Enhanced Fenton-like Degradation of Trichloroethylene by Hydrogen Peroxide Activated with Nanoscale Zero Valent Iron Loaded on Biochar

Composite of nanoscale Zero Valent Iron (nZVI) loaded on Biochar (BC) was prepared and characterized as hydrogen peroxide (H2O2) activator for the degradation of trichloroethylene (TCE). nZVI is homogeneously loaded on lamellarly structured BC surfaces to form nZVI/BC with specific surface area (SBET) of 184.91 m2 g−1, which can efficiently activate H2O2 to achieve TCE degradation efficiency of 98.9% with TOC removal of 78.2% within 30 min under the conditions of 0.10 mmol L−1 TCE, 1.13 g L−1 nZVI/BC and 1.50 mmol L−1 H2O2. Test results from the Electron Spin Resonance (ESR) measurement and coumarin based fluorescent probe technology indicated that ∙OH radicals were the dominant species responsible for the degradation of TCE within the nZVI/BC-H2O2 system. Activation mechanism of the redox action of Fe2+/Fe3+ generated under both aerobic and anaerobic conditions from nZVI and single electron transfer process from BC surface bound C–OH to H2O2 promoted decomposition of H2O2 into ∙OH radicals was proposed.

As one of the most commonly used chemicals in industry, chlorinated solvents such as trichloroethylene (TCE) has been frequently encountered in subsurface environments as dense non-aqueous phase liquids (DNAPLs) at many industrial sites 1,2 . Due to its high toxicity and adverse effects on liver and kidney, TCE has been classified as a potential human carcinogen and listed as a priority pollutant by the United States Environmental Protection Agency (U.S. EPA) 3 , and the Safe Drinking Water Act in the USA defines the maximum contaminant level of TCE at 5 μg L −1 for drinking water 4 . Therefore, effective remediation and complete mineralization of TCE in aquifers is urgently required to reduce its adverse effects on the environment and human health.
Advanced oxidation processes (AOPs) have been emerged as the most efficient alternative to degrade various organic pollutants for the generation of reactive radicals 5 . Among AOPs, Fenton (i.e., the reaction between Fe 2+ and H 2 O 2 ) is a powerful oxidant generating the hydroxyl radicals (•OH, E 0 = 2.80 V), which react with various organic compounds at the near-diffusion controlled rates 6,7 , leading to an effective degradation and mineralization of organic pollutants 8 . However, it should to be operated at pH < 3.0 and the generated iron sludge limited the wide application of the homogeneous Fenton process 9 . Heterogeneous Fenton-like activator of zero valent iron (ZVI) was developed, which could be used over a wide pH range to decompose H 2 O 2 instead of homogeneous ferrous iron 10,11 . In addition, nanoscale zero valent iron (nZVI) could enhance H 2 O 2 activation due to the small particle size and high reactivity. Xu et al. 12 reported that the heterogeneous Fenton-like system using nZVI as catalyst was effective for the removal of biocide 4-chloro-3-methyl phenol in the presence of H 2 O 2 , and the reaction was induced through following reactions in the heterogeneous system of nZVI/H 2 O 2 : Though nZVI has performance for H 2 O 2 activation, it tends to aggregate into forming microscale particles due to its high surface energy and strong magnetic interaction, leading to the reduced reactivity 13 . To overcome the problem, granular activated carbon 14 , bentonite 15 , and rectorite 16 were introduced as a support for nZVI to gain better distribution. Biochar (BC) is a promising environmental friendly material pyrolyzed under low oxygen conditions. It possesses large surface area with porous structure and has oxygen containing functional groups 17 . Thus, it is anticipated that nZVI loaded uniformly on BC surface to form nZVI/BC composite will effectively prevent the aggregation of nZVI with significantly enhanced performance.
In this study, the composite of nZVI loaded on BC sheets was synthesized and characterized as H 2 O 2 activator for the degradation of TCE. The presence of C−OH groups on BC surface could be used to activate H 2 O 2 through electron transfer process 18 . Thus, both the dispersive nature and H 2 O 2 activator of BC will be simultaneously achieved for nZVI/BC. The present work aims to (1) synthesize a novel composite of nZVI/BC, where nZVI was loaded on BC surface uniformly and the aggregation of nZVI was prevented effectively, (2)

Results and Discussion
Characterization of nZVI/BC. SEM analyses were firstly conducted to observe the morphologies of the prepared nZVI, BC and nZVI/BC, respectively. As illustrated in Fig. 1a, nZVI was spherical with diameters of about 30 nm, and the agglomeration of nZVI was observed due to the nanometer effect and magnetic properties of nZVI. In addition, lamellarly structured BC of rough surface morphologies was obtained, and nZVI was homogeneously loaded on BC surface from the SEM image of nZVI/BC composite. XRD was also conducted, and the data were shown in Fig. 1d. It can be seen that the XRD pattern of nZVI revealed a highly crystalline and single phase structure by diffraction peaks at 45.0° (JCPD 01-087-0721) 19 . The crystallite size of nZVI was estimated to be 29.7 nm derived from the Debye-Sherrer equation (D = Kλ/(βcosθ), where K is the Sherrer constant (0.89), λ is the X-ray wavelength (0.15418 nm), β is the full peak width at half maximum and θ is the Bragg diffraction angle), which was in accordance with the SEM image. The broad reflection peak of BC in XRD indicated the amorphous BC, suggesting that nZVI was successfully loaded on BC surface from XRD spectrum of nZVI/BC (Fig. 1d).
The BET surface areas were measured by using N 2 adsorption method as shown in Fig. 1e. The S BET values were calculated according to where V is the volume of nitrogen adsorbed per gram, V m is the monolayer capacity and C is related to the heat of adsorption. From the results, the S BET value of bare nZVI was 26.61 m 2 g −1 , and was increased to 184.91 m 2 g −1 for the nZVI/BC composite after nZVI was loaded onto BC surface (S BET value of 205.35 m 2 g −1 ). In the FTIR spectrum, the band at about 3400 cm −1 was belonged to the vibration of hydroxyl groups (−OH). The signal at 1590 cm −1 was ascribed to C=O stretching vibration, the peak at 1100 cm −1 was assigned to the C−O groups, and the absorption peak at 802 cm −1 was corresponded to the vibration of aromatic C−H 17 . The weak adsorption in nZVI/BC spectrum at 561 cm −1 was observed in the Fig. 1f, indicating the Fe−O bond formed between BC and nZVI 20 .
Heterogeneous fenton-like degradation of TCE in nZVI/BC-H 2 O 2 system. The performances of TCE degradation by H 2 O 2 activated with nZVI, BC and nZVI/BC were investigated and presented in Fig. 2a. The control test suggested the TCE (0.10 mmol L −1 ) loss was less than 2% due to the volatilization during the experimental period under all the tested conditions. With the effect of 1.50 mmol L −1 H 2 O 2 , TCE was hardly degraded in the absence of any activators within 30 min, but its degradation efficiencies were enhanced to 39.1%, 6.5% and 98.9% with H 2 O 2 in the presence of nZVI, BC and nZVI/BC, respectively. Under all the tested conditions, the TCE degradation kinetics approximately followed pseudo-first-order reaction of ln (c 0 /c t ) = kt + b, where c 0 and c t are the TCE concentrations at initial time (t = 0) and reaction time (t = t), k is the apparent reaction rate constant (min −1 ), t is reaction time (min), and b is a constant. As illustrated in Fig. 2b by adding 0.19 g L −1 nZVI, the apparent reaction rate constant of TCE was 0.0128 min −1 , and the relative small k value of TCE degradation might be due to the aggregation of prepared nZVI with a small S BET value. The k value of 0.0002 min −1 for TCE degradation was observed when 0.94 g L −1 BC was added, indicating the weak activation ability of BC for H 2 O 2 , which was in accordance with the results reported by previous studies 21,22 . Loading nZVI onto BC significantly increased k values almost by 11 times from 0.0128 to 0.136 min −1 from Fig. 2a, demonstrating that nZVI/BC was more efficient than nZVI for H 2 O 2 activation. nZVI were distributed on the BC surface homogeneously from SEM image, and nZVI/BC has higher S BET value than that of the raw nZVI. The increased surface area of nZVI/BC enhanced the amount of H 2 O 2 activation sites and TCE adsorption, which might significantly lead to the increase in the H 2 O 2 activation ability and TCE degradation rate constant.
As shown in Fig. 2b and c, effects of nZVI/BC dosage and initial H 2 O 2 concentration were investigated. When nZVI/BC dosage was increased from 0.372 to 1.13 g L −1 , the k value for TCE degradation was almost linearly increased from 0.022 to 0.136 min −1 . The increased nZVI/BC dosage provided a large amount of H 2 O 2 active sites, and thus more •OH radicals generated 23 . However, the k value was decreased to 0.101 min −1 when the dosage of nZVI/BC was further increased to 1.13 g L −1 . The decreased k value may be ascribed to the decrease of •OH radicals for TCE degradation due to the scavenging effect by excessive Fe 2+ species (generated from Eqs 1 and 2) in according to Eq. 4 24 .
Scientific RepoRts | 7:43051 | DOI: 10.1038/srep43051 In consideration of the effect of H 2 O 2 concentration, the k values were increased quickly from 0.033 min −1 to 0.136 min −1 as H 2 O 2 concentrations were increased from 0.33 mmol L −1 to 1.50 mmol L −1 . H 2 O 2 is the precursor for •OH generation, relative high H 2 O 2 concentration induced more •OH radicals accounted for TCE degradation, and hence the increased k value was obtained. When the concentration of H 2 O 2 was beyond 1.50 mmol L −1 , the k value was decreased due to the reaction between •OH and excessive H 2 O 2 (Eq. 5) 8,25 . Therefore, the nZVI/ BC dosage and the initial H 2 O 2 concentration were fixed to be 1.13 g L −1 and 1.50 mmol L −1 respectively for the degradation of 0.10 mmol L −1 TCE. Figure 2d showed the effect of the initial solution pH on TCE degradation in the presence of nZVI/BC and H 2 O 2 . The ability of nZVI/BC to activate H 2 O 2 was decreased with the increase of the initial solution pH. However, the k value of 0.059 min −1 was observed when the solution pH was as high as 10.0, which indicated that the nZVI/BC-H 2 O 2 system could be effective even in alkaline pH conditions with no adjustment of the pH value being needed for the effective TCE degradation.   Fig. 3a.
2 2 degradation 2 2 decomposition As shown in Fig. 3b in the presence of 1.50 mmol L −1 H 2 O 2 , the TOC removal after 30 min was 5.1%, 32.6% and 78.2% with the addition of BC, nZVI and nZVI/BC, corresponding to the TCE degradation efficiencies of 6.5%, 39.0% and 98.9% respectively from GC results (Fig. 2a)  As shown in Fig. 4a, no signals were observed in the presence of H 2 O 2 . The measured ESR spectra in the three systems mentioned above illustrated the four-fold characteristic peak with an intensity ratio of 1:2:2:1, which were in accordance with the pattern of typical DMPO-•OH adduct 33 . Possibly due to the weak activation ability of BC and high adsorption of DMPO-•OH adduct by BC in BC-H 2 O 2 system 17 , relative low concentration of DMPO-•OH adduct was detected in aqueous solution.
The intensity of DMPO-•OH adduct in the nZVI/BC-H 2 O 2 system was much higher than that in the nZVI-H 2 O 2 system, indicating more •OH radicals being generated. In addition, due to the instability of the O 2 • − /HO 2 • radicals in the solution, six-fold characteristic peak of the O 2 • − /HO 2 • radicals adduct by using dimethyl sulfoxide as solvent was measured 34 , with no signal being detected (data were not shown). These results suggested that •OH radicals were the main ROSs generated from the decomposition of H 2 O 2 responsible for TCE degradation. Coumarin based fluorescent probe technology was used to measure the generation of •OH radicals. As illustrated in Fig. 4b the fluorescence intensities were increased quickly in the first few minutes in both nZVI/ BC-H 2 O 2 and nZVI-H 2 O 2 systems, indicating the fast generation of •OH radicals. The fluorescence intensity in the nZVI/BC-H 2 O 2 system was consistently higher than that in the nZVI-H 2 O 2 system, hinting the excellent property of nZVI/BC for H 2 O 2 activation. The data obtained here was well coincided with ESR results.
Discussion on reaction mechanism. nZVI particles could be oxidized to Fe 2+ under anaerobic or aerobic conditions, and it is known that homogeneous Fe 2+ plays a critical role in the activation of H 2 O 2 to generate •OH radicals 35,36 . Thus, the dissolved Fe 2+ and Fe 3+ concentrations in the presence of nZVI/BC with and without H 2 O 2 were measured after reaction, respectively. As shown in Fig. 5a, the concentrations of Fe 2+ and Fe 3+ were 0.73 and 3.15 mg L −1 respectively with only the nZVI/BC in aqueous solution. However, in the nZVI/BC-H 2 O 2 system, the concentrations of Fe 2+ and Fe 3+ were 7.70 and 45.59 mg L −1 respectively, being much higher than those in the  Reaction conditions: the concentration of nZVI itself or in nZVI/BC composite was 0.19 g L −1 , the concentration of nZVI/BC was 1.13 g L −1 , the dosage of BC was 0.94 g L −1 , the concentration of H 2 O 2 was 1.50 mmol L −1 , the concentration of TCE was 0.10 mmol L −1 , and the initial pH was 6.2.
absence of H 2 O 2 . The results indicated that the oxidation of nZVI to Fe 2+ was occurred initially, and the dissolved Fe 2+ directly activated H 2 O 2 to generate •OH radicals subsequently. The consumption of Fe 2+ for H 2 O 2 activation accelerated nZVI transformation and Fe 3+ formation in accordance with Eq. 3.
In the BC-H 2 O 2 system, the degradation efficiency of TCE was 6.5%, indicating that BC has also acted as an electron-transfer mediator to activate H 2 O 2 . BC characteristics of porosity, specific surface area, surface inertness and surface functional groups might significantly affect the catalytic activity for H 2 O 2 decomposition 37,38 . XPS spectra of nZVI/BC were measured to better understand the roles of BC in the activation of H 2 O 2 before and after the reaction. As illustrated in Fig. 5b, the peaks of C (1 s) at 284.5, 286.5 and 289.0 eV were attributed to C-C, C-OH and COOH, respectively. From the spectra, the proportion of C-C, C-OH and COOH peaks were 71.8%, 21.7% and 6.5% respectively for fresh nZVI/BC before the reaction. However, the proportion of the three peaks mentioned above were 70.0%, 17.3% and 12.7% after the H 2 O 2 activation. The decrease in surface bound C-OH proportion of BC after the reaction suggested that BC might act as an electron transfer medium participated in the H 2 O 2 activation 39,40 , and the increase in COOH was due to the reaction between C-OH and H 2 O 2 in accordance with Eq. 8 18 . By releasing organic radicals of CO• from C-OH through single electron transfer process, •OH radicals were generated (Eq. 9) 41 .
As derived from the above discussions, the activation mechanism of H 2 O 2 in the presence of nZVI/BC was proposed in Fig. 6. Firstly, nZVI particles were oxidized to Fe 2+ , and the redox reaction of Fe 2+ /Fe 3+ was accounted for the generation of •OH radicals. Secondly, as an electron transfer mediator to H 2 O 2 , BC surface bound C-OH decomposed H 2 O 2 into •OH radicals by releasing CO• radicals. Once •OH radicals having been  produced, it would rapidly react with TCE. GC-MS was utilized to monitor the process of TCE degradation, however, no intermediate products was detected except for the undegraded TCE. Though TCE would be transformed into low molecule weight organic acids with the effect of •OH radicals initially, only CO 2 and Cl − were measured during the oxidative process in the nZVI/BC-H 2 O 2 -TCE system.

Conclusions
The nZVI/BC composite was successfully synthesized and characterized as an efficient H 2 O 2 activator for the degradation of TCE. nZVI loaded on lamellarly structured BC surface prevented its agglomeration behaviour, whcih significantly enhanced the generation of •OH radicals. The redox effect of Fe 2+ /Fe 3+ and the single electron transfer process from BC surface bound C-OH to H 2 O 2 were accounted for the promoted generation of •OH radicals, leading to rapid TCE degradation. The enhanced Fenton-like activation of H 2 O 2 using nZVI/BC presents the great potential for TCE degradation in aqueous solution.
Synthesis of nZVI/BC composite. Biochars were prepared by the pyrolysis of rice hull collected locally.
Firstly, the rice hull was washed with ultrapure water and dried in oven at 80 °C. Secondly, the dried rice hull was pyrolyzed in muffle furnace under oxygen limited condition at a temperature of 350 °C for 6 h. Finally, the BC were obtained after treating with 1.0 mol L −1 HCl and washed with ultrapure water.
For the synthesis of nZVI/BC composite, 3.78 g biochar was dispersed in 250 mL oxygen free ultrapure water. Then, 0.0135 mol FeSO 4 • 7H 2 O was added at pH 5.0. With mechanical stirring and ultrasonic, nZVI was formed and loaded on the surface of BC by addition of 100 mL 0.27 mol L −1 NaBH 4 at a velocity of 20 mL min −1 . Following 2 h reaction, the nZVI/BC composite were separated and washed with deoxygenized ultrapure water, and finally vacuum dried. The preparation of nZVI was described as Eq. 10, and the schematic for the preparation of nZVI/BC composite was shown in Fig. 7. Characterization. X-ray diffraction (XRD, X'TRA, Swiss) analysis was conducted to determine the crystal structure and crystallinity of the prepared composites using Cu K α radiation with 2θ collection range from 10° to 80°. X-ray photoelectron spectra (XPS) were recorded on Axis Ultra spectrometer (Kratos) using Al Kα radiation excitation source. The infrared spectrum was recorded on Fourier transform infrared spectroscopy (FT-IR) spectra from 400 to 4000 cm −1 (NICOLET iN10 MX, Thermo Scientific, USA). The morphology of the composites were observed using scanning electron microscope (SEM, Hitachi S-4800, Japan) with 10 kV accelerating voltage, and the Brunauer-Emmett-Teller (BET) specific surface areas (S BET ) were measured with ASAP 2020M + C (Micromeritics, USA) from N 2 adsorption method. Procedures and analysis. In a typical sacrificial batch experiment, a 20 mL cylindrical glass vessel was fully filled with 0.10 mmol L −1 TCE, appropriate amounts of H 2 O 2 and activators (i.e., nZVI, BC or nZVI/BC), and the vessel was tightly sealed with Teflon reaction head successively. Then, the reaction was initiated in a rotary shaker at 298 K and 150 rpm. Control test was also carried out under the same condition without H 2 O 2 and activators. Samples were taken at the desired reaction time intervals and filtered through 0.2 μm membrane prior to the analysis. Samples were conducted in triplicate and the mean value was obtained. The concentration of TCE was analyzed by headspace Gas Chromatograph Mass Spectrometer (GC-MS, Agilent 7890A and 5975C) using DB-624 chromatographic column (30.0 m × 0.25 mm × 1.4 μm). Dissolved ferrous ion was quantified through 1,10-Phenanthrolinemonohydrate Spectrophotometry by using a Cary 50 UV-vis spectrophotometer (Varian Cary 50, USA). The concentration of H 2 O 2 was measured with the DPD method using UV-vis spectrophotometer (Varian Cary 50, USA). The generated radical species was detected with electron spin resonance spectrometer (Bruker ESR 300E, Germany) with microwave bridge (receiver gain, 1 × 10 5 ; modulation amplitude, 2 Gauss; microwave power, 10 mW; modulation frequency, 100 kHz) using DMPO as radical spin-trapping reagent, and fluorescence spectra were measured on fluorescence spectrophotometer (Jasco FP-6200, Japan). Total organic carbon (TOC) was recorded with a multi N/C model TOC analyzer (Analytik Jena, multi N/C 2100, Germany).