Coke composite as a heterogeneous Fenton-like catalyst

STOTEN-20654; No of Pages 9 Science of the Total Environment xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst Dong Wan, Wenbing Li ⁎, Guanghua Wang, Lulu Lu, Xiaobi Wei School of Chemistry and Chemical Engineering, Wuhan University of Science and Technology, Wuhan 430081, China

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Fe0/Fe3O4/Coke composite was synthesized and used as the heterogeneous Fenton catalyst. • Response surface methodology was used to study the degradation of p-Nitrophenol in Fe0/Fe3O4/Coke-H2O2 system. • A possible degradation pathway of pNitrophenol in Fe0/Fe3O4/Coke-H2O2 process was proposed. • The obtained Fe0/Fe3O4/Coke composite was low-cost and environment-friendly and showed good stability for repeated use.

a r t i c l e

i n f o

Article history: Received 5 July 2016 Received in revised form 2 August 2016 Accepted 6 August 2016 Available online xxxx Editor: Jay Gan Keywords: Magnetic composite Coke p-Nitrophenol Heterogeneous degradation Response surface methodology

a b s t r a c t A Coke supported Fe3O4 and Fe0 composite (Fe0/Fe3O4/Coke) was prepared for the first time with the aim of evaluating its ability to be used as heterogeneous catalyst for the Fenton degradation of p-Nitrophenol (p-NP). A four factor Box–Behnken design (BBD) coupled with response surface methodology (RSM) was applied to evaluate the effects of several operating parameters, namely Fe0/Fe3O4/Coke dosage, reaction temperature, initial pH and H2O2 concentration, on the removal efficiency of p-NP. A significant quadratic model (p-value b 0.0001, R2 = 0.9952) was derived using analysis of variance (ANOVA). Optimum conditions were determined to be 1.3 g/L catalyst, 32 °C, pH 3.1 and 11.3 mM H2O2. 100% of p-NP (100 mg/L) conversion and 81% of COD removal were achieved after 120 min of reaction time, respectively, under the optimum conditions, which agreed well with the modeling prediction. The recyclability of Fe0/Fe3O4/Coke was also investigated after three successive runs, in which p-NP degradation performances showed a slight difference with the first oxidation cycle with an acceptable iron leaching. Moreover, according to the main intermediate products identified by gas chromatography–mass spectrometry (GC–MS), a possible pathway of p-NP degradation was proposed based on hydrogen radicals ([H]) or hydroxyl radicals (•OH) mechanism. © 2016 Elsevier B.V. All rights reserved.

1. Introduction ⁎ Corresponding author. E-mail address: [email protected] (W. Li).

As an important organic chemical, p-Nitrophenol (p-NP) is widely employed as raw materials or intermediates in the preparation of

http://dx.doi.org/10.1016/j.scitotenv.2016.08.042 0048-9697/© 2016 Elsevier B.V. All rights reserved.

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042

2

D. Wan et al. / Science of the Total Environment xxx (2016) xxx–xxx

explosives, preservatives and dyes (Bae et al., 2016; Chen et al., 2016; Pang and Lei, 2016). However, p-NP has been listed as a priority pollutant by the US Environmental Protection Agency due to its high toxicity, potential carcinogenicity and non-biodegradation (Sun and Lemley, 2011). Therefore, an efficient treatment of effluents containing p-NP is essential before their final discharge into the aquatic environment. More recently, a number of methods have been used to address p-NP contamination in water, including adsorption (Liu et al., 2016), Fenton-like (Sun and Lemley, 2011; Zhang et al., 2012), microwave catalysis (Pang and Lei, 2016; Zhou et al., 2016), thermally activated persulfate (Chen et al., 2016), photocatalytic (Li et al., 2016), electrochemical process (Xie et al., 2016) and ozonation (Kuosa et al., 2015). Among these methods, heterogeneous Fenton-like reactions have shown great potential. The process operates at near ambient temperature and pressure with a relatively simple operation. In the presence of H2O2 and heterogeneous catalyst, hydroxyl radical (•OH) was generated as oxidant, which could oxidize nonselectively most of the organic molecules to CO2, H2O and harmless products, avoiding secondary pollution (Nidheesh, 2015; Shahidi et al., 2015). Thus, in order to enhance the activation of H2O2, the development of more active and stable heterogeneous catalysts has been gaining importance. In the last few years, the development of magnetite (Fe3O4)-based catalysts has attracted significant attention, which provides an interesting and easy separation method upon the application of a magnetic field, simplifying its recovery and reusability (Ambashta and Sillanpää, 2010; Munoz et al., 2015). On the other hand, Fe3O4 nanoparticles were found to possess intrinsic peroxidase-like activity, allowing the magnetite be an efficient heterogeneous catalyst for Fenton oxidation (Gao et al., 2007; Zhang et al., 2008; Zhang et al., 2009; Xu and Wang, 2012). So far, carbon material has been the most studied support used for the preparation of Fe3O4-based catalysts including carbon nanotube (Hu et al., 2011; Yu et al., 2015), graphene oxide (Zubir et al., 2014), activated carbon (Kakavandi et al., 2016) and other carbon material (Chun et al., 2012; Zhang et al., 2013). However, the synthesis of these materials could be quite tedious in which required hazardous chemicals or harsh reaction conditions and the cost of these materials may not reasonable in practical application. Therefore, the use of an environmentally friendly and low cost material is important to make the heterogeneous Fenton-like treatment more affordable as a general decontamination technology. Coal Coke, a relatively inexpensive and abundant feedstock, was usually used as fuel or waste materials in many coking plants (Kraines et al., 2002). The potential utilization value of Coke, unfortunately, has not been exploited efficiently. As a kind of carbon materials, Coke also has a lot of excellent properties such as good adsorption, strong stability and, especially, low cost (Zubot et al., 2012; Niasar et al., 2016). What's more, the Coke could exhibit graphene-like characteristics due to its good surface properties which consists of fused six-membered polyaromatic ring systems (Wauters and Marin, 2001; Catak et al., 2011; Sierra et al., 2016), allowing the Coke may be a suitable material for the practical wastewater treatment. However, there have been few attempts to investigate the coal Coke been used in heterogeneous Fenton system. In previous study (Lu et al., 2016), Fe3O4/Coke composite was successfully synthesized and investigated, which revealed the good performance of Coke for the immobilization of Fe3O4 nanoparticles. In addition, recent works demonstrated that bringing Fe0 particles into contact with Fe3O4 particles increases its activity for H2O2 decomposition (Moura et al., 2005; Hou et al., 2014). On the other hand, the existence of Fe0 in Fe3O4/Coke may lead to the formation of Fe–C micro-electrolysis system, in which Fe2+ and H⁎ with high chemical activity could be generated, promoting the degradation and removal of organic compounds (Zhu et al., 2014; Lin et al., 2016). Herein, Fe0/Fe3O4/Coke composite has been prepared and characterized in the present work. For the first time, the Fe0/Fe3O4/Coke composite has been tested as a heterogeneous catalyst for the Fenton-like degradation of p-NP. The influence of parameters in the Fenton-like

removal of p-NP was studied in detail using Response Surface Methodology (RSM) based on the Box–Behnken design. Furthermore, H2O2 decomposition and iron leaching during the reaction process as well as the mechanism of p-NP degradation and the reusability of the catalyst were investigated. 2. Experimental 2.1. Chemicals and materials The p-Nitrophenol used as the substrate was acquired from Sinopharm Chemical Reagent Co., Ltd (China). The Coke breeze, purchased from coking plant of Wuhan Iron and Steel Group, was ground and sieved to 200 mesh size without further purification. The Proximate Analysis and Elemental Analysis of the Coke are given in Table S1 and S2 in Supplementary data. The surface area of the Coke was 3.73 m2/g (Fig. S1). All chemicals were of analytical grade and used as received. All aqueous solutions were prepared with distilled water. 2.2. Preparation of Fe0/Fe3O4/Coke composite Fe3O4/Coke with Fe3O4: Coke mass ratio of 2:1 was firstly synthesized by in situ oxidation precipitation method according to the previous work (Lu et al., 2016). The Fe0/Fe3O4/Coke composite was prepared by borohydride reduction at a Fe0:Fe3O4:Coke mass ratio of 3:4:2, using FeCl2·4H2O as iron source and NaBH4 as reductant. After ground and sieved to 200 mesh size, 2 g Fe3O4/Coke powders was added to an ethanol/water mixture (100 mL ethanol + 25 mL distilled water) containing 3.6 g FeCl2·4H2O. The mixture was subsequently left in an ultrasonic shaker for 30 min. Meanwhile, 2 g NaBH4 was dissolved in 50 mL distilled water to form 1 M solution. Afterwards, the borohydride solution was added dropwise to the above mixture and further stirred for 30 min. The formed Fe0/Fe3O4/Coke solid were collected under the assistance of a magnet and washed at least three times with absolute ethanol, and then dried in a vacuum oven at 80 °C for 12 h. Fe3O4 and Fe0 were synthesized as above procedure without adding Coke. 2.3. Characterization The structure and morphology of the samples were characterized by X-ray powder diffractometer (XRD, Rigaku D/max-RB, Japan) with Cu Kα radiation and scanning electron microscopy (SEM, FEI Nova400, USA). X-ray Photoelectron Spectroscopy (XPS) was performed with a MULTILAB2000 X-ray photoelectron spectrometer (VG, USA). Curve fittings were achieved with the XPSPeak 4.1 program. Magnetic characterization of the samples were performed using a JDAW-2000D vibrating sample magnetometer (VSM) at room temperature. 2.4. Heterogeneous Fenton-like reaction The degradation of p-NP was conducted in a conical flask (200 mL of solution containing 100 mg/L p-NP) with continuously vigorous stirring in the dark. After the addition of determined Fe0/Fe3O4/Coke, the initial pH of the reaction suspension was adjusted by H2SO4. The degradation reaction was initiated by adding a certain amount of H2O2 to the solution. At selected time intervals, 5 mL aliquots were sampled, clarified by a strong outer permanent magnet prior to analysis. All experiments were repeated at least two times, and the results are presented as average. 2.5. Experimental design and data analysis by RSM RSM based on the Box–Behnken design (BBD) was applied to optimize the experimental conditions for p-NP degradation. The independent variables were Fe0/Fe3O4/Coke dosage (A), reaction temperature

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042

D. Wan et al. / Science of the Total Environment xxx (2016) xxx–xxx

(B), initial pH (C) and H2O2 concentration (D). Based on the preliminary assays, the ranges of four parameters chosen in this study are summarized in Table S3. The response analyzed was p-NP removal (Y) after 60 min of reaction, which was measured by the reduction of C/C0. The experimental data were analyzed with the help of Design Expert 8.06 software and were fitted with a second-order polynomial equation. A total number of 29 experiments were employed for response surface modeling, as showed in Table S4. The optimized parameters for maximizing the p-NP removal were also obtained by using the established equation. 2.6. Analytical methods The p-NP concentration was determined using UV–Vis spectroscopy (Ultrospec 3300 pro, GE Healthcare Bio-Sciences China Ltd) at a wavelength of 400 nm (pH N 11) (Wang et al., 2014). The intermediate products after p-NP degradation was identified by a GC–MS system (Agilent GC7890A-MSD5975C, USA). The chemical oxygen demand (COD) of the p-NP solution was measured by dichromate method (Neamţu et al., 2004). The concentration of iron leaching in the solution was measured by the 1,10-phenatroline spectrophotometric method (Chen et al., 2011). Hydrogen peroxide was analyzed spectrophotometrically with titanium oxalate (Sellers, 1980). 3. Results and discussion 3.1. Characterization of Fe0/Fe3O4/Coke X-ray diffraction (XRD) was conducted to identify the structures and crystallinity of as-prepared products. As showed in Fig. 1, diffraction peaks assigned to Coke at 2θ = 26° can be seen for pure Coke and the nanocomposites with no sharp diffraction peak appears, indicating that the Coke has a similar structure with activated carbon (Kakavandi et al., 2016). The presence of metallic iron is indicated by the major reflection at 2θ = 44.7° (Costa et al., 2008; Hou et al., 2014). The crystal structure of Fe3O4 nanoparticles was confirmed by the presence of diffraction peaks at 18.3, 30.1, 35.5, 43.1, 53.5, 57.0 and 62.6° according to the standard JCPDS data (card No. 01-1111) (Hou et al., 2014). Similar XRD patterns were also noted on Fe3O4/Coke and Fe0/Fe3O4/Coke nanocomposites, which confirmed the successful deposition of Fe3O4

3

nanoparticles onto the Coke surface. The reduction of the peak at 2θ = 44.7° in the reused Fe0/Fe3O4/Coke may due to the reaction between Fe0 and Fe3O4, which will be discussed in the sections below. The morphology of the as-prepared products was further studied with SEM. It can be seen from Fig. 2a that Coke shows a large particle and relatively smooth nonporous surface. The approximated size of the Fe3O4 and Fe0 nanoparticles are around 100 nm and 20 nm, respectively (Fig. 2b and c). SEM image in Fig. 2d showed the spherical-like Fe3O4 and Fe0 nanoparticles growing on the Coke surface regularly with well distributed and little agglomeration. The samples have been repeated washing in water and methanol under ultrasonication, but almost all magnetic nanoparticles were still found on the Coke surface, which indicated the strong interaction between Coke and magnetic nanoparticles. The EDS spectra of Fe0/Fe3O4/Coke is illustrated in Fig. 2f. It can be found that the nanocomposite is mainly consists of C, O, Al, Si, S, Ca and Fe, which is in conformity with the Elemental Analysis of the Coke. XPS was then employed to further verify the phase of the prepared nanocomposites. As showed in Fig. 3a, the binding energies (BE) of about 284, 530 and 711 eV indexed to C 1s, O 1s and Fe 2p, respectively (Hu et al., 2011). Fig. 3b shows the high resolution Fe 2p scan, as indicated by the vertical marks, the Fe 2p peaks at BE of 711.2 and 725.0 eV are characteristic of Fe3+, while the peaks at BE of 709.9 and 723.4 eV are the characteristic of Fe2+ (Xu and Wang, 2012; Hu et al., 2011). Similar peak patterns were attained with Fe3O4 and Fe0/Fe3O4/Coke, indicating the formation of Fe3O4 phase in the nanocomposites. Besides, it is obviously in the Fe 2p scan of Fe0/Fe3O4/Coke that a higher BE shoulder was fitted around 713.4 eV, which may be related to an interaction between Fe0 and Fe3O4, being similar to the reported interaction between Fe core and Fe2O3 shell (Lu et al., 2007). The O 1s spectra of Fe3O4 and Fe0/ Fe3O4/Coke were shown in Fig. 3c. The photoelectron line of Fe0/ Fe3O4/Coke at BE of about 531.7 and 533.2 eV were attributed to oxygen in carboxylate and/or carbonyl (O\\C_O; C_O) and in the epoxy and/ or hydroxyl (C\\O\\C; C\\OH) of the Coke, respectively (Hu et al., 2011; Zubir et al., 2014). The peak around 529.6 eV is characteristic of the O 1s belonging to lattice oxygen in Fe3O4 (Fe\\O) (Geng et al., 2012), which was found in the photoelectron line of both Fe3O4 and Fe0/Fe3O4/Coke. The new peak at about 530.6 eV in the Fe0/Fe3O4/ Coke spectra is assigned to the deposition of Fe3O4 onto Coke, which is possibly via Fe\\O\\C bonds (Zubir et al., 2014). Similar findings were also reported that the formation of metal\\O\\C bonds in nanocomposites could be shifted to higher binding energy from the metal\\O bonds of metal oxides (Zubir et al., 2014; Zhou et al., 2011). The C 1s spectra of Fe0/Fe3O4/Coke is shown in Fig. 3d, which can be deconvoluted into five components corresponding to C_C sp2 (284.6 eV), C\\C sp3 (285.1 eV), C\\OH/C\\O\\C (286.7 eV) and O\\C_O (289.0 eV) (Variava et al., 2012). Magnetic properties of the synthesized Fe3O4, Fe3O4/Coke, Fe0 and 0 Fe /Fe3O4/Coke were measured (Fig. S2). As can be viewed, the saturation magnetization (Ms) of Fe3O4/Coke (62.6 emu/g) is much smaller than that of the unmodified Fe3O4 (82.9 emu/g), which may result from the existence of nonmagnetic Coke. However, the Ms value of Fe0/Fe3O4/Coke reached up to 99.3 emu/g, which could attribute to the formation of Fe0 (Ms = 158.6 emu/g). Moreover, room-temperature magnetic response of the Fe0/Fe3O4/Coke nanocomposite was also examined by an external magnet (Fig. S2 inset). The result shows a potential advantage for the separation, recovery and reuse of the synthesized Fe0/Fe3O4/Coke nanocomposite. 3.2. Degradation of p-NP by various processes

Fig. 1. XRD patterns of the samples.

The performance of p-NP degradation by different processes (Fig. 4a) was evaluated under identical experimental conditions (100 mg/L p-NP, temperature 30 °C and pH 3.0). As can be seen, without H2O2, the decrease of the p-NP concentration was fast in the first 15 min and then reached a steady state of 12.5% removal, which may due to

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042

4

D. Wan et al. / Science of the Total Environment xxx (2016) xxx–xxx

Fig. 2. SEM photographs of Coke (a), Fe3O4 (b), Fe0 (c), Fe0/Fe3O4/Coke (d) and reused Fe0/Fe3O4/Coke (e) and EDS spectra of Fe0/Fe3O4/Coke (f).

the adsorption of p-NP on the surface of the Fe0/Fe3O4/Coke catalyst, corresponding to the adsorption capacity of the Fe0/Fe3O4/Coke sample for p-NP (Fig. S3, 14.6 mg/g). Depending on Fig. 4a, application of H2O2 alone led to a slight p-NP removal of 6.7% within 120 min, indicating that the oxidation ability of H2O2 is limited. In the presence of both catalyst and H2O2, the p-NP conversion reached 73.4% (Fe3O4–H2O2) and 92.1% (Fe3O4/Coke–H2O2) after a 120 min reaction, respectively. This result was attributed to the peroxidase-like activity of Fe3O4 nanoparticles, which could produce highly active hydroxyl radicals (•OH) as described in Eqs. (1) and (2) (Munoz et al., 2015; Xu and Wang, 2012), resulting in the degradation of p-NP. The more efficient performance of Fe3O4/ Coke than bare Fe3O4 may owing to the better dispersity of Fe3O4 on Coke, which makes more active sites of Fe3O4 exposing. In similar experimental conditions, the catalytic ability of Fe0/Fe3O4/Coke composite in presence of H2O2 showed a significant p-NP reduction, which was almost completely removed within only 60 min, indicating the high catalytic ability toward H2O2 activation for Fe0/Fe3O4/Coke. We also investigated the H2O2 decomposition with time under different conditions, as shown in Fig. 4b. It can be observed from Fig. 4b that H2O2 was not converted in the absence of catalyst. However, by using the Fe3O4, Fe3O4/Coke and Fe0/Fe3O4/Coke catalysts, we observed that 42%, 49% and 63% H2O2 were decomposed in the entire experiment, respectively. Taking the efficiencies of p-NP degradation into consideration, it can be concluded that the H2O2-activating efficiency was the main evaluation ability for the heterogeneous Fenton catalyst. According to previous studies, the combination of Fe0 and Fe3O4 could make a significant increase for H2O2 decomposition (Moura et al., 2005; Costa et al., 2008). The Fe0 could act as an electron transfer agent in

Fe 0/Fe 3 O4 system, which could easily reduce Fe 3 + species in the Fe 3O 4 phase to regenerate the active Fe 2 + species, as seen in Eq. (3) (Moura et al., 2005). On the other hand, there may be a Fe–C micro-electrolysis system formed between Fe0 and Coke, prompting the formation of free hydrogen radicals [H], as indicated in Eqs. (4)–(5) (Lin et al., 2016). The organics with higher redox potential could be also decomposed into the smaller molecules by [H]. Thus, the H2O2-activating ability of Fe0/Fe3O4/Coke composite was strongly improved, resulting in a significant increase in activity for the oxidation of organics. ≡ Fe2þ þ H2 O2 → ≡ Fe3þ þ HO− þ ⋅ OH

ð1Þ

≡ Fe3þ þ H2 O2 → ≡ Fe2þ þ ⋅ O2 H þ H2 O

ð2Þ

Fe0 þ `Fe3þ →Fe2þ þ `Fe2þ

ð3Þ

Fe−2e− →Fe2þ

ð4Þ

2Hþ þ 2e− →2½H→H2

ð5Þ

p−NP þ ⋅ OH=½H→products

ð6Þ

3.3. Regression model and analysis of variance The full factorial BBD with four factors in three levels along with the observed response are listed in the Table S4. A total of 29 batch runs

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042

D. Wan et al. / Science of the Total Environment xxx (2016) xxx–xxx

5

Fig. 3. Wide scan XPS spectra (a) and XPS curve fit of Fe 2p (b), O 1 s (c) and C1 s (d) spectra of Fe0/Fe3O4/Coke and Fe3O4.

regarding the BBD-designed experiments with five replicates at central point were investigated to evaluate the effects of the independent factors on the responses. As presented in Table S4, the observed removal efficiencies of p-NP with 60 min varied between 78.23 and 99.32%. On the basis of RSM results, a second-order polynomial equation was

obtained for the process of p-NP removal as shown below: Y ð%Þ ¼ 98:31 þ 0:41A þ 0:51B−9:00C þ 0:98D þ 1:20AB þ 0:82AC þ 1:21AD þ 0:053BC þ 2:17BD−0:53CD−0:40A2 −1:06B2 −7:63C 2 −2:66D2 ð7Þ

Fig. 4. Removal of p-NP (a) and decomposition of H2O2 (b) along time by various processes (C0 = 100 mg/L, [catalyst] = 1 g/L, [H2O2] = 10 mM, T = 30 °C, pH = 3.0).

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042

6

D. Wan et al. / Science of the Total Environment xxx (2016) xxx–xxx

Analysis of variance (ANOVA) was performed to estimate the goodness of the fit of the equation, and the results are summarized in Table S5. The very low p-value (b0.0001) obtained in this work fully confirms that the model is significant. The determination coefficients R2 of the quadratic regression model were obtained as 0.9952 for p-NP removal, which indicated that the equation are highly reliable. The corresponding adjusted R2 values were 0.9904 for this model, which ensured a satisfactory adjustment of the quadratic models to the experimental data. In addition, the low value of the coefficient of variation (C.V. = 0.75%) suggested that variation in the mean value was acceptable and satisfactory in this work (Li et al., 2015). Therefore, the response surface models were accurately employed for predicting the relationships between these four parameters and the response. Fig. S4a presents the normal probability plot of the studentised residuals for the p-NP removal process. As can be viewed, the points are fairly close to a straight line with no deviation of variance, indicating that the residuals follow a normal distribution (Li et al., 2015). Fig. S4b shows the comparison between the predicted values (obtained according to Eq. (7)) of p-NP removal and the experimental points, which are distributed relatively near to the regression line, further suggesting that the model is reliable to describe the heterogeneous Fenton-like degradation of p-NP. The significance of each independent variable was also evaluated according to its p-value (significant at p b 0.05, not significant at p N 0.10). As can be seen in Table S5, factors C, BD, C2 and D2 had the highest impact on the process of p-NP removal. Besides, except for the factors A, BC, CD and A2, other factors were highly significant. 3.4. Three-dimensional (3D) surface plots and optimization analysis To gain the better comprehensive of p-NP removal process, the three dimensional response surface plots were analyzed, as showed in Fig. S5. It can be seen that the initial pH appears to be the most important variable governing p-NP degradation. Within the pH ranged from 4.0 to 2.4, the p-NP removal was significantly improved, which can be attributed to the higher oxidation potential of •OH and the stability of H2O2 in acidic solution (Zhang et al., 2008; Hu et al., 2011). According to Eq. (5), factors A (catalyst loading), B (reaction temperature) and D (H2O2 concentration) had positive effects on the p-NP degradation process. Based on Eqs. (1) and (2), the amount of •OH is dependent on the catalyst loading and H2O2 concentration. In other words, higher catalyst loading and H2O2 concentration could increase the amount of active sites on the catalyst surface for H2O2 activation, resulting in more •OH produced. However, with further increasing the catalyst loading and H2O2 concentration, the degradation of p-NP does not enhance but slightly decrease (Fig. S5), which can be explained by the scavenging effect of •OH by iron species and H2O2 through undesirable reactions (8) and (9) (Hu et al., 2011). Similar result was observed for the variable of reaction temperature. The reaction rate between H2O2 and the catalyst is increased with the increase of reaction temperature, while high temperature could accelerate the decomposition of H2O2 to H2O and O2 (Xu and Wang, 2012). Fe2þ þ ∙ OH→OH– þ Fe3þ

ð8Þ

H2 O2 þ ∙ OH→H2 O þ ∙ OOH

ð9Þ

Optimization process was carried out to determine the optimum value of p-NP removal using Design Expert 8.06 software along with a multiple-response method. According to the model prediction, the optimum conditions for the degradation of p-NP by heterogeneous Fenton reactions on Fe0/Fe3O4/Coke were determined to be 1.3 g/L catalyst, 32 °C, pH 3.1 and 11.3 mM H2O2, with the predicted p-NP removal of 99.42%. In order to verify the validity of the model for predicting the p-NP removal, five additional experiments were performed under the above conditions. The mean p-NP removal was determined to be 99.37%. Therefore, it can be concluded that the p-NP removal for the

response of parameters obtained from the experiments is in close agreement with the model prediction. It also confirms that the strategy to optimize the operational conditions of p-NP removal in this study is successful. 3.5. Iron leaching and H2O2 decomposition The concentrations of H2O2 and dissolved iron during p-NP degradation were investigated under the optimum reaction conditions (1.3 g/L catalyst, 32 °C, pH 3.1 and 11.3 mM H2O2), which was illustrated in Fig. 5a. As can be viewed, the concentration of total dissolved iron increased with the reaction time, and about 6.7 mg/L of the iron dissolved into solution after 120 min. In our previous study (Lu et al., 2016), the total dissolved iron was about 2.4 mg/L using Fe3O4/Coke as the catalyst. The increased dissolved iron in the new system may be explained by the continuous leaching of iron from Fe0 and Fe3O4. It can be clearly found that the most dissolved iron was Fe2+, which was different with the similar reaction in previous reports that Fe2+ contributed a very small percentage to the dissolved iron (Xu and Wang, 2012; Xu and Wang, 2015). The high Fe2+ concentration in this system may be coursed by the reduction of Fe3+ into Fe2+ by Fe0, as described in Section 3.2. The concentration of Fe2+ increased rapidly in the first 60 min, which was close to the time of complete removal of p-NP. However, a descending period of Fe2+ concentration occurred after 90 min, which could be attributed to the remaining H2O2 and •OH that oxidized the dissolved ferrous ions to ferric ions. It can be observed in the Fig. 6a that the concentration of H2O2 decreased rapidly in the first 60 min then gradually from 60 min to 120 min. The decomposition rate of H2O2 was mainly depend on ferrous ions, as expressed in Eq. (1). Owing to the existence of Fe0, the Fe2+ concentration was relatively high in our system. It is the high Fe2+ concentration that improved the decomposition of H2O2, thus promoting the degradation of p-NP. In order to investigate the utilization efficiency of H2O2 in our system, COD of the p-NP solution during the reaction was measured. As observed in Fig. 5b, about 81% COD removal was achieved after 120 min.

Fig. 5. (a) H2O2 decomposition, iron dissolution, (b) p-NP conversion and COD removal during p-NP degradation under the optimum reaction conditions.

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042

D. Wan et al. / Science of the Total Environment xxx (2016) xxx–xxx

Defining the efficiency of H2O2 (E) as the amount COD converted per unit mass of H2O2 decomposed: Eð%Þ ¼

ΔCOD  100 0:47  Δ½H2 O2 

ð10Þ

where ΔCOD is the removable COD value (mg/L), Δ[H2O2] is the decomposed H2O2 value (mg/L), and 0.47 is the conversion factor (Gao et al., 2015). The amount of the consumed H2O2 was about 8.2 mM during 120 min reaction (Fig. 5a). Therefore, the utilization efficiency of H2O2 was calculated as 96% in our system. Such high utilization efficiency of H2O2 was pretty good in comparison to the studies under similar conditions reported before (Xu and Wang, 2015; Gao et al., 2015). 3.6. Possible degradation mechanism of p-NP The p-NP could be removed efficiently under the optimum reaction conditions. However, it can be observed from Fig. 5b that the COD removal of p-NP is much slower than its decomposition. So it could be inferred that some achromatic intermediate products have not been fully degraded. The UV–Vis spectra recorded during the degradation of p-NP was shown in Fig. 6. The characteristic absorption peak at 400 nm was primarily due to the formation of p-nitrophenolate ions in alkaline condition (Bae et al., 2016). As can be seen, the intensity of the 400 nm absorption characteristic peak decreases rapidly with the reaction proceeding, which indicated that the \\NO2 group on the molecular structure of p-NP was reduced effectively. At the same time, two new peaks were appeared at the wavelength of 227 nm and 319 nm with the increasing of the reaction time, which can be ascribed to the absorption characteristic of p-aminophenol and some intermediate products (Tan et al., 2015). The new peaks were also decreased with increasing time, demonstrating that the new products generated could continue to be degraded in this process. However, the absorption peaks in the ultraviolet region did not completely disappear, related to the uncompleted destruction of some aromatic group, which well corresponds to the COD result. To further identify the intermediate products during the degradation process (after degradation for 30 min and 120 min), GC–MS analysis was employed and the corresponding molecules were summarized in Fig. S6. As can been observed from Fig. S6(a), the major intermediate products were: p-benzoquinone (11.304 min), hydroquinone (12.082 min), p-aminophenol (22.149 min) and p-NP (23.035 min). Different from some works on the degradation of p-NP by other kinds of advanced oxidation processes, p-aminophenol was detected in our

Fig. 6. UV–Vis spectra of p-NP as a function of reaction time under the optimum reaction conditions.

7

system. Most studies on the degradation pathways of the p-NP generally considered that the hydroxylation of p-NP turned into hydroquinone, which was further oxidized to benzoquinone (Sun and Lemley, 2011; Xie et al., 2016; Kuosa et al., 2015). However, due to the formation of free hydrogen radicals [H] (Eqs. (4)–(5)) in our system, p-NP was likely reduced to p-aminophenol, which was in agreement with the result of UV–Vis spectra. After the aromatic rings being broken, some esters and alcohols were generated such as ethyl 3-hydroxybutyrate (3.512 min), n-butyl acetate (4.536 min) and methyl acetate (5.464 min), which was concluded in Fig. S6(b). Based on the identified major aromatic intermediates, the possible degradation scheme of p-NP by Fe0/Fe3O4/Coke-H2O2 process was proposed, as indicated in Fig. 7. Firstly, •OH or [H] was generated on the surface of Fe3O4 or Fe0. On one hand, p-NP was reduced to p-aminophenol by direct reduction on the surface of catalyst or indirect reduction of [H], then further oxidized to p-benzoquinone by •OH (Tang et al., 2015). On the other hand, p-NP was oxidized to hydroquinone and then p-benzoquinone by •OH (Sun and Lemley, 2011). Secondly, the p-benzoquinone was further oxidized by •OH to form small molecule intermediates, which could be decomposed into CO2 and H2O. Therefore, p-NP was effectively degraded in Fe0/Fe3O4/Coke-H2O2 process.

3.7. Recyclability of Fe0/Fe3O4/Coke composite The potential to reuse Fe0/Fe3O4/Coke composite was tested as seen in Fig. S7. The removal of p-NP gradually decreased by repeated reuse of Fe0 /Fe3 O 4/Coke under the optimum reaction conditions, which could be explained by the leaching of iron from the catalyst surface detected in Section 3.5. However, the p-NP conversion still remained high after 120 min reaction, which indicated that satisfying catalytic activity was maintained after three runs. Furthermore, the homogeneous catalytic experiment was performed with a similar content of Fe 2 + and Fe3 + (Fig. 6, 4.0 mg/L Fe2 + and 2.7 mg/L Fe 3 +). Limited p-NP conversion (28.3%) was obtained, indicating that the homogeneous reaction just makes a very small contribution to the p-NP degradation. In addition, XRD patterns (Fig. 1) and SEM images (Fig. 2e) of the reused Fe0 /Fe3 O 4/Coke composite showed that the morphology, dimension and structure of the composite remained unchanged. This indicates that the obtained Fe 0/Fe 3O 4 / Coke composite was stable and exhibited excellent reusability in the catalysis process.

4. Conclusion A Coke supported Fe3O4 and Fe0 composite (Fe0/Fe3O4/Coke) was prepared and used for the first time as heterogeneous Fenton catalyst for the degradation of p-Nitrophenol. Compared with some similar magnetic catalysts reported previously, such as Fe3O4-MWCNTs and graphene oxide–Fe3O4, the present novel composite was much more economical with excellent magnetic performance and catalytic activity. The degradation of p-NP (100 mg/L) has been optimized by response surface methodology based on the Box–Behnken design. The proposed regression model was highly reliable. Under optimal conditions (1.3 g/L catalyst, 32 °C, pH 3.1 and 11.3 mM H 2O2), the p-NP conversion and COD removal were found to be 100% and 81%, respectively, after 120 min. A degradation pathway of p-NP was proposed based on the identified intermediate products. It is demonstrated that p-NP was first reduced to p-aminophenol by hydrogen radicals or oxidized to hydroquinone by hydroxyl radicals and further oxidized to p-benzoquinone. We believe that Fe0/Fe3O4/Coke composite will be of huge potential application in heterogeneous Fenton catalysis for the p-NP wastewater treatment due to its lowcost, environment-friendly, good structural stability and simple separation.

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042

8

D. Wan et al. / Science of the Total Environment xxx (2016) xxx–xxx

Fig. 7. The possible degradation pathway of p-NP by Fe0/Fe3O4/Coke-H2O2 process.

Acknowledgements This project was supported in part by the Natural Science Foundation of Hubei Province (2014CFB810), Specialized Research Fund for the Doctoral Program of Higher Education of China (20114219110002) and Science and Technology Project (Major) of Jiangxi Province (20152ACG70003). Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.scitotenv.2016.08.042. References Ambashta, R.D., Sillanpää, M., 2010. Water purification using magnetic assistance: a review. J. Hazard. Mater. 180, 38–49. Bae, S., Gim, S., Kim, H., Hanna, K., 2016. Effect of NaBH4 on properties of nanoscale zerovalent iron and its catalytic activity for reduction of p-nitrophenol. Appl. Catal. B 182, 541–549. Catak, S., Hemelsoet, K., Hermosilla, L., Waroquier, M., Speybroeck, V.V., 2011. Competitive reactions of organophosphorus radicals on Coke surfaces. Chem. Eur. J. 17, 12027–12036. Chen, L., Deng, C., Wu, F., Deng, N., 2011. Decolorization of the azo dye Orange II in a montmorillonite/H2O2 system. Desalination 281, 306–311. Chen, X., Murugananthan, M., Zhang, Y., 2016. Degradation of p-nitrophenol by thermally activated persulfate in soil system. Chem. Eng. J. 283, 1357–1365. Chun, J., Lee, H., Lee, S.H., Hong, S.W., Lee, J., Lee, C., Lee, J., 2012. Magnetite/mesocellular carbon foam as a magnetically recoverable Fenton catalyst for removal of phenol and arsenic. Chemosphere 89, 1230–1237. Costa, R.C.C., Moura, F.C.C., Ardisson, J.D., Fabris, J.D., Lago, R.M., 2008. Highly active heterogeneous Fenton-like systems based on Fe0/Fe3O4 composites prepared by controlled reduction of iron oxides. Appl. Catal. B 83, 131–139. Gao, L., Zhuang, J., Nie, L., Zhang, J., Zhang, Y., Gu, N., Wang, T., Feng, J., Yang, D., Perrett, S., Yan, X., 2007. Intrinsic peroxidase-like activity of ferromagnetic nanoparticles. Nat. Nanotechnol. 2, 577–583. Gao, Y., Wang, Y., Zhang, H., 2015. Removal of rhodamine B with Fe-supported bentonite as heterogeneous photo-Fenton catalyst under visible irradiation. Appl. Catal. B 178, 29–36. Geng, Z., Lin, Y., Yu, X., Shen, Q., Ma, L., Li, Z., Pan, N., Wang, X., 2012. Highly efficient dye adsorption and removal: a functional hybrid of reduced graphene oxide–Fe3O4 nanoparticles as an easily regenerative adsorbent. J. Mater. Chem. 22, 3527–3535. Hou, L., Zhang, Q., Jérôme, F., Duprez, D., Zhang, H., Royer, S., 2014. Shape-controlled nanostructured magnetite-type materials as highly efficient Fenton catalysts. Appl. Catal. B 144, 739–749. Hu, X., Liu, B., Deng, Y., Chen, H., Luo, S., Sun, C., Yang, P., Yang, S., 2011. Adsorption and heterogeneous Fenton degradation of 17α-methyltestosterone on nano Fe3O4/ MWCNTs in aqueous solution. Appl. Catal. B 107, 274–283. Kakavandi, B., Takdastan, A., Jaafarzadeh, N., Azizi, M., Mirzaei, A., Azari, A., 2016. Application of Fe3O4@C catalyzing heterogeneous UV-Fenton system for tetracycline removal with a focus on optimization by a response surface method. J. Photochem, Photobiol. A 314, 178–188. Kraines, S., Akatsuka, T., Crissman, L.W., Polenske, K.R., Komiyama, H., 2002. Pollution and cost in the Coke-making supply chain in Shanxi Province, China. J. Ind. Ecol. 6, 161–184.

Kuosa, M., Kallas, J., Häkkinen, A., 2015. Ozonation of p-nitrophenol at different pH values of water and the influence of radicals at acidic conditions. J. Environ. Chem. Eng. 3, 325–332. Li, H., Li, Y., Xiang, L., Huang, Q., Qiu, J., Zhang, H., Sivaiah, M.V., Baron, F., Barrault, J., Petit, S., Valange, S., 2015. Heterogeneous photo-Fenton decolorization of Orange II over Alpillared Fe-smectite: response surface approach, degradation pathway, and toxicity evaluation. J. Hazard. Mater. 287, 32–41. Li, L., Feng, Y., Liu, Y., Wei, B., Guo, J., Jiao, W., Zhang, Z., Zhang, Q., 2016. Titanium dioxide nanoparticles modified by salicylic acid and arginine: structure, surface properties and photocatalytic decomposition of p-nitrophenol. Appl. Surf. Sci. 363, 627–635. Lin, H., Lin, Y., Liu, L., 2016. Treatment of dinitrodiazophenol production wastewater by Fe/C and Fe/Cu internal electrolysis and the COD removal kinetics. J. Taiwan Inst. Chem. E 58, 148–154. Liu, F., Wu, Z., Wang, D., Yu, J., Jiang, X., Chen, X., 2016. Magnetic porous silica–graphene oxide hybrid composite as a potential adsorbent for aqueous removal of p-nitrophenol. Colloids Surf. A 490, 207–214. Lu, L., Ai, Z., Li, J., Zheng, Z., Li, Q., Zhang, L., 2007. Synthesis and characterization of Fe– Fe2O3 core–shell nanowires and nanonecklaces. Cryst. Growth Des. 7, 459–464. Lu, L., Li, W., Wang, G., Zhang, Z., Wan, D., Lü, L., Liu, N., Chen, B., Jiang, S., 2016. Synthesis and characterization of biomimetic Fe3O4/Coke magnetic nanoparticles composite material. J. Wuhan Univ. Technol. 31, 254–259. Moura, F.C.C., Araujo, M.H., Costa, R.C.C., Fabris, J.D., Ardisson, J.D., Macedo, W.A.A., Lago, R.M., 2005. Efficient use of Fe metal as an electron transfer agent in a heterogeneous Fenton system based on Fe0/Fe3O4 composites. Chemosphere 60, 1118–1123. Munoz, M., de Pedro, Z.M., Casas, J.A., Rodriguez, J.J., 2015. Preparation of magnetite-based catalysts and their application in heterogeneous Fenton oxidation – a review. Appl. Catal. B 176–177, 249–265. Neamţu, M., Zaharia, C., Catrinescu, C., Yediler, A., Macoveanu, M., Kettrup, A., 2004. Fe-exchanged Y zeolite as catalyst for wet peroxide oxidation of reactive azo dye Procion Marine H-EXL. Appl. Catal. B 48, 287–294. Niasar, H.S., Li, H., Kasanneni, T.V.R., Ray, M.B., Xu, C., 2016. Surface amination of activated carbon and petroleum Coke for the removal of naphthenic acids and treatment of oil sands process-affected water (OSPW). Chem. Eng. J. 293, 189–199. Nidheesh, P.V., 2015. Heterogeneous Fenton catalysts for the abatement of organic pollutants from aqueous solution: a review. RSC Adv. 5, 40552–40577. Pang, Y., Lei, H., 2016. Degradation of p-nitrophenol through microwave-assisted heterogeneous activation of peroxymonosulfate by manganese ferrite. Chem. Eng. J. 287, 585–592. Sellers, R.M., 1980. Spectrophotometric determination of hydrogen peroxide using potassium titanium (IV) oxalate. Analyst 105, 950–954. Shahidi, D., Roy, R., Azzouz, A., 2015. Advances in catalytic oxidation of organic pollutants – prospects for thorough mineralization by natural clay catalysts. Appl. Catal. B 174– 175, 277–292. Sierra, U., Álvarez, P., Blanco, C., Granda, M., Santamaría, R., Menéndez, R., 2016. Cokes of different origin as precursors of graphene oxide. Fuel 166, 400–403. Sun, S.P., Lemley, A.T., 2011. p-Nitrophenol degradation by a heterogeneous Fenton-like reaction on nano-magnetite: process optimization, kinetics, and degradation pathways. J. Mol. Catal. A 349, 71. Tan, S.T., Umar, A.A., Salleh, M.M., 2015. (001)-Faceted hexagonal ZnO nanoplate thin film synthesis and the heterogeneous catalytic reduction of 4-nitrophenol characterization. J. Alloys Compd. 650, 299–304. Tang, L., Tang, J., Zeng, G., Yang, G., Xie, X., Zhou, Y., Pang, Y., Fang, Y., Wang, J., Xiong, W., 2015. Rapid reductive degradation of aqueous p-nitrophenol using nanoscale zerovalent iron particles immobilized on mesoporous silica with enhanced antioxidation effect. Appl. Surf. Sci. 333, 220–228. Variava, M.F., Church, T.L., Harris, A.T., 2012. Magnetically recoverable FexOy–MWNT Fenton's catalysts that show enhanced activity at neutral pH. Appl. Catal. B 123– 124, 200–207. Wang, N., Zheng, T., Jiang, J., Lung, W., Miao, X., Wang, P., 2014. Pilot-scale treatment of pnitrophenol wastewater by microwave-enhanced Fenton oxidation process: effects of system parameters and kinetics study. Chem. Eng. J. 239, 351–359.

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042

D. Wan et al. / Science of the Total Environment xxx (2016) xxx–xxx Wauters, S., Marin, G.B., 2001. Computer generation of a network of elementary steps for Coke formation during the thermal cracking of hydrocarbons. Chem. Eng. J. 82, 267–279. Xie, F., Xu, Y., Xia, K., Jia, C., Zhang, P., 2016. Alternate pulses of ultrasound and electricity enhanced electrochemical process for p-nitrophenol degradation. Ultrason. Sonochem. 28, 199–206. Xu, L., Wang, J., 2012. Fenton-like degradation of 2,4-dichlorophenol using Fe3O4 magnetic nanoparticles. Appl. Catal. B 123–124, 117–126. Xu, L., Wang, J., 2015. Degradation of 2,4,6-trichlorophenol using magnetic nanoscaled Fe3O4/CeO2 composite as a heterogeneous Fenton-like catalyst. Sep. Purif. Technol. 149, 255–264. Yu, L., Yang, X., Ye, Y., Wang, D., 2015. Efficient removal of atrazine in water with a Fe3O4/ MWCNTs nanocomposite as a heterogeneous Fenton-like catalyst. RSC Adv. 5, 46059–46066. Zhang, J., Zhuang, J., Gao, L., Zhang, Y., Gu, N., Feng, J., Yang, D., Zhu, J., Yan, X., 2008. Decomposing phenol by the hidden talent of ferromagnetic nanoparticles. Chemosphere 73, 1524–1528. Zhang, S., Zhao, X., Niu, H., Shi, Y., Cai, Y., Jiang, G., 2009. Superparamagnetic Fe3O4 nanoparticles as catalysts for the catalytic oxidation of phenolic and aniline compounds. J. Hazard. Mater. 167, 560–566. Zhang, A., Wang, N., Zhou, J., Jiang, P., Liu, G., 2012. Heterogeneous Fenton-like catalytic removal of p-nitrophenol in water using acid-activated fly ash. J. Hazard. Mater. 201–202, 68–73.

9

Zhang, C., Mo, Z., Zhang, P., Feng, C., Guo, R., 2013. Facile synthesis of porous carbon@ Fe3O4 composites and their applications in wastewater treatment. Mater. Lett. 106, 107–110. Zhou, J., Song, H., Maab, L., Chen, X., 2011. Magnetite/graphene nanosheet composites: interfacial interaction and its impact on the durable high-rate performance in lithium-ion batteries. RSC Adv. 1, 782–791. Zhou, H., Hua, L., Wan, J., Yang, R., Yu, X., Li, H., Chen, J., Wang, L., Lu, X., 2016. Microwaveenhanced catalytic degradation of p-nitrophenol in soil using MgFe2O4. Chem. Eng. J. 284, 54–60. Zhu, Q., Guo, S., Guo, C., Dai, D., Jiao, X., Ma, T., Chen, J., 2014. Stability of Fe–C microelectrolysis and biological process in treating ultra-high concentration organic wastewater. Chem. Eng. J. 255, 535–540. Zubir, N.A., Yacou, C., Motuzas, J., Zhang, X., da Costa, J.C.D., 2014. Structural and functional investigation of graphene oxide–Fe3O4 nanocomposites for the heterogeneous Fenton-like reaction. Sci. Rep. 4, 101–104. Zubot, W., MacKinnon, M.D., Chelme-Ayala, P., Smith, D.W., 2012. Mohamed Gamal El-Din, petroleum Coke adsorption as a water management option for oil sands process-affected water. Sci. Total Environ. 427–428, 364–372.

Please cite this article as: Wan, D., et al., Degradation of p-Nitrophenol using magnetic Fe0/Fe3O4/Coke composite as a heterogeneous Fenton-like catalyst, Sci Total Environ (2016), http://dx.doi.org/10.1016/j.scitotenv.2016.08.042