Development of nanoscale zirconium molybdate embedded anion exchange resin for selective removal of phosphate

Water Research 134 (2018) 22e31

Contents lists available at ScienceDirect

Water Research journal homepage: www.elsevier.com/locate/watres

Development of nanoscale zirconium molybdate embedded anion exchange resin for selective removal of phosphate Trung Huu Bui a, Sung Pil Hong a, Jeyong Yoon a, b, * a School of Chemical and Biological Engineering, College of Engineering, Institute of Chemical Process, Seoul National University (SNU), Gwanak-gu, Daehak-dong, Seoul, 151-742, Republic of Korea b Asian Institute for Energy, Environment & Sustainability(AIEES), Seoul National University (SNU), Gwanak-gu, Daehak-dong, Seoul, 151-742, Republic of Korea

a r t i c l e i n f o

a b s t r a c t

Article history: Received 13 December 2017 Received in revised form 22 January 2018 Accepted 25 January 2018

Development of a selective adsorbent with an enhanced removal efficiency for phosphate from wastewater is urgently needed. Here, a hybrid adsorbent of nanoscale zirconium molybdate embedded in a macroporous anion exchange resin (ZMAE) is proposed for the selective removal of phosphate. The ZMAE consists of a low agglomeration of zirconium molybdate nanoparticles (ZM NPs) dispersed within the structure of the anion exchange (AE) resin. As major results, the phosphate adsorption capacity of the ZMAE (26.1 mg-P/g) in the presence of excess sulfate (5 mM) is superior to that of the pristine AE resin (1.8 mg-P/g) although their phosphate uptake capacity was similar in the absence of sulfate and these results were supported by the high selectivity coefficient of the ZMAE toward phosphate over sulfate (SPO4/SO4) more than 100 times compared to the pristine AE resin. This superior selective performance of the ZMAE for phosphate in the presence of sulfate ions is well explained by the role of the ZM NPs that contributed to 69% of the phosphate capacity which is based on an observation that the phosphate adsorption capacity of the ZM NPs is not affected by the presence of sulfate. In addition, the behavior of the selective phosphate removal by the ZMAE was well demonstrated by not only in the batch mode experiment with simulated Mekong river water and representative wastewater effluent but also in a column test. © 2018 Elsevier Ltd. All rights reserved.

Keywords: Selective removal of phosphate Zirconium molybdate Anion exchange resin Hybrid Phosphate removal

1. Introduction The occurrence of excess phosphate is one of the main sources causing the eutrophication of lakes, rivers, and coastal regions. Eutrophication has become a widespread problem deteriorating water quality which has led to an accumulation of organic maters (plants and algae) and the development of color and odor in water as well (Su et al., 2013; Yang et al., 2014). Therefore, developing effective technologies for reducing phosphate in wastewater going into river and lake waters has been one of the urgent strategies for controlling the eutrophication problem in recent decades. Among current technologies, adsorption technologies have been an alternative to biological and

* Corresponding author. School of Chemical and Biological Engineering, College of Engineering, Institute of Chemical Process, Seoul National University (SNU), Gwanak-gu, Daehak-dong, Seoul, 151-742, Republic of Korea. E-mail address: [email protected] (J. Yoon). https://doi.org/10.1016/j.watres.2018.01.061 0043-1354/© 2018 Elsevier Ltd. All rights reserved.

chemical precipitation processes due to their simple operation, environmental friendliness, and effective removal of low level phosphate concentrations (Li et al., 2014; Su et al., 2013). Thus, developing selective adsorbents for improving the efficiency of the adsorption process has become a critical issue. In recent years, organic-inorganic nanocomposites based adsorbent, which consist of inorganic nanoparticles (NPs) and organic support matrix, have been considered as alternative adsorbents due to their selective performance toward target ions. In addition, this approach for hybrid adsorbents is presumed to be one of the solutions for overcoming the inherent problems of NPs, such as poor mechanical/hydraulic properties, clogging the column, and separation difficulty (Pan et al., 2009a, 2014). For this purpose, various composite adsorbents have been synthesized and examined for phosphate adsorption, which consist of metal oxides NPs (e.g., hydrated ferric oxide and zirconium oxide) dispersed into a porous structure of an organic host (e.g., activated carbon/graphite (Xu et al., 2015; Yao et al., 2013; Zhang et al., 2015), mesoporous silica (Zheng et al., 2016), chitosan

T.H. Bui et al. / Water Research 134 (2018) 22e31

(Jiang et al., 2013; Liu and Zhang, 2015), zeolites (Guaya et al., 2015; Xie et al., 2015), or functional polymer (Blaney et al., 2007; Pan et al., 2009c; Song et al., 2016; You et al., 2016)). Particularly, ion exchange polymers have been reported as good candidates for a polymer matrix of a hybrid adsorbent due to their Donnan effect resulting from the fixed charges (Cumbal and SenGupta, 2005; Pan et al., 2013). Despite the limited success of the anion exchange resin as a matrix for the enhancing adsorption capacity and efficiency for phosphate removal (Cumbal and SenGupta, 2005; Du et al., 2013), selective adsorption for phosphate removal remains a challenging issue because the anion exchanger resin in a hybrid adsorbent rapidly loses its adsorption capacity in the presences of excess sulfate ions, which is inevitable in wastewater conditions (Awual and Jyo, 2011; Pan et al., 2009b). Thus, developing a selective adsorbent is urgently required to achieve a high removal efficiency of phosphate from wastewater (Pan et al., 2009b). Therefore, the aim of this study was to develop a novel molybdate-based nanocomposite adsorbent for selective phosphate removal, which was a hybrid of nanoscale zirconium molybdate and a macroporous anion exchange resin (ZMAE). The performance of the ZMAE was evaluated by adsorption studies in the absences/presence of excessive sulfate (5 mM) compared with that of the pristine AE resin. Furthermore, the ZMAE was used with simulated Mekong river water in batch mode and simulated waste effluent in a fixed-bed column test to test its practical use.

2. Experiments and methods 2.1. Materials and chemicals All chemicals including the (pristine) AE resin used in this study were purchased from Sigma-Aldrich (St. Louis, MO, USA). Zirconium molybdate (ZM) NPs, ammonium heptamolybdate tetrahydrate ((NH4)6Mo7O24$4H2O) and zirconium (IV) oxychloride octahydrate (ZrOCl2$8H2O) were used as received. The anion exchange (AE) resin used in this study was a macroporous strong base anion exchange resin with a polystyrene-divinylbenzene matrix (a Dowex™ Marathon™ MSA, 640 ± 50 mm). Prior to use, the resin was washed with a hydrochloric acid solution replacing the anionic form with chloride ions and then rinsed with ethanol for 24 h and dried at room temperature.

2.3. Analytical methods The phosphate concentration was analyzed with the molybdenum blue method using a UVeVis spectrometer (8453 E UV/Vis, Agilent, USA) (Murphy and Riley, 1962; Tsang et al., 2007). Low concentrations of phosphate, below 0.05 mg/L, were analyzed with an ICP analyzer (ICP-MS, Varian 820-MS, Varian, Australia). The contents of Zr and Mo loaded in the hybrid adsorbents (ZMAE and ZAE) were determined with ICP-AES after digesting with mixed acids (HCl, H2SO4, and HClO4). The morphology and surface characteristics of the ZMAE were analyzed with a scanning electron microscope (SEM, JSM-6700F, Jeol, Japan), high resolution transmission electron microscope (HR-TEM, JEM 3010, Jeol, Japan), high resolution X-ray diffractometer (XRD, D8 Discover, Bruker, Germany), Fourier Transform Infrared Spectroscopy (FT-IR 200, Jasco, Japan), and X-ray photoelectron spectroscopy (XPS, Sigma Probe, ThermoVG, U.K.). The specific surface area and pore volume of the ZMAE adsorbent were measured with a BET analyzer (ASAP, 2000; Micromeritics, USA). The pHpzc (point of zero charge) of the ZMAE was determined by the drift method (Fan et al., 2011). 2.4. Batch adsorption experiments Phosphate adsorption isotherm experiments with the ZMAE (0.5 g/L) were carried out in the batch mode with a shaking glass bottle (100 mL) containing 50 mL of a phosphate solution (5e50 mg/L of P-PO4) at an initial pH of 5.5 (without adjusting during adsorption) for 24 h and compared with the ZAE and the pristine AE resin. All batch mode experiments were conducted in a shaker at a constant temperature (25
C). These experiments were conducted in the absence and presence of excess sulfate ions (5 mM). In addition, a kinetic study with ZMAE (0.5 g/L) in the absence and presence of high sulfate (5 mM) was performed in phosphate solution (10 mg/L of P-PO4 (0.31 mM), 200 mL) and compared with the pristine AE resin. At the predetermined time, 0.5e2 mL of this solution were withdrawn to measure the phosphate concentration. Regeneration of the ZMAE was successively examined with adsorption-desorption experiments of phosphate ion. For adsorption step, the experiments with the ZMAE (50 mg/50 mL) were carried out in the batch mode containing 10 mg/L P-PO4 with shaking (200 rpm) at pH of 5.5 for 24 h. The desorption step was performed in the condition of 0.1 M NaOH (5 mL) for 2 h at 200 rpm shaking. The adsorption and desorption efficiencies were calculated as follows:

2.2. Synthesis of the ZMAE adsorbent In this study, the ZMAE was fabricated as follows: To fabricate the ZMAE, 10 g of the AE resin was added to a 100 mL of 0.05 M (NH4)6Mo7O24 solution with shaking for 8 h. This treated AE resin was immersed in a 100 mL of 0.2 M ZrOCl2 solution (containing 5% sodium chloride) with shaking for 24 h to precipitate the ZM NPs inside the AE resin. This AE resin containing the ZM NPs was washed with deionized water and ethanol. Finally, the ZMAE was obtained after drying at 50e55
C. The ZMAE was compared with the zirconium oxide embedded AE resin (ZAE) for the purpose of examining the potential role of Mo component in the ZMAE for the phosphate adsorption. The ZAE only containing zirconium oxide instead of zirconium molybdate was synthesized, following the method reported in the previous studies (Padungthon et al., 2015; Pan et al., 2013), which was similar with the synthetic procedure for the ZMAE. The only difference was that the hydrated zirconium oxide was precipitated with a 5% NaOH solution instead of the (NH4)6Mo7O24 solution.

23

PO4 adsorbedð%Þ ¼

½PO4 removed  100 ½PO4 initial

PO4 desorbedð%Þ ¼

½PO4 desorbed  100 ½PO4 removed

2.5. Determination of the selectivity coefficient of ZMAE The selectivity of the ZMAE (0.5 g/L) for phosphate ions over sulfate ions was measured in mixed solutions of equal concentrations of phosphate and sulfate ions for three different conditions (0.25, 0.75, and 2.0 mM) at pH 5 for 24 h (equilibrium condition) and compared with that of the pristine AE resin. The selectivity coefficient (SPO4/SO4) for either the ZMAE or the pristine AE resin was defined as the ratio of the distribution coefficients of the two respective ions between solution and

24

T.H. Bui et al. / Water Research 134 (2018) 22e31

adsorbent (Liu et al., 2013; Petrova et al., 2015):

SPO4=SO4 ¼

DPO4 DSO4

where DPO4 or DSO4 (L of solution/g of adsorbent) is the distribution coefficient of phosphate and sulfate ions, respectively. The distribution coefficient (D) was defined as follows:

DPO4 ¼

qePO4 q or DSO4 ¼ eSO4 CePO4 CeSO4

where qe (mmol/g) and Ce (mmol/L) are the adsorption capacity and the concentration of the phosphate or sulfate in the equilibrated solution, respectively.

AE resin within the ZMAE was estimated by the ion exchange capacity (IEC, meq/g) with a conventional titration method using chloride (Karas et al., 2014). If the ion exchange behavior of the R4Nþ of the AE resin inside the ZMAE is identical with that of the pristine AE resin, the fraction of the number of R4Nþ functional groups of the AE resin within the ZMAE over that of the pristine AE resin (f) (the mass fraction of the AE within the ZMAE) was obtained by Equation (1). This assumption is supported by that the chloride adsorption on the ZM NPs was negligible (refer to Fig. S2 in SI):

f ¼

qR4Nþðin ZMAEÞ qmAE ðin ZMAEÞ ¼ qR4Nþðin AEÞ qmAE

qmZMAE ¼ qmNPsðin ZMAEÞ þ qmAEðin ZMAEÞ 2.6. Quantitative determination of the selective phosphate property of the ZMAE To evaluate the contribution of the ZM NPs to the selectivity of the ZMAE, the phosphate capacity of two different sites within the ZMAE ((i) the quaternary ammonium groups (R4Nþ) of the AE resin and (ii) the ZM NPs) were obtained separately in the following manner with the assumption that the ZMAE has only these two different types of sites for phosphate adsorption (refer to the TEM image in Fig. 1(a). First, the number of R4Nþ groups of the pristine AE resin and the

(1)

(2)

where

C F is the fraction of the number of R4Nþ functional groups in the AE resin within the ZMAE over that of the pristine AE resin; C q R4Nþ (in ZMAE) is the number of R4Nþ groups of the AE resin within the ZMAE per weight of the ZMAE (meq/g); C q R4Nþ (in AE) is the number of R4Nþ groups per weight of the pristine AE resin (meq/g); C qm-AE (in ZMAE) is the maximum phosphate capacity of the AE   of P resin inside the ZMAE per weight of the ZMAE g mg ; of ZMAE

Fig. 1. Characteristics of the nanoscale zirconium molybdate embedded within a commercial anion exchange resin (ZMAE): (a) HR-TEM image of the ZMAE showing the zirconium molybdate nanoparticles (the ZM NPs, darker spots) dispersed in the anion exchange resin (AE) support and (b) FTIR pattern of the ZMAE compared with the pristine AE resin.

T.H. Bui et al. / Water Research 134 (2018) 22e31

C qm-AE is the maximum phosphate capacity of the pristine AE   of P resin mg ; g of AE C qm-ZMAE is the maximum phosphate capacity of the ZMAE   mg of P ; g of ZMAE C qm-NPs (in ZMAE) is the maximum phosphate capacity of the ZM   of P NPs inside the ZMAE per weight of the ZMAE g mg of Z MAE

25

In addition, a fixed-bed column experiment was conducted with a glass column (12  50 mm in diameter and length) packed with the ZMAE (5 mL) and feed water at a constant temperature (25
C) using a water bath. The composition of the feed water was as follows: phosphate ([P-PO4] ¼ 2.0 mg/L) and other co-existing anions 2 ([Cl] ¼ 110 mg/L, [HCO 3 ] ¼ 150 mg/L, [SO4 ] ¼ 120 mg/L at pH 4.3e4.4) with a constant flow rate of 1.25 mL/min (empty bed contact time (EBCT) ¼ 4 min). 3. Results and discussion

Second, the maximum phosphate capacity of the ZM NPs within the ZMAE (qmNPsðin ZMAEÞ ) was obtained by subtracting the maximum phosphate capacity of the AE resin within the ZMAE ( qmAEðin ZMAEÞ ) from the maximum phosphate capacity of the ZMAE (qmZMAE ) shown in Equation (2). Both the qmZMAE and qmAEðin ZMAEÞ values were obtained from the phosphate isotherm experiments using the ZMAE and AE resin with the estimation of Langmuir isotherm model. On the other hand, in the presence of excess sulfate, the maximum phosphate capacity of the AE in the ZMAE can be estimated with Equation (3):

q*mAEðin ZMAEÞ ¼ q*mZMAE  q*mNPsðin ZMAEÞ

(3)

where the additional “*” in (q*mZMAE , q*mNPsðin ZMAEÞ and q*mAEðin ZMAEÞ ) denotes the condition with the presence of excess sulfate, respectively. The value of q*m-ZMAE (mg$P/g$ZMAE) was obtained in the same manner as done with qm-NPs (in ZMAE). The q*mNPs (in ZMAE) value was assumed to be identical with qm-NPs (in ZMAE) because the experimental results of ZM NPs for the phosphate adsorption was not affected at all by the presence of excess sulfate ions (refer to Fig. S3 in SI). 2.7. pH effect on the phosphate adsorption and stability of the ZMAE To examine the pH effect on the phosphate adsorption of the ZMAE, a batch adsorption experiment was conducted in a pH range from 2 to 11 under the same condition. Additionally, to evaluate the stability of the ZMAE with respect to the pH, the amounts of zirconium and molybdenum species from the ZMAE were measured. A solution containing 1 g/L of the ZMAE was stirred for 48 h with a varying pH range from 1 to 13. Then, the filtered supernatant solution was used to measure the zirconium and molybdenum concentration by ICP-AES. The solution pH was maintained by adding HCl or NaOH solution during the stability experiments. 2.8. Application of the ZMAE with simulated Mekong River water and wastewater effluent To examine the feasibility of the ZMAE for practical applications, batch adsorption and fixed-bed column experiments were conducted with simulated Mekong River water and typical wastewater effluent with the excess presence of phosphate and other coexisting ions, respectively. The pH condition was adjusted to an acidic condition around pH 4.4 to minimize the release of the molybdate species from the ZMAE. The composition of the simulated Mekong River water was as 2 follows: [Cl] ¼ 50 mg/L, [HCO 3 ] ¼ 150 mg/L, and [SO4 ] ¼ 70 mg/L as competing ions, and [P-PO4] ¼ 10 mg/L at pH 4.5 (Huang et al., 2009). The simulated wastewater effluent with an excess presence of phosphate was as follows: [Cl] ¼ 150 mg/L, 2 [HCO 3 ] ¼ 150 mg/L, and [SO4 ] ¼ 200 mg/L with 10 mg/L of P-PO4 at pH 4.5 (Blaney et al., 2007).

3.1. Characterizations of the ZMAE Fig. 1 and Table 1 show the physicochemical characterization of the ZMAE fabricated in this study. As shown in Fig. 1(a) and Table 1, the ZM NPs were successfully loaded into the structure of the AE resin with low agglomeration, and their size was about 8.9 ± 1.7 nm (N ¼ 50 particles). The ZM NPs with a poor crystalline in nature and the higher BET surface area of the ZMAE compared to the pristine AE resin (Table 1) are presumed to be favorable for phosphate adsorption (refer to Fig. S1 (a) in SI) (Cumbal and SenGupta, 2005; Pan et al., 2009c). The FTIR spectra of the ZMAE in contrast with the pristine AE resin show the specific peaks characterizing the zirconium molybdate embedded inside the ZMAE (Fig. 1 (b)). For example, the peaks at the region of 800e950 cm1 indicate the presence of the molybdate group for the loaded NPs (Inamuddin and Ismail, 2010; Nabi et al., 2010). Additionally, the FTIR spectrum of the ZMAE was distinguishable from the pristine AE resin by the peaks in the range of 450e550 cm1, indicating the stretching vibration of the metaleoxygen (ZreO and MoeO) (Nabi et al., 2010; Uhlrich et al., 2011). The inherent characteristic of the quaternary ammonium functional in the AE support resin was observed in the vibration of the merging stretching hydrogen bonded hydroxyl groups and hydroxyl groups of adsorbed water (~3500e3300 cm1), CeH peak (3000 cm1), and C¼C peak (1650 cm1) (Kociolek-Balawejder et al., 2016; Shah and Chudasama, 2014). These results indicate that the ZMAE possesses not only the specific properties of the loaded ZM NPs but also the anion exchange properties of the R4Nþ functional groups in the AE resin. Furthermore, the presence of zirconium and molybdenum in the ZMAE was confirmed by ICPAES with weight percent of 4.3 and 14.7 (Table 1) and by XPS spectra with their oxidation states of Zrþ4 and Moþ6 (refer to Fig. S1 (b) in SI), respectively. As shown in Fig. 1 (b), the structure of the polymer matrix and functional group in the ZMAE appeared to be very similar with that of the pristine AE resin. However, the surface area of the ZMAE containing the ZM NPs within the AE resin was increased from 22.2 to 28.3 m2/g, which is explained by the poor crystalline nature of the ZM NPs (Table 1). 3.2. Selective phosphate adsorption of the ZMAE 3.2.1. Effective phosphate adsorption of the ZMAE Fig. 2 shows the phosphate adsorption performance of the ZMAE compared with the pristine AE resin in terms of the kinetic behavior (a) and adsorption isotherm (b) (5 mM of sulfate), and the effect of a competing ion concentration (c). As shown in Fig. 2 (a), in the absence of competing ions (sulfate ions), the difference in the phosphate uptake between the ZMAE and the pristine AE resin is not much (64% and 62% removal for the ZMAE and the AE for 90 min, respectively). One difference is that the phosphate uptake of the ZMAE steadily occurred even up to 9 h, whereas the pristine AE resin quickly reached equilibrium.

26

T.H. Bui et al. / Water Research 134 (2018) 22e31 Table 1 Summary of the physicochemical characteristics of the ZMAE (nanoscale zirconium molybdate embedded within a commercial anion exchange resin) compared with those of the anion exchange resin (pristine AE). Adsorbent

ZMAE

Pristine AE

Matrix structure Functional group BET surface area (m2/g) Average pore diameter (nm) Size of nanoparticle (nm) Metals content (% mass)

Poly (styrene-divinylbenzene) R-Nþ(CH3)3 28.3 30.2 8.9 ± 1.7 Zr: 4.3 Mo: 14.7 1: 3.2

Poly (styrene-divinylbenzene) R-Nþ(CH3)3 22.2 32.5

Mol ratio of Zr: Mo

e 0

Fig. 2. Phosphate adsorption of the nanoscale zirconium molybdate embedded within a commercial anion exchange resin (ZMAE) compared with the pristine anion exchange resin
(the pristine AE): (a) Phosphate uptake with a contact time up to 9 h ([P-PO4] ¼ 10 mg/L, [ZMAE] or [pristine AE] ¼ 0.5 g/L, [SO2 4 ] ¼ 480 mg/L (5 mM), initial pH ¼ 5.5 at 25 C); (b)
Equilibrium adsorption ([ZMAE] or [pristine AE] ¼ 0.5 g/L, [SO2 4 ] ¼ 480 mg/L (5 mM), initial pH ¼ 5.5 at 25 C, 24 h); and (c) Effect of sulfate concentration ([P-PO4] ¼ 6.2 mg/L (0.2 mM), initial pH ¼ 5.5 at 25
C, 24 h). Please note that the pH value was not adjusted during the adsorption process.

On the other hand, as shown in Fig. 2(a), in the presence of excess sulfate ions (5 mM), the phosphate uptake of the pristine AE resin became negligible even after 9 h (only 3% removal), possibly due to the interfering effect from the presence of excess competing ions (5 mM of sulfate) (Yang et al., 2015), whereas the ZMAE exhibited still a much higher phosphate uptake capability (48% removal at 9 h, showing that the ZMAE had not only a higher phosphate removal efficiency in the absence of sulfate ions but also a selective property even in the presence of sulfate ions compared with the AE resin. Fig. 2 (b) shows the phosphate adsorption isotherm of the ZMAE compared with the pristine AE resin. For the whole range of

phosphate concentrations in Fig. 2 (b), the phosphate adsorption of the ZMAE was higher than that of the pristine AE resin regardless of the presence of excess sulfate, as expected from the result of Fig. 2 (a). However, these differences between the ZMAE and AE became more drastic in the presence of excess sulfate because phosphate adsorption by the AE resin becomes greatly hindered in the presence of sulfate. For example, in case of the ZMAE, the maximum adsorption capacity (qm) value (obtained by Langmuir isotherm model) was decreased by 38%, (from 42.2 to 26.1 mg-P/g) in the presence of excess sulfate, whereas the qm of the pristine AE resin was significantly diminished by 96% (from 43.1 to 1.8 mg-P/g). Fig. 2 (c) shows the phosphate adsorption with respect to

T.H. Bui et al. / Water Research 134 (2018) 22e31

increasing sulfate concentration compared with the AE resin. As shown in Fig. 2 (c), phosphate removal by the ZMAE was maintained at over about 55% of the removal even at an extremely high concentration of sulfate, although it decreased slowly with an increasing sulfate concentration, whereas the phosphate removal of the pristine AE resin quickly dropped to almost zero. 3.2.2. Selectivity coefficient of the ZMAE for phosphate ions relative to sulfate ions Table 2 shows the distribution coefficients of phosphate and sulfate (DPO4 and DSO4) and the selectivity coefficients of the ZMAE for phosphate removal over sulfate (SPO4/SO4) compared with the pristine AE resin, which were examined in a mixed solution of identical concentrations of phosphate and sulfate (0.25 mMe2.0 mM) at pH 5 for 24 h (equilibrium condition). As shown in Table 2, the SPO4/SO4 value of the ZMAE was 7.0 in the mixed solution of phosphate and sulfate (0.25 mM each), indicating that the DPO4 was 7 times higher compared to the DSO4. Under the same condition, the SPO4/SO4 value of the pristine AE resin was 0.086, indicating that the pristine AE resin adsorbed sulfate much more. Therefore, this result of selectivity compared between the ZMAE and pristine AE resin supports the great performance of the ZMAE for phosphate adsorption in the presence of excess sulfate ions shown in Fig. 2. 3.2.3. Chemical characterization for selective phosphate adsorption of the ZMAE Fig. 3 shows the results of the XPS analysis in terms of the O1s (a) and P2p (b) spectra with respect to phosphate adsorption. As shown in Fig. 3 (a), the clear change of the O1s spectra of the ZMAE indicates a specific interaction between the loaded ZM NPs and phosphate ions because of phosphate adsorption. For example, the change in quantity of the two overlapped peaks of the O1s XPS spectra at 530.5 and 531.9 eV corresponding to oxide oxygen (eOe) and hydroxyl group (eOH) after the phosphate adsorption reveals a new bonding formation between the surface ZM NPs and phosphate species (Su et al., 2013). The initial content of the eOH peak in the O1s total peak dropped after reacting with phosphate corresponding to the relative increase of the eOe peak content. This change in the O1s spectra means that the hydroxyl surface group of the ZM NPs in the hybrid is replaced by the phosphate group during the adsorption process. On the other hand, the P2p XPS spectra of the ZMAE in Fig. 3 (b) (appeared in two main peaks at 133.8 and 132.8 eV corresponding to the P5þ state) show the specific interaction between the phosphate and the surface of the ZM NPs (Bai et al., 2016; Yao et al., 2013), as a result of phosphate adsorption onto the ZMAE. Fig. 4 shows the FTIR spectra of the ZMAE with respect to phosphate adsorption. As shown in Fig. 4, the adsorption of phosphate produces a slightly negative shift in the stretching vibration of the molybdate component from the region of 800e946 cm1 to

27

750e930 cm1, possibly resulting from a specific interaction between the phosphate ion and molybdate component of the ZMAE forming a complex of phosphomolybdate (Berchmans et al., 2011; Zezza et al., 2012). In addition, a new broad and intense peak was observed at 1052 cm1, indicating the asymmetry vibration of the PeO bond of the adsorbed phosphate after the phosphate adsorption (Su et al., 2013). The appearance of this peak confirms the surface adsorption of the phosphate onto the loaded NPs (Su et al., 2013; Yao et al., 2013), as demonstrated in the XPS analysis (Fig. 3). 3.3. Interpretation of the selective phosphate selectivity of the ZMAE Fig. 5 shows the phosphate adsorption distribution of the ZMAE (qm-AE (in ZMAE), qm-NPs (in ZMAE), q*m-NPs (in ZMAE), and q*m-AE (in ZMAE)) compared with that of the pristine AE resin with or without the excess sulfate (qm-AE and q*m-AE)). Note that the values of qm-AE, q*m-AE, qm-ZMAE, and q*m-ZMAE were previously mentioned in Fig. 2. As previously described in the experimental section, the contributions of the ZM NPs in the ZMAE for phosphate capacity (qmNPs (in ZMAE)) were obtained from Equations (1) and (2). These contributions of the ZM NPs in the ZMAE for phosphate capacity were calculated based on the observation of the negligible effect of the excess sulfate on the phosphate adsorption capacity of the ZM NPs (refer to Fig. S3 in SI). The f value (Equation (1), the fraction of the number of R4Nþ functional groups in the AE resin within the ZMAE over the pristine AE resin) was obtained as 0.56 (q R4Nþ (in AE) ¼ 3.42 meq/g, and q R4Nþ (in ZMAE) ¼ 1.92 meq/g), indicating that the AE resin within the ZMAE occupied around 56 wt% of the ZMAE, which has a phosphate capacity of 24.1 mg P/g (¼ qm-AE (in ZMAE), Fig. 5). It was noted that the phosphate capacity (24.1 mg/g or 0.78 meq/g) calculated from experimental data is less than the qR4Nþ value (1.92 meq/g). One of the explanations for this discrepancy can be from the multiprotic property of phosphoric acid (pKa1 ¼ 2.1, pKa2 ¼ 7.2, pKa3 ¼ 12.3 (Chitrakar et al., 2006)) inside the AE resin. Phosphate may be present partly as HPO2 4 and therefore occupy two R4Nþ sites in the resin. On the other hand, it can be explained by that H2PO 4 ion may be not effectively exchanged with chloride ion of the AE resin due to the similar affinity of R4Nþ inside the AE  resin toward both H2PO 4 and Cl (Awual and Jyo, 2011). Then, qmbecame 18.1 mg P/g which occupied 43% of the qm-ZMAE. NPs (in ZMAE) This is explained in Case I shown in Fig. 5. The contribution of qmNPs (in ZMAE) was increased to 69% in the presence of excess sulfate because the maximum phosphate capacity of the AE within the ZMAE was decreased from 24.1 to 8.0 mg P/g. In contrast with Case I, the second approach was made (Case II) to characterize the phosphate adsorption distribution of the ZMAE in the presence of excess sulfate. In this approach, the maximum phosphate capacity of the pristine AE resin (q*m-AE: (mg$P/g)) was maintained in the maximum phosphate capacity of the AE resin

Table 2 Selectivity coefficients (S) and distribution (D) of the ZMAE (nanoscale zirconium molybdate embedded within a commercial anion exchange resin) for phosphate adsorption over sulfate at equilibrium compared with the pristine AE (pristine anion exchange resin). Adsorbents

C0 (mM)

Ce-SO4 (mM)

Ce-PO4 (mM)

SPO4/SO4

DPO4 (L/g)

DSO4 (L/g)

qe-SO4 (mmol/g)

qe-PO4 (mmol/g)

ZMAE

0.25 0.75 2.0 0.25 0.75 2.0

0.131 0.563 1.73 0 0.071 1.19

0.039 0.35 1.25 0.081 0.69 1.96

7.0 4.7 4.6 0.086 0.012 0.038

10.9 2.27 0.91 4.18 0.17 0.045

1.56 0.47 0.19 48.4 14.1 1.17

0.219 0.290 0.358 0.479 1.314 1.474

0.425 0.796 1.248 0.431 0.120 0.088

Pristine AE

C0: Initial concentration of the phosphate and sulfate ions (mmol/L) PO4 SPO4/SO4.¼ D DSO4 Adsorbent dose 25 mg/50 mL.

28

T.H. Bui et al. / Water Research 134 (2018) 22e31

Fig. 3. O1s (a) and P2p (b) XPS spectra of the ZMAE (nanoscale zirconium molybdate embedded within a commercial anion exchange resin) in terms of phosphate adsorption.

indicate that the major contribution of the ZM NPs in the ZMAE is for the selective phosphate removal. In addition, the advantages of zirconium molybdate (in the ZMAE) over zirconium oxide (in the ZAE) for phosphate adsorption are well demonstrated in Fig. S4 (SI), showing that the ZMAE exhibited the better selectivity (C, the maximum adsorption capacity of 26.1 mg/g estimated by Langmuir model) than the ZAE (-, the maximum adsorption capacity of 8.4 mg/g estimated by Langmuir model) for almost all the concentration range of phosphate in the presence of excess sulfate ion, although the adsorption capacities of the ZMAE (B, 42.2 mg/g, estimated by Langmuir model) and the ZAE (,, 46.7 mg/g, estimated by Langmuir model) appeared to be similar in the absence of excess sulfate ion. This is well explained by the specific reaction of molybdate with phosphate to form phosphomolybdate complexes (Hori, 1977; Zhao et al., 1996), in which the complex of 12-phosphomolybdic acid has been specifically applied for trace analysis of phosphate for many decades (Molybdenum Blue method), (Bobtelsky and Barzily, 1963; Ma et al., 2014). Note that the Zr contents of the ZMAE and the ZAE was similar as 4.3 and 4.9 wt%, respectively for this comparison. 3.4. The pH effect on the phosphate adsorption capacity and chemical stability of the ZMAE Fig. 4. FTIR spectra of the ZMAE (nanoscale zirconium molybdate embedded within a commercial anion exchange resin) in terms of phosphate adsorption.

within the ZMAE (qm-AE (in ZMAE)), assuming the consistency of the adsorption behavior of R4Nþ for both the pristine AE resin and AE resin within the ZMAE. Then, q*m-NPs (in ZMAE) was obtained as 25.1 mg P/g from a q*m-ZMAE of 26.1 mg P/g because the AE resin inside the ZMAE occupied around 56 wt% of the ZMAE, and q*m-AE in ZMAE became 1.0 mg P/g, showing the higher contributions of the ZM NPs in the ZMAE for phosphate capacity compared with the Case I approach. Consequently, both approaches (Case I & II)

Fig. 6 shows the pH effect on the phosphate adsorption capacity of the ZMAE (a) and on the chemical stability of the ZMAE focusing on the release of the molybdenum and zirconium species (b). As shown in Fig. 6 (a), the phosphate adsorption capacity appears to be quite dependent upon the pH conditions. The phosphate capacity was the highest in the pH range of 3e5 possibly due to the favorable complexation of the molybdate and phosphate at acidic pH and the electrostatic attraction between the R4Nþ functional groups in the AE resin inside the ZMAE and H2PO 4 (pKa2 of 7.20 (Chitrakar et al., 2006),) (Hori, 1977; Zhao et al., 1996). Meanwhile, the phosphate adsorption capacity decreased as the pH increased over 5 even though the amount of negatively charged phosphate was increased.

T.H. Bui et al. / Water Research 134 (2018) 22e31

29

Fig. 5. The phosphate adsorption distributions of the ZMAE (nanoscale zirconium molybdate embedded within a commercial anion exchange resin) compared with the pristine AE (pristine anion exchange resin), depending upon the excess presence of sulfate (5 mM). The phosphate adsorption capacity of the ZM NPs (the zirconium molybdate nanoparticles) and that of the quaternary ammonium functional group (R4Nþ) of the polymer matrix are displayed as checked and gray bars, respectively; qm-AE (in ZMAE): the maximum phosphate capacity of the AE resin within the ZMAE (mg$P/g); qm-NPs (in ZMAE): the maximum phosphate capacity of the ZM NPs within the ZMAE (mg$P/g); qm-ZMAE: the maximum phosphate capacity of the ZMAE; qm-AE: the maximum phosphate capacity of the pristine AE resin (mg$P/g), and the superscript * mark: the presence of excess sulfate. In Case I, the maximum phosphate capacity of the ZM NPs in the ZMAE is maintained even in the presence of excess sulfate. In Case II, the maximum phosphate capacity of the AE resin in the presence of excess sulfate is maintained by the AE within the ZMAE.

Fig. 6. (a) Phosphate adsorption of the ZMAE (nanoscale zirconium molybdate embedded within a commercial anion exchange resin) with respect to the pH and (b) release of wt% Mo and Zr from the ZMAE with respect to the pH (+: wt% of Zr release, and C: wt% of Mo release); the experimental condition for the adsorption experiments was [P-PO4] ¼ 10 mg/ L, [ZMAE] ¼ 0.5 g/L, and 4.5 h; the experimental condition for the stability test was [ZM NPs] ¼ 1.0 g/L and 48 h (pH was maintained at the initial value during the adsorption experiments).

On the other hand, as shown in Fig. 6 (b), molybdenum species started to be released beyond pH 6, which is a disadvantage for practical applications at neutral or basic conditions, whereas no zirconium species were released from the ZMAE even at basic conditions. Further improvement is required to prevent the release of the molybdenum species at neutral or basic conditions, although the ion exchange polymer matrix inside the ZMAE contributed to delaying the release of the molybdenum species inside the ZMAE

compared with the case of the bare ZM NPs (refer to Fig. S5 in SI). In addition, it is necessary to maintain the pH of water at not higher than pH 6 in order to maintain the stability of the ZMAE.

3.5. The regeneration of the ZMAE Fig. 7 shows the results about the regeneration of the ZMAE with the five successive cycles of phosphate adsorption-desorption. As

30

T.H. Bui et al. / Water Research 134 (2018) 22e31

Fig. 7. Regeneration efficiency of the ZMAE adsorbent for phosphate removal for successive five cycles of adsorption-desorption. Adsorption step: [P-PO4] ¼ 10 mg/L, [ZMAE] ¼ 50 mg/50 mL, pH 5.5, 24 h; Desorption step: regenerant solution: 0.1 M NaOH (5 mL), 2 h.

shown in Fig. 7, the regeneration efficiency (or re-adsorption efficiency, black bar) maintained more than 70% up to the 5th cycle of adsorption-desorption experiments. The lowering regeneration efficiency with respect to subsequent cycle is explained by the Mo leaching during the desorption process at alkaline condition as previously mentioned in Fig. 6(b). On the other hand, the phosphate-adsorbed adsorbents was effectively regenerated by 0.1 M NaOH solution (92.0 ± 3.0% of phosphate desorption efficiency). Although, more efforts were made to find out the regenerant solution other than alkaline solution (0.1 M NaOH) in order to overcome the problem of Mo leaching during the desorption process at alkaline condition, it was noted that perchlorate/perchloric acid, sulfate/sulfuric acid were not effective at all as regenerant solutions. 3.6. Application of the ZMAE for phosphate removal in simulated water The batch adsorption experiments and column experiments with the ZMAE in a slightly acidic pH were carried out to examine the feasibility of the ZMAE for use in potential applications and compared with the pristine AE resin (Fig. 8). As shown in Fig. 8(a), the phosphate adsorption capacity of the ZMAE was relatively sustained compared with the pristine AE resin when using the simulated Mekong river water and a representative wastewater. For example, the qe values of the ZMAE were about 15 and 13 mg P/g for the simulated Mekong river water and wastewater effluent, respectively, even in the presence of excess sulfate, whereas, the that of the pristine AE resin appeared to be very low for the simulated Mekong river water and wastewater effluent, showing the capability of the potential applications of the ZMAE for selective phosphate removal in common water/wastewater within the presence of multi-anions other than phosphate ions. Fig. 8(b) shows the result of the column study of the ZMAE with a simulated water compared with that of the pristine AE resin. As shown in Fig. 8(b) ZMAE had an extremely effective removal of phosphate compared with that of the pristine AE resin. For instance, ZMAE achieved about 4400 BVs (considering the allowance of the phosphate concentration of discharging water (0.5 mg,P-PO4/L)), whereas the pristine AE resin achieved only about 70 BV, demonstrating the great potential of the ZMAE application for selective phosphate removal from various waters.

Fig. 8. Phosphate uptake of the ZMAE (nanoscale zirconium molybdate embedded within a commercial anion exchange resin) compared with the pristine AE (anion exchange resin) at 25
C for both: (a) Batch experiment with simulated Mekong (MK) river water in the presence of excess phosphate ([Cl] ¼ 50 mg/L, [HCO 3 ] ¼ 150 mg/L,   and [SO2 4 ] ¼ 70 mg/L) and effluent wastewater ([Cl ] ¼ 150 mg/L, [HCO3 ] ¼ 150 mg/L and [SO2 ] ¼ 200 mg/L) at pH 4.5 ([P-PO ] ¼ 10 mg/L, [ZMAE] or [pristine AE] ¼ 0.5 g/L, 4 4 24 h); (b) Fixed bed column experiment with simulated wastewater effluent (EBCT (empty bed contact time) 4 min), [P-PO4] ¼ 2.0 mg/L, [SO2 4 ] ¼ 120 mg/L, [Cl] ¼ 110 mg/L, and [HCO 3 ] ¼ 150 mg/L at pH 4.3e4.4).

4. Conclusion It was successfully fabricated a hybrid ZMAE, which is nanoscale zirconium molybdate embedded within a commercial macroporous anion exchange resin. The ZMAE showed excellent selectivity toward phosphate, which much superior to that of the hybrid ZAE or pristine AE resin, supported by specific interaction of phosphate and the molybdate component of the loaded ZM NPs. There was about 62% of the phosphate capacity of the ZMAE served the selective adsorption in the presence of excess sulfate ion. The great selectivity of phosphate of the ZMAE was given by the attribution of the ZM NPs (69e96% of the selective capacity), which not only exhibited itself excellent selectivity toward phosphate ion but also contributed to enhance the selective phosphate adsorption of the supported AE resin. Application of the ZMAE for removal of phosphate from synthetic water showed a great potential of treatment with over 4400 BVs of effective treatment capacity (from

T.H. Bui et al. / Water Research 134 (2018) 22e31

2 mg/L P-PO4 to below 0.5 mg/L P-PO4) compared with the pristine AE resin (less than 70 BVs). These results promise a high potential of field application of the ZMAE for effective removal of phosphate from wastewater. Acknowledgement The authors would like to thank the anonymous reviewers for their comments and suggestions to improve the quality of this paper. This subject is supported by Korea Ministry of Environment as “Global Top Project (E617-00211-0608-0). Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.watres.2018.01.061. References Awual, M.R., Jyo, A., 2011. Assessing of phosphorus removal by polymeric anion exchangers. Desalination 281, 111e117. Bai, Z., Zhou, C., Gao, N., Pang, H., Ma, H., 2016. A chitosanePt nanoparticles/carbon nanotubesdoped phosphomolybdate nanocomposite as a platform for the sensitive detection of nitrite in tap water. RSC Adv. 6, 937e946. Berchmans, S., Karthikeyan, R., Gupta, S., Poinern, G.E.J., Issa, T.B., Singh, P., 2011. Glassy carbon electrode modified with hybrid films containing inorganic molybdate anions trapped in organic matrices of chitosan and ionic liquid for the amperometric sensing of phosphate at neutral pH. Sensor. Actuator. B Chem. 160 (1), 1224e1231. Blaney, L.M., Cinar, S., SenGupta, A.K., 2007. Hybrid anion exchanger for trace phosphate removal from water and wastewater. Water Res. 41 (7), 1603e1613. Bobtelsky, M., Barzily, I., 1963. A rapid and precise heterometric method for the determination of traces of phosphoric or arsenic acid with nitron and molybdate. Anal. Chim. Acta 28, 118e126. Chitrakar, R., Tezuka, S., Sonoda, A., Sakane, K., Ooi, K., Hirotsu, T., 2006. Selective adsorption of phosphate from seawater and wastewater by amorphous zirconium hydroxide. J. Colloid Interface Sci. 297 (2), 426e433. Cumbal, L., SenGupta, A.K., 2005. Arsenic removal using polymer-supported hydrated iron(III) oxide Nanoparticles: role of Donnan membrane effecty. Environ. Sci. Technol. 39 (17), 6508e6515. Du, Q., Zhang, S., Pan, B., Lv, L., Zhang, W., Zhang, Q., 2013. Bifunctional resin-ZVI composites for effective removal of arsenite through simultaneous adsorption and oxidation. Water Res. 47 (16), 6064e6074. Fan, H.T., Sun, T., Xu, H.B., Yang, Y.J., Tang, Q., Sun, Y., 2011. Removal of arsenic(V) from aqueous solutions using 3-2-(2-aminoethylamino)ethylamino propyltrimethoxysilane functionalized silica gel adsorbent. Desalination 278 (1e3), 238e243. Guaya, D., Valderrama, C., Farran, A., Armijos, C., Luis Cortina, J., 2015. Simultaneous phosphate and ammonium removal from aqueous solution by a hydrated aluminum oxide modified natural zeolite. Chem. Eng. J. 271, 204e213. Hori, T., 1977. Formation of colourless molybdate complexes of phosphorus compounds in aqueous solution. J. Inorg. Nucl. Chem. 39 (12), 2173e2177. Huang, X., Sillanpaa, M., Gjessing, E.T., Vogt, R.D., 2009. Water quality in the Tibetan Plateau: major ions and trace elements in the headwaters of four major Asian rivers. Sci. Total Environ. 407 (24), 6242e6254. Inamuddin, Ismail, Y.A., 2010. Synthesis and characterization of electrically conducting poly-o-methoxyaniline Zr(1V) molybdate Cd(II) selective composite cation-exchanger. Desalination 250 (2), 523e529. Jiang, H., Chen, P., Luo, S., Tu, X., Cao, Q., Shu, M., 2013. Synthesis of novel nanocomposite Fe3O4/ZrO2/chitosan and its application for removal of nitrate and phosphate. Appl. Surf. Sci. 284 (Suppl. C), 942e949. Karas, F., Hn at, J., Paidar, M., Schauer, J., Bouzek, K., 2014. Determination of the ionexchange capacity of anion-selective membranes. Int. J. Hydrogen Energy 39 (10), 5054e5062. Kociolek-Balawejder, E., Stanislawska, E., Jacukowicz-Sobala, I., 2016. Synthesis and characterization of CuO-loaded macroreticular anion exchange hybrid polymer. React. Funct. Polym. 100, 107e115. Li, G., Gao, S., Zhang, G., Zhang, X., 2014. Enhanced adsorption of phosphate from aqueous solution by nanostructured iron(III)ecopper(II) binary oxides. Chem. Eng. J. 235, 124e131. Liu, B., Lv, X., Meng, X., Yu, G., Wang, D., 2013. Removal of Pb(II) from aqueous solution using dithiocarbamate modified chitosan beads with Pb(II) as imprinted ions. Chem. Eng. J. 220, 412e419. Liu, X., Zhang, L., 2015. Removal of phosphate anions using the modified chitosan beads: adsorption kinetic, isotherm and mechanism studies. Powder Technol.

31

277, 112e119. Ma, J., Sengupta, M.K., Yuan, D., Dasgupta, P.K., 2014. Speciation and detection of arsenic in aqueous samples: a review of recent progress in non-atomic spectrometric methods. Anal. Chim. Acta 831, 1e23. Murphy, J., Riley, J.P., 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta 27, 31e36. Nabi, S.A., Bushra, R., Naushad, M., Khan, A.M., 2010. Synthesis, characterization and analytical applications of a new composite cation exchange material poly-otoluidine stannic molybdate for the separation of toxic metal ions. Chem. Eng. J. 165 (2), 529e536. Padungthon, S., German, M., Wiriyathamcharoen, S., SenGupta, A.K., 2015. Polymeric anion exchanger supported hydrated Zr(IV) oxide nanoparticles: a reusable hybrid sorbent for selective trace arsenic removal. React. Funct. Polym. 93 (Suppl. C), 84e94. Pan, B., Li, Z., Zhang, Y., Xu, J., Chen, L., Dong, H., Zhang, W., 2014. Acid and organic resistant nano-hydrated zirconium oxide (HZO)/polystyrene hybrid adsorbent for arsenic removal from water. Chem. Eng. J. 248, 290e296. Pan, B., Pan, B., Zhang, W., Lv, L., Zhang, Q., Zheng, S., 2009a. Development of polymeric and polymer-based hybrid adsorbents for pollutants removal from waters. Chem. Eng. J. 151 (1e3), 19e29. Pan, B., Wu, J., Pan, B., Lv, L., Zhang, W., Xiao, L., Wang, X., Tao, X., Zheng, S., 2009b. Development of polymer-based nanosized hydrated ferric oxides (HFOs) for enhanced phosphate removal from waste effluents. Water Res. 43 (17), 4421e4429. Pan, B., Wu, J., Pan, B., Lv, L., Zhang, W., Xiao, L., Wang, X., Tao, X., Zheng, S., 2009c. Development of polymer-based nanosized hydrated ferric oxides (HFOs) for enhanced phosphate removal from waste effluents. Water Res. 43 (17), 4421e4429. Pan, B., Xu, J., Wu, B., Li, Z., Liu, X., 2013. Enhanced removal of fluoride by polystyrene anion exchanger supported hydrous zirconium oxide nanoparticles. Environ. Sci. Technol. 47 (16), 9347e9354. Petrova, Y.S., Pestov, A.V., Usoltseva, M.K., Neudachina, L.K., 2015. Selective adsorption of silver(I) ions over copper(II) ions on a sulfoethyl derivative of chitosan. J. Hazard Mater. 299, 696e701. Shah, B., Chudasama, U., 2014. Synthesis and characterization of a novel hybrid material as amphoteric ion exchanger for simultaneous removal of cations and anions. J. Hazard Mater. 276, 138e148. Song, L., Huo, J., Wang, X., Yang, F., He, J., Li, C., 2016. Phosphate adsorption by a Cu(II)-loaded polyethersulfone-type metal affinity membrane with the presence of coexistent ions. Chem. Eng. J. 284, 182e193. Su, Y., Cui, H., Li, Q., Gao, S., Shang, J.K., 2013. Strong adsorption of phosphate by amorphous zirconium oxide nanoparticles. Water Res. 47 (14), 5018e5026. Tsang, S., Phu, F., Baum, M.M., Poskrebyshev, G.A., 2007. Determination of phosphate/arsenate by a modified molybdenum blue method and reduction of arsenate by S2O2 4 . Talanta 71 (4), 1560e1568. Uhlrich, J.J., Sainio, J., Lei, Y., Edwards, D., Davies, R., Bowker, M., Shaikhutdinov, S., Freund, H.J., 2011. Preparation and characterization of ironemolybdate thin films. Surf. Sci. 605 (15e16), 1550e1555. Xie, J., Lai, L., Lin, L., Wu, D., Zhang, Z., Kong, H., 2015. Phosphate removal from water by a novel zeolite/lanthanum hydroxide hybrid material prepared from coal fly ash. J. Environ. Sci. Health - Part A Toxic/Hazard. Subst. Environ. Eng. 50 (12), 1298e1305. Xu, X., Song, W., Huang, D., Gao, B., Sun, Y., Yue, Q., Fu, K., 2015. Performance of novel biopolymer-based activated carbon and resin on phosphate elimination from stream. Colloid. Surface. Physicochem. Eng. Aspect. 476, 68e75. Yang, W., Yu, Z., Pan, B., Lv, L., Zhang, W., 2015. Simultaneous organic/inorganic removal from water using a new nanocomposite adsorbent: a case study of pnitrophenol and phosphate. Chem. Eng. J. 268, 399e407. Yang, Y., Lohwacharin, J., Takizawa, S., 2014. Hybrid ferrihydrite-MF/UF membrane filtration for the simultaneous removal of dissolved organic matter and phosphate. Water Res. 65, 177e185. Yao, Y., Gao, B., Chen, J., Yang, L., 2013. Engineered biochar reclaiming phosphate from aqueous solutions: mechanisms and potential application as a slowrelease fertilizer. Environ. Sci. Technol. 47 (15), 8700e8708. You, X., Farran, A., Guaya, D., Valderrama, C., Soldatov, V., Cortina, J.L., 2016. Phosphate removal from aqueous solutions using a hybrid fibrous exchanger containing hydrated ferric oxide nanoparticles. J. Environ. Chem. Ecotoxicol. 4 (1), 388e397. Zezza, T.R.C., Castilho, M.d.S., Stradiotto, N.R., 2012. Determination of phosphorus in biodiesel using 1:12 phosphomolybdic modified electrode by cyclic voltammetry. Fuel 95, 15e18. Zhang, L., Gao, Y., Li, M., Liu, J., 2015. Expanded graphite loaded with lanthanum oxide used as a novel adsorbent for phosphate removal from water: performance and mechanism study. Environ. Technol. 36 (8), 1016e1025. Zhao, Y.C., Zouboulis, A.I., Matis, K.A., 1996. Flotation of molybdate oxyanions from dilute solutions Part II. Selective separation from phosphates, arsenates and silicates. Hydrometallurgy 43 (1e3), 155e167. Zheng, X., Pan, J., Zhang, F., Liu, E., Shi, W., Yan, Y., 2016. Fabrication of free-standing bio-template mesoporous hybrid film for high and selective phosphate removal. Chem. Eng. J. 284, 879e887.