hybrid catalytic degradation processes

Accepted Manuscript Controllable Mullite Bismuth Ferrite Micro/Nanostructures with Multifarious Catalytic Activities for Switchable/Hybrid Catalytic Degradation Processes Zhong-Ting Hu, Wen-Da Oh, Yiquan Liu, En-Hua Yang, Teik-Thye Lim PII: DOI: Reference:

S0021-9797(17)31058-5 http://dx.doi.org/10.1016/j.jcis.2017.09.035 YJCIS 22786

To appear in:

Journal of Colloid and Interface Science

Received Date: Revised Date: Accepted Date:

22 June 2017 18 August 2017 7 September 2017

Please cite this article as: Z-T. Hu, W-D. Oh, Y. Liu, E-H. Yang, T-T. Lim, Controllable Mullite Bismuth Ferrite Micro/Nanostructures with Multifarious Catalytic Activities for Switchable/Hybrid Catalytic Degradation Processes, Journal of Colloid and Interface Science (2017), doi: http://dx.doi.org/10.1016/j.jcis.2017.09.035

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Controllable Mullite Bismuth Ferrite Micro/Nanostructures with Multifarious Catalytic Activities for Switchable/Hybrid Catalytic Degradation Processes Zhong-Ting Hu, *,a,b Wen-Da Oh,c Yiquan Liu,d En-Hua Yang,b Teik-Thye Lim*,b,e a

College of Environment, Zhejiang University of Technology, Hangzhou 310014, China School of Civil and Environmental Engineering, Nanyang Technological University (NTU), 50 Nanyang Avenue, Singapore 639798 c Nanyang Environment and Water Research Institute (NEWRI), NTU, 1 Cleantech Loop, CleanTech One, Level 6, Singapore 637141 d Energy Research Institute @ NTU (ERI@N), Research Techno Plaza, X-Frontier Block, Level 5, 50 Nanyang Avenue, Singapore 639798 e Environmental Chemistry & Materials Centre, Nanyang Environment and Water Research Institute (NEWRI), NTU, 1 Cleantech Loop, CleanTech One, Level 6, Singapore 637141 *E-mail: [email protected]; [email protected] Table of Contents b

ABSTRACT: In this work, controllable preparation of micro/nanostructured bismuth ferrites (BFOs) were used to investigate multifarious heterogeneous catalyses, including Fenton/Fenton-like reaction, photocatalysis, photo-Fenton oxidation, and peroxymonosulfate

1

(PMS) activation. Results showed that BFO can be used as a novel catalyst to activate switchable catalytic degradation of organic matters. Additionally, a novel catalytic system for degradation of organic pollutants, which integrating all-above heterogeneous catalyses is denoted as BFO/H2O2/PMS hybrid reaction, is introduced for the first time. BFO/H2O2/PMS system effectively degraded > 99% for both methyl orange (MO) and sulfamethoxazole (SMX) within 60 min, which shows better efficiency than above BFO-driven catalyses. The major SMX degradation pathway in BFO/H2O2/PMS system is proposed via detecting intermediates using LC/MS/MS. It was found that catalytic activities of BFOs are in the order of BFO-L (co-precipitation, micro/nanosize, single crystals exposing facet (001)) > BFO-H (hydrothermal, nanocluster with a higher surface area than other BFOs) > BFO-C (fabricated using calcination process, microsize), which demonstrated that crystallographic orientation is more significant in heterogeneous catalyses than specific surface area at micro/nanoscale. Besides, the required H2O2 consumption for achieving 99% TOC removal was identified in BFO-driven photo-Fenton oxidation. The other effects on degradation efficiency, such as H2O2 dosage and pH, were investigated as well. In Fenton/Fenton-like reaction, reaction conditions suggested are ~61.5 mM H2O2 dosage and pH  4.5 to avoid quenching of HO • into HO2• by excessive H2O2 and Fe leaching. KEYWORDS: Hybrid catalytic oxidation, Wastewater purification, Switchable catalytic activity, Fenton, Sulfate radical, Micro/nanostructures

1. INTRODUCTION Advanced oxidation processes (AOPs) are one of the most effective remediation technologies for removing refractory organic pollutants in water/wastewater treatment. The key AOPs include UV photochemical oxidation, peroxone reaction (H 2O2/O3), ozonation, conventional Fenton oxidation, electrochemical oxidation, ultrasonication, supercritical water oxidation (SCWO), heterogeneous photo-Fenton oxidation, and photocatalysis [1-4]. The

2

AOP based on heterogeneous catalysis has received increasing attention as a promising technology for the advantages of 1) eco-friendly, 2) cost-effective, 3) ease of catalyst recovery, 4) simpler post-treatment, and 5) energy efficient. From the viewpoint of environmental applications, it is critical to develop a solid catalyst with nontoxicity, good physiochemical stability, and high catalytic activity for effectively activating heterogeneous catalysis. Over the past decades, metal oxides have attracted considerable attention as heterogeneous catalysts for various catalytic processes. For examples, TiO 2 has been extensively studied as an UV-driven heterogeneous catalyst for photocatalytic removal of recalcitrant pollutants [5, 6]. The Fe2O3/Fe3O4 has been extensively developed as catalyst for Fenton/Fenton-like and photo-Fenton reaction [7]. As compared with the Fenton/Fenton-like reaction, photo-Fenton reaction by virtue of absorbing light energy can enhance the formation of radical species in the presence of illumination [8, 9]. Moreover, Co3O4 is widely used to generate sulfate radical via catalytic activation of peroxymonosulfate (PMS, HSO 5-) [10]. However, there are some disadvantages associated with the application of these metal oxides as heterogeneous catalyst for water treatment, such as low efficiency and poor stability of catalysts, strict conditions required for reaction systems (e.g., lower pH), or toxic metal leaching (e.g., Co 2+). As such, a robust catalyst without these disadvantages is desirable. In ternary metal oxides, redox coupling effect between different polyvalent metal cations within a crystalline structure is beneficial to obtaining a robust catalyst with high activity (via strong electron-electron interactions) and chemical stability [11]. Recent studies revealed that ternary metal oxides can be utilized as heterogeneous catalysts for various applications including photocatalysis, photo-Fenton and Fenton/Fenton-like reactions, and PMS activation in water/wastewater treatment. In (photo-)catalytic oxidations, the mechanism of reaction can be ascribed to the redox reaction involving the interconversion of M (n+1)+/Mn+ (M = metal

3

element) state within the metal oxide catalyst along with the formation of various reactive oxygen species (e.g. H2O2 converted to HO• or HO2•, PMS to SO4•- or SO5•-). The reported ternary metal oxides as catalysts for water treatment include Bi2Fe4O9, BiFeO3, CuFeO2, LaFeO3, MFe2O4 (M = Fe, Mn, Co, Ni, Cr, Cu, Mg, Zn, or Ti), and M’xM’’yOz (M’ = Co, Cu, Mn, Zn, Ni, or Ru; M’’ = Bi, Al, Ru, or Cr) [10, 12-22]. It has been reported that nanostructured bismuth ferrite (e.g., Bi2Fe4O9) exhibits remarkable visible-light-driven photo-Fenton and photocatalysis for the degradation of refractory organic pollutants in aqueous systems due to its narrow bandgap (~2.2 eV), iron-rich characteristics, distorted structure and spontaneous polarization [23, 24]. In perspective, a catalyst possessing more than one desired catalytic abilities (or namely multifunctionalized catalysts) can be one of the most attractive candidates for degradation of organic matter in wet atmosphere. Moreover, the catalytic activity of metal oxides has a close relationship with material characteristics in terms of crystallographic orientation, crystallinity, nanostructure, morphology, particle size and surface property (namely structure-activity relationship) [25-29]. However, the daunting challenges in synthesis of Bi2Fe4O9 are preventing formation of impurity phases and tuning its morphology, nanostructure or particle size simultaneously since Bi2 Fe4O9 is sensitive to synthesis parameters such as temperature and oxygen pressure [30]. In this study, different synthesis methods were used to successfully fabricate various bismuth ferrite (BFO)-based catalysts which possess different material characteristics such as single-crystalline or polycrystalline structures, particle sizes, nanostructures, or surface properties. For the first time, BFO was used to investigate the influences of material characteristics and different catalytic oxidation systems on the efficiency of heterogeneous catalyses in water treatment. The effects of pH value and type of chemical agents (H 2O2, PMS) on the performance of the catalytic oxidation system were investigated. The required dosages of H2O2 in Fenton/Fenton-like and photo-Fenton reaction were determined. Besides

4

methyl orange (MO) dye with the TOC removal, the degradation of organic matters were investigated using sulfamethoxazole (SMX) antibiotics. The SMX degradation pathway in BFO-L/H2O2/PMS hybrid reactions was proposed based on the detected reaction intermediates using LC/MS/MS. A plausible mechanism of BFO-driven switchable catalytic activity is proposed. 2. EXPERIMENTAL SECTION 2.1 Chemicals and Materials All the chemicals were of analytical grade and used without further purification. The chemicals are bismuth(III) nitrate pentahydrate (≥ 98%, VWR), ferric(III) nitrate nonahydrate (≥ 99%, Merck), sodium hydroxide (pellet, Schedelco), nitric acid (1N, Merck), citric acid (≥ 99.5%, Merck), urea (99%, Sigma-Aldrich), methanol (LC grade, Merck), absolute ethanol (99.9%, Fisher chemical), methyl orange (85%, Sigma), PMS (in the form of Oxone ®, 2KHSO5∙KHSO4∙K2SO4, Alfa Aesar), and hydrogen peroxide (35% w/w, Alfa Aesar). Milli-Q ultrapure water (18.2 MΩ cm) was used for all experiments.

Scheme 1. Schematic illustration of the formation methods of various as-prepared catalysts. 2.2 Preparation of Materials

5

The schematic illustration of the different synthesis methods is presented as shown in Scheme 1. Typically, Bi(NO3)3∙5H2O (1.21 g) and Fe(NO3)3∙9H2O (2.02 g) were dissolved in 2 ml of 2 M HNO3 and citric acid (3.2 g) was dissolved in 5 mL of water, respectively. A transparent solution could be obtained after mixing the two solutions together in Teflon vessel. 33 mL of 12 M NaOH was then instantly added into the solution with vigorous stirring. After stirring for 1 h, the Teflon vessel containing the deep-brown slurry (Bi/Fe hydroxide precipitate) was transferred to an oil bath and heated at 95°C with constant stirring. After 12 h, the reaction was cooled down naturally to room temperature. Bismuth ferrite nanoparticles (surface-modified BFO NPs) could be prepared via a dehydration process of Bi/Fe hydroxides. The hydroxides were successfully deagglomerated using citrates as the chelating agent. The surface-modified BFO NPs were used to prepare microstructured bismuth ferrite (BFO-C) by a conventional calcination process with heating rate of 5°C min-1, holding temperature at 700°C and holding time of 24 h. For comparison, different reaction conditions were investigated at temperature range of 200-700°C, and holding times of 20 min to 24 h. As shown in Scheme 1, the surface-modified BFO NPs also could be used to fabricate bismuth ferrite clusters (BFO-H) at 200°C for 10 min via a delicate synthesis process including hydrothermal treatment in methanol/water co-solvent system. Moreover, hierarchical structures (BFO-S) were fabricated using another bismuth ferrite cluster at 200°C for 3 d by a facile solvothermal treatment in methanol system. The detailed synthesis process can refer to our previously published report [20]. At last, a single-crystalline bismuth ferrite (BFO-L) was prepared without adding citric acid via a co-precipitation method at 95°C for 36 h in an oil bath [31], in which the evolution process of BFO-L mainly includes bismuth ferrite nanoparticle formation and crystal growth with self-assembly. 2.3 Material Characterization

6

The resultant products were characterized by various techniques. The morphologies and/or structures were investigated using field emission scanning electron microscopy (FESEM) (JEOL JSM-7600F), and transmission electron microscopy (TEM) (JEOL JEM-2010 UHR). The crystal structure was studied with X-ray powder diffraction (XRD) analysis (Bruker, D8 Advance, Cu-Ka, λ = 1.5418 Å). Moreover, the chemical composition and elemental distribution were examined by X-ray Energy Dispersive Spectroscopy (EDS, Oxford Xmax80 LN2 Free). The material was further characterized using UV-vis spectroscopy (UV-2600, Shimadzu), X-ray photoelectron spectroscopy (XPS) (Phoibos 100 Spectrometer equipped with a monochromatic Mg X-ray source, SPECS, Germany), and Fourier

transform-infrared

spectroscopy

(FT-IR)

(PerkinElmer

GX).

The

N2

adsorption/desorption isotherm was obtained using a Quantachrom Quadrasorb SI. 2.4 Performance Evaluation A series of experiments were conducted to investigate the effect of the catalysts on the degradation

of

visible-light-driven

organic

matters

photocatalysis,

via

various

heterogeneous

Fenton/Fenton-like

reaction,

catalyses

including

visible-light-driven

photo-Fenton oxidation, sulfate radical-based oxidation, and their hybrid reactions. In a typical experiment, 50 mL of 3 mg L -1 of MO (or SMX) was introduced into a reactor. The pH of the solution was adjusted to the pre-determined pH of 3, 4.5, 6, 7 or 9 using 0.25 M H2SO4 or 0.5 M NaOH. A certain amount of the catalyst was added into the solution followed by continuous stirring for 1 h in dark to achieve adsorption-desorption equilibrium. Thereafter, for the Fenton/Fenton-like reaction experiment, the desired amount of H2O2 of 10, 20, 40, 60 or 80 mmol L-1 was added into the reactor to commence the catalytic reaction in dark. For the visible-light-driven photo-Fenton oxidation, the reaction was carried out under the simulated visible light (420-630 nm) using a solar simulator equipped with dichroic 7

mirror (Newport, 150 W Xenon arc lamp). For visible-light-driven photocatalysis, there is no addition of H2O2. There are few differences for experiments between Fenton/Fenton-like reaction and sulfate radical-based oxidation. Since the PMS is an acidic chemical, the pH of the solution was adjusted to the pre-determined pH after adding PMS into MO solution and the amount required for 0.25 M H2SO4 or 0.5 M NaOH was recorded. The experimental procedure is similar to that of Fenton/Fenton-like reaction. The desired amount of PMS was added into the reactor to activate the catalytic reaction. At each designated time interval, a certain amount of solution was sampled from the reactor. After the catalyst was separated, the supernatant solution for MO was analyzed for UV-vis absorption using UV-vis spectrometer (Shimadzu, UV-1800) and the SMX concentration was quantified using a High-Performance Liquid Chromatography (HPLC, Perkin Elmer Series200) equipped with Hypersil GOLD C18 column. The HPLC was operated using the following conditions: a mobile phase of methanol/water (60/40, v/v), a flow rate (0.8 mL min-1) and detector wavelength of 256 nm. The total organic carbon (TOC) remaining in the reaction solution was measured using a TOC analyzer (Shimadzu ASI-V TOC). The degradation pathway of SMX in hybrid reactions was studied using LC/MS/MS system (Agilent 6460 Triple Quad LC/MS). The Fe leaching from the BFOs in reaction system was determined using Inductively Coupled Plasma-Mass Spectrometry (ICP-MS, Perkin Elmer Elan DRC-e). 3. RESULTS AND DISCUSSION 3.1 Characteristics of Catalysts Figure 1 presents the standard XRD patterns of Bi2Fe4O9 [23], Bi25FeO40 [32] and BiFeO3 [33], and the experimental XRD patterns of the as-prepared catalysts. In this study, the BFO NPs are pre-fabricated as the starting precursor to prepare mullite Bi2 Fe4O9 via various methods, namely direct calcination (BFO-C), low temperature co-precipitation (BFO-L), hydrothermal (BFO-H), and solvothermal (BFO-S). The BFO NPs are mainly composed of

8

BiFeO3 and a relatively small percentage (~4%) of Bi2 Fe4O9 (Figure S1). For BFO-C preparation, the calcination temperature plays a critical role in controlling the crystal phase of the material. The corresponding major characteristic diffraction peaks for the products at different calcination temperatures are presented in the Figure 1a. When the calcination temperature below 500°C is employed, crystalline phase of BFO consists of mainly BiFeO 3 formed. At 600°C, the Bi2Fe4O9 and sillenite crystal structure (Bi25FeO40) can be obtained. The single phase mullite Bi2 Fe4O9 (denoted as BFO-C) can be obtained after thermal treatment of BFO NPs at 700°C. The main diffraction peaks of Bi2Fe4O9 are attributed to its different facets, such as (001), (121), (211), (220), (112), (330), and (332), which are shown in Figure 1a. The time-elapsed growth of BFO-C at 700°C is further investigated by obtaining the XRD pattern at various holding time intervals (20 min to 24 h). The results indicate that there is no significant influence of holding time from 12 to 24 h on the resultant crystal structures of BFO-C (Figure 1b). However, when the holding time was reduced to 20 min, the resultant BFO becomes a mixture of BiFeO3 and Bi2 Fe4O9. Based on the results, the formation mechanism of Bi2 Fe4O9 is proposed as follow: BiFeO3

Bi25FeO40

Bi2 Fe4O9

These evidences indicate that the crystal structure of bismuth ferrite as a ternary metal oxide is very sensitive to reaction temperature [30] and the calcination process for preparing single phase mullite bismuth ferrite requires a relatively high temperature of 700°C.

9

Figure 1. XRD patterns of as-prepared catalysts at different temperature with holding time of 24 h (a), at different holding time with a temperature of 700°C (b), and as-prepared BFO-C (700°C for 24 h), BFO-L, BFO-H and BFO-S (c). The XRD pattern of BFO-L in Figure 1c can be indexed to single phase Bi2 Fe4O9 while the XRD patterns of BFO-H indicates that it consists mainly of Bi2Fe4O9 as the main crystalline product with BiFeO3 as the secondary product. The XRD/Rietveld analysis shows that the BFO-H catalyst has a Bi2Fe4O9:BiFeO3 weight ratio of 93:7. It can be observed that the XRD peaks of Bi2 Fe4O9 in BFO-C, BFO-H and BFO-L are similar except for the 10

difference in the diffraction peak intensities (particularly at 001 and 002), suggesting that the BFOs have different preferential crystal growth orientation [31]. The XRD pattern of BFO-S has different intensities and peak positions of the diffraction peaks compared to other as-prepared BFOs (more information refers to Supporting Information in Figure S2), implying that the physiochemical properties are different from other BFOs. The FESEM micrographs in Figure 2a and b indicate that BFO-C consists of microstructured particles with irregular morphology. Meanwhile, the BFO-L has a pad-like morphology with an average side length of 1.5 μm and thickness of 170 nm (Figure 2c). The selected area electron diffraction (SAED) pattern of BFO-L along the zone axis [001] in Figure 2d (inset) shows distinct sharp diffraction spots, indicating that BFO-L is single-crystalline with an exposed crystal facet of (001) plane. The high-resolution TEM (HRTEM) image displays clear lattice fringes (Figure S3). The EDX spectrum shows that the BFO-L has the element compositions of Bi, Fe and O (Figure 3a) with Bi:Fe molar ratio of 0.48:1 which is close to the theoretical Bi:Fe molar ratio of 0.5:1. The BFO-H has a cube-like morphology with the side length of 400-450 nm (Figures 2e and f). The TEM micrograph indicates that the BFO-H is a nanocluster consisting of smaller nanocrystals, which is further confirmed by the SAED pattern (Figure 2e, inset). The EDX elemental distribution mapping shows that Bi, Fe and O are homogeneously distributed within BFO-H (Figure S4) with Bi:Fe molar ratio of 0.51:1. The BFO-S has hierarchical nanostructures with ultrathin nanoflakes (Figure 2g and h). The XPS survey spectrum (Figure 3b) and the corresponding high-resolution XPS spectra (Figure S5) of BFO-S displays the characteristic peaks of Fe 2p and Bi 4f. As compared with the as-prepared BFO nanoclusters (Figure S6), the characteristic peaks of Fe 2p and Bi 4f in BFO-S are stable at binding energies of 725 eV (Fe 2p1/2), 710 eV (Fe 2p3/2), 164 eV (Bi 4f5/2) and 159 eV (Bi 4f7/2), respectively.

11

Figure 2. FESEM images of BFO-C (a, b). FESEM image of BFO-L and its TEM image (c) and SAED pattern (d). FESEM images of BFO-H (e) and its TEM image (f) and SAED pattern (inset in e). FESEM images of BFO-S and its TEM image (h).

12

Figure 3. (a) EDX spectrum of BFO-L. (b) XPS survey spectrum of BFO-S. (c) FT-IR transmittance spectra of BFO-C, BFO-L, BFO-H and BFO-S in the 400-4000 cm-1. (d) Nitrogen adsorption/desorption isotherm of BFO-C, BFO-L, BFO-H and BFO-S. (e) UV-vis absorption spectra of BFO-C, BFO-L, BFO-H and BFO-S and (f) the corresponding Kubelka-Munk transformed reflectance spectra. Figure 3c shows the FT-IR transmittance spectra of the as-prepared catalysts. In the region of 400-1100 cm-1, the characteristic peaks are attributed to the metal-oxygen bond of 13

Bi-O and Fe-O [31]. Except for the BFO-L, all the catalysts have a distinct peak at 1402 cm-1 attributed to the carboxyl C-OH stretching vibration which arises due to the residual organic compounds in the as-prepared catalysts. The peaks at 3420 and 1630 cm-1 can be assigned to the –OH stretching vibration and H-O-H bending stretching of the adsorbed H2O, respectively. In particular, it is worth noting that the peaks corresponding to the adsorbed moisture and the residual organic compounds on the surface of BFO-L is not significant, which could be due to the following two reasons: 1) there is no organic chemical used during the synthesis of BFO-L, and 2) BFO-L has exposed unique crystal facets of (001) plane with a higher surface energy and lower moisture adsorption. In the BFO-S, the additional peaks at 1479 and 1572 cm-1 are observed which could be attributed to the asymmetric C=O stretching in the COO- group and C=C vibration in the carbon residual formed during solvothermal synthesis, respectively [20, 34]. The results provide evidence that the presence of citric acid during synthesis would influence the surface properties of the prepared catalysts. As shown in Figure 3d, the BET specific surface areas of BFO-C, BFO-L, BFO-H and BFO-S are 0.37, 5.80, 7.07 and 13.84 m2 g-1, respectively. The surface area of the catalyst is influenced by the particle sizes, micro/nanostructures, and morphologies of the catalysts. Figure 3e shows the properties of photo-response of the as-prepared catalysts at the visible light region from 400 to 700 nm. The light absorption efficiencies are in the order of BFO-S > BFO-C > BFO-L ≈ BFO-H. The corresponding optical bandgaps are 2.15, 1.96, 2.16 and 2.14 eV respectively (Figure 3f). This indicates that all the catalysts synthesized can be used as visible-light active catalyst. A summary of the properties of the as-prepared BFO are shown in Table S1. 3.2 Performance of Various Catalysts in Varied Catalyses Heterogeneous Fenton/Fenton-like Reaction In the heterogeneous Fenton/Fenton-like reaction, the interconversion between fictitious BFO(Fe3+)/BFO(Fe2+) cations takes place within the solid catalyst with the formation of 14

reactive species and consumption of H2O2 as follow: BFO(Fe3+) + H2O2  BFO(Fe2+) + H+ +HO2•

(1)

BFO(Fe2+) + H2O2  BFO(Fe3+) + HO• + OH-

(2)

The control experiment with H2O2 shows that only < 1.5% of MO is degraded in 5 h without catalyst addition. Preliminary experiments were conducted using BFO-H to investigate the effects of pH (Figure 4) and H2O2 dosage (Figure 5). Typically, all the experimental data were fitted with the pseudo-first order kinetics, ln (C/Co) = - kappt, where kapp is the apparent degradation rate constant, Co and C are the concentration of MO at initial and at a certain reaction time t, respectively. In this study, the corresponding kinetic constants (kapp) are shown in Table 1. The results in Figure 4 indicate that the catalytic activity of BFO-H are the lowest at pH 7 with only ~ 20% removal efficiency (kapp = 1.01 × 10-3 min-1) at t = 300 min. The highest MO removal efficiency of > 90% (kapp = 7.92 × 10-3 min-1) can be achieved at pH 3. Considering the point of zero charge of the catalyst is 5.5, the surface of the catalyst shows a net negative charge at pH > 5.5. This leads to excessive interfacial repulsion between the catalyst surface and the negatively-charged MO (pKa = 3.47). When the pH value is decreased from 6 to 3, the MO adsorption rate gradually increases (Figure 4). This is mainly attributed to the stronger electrostatic attraction between the positively charge catalyst surface and the negatively charged MO. It should be noted that the molecular form of MO would be converted into its protonated form at pH < 3.47 (Figure S7), which reduces the net negative charges of MO via neutralization effect. The physiochemical properties of MO, such as color of solution and characteristic absorption peak (position and intensity), had been changed as shown in Supporting Information (Figure S8 and Table S2). However, this does not affect the MO quantification in this study. The Fe leaching from BFO-H to the solution after catalytic reaction at pH 3 is ~ 40 µg L-1 (0.03%, w/w) while for experiments conducted at other pH values (Table S1), no Fe is detected. The excellent performance of BFO-H at pH 3 can be 15

ascribed to 1) synergistic adsorption and oxidation of MO, and 2) homogeneous Fenton/Fenton-like reactions. Typically, at pH ~3, free iron ions can effectively activate homogeneous Fenton/Fenton-like reactions through Eqs. 3 and 4 [35, 36]. Fe3+ + H2O2  Fe2+ + H+ +HO2•

(3)

Fe2+ + H2O2  Fe3+ + HO• + OH-

(4)

Henceforth, the subsequent performance investigation is carried out at pH ~4.5 to avoid Fe leaching. There is no Fe leaching for the other as-prepared catalysts (Table S1).

Figure 4. Effect of pH on MO degradation in Fenton-like reaction with BFO-H catalyst. Figure 5a shows the effect of H2O2 dosage (from 20 to 80 mmol L-1) on MO degradation with BFO-H catalyst. The MO degradation efficiency gradually increases and peaks at 60 mmol L-1 with 54% MO removal efficiency and kapp = 2.06 × 10-3 min-1 (Figure 5b). Further increase of H2O2 dosage to 80 mmol L-1 leads to the decreased in MO degradation rate and efficiency (kapp = 1.74 × 10-3 min-1). Typically, the processes of Fenton/Fenton-like reaction can be classified into three stages, namely radical generation, radical consumption and organic degradation [19]. A reaction system with a higher concentration of H 2O2 up to 60 mmol L-1 enhances diffusion of H2O2 and formation of more radicals (i.e., HO• and HO2•) for MO degradation. However, excessive H2O2 acts as a HO• scavenger and transforms the HO• 16

into a less active radical species (Eq. 5) via the following termination steps: HO• + H2O2  HO2• + H2O

(5)

HO2• + HO2•  H2O2 + O2

(6)

Figure 5. (a) Effect of H2O2 dosage on MO degradation in Fenton-like reaction with BFO-H catalyst. (b) MO removal as a function of H2O2 dosage. The three-dimensional response surface and two-dimensional contour plot (Figures 6a and b) were constructed by plotting MO removal efficiency against H 2O2 dosage and pH using MATLAB R2015b (R2 is 0.9905). Experimental results for MO removal ratio at different conditions are showed in Supporting Information (Table S3). As can be seen, the optimum H2O2 dosage is 61.5 mmol L-1 at pH 4.5 (t = 300 min, catalyst loading of 0.12 g L -1). As such, this condition is used for the subsequent study.

Figure 6. (a) 3D surface response and (b) contour plot for MO removal against pH and H2O2 dosage in Fenton-like reaction at BFO-H loading of 0.12 g L-1.

17

As shown in Figure 7a, all the as-prepared catalysts (i.e., BFO-H, BFO-C and BFO-L) except BFO-S can be used to activate H2O2 to degrade MO. The poor performance of BFO-S can be ascribed to the poor adsorptive interaction between the catalyst and MO. The surface of BFO-S seems to be passivated/deactivated at pH 4.5, which loses the adsorption ability on MO (Figure S9). This may be attributed to the residual compounds, such as the organics (Figure 3c) and carbon residuals (Figure 3b and c), on its surface. The performance of BFOs as Fenton/Fenton-like catalyst is in the following order: BFO-L (82.3%) > BFO-H (54.6%) > BFO-C (19.5%). In comparison, the commercial TiO2 (Evonik P25) cannot be a catalyst to activate Fenton/Fenton-like reaction (Figure S10a). Compared with BFO-C (microstructure), BFO-L (micro/nano-structure) and BFO-H (nanocluster) with smaller particle sizes show better MO degradation efficiencies. This indicates that the catalytic activity could be influenced by the micro/nanostructure and specific surface area of the catalyst. In principle, the catalytic activity of a catalyst is directly proportional to its specific surface area [37]. Because heterogeneous catalysis is an interfacial phenomenon which relies on the adsorption of pollutant prior to degradation, higher specific surface area provides more active sites for adsorption and catalysis [6, 37]. In addition, the surface structure is important in providing specific crystallographic orientation with highly reactive facets. Figure 7a shows that BFO-L exhibits a better MO degradation efficiency than BFO-H and their adsorption showed a very similar performance even though BFO-L (5.8 m2 g-1) has a lower specific surface area than BFO-H (7.1 m2 g-1). This suggests that the specific crystallographic orientation (i.e., an exposed crystal facet) is more important in determining the catalytic performance of the catalyst than its specific surface area possibly due to the variation of adsorption process, electron transfer pathway, surface energy, array of ions, etc. Therefore, the catalytic activity of a solid catalyst is significantly influenced by its material characteristics such as

18

crystallographic orientation,

single-crystalline/polycrystalline structure, particle size,

nanostructure, morphology, and surface property [25-29].

Figure 7. Effect of various as-prepared catalysts on MO degradation in (a) Fenton/Fenton-like reaction, (b) visible light photocatalysis, and (c) visible light photo-Fenton oxidation.

19

Visible-Light-Driven Photocatalysis In general, when the BFO is illuminated with photon containing energy (hv) equal to or greater than its bandgap energy (~2.2 eV), the electron from the valence band can be promoted to the conduction band. The generated electron (e -) reacts with the electron acceptor (e.g., O2) to produce reactive radicals (e.g., O2•-) which oxidizes pollutant (Eq. 8). The generated hole (h+) with a weak valence band hole (VBH, 1.3 V at pH 7) is unable to oxidize H2O to HO• (Eo (HO•/H2O) = 2.27 V at pH 7) [31]. BFO + hv (λ > 420 nm)  h+ + e-

(7)

e- + O2 → O2•-

(8)

O2•- + H+  HO2•

(pH 4.5)

(9)

MO degradation efficiency via photolysis under visible light irradiation (Figure 7b) is insignificant (< 1%), while it is only 6% using P25-driven photocatalysis (Figure S10b) due to limited photoresponse of TiO2 at   420 nm. In the presence of a catalyst under visible light irradiation, the photocatalytic activity of the catalysts is in the following order: BFO-L (kapp = 0.714 × 10-3 min-1) > BFO-H (kapp = 0.596 × 10-3 min-1) > BFO-C (kapp = 0.057 × 10-3 min-1). Similar to the heterogeneous Fenton/Fenton-like reaction, the performance of the catalyst is highly influenced by the material characteristics of the catalyst. The solar-light-driven photocatalysis exhibits a lower MO degradation efficiency compared with the heterogeneous Fenton/Fenton-like reaction system because the main reactive radical (O2•-, Eo = - 0.28 V at pH 7) generated from the visible-light-driven photocatalysis is much weaker than the HO• (Eo = 2.27 V at pH 7). However, the MO degradation efficiency in BFO-H/photocatalysis system is similar to that of BFO-L/photocatalysis system. A plausible reason is that the residual organic functional group on the surface of BFO-H (Figure 3c)

20

enhances the photocatalytic activity in the formation of (organic functional group) sensitized BFO-H [38]. Heterogeneous Photo-Fenton Oxidation As compared to the Fenton/Fenton-like and photocatalysis reactions, the photo-Fenton reaction shows an accelerated MO degradation rate in the presence of BFO, H 2O2 and visible-light irradiation (Figure 7c). The mechanism of HO• generation by the photo-Fenton reaction is shown as follows [31]: BFO(Fe3+) + 2H2O2 + hv  BFO(Fe2+) + HO• + H+ + H2O + O2 (10) The BFO-L can achieve > 99% of MO degradation efficiency (kapp = 60.4 × 10-3 min-1) within 90 min (Figure 7c), while BFO-H and BFO-C achieve ~80 and ~50% of MO degradation efficiency, respectively. Since the photo-Fenton oxidation relies on both absorbing light energy and consuming H2O2, the importance of the strength of the catalyst’s catalytic activity becomes relatively minor. This is evidenced by the observation that BFO-C (Figure 7c) also exhibits a relatively good performance on MO degradation in photo-Fenton oxidation than in other heterogeneous catalysis. It indicates that requirement of catalyst quality can be lowered when supplying additional chemicals and/or energy in heterogeneous catalysis.

Figure 8. Performance of BFO-L for MO degradation through varied heterogeneous catalysis. 21

Figure 8 shows a comparison of the various heterogeneous catalytic reactions driven by BFO-L. The MO removal rate in different catalytic processes is in the order of photo-Fenton oxidation (kapp = 60.4 × 10-3 min-1) > Fenton/Fenton-like reaction (kapp = 4.76 × 10-3 min-1) > photocatalysis (kapp = 0.714 × 10-3 min-1). It is obvious that photo-Fenton shows not only an enhanced Fenton/Fenton-like reaction via absorbing light energy but also a great synergistic effect between Fenton/Fenton-like reaction and photocatalysis in the presence of H2O2 and visible light illumination.

Figure 9. (a) Effect of different H2O2 dosage on degradation of MO in photo-Fenton reaction with BFO-L and the corresponding removal of TOC at 2 h (inset). (b) Time-dependent TOC removal and H2O2 residual in photo-Fenton reaction with [H2O2] = 40 mmol L-1. Figure 9a shows the effect of H2O2 dosage on MO degradation efficiency and the corresponding TOC removal with BFO-L catalyst in photo-Fenton reaction. With the decrease of H2O2 dosage, the BFO-L can still achieve ~ 99% of MO degradation efficiency within 120 min. The corresponding TOC removal efficiencies at reaction time of 2 h are 54.7, 77.4 and 80.7% for H2O2 dosages of 20, 40 and 61.5 mM, respectively. When the reaction time at H2O2 dosage of 40 mM (Figure 9b) was extended to 4 h, the percentage of TOC removal could reach up to 99%. The corresponding consumed H2O2 was ~ 40% (Figure S11), which indicates that BFO-L driven photo-Fenton oxidation in the presence of 16 mM H2O2 with a sufficient reaction time is possible for 99% TOC removal

22

Heterogeneous Sulfate Radical-based Oxidation In heterogeneous sulfate radical-based oxidation, SO4•- and SO5•- can be formed through the activation of peroxymonosulfate (PMS, HSO5-) via the interconversion of Fe3+/Fe2+ states on the surface of catalysts as shown in Eqs. 11-12 [21]. Catalyst(Fe3+) + HSO5- → Catalyst(Fe2+) + SO5•- + H+

(11)

Catalyst(Fe2+) + HSO5-  Catalyst(Fe3+) + SO4•- + OH-

(12)

Figure 10. Effect of various as-prepared catalysts on MO degradation in sulfate radical-based oxidation. Figure 10 shows the MO degradation in the presence of catalyst and PMS at various time intervals, which indicates the as-prepared catalysts could be used to activate PMS efficiently. The PMS is a relatively strong oxidant (Eo = 1.76 V) and can remove up to 26% of MO in 60 min (control experiment with PMS only). The MO degradation rate and removal efficiency (in 40 min) in the presence of different catalysts are in the order of BFO-H (99%, kapp = 139 × 10-3 min-1) > BFO-L (71%, kapp = 104 × 10-3 min-1 ) > BFO-C (36%, kapp = 10.2 × 10-3 min-1) > BFO-S (31%, kapp = 8.9 × 10-3 min-1). When the catalyst loading is increased to 0.12 g L-1, the

23

MO degradation could be completed within 5 min (Figure S12). This result is commensurate with the trend of the specific surface area of the catalyst whereby the microstructured BFO-C with the lowest surface area exhibits the lowest catalytic activity. The crystallographic orientation is less influential compared with the specific surface area. It is because larger specific surface area provides more catalytic sites for reaction. The other plausible reason is that the importance of a specific crystal facet with a higher surface energy becomes relatively minor since PMS with asymmetrical molecular structure has very strong activity and is easily transformed into sulfate radicals only by virtue of a minor catalytic activity of applied catalyst [39-41].

Figure 11. Schematic illustration of the micro/nanostructured bismuth ferrites exhibiting switchable catalytic activities for degradation of methyl orange. Based on the above experimental results, a plausible mechanism of BFO-driven switchable catalytic activity is proposed (Figure 11). The interconversion of valence state of Fe atoms takes place internally within BFOs along with both formation of reactive oxygen species (ROSs) and consumption of related chemical agents (e.g., H2O2, PMS). In the presence/absence of visible light illumination, the generated ROSs can be SO5•-, SO4•-, HO•, 24

and HO2•, while it can also be O2•- as catalytic degradation process switched into photocatalysis. The formed ROSs on the surface of BFOs can be used to degrade the adsorbed MO subsequently. The reusability of BFOs as photocatalyst, Fenton/Fenton-like catalyst, photo-Fenton catalyst and PMS activator for MO degradation are conducted for 3 cycles. The results show that the as-prepared catalysts (BFO-L for photocatalyst, Fenton/Fenton-like catalyst, photo-Fenton catalyst, and BFO-H for PMS activation) exhibit good reusability (Figure S13). The catalysts are also stable without detectable metal leaching during reaction. This suggests that BFOs can be used as multifunctional catalysts for various environmental applications. 3.3 Hybrid Advanced Oxidation Processes (HAOPs) Figure 12a shows the MO degradation efficiencies in BFO-L/H2O2, BFO-L/PMS and BFO-L/H2O2/PMS systems. As can be seen, BFO-L/H2O2/PMS shows higher MO degradation efficiency than others with a good reusability (Figure S13). In addition, SMX could be degraded as well in BFO-L/H2O2/PMS system (Figure 12a), which is unequivocally confirmed in LC/MS/MS analysis (Figure S14). Figure 12b presents the proposed main SMX degradation pathway in the BFO-L/H2O2/PMS system while the mass spectra of the reaction intermediates are presented in Supporting Information (Figure S15). The SMX forms monohydroxylated sulfamethoxazole (OH-SMX, m/z (+) = 270) by direct attack of HO• on the aromatic moiety of SMX and/or hydrolysis of unstable radical cation SMX•+ formed by interaction with SO4•- [42]. Subsequently, the sulfonamide bond is cleaved by reactive oxygen species to produce monohydroxylated sulfanilic acid (OH-SAA, m/z (+) = 190) and 3-amino-5-methyl-isoxazole (AMI, m/z (+) = 99) [43, 44]. However, the OH-SAA intermediates were not detected in this study as a possible result of their quick transformation. It is noteworthy that monohydroxylated aniline (OH-AN, m/z (-) = 109) was identified, which provides an evidence that OH-SAA was produced during the degradation of 25

SMX. Furthermore, the reaction by-products can be further degraded into inorganic ions and organic acids via ring opening reactions [42, 44].

Figure 12. (a) Performance of BFO-L for degradation of organic matters through BFO-L/H2O2/PMS hybrid reactions and (b) Proposed SMX degradation pathway in the BFO-L/H2O2/PMS system. 4. CONCLUSIONS Various

BFOs

(mullite

bismuth

ferrites),

with

different

particle

sizes,

micro/nanostructures, and single-crystalline and polycrystalline states, were successfully prepared via different synthesis methods. One of the important findings in this study demonstrated for the first time that BFO is a novel multifunctional catalyst for both switchable catalyses and hybrid reactions (i.e., BFO/H2O2/PMS system) under visible light irradiation. A plausible mechanism of BFO-driven switchable catalytic activity is proposed in this study. The developed hybrid reaction shows the best pollutant degradation efficiency in environmental decontamination compared with BFO/PMS, visible light/BFO/H2O2 or other systems (e.g., Fenton/Fenton-like reaction, photocatalysis). Through investigation of various catalytic systems, the results consistently demonstrated that BFOs with smaller particle size show better catalytic activities while single-crystalline BFO with specific crystallographic orientation is more important in determining the catalytic activity than specific surface area at 26

micro/nanoscale. The results obtained in the application study reveal that (1) catalytic activity of BFO becomes weaker at pH 7 and its iron leaching was detected at pH 3; (2) excessive H2O2 inhibits the MO degradation efficiency in Fenton/Fenton-like reaction; (3) the TOC removal increases with increasing reaction time (or H2O2 dosage) and H2O2 consumption in photo-Fenton reaction. The major SMX degradation pathway in BFO/H 2O2/PMS system is proposed via detecting intermediates directly. This study signifies that BFO-driven switchable catalyses or their hybrid reactions could be used as a novel technology for effective degradation of organic pollutants in the aquatic system.

ACKNOWLEDGEMENTS The authors would like to acknowledge financial support from Centre of Infrastructure, School of Civil and Environmental Engineering, Nanyang Technological University, Singapore (M060030001). The authors are grateful to the laboratory staff of the Central Environmental Science and Engineering Laboratory (CESEL) and FACTS (Facility for Analysis, Characterisation Testing and Simulation) for their kind assistance.

REFERENCES [1] C. Comninellis, A. Kapalka, S. Malato, S.A. Parsons, I. Poulios, D. Mantzavinos, Journal of Chemical Technology and Biotechnology 83(6) (2008) 769-776. [2] A. Stasinakis, Global NEST journal 10(3) (2008) 376-385. [3] H. Barner, C. Huang, T. Johnson, G. Jacobs, M. Martch, W. Killilea, Journal of hazardous materials 31(1) (1992) 1-17. [4] A. Matilainen, M. Sillanpää, Chemosphere 80(4) (2010) 351-365. [5] A. Fujishima, X. Zhang, Comptes Rendus Chimie 9(5-6) (2006) 750-760. [6] N. Mizuno, M. Misono, Chemical Reviews 98(1) (1998) 199-218. [7] J.J. Pignatello, E. Oliveros, A. MacKay, Critical Reviews in Environmental Science and Technology 36(1) (2006) 1-84. [8] E. Graf, J.R. Mahoney, R.G. Bryant, J.W. Eaton, Journal of Biological Chemistry 259(6) (1984) 3620-4. [9] S. Giannakis, M.I.P. López, D. Spuhler, J.A.S. Pérez, P.F. Ibáñez, C. Pulgarin, Applied Catalysis B: Environmental 198 (2016) 431-446. [10] P. Hu, M. Long, Applied Catalysis B: Environmental 181 (2016) 103-117. [11] Y. Mao, T.J. Park, S.S. Wong, Chemical Communications (46) (2005) 5721-5735. [12] X. Li, Y. Huang, C. Li, J. Shen, Y. Deng, Chemical Engineering Journal 260 (2015) 28-36. [13] W. Liu, Y. Wang, Z. Ai, L. Zhang, ACS applied materials & interfaces 7(51) (2015) 28534-28544. [14] K. Rusevova, F.-D. Kopinke, A. Georgi, Journal of Hazardous Materials 241–242 (2012) 433-440. 27

[15] L. Xu, J. Wang, Environmental science & technology 46(18) (2012) 10145-10153. [16] S.R. Pouran, A.A.A. Raman, W.M.A.W. Daud, Journal of Cleaner Production 64 (2014) 24-35. [17] X. Zhang, Y. Ding, H. Tang, X. Han, L. Zhu, N. Wang, Chemical Engineering Journal 236 (2014) 251-262. [18] K. Rusevova, R. Köferstein, M. Rosell, H.H. Richnow, F.-D. Kopinke, A. Georgi, Chemical Engineering Journal 239 (2014) 322-331. [19] M. Munoz, Z.M. de Pedro, J.A. Casas, J.J. Rodriguez, Applied Catalysis B: Environmental 176 (2015) 249-265. [20] Z.-T. Hu, Z. Chen, R. Goei, W. Wu, T.-T. Lim, Nanoscale 8(25) (2016) 12736-12746. [21] W.D. Oh, Z. Dong, T.T. Lim, Applied Catalysis B: Environmental 194 (2016) 169-201. [22] Y. Yao, Y. Cai, F. Lu, F. Wei, X. Wang, S. Wang, Journal of hazardous materials 270 (2014) 61-70. [23] E. Kostiner, G.L. Shoemaker, Journal of Solid State Chemistry 3(2) (1971) 186-189. [24] Z.-T. Hu, S.K. Lua, X. Yan, T.-T. Lim, RSC Advances 5(106) (2015) 86891-86900. [25] Q. Kuang, S. Yang, ACS Applied Materials and Interfaces 5(9) (2013) 3683-3690. [26] T.P. Chou, Q. Zhang, B. Russo, G.E. Fryxell, G. Cao, The Journal of Physical Chemistry C 111(17) (2007) 6296-6302. [27] J.X. Wang, H. Inada, L. Wu, Y. Zhu, Y. Choi, P. Liu, W.-P. Zhou, R.R. Adzic, Journal of the American Chemical Society 131(47) (2009) 17298-17302. [28] S. Kango, S. Kalia, A. Celli, J. Njuguna, Y. Habibi, R. Kumar, Progress in Polymer Science 38(8) (2013) 1232-1261. [29] A. Verma, F. Stellacci, Small 6(1) (2010) 12-21. [30] M.M. Murshed, G. Nénert, M. Burianek, L. Robben, M. Mühlberg, H. Schneider, R.X. Fischer, T.M. Gesing, Journal of Solid State Chemistry 197(0) (2013) 370-378. [31] Z.-T. Hu, B. Chen, T.-T. Lim, RSC Advances 4(53) (2014) 27820-27829. [32] C.E. Infante, B. Carrasco, Materials Letters 4(4) (1986) 194-197. [33] P. Ravindran, R. Vidya, A. Kjekshus, H. Fjellvåg, O. Eriksson, Physical Review B 74(22) (2006) 224412. [34] J. Liu, H. Jeong, K. Lee, J.Y. Park, Y.H. Ahn, S. Lee, Carbon 48(8) (2010) 2282-2289. [35] M. Kitis, C.D. Adams, G.T. Daigger, Water Research 33(11) (1999) 2561-2568. [36] J. Yoon, Y. Lee, S. Kim, Investigation of the reaction pathway of OH radicals produced by Fenton oxidation in the conditions of wastewater treatment, 2001, pp. 15-21. [37] G.A. Somorjai, Y. Li, Introduction to surface chemistry and catalysis, John Wiley & Sons2010. [38] M.A. Fox, M.T. Dulay, Chemical reviews 93(1) (1993) 341-357. [39] X. Chen, W. Wang, H. Xiao, C. Hong, F. Zhu, Y. Yao, Z. Xue, Chemical Engineering Journal 193 (2012) 290-295. [40] Y. Ren, L. Lin, J. Ma, J. Yang, J. Feng, Z. Fan, Applied Catalysis B: Environmental 165 (2015) 572-578. [41] A. Khan, Z. Liao, Y. Liu, A. Jawad, J. Ifthikar, Z. Chen, Journal of Hazardous Materials (2017). [42] M.W. Lam, S.A. Mabury, Aquatic Sciences 67(2) (2005) 177-188. [43] Y. Ji, Y. Fan, K. Liu, D. Kong, J. Lu, Water Research 87 (2015) 1-9. [44] W.-D. Oh, V.W. Chang, T.-T. Lim, Environmental Science and Pollution Research (2017) 1-10.

28

List of Figures

Scheme 1. Schematic illustration of the formation methods of various as-prepared catalysts. Figure 1. XRD patterns of as-prepared catalysts at different temperature with holding time of 24 h (a), at different holding time with a temperature of 700°C (b), and as-prepared BFO-C (700°C for 24 h), BFO-L, BFO-H and BFO-S (c). Figure 2. FESEM images of BFO-C (a, b). FESEM image of BFO-L and its TEM image (c) and SAED pattern (d). FESEM images of BFO-H (e) and its TEM image (f) and SAED pattern (inset in e). FESEM images of BFO-S and its TEM image (h). Figure 3. (a) EDX spectrum of BFO-L. (b) XPS survey spectrum of BFO-S. (c) FT-IR transmittance spectra of BFO-C, BFO-L, BFO-H and BFO-S in the 400-4000 cm-1. (d) Nitrogen adsorption/desorption isotherm of BFO-C, BFO-L, BFO-H and BFO-S. (e) UV-vis absorption spectra of BFO-C, BFO-L, BFO-H and BFO-S and (f) the corresponding Kubelka-Munk transformed reflectance spectra. Figure 4. Effect of pH on MO degradation in Fenton-like reaction with BFO-H catalyst. Figure 5. (a) Effect of H2O2 dosage on MO degradation in Fenton-like reaction with BFO-H catalyst. (b) MO removal as a function of H2O2 dosage. Figure 6. (a) 3D surface response and (b) contour plot for MO removal against pH and H2O2 dosage in Fenton-like reaction at BFO-H loading of 0.12 g L-1. Figure 7. Effect of various as-prepared catalysts on MO degradation in (a) Fenton/Fenton-like reaction, (b) visible light photocatalysis, and (c) visible light photo-Fenton oxidation. Figure 8. Performance of BFO-L for MO degradation through varied heterogeneous catalysis. Figure 9. (a) Effect of different H2O2 dosage on degradation of MO in photo-Fenton reaction with BFO-L and the corresponding removal of TOC at 2 h (inset). (b) Time-dependent TOC removal and H2O2 residual in photo-Fenton reaction with [H2O2] = 40 mmol L-1. Figure 10. Effect of various as-prepared catalysts on MO degradation in sulfate radical-based oxidation. Figure 11. Schematic illustration of the micro/nanostructured bismuth ferrites exhibiting switchable catalytic activities for degradation of methyl orange. Figure 12. (a) Performance of BFO-L for degradation of organic matters through BFO-L/H2O2/PMS hybrid reactions and (b) Proposed SMX degradation pathway in the BFO-L/H2O2/PMS system.

29

Table 1 Kinetic constants using various as-prepared catalysts at different heterogeneous catalysis. Photocatalysis a

AOPs

H2O2 e kapp -1 (mmol L ) (× 10-3 min-1)

Photo-Fenton oxidation c

Sulfate radical-based d oxidation

kapp (× 10-3 min-1)

R

kapp (× 10-3 min-1)

R

kapp (× 10-3 min-1)

R2

2

2

Sample

pH

BFO-C

4.5 ±0.1 61.5

0.057

0.755

0.613

0.911

9.91

0.998

10.2

0.983

BFO-H

3 ±0.1

60

-

-

7.92

0.934

-

-

-

-

4.5 ±0.1 60

-

-

2.06

0.998

-

-

-

-

4.5 ±0.1 61.5

0.596

0.939

2.15

0.862

26.9

0.952

193

0.838

6 ±0.1

60

-

-

1.67

0.968

-

-

-

-

7 ±0.1

60

-

-

1.01

0.959

-

-

-

-

9 ±0.1

60

-

-

1.93

0.968

-

-

-

-

4.5 ±0.1 10

-

-

0.864

0.952

-

-

-

-

4.5 ±0.1 20

-

-

0.881

0.961

-

-

-

-

4.5 ±0.1 40

-

-

1.28

0.941

-

-

-

-

4.5 ±0.1 80

-

-

1.74

0.939

-

-

-

-

4.5 ±0.1 61.5

0.714

0.992

4.76

0.903

60.4

0.948

104

0.907

BFO-L a

R

2

Fenton-like reaction b

-1

conditions: catalyst loading of 0.12 g L , visible light region from 420 to 630 nm conditions: catalyst loading of 0.12 g L-1, H2O2 dosage of 61.5 mmol L-1, reaction in dark c conditions: catalyst loading of 0.12 g L-1, H2O2 dosage of 61.5 mmol L-1, reaction under illumination of visible light d conditions: catalyst loading of 0.06 g L-1, Oxone dosage of 7.5 mg L-1 e H2O2 only applied into Fenton-like reaction and Photo-Fenton oxidation b

30