was employed. Once equilibrium between the matrix and the gaseous phase was reached, a 1-mL vapor sample was injected into the system. An HP-5 MS capillary column (30 m length×0.25 mm i.d.×0.25 μm film thickness) was employed for GC separation. The GC equipment was operated in a temperature-programmed mode with an initial temperature of 40 °C held for 2 min, then ramping first to 100 °C at 10 °C min−1 and then further to 200 °C at 20 °C min−1, and finally holding at this temperature for 5 min. The injector and transfer-line temperature was 200 °C. The injector was operated in splitless mode. Helium was used as carrier gas at a flow rate of 5.5 mL/min. The MS-detected scan field was 2–500m/z, and the ion source temperature was 200 °C. The concentration of SDS was determined by the methylene blue method (Kenshi 1975). Methylene blue was first added to the supernatant liquid, then the anionic surfactant– cationic dye complex was extracted with chloroform, and the absorption of the chloroform phase was measured at λmax = 655 nm using a Rayleigh UV-9100 spectrophotometer (Rayleigh Co., China). From this figure, the concentration of SDS could be obtained. Experiment design and data analysis Box–Behnken design (BBD) was applied to investigate the effects of three independent variables, namely, SDS concentration (X1), persulfate concentration (X2), and Fe2+ concentration (X3), on the response function. The low, middle, and high levels of each variable were coded as −1, 0, and +1, respectively, as illustrated in Table 1. The corresponding
Table 1
Experimental range and level of independent variables
Variables
Symbol
SDS concentration (w/v), A Persulfate concentration (mM), B Fe2+ concentration (mM), C
X1 X2 X3
Range and level −1
0
1
0.5 10 5
1 20 10
1.5 30 15
actual values were selected based on available data and preliminary experiments. The dependent variable or objective function was the pseudo-second-order removal rate constant (k2 value). Each independent variable xi was coded as Xi according to the following equation (Singh et al. 2012): X i ¼ ðxi − x0 Þ=Δxi
ð3Þ
where x0 is the value of xi at the center point and Δxi represents the step change value. Thus, a total of 15 experiments based on three levels and three variables was required, including three replicates at the center point in order to estimate the pure error sum of squares. All 15 experiments were conducted in random order. Multiple regression analysis is expressed by the second-order polynomial model as follows (Nair and Ahammed 2014): Y ¼ β0 þ
3 X
βi X i þ
i¼1
3 X i¼1
βii X 2i
þ
2 3 X X
βi j X i X j
ð4Þ
i¼1 j¼iþ1
where Y is the predicted response by the model (k), Xi and Xj are the independent variables, and β0, βi, βii, and βij are the regression coefficients of the model. The validity of the predicted model was verified by analysis of variance (ANOVA), and the second-order model quality was assessed by the determination coefficient (R2). The results were analyzed by a Fischer trial and the probability value (with 95 % confidence level). Finally, the optimal values of the tested variables were obtained by analyzing the surface curves.
Results and discussion Single-factor tests The toluene removal efficiency depends on several main factors, such as the concentrations of SDS, persulfate, and Fe2+ in the electro/Fe2+/persulfate process. Some preliminary runs were performed to identify suitable single factors. Kinetic analysis illustrated that pseudo-second-order kinetics fitted the experimental data better than pseudo-first-order kinetics. As shown in Tables SM-1 and SM-2, the correlation coefficients (R2) for pseudo-second-order kinetics ranged from
Environ Sci Pollut Res
0.9683 to 0.9924, whereas those for pseudo-first-order kinetics were in the range 0.8035–0.9564. When the SDS concentration was increased from 0.5 to 2 %, the toluene removal rate decreased, as indicated in Fig. 1a. The values of rate constant k2 were 11.54×10−3, 8.28×10−3, and 3.02×10−3 L mg−1 min−1 at SDS concentrations of 0.5, 1, and 2 %, respectively. The results indicated that : there were competing reactions of SO4 − with toluene and SDS. A high concentration of SDS probably prevented the activated radicals from reaching the toluene. Moreover, a higher available Fe2+ concentration improved the decomposi: tion of persulfate to generate SO4 −. As can be seen in Fig. 1b, the toluene removal rate increased when the Fe2+ concentration was increased from 2 to 10 mM. The results indicated that the Fe3+ produced in the reaction could be effectively converted to Fe2+ by electro-reduction, further activating persulfate : for SO4 − generation.
Fig. 1 The influence of different parameters on the rate constants under electro/Fe2+/persulfate process: a the effect of Fe2+ concentration (C0 = 1 mM, [persulfate]=30 mM); b the effect of SDS concentration (C0 = 1 mM, [persulfate]=30 mM, [Fe2+]=10 mM)
In order to further investigate the effect of persulfate on the removal of toluene, a persulfate concentration of 20 mM was used. Under the same reaction conditions, the effect of Fe2+ concentration on the removal of toluene was also investigated by conducting experiments at 0, 2, 5, 10, and 20 mM when the SDS concentration was 2 %. As shown in Fig. 2a, b, the k2 value increased from 2.33×10−4 to 23.2×10−4 L mg−1 min−1 as the Fe2+ concentration was increased from 0 to 20 mM. That is to say, increasing the Fe2+ dosage corresponds to a higher available Fe2+ concentration under the electrochemical conditions. Moreover, according to the above results, the k2 value increased as the persulfate concentration was increased from 20 to 30 mM when the Fe2+ concentration was 5 or 10 mM. : This was because persulfate is a source of SO4 − in the system, and so, more of these reactive radicals were generated to degrade toluene at higher persulfate concentrations. The k2 value was 2.33×10−4 L mg−1 min−1 in the absence of Fe2+. : This result indicated that persulfate could generate SO4 − by
Fig. 2 The influence of Fe2+ concentration on the rate constants under electro/Fe2+/persulfate process (C0 =1 mM, [SDS]=2 %, [persulfate]= 20 mM)
Environ Sci Pollut Res
an electron-transfer reaction (Eq. (5)) (Lin et al. 2013; Wu et al. 2012b). − •− 2− S2 O2− 8 þ e →SO4 þ SO4
ð5Þ
Response surface evaluation The full factorial analysis with three factors at three levels is presented in Table 2, along with the k2 values for the respective runs. Based on the data in Table 2, the main effects plot and the interaction plots for k2 values at 30 min were developed. The main effects plot indicated that the effect of the Fe2+ or persulfate concentration on the k2 value was positive, whereas that of the SDS concentration was negative (Fig. SM-1). Thus, a high level (+1) of Fe2+ or persulfate concentration, but a low level (−1) of SDS concentration would lead to a higher k2 value (Wu et al. 2012a; Zhang et al. 2009, 2010). The pretreatment conditions were optimized by employing a Box–Behnken design (BBD) and a polynomial equation describing the k2 value over 30 min as a simultaneous function of SDS concentration (X1), persulfate concentration (X2), and Fe2+ concentration (X3), as shown in Eq. (6): −3
Y ¼ 10
ð1:238−1:850X 1 þ 1:088X 2 þ 1:863X 3
þ
þ
Run order
Full factorial BBD matrix for the rate constants X1
X2
X3
ð7Þ
−0:475X 1 X 2 −0:625X 1 X 3 þ 0:7X 2 X 3
ð6Þ
0:121X 23 Þ
The effects of two-factor interactions were investigated with one factor fixed at a high (+1) or low (−1) level while Table 2
Y ¼ 10−3 ð1:238−1:850X 1 þ 1:088X 2 þ 1:863X 3 þ 0:163X 21 þ 0:121X 23 Þ
−0:475X 1 X 2 −0:625X 1 X 3 þ0:7X 2 X 3 þ 0:163X 21 0:008X 22
the other was varied (Gong et al. 2010; Rastegar et al. 2011). Figure SM-2 shows interaction plots, which indicate whether or not there was an interaction between the factors. Here, the interaction between each pair of variables, namely, SDS, persulfate, and Fe2+ concentrations, was significant, as indicated by the curves crossing. That is to say, a change in one variable would affect the other. This was also confirmed by the high probability value ((Prob>F)>0.1) obtained through analysis of variance (ANOVA). The ANOVA results in Table 3 were used to assess the significance and adequacy of the model. The value of (Prob>F) was used to determine the significance of each model term. Corresponding model terms are more significant with decreasing value of (Prob>F), whereas a value over 0.1 implies insignificance (Zhang et al. 2011a). According to the selected confidence level ((Prob > F) > 0.1), the insignificant model term (X 2 2 ) was removed from the RSM models; Eq. (6) could then be simplified to Eq. (7), and the ultimate RSM models, in terms of the coded factors, were determined to calculate the rate constants.
Second-order rate constant (k2 value)×10−3 Observed
Predicted
1 2 3 4 5 6 7
0 −1 0 0 1 1 1
−1 0 0 1 0 1 −1
1 −1 0 1 −1 0 0
2.3 3.3 1.2 5.9 1.0 1.7 0.5
2.5 3.4 1.2 6.1 0.99 1.6 0.4
8 9 10 11 12 13 14 15
1 0 0 −1 0 −1 −1 0
0 0 0 −1 −1 0 1 1
1 0 0 0 −1 1 0 −1
3.6 1.3 1.2 3.1 0.4 8.7 6.2 1.2
3.5 1.2 1.2 3.2 0.2 8.4 6.3 1.0
The very low (Prob>F) values (F)
Model X1 X2 X3 X1X2
8.3×10−5 2.7×10−5 9.5×10−6 2.8×10−5 9.0×10−7
9 1 1 1 1
9.3×10−6 2.7×10−5 9.5×10−6 2.8×10−5 9.0×10−7
111.82 330.54 114.22 335.03 10.90
