- Email: [email protected]
Enhanced Fenton-like catalysis by iron-based metal organic frameworks for degradation of organic pollutants
Journal of Catalysis 356 (2017) 125–132
Contents lists available at ScienceDirect
Journal of Catalysis journal homepage: www.elsevier.com/locate/jcat
Enhanced Fenton-like catalysis by iron-based metal organic frameworks for degradation of organic pollutants Cong Gao, Shuo Chen, Xie Quan ⇑, Hongtao Yu, Yaobin Zhang Key Laboratory of Industrial Ecology and Environmental Engineering (Ministry of Education), School of Environmental Science and Technology, Dalian University of Technology, Linggong Road 2, Dalian 116024, China
a r t i c l e
i n f o
Article history: Received 12 July 2017 Revised 16 September 2017 Accepted 19 September 2017
Keywords: Heterogeneous Fenton-like catalysts Iron-based MOFs MIL-88B-Fe Hydroxyl radicals Mechanism
a b s t r a c t The Fenton reaction is an efficient technology for degrading refractory organic pollutants in water. Heterogeneous Fenton-like catalysts have been demonstrated to be promising alternatives to homogeneous catalysts because of their reusability and lack of sludge production. These catalysts, however, generally show low activity for generating OH due to their limited exposed active sites and difficulty in the reduction of Fe(III) to Fe(II). Here, enhanced catalytic performance was achieved by using an iron-based metal organic framework (MIL-88B-Fe) as a heterogeneous Fenton-like catalyst over a wide pH range (4– 6). The catalytic activity of MIL-88B-Fe was about 1–3 orders of magnitude higher than that of three other conventional catalysts (Fe2O3, a-FeOOH, and Fe3O4) and two other iron-based MOFs (MIL-53-Fe and MIL101-Fe). The superior activity of MIL-88B-Fe could originate from the abundance of active sites, the flexible structure, and facilitated reduction of Fe(III) to Fe(II). Hydroxyl radicals generated from reaction between MIL-88B-Fe and H2O2 were the main reactive oxidative species for phenol degradation. Ó 2017 Elsevier Inc. All rights reserved.
1. Introduction The Fenton reaction is an efficient advanced oxidation process for refractory organic wastewater treatment because of its ability to produce highly reactive hydroxyl radicals (OH, E0 = 2.80 V), which are able to degrade most organic compounds without selectivity. Nevertheless, the traditional homogeneous catalysts, Fe2+/ Fe3+ ions, are difficult to recover and are restricted to acidic conditions (pH 3) [1]. In the past decades, various iron oxides (Fe2O3 and FeOOH) and supported iron catalysts have been applied in heterogeneous Fenton-like reactions to replace dissolved Fe2+/ Fe3+ catalysts, since they can be recycled and the processes can be carried out over a wide pH range [2–4]. These catalysts, however, generally show low activity for generating OH. For heterogeneous Fenton-like reactions, catalytic decomposition of H2O2 into OH mainly contains two steps: initially, the interaction of H2O2 with the active sites; second, reversible electron transfer between H2O2 and active sites (Fe(II)/Fe(III)). In the overall reaction, the reaction rate is strongly dependent on the number of exposed active sites and the reduction of Fe(III) to Fe(II) by H2O2 (0.001– 0.02 M1 s1), which has been demonstrated to be the rate-limiting step in homogeneous and heterogeneous Fentonlike reactions [4]. ⇑ Corresponding author. E-mail address: [email protected] (X. Quan). https://doi.org/10.1016/j.jcat.2017.09.015 0021-9517/Ó 2017 Elsevier Inc. All rights reserved.
In the past decades, porous solid matrices such as activated carbon, zeolites, silica, and clays have been used for immobilizing active iron species to enhance their catalytic activity [5–8]. However, the active species are generally unevenly distributed and the surface area decreases, attributable to pore blockage. The photo-Fenton or electro-Fenton process has been found to accelerate the reduction of Fe(III) to Fe(II) and improve the catalytic performance with the assistance of light or electricity [9,10]. However, the need for specific equipment and energy consumption will increase the cost of these processes. Thus, exploring highly effective heterogeneous Fenton-like catalysts with large numbers of exposed active sites and effective reduction from Fe(III) to Fe(II) is of considerable significance. Iron-based metal organic frameworks (MOFs) are hybrid solids with periodic network structures constructed from metal ions/clusters and organic ligands. The specific textural properties of MOFs, such as wide distribution of single iron sites, porous structures, and large surface areas, endow them with abundant exposed active sites and favorable access of reactants to active sites, making them promising alternatives to traditional heterogeneous catalysts. Recently, MIL-53-Fe and FeII@MIL-100(Fe) have been applied in the Fenton-like reaction for dye degradation [11,12], yet the catalytic efficiency of MOFs still needs to be improved and the reaction mechanism is unclear. Generally, OH is generated from electron transfer between the complex of H2O2 and iron sites. Thus, the formation of the H2O2–iron complex on the surfaces of
126
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
catalysts is the key step in the generation of OH. As a Lewis base, H2O2 tends to adsorb onto the Lewis acid sites. Therefore, it is deduced that the iron, coordinatively unsaturated or occupied by some replaceable ligands such as water, is the active site on the surfaces of heterogeneous Fenton-like catalysts. Recently, open iron sites, which are accessible and coordinatively unsaturated, were found in MIL-88B-Fe [13]. MIL-88B-Fe, a 3D porous MOF, is constructed from 1,4-benzenedicarboxylic acid (BDC) and trimeric Fe octahedral (Fe3-l3-oxo) clusters. Within the trimeric (Fe3-l3-oxo) clusters, Fe atoms exhibit an octahedral environment with open iron sites, which are terminated by some nonbridging ligands such as water or halogen/hydroxide anionic ligands [13]. Previous studies revealed that the nonbridging ligands could be replaced by Lewis bases [14,15]. Thus, H2O2 can be adsorbed onto the iron sites of MIL-88B-Fe by displacing these ligands, which is the essential step in the Fenton-like reaction. The electron-rich organic ligands (BDC) could donate electrons to the center Fe ions [16], which may increase the redox potential of Fe(II)/Fe(III) and thus enhance reduction of Fe(III) to Fe(II) [17]. Moreover, MIL-88B-Fe undergoes a very large swelling (85% increase of its cell volume) upon exposure to polar solvents such as water and can maintain its open-framework topology [18]. This property is favorable for reactant transfer in MIL-88B-Fe. From these advantages, MIL-88B-Fe is expected to be a promising Fenton-like catalyst. In this work, MIL-88B-Fe with chemical composition Fe3O[C6H4(CO2)2]3XnH2O (X = Cl or OH) was employed as a heterogeneous Fenton-like catalyst. The influence of reaction parameters (e.g., pH and catalyst and H2O2 concentration) on the catalytic activity of MIL-88B-Fe and the H2O2 catalytic decomposition mechanism was investigated. Moreover, the catalytic activity of MIL-53-Fe and MIL-101-Fe, which contain the same organic ligand but different structures from MIL-88B-Fe, was tested to explore the relationship between the catalytic activity of iron-based MOFs and their structures.
2. Materials and methods 2.1. Chemicals and materials 1,4-Benzenedicarboxylic acid (BDC), benzoic acid, Fe2O3,
a-FeOOH, and Fe3O4 were obtained from the Aladdin Industrial
Corporation (China). H2O2 (30%), FeCl36H2O, N,N-dimethyl formamide, methanol, isopropyl alcohol, Na2SO4, NaOH, and H2SO4 were purchased from the Damao Chemical Reagent Factory (Tianjin, China) and the Fuyu Chemical Co., Ltd. (Tianjin, China). T-Butyl alcohol (TBA) was bought from the Bodi Chemical Co., Ltd. (Tianjin, China). All chemicals were analytical grade and used without further purification. Ultrapure water (18.2 MX) was used throughout the experiments.
2.2. Synthesis of catalyst MIL-88B-Fe was prepared by a solvothermal method according to the procedure in the literature [19]: briefly, hydrothermal treatment of FeCl36H2O (270 mg) and 1,4-benzenedicarboxylic acid (116 mg) in N,N-dimethyl formamide (5 mL) with NaOH (2 M, 0.4 mL) at 100 °C for 12 h. After solvothermal treatment, the assynthesized MIL-88B-Fe was collected by filtration and washed with DMF, methanol, and water at 50 °C until the supernatant became colorless. Finally, it was activated overnight at 110 °C. The obtained catalysts were stored in a glass desiccator.
2.3. Characterization of catalyst Powder X-ray diffraction (XRD) of catalysts was performed on an EMYPREAN diffractometer with Cu Ka radiation (k = 1.54056 Å) over a 2h range of 5°–40°. Scanning electron microscopy (SEM) images were obtained on an S-4800 type SEM (Hitachi, Japan) and transmission electron microscopy (TEM) images were obtained on a Tecnai G2 F30 S-Twin type TEM (FEI Company, USA). The Brunauer–Emmett–Teller (BET) specific surface area was determined by nitrogen adsorption in Quantachrome SI (USA) at 77 K. The thermal stability was measured on a TG-DTG 6300 thermogravimetric analyzer (EXSTAR, Japan) by heating samples to 600 °C at a rate of 5 °C/min in 60 mL/min air flow. The organic groups on the surfaces of catalysts were investigated by infrared spectroscopy (IR), which was recorded on a Bruker VERTEX 70 FTIR spectrometer in the wavenumber range from 4000 to 500 cm1 with a resolution of 4 cm1 at room temperature. The coordinatively unsaturated irons in catalysts were tested by CO adsorption infrared spectra [20]. Elemental analysis of catalysts and the concentrations of leached irons during Fenton reaction were tested by inductively coupled plasma optical emission spectroscopy (ICP, Perkin Elmer Optima 2000, USA). The elemental valence state of catalysts was investigated by X-ray photoelectron spectroscopy, which was performed on an ESCALAB 250XI (Thermo Fisher Scientific, USA). The C1s peak at 284.6 eV was used to calibrate peak positions. 2.4. Catalytic activity test Phenol was selected as a target pollutant to evaluate the catalytic activity of MIL-88B-Fe, as it is a ubiquitous pollutant and an important intermediate in industrial processes. The degradation experiments were carried out in a batch mode using a threenecked flask (250 mL) at room temperature. The pH was adjusted with NaOH (0.1 M) or H2SO4 (0.1 M). In a typical experiment, MIL-88B-Fe was dispersed in phenol solution (50 mg/L, 150 mL) at a specific pH by sonication for 2 min. Then the degradation reaction was initiated by adding H2O2 under magnetic stirring. At predetermined time intervals, samples were withdrawn and filtered through 0.22 lm membrane filters to remove suspended MOFs. Meanwhile, an aliquot of 1 M isopropyl alcohol was immediately added to quench the reaction, and then the concentration of phenol was analyzed. After reaction, MIL-88B-Fe was separated by filtration, washed with water and methanol, and then reused in a new reaction after being dried at 110 °C. This process was repeated several times to test the reusability of MIL-88B-Fe. The performance of H2O2 oxidation without MIL-88B-Fe and the amount of phenol adsorbed by MIL-88B-Fe were tested as control experiments. Each degradation experiment was run in triplicate. The reported data are arithmetic means of three measured values. 2.5. Analytical methods The concentration of phenol was analyzed on a highperformance liquid chromatograph (Agilent 1200) equipped with a diode array detector (DAD) and a C18 reversed-phase column (5 lm, 4.6 mm 150 mm). The mobile phase was a mixture of methanol and water (55:45, v/v) at a flow rate of 1.0 mL/min with a column temperature of 30 °C, and the analytical wavelength was 270 nm. The total organic carbon (TOC) was measured using a multi N/C 2100S (Analytikjena, Germany). The main reactive oxidative species generated in the systems were tested by electron paramagnetic resonance (EPR) and radical quencher experiments. For the EPR test, 20 lL of 0.1 M DMPO was added rapidly into a 1 mL sample after filtration to form a DMPO–radicals adduct. The EPR spectra were obtained on a Bruker 300E spectrometer
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
(Germany). The generation rates of OH were analyzed by monitoring the oxidation rates of the benzoic acid heterogeneous Fentonlike reaction [14,21]. The information for ATR-FTIR, in situ Raman spectroscopy, and linear sweep voltammetry (LSV) characterizations are shown in the Supporting Information. 3. Results and discussion 3.1. Characterization As shown in Fig. 1A, the XRD pattern of as-synthesized MIL88B-Fe was well matched with the [0 0 2], [1 0 1], [1 0 3], [2 0 2], and [2 1 1] diffraction planes of MIL-88B-Fe [18,22,23]. The assynthesized MIL-88B-Fe was of high purity, as no diffraction peaks of other substances were found. It should be noted that the slight differences of peak positions and their intensity were possible because of the structural flexibility [24]. The SEM and TEM images of the as-synthesized MIL-88B-Fe demonstrated a needlelike morphology with a uniform size of 0.8–1 lm in length and 100 nm in width (Fig. 1B and C). The as-synthesized MIL-88B-Fe gave a BET surface area of 165.4 m2 g1 with a pore volume of 0.2 cm3 g1. As shown in Fig. S1B in the Supporting Information, the IR band at 1656 or 1660 cm1 representing DMF in MIL-88B-Fe was scarcely found, indicating that only water was trapped in MIL-88B-Fe after activation [25,26]. The TG curve revealed that MIL-88B-Fe possessed good thermal stability. The structure collapsed (sharp weight loss) when the temperature was above 350 °C. As shown in Fig. S1C, the weight loss in the range 25–150 °C was 3% (the departure of coordinated water), which was lower than the reported 5–15% loss [26,27]. It was caused by partial water removal during the activation process (110 °C for 12 h after preparation). 3.2. Catalytic activity of MIL-88B-Fe Control experiments were conducted to compare the efficiency of phenol removal by various processes. As can be seen from Fig. S2, in the absence of MIL-88B-Fe, phenol was hardly degraded by the H2O2 oxidation process. With 0.1 g/L MIL-88B-Fe alone, the removal of phenol was fast in the first 5 min and then reached a steady state of 11% removal efficiency after reaction for 30 min, which was attributed to the adsorption of phenol onto the surface of MIL-88B-Fe. With the simultaneous presence of 0.1 g/L MIL-88BFe and 15 mM H2O2, 99% of phenol was removed. It could be concluded that MIL-88B-Fe possessed catalytic activity for Fenton-like reactions.
127
Since the performance of Fenton-like reactions is dependent on the initial solution pH values, the temperature, and the concentrations of catalysts and H2O2, the influence of these parameters on the degradation of phenol in MIL-88B-Fe catalyzed Fenton-like reactions was investigated. The removal of phenol followed a pseudo-first-order reaction in kinetics under all the tested experimental conditions. As shown in Fig. 2A and S3A, the removal of phenol was faster at lower pH values than at higher pH values. The lower removal efficiency of phenol at higher pH values could be ascribed to the auto-decomposition of H2O2 into oxygen and water (2 H2O2 ? 2 H2O + O2) and the lower oxidation potential of OH at higher pH values (pH 0: 2.8 V; pH 7: 1.9 V) [28]. The pseudo-first-order kinetic constant (k) of phenol degradation at pH 4 was 0.15 min1, which had a level of magnitude similar to that of the homogeneous Fenton reaction reported by Esplugas et al. [29]. Moreover, at near-neutral pH, MIL-88B-Fe demonstrated good catalytic activity, while the homogeneous and some previously reported heterogeneous Fenton-like catalysts could hardly work [3,30,31]. The pseudo-first-order kinetic constant of phenol degradation at pH 5 by a MIL-88B-Fe catalyzed Fenton-like reaction was one order of magnitude higher than those for many reported heterogeneous Fenton-like catalysts, as shown in Table S1. The effect of H2O2 concentration on phenol degradation is illustrated in Figs. 2B and S3B. When the H2O2 concentration was increased from 5 to 15 mM, k increased from 0.07 to 0.15 min1 and the removal efficiency of phenol was increased from 90 to 99%. The enhanced removal of phenol might be attributed to the increased H2O2 concentration leading to more OH generation in the Fenton-like reaction. However, further increasing the H2O2 concentration (20 mM) resulted in reduction of k (0.11 min1) and the removal efficiency of phenol (97%). This might be related to the scavenging of OH by excessive H2O2 to form less reactive species such as hydroperoxyl radicals, which could hardly degrade organic pollutants [30]. The effect of catalyst dosage on phenol degradation was also studied. As shown in Figs. 2C and S3C, with increasing dosage of catalyst from 0.05 to 0.15 g/L, both the removal efficiency of phenol and the k value increased (86–99%, 0.06–0.17 min1), which might be ascribed to the fact that the increasing number of active sites accelerated the production of OH. Nevertheless, the phenol degradation was not enhanced but slightly decreased (98% phenol removal with k of 0.14 min1) when the catalyst dosage further increased to 0.2 g/L. This phenomenon was probably attributable to the agglomeration of catalysts, which limited the diffusion of reactants to the surfaces of catalysts (Fig. S4). Moreover, the excess
Fig. 1. XRD pattern (A) and SEM (B) and TEM images (C) of as-synthesized MIL-88B-Fe.
128
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
Fig. 2. Factorial effects of reaction conditions on the catalytic degradation of phenol: initial pH value (A), H2O2 dosage (B), catalyst dosage (C), and temperature (D). Except for the investigated parameters, other parameters were fixed: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20°C.
of catalysts added may scavenge OH, which reduced the amount of OH and hindered the degradation of phenol [32]:
OH + BFe(II) ! OH + BFe(III).
ð1Þ
As shown in Figs. 2D and S3D, a higher reaction rate was achieved at an elevated temperature. k at 20, 30, and 40 °C was 0.15, 0.16, and 0.17 min1, respectively. The catalytic activity of as-synthesized MIL-88B-Fe was compared with that of the other three conventional heterogeneous Fenton-like catalysts (Fe2O3, a-FeOOH, and Fe3O4). As shown in Fig. 3A and Table S2, the pseudo-first-order kinetic rate constant of phenol degradation in the MIL-88B-Fe catalyzed Fenton-like reaction was 1–3 orders of magnitude higher than that with Fe3O4, Fe2O3, or a-FeOOH. The mineralization of phenol was also
measured. As shown in Fig. 3B, 44% of TOC was removed in the MIL-88B-Fe catalyzed Fenton-like reaction after 30 min, which was much higher than those for the conventional catalysts (<5% TOC removal). Considering that the organic ligands in MIL-88B-Fe may influence its redox properties, the LSV measurement was performed to measure the reduction potential of MIL-88B-Fe. As shown in Fig. 3C, compared with the traditional catalysts, higher reduction peaks representing reduction from Fe(III) to Fe(II) were observed on MIL-88B-Fe, indicating that the reduction of Fe(III) was easier in MIL-88B-Fe than in the traditional Fenton catalysts. Moreover, the reduction potential of Fe3+ ions was tested. The reduction potential of Fe3+ ions was 0.14 V, which was much lower than that for MIL-88B-Fe (0.32 V), implying that the reduction of MIL-88B-Fe was easier. These phenomena could be attributed to
Fig. 3. Degradation of phenol (A), TOC removal (B) in Fe2O3, a-FeOOH, and Fe3O4 catalyzed Fenton-like reactions, and (C) LSV collected with MIL-88B-Fe, Fe2O3, a-FeOOH, Fe3O4, and Fe3+ (the concentration of Fe in MIL-88B-Fe) in N2-saturated 0.1 M Na2SO4 electrolyte at room temperature. Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20 °C.
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
the centering of the organic ligands donor electrons to Fe, leading to enhanced reduction of Fe(III) to Fe(II). These results indicated that MIL-88B-Fe was promising as a catalyst for Fenton-like reactions. 3.3. Stability and reusability of MIL-88B-Fe Fig. 4A displays the concentrations of the leached iron ions from MIL-88B-Fe versus reaction time at pH 4–5. The concentration of leached iron increased with reaction time. The highest concentration of leached iron at pH 4 was 0.68 mg/L at 30 min, which was below the environmental standard (2 mg/L) imposed by the European Union. To exclude the possibility that the observed catalytic activity was caused by the leaching iron ions, 0.1 g/L MIL-88B-Fe was added into 50 mg/L phenol solution at pH 4 for 30 min, and then H2O2 was added to the solution after the MIL-88B-Fe was removed by filtration. As shown in Fig. S5, the phenol degradation catalyzed by the leached iron ions was negligible. Thus, the catalytic activity of MIL-88B-Fe was dominated by heterogeneous catalysis rather than homogeneous catalysis resulting from the leached iron. Moreover, the used MIL-88B-Fe after the reaction was characterized by SEM, XRD, and IR (Figs. S6 and S7). The XRD patterns, morphology, and IR spectra of fresh and used catalysts revealed no obvious change. To evaluate the reusability of MIL88B-Fe, four successive cycles of degradation testing were performed. As shown in Fig. 4B, the removal efficiency of phenol remained almost constant in every run, indicating the excellent stability of MIL-88B-Fe. The concentration of leaching iron ions after each cycle was measured. As shown in Fig. S8A, the concentration of leached iron after each run was below the environmental standard (2 mg/L) imposed by the European Union. Moreover, the catalytic performance of the leached iron was investigated (Fig. S8B). It was found that the leached iron exhibited negligible catalytic activity. These results suggested that MIL-88B-Fe was stable enough as a heterogeneous Fenton-like catalyst. 3.4. Reactive oxidative species and mechanism discussion The reactive oxidative species (ROS) generated in MIL-88B-Fe catalyzed Fenton-like reactions were measured by an EPR spintrap technique (with DMPO). The EPR spectrum of the DMPOROS adducts generated during the Fenton-like reaction at 5 min is shown in Fig. 5A. A four-fold characteristic peak with an intensity ratio of 1:2:2:1 that belonged to the DMPO–OH signal was found [15]. However, the DMPO–OH signal was not observed in the presence of 300 mM TBA (OH scavenger). Thus, OH was the main ROS generated in MIL-88B-Fe catalyzed Fenton-like reac-
129
tions. The effect of OH on phenol degradation was measured by determining the influence of TBA as OH scavenger. As shown in Fig. 5B, the removal efficiency of phenol declined from 99 to 83% in the presence of 10 mM TBA at 30 min, and it continuously decreased to 5% when the concentration of TBA was further increased to 300 mM. This result indicated that OH played a key role in phenol degradation in MIL-88B-Fe catalyzed Fenton-like reactions. To investigate the active sites in MIL-88B-Fe, the catalytic activity of MOF-5-Zn (characterization results shown in Figs. S9 and S10) that is constructed from the same organic ligands (BDC) as MIL-88B-Fe but with a different metal (Zn, which is unable to catalyze Fenton-like reactions) in Fenton-like reactions was studied. As can be seen in Fig. 6A, the concentration of phenol was almost unchanged in 30 min, indicating that MOF-5 could not catalyze Fenton-like reactions. The result suggested the iron clusters were the active sites in MIL-88B-Fe. This conclusion was further confirmed by the low efficiency of phenol removal (2%) in the organic ligands (BDC) catalyzed Fenton-like reaction (Fig. S11). Phosphate (a Lewis base, pKa(H2O2) < pKa(PO3 4 )), which has a stronger affinity to the surface metal sites (coordinatively unsaturated metal sites) than H2O2, water, and some other nonbridging anionic ligands (Cl1 or OH1), was added to MIL-88B-Fe catalyzed Fenton reactions [33]. As shown in Fig. 6A, the degradation of phenol was totally inhibited by the addition of phosphate, suggesting that combination of H2O2 with metal sites was probably hindered by phosphate. To confirm this hypothesis, ATR-FTIR spectroscopy was employed to detect the adsorption of H2O2 on MIL-88B-Fe in the presence of phosphate. As shown in Fig. 6B, no significant new peaks that can be ascribed to the m(OAO) wavenumber (1400, 1195, and 1080 cm1) appeared after addition of H2O2 [34,35]. Therefore, it could be concluded that the coordinatively unsaturated iron atoms were the active sites and their direct contact with H2O2 was the essential step in the MIL-88B-Fe catalyzed Fenton-like catalytic reaction. In addition, the interaction of H2O2 and catalysts was investigated by in situ Raman spectroscopy. As can be seen in Fig. 6C, before addition of H2O2, the suspensions of MIL-88B-Fe showed a band at 865 cm1, which was attributed to the out-of-plane deformation modes of the CAH bond [36]. After the addition of H2O2, a new band located at 875 cm1 that could be assigned to the OAO stretching of peroxo complexes appeared, indicating that H2O2 was coordinated to the iron sites of MIL-88B-Fe [37,38]. With increasing reaction time, the above-mentioned band was gradually decreased, suggesting that H2O2 was decomposed on MIL-88B-Fe. XPS characterization was conducted to investigate the atomic composition and chemical state of the iron species on the surface
Fig. 4. Variation of leached iron concentration during reaction at pH = 4–5 (A) and reusability of MIL-88B-Fe in Fenton-like reaction (B). Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20 °C.
130
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
Fig. 5. DMPO-ROS adducts generated from MIL-88B-Fe catalyzed Fenton-like reactions at 5 min (A) and influence of radical scavenger (TBA) concentration on the catalytic degradation of phenol (B). Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, [TBA] = 300 mM, and 20 °C.
Fig. 6. Degradation of phenol in MOF-5 and MIL-88B-Fe catalyzed Fenton-like reactions in the presence of 10mM phosphate (A), ATR-FTIR spectra of MIL-88B-Fe (B), in situ Raman spectra of MIL-88B-Fe with H2O2 (C), and Fe2p XPS survey spectra of fresh and used MIL-88B-Fe (D). Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20 °C.
of MIL-88B-Fe. As shown in Table S3, 1.65% of Cl existed on MIL88-Fe, which was attributed to the nonbridging chlorine anionic ligands on Fe clusters in MIL-88B-Fe. After reaction, the atomic ratio of Cl was decreased to 0.15%. Considering the weak interaction between Cl1 and Fe, it was speculated that Cl1 could be replaced by H2O2 in the reaction [39]. To confirm this hypothesis, XPS spectra of MIL-88B-Fe after addition of H2O2 for 10 min were tested. The ratio of Cl in MIL-88B-Fe after addition of H2O2 was changed to 0.21%, which was similar to the 0.15% of Cl in MIL-88B-Fe after reaction, indicating that the nonbridging halogen
anionic ligand (Cl1) on Fe clusters could be replaced with H2O2 during reaction. As generally recognized, electron exchange between Fe(II)/Fe (III) and H2O2 induces OH generation in homogeneous and heterogeneous Fenton-like reactions [40]. The chemical states of the iron species on MIL-88B-Fe before and after reaction were measured. Fig. 6D illustrates XPS spectra of Fe2p in fresh and used MIL-88BFe. The fresh MIL-88B-Fe displays the spin–orbit with binding energy (BE) of 725.7 eV (Fe2p1/2) and 711.8 eV (Fe2p3/2). The Fe2p3/2 peak could be fitted into three contributions at 711.3,
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
131
3.5. The relationship between the catalytic activity and the structures of iron-based MOFs
Fig. 7. Schematic diagram of the reaction mechanism of H2O2 activation by MIL88B-Fe.
712.5, and 713.8 eV, respectively, indicating that only Fe(III) existed on fresh MIL-88B-Fe. The weak shake-up satellite peak at about 717.4 eV, which is observed in the Fe2p spectrum, further confirmed that iron was predominantly in the Fe(III) state. After the Fenton-like reaction, the BE of Fe2p was moved slightly to lower BE values (724.0 and 710.9 eV). Deconvolution of the Fe2p3/2 peak of used MIL-88B-Fe after 10 min showed that Fe(II) appeared on MIL-88B-Fe after reaction, indicating that some Fe (III) was reduced to Fe(II) during the Fenton-like reaction. According to the above discussion, the reaction mechanism was shown in Fig. 7: first, H2O2 was attached to the iron sites, as H2O2 is a strong Lewis base that has stronger affinity to iron sites than water and nonbridging anionic ligands,
BFe(III)— L + H2 O2 ! BFe(III) (H2 O2 ) + L
ð2Þ
ðL ¼ H2 O or nonbridging anionic ligandsÞ; and then the electron/charge was transferred between H2O2 and MIL-88B-Fe, resulting in the reduction of Fe(III) to Fe(II) on MIL88B-Fe:
BFe(III) (H2 O2 ) ! BFe(II) + O2 H + Hþ .
ð3Þ
Next, the generated surface Fe(II) reacted with H2O2 to form OH:
BFe(II) + H2 O2 ! BFe(III) + OH + OH .
ð4Þ
BFe(III) + O2 H ! BFe(II) + O2 + Hþ .
ð5Þ
Eventually, organic pollutants were oxidized by surface-bonded OH or some diffused OH in the bulk solution:
Phenol +
OH / ! CO2 + H2 O.
ð6Þ
To explore the effect of structure of iron clusters on the Fenton catalytic activity of iron-based MOFs, the Fenton-like catalytic activity of MIL-53-Fe, which consists of the same organic ligand as MIL-88B-Fe but possesses different structures of iron clusters, was tested (detailed characterizations are given in Figs. S12– S15). As shown in Fig. 8A and B, only 11% of phenol was degraded after 30 min and the kinetic constant of benzoic acid degradation was 2.36 104 min1, which were obviously much lower than those for MIL-88B-Fe. These results may be owing to the different structures of iron clusters in MOFs. CO adsorption infrared spectra were applied to determine the amount of coordinatively unsaturated iron sites of the MIL-88B-Fe and MIL-53-Fe [20]. The intensity of the peak around 2172–2179 cm1 (CO adsorption on the coordinatively unsaturated iron sites in MOFs) was proportional to the amount of coordinatively unsaturated iron in catalysts. As shown in Fig. 8C, after CO adsorption for 20 min, the intensity of the signal around 2172–2179 cm1 of MIL-88B-Fe is higher than that of MIL-53-Fe. It could be deduced that the number of coordinatively unsaturated irons in MIL-88B-Fe was much higher than for MIL-53-Fe. Thus, the excellent Fenton catalytic activity of MIL-88B-Fe was related to the large number of coordinatively unsaturated iron sites. The catalytic activity of MIL-101-Fe, which consists of the same building units as MIL-88B-Fe but possesses a rigid topology (does not swell upon exposure to polar solvents), was tested to explore the effect of the flexibility of iron-based MOFs on the Fenton-like catalytic activity. The detailed characterization results for MIL101 are given in Figs. S10–S13. As shown in Fig. 8A and B, 62% of phenol was degraded within 30 min and the benzoic acid degradation kinetic constant was 1.54 103 min1, demonstrating that the catalytic activity of MIL-101-Fe was lower than that of MIL88-Fe. According to the characterizations of BET and CO-FTIR, the surface area, pore size, and number of coordinatively unsaturated iron sites for MIL-101-Fe were similar to those of MIL-88B-Fe, implying that the difference in catalytic activity between MIL-101-Fe and MIL-88B-Fe was not related to these mentioned structural characters, while other factors may affect the catalytic performance of iron-based MOFs (Table S4 and Fig. 8C). In view of the flexible behavior of MIL-88B-Fe, which possesses a very large volume swelling (increasing by 85% of its cell volume from 1135 to 2110 Å3) upon exposure to polar solvents, it is more favorable for reactant transfer in MIL-88B-Fe than in MIL-101-Fe. Thus, the superior activity of MIL-88B-Fe was probably associated with the greater accessibility of the active sites with H2O2, which resulted from its flexible topology. Based on the above results, the catalytic
Fig. 8. Catalytic performance of MIL-53-Fe, MIL-88B-Fe, and MIL-101-Fe: phenol degradation (A), kinetics of benzoic acid catalytic degradation (B), and CO adsorption infrared spectra of MIL-53-Fe, MIL-88B-Fe, and MIL-101-Fe (catalyst 0.1 g, v(CO) 50 mL/min) (C). Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20 °C.
132
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
activity of iron-based MOFs was strongly dependent on the number and accessibility of the coordinatively unsaturated irons. 4. Conclusions In summary, MIL-88B-Fe exhibited high activity and stability with a wide operating range of pH (46) in Fenton-like reactions. The catalytic activity of MIL-88B-Fe was better than that of three conventional catalysts (Fe2O3, a-FeOOH, Fe3O4) and was even comparable to that of the homogenous Fenton catalyst. The superior activity of MIL-88B-Fe was mainly attributed to the abundance of coordinatively unsaturated irons, the flexible structure, and the enhanced Fe(III)/Fe(II) redox properties. This work could provide new insights into the rational design and development of highly effective Fenton-like catalysts. Acknowledgments This work was supported financially by the National Natural Science Foundation of China (No. 21590813) and the Programme of Introducing Talents of Discipline to Universities (B13012). Conflict of interest The authors declare no competing financial interests. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.jcat.2017.09.015. References [1] A. Dhakshinamoorthy, S. Navalon, M. Alvaro, H. Garcia, Chem. Sust. Chem. 5 (2012) 46–64. [2] N. Panda, H. Sahoo, S. Mohapatra, J. Hazard. Mater. 185 (2011) 359–365. [3] L. Xu, J. Wang, Appl. Catal. B Environ. 123–124 (2012) 117–126. [4] A.M. Mesquita, I.R. Guimarães, G.M.M. de Castro, M.A. Gonçalves, T.C. Ramlho, M.C. Guerreiro, Appl. Catal. B Environ. 192 (2016) 286–295. [5] M.H. Do, N.H. Phan, T.D. Nguyen, T.T. Suong Pham, V.K. Nguyen, T.T. Trang Vu, T.K. Phuong Nguyen, Chemosphere 85 (2011) 1269–1276. [6] H. Xin, A. Koekkoek, Q. Yang, R. van Santen, C. Li, Emie J.M. Hensen, Chem. Commun. 48 (2009) 7590–7592. [7] L. Xiang, S. Royer, H. Zhang, J.-M. Tatibouët, J. Barrault, S. Valange, J. Hazard. Mater. 172 (2009) 1175–1184.
[8] M. Bobu, A. Yediler, I. Siminiceanu, S. Schulte-Hostede, Appl. Catal. B Environ. 83 (2008) 15–23. [9] Y. Baba, T. Yatagai, T. Harada, Y. Kawase, Chem. Eng. J. 277 (2015) 229–241. [10] E. Brillasm, I. Sirés, M.A. Oturan, Chem. Rev. 109 (2009) 6570–6631. [11] L.H. Ai, C.H. Zhang, L.L. Li, J. Jiang Appl. Catal. B Environ. 148–149 (2014) 191– 200. [12] H.L. Lv, H.Y. Zhao, T.C. Cao, L. Qian, Y.B. Wang, G.H. Zhao, J. Mol. Catal. A Chem. 400 (2015) 81–89. [13] M. Ma, H. Noei, B. Mienert, J. Niesel, E. Bill, M. Muhler, R.A. Fischer, Y. Wang, U. Schatzshneider, N. Metzler-Nolte, Chem. Eur. J. 19 (2013) 6785–6790. [14] S. Huo, X. Yan, J. Mater. Chem. 22 (2012) 7449–7455. [15] X. Yang, X. Xu, J. Xu, Y. Han, J. Am. Chem. Soc. 135 (2013) 16058–16061. [16] E. Galardon, S. Ramdeehul, J.M. Brown, A. Cowley, K.K. Hii, A. Jutand, Angew. Chem. Int. Ed. 41 (2002) 1760. [17] J.K. Li, S. Ghoshal, W.T. Liang, M.T. Sougrati, F. Jaouen, B. Halevi, S. Mckinney, G. McCool, C.R. Ma, X.X. Yuan, Z.F. Ma, S. Mukerijee, Q.Y. Jia, Energy Environ. Sci. 9 (2016) 2418–2432. [18] C. Mellot-Draznieks, C. Serre, S. Surblé, N. Audebrand, G. Férey, J. Am. Chem. Soc. 127 (2005) 16273–16278. [19] P. Horcajada, F. Salles, S. Wuttke, T. Devic, D. Heurtaux, G. Maurin, A. Vimont, M. Daturi, O. David, E. Magnier, N. Stock, Y. Filinchuck, D. Popov, C. Riekel, G. Férey, C. Serre, J. Am. Chem. Soc. 133 (2011) 17839–17847. [20] F. Vermoortele, R. Ameloot, L. Alaerts, R. Matthessen, B. Carlier, E.V. Ramos Fernandez, J. Gascon, F. Kapteijn, D.E. De Vos, J. Mater. Chem. 22 (2012) 10313– 10321. [21] W.P. Kwan, B.M. Voelker, Environ. Sci. Technol. 37 (2003) 1150–1158. [22] M. Pham, G. Vuong, A. Vu, T. Do, Langmuir 27 (2011) 15261–15267. [23] K.G.M. Laurier, F. Vermoortele, R. Ameloot, D.E. De Vos, J. Hofkens, M.B.J. Roeffaers, J. Am. Chem. Soc. 135 (2013) 14488–14491. [24] C. Scherb, A. Schödel, T. Bein, Angew. Chem. Int. End. 47 (2008) 5777–5779. [25] G. Vuong, M. Pham, T. Do, Cryst. Eng. Comm. 15 (2013) 9694–9703. [26] L. Paseta, B. Seoane, D. Julve, V. Sebastián, C. Téllez, J. Coronas, A.C.S. Appl, Mater. Interfaces 5 (2013) 9405–9410. [27] N.A. Ramsahye, T. Trung, L. Scott, F. Nouar, T. Devic, P. Horcajada, E. Magnier, O. David, C. Serre, P. Trens, Chem. Mater. 25 (2013) 479–488. [28] N.K. Daud, B.H. Hameed, J. Hazard. Mater. 176 (2010) 938–944. [29] S. Esplugas, J. Giméne, S. Contreras, E. Pascual, M. Rodríguez, Water Res. 36 (2002) 1034–1042. [30] L. Xu, J. Wang, J. Hazard. Mater. 186 (2011) 256–264. [31] J. Wang, Z. Yao, M. Yang, Y. Wang, Q. Xia, Z. Jiang, Environ. Sci. Pollut. Res. 23 (2016) 14927–14936. [32] A. Khataee, P. Gholami, M. Sheydaei, S. Khorram, S.W. Joo, New J. Chem. 40 (2016) 5221–5230. [33] T. Zhang, J. Ma, J. Mol. Catal. A. Chem. 279 (2008) 82–89. [34] Y. Nie, L. Zhang, Y. Li, C. Hu, J. Hazard. Mater. 294 (2015) 195–200. [35] K. Kunimatsu, T. Yoda, D.A. Tryk, H. Uchida, M. Watanabe, Phys. Chem. Chem. Phys. 12 (2010) 621–629. [36] D. Saha, S. Deng, J. Colloid, Interface Sci. 348 (2010) 615–620. [37] L. Lyu, L. Zhang, Q. Wang, Y. Nie, C. Hu, Environ. Sci. Technol. 49 (2015) 8639– 8647. [38] S. Bordiga, A. Damin, F. Bonino, G. Ricchiardi, C. Lamberti, A. Zecchina, Angew. Chem. 41 (2002) 4734–4737. [39] J. Zhang, X.L. Jiao, Y.G. Xia, F.F. Liu, Y.P. Pang, X.F. Zhao, D.R. Chen, Chem. Eur. J. 22 (2016) 9321–9329. [40] X. Hu, B. Liu, Y. Deng, H. Chen, S. Luo, C. Sun, P. Yang, S. Yang, Appl. Catal. B Environ. 107 (2011) 274–283.
Contents lists available at ScienceDirect
Journal of Catalysis journal homepage: www.elsevier.com/locate/jcat
Enhanced Fenton-like catalysis by iron-based metal organic frameworks for degradation of organic pollutants Cong Gao, Shuo Chen, Xie Quan ⇑, Hongtao Yu, Yaobin Zhang Key Laboratory of Industrial Ecology and Environmental Engineering (Ministry of Education), School of Environmental Science and Technology, Dalian University of Technology, Linggong Road 2, Dalian 116024, China
a r t i c l e
i n f o
Article history: Received 12 July 2017 Revised 16 September 2017 Accepted 19 September 2017
Keywords: Heterogeneous Fenton-like catalysts Iron-based MOFs MIL-88B-Fe Hydroxyl radicals Mechanism
a b s t r a c t The Fenton reaction is an efficient technology for degrading refractory organic pollutants in water. Heterogeneous Fenton-like catalysts have been demonstrated to be promising alternatives to homogeneous catalysts because of their reusability and lack of sludge production. These catalysts, however, generally show low activity for generating OH due to their limited exposed active sites and difficulty in the reduction of Fe(III) to Fe(II). Here, enhanced catalytic performance was achieved by using an iron-based metal organic framework (MIL-88B-Fe) as a heterogeneous Fenton-like catalyst over a wide pH range (4– 6). The catalytic activity of MIL-88B-Fe was about 1–3 orders of magnitude higher than that of three other conventional catalysts (Fe2O3, a-FeOOH, and Fe3O4) and two other iron-based MOFs (MIL-53-Fe and MIL101-Fe). The superior activity of MIL-88B-Fe could originate from the abundance of active sites, the flexible structure, and facilitated reduction of Fe(III) to Fe(II). Hydroxyl radicals generated from reaction between MIL-88B-Fe and H2O2 were the main reactive oxidative species for phenol degradation. Ó 2017 Elsevier Inc. All rights reserved.
1. Introduction The Fenton reaction is an efficient advanced oxidation process for refractory organic wastewater treatment because of its ability to produce highly reactive hydroxyl radicals (OH, E0 = 2.80 V), which are able to degrade most organic compounds without selectivity. Nevertheless, the traditional homogeneous catalysts, Fe2+/ Fe3+ ions, are difficult to recover and are restricted to acidic conditions (pH 3) [1]. In the past decades, various iron oxides (Fe2O3 and FeOOH) and supported iron catalysts have been applied in heterogeneous Fenton-like reactions to replace dissolved Fe2+/ Fe3+ catalysts, since they can be recycled and the processes can be carried out over a wide pH range [2–4]. These catalysts, however, generally show low activity for generating OH. For heterogeneous Fenton-like reactions, catalytic decomposition of H2O2 into OH mainly contains two steps: initially, the interaction of H2O2 with the active sites; second, reversible electron transfer between H2O2 and active sites (Fe(II)/Fe(III)). In the overall reaction, the reaction rate is strongly dependent on the number of exposed active sites and the reduction of Fe(III) to Fe(II) by H2O2 (0.001– 0.02 M1 s1), which has been demonstrated to be the rate-limiting step in homogeneous and heterogeneous Fentonlike reactions [4]. ⇑ Corresponding author. E-mail address: [email protected] (X. Quan). https://doi.org/10.1016/j.jcat.2017.09.015 0021-9517/Ó 2017 Elsevier Inc. All rights reserved.
In the past decades, porous solid matrices such as activated carbon, zeolites, silica, and clays have been used for immobilizing active iron species to enhance their catalytic activity [5–8]. However, the active species are generally unevenly distributed and the surface area decreases, attributable to pore blockage. The photo-Fenton or electro-Fenton process has been found to accelerate the reduction of Fe(III) to Fe(II) and improve the catalytic performance with the assistance of light or electricity [9,10]. However, the need for specific equipment and energy consumption will increase the cost of these processes. Thus, exploring highly effective heterogeneous Fenton-like catalysts with large numbers of exposed active sites and effective reduction from Fe(III) to Fe(II) is of considerable significance. Iron-based metal organic frameworks (MOFs) are hybrid solids with periodic network structures constructed from metal ions/clusters and organic ligands. The specific textural properties of MOFs, such as wide distribution of single iron sites, porous structures, and large surface areas, endow them with abundant exposed active sites and favorable access of reactants to active sites, making them promising alternatives to traditional heterogeneous catalysts. Recently, MIL-53-Fe and FeII@MIL-100(Fe) have been applied in the Fenton-like reaction for dye degradation [11,12], yet the catalytic efficiency of MOFs still needs to be improved and the reaction mechanism is unclear. Generally, OH is generated from electron transfer between the complex of H2O2 and iron sites. Thus, the formation of the H2O2–iron complex on the surfaces of
126
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
catalysts is the key step in the generation of OH. As a Lewis base, H2O2 tends to adsorb onto the Lewis acid sites. Therefore, it is deduced that the iron, coordinatively unsaturated or occupied by some replaceable ligands such as water, is the active site on the surfaces of heterogeneous Fenton-like catalysts. Recently, open iron sites, which are accessible and coordinatively unsaturated, were found in MIL-88B-Fe [13]. MIL-88B-Fe, a 3D porous MOF, is constructed from 1,4-benzenedicarboxylic acid (BDC) and trimeric Fe octahedral (Fe3-l3-oxo) clusters. Within the trimeric (Fe3-l3-oxo) clusters, Fe atoms exhibit an octahedral environment with open iron sites, which are terminated by some nonbridging ligands such as water or halogen/hydroxide anionic ligands [13]. Previous studies revealed that the nonbridging ligands could be replaced by Lewis bases [14,15]. Thus, H2O2 can be adsorbed onto the iron sites of MIL-88B-Fe by displacing these ligands, which is the essential step in the Fenton-like reaction. The electron-rich organic ligands (BDC) could donate electrons to the center Fe ions [16], which may increase the redox potential of Fe(II)/Fe(III) and thus enhance reduction of Fe(III) to Fe(II) [17]. Moreover, MIL-88B-Fe undergoes a very large swelling (85% increase of its cell volume) upon exposure to polar solvents such as water and can maintain its open-framework topology [18]. This property is favorable for reactant transfer in MIL-88B-Fe. From these advantages, MIL-88B-Fe is expected to be a promising Fenton-like catalyst. In this work, MIL-88B-Fe with chemical composition Fe3O[C6H4(CO2)2]3XnH2O (X = Cl or OH) was employed as a heterogeneous Fenton-like catalyst. The influence of reaction parameters (e.g., pH and catalyst and H2O2 concentration) on the catalytic activity of MIL-88B-Fe and the H2O2 catalytic decomposition mechanism was investigated. Moreover, the catalytic activity of MIL-53-Fe and MIL-101-Fe, which contain the same organic ligand but different structures from MIL-88B-Fe, was tested to explore the relationship between the catalytic activity of iron-based MOFs and their structures.
2. Materials and methods 2.1. Chemicals and materials 1,4-Benzenedicarboxylic acid (BDC), benzoic acid, Fe2O3,
a-FeOOH, and Fe3O4 were obtained from the Aladdin Industrial
Corporation (China). H2O2 (30%), FeCl36H2O, N,N-dimethyl formamide, methanol, isopropyl alcohol, Na2SO4, NaOH, and H2SO4 were purchased from the Damao Chemical Reagent Factory (Tianjin, China) and the Fuyu Chemical Co., Ltd. (Tianjin, China). T-Butyl alcohol (TBA) was bought from the Bodi Chemical Co., Ltd. (Tianjin, China). All chemicals were analytical grade and used without further purification. Ultrapure water (18.2 MX) was used throughout the experiments.
2.2. Synthesis of catalyst MIL-88B-Fe was prepared by a solvothermal method according to the procedure in the literature [19]: briefly, hydrothermal treatment of FeCl36H2O (270 mg) and 1,4-benzenedicarboxylic acid (116 mg) in N,N-dimethyl formamide (5 mL) with NaOH (2 M, 0.4 mL) at 100 °C for 12 h. After solvothermal treatment, the assynthesized MIL-88B-Fe was collected by filtration and washed with DMF, methanol, and water at 50 °C until the supernatant became colorless. Finally, it was activated overnight at 110 °C. The obtained catalysts were stored in a glass desiccator.
2.3. Characterization of catalyst Powder X-ray diffraction (XRD) of catalysts was performed on an EMYPREAN diffractometer with Cu Ka radiation (k = 1.54056 Å) over a 2h range of 5°–40°. Scanning electron microscopy (SEM) images were obtained on an S-4800 type SEM (Hitachi, Japan) and transmission electron microscopy (TEM) images were obtained on a Tecnai G2 F30 S-Twin type TEM (FEI Company, USA). The Brunauer–Emmett–Teller (BET) specific surface area was determined by nitrogen adsorption in Quantachrome SI (USA) at 77 K. The thermal stability was measured on a TG-DTG 6300 thermogravimetric analyzer (EXSTAR, Japan) by heating samples to 600 °C at a rate of 5 °C/min in 60 mL/min air flow. The organic groups on the surfaces of catalysts were investigated by infrared spectroscopy (IR), which was recorded on a Bruker VERTEX 70 FTIR spectrometer in the wavenumber range from 4000 to 500 cm1 with a resolution of 4 cm1 at room temperature. The coordinatively unsaturated irons in catalysts were tested by CO adsorption infrared spectra [20]. Elemental analysis of catalysts and the concentrations of leached irons during Fenton reaction were tested by inductively coupled plasma optical emission spectroscopy (ICP, Perkin Elmer Optima 2000, USA). The elemental valence state of catalysts was investigated by X-ray photoelectron spectroscopy, which was performed on an ESCALAB 250XI (Thermo Fisher Scientific, USA). The C1s peak at 284.6 eV was used to calibrate peak positions. 2.4. Catalytic activity test Phenol was selected as a target pollutant to evaluate the catalytic activity of MIL-88B-Fe, as it is a ubiquitous pollutant and an important intermediate in industrial processes. The degradation experiments were carried out in a batch mode using a threenecked flask (250 mL) at room temperature. The pH was adjusted with NaOH (0.1 M) or H2SO4 (0.1 M). In a typical experiment, MIL-88B-Fe was dispersed in phenol solution (50 mg/L, 150 mL) at a specific pH by sonication for 2 min. Then the degradation reaction was initiated by adding H2O2 under magnetic stirring. At predetermined time intervals, samples were withdrawn and filtered through 0.22 lm membrane filters to remove suspended MOFs. Meanwhile, an aliquot of 1 M isopropyl alcohol was immediately added to quench the reaction, and then the concentration of phenol was analyzed. After reaction, MIL-88B-Fe was separated by filtration, washed with water and methanol, and then reused in a new reaction after being dried at 110 °C. This process was repeated several times to test the reusability of MIL-88B-Fe. The performance of H2O2 oxidation without MIL-88B-Fe and the amount of phenol adsorbed by MIL-88B-Fe were tested as control experiments. Each degradation experiment was run in triplicate. The reported data are arithmetic means of three measured values. 2.5. Analytical methods The concentration of phenol was analyzed on a highperformance liquid chromatograph (Agilent 1200) equipped with a diode array detector (DAD) and a C18 reversed-phase column (5 lm, 4.6 mm 150 mm). The mobile phase was a mixture of methanol and water (55:45, v/v) at a flow rate of 1.0 mL/min with a column temperature of 30 °C, and the analytical wavelength was 270 nm. The total organic carbon (TOC) was measured using a multi N/C 2100S (Analytikjena, Germany). The main reactive oxidative species generated in the systems were tested by electron paramagnetic resonance (EPR) and radical quencher experiments. For the EPR test, 20 lL of 0.1 M DMPO was added rapidly into a 1 mL sample after filtration to form a DMPO–radicals adduct. The EPR spectra were obtained on a Bruker 300E spectrometer
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
(Germany). The generation rates of OH were analyzed by monitoring the oxidation rates of the benzoic acid heterogeneous Fentonlike reaction [14,21]. The information for ATR-FTIR, in situ Raman spectroscopy, and linear sweep voltammetry (LSV) characterizations are shown in the Supporting Information. 3. Results and discussion 3.1. Characterization As shown in Fig. 1A, the XRD pattern of as-synthesized MIL88B-Fe was well matched with the [0 0 2], [1 0 1], [1 0 3], [2 0 2], and [2 1 1] diffraction planes of MIL-88B-Fe [18,22,23]. The assynthesized MIL-88B-Fe was of high purity, as no diffraction peaks of other substances were found. It should be noted that the slight differences of peak positions and their intensity were possible because of the structural flexibility [24]. The SEM and TEM images of the as-synthesized MIL-88B-Fe demonstrated a needlelike morphology with a uniform size of 0.8–1 lm in length and 100 nm in width (Fig. 1B and C). The as-synthesized MIL-88B-Fe gave a BET surface area of 165.4 m2 g1 with a pore volume of 0.2 cm3 g1. As shown in Fig. S1B in the Supporting Information, the IR band at 1656 or 1660 cm1 representing DMF in MIL-88B-Fe was scarcely found, indicating that only water was trapped in MIL-88B-Fe after activation [25,26]. The TG curve revealed that MIL-88B-Fe possessed good thermal stability. The structure collapsed (sharp weight loss) when the temperature was above 350 °C. As shown in Fig. S1C, the weight loss in the range 25–150 °C was 3% (the departure of coordinated water), which was lower than the reported 5–15% loss [26,27]. It was caused by partial water removal during the activation process (110 °C for 12 h after preparation). 3.2. Catalytic activity of MIL-88B-Fe Control experiments were conducted to compare the efficiency of phenol removal by various processes. As can be seen from Fig. S2, in the absence of MIL-88B-Fe, phenol was hardly degraded by the H2O2 oxidation process. With 0.1 g/L MIL-88B-Fe alone, the removal of phenol was fast in the first 5 min and then reached a steady state of 11% removal efficiency after reaction for 30 min, which was attributed to the adsorption of phenol onto the surface of MIL-88B-Fe. With the simultaneous presence of 0.1 g/L MIL-88BFe and 15 mM H2O2, 99% of phenol was removed. It could be concluded that MIL-88B-Fe possessed catalytic activity for Fenton-like reactions.
127
Since the performance of Fenton-like reactions is dependent on the initial solution pH values, the temperature, and the concentrations of catalysts and H2O2, the influence of these parameters on the degradation of phenol in MIL-88B-Fe catalyzed Fenton-like reactions was investigated. The removal of phenol followed a pseudo-first-order reaction in kinetics under all the tested experimental conditions. As shown in Fig. 2A and S3A, the removal of phenol was faster at lower pH values than at higher pH values. The lower removal efficiency of phenol at higher pH values could be ascribed to the auto-decomposition of H2O2 into oxygen and water (2 H2O2 ? 2 H2O + O2) and the lower oxidation potential of OH at higher pH values (pH 0: 2.8 V; pH 7: 1.9 V) [28]. The pseudo-first-order kinetic constant (k) of phenol degradation at pH 4 was 0.15 min1, which had a level of magnitude similar to that of the homogeneous Fenton reaction reported by Esplugas et al. [29]. Moreover, at near-neutral pH, MIL-88B-Fe demonstrated good catalytic activity, while the homogeneous and some previously reported heterogeneous Fenton-like catalysts could hardly work [3,30,31]. The pseudo-first-order kinetic constant of phenol degradation at pH 5 by a MIL-88B-Fe catalyzed Fenton-like reaction was one order of magnitude higher than those for many reported heterogeneous Fenton-like catalysts, as shown in Table S1. The effect of H2O2 concentration on phenol degradation is illustrated in Figs. 2B and S3B. When the H2O2 concentration was increased from 5 to 15 mM, k increased from 0.07 to 0.15 min1 and the removal efficiency of phenol was increased from 90 to 99%. The enhanced removal of phenol might be attributed to the increased H2O2 concentration leading to more OH generation in the Fenton-like reaction. However, further increasing the H2O2 concentration (20 mM) resulted in reduction of k (0.11 min1) and the removal efficiency of phenol (97%). This might be related to the scavenging of OH by excessive H2O2 to form less reactive species such as hydroperoxyl radicals, which could hardly degrade organic pollutants [30]. The effect of catalyst dosage on phenol degradation was also studied. As shown in Figs. 2C and S3C, with increasing dosage of catalyst from 0.05 to 0.15 g/L, both the removal efficiency of phenol and the k value increased (86–99%, 0.06–0.17 min1), which might be ascribed to the fact that the increasing number of active sites accelerated the production of OH. Nevertheless, the phenol degradation was not enhanced but slightly decreased (98% phenol removal with k of 0.14 min1) when the catalyst dosage further increased to 0.2 g/L. This phenomenon was probably attributable to the agglomeration of catalysts, which limited the diffusion of reactants to the surfaces of catalysts (Fig. S4). Moreover, the excess
Fig. 1. XRD pattern (A) and SEM (B) and TEM images (C) of as-synthesized MIL-88B-Fe.
128
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
Fig. 2. Factorial effects of reaction conditions on the catalytic degradation of phenol: initial pH value (A), H2O2 dosage (B), catalyst dosage (C), and temperature (D). Except for the investigated parameters, other parameters were fixed: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20°C.
of catalysts added may scavenge OH, which reduced the amount of OH and hindered the degradation of phenol [32]:
OH + BFe(II) ! OH + BFe(III).
ð1Þ
As shown in Figs. 2D and S3D, a higher reaction rate was achieved at an elevated temperature. k at 20, 30, and 40 °C was 0.15, 0.16, and 0.17 min1, respectively. The catalytic activity of as-synthesized MIL-88B-Fe was compared with that of the other three conventional heterogeneous Fenton-like catalysts (Fe2O3, a-FeOOH, and Fe3O4). As shown in Fig. 3A and Table S2, the pseudo-first-order kinetic rate constant of phenol degradation in the MIL-88B-Fe catalyzed Fenton-like reaction was 1–3 orders of magnitude higher than that with Fe3O4, Fe2O3, or a-FeOOH. The mineralization of phenol was also
measured. As shown in Fig. 3B, 44% of TOC was removed in the MIL-88B-Fe catalyzed Fenton-like reaction after 30 min, which was much higher than those for the conventional catalysts (<5% TOC removal). Considering that the organic ligands in MIL-88B-Fe may influence its redox properties, the LSV measurement was performed to measure the reduction potential of MIL-88B-Fe. As shown in Fig. 3C, compared with the traditional catalysts, higher reduction peaks representing reduction from Fe(III) to Fe(II) were observed on MIL-88B-Fe, indicating that the reduction of Fe(III) was easier in MIL-88B-Fe than in the traditional Fenton catalysts. Moreover, the reduction potential of Fe3+ ions was tested. The reduction potential of Fe3+ ions was 0.14 V, which was much lower than that for MIL-88B-Fe (0.32 V), implying that the reduction of MIL-88B-Fe was easier. These phenomena could be attributed to
Fig. 3. Degradation of phenol (A), TOC removal (B) in Fe2O3, a-FeOOH, and Fe3O4 catalyzed Fenton-like reactions, and (C) LSV collected with MIL-88B-Fe, Fe2O3, a-FeOOH, Fe3O4, and Fe3+ (the concentration of Fe in MIL-88B-Fe) in N2-saturated 0.1 M Na2SO4 electrolyte at room temperature. Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20 °C.
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
the centering of the organic ligands donor electrons to Fe, leading to enhanced reduction of Fe(III) to Fe(II). These results indicated that MIL-88B-Fe was promising as a catalyst for Fenton-like reactions. 3.3. Stability and reusability of MIL-88B-Fe Fig. 4A displays the concentrations of the leached iron ions from MIL-88B-Fe versus reaction time at pH 4–5. The concentration of leached iron increased with reaction time. The highest concentration of leached iron at pH 4 was 0.68 mg/L at 30 min, which was below the environmental standard (2 mg/L) imposed by the European Union. To exclude the possibility that the observed catalytic activity was caused by the leaching iron ions, 0.1 g/L MIL-88B-Fe was added into 50 mg/L phenol solution at pH 4 for 30 min, and then H2O2 was added to the solution after the MIL-88B-Fe was removed by filtration. As shown in Fig. S5, the phenol degradation catalyzed by the leached iron ions was negligible. Thus, the catalytic activity of MIL-88B-Fe was dominated by heterogeneous catalysis rather than homogeneous catalysis resulting from the leached iron. Moreover, the used MIL-88B-Fe after the reaction was characterized by SEM, XRD, and IR (Figs. S6 and S7). The XRD patterns, morphology, and IR spectra of fresh and used catalysts revealed no obvious change. To evaluate the reusability of MIL88B-Fe, four successive cycles of degradation testing were performed. As shown in Fig. 4B, the removal efficiency of phenol remained almost constant in every run, indicating the excellent stability of MIL-88B-Fe. The concentration of leaching iron ions after each cycle was measured. As shown in Fig. S8A, the concentration of leached iron after each run was below the environmental standard (2 mg/L) imposed by the European Union. Moreover, the catalytic performance of the leached iron was investigated (Fig. S8B). It was found that the leached iron exhibited negligible catalytic activity. These results suggested that MIL-88B-Fe was stable enough as a heterogeneous Fenton-like catalyst. 3.4. Reactive oxidative species and mechanism discussion The reactive oxidative species (ROS) generated in MIL-88B-Fe catalyzed Fenton-like reactions were measured by an EPR spintrap technique (with DMPO). The EPR spectrum of the DMPOROS adducts generated during the Fenton-like reaction at 5 min is shown in Fig. 5A. A four-fold characteristic peak with an intensity ratio of 1:2:2:1 that belonged to the DMPO–OH signal was found [15]. However, the DMPO–OH signal was not observed in the presence of 300 mM TBA (OH scavenger). Thus, OH was the main ROS generated in MIL-88B-Fe catalyzed Fenton-like reac-
129
tions. The effect of OH on phenol degradation was measured by determining the influence of TBA as OH scavenger. As shown in Fig. 5B, the removal efficiency of phenol declined from 99 to 83% in the presence of 10 mM TBA at 30 min, and it continuously decreased to 5% when the concentration of TBA was further increased to 300 mM. This result indicated that OH played a key role in phenol degradation in MIL-88B-Fe catalyzed Fenton-like reactions. To investigate the active sites in MIL-88B-Fe, the catalytic activity of MOF-5-Zn (characterization results shown in Figs. S9 and S10) that is constructed from the same organic ligands (BDC) as MIL-88B-Fe but with a different metal (Zn, which is unable to catalyze Fenton-like reactions) in Fenton-like reactions was studied. As can be seen in Fig. 6A, the concentration of phenol was almost unchanged in 30 min, indicating that MOF-5 could not catalyze Fenton-like reactions. The result suggested the iron clusters were the active sites in MIL-88B-Fe. This conclusion was further confirmed by the low efficiency of phenol removal (2%) in the organic ligands (BDC) catalyzed Fenton-like reaction (Fig. S11). Phosphate (a Lewis base, pKa(H2O2) < pKa(PO3 4 )), which has a stronger affinity to the surface metal sites (coordinatively unsaturated metal sites) than H2O2, water, and some other nonbridging anionic ligands (Cl1 or OH1), was added to MIL-88B-Fe catalyzed Fenton reactions [33]. As shown in Fig. 6A, the degradation of phenol was totally inhibited by the addition of phosphate, suggesting that combination of H2O2 with metal sites was probably hindered by phosphate. To confirm this hypothesis, ATR-FTIR spectroscopy was employed to detect the adsorption of H2O2 on MIL-88B-Fe in the presence of phosphate. As shown in Fig. 6B, no significant new peaks that can be ascribed to the m(OAO) wavenumber (1400, 1195, and 1080 cm1) appeared after addition of H2O2 [34,35]. Therefore, it could be concluded that the coordinatively unsaturated iron atoms were the active sites and their direct contact with H2O2 was the essential step in the MIL-88B-Fe catalyzed Fenton-like catalytic reaction. In addition, the interaction of H2O2 and catalysts was investigated by in situ Raman spectroscopy. As can be seen in Fig. 6C, before addition of H2O2, the suspensions of MIL-88B-Fe showed a band at 865 cm1, which was attributed to the out-of-plane deformation modes of the CAH bond [36]. After the addition of H2O2, a new band located at 875 cm1 that could be assigned to the OAO stretching of peroxo complexes appeared, indicating that H2O2 was coordinated to the iron sites of MIL-88B-Fe [37,38]. With increasing reaction time, the above-mentioned band was gradually decreased, suggesting that H2O2 was decomposed on MIL-88B-Fe. XPS characterization was conducted to investigate the atomic composition and chemical state of the iron species on the surface
Fig. 4. Variation of leached iron concentration during reaction at pH = 4–5 (A) and reusability of MIL-88B-Fe in Fenton-like reaction (B). Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20 °C.
130
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
Fig. 5. DMPO-ROS adducts generated from MIL-88B-Fe catalyzed Fenton-like reactions at 5 min (A) and influence of radical scavenger (TBA) concentration on the catalytic degradation of phenol (B). Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, [TBA] = 300 mM, and 20 °C.
Fig. 6. Degradation of phenol in MOF-5 and MIL-88B-Fe catalyzed Fenton-like reactions in the presence of 10mM phosphate (A), ATR-FTIR spectra of MIL-88B-Fe (B), in situ Raman spectra of MIL-88B-Fe with H2O2 (C), and Fe2p XPS survey spectra of fresh and used MIL-88B-Fe (D). Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20 °C.
of MIL-88B-Fe. As shown in Table S3, 1.65% of Cl existed on MIL88-Fe, which was attributed to the nonbridging chlorine anionic ligands on Fe clusters in MIL-88B-Fe. After reaction, the atomic ratio of Cl was decreased to 0.15%. Considering the weak interaction between Cl1 and Fe, it was speculated that Cl1 could be replaced by H2O2 in the reaction [39]. To confirm this hypothesis, XPS spectra of MIL-88B-Fe after addition of H2O2 for 10 min were tested. The ratio of Cl in MIL-88B-Fe after addition of H2O2 was changed to 0.21%, which was similar to the 0.15% of Cl in MIL-88B-Fe after reaction, indicating that the nonbridging halogen
anionic ligand (Cl1) on Fe clusters could be replaced with H2O2 during reaction. As generally recognized, electron exchange between Fe(II)/Fe (III) and H2O2 induces OH generation in homogeneous and heterogeneous Fenton-like reactions [40]. The chemical states of the iron species on MIL-88B-Fe before and after reaction were measured. Fig. 6D illustrates XPS spectra of Fe2p in fresh and used MIL-88BFe. The fresh MIL-88B-Fe displays the spin–orbit with binding energy (BE) of 725.7 eV (Fe2p1/2) and 711.8 eV (Fe2p3/2). The Fe2p3/2 peak could be fitted into three contributions at 711.3,
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
131
3.5. The relationship between the catalytic activity and the structures of iron-based MOFs
Fig. 7. Schematic diagram of the reaction mechanism of H2O2 activation by MIL88B-Fe.
712.5, and 713.8 eV, respectively, indicating that only Fe(III) existed on fresh MIL-88B-Fe. The weak shake-up satellite peak at about 717.4 eV, which is observed in the Fe2p spectrum, further confirmed that iron was predominantly in the Fe(III) state. After the Fenton-like reaction, the BE of Fe2p was moved slightly to lower BE values (724.0 and 710.9 eV). Deconvolution of the Fe2p3/2 peak of used MIL-88B-Fe after 10 min showed that Fe(II) appeared on MIL-88B-Fe after reaction, indicating that some Fe (III) was reduced to Fe(II) during the Fenton-like reaction. According to the above discussion, the reaction mechanism was shown in Fig. 7: first, H2O2 was attached to the iron sites, as H2O2 is a strong Lewis base that has stronger affinity to iron sites than water and nonbridging anionic ligands,
BFe(III)— L + H2 O2 ! BFe(III) (H2 O2 ) + L
ð2Þ
ðL ¼ H2 O or nonbridging anionic ligandsÞ; and then the electron/charge was transferred between H2O2 and MIL-88B-Fe, resulting in the reduction of Fe(III) to Fe(II) on MIL88B-Fe:
BFe(III) (H2 O2 ) ! BFe(II) + O2 H + Hþ .
ð3Þ
Next, the generated surface Fe(II) reacted with H2O2 to form OH:
BFe(II) + H2 O2 ! BFe(III) + OH + OH .
ð4Þ
BFe(III) + O2 H ! BFe(II) + O2 + Hþ .
ð5Þ
Eventually, organic pollutants were oxidized by surface-bonded OH or some diffused OH in the bulk solution:
Phenol +
OH / ! CO2 + H2 O.
ð6Þ
To explore the effect of structure of iron clusters on the Fenton catalytic activity of iron-based MOFs, the Fenton-like catalytic activity of MIL-53-Fe, which consists of the same organic ligand as MIL-88B-Fe but possesses different structures of iron clusters, was tested (detailed characterizations are given in Figs. S12– S15). As shown in Fig. 8A and B, only 11% of phenol was degraded after 30 min and the kinetic constant of benzoic acid degradation was 2.36 104 min1, which were obviously much lower than those for MIL-88B-Fe. These results may be owing to the different structures of iron clusters in MOFs. CO adsorption infrared spectra were applied to determine the amount of coordinatively unsaturated iron sites of the MIL-88B-Fe and MIL-53-Fe [20]. The intensity of the peak around 2172–2179 cm1 (CO adsorption on the coordinatively unsaturated iron sites in MOFs) was proportional to the amount of coordinatively unsaturated iron in catalysts. As shown in Fig. 8C, after CO adsorption for 20 min, the intensity of the signal around 2172–2179 cm1 of MIL-88B-Fe is higher than that of MIL-53-Fe. It could be deduced that the number of coordinatively unsaturated irons in MIL-88B-Fe was much higher than for MIL-53-Fe. Thus, the excellent Fenton catalytic activity of MIL-88B-Fe was related to the large number of coordinatively unsaturated iron sites. The catalytic activity of MIL-101-Fe, which consists of the same building units as MIL-88B-Fe but possesses a rigid topology (does not swell upon exposure to polar solvents), was tested to explore the effect of the flexibility of iron-based MOFs on the Fenton-like catalytic activity. The detailed characterization results for MIL101 are given in Figs. S10–S13. As shown in Fig. 8A and B, 62% of phenol was degraded within 30 min and the benzoic acid degradation kinetic constant was 1.54 103 min1, demonstrating that the catalytic activity of MIL-101-Fe was lower than that of MIL88-Fe. According to the characterizations of BET and CO-FTIR, the surface area, pore size, and number of coordinatively unsaturated iron sites for MIL-101-Fe were similar to those of MIL-88B-Fe, implying that the difference in catalytic activity between MIL-101-Fe and MIL-88B-Fe was not related to these mentioned structural characters, while other factors may affect the catalytic performance of iron-based MOFs (Table S4 and Fig. 8C). In view of the flexible behavior of MIL-88B-Fe, which possesses a very large volume swelling (increasing by 85% of its cell volume from 1135 to 2110 Å3) upon exposure to polar solvents, it is more favorable for reactant transfer in MIL-88B-Fe than in MIL-101-Fe. Thus, the superior activity of MIL-88B-Fe was probably associated with the greater accessibility of the active sites with H2O2, which resulted from its flexible topology. Based on the above results, the catalytic
Fig. 8. Catalytic performance of MIL-53-Fe, MIL-88B-Fe, and MIL-101-Fe: phenol degradation (A), kinetics of benzoic acid catalytic degradation (B), and CO adsorption infrared spectra of MIL-53-Fe, MIL-88B-Fe, and MIL-101-Fe (catalyst 0.1 g, v(CO) 50 mL/min) (C). Reaction conditions: pH = 4, [H2O2] = 15 mM, [catalyst] = 0.1 g/L, [phenol] = 50 mg/L, and 20 °C.
132
C. Gao et al. / Journal of Catalysis 356 (2017) 125–132
activity of iron-based MOFs was strongly dependent on the number and accessibility of the coordinatively unsaturated irons. 4. Conclusions In summary, MIL-88B-Fe exhibited high activity and stability with a wide operating range of pH (46) in Fenton-like reactions. The catalytic activity of MIL-88B-Fe was better than that of three conventional catalysts (Fe2O3, a-FeOOH, Fe3O4) and was even comparable to that of the homogenous Fenton catalyst. The superior activity of MIL-88B-Fe was mainly attributed to the abundance of coordinatively unsaturated irons, the flexible structure, and the enhanced Fe(III)/Fe(II) redox properties. This work could provide new insights into the rational design and development of highly effective Fenton-like catalysts. Acknowledgments This work was supported financially by the National Natural Science Foundation of China (No. 21590813) and the Programme of Introducing Talents of Discipline to Universities (B13012). Conflict of interest The authors declare no competing financial interests. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.jcat.2017.09.015. References [1] A. Dhakshinamoorthy, S. Navalon, M. Alvaro, H. Garcia, Chem. Sust. Chem. 5 (2012) 46–64. [2] N. Panda, H. Sahoo, S. Mohapatra, J. Hazard. Mater. 185 (2011) 359–365. [3] L. Xu, J. Wang, Appl. Catal. B Environ. 123–124 (2012) 117–126. [4] A.M. Mesquita, I.R. Guimarães, G.M.M. de Castro, M.A. Gonçalves, T.C. Ramlho, M.C. Guerreiro, Appl. Catal. B Environ. 192 (2016) 286–295. [5] M.H. Do, N.H. Phan, T.D. Nguyen, T.T. Suong Pham, V.K. Nguyen, T.T. Trang Vu, T.K. Phuong Nguyen, Chemosphere 85 (2011) 1269–1276. [6] H. Xin, A. Koekkoek, Q. Yang, R. van Santen, C. Li, Emie J.M. Hensen, Chem. Commun. 48 (2009) 7590–7592. [7] L. Xiang, S. Royer, H. Zhang, J.-M. Tatibouët, J. Barrault, S. Valange, J. Hazard. Mater. 172 (2009) 1175–1184.
[8] M. Bobu, A. Yediler, I. Siminiceanu, S. Schulte-Hostede, Appl. Catal. B Environ. 83 (2008) 15–23. [9] Y. Baba, T. Yatagai, T. Harada, Y. Kawase, Chem. Eng. J. 277 (2015) 229–241. [10] E. Brillasm, I. Sirés, M.A. Oturan, Chem. Rev. 109 (2009) 6570–6631. [11] L.H. Ai, C.H. Zhang, L.L. Li, J. Jiang Appl. Catal. B Environ. 148–149 (2014) 191– 200. [12] H.L. Lv, H.Y. Zhao, T.C. Cao, L. Qian, Y.B. Wang, G.H. Zhao, J. Mol. Catal. A Chem. 400 (2015) 81–89. [13] M. Ma, H. Noei, B. Mienert, J. Niesel, E. Bill, M. Muhler, R.A. Fischer, Y. Wang, U. Schatzshneider, N. Metzler-Nolte, Chem. Eur. J. 19 (2013) 6785–6790. [14] S. Huo, X. Yan, J. Mater. Chem. 22 (2012) 7449–7455. [15] X. Yang, X. Xu, J. Xu, Y. Han, J. Am. Chem. Soc. 135 (2013) 16058–16061. [16] E. Galardon, S. Ramdeehul, J.M. Brown, A. Cowley, K.K. Hii, A. Jutand, Angew. Chem. Int. Ed. 41 (2002) 1760. [17] J.K. Li, S. Ghoshal, W.T. Liang, M.T. Sougrati, F. Jaouen, B. Halevi, S. Mckinney, G. McCool, C.R. Ma, X.X. Yuan, Z.F. Ma, S. Mukerijee, Q.Y. Jia, Energy Environ. Sci. 9 (2016) 2418–2432. [18] C. Mellot-Draznieks, C. Serre, S. Surblé, N. Audebrand, G. Férey, J. Am. Chem. Soc. 127 (2005) 16273–16278. [19] P. Horcajada, F. Salles, S. Wuttke, T. Devic, D. Heurtaux, G. Maurin, A. Vimont, M. Daturi, O. David, E. Magnier, N. Stock, Y. Filinchuck, D. Popov, C. Riekel, G. Férey, C. Serre, J. Am. Chem. Soc. 133 (2011) 17839–17847. [20] F. Vermoortele, R. Ameloot, L. Alaerts, R. Matthessen, B. Carlier, E.V. Ramos Fernandez, J. Gascon, F. Kapteijn, D.E. De Vos, J. Mater. Chem. 22 (2012) 10313– 10321. [21] W.P. Kwan, B.M. Voelker, Environ. Sci. Technol. 37 (2003) 1150–1158. [22] M. Pham, G. Vuong, A. Vu, T. Do, Langmuir 27 (2011) 15261–15267. [23] K.G.M. Laurier, F. Vermoortele, R. Ameloot, D.E. De Vos, J. Hofkens, M.B.J. Roeffaers, J. Am. Chem. Soc. 135 (2013) 14488–14491. [24] C. Scherb, A. Schödel, T. Bein, Angew. Chem. Int. End. 47 (2008) 5777–5779. [25] G. Vuong, M. Pham, T. Do, Cryst. Eng. Comm. 15 (2013) 9694–9703. [26] L. Paseta, B. Seoane, D. Julve, V. Sebastián, C. Téllez, J. Coronas, A.C.S. Appl, Mater. Interfaces 5 (2013) 9405–9410. [27] N.A. Ramsahye, T. Trung, L. Scott, F. Nouar, T. Devic, P. Horcajada, E. Magnier, O. David, C. Serre, P. Trens, Chem. Mater. 25 (2013) 479–488. [28] N.K. Daud, B.H. Hameed, J. Hazard. Mater. 176 (2010) 938–944. [29] S. Esplugas, J. Giméne, S. Contreras, E. Pascual, M. Rodríguez, Water Res. 36 (2002) 1034–1042. [30] L. Xu, J. Wang, J. Hazard. Mater. 186 (2011) 256–264. [31] J. Wang, Z. Yao, M. Yang, Y. Wang, Q. Xia, Z. Jiang, Environ. Sci. Pollut. Res. 23 (2016) 14927–14936. [32] A. Khataee, P. Gholami, M. Sheydaei, S. Khorram, S.W. Joo, New J. Chem. 40 (2016) 5221–5230. [33] T. Zhang, J. Ma, J. Mol. Catal. A. Chem. 279 (2008) 82–89. [34] Y. Nie, L. Zhang, Y. Li, C. Hu, J. Hazard. Mater. 294 (2015) 195–200. [35] K. Kunimatsu, T. Yoda, D.A. Tryk, H. Uchida, M. Watanabe, Phys. Chem. Chem. Phys. 12 (2010) 621–629. [36] D. Saha, S. Deng, J. Colloid, Interface Sci. 348 (2010) 615–620. [37] L. Lyu, L. Zhang, Q. Wang, Y. Nie, C. Hu, Environ. Sci. Technol. 49 (2015) 8639– 8647. [38] S. Bordiga, A. Damin, F. Bonino, G. Ricchiardi, C. Lamberti, A. Zecchina, Angew. Chem. 41 (2002) 4734–4737. [39] J. Zhang, X.L. Jiao, Y.G. Xia, F.F. Liu, Y.P. Pang, X.F. Zhao, D.R. Chen, Chem. Eur. J. 22 (2016) 9321–9329. [40] X. Hu, B. Liu, Y. Deng, H. Chen, S. Luo, C. Sun, P. Yang, S. Yang, Appl. Catal. B Environ. 107 (2011) 274–283.