Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration

CLAY-03860; No of Pages 8 Applied Clay Science xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Applied Clay Science journal homepage: www.elsevier.com/locate/clay

Research paper

Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration Yu Takaki a, Xinhong Qiu a,b,⁎, Tsuyoshi Hirajima a, Keiko Sasaki a,⁎ a b

Department of Earth Resources Engineering, Kyushu University, Fukuoka 819-0395, Japan School of Chemistry and Environmental Engineering, Wuhan Institute of Technology, Wuhan 430073, PR China

a r t i c l e

i n f o

Article history: Received 2 September 2015 Received in revised form 6 May 2016 Accepted 9 May 2016 Available online xxxx Keywords: Arsenate Hydrocalumite Trimetallic layered double hydroxides Dissolution–reprecipitation Surface complexation

a b s t r a c t We investigated the influence of initial arsenate concentration (C0) in the 5th order of magnitude on removal of arsenate by hydrocalumite (bimetallic layered double hydroxide, LDH) and Mg-doped hydrocalumite (trimetallic LDH) from aqueous solution. These hydrocalumites were prepared by the microwave-assisted hydrothermal treatment. There is a trend that the larger adsorption density of arsenate (Qe) values is observed with bimetallic LDH under low C0 values and with trimetallic LDH under high C0 values. The transitional C0 values ranged at 2.10–2.96 mM. Comprehensively understanding characterization results for the solid residues after adsorption of arsenate by X-ray diffraction, 27Al-nuclear magnetic resonance, and scanning electron microscopy– energy dispersive X-ray, the mechanism to remove arsenate was dependent on arsenate concentrations. At low arsenate concentration, partial intercalation and dissolution–reprecipitation (DR) happened together. With increasing C0, full intercalation and DR happened to bring out one phase of arsenate-bearing hydrocalumite. Under the very high C0, DR mechanism happened at the edge sites of LDH sheets, leading that the newly formed massive precipitates block the further intercalation with nitrate. As a result, two phases of LDH were observed. The greater Qe with bimetallic LDH in low concentration comes from high crystallinity to enhance partial ion-exchange, and greater Qe with trimetallic LDH in high concentration is derived from more fragile properties to enhance DR mechanism. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Arsenic is a highly toxic pollutant. The presence of elevated concentrations of arsenic is regarded as one of the most serious environmental problems, especially in developing and underdeveloped countries (Bhumbla and Keefer, 1994; Plant et al., 2004). Long-term and excess uptake of arsenic contaminated drinking water causes many human health hazards, such as cancers, skin lesions, and nerve tissue injures (Bates et al., 1992; Morales et al., 2000). The World Health Organization and many countries have the strict guideline of b 10 μg/L arsenic in drinking water (WHO, 2006). Therefore, inexpensive and appropriate technologies for arsenate removal need to be developed. Many methods have been developed to remove excessive arsenic from water using techniques such as coagulation, electro-coagulation, precipitation, ion exchange, reverse osmosis and adsorption (Mohan and Pittman, 2007). Because of their high anion-exchange capacity and thermal stability, layered double hydroxides (LDHs) are currently receiving considerable ⁎ Corresponding authors. E-mail addresses: [email protected] (X. Qiu), [email protected] (K. Sasaki).

attention for a wide variety of applications in environmental remediation (Goh et al., 2008). The chemical composition of LDH can be defined as [M(II)1 − xM(III)x(OH)2]x+(An−)x/n·nH2O, in which M(II) and M(III) are divalent and trivalent metallic cations located in the host layers, An− is the interlayer anion, m is the number of interlayer water molecules, and x is the molar ratio of M(III)/(M(II) + M(III)) (Evans and Slade, 2006). LDH has different properties depending on the combination of divalent and trivalent cations and their molar ratio (Guo et al., 2005; Koilraj and Kannan, 2010; Guo et al., 2012; Lv et al., 2012; Zhou et al., 2015). There are several reports that trimetallic LDH influenced the uptake of anionic species. Zhou et al. explained that MgCaFe-LDH lowered adsorption capacity of selenite and chromate compared with MgFe-LDH due to collapse of trimetallic LDH after dissolution of Ca2+ (Zhou et al., 2015). Meanwhile Koilraj and Kannan observed that ZnAlZr-LDH enhanced adsorption capacity of phosphate than ZnAlLDH, and exchange precipitation of Zn(PO4)2·4H2O happened after dissolution of Zn2+ from ZnAlZr-LDH (Koilraj and Kannan, 2010). These findings apparently seem to be inconsistent, however, both phenomena is based on chemically fragile characteristics of host layers in trimetallic LDH. Further investigation is necessary in details in a wide range of target anion concentrations.

http://dx.doi.org/10.1016/j.clay.2016.05.010 0169-1317/© 2016 Elsevier B.V. All rights reserved.

Please cite this article as: Takaki, Y., et al., Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration, Appl. Clay Sci. (2016), http://dx.doi.org/10.1016/j.clay.2016.05.010

2

Y. Takaki et al. / Applied Clay Science xxx (2016) xxx–xxx

To date, tremendous researches have been conducted for hydrotalcite (MgAl-LDH) (Toraishi et al., 2002; Lazaridis et al., 2002; Doušová et al., 2003; Yang et al., 2005; Bhaumik et al., 2005; Das et al., 2006; Kiso et al., 2005; Peng et al., 2005; Yang et al., 2006; Violante et al., 2009; Goh et al., 2009; Grover et al., 2010; Caporale et al., 2011), while there have been only a few reports about hydrocalumite (CaAlLDH) (Grover et al., 2010; Choi et al., 2012; Guo and Tian, 2013). Hydrocalumite is easily prepared and commonly found in cement (Liu et al., 2011), and contains more regularly arranged anions, cations, and interlayer water molecules than hydrotalcite (Vieille et al., 2003). The crystallinity of LDH depends on involved metals, and also affects the adsorption characteristics of anionic species. It has been reported that adsorption density of arsenate was larger on hydrocalumite than hydrotalcite with the same molar ratio of M(II)/M(III), since the adsorption mechanism with hydrocalumite involved arsenic exchange as well as partial dissolution and recrystallization (Grover et al., 2010). In contrast to hydrotalcite, the removal mechanism of contaminants by hydrocalumite is through ion-exchange and dissolution– reprecipitation (DR) (Liu et al., 2011). In this process, oxyanion uptake by hydrocalumite should involve the disruption of the electrostatic interactions and hydrogen bonds between the hydroxide layers and the outgoing anions, followed by the reformation of these bonds with the incoming anions (Radha et al., 2005). This means that the structural stability of hydrocalumite in the solution is closely related to the mechanism and efficiency of anion removal. Similar to hydrocalumite, green rust is unstable and can be easily transformed into magnetite and Fe(OH)2, even under anoxic conditions (Radha et al., 2005). However, the stability of green rust is improved when the concentration of multivalent oxyanions (PO3− and AsO3− 4 4 ) reaches a certain level. The phosphate and arsenate ions adsorbed on the lateral faces may act as a barrier to inhibit the release of interlayer anions (Benali et al., 2001; Bocher et al., 2004; Mitsunobu et al., 2008; Goh and Lim, 2010). If arsenate functions to stabilize the structure of hydrocalumite, the process of arsenate adsorption onto hydrocalumite related to the removal mechanism is an interesting issue. In addition, pollutant removal by LDH has mainly focused on the bimetallic LDH, and there are few reports about pollutant removal by trimetallic LDH, such as Mg-doped hydrocalumite (trimetallic LDH). Differently from the hydrotalcite, Ca2+ is coordinated with six OH groups and the water molecules in the interlayer are also directly coordinated to Ca atoms. Because of this, each interlayer water molecule occupies a certain ordered position to create a well-defined anionic interlayer. More regularly arranged anions, cations, and interlayer water molecules are observed in Ca-Al-LDH compared with Mg-Al-LDH (Pfeiffer et al., 2011; Mora et al., 2011). In addition, the solubility product (Ksp) for hydrocalumite is 10−27.10 (Liu et al., 2011) while Ksp for hydrotalcite is 10−52.12 (Ravel and Newville, 2005). It is evident that the solubility of hydrotalcite is lower than that of hydrocalumite. Therefore, the adsorption mechanism of arsenate to bimetallic LDH may change after Mg is doped into the metallic layers. However, there is a lack of detailed discussion and study on immobilization of anionic species by trimetallic LDH. In this study, we investigated adsorption of arsenate on bimetallic and trimetallic LDHs in a wide range of arsenate concentrations with 5th orders of magnitude. Using multiple characterization methods, the detailed removal mechanism of arsenate by bimetallic and trimetallic LDHs and the influence of Mg in the metallic layers are discussed. 2. Methods 2.1. Preparation of LDH Microwave-assisted hydrothermal methods were used for the synthesis of bimetallic and trimetallic LDHs. For bimetallic LDH, a solution containing specific amounts of Ca(NO3)2·4H2O (2.36 g) and Al(NO3)3·9H2O (1.86 g) adjusted to the molar ratio Ca:Al = 2:1 was

added to 50 mL of 0.50 M NaNO3. The pH was adjusted to 12.0 by 2 M NaOH maintained for 2 h. All of the chemical reagents were in special grade, purchased from WAKO chemicals (Osaka, Japan) and used without purification. The resulting slurry was transferred into a Teflon vessel to supply for microwave irradiation with a Milestone Ethos Plus microwave (Sorisole, Italy). The temperature was increased to 150 °C within 10 min and then maintained for 3 h. The cooled seriflux was treated by solid–liquid separation by super-centrifugation at 10,000 rpm for 10 min, and washed several times with ultrapure water. The resulting product was freeze-dried and called as bimetallic LDH. Trimetallic LDH was prepared in the same manner to bimetallic LDH. The only difference was that the mixed solution contained Mg2+ ions with the molar ratio of Ca:Mg:Al = 1.6:0.4:1.0. 2.2. Characterization The crystalline phases within the various bimetallic and trimetallic LDHs were characterized using an X–ray diffractometer (Ultima IV, Rigaku, Akishima, Japan) with Cu Kα radiation. The accelerating voltage and applied current were 40 kV and 40 mA, respectively, with a scanning speed of 2°/min and a scanning step of 0.02°. Brunner–Emmett– Teller (BET) specific surface area measurements were performed using a BELSORP-MR6 surface analyzer (BEL JAPAN Inc., Toyonaka, Japan) with the nitrogen adsorption method. The samples were preheated at 150 °C for 3 h for degassing. The morphologies and the elemental compositions of the products were observed using a VE-9800 scanning electron microscope (Keyence, Osaka, Japan) equipped with an energy dispersive X–ray (EDX) spectrometer at 20 kV accelerating voltage. Elemental analysis was conducted at least 5 particles per one sample and at least 6 points for one particle by SEM-EDX. 27Al-NMR spectra were collected on a JNM-ECA 800 spectrometer (JEOL, Akishima, Japan) with Delta NMR software version 4.3 using 3.2 nm MQMAS probes and a single pulse method. The resonance frequency for 27Al was 208.5 MHz at a field strength of 18.8 T. Typical acquisition parameters were spinning speed 250 kHz, pulse length 2.14 μs, and relaxation delay 1 s. Extended X-ray absorption fine structure (EXAFS) spectra of the Ca K-edge were collected on BL06 at the Kyushu Synchrotron Light Research Center (SAGA-LS, Tosu, Japan). The spectra of the samples were collected using an ionization chamber in transmission mode. The photon energy was scanned in the range of 3.7–5.6 keV for the Ca K-edge. The samples were mixed with boron nitride (BN) and pressed into pellets. All of the spectra were averaged and normalized using IFFEFIT software version 1.2.11 (Ravel and Newville, 2005). The Ca, Mg, and Al contents in the products were determined using an inductively coupled plasma-atomic emission spectrometer (ICP-AES, Perkin Elmer 8500, Yokohama, Japan) after acid decomposition with 6.5 M HNO3. The C, H, and N contents were determined by CHN analysis using Yanaco CHN Corder MT-6 Elemental Analyzer (Tokyo, Japan). 2.3. Batch adsorption tests For the adsorption experiments, 0.100 g of product was added to 40 mL of 0.002–10.2 mM KH2AsO4 as the initial concentration (C0). The mixture was shaken at 100 rpm with an 8 cm stroke in a shaker (TB-16R, Takasaki Kagaku, Kawaguchi, Japan) at room temperature. At intervals, the supernatants were taken and filtered (0.20 μm) and the total Ca, Mg, and Al concentrations were determined by ICP-AES. Concentrations of As lower than 0.02 mM were determined by hydride generation-atomic absorption spectrometry (HG-AAS, Solaar AA series, Thermo Scientific, Yokohama, Japan). The solid residues after adsorption were collected for characterization. 3. Results and discussion Based on the elemental determination using CHN analysis and ICPAES, the possible chemical formula of bimetallic and trimetallic LDHs

Please cite this article as: Takaki, Y., et al., Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration, Appl. Clay Sci. (2016), http://dx.doi.org/10.1016/j.clay.2016.05.010

Y. Takaki et al. / Applied Clay Science xxx (2016) xxx–xxx

3

Table 1 The possible chemical formula of synthesized LDH. Notation

Chemical formula

Bimetallic LDH Trimetallic LDH

Ca0.66Al0.33(OH)2·(NO3)0.23(CO3)0.05·nH2O Ca0.48Mg0.22Al0.30(OH)2·(NO3)0.19(CO3)0.06·nH2O

are summarized in Table 1. The molar ratios of M(II)/M(III) were 2.0 in bimetallic LDH and 2.3 in trimetallic LDH. Small amounts of carbonate, which may come from air during the preparation and/or drying step, were present in both LDHs. The XRD patterns of the synthesized bimetallic and trimetallic LDHs are shown in Fig. 1(a), (b). Both XRD patterns show typical peaks of LDH with sharp and high reflections for the 003 and 006 planes (Grover et al., 2010; Xu et al., 2011). According to the diffraction plane of 003, the d-value of the bimetallic and trimetallic LDHs was calculated to be around 8.589 Å. In addition, the intensities of the reflection (e.g., d003), as a measure of crystallinity, were slightly smaller in trimetallic LDH than bimetallic one. This suggests that addition of Mg decreases the structural regularity of hydrocalumite. Fig. 1(c), (d) shows SEM images of the bimetallic and trimetallic LDHs. The morphologies of both LDHs were in hexagonal plates, which are characteristic of LDH (Tian and Guo, 2014). While the particles of bimetallic LDH show smooth surfaces with clear edges, the particles of trimetallic LDH show rough surfaces with less straight edges. This is also consistent with the much higher BET specific surface area of trimetallic LDH (32.3 m2/g) than bimetallic LDH (2.6 m2/g). The adsorption kinetics of arsenate on bimetallic and trimetallic LDHs are presented in Fig. 2. The concentration of arsenate decreased from 0.18 mM to less than 0.1 μM within 20 min in adsorption on bimetallic and trimetallic LDHs (Fig. 2(a) inset). When the C0 was 2.10 mM, high adsorption efficiency was found for trimetallic LDH within 5 min (Fig. 2(b)). This phenomenon was more significant when the C0 was increased to 4.91 mM. The changes of the Ca2+ and Al3+ concentrations and pH with time were similar for both LDHs (Fig. 3). The molar ratio of dissolved Ca/Al was around 2 in most cases, suggesting the

Fig. 2. Time courses of arsenate concentrations with time during sorption of (a) 0.18 mM and (b) 2.1 mM and 4.91 mM arsenate on bimetallic LDH and trimetallic LDH. Inset in (a) has the expanded y-axis. Symbols: ●, bimetallic; ■, trimetallic.

Fig. 1. XRD patterns for (a) bimetallic LDH and (b) trimetallic LDH and their SEM images in (c) and (d). Symbols (☆) in XRD indicate peak assignment of hydrocalumite (d003 = 8.589). Horizontal bars in SEM images indicate 500 nm.

stoichiometric dissolution of hydrocalumite. However, there was a clear difference between bimetallic and trimetallic LDHs in adsorption of 4.91 mM arsenate. That is, dissolution of Al was markedly suppressed in the case of adsorption of 4.91 mM arsenate on bimetallic LDH, resulting in non-stoichiometric dissolution with dissolved Ca/Al of 11.5 (Fig. 3(a), (b)). Release of Mg2 + ions from trimetallic LDH was not detected in any cases. It is also notable that the equilibrium pH was lower when the C0 was larger (Fig. 3(c)). This suggests that consumption of hydroxyl ions more easily happens under the higher arsenate concentration. Including the results derived from Fig. 2, the adsorption isotherms of arsenate on bimetallic and trimetallic LDHs were obtained in the 5th order of magnitude of arsenate concentration as shown in Fig. 4. There is a trend that the larger adsorption densities (Qe) are observed with bimetallic LDH under low C0 values and with trimetallic LDH under high C0 values. The transitional C0 values ranged at 2.10–2.96 mM. At C0 = 10.2 mM, the Qe with trimetallic LDH was around 1.65 mmol/g, while it was 1.38 mmol/g with bimetallic LDH (Fig. 4(c)). This result covers

Please cite this article as: Takaki, Y., et al., Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration, Appl. Clay Sci. (2016), http://dx.doi.org/10.1016/j.clay.2016.05.010

4

Y. Takaki et al. / Applied Clay Science xxx (2016) xxx–xxx

Fig. 3. Time courses of released (a) Ca and (b) Al concentrations from bimetallic and trimetallic LDHs and (c) pH during adsorption of 0.18–4.91 mM arsenate.

the wide range of arsenate concentration on adsorption isotherm with bimetallic and trimetallic LDHs. The previously reported Qe values on hydrocalumite, which are 1.14 mmol/g (Grover et al., 2010) and 1.13 mmol/g (Choi et al., 2012), are also consistent with the present results. The adsorption performance and mechanism of arsenate on bimetallic and trimetallic LDHs are reflected by the concentration range of arsenate. To investigate the adsorption mechanisms on the different LDH, the solid residues after adsorption of arsenate were characterized by several methods. The XRD patterns of the solid residues collected after adsorption of 0.10 mM arsenate on bimetallic and trimetallic LDHs showed two types of LDH with different interlayer space (d) (Fig. 5): 7.575 Å at 11.64° and 8.589 Å at 10.22°. The latter is the same as the original dvalue. Because the metallic layer of LDH is considered to be 4.80 Å in thickness (Sá et al., 2013), the interlayer spaces occupied by the anionic species in the two types of LDH were calculated to be approximately 2.775 Å and 3.789 Å, suggesting that two major anionic species were intercalated into the bimetallic and trimetallic LDHs. In addition, the new LDH with d003 = 7.575 Å was more obviously formed in trimetallic LDH than bimetallic one (Fig. 5(b)). Only one type of LDH with d003 = 7.575 Å was found in the solid residues after adsorption of 1.20– 2.96 mM arsenate on bimetallic LDH and 0.83–4.91 mM arsenate on trimetallic LDH. These peak intensities were always very low. In Fig. 4(b), (c), in case of C0 ≥ 2.10 mM, Qe values were always greater in trimetallic LDH than bimetallic one. This trend might be partially caused by the larger BET specific surface area in trimetallic LDH as above. However, the XRD results in Fig. 5 suggest that new hydrocalumite is formed after partial dissolution of the original hydrocalumite, so the large BET specific surface area would not be the predominant reason. When the

C0 is more than 4.91 mM on bimetallic LDH, two types of LDH with interlayer space of 7.575 Å and 8.589 Å were again present in the solid residues. This phenomenon occurred in trimetallic LDH when the C0 was more than 5.91 mM. In addition, when the C0 was up to 10.2 mM, the LDH with d003 = 7.575 Å almost disappeared in both bimetallic and trimetallic LDHs. 27Al-NMR spectra of some solid residues before and after adsorption of 0.10–10.2 mM arsenate on bimetallic and trimetallic LDHs are shown in Fig. S2. Guo and Tian (2013) suggested that adsorption of arsenate on hydrocalumite was mainly caused by the formation of arsenic–calcium containing minerals. However, based on the results of XRD (Fig. 5), secondary minerals, such as johnbaumite (Ca5(AsO4)3(OH)), sainfeldite (Ca5(AsO3OH)2(AsO4)2), and bulachite (Al2(AsO4)(OH)3) mentioned by Guo and Tian, were not found. The k3-weighted Ca K-edge EXAFS spectra and the corresponding radical structure functions (RSFs) for bimetallic and trimetallic LDHs before and after adsorption of 2.10 mM arsenate are shown in Fig. S1(a), (b). All of the spectra are quite similar, indicating that the coordination of Ca atoms before and after arsenate adsorption did not change. Two intense peaks were observed in the RSF at R + ΔR = 1.9 Å and 2.9 Å, corresponding to Ca–O and Ca–M (M = Ca or Al) single scattering, respectively. The heights of the Ca–M peaks significantly decreased after adsorption of arsenate (Fig. S1(b)), suggesting a decrease in crystallinity. Interestingly, the EXAFS spectrum of Ca3(AsO4)2 shows clearly different features for bimetallic and trimetallic LDHs in both k space and R space. It has been previously reported that adsorption of arsenate on hydrocalumite leads to partial structural transformation of LDH to Ca3(AsO4)2 (Guo and Tian, 2013). However, this phenomenon was not observed in the present work. Therefore, the formation of calcium arsenate-containing minerals is

Fig. 4. Adsorption isotherms of arsenate onto bimetallic and trimetallic LDHs at 25 °C. The C0 is (a) lower than 0.83 mM, (b) lower than 2.10 mM, and (c) lower than 10.2 mM. Dotted lines in (a) are the maximum concentration limit (MCL) of arsenic with 100 μg/L for industrial drainages and 10 μg/L for drinking waters. The numbers in the figure indicate the C0.

Please cite this article as: Takaki, Y., et al., Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration, Appl. Clay Sci. (2016), http://dx.doi.org/10.1016/j.clay.2016.05.010

Y. Takaki et al. / Applied Clay Science xxx (2016) xxx–xxx

5

Fig. 5. XRD patterns for the original bimetallic and trimetallic LDHs and their solid residues after adsorption of 0.10–10.2 mM arsenate. (a) Bimetallic LDH, (b) trimetallic LDH.

not the main mechanism of arsenate removal here. This may depend on the experimental conditions, such as pH, concentrations of metallic ions, and the crystallinity of bimetallic LDH from the preparation methods. The representative SEM-EDX results of the original bimetallic and trimetallic LDHs and solid residues after adsorption of 0.10–10.2 mM arsenate are presented in Fig. 6 and Table 2. At least 6 points on one hexagonal sheet, elemental determination of Ca, Mg, Al and As was conducted. The EDX results show the molar ratio of Al/Ca in bimetallic LDH is 0.37–0.59 (Table 2(A)), which is mostly consistent with results

of ICP-AES in Table 1. In trimetallic LDH, Mg contents are not always homogeneous in a specific particle but also may be dependent on particles. The molar ratio of Al/(Ca + Mg) is 0.27–0.51 in the observed particle (Table 2(E)), and the average of Al/(Ca + Mg) by EDX is 0.41 which is also consistent with the results in Table 1, showing that the ratio of M(II):M(III) is mostly kept to 3:1–2:1 in trimetallic LDH. For the solid residues after adsorption of 0.10 mM arsenate on both LDHs, the arsenate content was less than 1 atomic% at any spots as long as observed and sometimes under the detection limit (Table 2(B), (F)).

Fig. 6. SEM images for the original bimetallic and trimetallic LDHs and their solid residues after adsorption of 0.10–10.2 mM arsenate. Horizontal bars indicate 1 μm for (E) and (H), and 2 μm for others. EDX analysis was conducted at cross points.

Please cite this article as: Takaki, Y., et al., Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration, Appl. Clay Sci. (2016), http://dx.doi.org/10.1016/j.clay.2016.05.010

6

Y. Takaki et al. / Applied Clay Science xxx (2016) xxx–xxx

This causes to maintain the original layer distance with small amounts of arsenate-bearing hydrocalumite after intercalation with nitrate, resulting in two d003 values in XRD patterns (Fig. 5). Molar ratio of M(II):M(III) in host layers of LDH slightly changed due to dissolution (Fig. 3). The molar ratio of As/Al is varying from under the detection limit to 0.29 (0.10 ± 0.09, n = 8) in bimetallic LDH and from under the detection limit to 0.21 (0.08 ± 0.09, n = 7) in trimetallic LDH (Table 2(B), (F)). The formation of new LDH intercalating HAsO2− 4 should occur locally followed by dissolution. Considering the equilibrated water chemistry, the ionic activity products for hydrocalumite are larger than the expected solubility product. The formation of new hydrocalumite at least partially occurs through dissolution of the original hydrocalumite followed by re-precipitation of arsenate-bearing hydrocalumite (DR mechanism). When the C0 is 0.10 mM, partial intercalation should happen. The reason why the large fluctuation was observed by SEM-EDX is caused by heterogeneous distribution of arsenate in LDH particles. The DR mechanism has been found with hydrotalcite at the first time by Allada et al. (2002). The DR mechanism has been also demonstrated with hydrocalumite by Radha et al. (2005). Because of the relatively high solubility of hydrocalumite under pH 5 (Boclair and Braterman, 1995; Qian et al., 2012), bimetallic and trimetallic LDHs are dissolved to some extent as follows: Ca1 − x Alx ðOHÞ2  ðNO3 Þy ðCO3 Þz  nH2 O⇄ð1−xÞCa2þ þ xAlðOHÞ− 4 2− þ ð2−4xÞOH− þ yNO− 3 þ zCO3 þ nH2 O

ð3Þ The surface-complexation with M\\O would happen not only at the edges of the LDH sheets but also on the surfaces to contribute to the stabilization of metallic ions. The upper limit of C0, providing one d003 value of XRD for solid residues, is larger in trimetallic LDH than bimetallic one Table 2 Elemental compositions in atomic% and the molar ratios of Al/Ca and As/Ca before and after adsorption of arsenate on bimetallic and trimetallic LDHs obtained from SEM-EDX as shown in Fig. 6 Bimetallic LDH

Ca

(A) Starting material

(a) (b) (c) (d) (e) (f) (a) (b) (c) (d) (e) (f) (g) (h) (a) (b) (c) (d) (e) (f) (a) (b) (c) (d) (e) (f)

(B) 0.10 mM As

ð1Þ

where x ranged from 0 to 1 and equals to (y + 2z). During this process, the pH value increased to N 10.0 (Fig. 3(c)). Under this condition, the predominant species of arsenic is HAsO24 − (Jönsson and Sherman, 2008). The attractive force between arsenate and the metallic layer in the LDH is greater for multivalent anions than monovalent anions (Miyata, 1983). Thus, the interlayer space is more easily occupied by HAsO2− than NO− 4 3 inside newly precipitated LDH as shown in the following reaction, to give d003 of 7.575 in Fig. 5. − 2− ð1−xÞCa2þ þ xAlðOHÞ− 4 þ ð2−4xÞOH þ x=2HAsO4 þ nH2 O⇄Ca1 − x Alx ðOHÞ2 ðHAsO4 Þx=2  nH2 O

coordinated with arsenate which is inner-complexed with `Al\\OH as shown in Eq (3).

(C) 1.60 mM As

(D) 10.2 mM As

ð2Þ

The representative examples in a range of C0, where one d003 value was observed in Fig. 5, can be selected to 1.60 mM and 1.20 mM of C0 for bimetallic and trimetallic LDHs. According to the EDX results for these solid residues, As contents increased to 1.20–3.65 atomic% (Fig. 6(c), (g)). The molar ratio of As/Al is varying from 0.43 to 1.02 (0.68 ± 0.21, n = 6) in bimetallic LDH and from 0.26 to 1.41 (0.64 ± 0.38, n = 6) in trimetallic LDH (Table 2(C), (G)). The standard deviation is relatively smaller than the lower arsenate concentration range. This result suggests that there is a trend of the replacement of the original NO− 3 with arsenate, leading to the right shift of original d003 position in XRD (Fig. 5). In the SEM images for the solid residues after adsorption of 2.10 mM arsenate on bimetallic and trimetallic LDHs, small particles appeared to gather around the hexagonal sheets, which were partially maintained (Fig. S3). This supports that partial dissolution and new precipitation of hydrocalumite occurred as well as intercalation. Compared with lower arsenate concentrations, intercalation would be further enhanced and also DR reaction promotes to precipitate new hydrocalumite bearing arsenate because larger arsenate concentrations are present. As a result, only one d003 value, assigned to arsenate-bearing hydrocalumite, was observed with small intensities (Fig. 5). In addition, the surface complexation as below (Eq (3)) might also occur, when the C0 was 4.91 mM as shown in Fig. 3(a), (b), where the molar ratio of released Ca/Al was much smaller than 2. This phenomena was more obvious in bimetallic LDH than trimetallic one, but never observed when the C0 was lower than 2.10 mM. In bimetallic LDH, metallic layers are more stable, so large amount of arsenate is more favorable to coordinate with `Al\\OH sites in LDH, resulting in stabilization of Al. Meanwhile, Ca2 + ions are released and some of them might be

Trimetallic LDH (E) Starting material

(F) 0.10 mM As

(G) 1.20 mM As

(H) 10.2 mM As

Ca

(a) (b) (c) (d) (e) (f) (g) (h) (a) (b) (c) (d) (e) (f) (g) (a) (b) (c) (d) (e) (f) (a) (b) (c) (d) (e) (f) (g)

Al

Mg O

11.16 5.75 – 11.45 6.19 – 11.27 6.68 – 13.77 7.05 – 10.03 4.36 – 10.91 4.08 – 12.33 8.34 – 12.90 5.07 –– 13.38 5.23 – 11.44 5.24 – 7.48 3.43 – 12.15 4.80 – 11.32 4.24 – 11.99 6.19 –– 11.67 5.45 – 12.47 4.38 – 10.96 5.99 – 7.91 5.10 – 8.59 3.25 – 13.64 3.21 – 9.73 3.66 – 9.20 3.34 – 11.04 4.53 –– 5.88 2.83 – 10.76 11.16 – 11.99 6.06 – Al

Mg

O

83.08 82.37 82.05 79.19 85.61 85.02 78.82 82.03 81.38 82.38 88.11 82.68 84.13 81.23 80.02 79.50 80.10 84.78 85.61 79.86 86.61 87.02 84.43 87.54 72.06 81.18 As

As

Al/Ca As/Ca As/Al

– – –– – – – 0.50 n.d. n.d. 0.93 0.98 0.37 0.31 0.59 2.86 3.65 2.94 2.21 2.54 3.29 n.d. 0.44 n.d. 3.75 6.02 0.77

0.52 0.54 0.59 0.51 0.43 0.37 0.68 0.39 0.39 0.46 0.46 0.40 0.37 0.52 0.47 0.35 0.55 0.64 0.38 0.24 0.38 0.36 0.41 0.48 1.04 0.51

– –– – – – – 0.04 n.d. n.d. 0.08 0.13 0.03 0.03 0.05 0.25 0.29 0.27 0.28 0.30 0.24 n.d. 0.05 n.d. 0.64 0.56 0.06

Al/(Ca

As/(Ca

+ Mg)

+ Mg)

5.20 7.47 10.44 76.89 – 0.48 5.54 8.53 11.11 74.82 – 0.51 6.01 7.93 11.64 74.42 – 0.45 10.20 4.35 5.87 79.58 –– 0.27 7.55 4.57 4.58 83.30 – 0.38 5.31 6.23 9.72 78.74 – 0.41 7.58 7.66 11.79 72.97 – 0.40 9.54 6.19 7.67 76.61 – 0.36 11.34 6.41 9.12 72.79 0.34 0.31 7.40 3.14 5.14 84.32 n.d. 0.25 17.83 6.97 11.78 62.46 0.96 0.24 9.82 3.85 5.88 79.63 0.82 0.25 7.01 2.90 6.04 84.05 n.d. 0.22 15.80 10.23 15.49 58.49 n.d. 0.33 3.33 2.81 8.76 84.58 0.53 0.23 5.38 3.49 9.01 80.00 2.12 0.24 6.20 1.45 7.66 82.66 2.04 0.10 3.69 4.46 10.87 79.62 1.38 0.31 6.04 2.95 7.35 81.59 2.06 0.22 5.94 2.54 7.20 82.85 1.48 0.19 3.43 4.60 12.33 78.44 1.20 0.29 5.34 2.94 7.57 83.53 0.62 0.23 5.07 2.24 7.97 84.15 0.57 0.17 5.26 2.66 7.75 83.56 0.77 0.20 5.02 3.37 8.62 81.93 1.07 0.25 3.81 2.37 7.32 83.85 2.65 0.21 4.48 2.68 6.87 83.49 2.48 0.24 5.19 2.74 7.44 81.93 2.70 0.22

– – – – – – – – 0.02 n.d. 0.03 0.05 n.d. n.d. 0.04 0.15 0.15 0.09 0.15 0.11 0.08 0.05 0.04 0.06 0.08 0.24 0.22 0.21

– – – –– – – 0.06 n.d. n.d. 0.18 0.29 0.08 0.07 0.10 0.52 0.83 0.49 0.43 0.78 1.02 n.d. 0.13 n.d. 1.33 0.54 0.13 As/Al – – – – – – – – 0.05 n.d. 0.14 0.21 n.d. n.d. 0.19 0.61 1.41 0.31 0.70 0.58 0.26 0.21 0.25 0.29 0.32 1.12 0.93 0.99

–, not determined n.d., not detected.

Please cite this article as: Takaki, Y., et al., Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration, Appl. Clay Sci. (2016), http://dx.doi.org/10.1016/j.clay.2016.05.010

Y. Takaki et al. / Applied Clay Science xxx (2016) xxx–xxx

7

Fig. 7. Illustration for adsorption mechanism of arsenate with bimetallic and trimetallic LDHs depending on the C0.

(Fig. 5). The dominant reason is considered to be derived from more fragile property of trimetallic LDH to facilitate the super-saturation of newly forming arsenate-bearing hydrocalumite. When the C0 was 10.2 mM, the distribution of arsenic on both LDHs was significantly higher at edges than at the insides of sheets, as shown in Table 2(D-d, e) for bimetallic LDH and Table 2(H-e, f, g) for trimetallic LDH. The molar ratio of As/Al is varying from under the detection limit to 1.33 (0.35 ± 0.47, n = 6) in bimetallic LDH and from under the detection limit to 1.12 (0.59 ± 0.37, n = 7) in trimetallic LDH (Table 2). These LDH sheets are considered to still keep NO− 3 between layers, which cannot go out of the guest layers of LDH to exchange with HAsO2− 4 . In the presence of high HAsO2− concentrations, after dissolution of nitrate4 bearing hydrocalumite, arsenate-bearing hydrocalumite will be furthermore easily super-saturated, leading to massive precipitation. This reprecipitation process would happen more frequently at the edge site of large hexagonal sheets, where dissolution happens. Thus, the intercalation of nitrate with arsenate is blocked at the edge sides of the layers. This interpretation is consistent with XRD results (Fig. 5), where two d003 values of LDH including nitrate and arsenate are present together on solid residues after adsorption of 4.91–10.2 mM arsenate on bimetallic LDH and 5.91–10.2 mM arsenate on trimetallic LDH. Based on the above comprehensive discussion, the adsorption mechanisms are strongly dependent on the C0, and the bordering concentration of arsenate to transit the mechanism is shifted between bimetallic and trimetallic LDHs, as illustrated in Fig. 7. The arsenate concentration range, where full intercalation and reprecipitation happen, was wider in trimetallic LDH. Considering more fragile property of metallic layers in trimetallic LDH, full intercalation starts occurring at lower arsenate concentration and blocking edge sites by massive precipitation of new LDH starts happening at higher arsenate concentrations with trimetallic LDH.

trimetallic LDHs depending on C0. With low C0, bimetallic LDH showed better performance to remove arsenate. However, when the C0 was higher than 2.96 mM, trimetallic LDH showed greater adsorption density. Through the comprehensive interpretation of solid characterization results, at least three removal mechanisms of arsenate by LDH are possible: the intercalation, DR reaction and surface complexation. For bimetallic LDH, the partial intercalation is predominant to keep d003 value for nitrate-bearing hydrocalumite when the C0 was lower than 1.20 mM. The DR reaction should also happen in some extent based on its water chemistry. With increasing C0, both mechanisms should be enhanced leading that one phase of arsenate-bearing hydrocalumite is formed. In addition the surface-complexation of arsenate occurs to suppress the released Al ions. When the C0 is increased further, the DR reaction is more enhanced because the water chemistry easily reaches to the super-saturation to arsenate-bearing hydrocalumite at the edge of the LDH sheets and the newly formed precipitates act as barriers to prevent intercalation. Thus, the adsorption mechanisms are transitional and strongly dependent on the C0. That is, the range of C0, providing one phase of arsenate-bearing hydrocalumite in XRD for solid residues, is wider in trimetallic LDH than bimetallic one. The dominant reason is considered to be derived from more fragile property of trimetallic LDH to facilitate the super-saturation of newly forming arsenate-bearing hydrocalumite. The powdery hydrocalumite-like LDH can be practically applied to encapsulation in gel using high molecular polymers, for example, for remediation of arsenate pollutant waste waters. The fabrication techniques facilitate to separate the adsorbents after removal of arsenate, and also to reuse the adsorbents by repeating adsorption and desorption.

4. Conclusions

Financial support was provided to KS by Grant-in-Aid for Scientific Research 16H02435 from Japan Society of Promotion of Science (JSPS) and Funding Program for Progress 100 (entitled with "Development of novel adsorbents for oxoanions using organo-modified layered compounds") through the operating expense grants of the Ministry of Education, Culture, Sports, Science, and Technology. XAFS/EXAFS spectra

Bimetallic and trimetallic LDHs were synthesized by the microwaveassisted method and the results show that both are as effective adsorbents for arsenate. Based on the adsorption isotherms and kinetics, the adsorption mechanism was slightly different for bimetallic and

Acknowledgements

Please cite this article as: Takaki, Y., et al., Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration, Appl. Clay Sci. (2016), http://dx.doi.org/10.1016/j.clay.2016.05.010

8

Y. Takaki et al. / Applied Clay Science xxx (2016) xxx–xxx

were collected at SAGA Light Source Kyushu Synchrotron Light Source Center (no. 2013IIK022). 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.clay.2016.05.010 References Allada, R.K., Navrotsky, A., Berbeco, H.T., Casey, W.H., 2002. Thermochemistry and aqueous solubilities of hydrotalcite-like solids. Science 296, 721–723. Bates, M.N., Smith, A.H., Hopenhayn-Rich, C., 1992. Arsenic ingestion and internal cancers: a review. Am. J. Epidemiol. 135, 462–476. Benali, O., Abdelmoula, M., Refait, P., Génin, J.M.R., 2001. Effect of orthophosphate on the oxidation products of Fe(II)-Fe(III) hydroxycarbonate: the transformation of green rust to ferrihydrite. Geochim. Cosmochim. Acta 65, 1715–1726. Bhaumik, A., Samanta, S., Mal, N.K., 2005. Efficient removal of arsenic from polluted ground water by using a layered double hydroxide exchanger. Indian J. Chem. 44, 1406–1409. Bhumbla, D.K., Keefer, R.F., 1994. Arsenic Mobilization and Bioavailability in Soils. In: Nriagu, J.O. (Ed.), Arsenic in the Environment, Part I: Cycling and Characterization. John Wiley & Sons, Inc., pp. 53–82. Bocher, F., Géhin, A., Ruby, C., Ghanbaja, J., Abdelmoula, M., Génin, J.M.R., 2004. Coprecipitation of Fe(II–III) hydroxycarbonate green rust stabilised by phosphate adsorption. Solid State Sci. 6, 117–124. Boclair, J.W., Braterman, P.S., 1995. Layered double hydroxide stability. 1. Relative stabilities of layered double hydroxides and their simple counterparts. Chem. Mater. 11, 298–302. Caporale, A.G., Pigna, M., Dynes, J.J., Cozzolino, V., Zhu, J., Violante, A., 2011. Effect of inorganic and organic ligands on the sorption/desorption of arsenate on/rom Al-Mg and Fe-Mg layered double hydroxides. J. Hazard. Mater. 198, 291–298. Choi, W.H., Ghorpade, P.A., Kim, K.B., Shin, J.W., Park, J.Y., 2012. Properties of synthetic monosulfate as a novel material for arsenic removal. J. Hazard. Mater. 227, 402–409. Sá, F.P.D., Cunha, B.N., Nunes, L.M., 2013. Effect of pH on the adsorption of Sunset Yellow FCF food dye into a layered double hydroxide (CaAl-LDH-NO3). Chem. Eng. J. 215– 216, 122–127. Das, J., Patra, B.S., Baliarsingh, N., Parida, K.M., 2006. Adsorption of phosphate by layered double hydroxides in aqueous solutions. Appl. Clay Sci. 32, 252–260. Doušová, B., Machovč, V., Koloušek, D., Kovanda, F., Dorničák, V., 2003. Sorption of As(V) species from aqueous systems. Water Air Soil Pollut. 149, 251–267. Evans, D.G., Slade, R.C.T., 2006. Structural Aspects of Layered Double Hydroxides. In: Duan, X., Evans, D.G. (Eds.), Layered Double Hydroxides. Springer-Verlag, pp. 1–87. Goh, K.H., Lim, T.T., 2010. Influences of co-existing species on the sorption of toxic oxyanions from aqueous solution by nanocrystalline Mg/Al layered double hydroxide. J. Hazard. Mater. 180, 401–408. Goh, K.H., Lim, T.T., Dong, Z., 2008. Application of layered double hydroxides for removal of oxyanions: a review. Water Res. 42, 1343–1368. Goh, K.H., Lim, T.T., Dong, Z., 2009. Enhanced arsenic removal by hydrothermally treated nanocrystalline Mg/Al layered double hydroxide with nitrate intercalation. Environ. Sci. Technol. 43, 2537–2543. Grover, K., Komarneni, S., Katsuki, H., 2010. Synthetic hydrotalcite-type and hydrocalumite-type layered double hydroxides for arsenate uptake. Appl. Clay Sci. 48, 631–637. Guo, Q., Tian, J., 2013. Removal of fluoride and arsenate from aqueous solution by hydrocalumite via precipitation and anion exchange. Chem. Eng. J. 231, 121–131. Guo, Y., Zhang, H., Zhao, L., Li, G.D., Chen, J.S., Xu, L., 2005. Synthesis and characterization of Cd-Cr and Zn-Cd-Cr layered double hydroxides intercalated with dodecyl sulfate. J. Solid State Chem. 178, 1830–1836. Guo, Y., Zhu, Z., Qiu, Y., Zhao, J., 2012. Adsorption of arsenate on Cu/Mg/Fe/La layered double hydroxide from aqueous solutions. J. Hazard. Mater. 239, 279–288. Jönsson, J., Sherman, D.M., 2008. Sorption of As(III) and As(V) to siderite, green rust (fougerite) and magnetite: implications for arsenic release in anoxic groundwaters. Chem. Geol. 255, 173–181.

Kiso, Y., Jung, Y.J., Yamada, T., Nagai, M., Min, K.S., 2005. Removal properties of arsenic compounds with synthetic hydrotalcite compounds. Water Supply 5, 75–81. Koilraj, P., Kannan, S., 2010. Phosphate uptake behavior of ZnAlZr ternary layered double hydroxides through surface precipitation. J. Colloid Interface Sci. 341, 289–297. Lazaridis, N.K., Hourzemanoglou, A., Matis, K.A., 2002. Flotation of metal-loaded clay anion exchangers. Part II: The case of arsenates. Chemosphere 47, 319–324. Liu, Q., Li, Y., Zhang, J., Chi, Y., Ruan, X., Liu, J., Qian, G., 2011. Effective removal of zinc from aqueous solution by hydrocalumite. Chem. Eng. J. 175, 33–38. Lv, T., Ma, W., Xin, G., Wang, R., Xu, J., Liu, D., Liu, F., Pan, D., 2012. Physicochemical characterization and sorption behavior of Mg–Ca–Al (NO3) hydrotalcite-like compounds toward removal of fluoride from protein solutions. J. Hazard. Mater. 237–238, 121–132. Mitsunobu, S., Takahashi, Y., Sakai, Y., Inumaru, K., 2008. Interaction of synthetic sulfate green rust with antimony (V). Environ. Sci. Technol. 43, 318–323. Miyata, S., 1983. Anion-exchange properties of hydrotalcite-like compounds. Clay Clay Miner. 31, 305–311. Mohan, D., Pittman Jr., C.U., 2007. Arsenic removal from water/wastewater using adsorbents -a critical review. J. Hazard. Mater. 142, 1–53. Mora, M., López, M.I., Jiménez-Sanchidrián, C., Ruiz, J.R., 2011. Near- and mid-infrared spectroscopy study of synthetic hydrocalumites. Solid State Sci. 13, 101–105. Morales, K.H., Ryan, L., Kuo, T., Wu, M., Chen, C., 2000. Risk of internal cancers from arsenic in drinking water. Environ. Health Perspect. 108, 655–661. Peng, S., Yang, Y., Chen, T., Jiang, S., Xu, H., 2005. Adsorption of arsenate anion by Mg–Al anionic clay. J. Chin. Ceram. Soc. 33, 1023–1027. Pfeiffer, H., Ávalos-Rendón, T., Lima, E., Valente, J.S., 2011. Thermochemical and cyclability analyses of the CO2 absorption process on a Ca/Al layered double hydroxide. J. Environ. Eng. 137, 1058–1065. Plant, J.A., Kinniburgh, D.G., Smedley, P.L., Fordyce, F.M., Klinck, B.A., 2004. Arsenic and Selenium. In: Holland, H.D., Tunekian, K.K. (Eds.), Treatise on Geochemistry. Elsevier Ltd., pp. 17–66. Qian, G., Feng, L., Zhou, J.Z., Xu, Y., Liu, J., Zhang, J., Xu, Z.P., 2012. Solubility product (Ksp)controlled removal of chromate and phosphate by hydrocalumite. Chem. Eng. J. 181– 182, 251–258. Radha, A.V., Kamath, P.V., Shivakumara, C., 2005. Mechanism of the anion exchange reactions of the layered double hydroxides (LDHs) of Ca and Mg with Al. Solid State Sci. 7, 1180–1187. Ravel, Á., Newville, M., 2005. ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray absorption spectroscopy using IFEFFIT. J. Synchrotron Radiat. 12, 537–541. Tian, J., Guo, Q., 2014. Thermal decomposition of hydrocalumite over a temperature range of 400–1500 °C and its structure reconstruction in water. J. Chem. 2014, 1–8. 2− − Toraishi, T., Nagasaki, S., Tanaka, S., 2002. Adsorption behavior of IO− 3 by CO3 - and NO3 hydrotalcite. Appl. Clay Sci. 22, 17–23. Vieille, L., Rousselot, I., Leroux, F., Besse, J.P., Taviot-Guého, C., 2003. Hydrocalumite and its polymer derivatives. 1. Reversible thermal behavior of Friedel's salt: a direct observation by means of high-temperature in situ powder X-ray diffraction. Chem. Mater. 15, 4361–4368. Violante, A., Pucci, M., Cozzolino, V., Zhu, J., Pigna, M., 2009. Sorption/desorption of arsenate on/from Mg–Al layered double hydroxides: influence of phosphate. J. Colloid Interface Sci. 333, 63–70. WHO, 2006. Guidelines for Drinking-water Quality: First Addendum to Volume 1, Recommendations. World Health Organization. Xu, S., Zhang, B., Chen, Z., Yu, J., Evans, D.G., Zhang, F., 2011. A general and scalable formulation of pure CaAl-layered double hydroxide via an organic/water solution route. Ind. Eng. Chem. Res. 50, 6567–6572. Yang, L., Shahrivari, Z., Liu, P.K.T., Sahimi, M., Tsotsis, T.T., 2005. Removal of trace levels of arsenic and selenium from aqueous solutions by calcined and uncalcined layered double hydroxides (LDH). Ind. Eng. Chem. Res. 44, 6804–6815. Yang, L., Dadwhal, M., Shahrivari, Z., Ostwal, M., Liu, P.K.T., Sahimi, M., Tsotsis, T.T., 2006. Adsorption of arsenic on layered double hydroxides: effect of the particle size. Ind. Eng. Chem. Res. 45, 4742–4751. Zhou, J., Su, Y., Zhang, J., Xu, X., Zhao, J., Qian, G., Xu, Y., 2015. Distribution of OH bond to metal-oxide in Mg3 − xCaxFe-layered double hydroxide (x = 0–1.5): its role in adsorption of selenate and chromate. Chem. Eng. J. 262, 383–389.

Please cite this article as: Takaki, Y., et al., Removal mechanism of arsenate by bimetallic and trimetallic hydrocalumites depending on arsenate concentration, Appl. Clay Sci. (2016), http://dx.doi.org/10.1016/j.clay.2016.05.010