Organic–inorganic interpenetrated hybrids based on cationic polymer and hydrous zirconium oxide for arsenate and arsenite removal

Chemical Engineering Journal 287 (2016) 744–754

Contents lists available at ScienceDirect

Chemical Engineering Journal journal homepage: www.elsevier.com/locate/cej

Organic–inorganic interpenetrated hybrids based on cationic polymer and hydrous zirconium oxide for arsenate and arsenite removal Jehú Pérez a, Leandro Toledo a, Cristian H. Campos b, Bernabé L. Rivas a, Jorge Yañez c, Bruno F. Urbano a,⇑ a

Department of Polymer, Faculty of Chemical Science, University of Concepción, Chile Department of Organic Chemistry, Faculty of Chemical Science, University of Concepción, Chile c Department of Analytical and Inorganic Chemistry, Faculty of Chemical Sciences, University of Concepción, Chile b

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


Cationic polymer-hydrous zirconium

oxide hybrids retain arsenite and arsenate.
The higher the hydrous zirconium oxide content the more arsenite sorption.
Hybrid sorbent exhibits selectivity toward arsenite and arsenate sorption.
Inorganic precursor and monomer mole ratio produces changes in the surface composition.

a r t i c l e

i n f o

Article history: Received 19 August 2015 Received in revised form 11 November 2015 Accepted 16 November 2015 Available online 2 December 2015 Keywords: Hybrid Arsenic Sorption Hydrous zirconium oxide Selectivity

a b s t r a c t In this study, the synthesis of interpenetrated hybrid materials capable of removing arsenic from water is reported. These novel hybrid sorbents consisted of a polymeric matrix with quaternary ammonium groups and zirconium oxide as the organic and inorganic phase, respectively. Furthermore, a coupling agent was used to enhance the compatibility between the polymeric matrix and the metal oxide, creating a covalent bond between them. The hybrid sorbents were prepared by varying the mole ratio of the zirconium oxide precursor and the monomer to provide different surface compositions, and their effect on arsenic sorption was investigated. The hybrids were characterized by spectroscopic analysis such as, Fourier transform infrared spectroscopy (FTIR), 13C, and 29Si solid-state nuclear magnetic resonance (NMR), thermal analysis (TGA), transmission and scanning electron microscopy (TEM, SEM–EDS), and FTIR microspectroscopy. Arsenite and arsenate sorption experiments were conducted under different experimental conditions (i.e., initial arsenic concentration, kinetics, and selectivity). The hybrids with a higher hydrous zirconium oxide content exhibited greater sorption of arsenite, and the sorbents with the lowest content exhibited greater sorption of arsenate. Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction Diverse organic and inorganic pollutants are present in water sources for human consumption, and arsenic (organic and inorganic) exhibits one of the highest toxicities. The most popular ⇑ Corresponding author. Tel.: +56 041 2203538. E-mail address: [email protected] (B.F. Urbano). http://dx.doi.org/10.1016/j.cej.2015.11.051 1385-8947/Ó 2015 Elsevier B.V. All rights reserved.

methods on a larger scale involve adsorption and ion exchange (IEX). In general, these approaches are highly efficient, easy to operate and handle, inexpensive, and the process is sludge-free. Despite the advantages of adsorption processes, a high removal efficiency is only achieved when the predominant species in the effluent is pentavalent arsenic (H3AsO4: pKa1 = 2.3; pKa2 = 6.8; pKa3 = 11.6) because at the natural pH of water, the arsenate is in its anionic form, which allows for removal by ion exchange or

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

electrostatic interactions with an adsorbent. However, in natural effluents, arsenic is primarily found in two oxidation states (III and V), and natural pH trivalent arsenic species are non-charged (H3AsO3: pKa1 = 9.2; pKa2 = 12.1; pKa3 = 12.7), which decreases the efficiency of sorption [1]. This situation is predominant when the effluent comes from groundwater. To improve the efficiency, an oxidation step is performed prior to the adsorption processes to increase the concentration of arsenates [2]. Therefore, more variables are introduced into the process (e.g., by-product waste oxidation and additional infrastructure). In addition, an increase in the cost of installation and operation occurs. Although ion exchange (IEX) resins are the primary materials for industrial use, metal oxides, such as zero-valent iron (ZVI) [3], activated alumina (AA) [4], iron oxide [5], titanium oxide [6], cerium oxide [7], have been extensively studied in both academic and industrial fields, and some of these metal oxides possess the ability to remove arsenic species with high selectivity due to the formation of inner or outer sphere complexes [8]. It is important to note that these processes do not use large amounts of additional chemicals, are easy to install [9] and do not produce polluting by-products [10]. These advantageous properties of metal oxides have been combined with cross-linked polymers through the incorporation of metal oxide nanoparticles into IEX resins, resulting in the production of a novel sorbent composite that exhibits the beneficial properties of IEX resins (i.e., hydraulic, porosity, and kinetics) and the selectivity provided by the metal oxide [11–15]. Susuki et al. prepared a porous resin loaded with monoclinic or cubic hydrous zirconium oxide (HZO) by incorporation of ZrOCl28H2O into porous spherical polymer beads followed by hydrolysis and hydrothermal treatment of the zirconium salt. The sorbent composite exhibited strong adsorption for As(V) in a slightly acidic to neutral pH region. In addition, As(III) was favorably adsorbed at a pH of 9–10 [16]. Recently, Pan et al. prepared a sorbent nanocomposite by encapsulating nano-sized HZO particles inside a polystyrene anion exchanger D201. The arsenate sorption reached 88.7 mg g1 using the nanocomposite compared to 70.0 mg g1 using the D201 resin [17]. In a similar way, the D201-HZO nanocomposite was used for fluoride uptake achieving an enhanced performance of the nanocomposite compared with the D201 polystyrene-based resin [18]. Based on the characteristics and properties of adsorbents, the aim of this study was to develop an interpenetrated hybrid network consisting of a polymer phase (organic) and a metal oxide (inorganic), and this material will be applied to the direct and selective removal of arsenic. Interpenetrated network (IPN) hybrids are typically composed of two phases (i.e., organic and inorganic) that are mixed on the molecular level. The synthesis of interpenetrated hybrids occurs either by a sequential two-stage process where a secondary network is formed in the interior of a primary network or by simultaneous formation of both networks. The material obtained is microscopically composed of separated phases but macroscopically homogeneous [19]. These materials can be classified into two categories: (i) Class I corresponds to the hybrids where the phases are interacting by intermolecular forces, such as Van der Waals and hydrogen bonds and (ii) Class II consists of phases that are strongly linked by covalent bonds [20]. In the current present system, the organic phase will consist of a cationic polymer (i.e., poly(4-vinylbenzyl)trimethylammonium chloride), the same function can be found in strong anion exchange resins. For the inorganic phase, hydrous zirconium oxide will be employed because zirconium oxide has a high capacity for removal of arsenate [21–24], and few studies have been focused on the removal of arsenite [22,25,26] indicating that this compound has a high affinity for forming an inner sphere complex with the hydroxyl groups on the surface [8,27]. Unlike, those composite materials were

745

nanoparticles are embedded into a polystyrene matrix, these hybrids both inorganic and organic phases are intimately combined. We expect that the resulting material possesses a combination of properties that are specific to each independent phase: the polymer phase retains arsenate ions mainly through ion exchange process, while the inorganic phase through complexation process. 2. Materials and methods 2.1. Materials The interpenetrated hybrids were obtained using zirconium tetrabutoxide (TBZr, 80%, Aldrich) as the starting material, the [3-(methacryloyloxy)propyl]trimethoxysilane (MPS, 95%, Aldrich) coupling reagent, and the 4-(vinylbenzyl)trimethylammonium chloride (ClVBTA, 99%, Aldrich) monomer. Acetylacetone (acac, 99%, Merck) and benzoyl peroxide (BPO, 99%, Aldrich) were used as a retardant compound and initiator, respectively. The other reagent used in the synthesis and sorption experiments included 2-butanol (anhydrous, >99.5%, Merck) as a synthesis solvent, NaAsO2 (0.05 mol L1, Merck), Na3AsO4 (1000 mg L1, Merck), potassium chloride (Merck, 99%), hydrochloric acid (Merck, 37%) and sodium hydroxide (Merck, 99%). 2.2. Synthesis of hybrid materials The syntheses were according to a two-stage procedure involving radical polymerization and a sol–gel process [28,29]. TBZr was dissolved in 2-butanol (10 mL) and heated at 80 °C. Then, acetyl acetone, ClVBTA, MPS and BPO (2 mol-% based to ClVBTA) were added to the reaction mixture. TBZr/ClVBTA + TBZr mole ratios (/) of 0.8, 0.6, 0.4, and 0.2 were used, and 0.01 mol of ClVBTA was used. The MPS was fixed to 1:1 respect to TBZr and acac. Once all of the reagents were dissolved, the mixture was degassed with N2 and polymerized at 80 °C for 24 h. Next, the solution was transferred to a Teflon beaker and cooled to room temperature, and deionized water was added over the gel followed by mechanical mixing. The mixture was maintained at room temperature for 8 h followed by drying at 80 °C in an oven until a fine white powder was obtained. All of the composites were washed with deionized water under magnetic stirring for 12 h and dried at 70 °C to remove the unreacted reagents. It is important to note that control samples were prepared for comparison. In particular, HZO were obtained using the same procedure but without the addition of monomers. The (P[ClVBTA-co-MPS]) copolymer was obtained by radical polymerization of ClVBTA and MPS in the presence of a protic solvent, which favored hydrolysis and condensation of the methoxysilane groups, leading to the formation of a cross-linked polymer. The sorbents were ground and sieved to obtain a particle size of 100–180 lm for the sorption experiments. The hybrids are referred to as HC(/), where / represents the (TBZr/ClVBTA + TBZr) mole ratio. 2.3. Physicochemical characterization The morphology and structure of the hybrids were elucidated by infrared spectroscopy (FTIR, Perkin Elmer 1760-X spectrometer using a range of 4000–400 cm1 and KBr pellets) and scanning electron microscopy, (SEM, SEM-PROBE CAMECA model SU-30 equipped with an energy dispersive X-ray device, EDX). Nuclear magnetic resonance (Bruker AscendTM 400 MHz CP-MAS solid state NMR), FTIR microspectroscopy (Perkin Elmer, Spotlight 400 FTIR Imaging System model Spectrum Frontier mid-IR spectral range, reflectance mode) and the obtained images were analyzed using the Spectrum (version 10.03.06.0100) and SpectrumIMAGE

746

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

(version R1.7.1.0401) software packages. The thermal properties of the samples were investigated by thermogravimetric analysis (TGA, Netzsch 209 FI model IRIS heating rate 10 °C min1 from 30 °C to 550 °C, N2 atmosphere, with aluminum pans). The concentrations of the arsenic solution were determined using atomic absorption spectrometry (AAS Perkin Elmer, Pinaacle 900F). 2.4. Sorption experiments Arsenic sorption studies were performed under different experimental conditions. The effects of pH, isotherms, sorption kinetics, and interfering ions were studied as a function of the mole ratio. The sorption experiments were carried out according to the following general unless stated otherwise. 30 mg of sorbent were added to 5 mL of an arsenate solution in a thermoregulated water bath at 140 rpm for 24 h. Then, the samples were filtered, and the arsenic concentrations were determined. The sorption capacities were calculated by the difference between the initial and final arsenic concentrations. To study the pH effect, 50 mg L1 arsenic solutions were adjusted to a pH of 2, 4, 6, 8, 10 and 12 and added to the hybrid. The isotherms were determined at 25 °C using a fixed amount of resin and varying the concentration of arsenic in the 10–1000 mg L1 range. The effect of time on trivalent and pentavalent arsenic retention was also studied. 2.0 g of hybrid sorbent was added to 500 mL of 100 mg L1 arsenic on a plate heater at 25 °C with constant agitation at 140 rpm, and then, the 5.0 mL samples were withdrawn at different time intervals (1–240 min). These samples were filtered using a Millipore filter with a pore size of 0.2 lm. To evaluate the selectivity of the hybrids, a set of arsenate solutions containing sulfate anions was prepared. The arsenic concentration was maintained at 100 mg L1, and the SO2 concentration was varied 4 between 50 and 500 mgL1. Due to the hybrid containing a polymer with hydrophilic characteristics, the water absorption capacities (WAC) of the hybrid were determined using the tea-bag method. For this purpose, 30 mg of the hybrid were placed in a filter bag with a mesh size of 80 lm and submerged in water at a pH of 6.0 for 48 h. Finally, the swollen hybrid was weighed to determine the amount of water uptake. Additionally, samples of swollen hybrids were lyophilized to obtain SEM images after water absorption and evaluate the surface changes. With the aim to evaluate the chemical stability of the sorbents and forecast their long-term use, selected hybrids were contacted

with oxalic acid (OA) solution 3 mM at pH 3.0, 6.0 and 9.0 for 24 h. Subsequently, the treated sorbents were contacted with a 100 mg L1 of arsenite solution to evaluate the sorption performance of the hybrids and determine if the OA treatment affects the sorption. Elution studies were conducted by contacting the sorbent with a 500 mg L1 of arsenic solution at pH 6.0 for 24 h, later, the exhausted sorbents were dispersed in HNO3 1 M and NaOH 1 M solution to elute the arsenic oxyanions. 3. Results and discussion 3.1. Hybrid sorbent characterization The synthesis of the hybrid consisted of two stages. The first stage consisted of radical copolymerization between the monomers (i.e., ClVBTA and MPS). The presence of an anhydrous solvent prevents the condensation of methoxysilane groups and the oxide network, and the addition of acetylacetone hinders hydrolysis and condensation of zirconium alkoxides [30]. The second stage consisted of a sol–gel reaction that involves the hydrolysis of zirconium alkoxide to form hydroxyl groups followed by autocondensation to polymerize the oxide network. The addition of deionized water after polymerization drives the formation of the inorganic network and allows for the formation of SiAOAZr bonds via the hydrolysis reaction with the inorganic precursor to create the covalent bond between the organic and inorganic phases (see Fig. 1). After the hybrids were obtained, the materials were washed with distilled water to remove any unreacted reagents. In addition, this procedure helps to remove the polymer chains that were not bound to the inorganic matrix, and the loss of mass indicates that the polymer chains were truly bonded to the matrix. The percentages of mass loss after washing were 29.3, 31.7, 34.5, and 38.8 for the 0.8, 0.6, 0.4, and 0.2 mol ratios, respectively. These differences are due to the MPS comonomer providing covalent bonds between the inorganic and organic component. Because an equimolar TBZr: MPS ratio was used, the hybrids with a higher inorganic component will possess numerous bonds between the polymer and the oxide networks and therefore exhibit a smaller mass loss. The hydrophilicity provided by the quaternary amine allows for the uptake of water molecules. The water absorption capacity (WAC) of the hybrid composites was investigated using the tea-bag method, and the WAC increased as the inorganic content decreased (i.e., 1.6, 7.9, 37.8, and 29.3 g g1 for 0.8, 0.6, 0.4, and 0.2 mol

Fig. 1. Scheme of the hybrid composite synthesis: (1) radical copolymerization and (2) sol–gel process.

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

fractions, respectively). This trend may be due to the polymer favoring water uptake. However, it is important to take into account the inverse relationship between the cross-linking degree and the water uptake. A higher cross-linking degree implies lower water absorption. The large amount of MPS used to prepare the hybrids with a higher inorganic content also prevented water absorption. 3.1.1. Spectroscopic characterization Fig. 2 shows the FTIR spectra of the synthesized hybrids. The spectrum a) corresponds to the copolymer (P[ClVBTA-co-MPS]) used as a control, and the expected vibrational bands due to the MPS and ClVBTA were observed (i.e., 1710 cm1 (C@O, st), 1656 cm1 (C@C, aromatic, st), 1473 cm1 (CAN+, st)), and the signal at 1113 cm1 corresponded to the CASiAO vibration mode. The ester band was observed in all of the spectra, and intensity of this band increased as / increased, which is due to the large MPS content used in the feed. In contrast, the band corresponding to the aromatic moieties and the signal due to quaternary ammonium group decreased in intensity until they disappeared (overlapped) in HC(0.8). For hybrids with a major zirconium oxide content in the feed, new bands appeared and other bands increased in intensity. For example, the bands located at 1600 cm1 and 1524 cm1 were due to the OAH vibration of water molecules physically adsorbed on the oxide, and the bands located at 1371, 1035 and 659 cm1 corresponded to the bending vibration of hydroxyl groups on the metal oxide (ZrAOH) [31,32]. The observed vibrational bands confirm the existence of the functional groups in the hybrid material, and the change in the intensities is in agreement with the different mole ratios used in the feed. The 29Si-NMR and 13C-NMR analyses were performed to confirm the chemical structure/linkage on the IPNs. Fig. 3 shows the 13 C-NMR spectra of the IPNs and control polymer. The spectrum of the copolymer exhibits signals that correspond to both of the monomers used in the synthesis. The signals in the d = 10.4–22.7 ppm region are due to alkane carbons on the main polymer chain and the propyl chain of MPS. However, the signal at 68.7 ppm corresponded to the carbon of the methoxy groups bound to silicon due to the electronic effects of oxygen atoms. The alkene carbon of the ester group appears at 178.1 ppm. For the ClVBTA pendant group, the signals located at d = 52.9, 127, and 133.1 ppm are due to methyl groups of the ammonium and aromatic ring in ClVBTA. The spectra of the IPNs exhibit signal intensities that vary as the mole fraction changes. Interestingly, the signal that was initially located at 68.7 ppm shifted to 65.3 (associated with the methoxy groups). This shift can be explained

Fig. 2. Infrared spectra of hybrid materials: (a) P(ClVBTA-co-MPS), (b) HC(0.2), (c) HC(0.4), (d) HC(0.6), and (e) HC(0.8).

747

by the change in the chemical environment of the methoxy groups after the (RO)3x–(SiAOAZr)x covalent bond formation with the oxide network compared to the control copolymer (Fig. 3a), which was observed in the 29Si-NMR (vide infra). The changes observed in the 13C-NMR are consistent with the molar composition used in the feed and the structure of the copolymer obtained in the first synthesis step. The 29Si-NMR analysis provides insight into how the methoxysilane groups react with the inorganic network and allows for confirmation of the covalent bonding between the organic and inorganic networks. The alkoxysilane coupling agents that are anchored to the HZO surface exhibit signals called T1, T2 and T3 with are associated with the 1, 2, and 3 alkoxy chains, resulting in a 29Si-NMR spectrum with signals in the 45 to 65 ppm range [33]. Fig. 3 also shows the 29Si-NMR spectra for two selected IPNs. The spectrum contains two signals (slightly overlapped) located at 56 ppm and 64 ppm. The signals can be assigned to T2 and T3 anchoring, and the slight downfield shift in the signals may be due to the electronic effects of zirconium. These results confirm that both the inorganic components (HZO) and the organic polymer components are covalently bound by the organosilane reagent, demonstrating that these materials are Class II hybrids [34]. 3.1.2. Thermal studies The thermal performance of the IPN hybrids was studied by thermogravimetric analysis. Fig. S2 shows the TGA curve, and Table S1 lists the main parameters for the thermal properties. The HZO sample exhibited a small mass decomposition with a constant weight loss over the entire temperature interval that was due to dehydration. In contrast, the control copolymer exhibited its first weight loss up to 150 °C due to water desorption followed by a steep decomposition step that begins at 230 °C, which was due to degradation of the pendant groups on polymer chain. Then, the onset of a second decomposition step begins at 356 °C, which corresponds to thermolysis of the main polymer chain. The IPNs exhibited an intermediate behavior between the copolymer and zirconium oxide. When a high oxide precursor content was used in the synthesis, the thermal behavior of HC was more similar to that of the HZO control. Table S1 display that the first degradation temperature increased as the polymer content increased, and similar changes were observed for the second decomposition temperature. These results indicate that HZO affects the stability of the hybrid, which favors decomposition. Similar results have been observed for polyhedral oligomer silsesquioxane-based composites, where an increase in the polymer content decreased the thermal stability [35,36], and high internal phase emulsion polymer hybrids [37], which exhibits a behavior that is in contrast to the expected behavior for most composite materials [38,39]. To elucidate this effect, additional studies that are beyond the scope of this manuscript are required. 3.1.3. Surface characteristics The adsorption process is a surface phenomenon, and a detailed study of the surface characteristics of the adsorbent can provide important information that allows us to understand the sorption results and the mechanism involved. Most of the arsenic sorption studies using hydrophilic sorbents have revealed that the process primarily occurs on the surface of the adsorbent with significant mass transfer at the beginning of the experiment followed by a smaller sorption contribution due to the diffusion of the analyte through the pores. As previously mentioned, ZrO2 can interact with arsenite and arsenate species, and the polymer phase (at natural pH) can only remove arsenate ions. Therefore, their presence at the surface of hybrid particles should be evaluated. FTIR microspectroscopy is a powerful technique for determining the infrared spectrum of the surface, and then, by plotting

748

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

Fig. 3. NMR solid-state analyses of the IPNs hybrids. (1) 13C-NMR (a) HC(0.8), (b) HC(0.6), (c) HC(0.4), (d) HC(0.2), and (e) P(ClVBTA-co-MPS). (2) 29Si-NMR solid-state analyses of selected IPNs hybrids.

the absorbance intensity at a selected wavenumber as a function of position, distribution maps can be created for the selected frequency to gain insight into the distribution of a particular functional group on the surface of an adsorbent [40]. Fig. 4 shows the infrared microspectroscopy results for the hybrid samples, and the absorbance of the absorption band at 1371 cm1, which is due to the deformation vibration of ZrAOH, was evaluated [31,32]. White represents a high absorbance, and dark blue represents the lowest absorption, and white and dark blue correspond to a higher and lower concentration of zirconium oxide, respectively. The images for HC(0.8) exhibit larger areas with a higher HZO concentration due to the major proportion of TBZr in the feed, which is in agreement with the TGA and solidstate NMR results. Because / decreases, the absorbance becomes less intense. The presence of an oxide on the surface is due to the procedure used for preparing the hybrid. During the second step (sol–gel), the product formed in the first step is swollen with water, leading to hydrolysis and condensation reactions of the inorganic precursor. Therefore, the oxide network growth reaches the surface of the product, increasing it concentration. Then, the

subsequent milling and sieving processes expose the oxide phase at the surface. To confirm and quantify the change in the surface, the elemental composition was determined using a SEM–EDS technique. For this purpose, three scanning electron images were analyzed, and the average of each element was determined (see Table 1 and Fig. S3). The values clearly indicate a change in the surface composition of the final product that is consistent with the content used in the feed. Specifically, Zr and Si exhibit the same molar ratio changes due to an equimolar ratio of inorganic precursor and coupling agent being used. However, the amount of C and Cl increased as / decreased due to the increase in the polymer content. These results confirm those observed in the images of Fig. 4 and demonstrate that changing the mole ratio in the feed can modulate the surface composition of the hybrids. Scanning electron microscopy was also used to evaluate the changes in the surface characteristics after water absorption (see Fig. 5). The images show the appearance of pores due to water uptake. In general, as / decreases, water absorption increases, which is due to the higher polymer content providing

749

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

Fig. 4. Surface image of HC(/) sorbents obtained from FTIR microspectroscopy: (a) 0.8, (b) 0.6, (c) 0.4, and (d) 0.2. [sample area = 50  50 lm, reflectance mode, wavenumber = 1371 cm1].

Table 1 Surface elemental analysis by energy dispersive X-ray (EDX) (wt.%). Hybrid/Element

C

O

Cl

Si

Zr

HC(0.8) HC(0.6) HC(0.4) HC(0.2)

32.57 ± 0.07 35.72 ± 4.38 58.79 ± 0.20 71.31 ± 2.14

32.35 ± 1.05 35.78 ± 1.13 16.48 ± 1.10 8.23 ± 3.09

1.21 ± 0.09 1.46 ± 0.25 9.20 ± 0.75 15.15 ± 4.41

7.63 ± 0.11 6.18 ± 1.36 3.97 ± 0.04 1.94 ± 0.43

26.25 ± 0.96 20.93 ± 4.93 11.45 ± 0.15 3.40 ± 0.40

Fig. 5. Electron micrographs of the HC(/) sorbents. (1) Scanning electron micrographs of hybrid after water absorption (a) 0.8, (b) 0.6, (c) 0.4, and (d) 0.2. (2) Transmission electron microscopy images of (e) 0.8, (f) 0.6, (g) 0.4, and (h) 0.2.

750

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

hydrophilicity and flexibility to the material that favors the adsorption of the water molecules into the hybrid. In contrast, the hybrids in the xerogel state exhibit a rough surface with no significant or clear trend as a function of / (see Fig. S4). Fig. 5 also shows the transmission electron microscopy images of the hybrids, and the results indicate dispersion of zirconium oxide (dark areas). The HC(0.6) and HC(0.2) hybrids exhibit good dispersion of the inorganic phase, indicating the formation of a better-defined interpenetrated network. In addition, the HC(0.8) and HC(0.4) hybrids indicate the formation of inorganic zirconium domains. 3.2. Sorption studies 3.2.1. Equilibrium sorption experiments pH is an important factor in the sorption processes of metal ions. The hydronium concentration can change the properties of the adsorbent and speciation of the metal ion in solution, affecting the performance of the adsorbent. Inorganic oxides are capable of varying the pH in solutions, and the pH changes the speciation of arsenic anions. The change in pH was evaluated for hybrids with different inorganic contents. Fig. S5 shows the change in the pH at equilibrium, and the results indicate that all of the sorbents exhibit the same performance. Between an initial pH of 4 and 10, the sorbent changes and maintains a pH at values of approximately 4–6. This behavior may be due to the point of zero charge (PZC), which should be located at pH 4–6 for the current system (reported PZC for zirconium oxide is between 4.0 and 8.6 [41,42]). Therefore, when pH > pHpzc, the terminating hydroxyl groups at the surface deprotonate and reduce the solution pH due to a decrease in the OH concentration. Based on these results, the following sorption experiments were performed at a pH of 6.0. Equilibrium sorption isotherms are important for designing adsorption processes. Moreover, isotherms provide useful information regarding the interaction between the adsorbate and the adsorbent. Fig. 6 shows the arsenic sorption isotherms obtained at 25 °C. The experiments were carried out for all of the synthesized hybrids and the control sorbents (copolymer and HZO). Arsenite sorption reveals that the hybrids with a major inorganic content exhibit higher sorption. In fact, the control sorbents (i.e., P(ClVBTA-co-MPS)) containing only ammonium groups and HZO

exhibit minimum and maximum sorption of arsenite, respectively. These results indicate the importance of the presence of the inorganic component. Ammonium-based sorbents do not retain arsenite due to the absence of charge at the pH of natural effluents, and only IEX can retain arsenate ions, confirming that arsenite sorption is due of the inorganic phase. Arsenite can form inner and outer sphere complexes on the surface of metal oxides, such as titanium, aluminum, and zirconium oxide [8,43]. In particular, for zirconium, three types of hydroxyl groups can be present on its surface [24]. Mono-bridged groups are considered to be the most extended at the surface and least sterically hindered compared to the di- and tri-bridged groups. Eq. (1) shows the inner sphere complex for a mono-bridged surface hydroxyl with arsenite.

2  Zr  OH þ AsðOHÞ3 ! ð Zr  OÞ2 AsðOHÞ þ 2H2 O

ð1Þ

Arsenate removal exhibited the opposite trend, and the sorbent with a higher polymer content exhibited the highest sorption. In this case, the sorbent with a lower mole fraction tended to be more similar to the ammonium-based sorbent. The arsenate sorption is explained by an ion exchange process between the negatively charged arsenic species and the positively charged quaternary ammonium. At a higher ammonium content (polymer), the sorption is greater. It is important to note that zirconium oxide can also retain arsenate. Cui et al. studied the sorption of arsenite and arsenate onto amorphous ZrO2, and this sorbent with sorbed arsenate shifted the isoelectric point from 5.9 to 5.0, which was due to the formation of an inner sphere complex (i.e., arsenate-ZrO2) [23]. For our system, the arsenate sorption was carried out at a pH of 6.0, and the inorganic phase possesses a negative charge and exhibits repulsive forces with the mono-anionic arsenate species, suggesting that sorption primarily occurs on the polymer phase. Interestingly, the sorbents with / = 0.2 and 0.4 retained more arsenate than the polymer control, indicating a synergism between both phases toward arsenate sorption. However, further experiments are required to confirm this result. The experimental data were fitted to the Langmuir and Freundlich isotherm models. The Langmuir isotherm is valid for monolayer sorption onto a surface containing a finite number of identical sites and is commonly applied for homogeneous surfaces where the adsorbed molecules do not interact. The Langmuir isotherm is expressed as:

qe ¼

qm bC e 1 þ bC e

ð2Þ

where Ce is the equilibrium concentration (mg L1), qe is the amount of arsenic removed at equilibrium (mg g1) and qm and b are the Langmuir constants related to the sorption capacity and energy of sorption, respectively. The Freundlich isotherm was derived to model multilayer adsorption and adsorption on heterogeneous surfaces. The form of the Freundlich isotherm is expressed below:

qe ¼ kf ðC e Þ1=n

Fig. 6. Arsenic sorption isotherms using the hybrid sorbent: (a) arsenite and (b) arsenate. (24 h, 140 rpm, pH = 6.0, dosage = 5 mg mL1).

ð3Þ

where qe is the amount of adsorbed analyte per unit weight of the solid phase at the equilibrium concentration, Ce. kf is the Freundlich constant related to the sorption capacity and 1/n is the sorption intensity. If 1/n is equal to 1, the adsorption is linear, and the adsorption sites are homogenous in energy. In addition, no interactions occur between the adsorbed compounds. However, when the 1/n constant is lower, the system is more heterogeneous. The Freundlich constant (n) should have a value lying in the range of 1–10 for adsorption to be favorable. Table 2 lists the isotherm parameters obtained using a nonlinear fit. The isotherm curves for arsenite exhibit acceptable agreement with the Langmuir and Freundlich models based on

751

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754 Table 2 Isotherm parameters obtained from non-linear fitting. x

Langmuir qm (mg g

Freundlich

1

)

1

b (L g

)

r

2

n

k (mg g1)

r2

Arsenite ZrO2 0.8 0.6 0.4 0.2 P(ClVBTA-co-MPS)

130.0 225.1 207.7 127.6 122.1 –

1.054 0.032 0.041 0.029 0.028 –

0.96210 0.97690 0.95520 0.95620 0.95050 –

3.1 1.8 2.1 2.0 2.0 0.7

44.1 15.6 21.4 10.9 10.2 0.01

0.9451 0.9817 0.9852 0.9864 0.9776 0.9098

Arsenate ZrO2 0.8 0.6 0.4 0.2 P(ClVBTA–co-MPS)

46.7 48.7 69.9 95.6 111.8 93.8

0.14 0.63 0.57 0.59 0.59 0.15

0.9479 0.7466 0.9241 0.9218 0.9649 0.9762

2.7 3.6 4.1 4.6 4.0 2.4

9.091 16.73 26.07 41.65 44.03 16.91

0.9200 0.8802 0.8662 0.9588 0.9716 0.9259

the correlation coefficients. The qm values confirm the trend observed in the isotherm curves. As the x values decreased, the sorption capacity also decreased. The Langmuir constants (b) reveal a decrease in the affinity of arsenite toward the sorbent due to the polymer phase, which does not favorably interact with arsenite. Despite ZrO2 possessing the highest affinity, the hybrids with / = 0.8 and 0.6 exhibit a sorption capacity that is higher than that of the oxide due to the presence of the polymer, which provides flexibility to the sorbent and allows the arsenite to access the sorbent. For arsenate sorption, the experimental data are better described by the Langmuir isotherm. The values of qm indicate that the polymer is the main active phase for arsenate sorption. However, uptake on the oxide phase cannot be discarded. Arsenic sorption isotherms at 25, 35, and 45 °C were determined for the sorbents that exhibit the best results (i.e., HC(0.8) and HC(0.2)) for arsenite and arsenate, respectively (see Fig. 7). Arsenite sorption is affected by the temperature. As the temperature increased, the affinity of the arsenite species to the sorbent decreased, which was demonstrated by the values of the Langmuir constant (b) being 0.031, 0.0086, and 0.0047 L g1 at 25, 35, and 45 °C, respectively. In the same way, the sorption of arsenate anions also resulted in a decrease in the Langmuir constant (b values of 0.59, 0.21, and 0.17 L g1 at 25, 35, and 45 °C, respectively). These results suggest that arsenic sorption is exothermic, where the total energy absorbed in bond breaking is less than the total energy released by bond making between the adsorbate and the adsorbent, Therefore, the extra energy is released in the form of heat, and an increase in temperature does not favor this process. 3.2.2. Kinetic experiments Kinetics studies were conducted to evaluate the effect of time on arsenic sorption (see Fig. 8a and b). The studies were carried out for hybrids that exhibited the highest sorption arsenite/arsenate capacities for comparison to the control sorbents. Arsenite uptake reaches a maximum sorption after 25 min of contact with all of the studied sorbents, and as expected, the sorption capacity of the hybrid sorbent is between those of the control sorbents. In addition, arsenate exhibited nearly instantaneous sorption for HC (0.2) (similar to that exhibited by the copolymer), and the ZrO2 sorbent exhibited the slowest sorption where the maximum was achieved after 100 min of contact. To extract more information from these results, the experimental data were fitted to different kinetic models. The experimental data were studied based on non-linear fitting to pseudo-first and pseudo-second order models (PFO and PSO), which are expressed by the following equations:

Fig. 7. Arsenic sorption isotherms at different temperatures. (a) Arsenite-HC(0.8) and (b) arsenate-HC(0.2).

Pseudo-first order model

qt ¼ qe ð1  ek1 t Þ

ð4Þ

Pseudo-second order model

qt ¼

k2 q2e t 1 þ k2 qe t

ð5Þ

where qe and q are the amount of arsenic adsorbed (mg g1) at equilibrium and at time t, respectively, and k1 (min1) and k2 (g mg1 min1) are the rate constants for sorption.

752

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

Lagergren pseudo first-order kinetics typically do not represent experimental kinetic data for the entire sorption period and tends to exhibit good agreement only in the early stages of sorption but fails to predict equilibrium sorption. However, the pseudo-second order equation has been widely used due to the excellent fit of the experimental data for the entire sorption period in many systems. Table S2 compares the kinetic parameters obtained for both kinetic models. Both arsenic species exhibit a better fit with the pseudosecond order model, and the rate constant is higher for the hybrid system than the control sorbents. The HC(0.8) hybrid sorbent exhibits a rate constant for arsenite sorption that is similar to that of the control polymer and higher than that of the ZrO2 control, indicating a faster sorption. The highest rate constant corresponded to the sorption of arsenate onto the HC(0.2) sorbent, and this rate constant was 10- and 24-times higher than those of the polymer and ZrO2, respectively. Although these models provide useful insight into the sorption process, they cannot predict the rate-limiting step of the arsenic sorption or provide mechanistic information. In a heterogeneous process, such as the sorption of arsenic by hybrid materials, three consecutive steps can be used to represent solute transfer as follows: (1) transport of the arsenate anions from the bulk solution through the liquid film surrounding the external surface of the hybrid particle (film diffusion), (2) arsenic diffusion into the pores of the hybrids (intraparticle diffusion), and (3) reaction of arsenic (sorption) on the interior surface pores and capillary areas of the hybrid particles (chemical reaction). The rate of sorption is controlled by the slowest step, which typically corresponds to film diffusion or intraparticle diffusion. To elucidate the mechanism of arsenate sorption, the intraparticle diffusion model was fitted to the experimental data. The model can be described by the following equation:

qt ¼ kip t 0:5 þ C

ð6Þ

0.5 where kip (mg g1 ) is the rate constant for intraparticle difHC min 1 fusion and C (mg gHC) is related to the extent of the boundary layer thickness. In an ideal intraparticle diffusion process, a plot of qt as a function of t0.5 should result in a straight line passing through the

origin. When the intercept is not zero, some degree of external film mass transfer or boundary layer control exists. When the intercept is larger, the layer effect is greater [44]. Fig. 8c and d show the intraparticle diffusion curves that were obtained from experimental data for the sorption of arsenic at a pH of 6. The curves are not straight lines, and therefore, the process is not controlled by intraparticle diffusion. The first segment of the curve is due a high sorption in a short time period, which may be due to a significant mass transfer from the solution to the surface of the hybrid and surface pores (nearly total capacity). However, the second segment, which has a very low contribution to sorption, can be associated with the diffusion of arsenic species to the inner pores of the hybrid. In summary, the sorption process is controlled by film diffusion. 3.2.3. Arsenic sorption mechanism According to the surface complex model theory, metal hydroxyl groups on the surface of many metal oxides are the most abundant and provide reactive adsorption sites for anions. These vibrational bands can be detected by FTIR spectroscopy, and their change after sorption can be determined. The arsenic adsorption mechanism on a hybrid sorbent was further investigated by FTIR spectroscopy. Fig. 9 compares the FTIR spectrum of loaded and unloaded sorbent. The spectra of the hybrid with As(III) and As(V) indicate that the vibration bands due to surface hydroxyl groups (1371, 1035 cm1) decreased in intensity. In addition, the signal observed at 1524 cm1, which corresponds to physically adsorbed water, disappears, which may be due to the displacement of water molecules by arsenic anions. These results may be due to the formation of inner surface complexes of arsenic (III and V) with the metal oxide [8,45]. 3.2.4. Selectivity The sulfate anion interferes with arsenic sorption, especially in ion exchange processes, and the concentration of SO2 should be 4 <20 mg L1 to achieve an efficient process [46]. The affinity sequence 2 for strong-base anion exchange resins is SO2 4 > HAsO4 > 2 CO2 3  Cl > H2AsO4 with sulfate causing the greatest interference. The interference is due to both structural and charge similarities

Fig. 8. Sorption kinetic experiments (a) arsenite, (b) arsenate, (c) arsenite intraparticle diffusion model, and (d) arsenate intraparticle diffusion model. ([As] = 100 mg L1, 140 rpm, pH = 6.0, dosage = 5 mg mL1).

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

753

Fig. 9. FTIR of hybrid sorbent before and after arsenic sorption. (a) HC(0.4) + As(III), (b) HC(0.4) + As(V), and (c) HC(0.4).

with the arsenate anion [47]. To evaluate the selectivity of the sorbent, experiments were conducted in the presence of sulfate ions with concentrations ranging from 50 to 500 mg L1. Fig. 10 compares the sorption performance for arsenite and arsenate using the sorbent that exhibited the best results. The results of the arsenite experiments indicate that the sulfate decreased the sorption capacity of the hybrids (from 0 to 50 mg L1). Then, as the sulfate concentration increased, the sorption capacity remains constant. For arsenate sorption, the sulfate anions strongly affect the sorption of the copolymer sorbent (ion exchange mechanism), and a constant decrease in the sorption was observed as the sulfate concentration increased. However, the HC(0.2) hybrid exhibits a similar behavior for arsenite sorption. These results demonstrate the selectivity properties are provided by the metal oxide phase, especially for arsenate sorption, due to the different mechanisms involved in the process. Based on these results, a complex between the arsenate ion and zirconium oxide was formed. 3.2.5. Chemical stability and elution studies The stability of the sorbent was studied using selected hybrid sorbents following the procedure described previously by Pan et al. [17] with some modifications. The adsorbents / = 0.8 and 0.2 were contacted with a concentration of 3 mM of oxalic acid (OA) for 24 h at pH 3, 6 and 9. Oxalic acid was used due to its presence in natural waters streams and as a model compound due to the stable complexes which can form with metal ions. The pH and the OA species can promote leaching of zirconium, affecting the performance of the adsorbent for longterm use, especially for the arsenite sorption. For this reason, the absorbent after treatment were washed and contacted with a solution of arsenite to evaluate the effect of treatment on the arsenic sorption. The Fig. S6 shows that the sorbent with higher content of HZO not showed a significant decrease of absorption after treatment. A similar behavior was observed for the hybrid HC(0.2), however, the treatment performed at acidic medium (pH 3.0) produces a decrease in the absorption of arsenic. This decrease can be attributed to the loss of HZO adsorption sites due to the leaching of zirconium during treatment. As mentioned before molar ratio used of inorganic precursor and coupling agent (MPS) was 1:1, then the hybrid with fewer HZO (/ = 0.2) has less binding sites with polymer favoring leaching. Arsenic elution and sorbent regeneration is a critical consideration in the analysis of process costs and metal recovery in a concentrated form or further disposal. For effective reuse, a successful desorption process must restore the sorbent such that it exhibits properties that are close to its initial properties. This

Fig. 10. Arsenic sorption in the presence of SO42 anions ([As] = 100 mg L1, 24 h, 140 rpm, pH = 6.0, dosage = 5 mg mL1).

process must be performed when the sorbent is exhausted by the use of a suitable eluent. In the present study the hybrid with the highest and lowest HZO content were studied for the elution of arsenite and arsenate, respectively. Fig. S7 shows that arsenite can be eluted in a approximately 90% with acidic and alkaline media, whereas arsenate can be eluted better with an alkaline solution. A high concentration of OH groups can displace the adsorbed arsenic oxyanions or turn the HZO surface negatively charged increasing the electrostatic repulsion. On the contrary, at acidic medium the arsenic species can turn into uncharged species decreasing the sorption. 4. Conclusions Interpenetrated network hybrids were obtained in a two-step synthesis consisting of (i) radical copolymerization and (ii) a sol– gel process. By changing the mole ratio of the inorganic precursor and comonomer, sorbents with different surface compositions were obtained, which modulated the sorption process. As the hydrous zirconium oxide content increased, the sorption of arsenite also increased. However, at higher mole ratios, the arsenate sorption was greater. The mechanism that is involved in the sorption process consists of the formation of inner and outer sphere complexes for arsenite onto zirconium oxide. However, for arsenate uptake, the mechanism involves an ion exchange process on the polymer phase and formation of a complex on the metal oxide.

754

J. Pérez et al. / Chemical Engineering Journal 287 (2016) 744–754

In addition, this mechanism endows the hybrid sorbent with selectivity. Therefore, these hybrids have the potential for use as sorbents for arsenic (III) and (V) removal. Acknowledgements The authors thank FONDECYT Initiation No 11121291, FONDECYT Postdoc No 3140130, FONDECYT No 1110079, and CIPA-CONICYT Regional R08C1002. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.cej.2015.11.051. References [1] L. Dambies, Existing and prospective sorption technologies for the removal of arsenic in water, Sep. Sci. Technol. 39 (2005) 603–627. [2] E.P. Agency, Arsenic Treatment Technology Evaluation Handbook for Small Systems, EPA U.S., 2003. [3] Y. Mamindy-Pajany, C. Hurel, N. Marmier, M. Romeo, Arsenic (V) adsorption from aqueous solution onto goethite, hematite, magnetite and zero-valent iron: effects of pH, concentration and reversibility, Desalination 281 (2011) 93–99. [4] T.S. Singh, K.K. Pant, Equilibrium, kinetics and thermodynamic studies for adsorption of As(III) on activated alumina, Sep. Purif. Technol. 36 (2004) 139– 147. [5] S. Lunge, S. Singh, A. Sinha, Magnetic iron oxide (Fe3O4) nanoparticles from tea waste for arsenic removal, J. Magn. Magn. Mater. 356 (2014) 21–31. [6] M.E. Pena, G.P. Korfiatis, M. Patel, L. Lippincott, X. Meng, Adsorption of As(V) and As(III) by nanocrystalline titanium dioxide, Water Res. 39 (2005) 2327– 2337. [7] W. Sun, Q. Li, S. Gao, J.K. Shang, Exceptional arsenic adsorption performance of hydrous cerium oxide nanoparticles: Part B. Integration with silica monoliths and dynamic treatment, Chem. Eng. J. 185–186 (2012) 136–143. [8] D.A. Sverjensky, K. Fukushi, A predictive model (ETLM) for As(III) adsorption and surface speciation on oxides consistent with spectroscopic data, Geochim. Cosmochim. Acta 70 (2006) 3778–3802. [9] M. Jang, W. Chen, F.S. Cannon, Preloading hydrous ferric oxide into granular activated carbon for arsenic removal, Environ. Sci. Technol. 42 (2008) 3369– 3374. [10] Y. Zhang, M. Yang, X.-M. Dou, H. He, D.-S. Wang, Arsenate adsorption on an FeCe bimetal oxide adsorbent: role of surface properties, Environ. Sci. Technol. 39 (2005) 7246–7253. [11] I.N. Savina, C.J. English, R.L.D. Whitby, Y.S. Zheng, A. Leistner, S.V. Mikhalovsky, A.B. Cundy, High efficiency removal of dissolved As(III) using iron nanoparticle-embedded macroporous polymer composites, J. Hazard. Mater. 192 (2011) 1002–1008. [12] S. Sarkar, P.K. Chatterjee, L.H. Cumbal, A.K. SenGupta, Hybrid ion exchanger supported nanocomposites: sorption and sensing for environmental applications, Chem. Eng. J. 166 (2011) 923–931. [13] L. Cumbal, A.K. SenGupta, Arsenic removal using polymer-supported hydrated iron(III) oxide nanoparticles: a role of donnan membrane effect, Environ. Sci. Technol. 39 (2005) 6508–6515. [14] N. Horzum, M.M. Demir, M. Nairat, T. Shahwan, Chitosan fiber-supported zerovalent iron nanoparticles as a novel sorbent for sequestration of inorganic arsenic, RSC Adv. 3 (2013) 7828–7837. [15] X. Zhao, L. Lv, B. Pan, W. Zhang, S. Zhang, Q. Zhang, Polymer-supported nanocomposites for environmental application: a review, Chem. Eng. J. 170 (2011) 381–394. [16] T.M. Suzuki, J.O. Bomani, H. Matsunaga, T. Yokoyama, Preparation of porous resin loaded with crystalline hydrous zirconium oxide and its application to the removal of arsenic, React. Funct. Polym. 43 (2000) 165–172. [17] B. Pan, Z. Li, Y. Zhang, J. Xu, L. Chen, H. Dong, W. Zhang, Acid and organic resistant nano-hydrated zirconium oxide (HZO)/polystyrene hybrid adsorbent for arsenic removal from water, Chem. Eng. J. 248 (2014) 290–296. [18] B. Pan, J. Xu, B. Wu, Z. Li, X. Liu, Enhanced removal of fluoride by polystyrene anion exchanger supported hydrous zirconium oxide nanoparticles, Environ. Sci. Technol. 47 (2013) 9347–9354. [19] L.H. Sperling, Interpenetrating Polymer Networks: An Overview, Interpenetrating Polymer Networks, American Chemical Society, 1994. pp. 3–38.

[20] G. Kickelbick, Concepts for the incorporation of inorganic building blocks into organic polymers on a nanoscale, Prog. Polym. Sci. 28 (2003) 83–114. [21] X. Luo, C. Wang, L. Wang, F. Deng, S. Luo, X. Tu, C. Au, Nanocomposites of graphene oxide-hydrated zirconium oxide for simultaneous removal of As(III) and As(V) from water, Chem. Eng. J. 220 (2013) 98–106. [22] H. Cui, Y. Su, Q. Li, S. Gao, J.K. Shang, Exceptional arsenic (III, V) removal performance of highly porous, nanostructured ZrO2 spheres for fixed bed reactors and the full-scale system modeling, Water Res. 47 (2013) 6258–6268. [23] H. Cui, Q. Li, S. Gao, J.K. Shang, Strong adsorption of arsenic species by amorphous zirconium oxide nanoparticles, J. Ind. Eng. Chem. 18 (2012) 1418– 1427. [24] A. Bortun, M. Bortun, J. Pardini, S.A. Khainakov, J.R. García, Synthesis and characterization of a mesoporous hydrous zirconium oxide used for arsenic removal from drinking water, Mater. Res. Bull. 45 (2010) 142–148. [25] B.K. Biswas, J.-I. Inoue, K. Inoue, K.N. Ghimire, H. Harada, K. Ohto, H. Kawakita, Adsorptive removal of As(V) and As(III) from water by a Zr(IV)-loaded orange waste gel, J. Hazard. Mater. 154 (2008) 1066–1074. [26] S. Mandal, T. Padhi, R.K. Patel, Studies on the removal of arsenic (III) from water by a novel hybrid material, J. Hazard. Mater. 192 (2011) 899–908. [27] Z. Ren, G. Zhang, J. Paul Chen, Adsorptive removal of arsenic from water by an iron–zirconium binary oxide adsorbent, J. Colloid Interface Sci. 358 (2011) 230–237. [28] C.H. Campos, B.F. Urbano, B.L. Rivas, Synthesis and characterization of organicinorganic hybrid composites from poly(acrylic acid)-[3-(trimethoxysilyl) propyl methacrylate]-Al2O3, Compos. B 57 (2014) 1–7. [29] C.H. Campos, B.F. Urbano, B.L. Rivas, Hybrid composites from poly[(4vinylbenzyl)trimethylammonium chloride]–metal oxide using simultaneous radical polymerization/sol–gel synthesis, Mater. Lett. 131 (2014) 198–202. [30] Ł. John, P. Sobota, Alkoxide molecular precursors for nanomaterials: a one step strategy for oxide ceramics, in: W. Wunderlich (Ed.), Ceramic Materials, InTech, Rijeka, 2010, pp. 69–86. [31] X. Dou, D. Mohan, C.U. Pittman Jr, S. Yang, Remediating fluoride from water using hydrous zirconium oxide, Chem. Eng. J. 198–199 (2012) 236–245. [32] C. Hang, Q. Li, S. Gao, J.K. Shang, As(III) and As(V) adsorption by hydrous zirconium oxide nanoparticles synthesized by a hydrothermal process followed with heat treatment, Ind. Eng. Chem. Res. 51 (2011) 353–361. [33] F. Bauer, H.-J. Glasel, U. Decker, H. Ernst, A. Freyer, E. Hartmann, V. Sauerland, R. Mehnert, Trialkoxysilane grafting onto nanoparticles for the preparation of clear coat polyacrylate systems with excellent scratch performance, Prog. Org. Coat. 47 (2003) 147–153. [34] G. Kickelbick, Concepts for the incorporation of inorganic building blocks into organic polymers on a nanoscale, Prog. Polym. Sci. (Oxford) 28 (2003) 83–114. [35] W. Zhou, S. Yuan, L. Tan, Y. Chen, Y. Huang, Crystallization, morphology, and mechanical properties of poly(butylene succinate)/poly(ethylene oxide)polyhedral oligomeric silsesquioxane nanocomposites, Polym. Eng. Sci. 52 (2012) 2063–2070. [36] J. Mu, Y. Liu, S. Zheng, Inorganic–organic interpenetrating polymer networks involving polyhedral oligomeric silsesquioxane and poly(ethylene oxide), Polymer 48 (2007) 1176–1184. [37] J. Normatov, M.S. Silverstein, Porous interpenetrating network hybrids synthesized within high internal phase emulsions, Polymer 48 (2007) 6648– 6655. [38] S. Kango, S. Kalia, A. Celli, J. Njuguna, Y. Habibi, R. Kumar, Surface modification of inorganic nanoparticles for development of organic–inorganic nanocomposites—a review, Prog. Polym. Sci. 38 (2013) 1232–1261. [39] S. Pavlidou, C.D. Papaspyrides, A review on polymer–layered silicate nanocomposites, Prog. Polym. Sci. 33 (2008) 1119–1198. [40] R. Bhargava, S.Q. Wang, J.L. Koenig, FTIR microspectroscopy of polymeric systems, Adv. Polym. Sci. (2003) 137–191. [41] M. Kosmulski, The pH-dependent surface charging and points of zero charge: V. Update, J. Colloid Interface Sci. 353 (2011) 1–15. [42] M. Kosmulski, The significance of the points of zero charge of zirconium (hydr) oxide reported in the literature, J. Dispersion Sci. Technol. 23 (2002) 529–538. [43] G. Jegadeesan, S.R. Al-Abed, V. Sundaram, H. Choi, K.G. Scheckel, D.D. Dionysiou, Arsenic sorption on TiO2 nanoparticles: size and crystallinity effects, Water Res. 44 (2010) 965–973. [44] F.-C. Wu, R.-L. Tseng, R.-S. Juang, Initial behavior of intraparticle diffusion model used in the description of adsorption kinetics, Chem. Eng. J. 153 (2009) 1–8. [45] S. Goldberg, C.T. Johnston, Mechanisms of arsenic adsorption on amorphous oxides evaluated using macroscopic measurements, vibrational spectroscopy, and surface complexation modeling, J. Colloid Interface Sci. 234 (2001) 204– 216. [46] U.S. Epa, Arsenic Treatment Technologies for Soil, Waste, and Water, EPA/ National Service Center for Environmental Publications, U.S, 2002. [47] D.A. Clifford, G.L. Ghurye, Metal-oxide adsorption, ion exchange, and coagulation-microfiltration for arsenic removal from water, in: W.T. Frankenberger (Ed.), Environmental Chemistry of Arsenic, Marcel Dekker Inc., New York, 2002.