Zirconia functionalized SBA-15 as effective adsorbent for phosphate removal

Microporous and Mesoporous Materials 155 (2012) 192–200

Contents lists available at SciVerse ScienceDirect

Microporous and Mesoporous Materials journal homepage: www.elsevier.com/locate/micromeso

Zirconia functionalized SBA-15 as effective adsorbent for phosphate removal Yuqiong Tang, Enmin Zong, Haiqin Wan ⇑, Zhaoyi Xu, Shourong Zheng, Dongqiang Zhu State Key Laboratory of Pollution Control and Resource Reuse, Jiangsu Key Laboratory of Vehicle Emissions Control, School of the Environment, Nanjing University, Nanjing 210093, China

a r t i c l e

i n f o

Article history: Received 20 October 2011 Received in revised form 12 January 2012 Accepted 13 January 2012 Available online 31 January 2012 Keywords: ZrO2 functionalized SBA-15 Phosphate adsorption Adsorption kinetics

a b s t r a c t Phosphate adsorbents with molecular level dispersions of surface functionalities were prepared by covalent grafting of ZrO2 onto SBA-15. The synthetic adsorbents were characterized by X-ray diffraction, N2 adsorption/desorption, transmission electron microscope, UV–Vis diffuse reflectance spectroscopy, X-ray photoelectron spectroscopy and zeta-potential measurements in terms of surface chemistry and pore structure, and their adsorption properties toward phosphate were examined. All adsorption isotherms could be well described by the Freundlich model. Compared with SBA-15, markedly enhanced phosphate adsorption was observed on ZrO2 functionalized SBA-15, which was attributed to a combined effect of high surface exposure of the Zr–OH group and its strong inner-sphere complexing ability for phosphate. Phosphate adsorption was affected by the pH and ionic strength, wherein the adsorption capacity of the adsorbent increased at low pH and high ionic strength. Additionally, the adsorption process obeyed the pseudo-second-order kinetics and the rate constant decreased with initial phosphate concentration. Findings in this study highlight the potential of using functionalized SBA-15 with highly dispersed ZrO2 as an effective adsorbent for phosphate removal. Ó 2012 Elsevier Inc. All rights reserved.

1. Introduction The presence of excessive nutrients (nitrogen and phosphorus) may cause eutrophication of lakes, lagoons, rivers and sea [1]. As one of the nutrients for water eutrophication, phosphate can be released from many non-point and point sources, such as agricultural fertilizers, municipal wastewater, and effluents from chemical and mineral processing industries [2]. Due to more stringent legislation for phosphate discharge, it is thus highly desirable to develop effective treatment methods for phosphate removal from phosphate bearing wastewater prior to its discharge into the aquatic environment. Adsorption method has been recognized as a simple approach to remove phosphate in water. A variety of metal hydroxides were found to be effective in adsorption of phosphate [3–5]. Among the metal hydroxides, zirconium hydroxide is a superior adsorbent due to its strong surface complexing ability for phosphate and high chemical stability under acidic and basic conditions. For example, Chitrakar et al. [6] reported much higher adsorption capacity of phosphate on zirconium hydroxide than that on layered double hydroxides. Liu et al. [7] attributed the adsorption of phosphate on mesoporous ZrO2 to a mechanism of anion exchange with surface Zr–OH groups.

⇑ Corresponding author. Tel.: +86 25 89680369; fax: +86 25 89680596. E-mail address: [email protected] (H. Wan). 1387-1811/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.micromeso.2012.01.020

Alternatively, dispersion of metal hydroxides over inert supports with high surface areas is an effective method to enhance the adsorption capacities of active species. Liao et al. [8] prepared ZrO2-loaded collagen fiber and observed effective phosphate removal over the adsorbents. In parallel, Yuchi et al. [9] studied phosphate adsorption over ZrO2-loaded polymer gel and found that the adsorbent was effective for the removal of phosphate at a sub ng ml1 level. Notably, in most cases the active ZrO2 moieties of supported adsorbents exit in nano-particles; accordingly, only ZrO2 surface is accessible and acts as active adsorption sites. Hence, it is speculated that loading of ZrO2 moieties on support surface via chemical bonding likely provides a molecular level dispersion of ZrO2 functionality, giving rise to high surface exposure of adsorption site and thus high adsorption capacity for phosphate. Since first synthesized in 1992 [10,11], mesoporous SiO2 has attracted widespread interest due to its large surface area, high pore volume and ordered mesostructure. Because mesoporous SiO2 matrix is inert in nature, surface functionalization may further extend its applications in adsorption, separation and catalysis [12– 14]. Moreover, mesoporous SiO2 has abundant surface silanol groups, which are susceptible to surface functionalization [13,14]. Considering the high complexing ability of ZrO2 toward phosphate, it is hypothesized that ZrO2 functionalized mesoporous SiO2 prepared by the covalent grafting method may have high exposure of surface functionality and display superior performance for phosphate adsorption. However, thus far few studies have been conducted.

Y. Tang et al. / Microporous and Mesoporous Materials 155 (2012) 192–200

The objective of this study is to explore the feasibility of ZrO2 functionalized SBA-15 adsorbents for adsorptive removal of phosphate in water. ZrO2 functionalized SBA-15 adsorbents were prepared by the post-grafting method, and were characterized by X-ray diffraction (XRD), infrared spectroscopy (IR), transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), N2 adsorption/desorption, and zeta potential measurements. Phosphate adsorption to the adsorbents was investigated using the batch technique. 2. Experimental 2.1. Material preparation SBA-15 was prepared using Pluronic P123 as the structuredirecting agent and tetraethoxysilane (TEOS) as the silica source [15]. Briefly, 8.0 g of Pluronic P123 (Aldrich) was dissolved in 300 ml of 2.0 M HCl solution at 40 °C and 17.6 g of TEOS (98%, Shanghai Chemical Co.) was then added. The final molar composition of the gel mixture was P123:HCl:H2O:TEOS = 1:60:10073:435. After stirring at 40 °C for 24 h the solution was transferred to a Teflon-lined autoclave which was kept at 100 °C for 24 h. The resulting material was recovered by filtration, washing repeatedly with distilled water. The organic template (Pluronic P123) was removed by calcination at 550 °C for 6 h in air. ZrO2 functionalized SBA-15 was prepared by the post-grafting method [16,17]. Briefly, 2.0 g of SBA-15 was suspended in 40.0 ml of dried toluene, to which 3.1 ml of 70% zirconium isopropoxide (Zr(OC3H7)4) solution in propanol (Aldrich) was added. The suspension was then refluxed at 110 °C for 3 h. After configuration, the resulting material was washed with toluene, ethanol and distilled water, following by vacuum drying at 110 °C for 12 h. Two- or three-cycle functionalization of SBA-15 was conducted by repeating the above described procedure two or three times. The samples prepared by single-, two- and three-cycle functionalization are denoted as SBA-15-Zr1, SBA-15-Zr2 and SBA-15-Zr3, respectively. For comparison, ZrO2 was prepared by the precipitation method [18]. Briefly, 400 ml of 0.5 M ZrOCl28H2O aqueous solution was added dropwise to 150 ml of 2.0 M ammonia solution under vigorous stirring followed by aging at room temperature for 2 h. ZrO2 was obtained by filtration, repeated washing with distilled water, drying at 105 °C for 6 h, and calcination at 400 °C for 4 h. 2.2. Adsorbent characterization XRD patterns were collected in a range of 0.6–5° for small-angle XRD and 10–80° for wide-angle XRD from a Rigaku D/max-RA power diffraction-meter using Cu Ka radiation. TEM images of the samples were recorded with a JEM-200CX electron microscope. N2 adsorption/desorption isotherms were obtained on a Micrometrics ASAP 2020 apparatus at 196 °C (77 K). ZrO2 contents in the samples were determined on an ARL9800XP X-ray fluorescence (XRF) spectrometer. The UV–Vis spectra of the samples were collected on a SHIMAD UV-2401PC UV/Vis spectrometer using BaSO4 as the reference. XPS analysis was performed on a PHI 550 ESCA/ SAM X-ray photoelectron spectroscopy equipped with a monochromatized Al Ka X-ray source (hm = 1486.6 eV). The C 1s peak (284.6 eV) was used for the calibration of binding energy values. To verify the distribution of ZrO2 in the samples, the ion sputtering model was used in XPS measurement with Ar as the ion source, under 1000 eV and middle speed sputtering 60 s. The surface zeta potentials of the samples were measured using a Zeta Potential Analyzer (Zeta PALS, Brookhaven Instruments Co.). The IR spectra of the samples were recorded at 4 cm1 resolution using a Nicolet 380 FTIR spectrometer in a vacuum IR system.

193

The sample was pressed into a self-supporting wafer which was placed in an IR cell connected to a vacuum system. The sample was activated under vacuum (<2  105 mbar) by heating to 300 °C at a rate of 10 °C min1 and held at this temperature for 4 h. After cooling to 50 °C, IR spectra were collected. 2.3. Phosphate adsorption Phosphate adsorption isotherms were determined by batch adsorption experiments. Briefly, 20.0 mg of adsorbent was introduced into 40 ml glass vials with polytetrafluoroethylene-lined screw caps receiving 40 ml of phosphate solution with varied initial concentrations. The initial pH of phosphate solution was adjusted using 0.1 M HCl to assure the final pH close to 6.2. The samples were shaken by an orbital shaker at 25 °C for 72 h. The time period was sufficient to reach apparent adsorption equilibrium (no further uptake) based on preliminarily determined adsorption kinetics (data not shown). After filtration using 0.45 lm filters, the residual phosphate concentration was determined using UV–Vis spectrometry according to the molybdenum blue method with a detecting wavelength of 700 nm [9,19]. The equilibrium adsorption amount of phosphate was calculated according to Eq. (1):

qe ¼

ðC 0  C e ÞV m

ð1Þ

where qe is the equilibrium adsorption amount, C0 is the initial phosphate concentration, Ce is the equilibrium concentration, V is the solution volume and m is the adsorbent mass. The effect of ionic strength on phosphate adsorption to SBA-15Zr2 was investigated in the presence of NaCl solutions with varied concentrations (0–0.1 mol l1) at pH 6.2 and 25 °C. Separate sets of experiments were conducted to test the effects of pH on phosphate adsorption to the sorbents. In the pH experiments, the initial pH of phosphate solution was preadjusted using 0.1 M HCl and 0.1 M NaOH and the pH of all samples was measured after adsorption equilibrium. All sorption experiments were conducted in duplicate. 2.4. Phosphate adsorption kinetics Phosphate adsorption kinetics was carried out at initial concentrations of 5.0, 12.5 and 30.0 mg P l1. Briefly, 0.25 g of SBA-15-Zr2 was added into a flask containing 500 ml of 5.0, 12.5 or 30.0 mg P l1 phosphate solution, which was strongly stirred in an incubator at 25 °C. During the adsorption process, about 2.5 ml of sample was withdrawn at preset time intervals. After fast filtration using 0.45 lm filters, the residual concentration of phosphate in the solution was determined spectrophotometrically. The adsorption amount was calculated as follows:

qt ¼

ðC 0  C t ÞV m

ð2Þ

where qt is the adsorption amount at time t, C0 is the initial phosphate concentration, Ct is the concentration at time t, V is the volume of phosphate solution and m is the adsorbent mass. 3. Results and discussion 3.1. Material characterization The wide- and small-angle XRD patterns of SBA-15, SBA-15-Zr1, SBA-15-Zr2 and SBA-15-Zr3 are presented in Fig. 1. In the wideangle XRD patterns of SBA-15 and SBA-15-Z1, only wide diffraction peaks were observed with 2h around 23°, characteristic of amorphous SiO2 [20]. Increasing ZrO2 amount led to a shift of the wide diffraction peak to higher 2h, reflecting that amorphous ZrO2 moieties are dominant on SBA-15 surface. Moreover, after calcina-

194

Y. Tang et al. / Microporous and Mesoporous Materials 155 (2012) 192–200

(a) Bulk ZrO2 Calcined SBA-15-Zr3 Intensity (a.u.)

Calcined SBA-15-Zr2 Calcined SBA-15-Zr1 SBA-15

SBA-15-Zr1

SBA-15-Zr3 SBA-15-Zr2 SBA-15-Zr1 SBA-15 10

20

30

40

50

60

70

80

2θ (degree) SBA-15-Zr2

(b)

SBA-15-Zr3

Fig. 2. TEM images of the samples.

SBA-15-Zr3

Amount adsorbed (cm3· g-1)

Intensity (a.u.)

(a) 1000 SBA-15-Zr2

SBA-15-Zr1

SBA-15 1

2

3 2θ (degree)

4

SBA-15

800

SBA-15-Zr1 600

SBA-15-Zr2 400

SBA-15-Zr3 200

0

5

0

0.4

0.6

0.8

1

Relative pressure (p/p0)

Fig. 1. (a) Wide angle and (b) small angle XRD patterns of the samples.

dV/dD (cm3·g-1·nm-1)

(b) tion at 500 °C for 4 h the diffraction peaks characteristic of crystalline ZrO2 were not detected in the samples, implying that ZrO2 is highly dispersed on SBA-15 surface. As for small-angle XRD patterns, three distinct diffraction peaks with 2h at 0.91°, 1.56° and 1.78°, respectively, characteristic of (1 0 0), (1 1 0) and (2 0 0) planes, were identified on SBA-15, indicative of the ordered mesoporous structure of SBA-15 with p6 mm hexagonal symmetry [15,21]. Similar to SBA-15, strong (1 0 0), (1 1 0) and (2 0 0) diffraction peaks were observed on SBA-15-Zr1, SBA-15-Zr2 and SBA15-Zr3. However, ZrO2 functionalization led to the decrease of peak intensity, likely due to the decreased structural ordering or/and the contrast matching between the SiO2 framework and ZrO2 moieties grafted to SBA-15 surface [22,23]. The ordered pore structures of SBA-15 and its functionalized counterparts with uniform dimension and hexagonal arrangement could also be clearly visualized by their TEM images (results presented in Fig. 2). N2 adsorption/desorption isotherms and the pore size distributions of the samples are shown in Fig. 3. For all samples, typical capillary condensation was observed within a relative pressure range of 0.4–0.8, reflecting the presence of mesopores. The calculated structural parameters, BET surface areas and pore volumes of the samples are listed in Table 1. The BET surface areas of the samples were 718, 603, 578, 553 m2 g1 for SBA-15, SBA-15-Zr1, SBA-15-Zr2 and SBA-15-Zr3, respectively, reflecting decreased surface area with functionalization cycles. Compared with SBA-15, the pore

0.2

0.30 0.25 0.20 0.15

SBA-15 0.10

SBA-15-Zr1 SBA-15-Zr2

0.05

SBA-15-Zr3 0.00 2

7

12

17

22

Pore diameter (nm) Fig. 3. (a) N2 adsorption/desorption isotherms and (b) pore size distributions of the samples.

volume of SBA-15-Zr3 decreased from 0.92 to 0.39 cm3 g1. Moreover, ZrO2 functionalization led to substantial changes in the pore size distributions. In comparison with SBA-15, single-cycle functionalization caused a slightly widened pore size distribution (see Fig. 3b), whereas two- and three-cycle functionalization resulted in a marked shift of the most probable pore size to smaller pore size, indicative of the continuous deposition of ZrO2 in the pore channel.

195

Y. Tang et al. / Microporous and Mesoporous Materials 155 (2012) 192–200

(a)

c d

ZrO2 contenta (wt.%)

a0b (nm)

SBET (m2g-1)

Vpc (cm3g-1)

Vmid (cm3g-1)

SBA-15 SBA-15-Zr1 SBA-15-Zr2 SBA-15-Zr3

2.1 8.8 19.6

10.0 10.0 10.0 10.0

718 603 578 553

0.92 0.70 0.52 0.39

0.053 0.037 0.030 0.025

ZrO2

Determined by X-ray fluorescence. p Unit cell, a0 = 2d100 3. Total pore volume, determined at P/P0 = 0.97. Micropore volume, calculated using the as-plots method.

Similar results were observed previously [24,25]. Notably, ZrO2 functionalization may lead to the change in the micropority. Hence, the micropore volumes were calculated using the as-plots method [26,27] and the results are listed in Table 1. The micropore volume of SBA-15, SBA-15-Zr1, SBA-15-Zr2 and SBA-15-Zr3 are 0.053, 0.037, 0.030 and 0.025 cm3 g1, respectively, indicative of decreased Vmi with functionalization cycles likely due to micropore clogging by grafted ZrO2. The IR spectra of SBA-15 and ZrO2 functionalized SBA-15 samples are presented in Fig. 4. For SBA-15, a sharp peak at 3742 cm1 and a broad peak around 3750 to 3300 cm1 were observed, which are assigned to the isolated and hydrogen bonded hydroxyl groups, respectively [28]. As for SBA-15-Zr1, the intensity of the IR band at 3742 cm1 decreased and a new IR band characteristic of Zr–OH groups appeared around 3662 cm1 [29,30], reflecting that ZrO2 moieties are successfully grafted on SBA-15 surface via a condensation reaction between Zr(OC3H7)4 and surface silanol groups. Additionally, two- or three-cycle functionalization led to a further increase in the intensity of Zr–OH group at the expense of silanol group, again confirming the gradual replacement of surface silanol groups by Zr–OH groups during the functionalization process. The UV–Vis diffuse reflectance spectra of the samples are compared in Fig. 5a. For ZrO2, the absorbance edge was found to be around 248 nm, characteristic of the ligand-to-metal charge transfer (LMCT) from O2 to Zr4+ with an octahedral coordination state [31]. No absorbance was observed on SBA-15 because SiO2 is transparent in the test UV region [32]. In contrast to SBA-15, functionalized SBA-15 samples displayed distinct absorbance in the UV range and absorbance intensity increased with functionalization cycle. Moreover, a substantial blue shift in absorbance was observed on the functionalized samples as compared with ZrO2. For the samples, the absorbance band edge energy could be further quantified as follows [33]:

Absorbance (a.u.)

SBA-15-Zr3

SBA-15-Zr2

SBA-15-Zr1

SBA-15

4000

3500

3000

2500

2000

1500

Wavenumber(cm -1) Fig. 4. IR spectra of SBA-15, SBA-15-Zr1, SBA-15-Zr2 and SBA-15-Zr3.

SBA-15-Zr3 SBA-15-Zr2 SBA-15-Zr1 SBA-15

200

300

400

500

600

700

800

Wavelength (nm)

(b)

40

30 (ahv)2 (eV2)

a b

Sample

Absorbance (a.u.)

Table 1 Structural properties of the samples.

ZrO2 SBA-15-Zr3

20 SBA-15-Zr2 SBA-15-Zr1

10

SBA-15

0 0

2

4

6

hv (eV) Fig. 5. (a) UV–Vis spectra and (b) plots of (ahv)2 versus hv of the samples.

ahm ¼ Aðhm  Eg Þ1=2

ð3Þ

where a is the absorbance intensity at light frequency of v, A is the absorbance constant and Eg is the absorbance edge energy, respectively. The plots of (ahv)2 versus hv are compiled in Fig. 5b. Eg is the intercept of linear extrapolation to hv axis. The absorbance edge energies for ZrO2, SBA-15-Zr1, SBA-15-Zr2, and SBA-15-Zr3 are calculated to be 4.9, 5.8, 5.5, and 5.4 eV, respectively. Compared with ZrO2, the much higher absorbance edge energies of ZrO2 functionalized SBA-15 samples result from LMCT from O2 to Zr4+ with lower coordination [31], likely due to the formation of Si–O–Zr linkage and surface Zr– OH groups. Consistently, the monotonically decreased absorbance edge energy with functionalization cycle can be ascribed to gradually increased Zr4+ coordination state, resulting from the formation of Zr– O–Zr linkage upon multiple-cycle functionalization. XPS analysis is a surface sensitive technique which can provide useful information about the oxidation states and compositions of surface elements. The XPS spectra of the samples in the Zr 3d and Si 2p regions are presented in Fig. 6. For ZrO2, the binding energy of Zr core line 3d5/2 was found to be 182.4 eV, similar to the reported values in the literatures [34]. In contrast, a much higher binding energy of Zr 3d5/2 was observed on SBA-15-Zr1, implying the presence of different Zr species from bulk ZrO2 upon surface functionalization. In comparison with bulk ZrO2, the higher binding energy of Zr 3d5/2 in SBA-15-Zr1 can be attributed to the formation of Si–O–Zr bond due to the higher electronegativity of Si atom as

196

Y. Tang et al. / Microporous and Mesoporous Materials 155 (2012) 192–200

(a)

Raw data Fitting curve Zr-O-Zr Si–O-Zr

—— —— ——

Zr3d

5/2

Intensity (a.u.)

Zr 3d 3/2

187

186

185

184

183

182

181

180

Binding energy (eV)

(b) —— —— ——

Raw data Fitting curve Si-O-Si Si-O-Zr Si2p

Intensity (a.u.)

SBA-15-Zr1

SBA-15-Zr2

ties on SBA-15 surface. To further characterize the distribution of ZrO2 moieties in the samples, the ion sputtering model was used in XPS measurement and the results are presented in Fig. 1S and Table 1S. Upon ion sputtering, for functionalized samples the binding energy of Zr 3d5/2 as well as Zr concentration is similar to that without ion sputtering, reflecting a homogeneous distribution of ZrO2 moieties in the samples. To further quantify the relative contents of surface species in the functionalized samples, the XPS profiles were deconvoluted and the fitting parameters are listed in Table 2. As shown in the XPS spectra in the Si 2p region (see Fig. 6b), the functionalized samples mainly consist of two Si species – Si1 with a binding energy of 102.4 eV assigned to Si–O–Zr and Si2 with a binding energy of 103.3 eV characteristic of Si–O–Si. The contents of Si1 are calculated to be 34%, 60% and 84% for SBA-15-Zr1, SBA-15-Zr2 and SBA-15-Zr3, respectively. The gradually increased Si1 content confirms the continuous reaction between Zr(OC3H7)4 and surface silanol group with functionalization cycle, as also reflected by the results of IR and UV–Vis spectroscopy. For the XPS spectra in the Zr 3d region (see Fig. 6a), surface Zr moieties are mainly composed of Zr1 from Zr–O–Si and Zr2 from Zr–O–Zr. In comparison with SBA-15-Zr1, the content of Zr1 in SBA-15-Zr3 decreases to 32%, suggesting that during multiple-cycle functionalization process both silanol and Zr–OH groups simultaneously react with Zr(OC3H7)4. The characterization results further indicate that for SBA-15-Zr1 ZrO2 moieties are dispersed on SBA-15 surface in a monolayer form, whereas two- and three cycle functionalization results in the presence of multi-layer ZrO2 moieties. Surface zeta potentials of SBA-15, SBA-15-Zr1, SBA-15-Zr2 and SBA-15-Zr3 as a function of pH are shown in Fig. 7. For all samples, zeta potentials prominently decreased over a pH range from approximately 3.0 to 7.0, but did not change much at higher pH. The isoelectric point (IEP) of SBA-15 was found to be 1.9, identical to that of amorphous SiO2 [38,39]. The IEPs were 5.1 for SBA-15Zr1 and 5.7 for SBA-15-Zr2 and SBA-15-Zr3, indicative of markedly increased IEP upon ZrO2 functionalization. The much higher IEP of functionalized SBA-15 than that of SBA-15 is mainly due to the replacement of silanol groups by Zr–OH groups on SBA-15 surface. Notably, the IEPs of ZrO2 functionalized SBA-15 samples were still lower than that of zirconium hydroxide (IEP = 6.7) [40], implying the coexistence of silanol and Zr–OH groups even after three-cycle functionalization. Consistent results were also observed by IR spectroscopy.

SBA-15-Zr3

3.2. Phosphate adsorption

SBA-15 108

107

106

105

104

103

102

101

100

99

98

Binding energy (eV) Fig. 6. XPS spectra of the samples in the Zr 3d region (a) and (b) Si 2p region.

compared to Zr atom [34,35]. However, two- or three-cycle functionalization led to a reverse shift of the binding energy of Zr 3d5/ 2, likely resulting from the formation of covalent Zr–O–Zr linkages via a reaction between Zr(OC3H7)4 and surface Zr–OH groups. The conclusion is further confirmed by the variation of the binding energy of Si 2p with functionalization cycle. For SBA-15, the binging energy of Si core line 2p was 103.3 eV, similar to previous reports [36,37]. As for SBA-15-Zr1, the binding energy of Si 2p shifted to 103.0 eV due to the formation of surface Si–O–Zr linkage. Moreover, multiple-cycle functionalization led to further decrease of the binging energy of Si 2p, indicative of continuous grafting of ZrO2 moie-

The phosphate adsorption isotherms over the adsorbents at 25 °C and pH 6.2 are compared in Fig. 8. SBA-15 exhibited negligible phosphate adsorption, while enhanced phosphate adsorption was identified on functionalized adsorbents. Moreover, phosphate adsorption increased with functionalization cycle, reflecting the crucial role of surface ZrO2 moieties for phosphate removal. The adsorption data were fitted to the Freundlich adsorption model:

qe ¼ KC ne

ð4Þ

where qe is the equilibrium adsorption amount at equilibrium phosphate concentration Ce, K is Freundlich coefficient characteristic of the adsorption affinity of the adsorbent, and n is the linearity index. The fitting parameters are listed in Table 3. It is shown that phosphate adsorption to the adsorbents can be well described by the Freundlich model with R2 values higher than 0.96. The K values were calculated to be 5.20, 10.60 and 10.95 mg P1n ln g1 for SBA-15Zr1, SBA-15-Zr2 and SBA-15-Zr3, respectively, confirming a positive correlation between phosphate adsorption capacity and ZrO2 con-

197

Y. Tang et al. / Microporous and Mesoporous Materials 155 (2012) 192–200 Table 2 Fitting parameters of XPS spectra of the samples. Samples

Si 2p

Zr 3d5/2

Binding energy(e.V)

SBA-15 ZrO2 SBA-15-Zr1 SBA-15-Zr2 SBA-15-Zr3

Si1/(Si1 + Si2) (%)

Si1(Si–O–Zr)

Si2(Si–O–Si)

  102.4 102.4 102.4

103.30  103.3 103.3 103.3

0  34 60 84

50

(a)

40

SBA-15

30

SBA-15-Zr1

  182.8 182.8 182.8

 182.4  182.4 182.4

 0 100 36 32

16 14

SBA-15-Zr3

10

Zr2(Zr–O–Zr)

18

SBA-15-Zr2

20

Zr1/(Zr1 + Zr2) (%)

Zr1(Zr–O–Si)

20

qe (mg P· g-1)

Zeta potential (mV)

Binding energy(e.V)

0 -10

12 10 8

-20

6

-30

4

-40

2

: SBA-15 : SBA-15-Zr1 : SBA-15-Zr2 : SBA-15-Zr3

0

-50

0 1

2

3

4

5

6

7

8

9

10

5

10

11

pH Fig. 7. Zeta potentials of the samples as a function of solution pH.

15

20

25

30

35

Ce (mg P· l-1)

(b)

450 400

3.3. Influence of pH and ionic strength on phosphate adsorption The influence of pH on phosphate adsorption to SBA-15-Zr2 is shown in Fig. 9a. Over the examined pH range, increasing pH led to monotonically decreased phosphate adsorption. This trend can be well interpreted in terms of the electrostatic interaction. Phosphate is a polyacid with two pKa values (7.2 and 12.3 [41]), and can undergo protonation/depronation with pH. In the test pH range (3–9), HPO2 and H2 PO 4 4 are dominant anionic species [42]. At pH below 5.7 the Zr–OH groups of SBA-15-Zr2 are positively charged via protonization; therefore, anionic phosphate can be

350 qe (mg P· gZrO2-1)

tent. Considering that the adsorbents have varied ZrO2 contents, a comparison of the adsorption data normalized by ZrO2 content can provide a deeper insight into structure dependent adsorption of phosphate to the adsorbents. After normalization, the adsorption capacity of SBA-15-Zr1 for phosphate is 319.7 mg P gZrO21 at an equilibrium concentration of 5.2 mg P l1. Notably, such high adsorption capacity has not been previously reported. Moreover, ZrO2-normalized phosphate adsorption declines in the order of SBA-15-Zr1 > SBA-15-Zr2 > SBA-15-Zr3. Consistently, the Kn values are found to be 253.6, 120.6, and 55.9 mg P1n ln gZrO21 for SBA15-Zr1, SBA-15-Zr2 and SBA-15-Zr3, respectively. As proved by XPS, IR and UV–Vis, covalent grafting of ZrO2 leads to a molecular level dispersion of surface functionality. Particularly, ZrO2 moieties in SBA-15-Zr1 present in a monolayer form, which are fully accessible for phosphate adsorption. On the contrary, multiple-cycle functionalization results in multi-layer ZrO2 moieties, in which the inner layer ZrO2 is inaccessible. Therefore, it is understandable that SBA15-Zr1 exhibits a higher ZrO2 normalized adsorption capacity as compared to SBA-15-Zr2 and SBA-15-Zr3.

: SBA-15-Zr1 : SBA-15-Zr2 : SBA-15-Zr3

300 250 200 150 100 50 0 0

5

10

15

20

Ce (mg P·

25

30

35

l-1)

Fig. 8. (a) Phosphate adsorption isotherms and (b) ZrO2 content normalized phosphate adsorption isotherms of the samples.

effectively adsorbed by positively charged Zr–OH groups via attractive electrostatic interactions. However, increasing pH leads to the formation of negatively charged Zr–OH groups via deprotonation, suppressing phosphate adsorption due to the repulsive electrostatic interaction. Moreover, at alkaline pH OH may compete with phosphate for surface adsorption sites, also accounting for the decreased phosphate adsorption with the pH. The observed pH dependent phosphate adsorption of the samples suggests that the adsorbent can be regenerated under basic conditions. The influence of ionic strength on phosphate adsorption is presented in Fig. 9b. Enhanced phosphate adsorption was observed with ionic strength. Similar results were also observed by Liu et al. and Barrow et al. [7,43]. The enhanced phosphate adsorption

198

Y. Tang et al. / Microporous and Mesoporous Materials 155 (2012) 192–200

Table 3 Fitting parameters of phosphate adsorption isotherms over the adsorbents.

SBA-15Zr1 SBA-15Zr2 SBA-15Zr3 a

Kn a (mg P1n ln gZrO 2 1)

n

R2

5.20 ± 0.11

253.6 ± 5.46

0.140 ± 0.009

0.96

10.60 ± 0.10

120.6 ± 1.15

0.100 ± 0.004

0.98

0.140 ± 0.004

0.99

K (mg P1n ln g1)

16 14 12

10.95 ± 0.13

55.9 ± 0.684

qt (mg P· l-1)

Sample

ZrO2 content normalized Freundlich affinity coefficient.

(a)

10 8 6

28

4

: C0=5.0 mg P· l-1

2

: C0=12.5 mg P· l-1

0

24 20

qe (mg P·g-1)

: C0=30.0 mg P· l-1

0 2000

4000 t (min)

6000

Fig. 10. Time resolved phosphate adsorption to SBA-15-Zr2 at varied initial phosphate concentrations.

16 12

(a)

8

1.5

4

1

0 2

3

4

5

6

7

8

9

10

0.5

(b)

log(qexp-qt)

pH 24 20

0

qe(mg P·g-1)

-0.5 16

: C0=5.0 mg P· l-1

-1

: C0=12.5 mg P· l-1

12

: C0=30.0 mg P· l-1

-1.5

8

0

: 0.0 M NaCl : 0.01M NaCl : 0.1M NaCl

4

5

10

15

4000

6000

8000

6000

8000

t (min)

0 0

2000

20

25

(b)

Ce (mg P·l-1)

500

Fig. 9. Influence of (a) solution pH and (b) ionic strength on phosphate adsorption to SBA-15-Zr2.

400

t/qt

with ionic strength further suggests that phosphate is adsorbed predominantly via inner sphere complexation, wherein the negatively charged complexes formed by phosphate adsorption can be efficiently compensated by cation at high ionic strength [44], giving rise to enhanced phosphate adsorption.

600

300

200

: C0=5.0 mg P· l-1

100

: C0=12.5 mg P· l-1 : C0=30.0 mg P· l-1

3.4. Adsorption kinetics 0

Phosphate adsorption to SBA-15-Zr2 with initial concentrations of 5.0, 12.5 and 30.0 mg P l1 are compiled in Fig. 10. At initial concentration of 5.0 and 12.5 mg P l1 phosphate adsorption reached equilibrium within 1500 and 2500 min, respectively, while phosphate adsorption at an initial concentration of 30.0 mg l1 was

0

2000

4000

t (min) Fig. 11. Fitting of phosphate adsorption to SBA-15-Zr2 using (a) pseudo-first-order kintics and (b) pseudo-second-order kinetics.

199

Y. Tang et al. / Microporous and Mesoporous Materials 155 (2012) 192–200 Table 4 Fitting parameters of phosphate adsorption to SAB-15-Zr2 using pseudo-first-order and pseudo-second-order kinetics models. Initial concentration (mg l1)

Pseudo first-order kinetic model qexp (mg P g

5.0 12.5 30.0

1

7.68 ± 0.14 11.14 ± 0.23 13.77 ± 0.26

)

1

qe (mg P g

Pseudo second-order kinetic model 1

)

k1 (min

4.92 ± 0.33 5.46 ± 0.50 8.07 ± 0.44

)

0.017 ± 0.003 0.009 ± 0.002 0.002 ± 0.0003

2

R

qe (mg P g1)

k2 (g mg1 min1)

R2

0.92 0.86 0.93

8.24 ± 0.051 12.07 ± 0.11 14.71 ± 0.15

0.0033 ± 0.0009 0.0018 ± 0.0005 0.0006 ± 0.0001

0.99 0.99 0.99

much slow, characteristic of a concentration dependent adsorption process. To evaluate the mass transfer process during the phosphate adsorption, the pseudo-first-order and pseudo-second-order kinetic models were applied. Lagergren’s rate equation was usually used for pseudo-first-order kinetics [45]:

ZrO2 functionality. Low pH and high ionic strength favors phosphate adsorption. Phosphate adsorption to the adsorbent follows pseudo-second-order kinetics. In addition, increasing initial concentration results in retarded phosphate adsorption rate due to prolonged diffusion pathway.

logðqe  qt Þ ¼ log qe  k1 t=2:303

Acknowledgements

ð5Þ

The pseudo-second-order kinetics can be expressed as follows [46]:

t=qt ¼ 1=ðk2 q2e Þ þ t=qe

ð6Þ

where qe is the equilibrium adsorption amount, qt is the adsorption amount at time t, k1 and k2 are pseudo-first-order and pseudo-second-order rate constants, respectively. The plots of lg(qe  qt) versus t based on pseudo-first-order kinetics and t/qt versus t based on the pseudo-second-order kinetics are compared in Fig. 11 and the fitting parameters are listed in Table 4. In comparison with the pseudo-first-order kinetic model, the pseudo-second-order kinetic model fitted the experimental data reasonably with R2 higher than 0.99. Additionally, the equilibrium adsorption amounts (qe) calculated from fitting results are nearly identical to the experimental data (see Table 4), reflecting that phosphate adsorption to SBA-15-Zr2 follows the pseudosecond-order kinetics. At initial concentrations of 5.0, 12.5 and 30.0 mg P l1, the k2 values are calculated to be 3.3  103, 1.8  103 and 6.0  104 g mg1 min1, respectively, confirming a retarded adsorption process with the increase of initial phosphate concentration. For SBA-15-Zr2, the active sites for phosphate adsorption are located on the external surface, pore mouth region and pore channel of SBA-15. At a lower initial concentration, fewer adsorption sites are required and phosphate is likely adsorbed on the external surface or/and pore mouth region of the adsorbent. At a higher initial concentration, however, the adsorption sites on the external surface and pore mouth region are fully occupied and phosphate has to diffuse a longer distance to reach the adsorption sites located in the pore channel of SBA-15, which results in a relatively slow adsorption process. 4. Conclusions In this study, ZrO2 functionalized SBA-15 samples were prepared by the post-grafting method and the adsorption of phosphate over the adsorbents were investigated. Characterization results reveal that ZrO2 moieties are convalently bonded to SBA-15 surface with a molecular level dispersion. Multiple-cycle functionalization leads to increased ZrO2 content and to the gradual replacement of silanol groups of SBA-15 by Zr–OH groups. Moreover, mono-layer ZrO2 dispersion is formed after single-cycle functionalization, while multi-layer ZrO2 dispersion presents upon two- and three-cycle functionalization. ZrO2 functionalized SBA-15 adsorbents exhibit enhanced phosphate adsorption as compared with SBA-15 and phosphate adsorption increases with functionalization cycle. However, ZrO2 content normalized phosphate adsorption decreases with functionalization cycle due to the formation of multi-layer

The financial supports from Key Laboratory of Yangtze River Water Environment (No. YRWEF201005), the Natural Science Foundation of China (No. 20877039 and 21077050) and the Natural Science Foundation of Jiangsu province of China (No. BK2010378) are gratefully acknowledged. We are indebted to the Modern Analytical Center, Nanjing University for the material characterization. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.micromeso.2012.01.020. References [1] V.H. Smith, G.D. Tilman, J.C. Nekola, Environ. Pollut. 100 (1999) 179. [2] W. Viessman Jr., M.J. Hammer, Water Supply and Pollution Control, sixth ed., Addison-Wesley Longman Inc., Menlo Park, CA, 1998. [3] J. Das, B.S. Patra, N. Baliarsingh, K.M. Parida, Appl. Clay Sci. 32 (2006) 252. [4] L.M. Blaney, S. Cinar, A.K. SenGupta, Water Res. 41 (2007) 1603. [5] N. Boujelben, J. Bouzid, Z. Elouear, M. Feki, F. Jamoussi, A. Montiel, J. Hazard. Mater. 151 (2008) 103. [6] R. Chitrakar, S. Tezuka, A. Sonoda, K. Sakane, K. Ooi, T. Hirotsu, J. Colloid Interf. Sci. 297 (2006) 426. [7] H.L. Liu, X.F. Sun, C.Q. Yin, C. Hu, J. Hazard. Mater. 151 (2008) 616. [8] X.P. Liao, Y. Ding, B. Wang, B. Shi, Ind. Eng. Chem. Res. 45 (2006) 3896. [9] A. Yuchi, A. Ogiso, S. Muranaka, T. Niwa, Anal. Chim. Acta 494 (2003) 81. [10] J.S. Beck, J.C. Vartuli, W.J. Roth, M.E. Leonowicz, C.T. Kresge, K.D. Schmitt, C.T.W. Chu, D.H. Olson, E.W. Sheppard, J. Am. Chem. Soc. 114 (1992) 10834. [11] C.T. Kresge, M.E. Leonowicz, W.J. Roth, J.C. Vartuli, J.S. Beck, Nature 359 (1992) 710. [12] U. Ciesla, F. Schuth, Micropor. Mesopor. Mater. 27 (1999) 131. [13] F. Hoffmann, M. Cornelius, J. Morell, M. Froba, Angew. Chem. Int. Ed. 45 (2006) 3216. [14] N. Impens, P. van der Voort, E.F. Vansant, Micropor. Mesopor. Mater. 28 (1999) 217. [15] D.Y. Zhao, J.L. Feng, Q.S. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky, Science 279 (1998) 548. [16] J.N. Zhang, Z. Ma, J. Jiao, H.F. Yin, W.F. Yan, E.W. Hagaman, J.H. Yu, S. Dai, Micropor. Mesopor. Mater. 129 (2009) 200. [17] Q. Tao, Z.Y. Xu, J.H. Wang, F.L. Liu, H.Q. Wan, S.R. Zheng, Micropor. Mesopor. Mater. 131 (2010) 177. [18] Z.Y. Xu, L.Q. Chen, Y. Shao, D.Q. Yin, S.R. Zheng, Ind. Eng. Chem. Res. 48 (2009) 8356. [19] Z.Q. He, C.W. Honeycutt, Commun. Soil Sci. Plan. 36 (2005) 1373. [20] S. Chytil, W.R. Glomm, E.A. Blekkan, Catal. Today 147 (2009) 217. [21] D.Y. Zhao, Q.S. Huo, J.L. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120 (1998) 6024. [22] X.G. Wang, K.S.K. Lin, J.C.C. Chan, S.F. Cheng, J. Phys. Chem. B 109 (2005) 1763. [23] L. Mercier, T.J. Pinnavaia, Chem. Mater. 12 (2000) 188. [24] J. Sauer, F. Marlow, B. Spliethoff, F. Schüth, Chem. Mater. 14 (2002) 217. [25] L.H. Hu, S.F. Ji, T.C. Xiao, C.X. Guo, P.Y. Wu, P.Y. Nie, J. Phys. Chem. 111 (2007) 3599. [26] A. Sayari, P. Liu, M. Kruk, M. Jaroniec, Chem. Mater. 9 (1997) 2499. [27] M. Jaroniec, M. kruk, J.P. Olivier, Langmuir 15 (1999) 5410. [28] C. Setzer, G. van Essche, N. Pryor, in: F. Schuth, K.S.W. Sing, J. Weitkamp (Eds.), Handbook of Porous Solids, Weily-VCH, Weinheim, 2002. [29] A. Kytokivi, E.L. Lakomaa, A. Root, H. Osterholm, J.P. Jacobs, H.H. Brongersmall, Langmuir 13 (1997) 2717.

200

Y. Tang et al. / Microporous and Mesoporous Materials 155 (2012) 192–200

[30] T. Merle-Mejean, P. Barberis, S.B. Othmane, F. Nardou, P.E. Quintard, J. Eur. Ceram. Soc. 18 (1998) 1579. [31] B.L. Newalkar, J. Olanrewaju, S. Komarneni, J. Phys. Chem. B 105 (2001) 8356. [32] T. Klimova, O. Gutierrez, L. Lizama, J. Amezcua, Micropor. Mesopor. Mater. 133 (2010) 91. [33] W.M. Wendl, G.H. Harry, Reflectance Spectroscopy, Interscience Publishers, New York, 1996. [34] W. Xia, Y.M. Wang, V. Hagen, A. Heel, G. Kasper, U. Patil, A. Devi, M. Muhler, Chem. Vapor Depos. 13 (2007) 37. [35] S. Damyanova, P. Grange, B. Delmon, J. Catal. 168 (1997) 421. [36] H.J. Song, H.J. Shin, Y. Chung, J.C. Lee, M.K. Lee, J. Appl. Phys. 97 (2005) 113711. [37] M. Ouyang, C. Yuan, R.J. Muisener, A. Boulares, J.T. Koberstein, Chem. Mater. 12 (2000) 1591.

[38] R.K. Iler, The Chemistry of Silica, Wiley, New York, 1979. [39] J.M. Rosenholm, T. Czuryszkiewicz, F. Kleitz, J.B. Rosenholm, M. Linden, Langmuir 23 (2007) 4315. [40] S. Mattson, A.J. Pugh, Soil Sci. 38 (1934) 299. [41] K.H. Yeon, H. Park, S.H. Lee, Y.M. Park, S.H. Lee, M. Iwamoto, Korean J. Chem. Eng. 25 (2008) 1040. [42] D.G. Grubb, M.S. Guimaraes, R. Valencia, J. Hazard. Mater. 76 (2000) 217. [43] N.J. Barrow, J.W. Bowden, A.M. Posner, J.P. Quirk, Aust. J. Soil. Res. 18 (1980) 395. [44] M.B. Mcbride, Clays Clay Miner. 45 (1997) 598. [45] H.C. Trivedi, V.M. Patel, R.D. Patel, Eur. Polym. J. 9 (1973) 525. [46] Y.S. Ho, G. Mckay, Process Biochem. 34 (1999) 451.