Silica-gel Particles Loaded with an Ionic Liquid for Separation of Zr(IV) Prior to Its Determination by ICP-OES

A new ionic liquid loaded silica gel amine (SG-APTMS-N,N-EPANTf2) was developed, as an adsorptive material, for selective adsorption and determination of zirconium, Zr(IV), without the need for a chelating intermediate. Based on a selectivity study, the SG-APTMS-N,N-EPANTf2 phase showed a perfect selectivity towards Zr(IV) at pH 4 as compared to other metallic ions, including gold [Au(III)], copper [Cu(II)], cobalt [Co(II)], chromium [Cr(III)], lead [Pb(II)], selenium [Se(IV)] and mercury [Hg(II)] ions. The influence of pH, Zr(IV) concentration, contact time and interfering ions on SG-APTMS-N,N-EPANTf2 uptake for Zr(IV) was evaluated. The presence of incorporated donor atoms in newly synthesized SG-APTMS-N,N-EPANTf2 phase played a significant role in enhancing its uptake capacity of Zr(IV) by 78.64% in contrast to silica gel (activated). The equilibrium and kinetic information of Zr(IV) adsorption onto SG-APTMS-N,N-EPANTf2 were best expressed by Langmuir and pseudo second-order kinetic models, respectively. General co-existing cations did not interfere with the extraction and detection of Zr(IV). Finally, the analytical efficiency of the newly developed method was also confirmed by implementing it for the determination of Zr(IV) in several water samples.


Introduction
Room temperature ionic liquids (RTILs) have been considerably used and there is growing interest in them in different chemistry fields. Organic, inorganic, analytical and electrochemistry are common areas for application of ionic liquids (ILs) [1,2]. In addition, ILs have received significant attention in analytical chemistry as environmentally-friendly solvents, replacing volatile, toxic hazardous and flammable organic compounds [3]. Thus, they have been used in solvent extractions [4], solid phase extraction [5][6][7], gas chromatography [8], liquid chromatography [9], capillary electrophoresis [10], mass spectrometry [11], electrochemistry [12], sensors [13] and spectroscopy [14]. Several research studies have been consecrated to examining and studying the spectroscopic and physico-chemical properties of ILs and designing of new families of chiral and achiral ILs [15][16][17][18][19]. Furthermore, ILs have many fascinating properties resulted in a wide range of promising applications in which they have been employed in various fields [20,21].

Preparation of N,N-EPANTf 2 Ionic Iiquid
The synthesis of N,N-EPANTf 2 ionic liquid was previously reported by Marwani [15,43]. Briefly, an aqueous solution of N,N-EPACl (1.00 g) was mixed with a solution of LiNTf 2 (1.43 g) in doubly distilled deionized water. The mixture was stirred for 2 h at room temperature, and the reaction produced two layers. Finally, separation of the two layers was carefully achieved, and the lower layer was dried under vacuum overnight.

Activation of SG
SG powder (25.0 g) was suspended in 50% (v/v) HCl solution (200 mL), and the mixture was refluxed under stirring for 8 h, to avoid the existence of nitrogenous impurities or metal oxides as well as enhancing silanol group content on the SG surface. The activated SG powder was filtered and washed several times with doubly distilled deionized water until neutralization. Finally, it was dried in an oven at 120˝C for 5 h for removing any adsorbed water on the SG surface.

Synthesis of SG-APTMS
For preparation of SG chemically bound amine, activated SG (8 g) was suspended in toluene (100 mL) in a round bottom flask, and then 3-aminopropyltrimethoxysilane (APTMS, 6 mL) was gradually added to this suspension as a silane coupling agent. The reaction mixture was allowed to reflux with continuous stirring for 12 h at~100˝C. The final product was filtered and washed with toluene, ethanol and diethyl ether, then, it was dried in an oven at 80˝C for 8 h to obtain SG chemically bound amine (SG-APTMS).
2.2.4. Synthesis of SG-APTMS-N,N-EPANTf 2 SG-APTMS-N,N-EPANTf 2 was prepared by functionalizing the hydrophobic ionic liquid (N,N-EPANTf 2 ) on the surface of SG-APTMS. An amount of N,N-EPANTf 2 (1.5 g) was completely dissolved by warming in ethanol (50 mL) and added to a suspension of SG-APTMS (5 g) in ethanol (60 mL). The reaction mixture was then stirred for 12 h at 60˝C. The modified phase (SG-APTMS-N,N-EPANTf 2 ) was filtered and washed with ethanol, doubly distilled deionized water and diethyl ether. The SG-APTMS-N,N-EPANTf 2 phase was then dried in an oven at 60˝C for 8 h and kept in a desiccator. Scheme 1 illustrates the synthetic route used to prepare the SG-APTMS-N,N-EPANTf 2 adsorbent.

Batch Procedure
In this investigation, the effect of various parameters including pH, initial concentration of Zr(IV) and contact time on the adsorption of SG-APTMS-N,N-EPANTf 2 towards Zr(IV) were studied For the effect of pH on the SG-APTMS-N,N-EPANTf 2 selectivity towards different metal ions, 2 mg¨L´1 standard solutions were prepared with pH values ranging from 1.0 to 9.0 with a series of appropriate buffer solutions, (0.2 mol¨L´1 HCl/KCl) for pH 1.0 and 2.0, (0.1 mol¨L´1 CH 3 COOH/CH 3 COONa) for pH 3.0-6.0 and (0.1 mol¨L´1 Na 2 HPO 4 /HCl) for pH 7.0-9.0. Then, an amount of SG-APTMS-N,N-EPANTf 2 (20 mg) was individually mixed with each standard solution and mechanically shaken for 1 h by use of a mechanical shaker at 150 rpm and 25˝C.
The supernatant concentrations of residual metal ions were determined by ICP-OES after filtration. For estimating the uptake capacity of SG-APTMS-N,N-EPANTf 2 towards Zr(IV),  (0, 5, 10, 20, 30, 40, 60, 80, 100, 125, 150 and 200 mg¨L´1) of Zr(IV) were prepared at the optimum pH value of 4.0 and individually mixed with SG-APTMS-N,N-EPANTf 2 (20 mg). The mixtures were mechanically shaken for 1 h at 25˝C. In addition, the effect of contact time on the SG-APTMS-N,N-EPANTf 2 adsorption capacity for Zr(IV) was investigated as above at equilibrium periods of 2.5, 5, 10, 20, 30, 40, 50, 60, 75 and 90 min. and washed several times with doubly distilled deionized water until neutralization. Finally, it was dried in an oven at 120 °C for 5 h for removing any adsorbed water on the SG surface.

Synthesis of SG-APTMS
For preparation of SG chemically bound amine, activated SG (8 g) was suspended in toluene (100 mL) in a round bottom flask, and then 3-aminopropyltrimethoxysilane (APTMS, 6 mL) was gradually added to this suspension as a silane coupling agent. The reaction mixture was allowed to reflux with continuous stirring for 12 h at ~100 °C. The final product was filtered and washed with toluene, ethanol and diethyl ether, then, it was dried in an oven at 80 °C for 8 h to obtain SG chemically bound amine (SG-APTMS).

Instrumentation
A Perkin Elmer Spectrum 100 series FT-IR spectrometer (Beaconsfield, Bucks, UK), SEM on a field emission-scanning electron microscope (FE-SEM, QUANT FEG 450, Amsterdam, The Netherlands) and a Barnstead Thermolyne 47900 benchtop muffle furnace (Waltham, MA, USA) were used to characterize the newly synthesized phase. A Jenway model 3505 laboratory pH meter (CamLab, Cambridge, UK) was used for the pH measurements. In addition, a Perkin Elmer ICP-OES model Optima 4100 DV (Norwalk, CT, USA) was used to perform measurements of metal ions concentration at the selected wavelengths of 267.50, 327.39, 238.89, 343.82, 267.72, 220. 35,196.026 and 194.17 nm for Au(III), Cu(II), Co(II), Zr(IV), Cr(III), Pb(II), Se(IV) and Hg(II), respectively. All instrumental parameters used in this study were similar to those applied in our previous study [44]. variation of sample masses. The concentration of N,N-EPANTf 2 on SG-APTMS surface was found to be 0.46 mmol¨g´1, based on this thermal desorption method.

FT-IR Analysis
The structure of activated SG and SG-APTMS-N,N-EPANTf 2 was studied and evaluated by the use of FT-IR spectroscopy. The FT-IR spectrum of newly modified adsorbent exhibited characteristic peaks, related to both APTMS and the N,N-EPANTf 2 ionic liquid. From Figure 1, it can be observed that several bands appeared at positions of about 735 cm´1 for (C-H), 1497 cm´1 for (C-S), 1577 cm´1 for (C=C) bonds. In addition, a characteristic stretching vibration band for (NH) bonds appeared at 2950 cm´1, confirming the presence of the anchored propyl group on SG-APTMS-N,N-EPANTf 2 adsorbent [43].

FT-IR Analysis
The structure of activated SG and SG-APTMS-N,N-EPANTf2 was studied and evaluated by the use of FT-IR spectroscopy. The FT-IR spectrum of newly modified adsorbent exhibited characteristic peaks, related to both APTMS and the N,N-EPANTf2 ionic liquid. From Figure 1, it can be observed that several bands appeared at positions of about 735 cm −1 for (C-H), 1497 cm −1 for (C-S), 1577 cm −1 for (C=C) bonds. In addition, a characteristic stretching vibration band for (NH) bonds appeared at 2950 cm −1 , confirming the presence of the anchored propyl group on SG-APTMS-N,N-EPANTf2 adsorbent [43].

SEM Analysis
In order to obtain more information about the changes in the surface characteristics of activated SG and SG-APTMS-N,N-EPANTf2, SEM images were obtained as presented in Figure 2a,b. As displayed in Figure 2b, N,N-EPANTf2 particles that covered the surface of SG chemically bound amine (SG-APTMS) were clearly observed. In addition, SG-APTMS particles were completely coated by N,N-EPANTf2. This can be regarded as evidence that the hydrophobic ionic liquid N,N-EPANTf2 was successfully loaded on SG-APTMS surface.

SEM Analysis
In order to obtain more information about the changes in the surface characteristics of activated SG and SG-APTMS-N,N-EPANTf 2 , SEM images were obtained as presented in Figure 2a,b. As displayed in Figure 2b, N,N-EPANTf 2 particles that covered the surface of SG chemically bound amine (SG-APTMS) were clearly observed. In addition, SG-APTMS particles were completely coated by N,N-EPANTf 2 . This can be regarded as evidence that the hydrophobic ionic liquid N,N-EPANTf 2 was successfully loaded on SG-APTMS surface. The structure of activated SG and SG-APTMS-N,N-EPANTf2 was studied and evaluated by the use of FT-IR spectroscopy. The FT-IR spectrum of newly modified adsorbent exhibited characteristic peaks, related to both APTMS and the N,N-EPANTf2 ionic liquid. From Figure 1, it can be observed that several bands appeared at positions of about 735 cm −1 for (C-H), 1497 cm −1 for (C-S), 1577 cm −1 for (C=C) bonds. In addition, a characteristic stretching vibration band for (NH) bonds appeared at 2950 cm −1 , confirming the presence of the anchored propyl group on SG-APTMS-N,N-EPANTf2 adsorbent [43].

SEM Analysis
In order to obtain more information about the changes in the surface characteristics of activated SG and SG-APTMS-N,N-EPANTf2, SEM images were obtained as presented in Figure 2a,b. As displayed in Figure 2b, N,N-EPANTf2 particles that covered the surface of SG chemically bound amine (SG-APTMS) were clearly observed. In addition, SG-APTMS particles were completely coated by N,N-EPANTf2. This can be regarded as evidence that the hydrophobic ionic liquid N,N-EPANTf2 was successfully loaded on SG-APTMS surface.

Effect of pH and Selectivity Study
Metal adsorption processes can be influenced by either protonation of binding sites of the adsorbent in an acidic solution or precipitation of many metal ions by hydroxide ions in a basic solution. Therefore, the effect of the pH of aqueous solutions was the first parameter to be evaluated. For investigation of the pH effect on the selectivity of SG-APTMS-N,N-EPANTf 2 towards various metal ions, including Au(III), Cu(II), Co(II), Zr(IV), Cr(III), Pb(II), Se(IV) and Hg(II), 25 mL of 2 mg¨L´1 of each selected metal ion sample solution was individually mixed with 20.0 mg of SG-APTMS-N,N-EPANTf 2 at different pH values (1.0-9.0). As can be noticed from Figure 3, there was no remarkable change in the % extraction of Co(II) and Hg(II). In addition, an increase followed by a subsequent decrease is observed in the % extraction of Au(III), Cu(II), Cr(III), Pb(II) and Se(IV) with an increase of the pH. However, it was interesting to note that the % extraction of Zr(IV) dramatically increased with an increase of the pH value. In sense, the Zr(IV) was found to give the highest quantitative % extraction within the pH range 4.0-9.0. In addition, it can be clearly observed that SG-APTMS-N,N-EPANTf 2 is most selective towards Zr(IV) at pH 4.0. Therefore, pH 4.

Effect of pH and Selectivity Study
Metal adsorption processes can be influenced by either protonation of binding sites of the adsorbent in an acidic solution or precipitation of many metal ions by hydroxide ions in a basic solution. Therefore, the effect of the pH of aqueous solutions was the first parameter to be evaluated. For investigation of the pH effect on the selectivity of SG-APTMS-N,N-EPANTf2 towards various metal ions, including Au(III), Cu(II), Co(II), Zr(IV), Cr(III), Pb(II), Se(IV) and Hg(II), 25 mL of 2 mg·L −1 of each selected metal ion sample solution was individually mixed with 20.0 mg of SG-APTMS-N,N-EPANTf2 at different pH values (1.0-9.0). As can be noticed from Figure 3, there was no remarkable change in the % extraction of Co(II) and Hg(II). In addition, an increase followed by a subsequent decrease is observed in the % extraction of Au(III), Cu(II), Cr(III), Pb(II) and Se(IV) with an increase of the pH. However, it was interesting to note that the % extraction of Zr(IV) dramatically increased with an increase of the pH value. In sense, the Zr(IV) was found to give the highest quantitative % extraction within the pH range 4.0-9.0. In addition, it can be clearly observed that SG-APTMS-N,N-EPANTf2 is most selective towards Zr(IV) at pH 4.0. Therefore, pH 4.  In addition, results obtained from calculated distribution coefficient (Kd) for each metal ion, as illustrated in Table 1, strongly supported that SG-APTMS-N,N-EPANTf2 phase was the highest in selectivity for Zr(IV) over other metal ions included in this study. The distribution coefficient (Kd) can be expressed as follows [45]: where Ci and Ce represent the initial and final concentrations (mg·L −1 ), respectively, V refers to the volume of solution (mL), and m indicates the mass of adsorbent (g). From Table 1, it can be clearly noted that the highest Kd value was acquired for Zr(IV), 289.45  10 3 mL·g −1 , as a representative In addition, results obtained from calculated distribution coefficient (K d ) for each metal ion, as illustrated in Table 1, strongly supported that SG-APTMS-N,N-EPANTf 2 phase was the highest in selectivity for Zr(IV) over other metal ions included in this study. The distribution coefficient (K d ) can be expressed as follows [45]: where C i and C e represent the initial and final concentrations (mg¨L´1), respectively, V refers to the volume of solution (mL), and m indicates the mass of adsorbent (g). From Table 1, it can be clearly noted   Figure 4 displays the uptake capacity of SG-APTMS-N,N-EPANTf 2 towards Zr(IV) obtained from the adsorption isotherm experiment. As can be deduced from Figure 4, the adsorption capacity of SG-APTMS-N,N-EPANTf 2 was determined to be 130.53 mg of zirconium per gram of adsorbent. This value was also compared with those previously reported for the adsorption capacity of Zr(IV) in other studies (1.15 [27], 50 [46], 86 [47] and 179 [48] mg¨g´1). Moreover, the adsorption capacity of Zr(IV) on activated SG was calculated to be 73.07 mg¨g´1 (Figure 4), providing that the uptake capacity for Zr(IV) was improved by 78.64% with newly modified SG-APTMS-N,N-EPANTf 2 adsorbent.    Figure 4 displays the uptake capacity of SG-APTMS-N,N-EPANTf2 towards Zr(IV) obtained from the adsorption isotherm experiment. As can be deduced from Figure 4, the adsorption capacity of SG-APTMS-N,N-EPANTf2 was determined to be 130.53 mg of zirconium per gram of adsorbent. This value was also compared with those previously reported for the adsorption capacity of Zr(IV) in other studies (1.15 [27], 50 [46], 86 [47] and 179 [48] mg·g −1 ). Moreover, the adsorption capacity of Zr(IV) on activated SG was calculated to be 73.07 mg·g −1 (Figure 4), providing that the uptake capacity for Zr(IV) was improved by 78.64% with newly modified SG-APTMS-N,N-EPANTf2 adsorbent.

Adsorption Isotherm Models
Adsorption isotherm studies were mainly applied in order to obtain an equation that accurately explained the results. Based on that, well known models describing the adsorption nature on adsorbent were evaluated. The Langmuir adsorption isotherm represents a monolayer adsorption on active sites of the adsorbent. The Langmuir adsorption isotherm equation can be obtained as follows [49]:

Adsorption Isotherm Models
Adsorption isotherm studies were mainly applied in order to obtain an equation that accurately explained the results. Based on that, well known models describing the adsorption nature on adsorbent were evaluated. The Langmuir adsorption isotherm represents a monolayer adsorption on active sites of the adsorbent. The Langmuir adsorption isotherm equation can be obtained as follows [49]: where, C e is the concentration of Zr(IV) in solution at equilibrium (mg¨mL´1), and q e refers to the amount of Zr(IV) per gram of SG-APTMS-N,N-EPANTf 2 adsorbent at equilibrium (mg¨g´1). Langmuir constants Q o and b correspond to related to the maximum Zr(IV) adsorption capacity (mg¨g´1) and affinity parameter (L¨mg´1), respectively. These constants can be estimated from a linear plot of C e /q e versus C e . Moreover, a dimensionless constant separation factor, R L , represents intrinsic characteristics of the Langmuir isotherm and can be obtained as follows:

Effect of Contact Time
The shaking time is an essential factor for estimating the time required for the adsorption process to attain equilibrium. In order to determine the equilibrium time of Zr(IV) on SG-APTMS-N,N-EPANTf2, a batch procedure was carried out at different contact times (2.5, 10, 20, 30, 40, 50, 60, 75 and 90 min). Figure 6 presents the contact time profile vs. milligrams of Zr(IV) adsorbed per gram SG-APTMS-N,N-EPANTf2. As shown in Figure 6, the adsorption of Zr(IV) increased when the shaking time increased, and equilibrium is attained after 60 min. As can be clearly observed from Figure 6

Effect of Contact Time
The shaking time is an essential factor for estimating the time required for the adsorption process to attain equilibrium. In order to determine the equilibrium time of Zr(IV) on SG-APTMS-N,N-EPANTf 2 , a batch procedure was carried out at different contact times (2.5, 10, 20, 30, 40, 50, 60, 75 and 90 min). Figure 6 presents the contact time profile vs. milligrams of Zr(IV) adsorbed per gram SG-APTMS-N,N-EPANTf 2 . As shown in Figure 6, the adsorption of Zr(IV) increased when the shaking time increased, and equilibrium is attained after 60 min. As can be clearly observed from

Kinetic Models
The study of the kinetic mechanism that controls the adsorption process is important for assessment of the time course for the adsorption process. Therefore, experimental data of Zr(IV) adsorption on SG-APTMS-N,N-EPANTf2 were analyzed in terms of both pseudo first-and second-order kinetic equations. The pseudo first-order equation can be represented by the following equation

Kinetic Models
The study of the kinetic mechanism that controls the adsorption process is important for assessment of the time course for the adsorption process. Therefore, experimental data of Zr(IV) adsorption on SG-APTMS-N,N-EPANTf 2 were analyzed in terms of both pseudo first-and second-order kinetic equations. The pseudo first-order equation can be represented by the following equation [51]: where, q e (mg¨g´1) and q t (mg¨g´1) correspond to the amount of Zr(IV) adsorbed at equilibrium and at time (min), respectively. The adsorption rate constant k 1 (min´1) and adsorption capacity q e can be calculated from the linear plot of log(q e -q t ) versus t. The pseudo second-order equation can be displayed as follows [51]: t{q t " 1{ν˝`p1{q e qt where, ν˝" k 2 q 2 e is the initial adsorption rate (mg¨g´1¨min´1) and k 2 (g¨mg´1¨min -1 ) refers to the rate constant, q e (mg¨g´1) corresponds to the amount of Zr(IV) adsorbed at equilibrium, and q t (mg¨g´1) denotes the amount of Zr(IV) on the SG-APTMS-N,N-EPANTf 2 surface at any time (min). From the linear plot of t/q t versus t, kinetic parameters of q e and ν˝can be estimated.

Kinetic Models
The study of the kinetic mechanism that controls the adsorption process is important for assessment of the time course for the adsorption process. Therefore, experimental data of Zr(IV) adsorption on SG-APTMS-N,N-EPANTf2 were analyzed in terms of both pseudo first-and second-order kinetic equations. The pseudo first-order equation can be represented by the following equation   From the results, the pseudo second-order equation was found to better represent the adsorption kinetics data than the pseudo first-order equation. From the pseudo second-order equation, the correlation coefficient factor of (R 2 ) = 0.999 (Figure 7). In addition, parameters q e , ν˝and k 2 were determined to be 133.1 mg¨g´1, 69.28 mg¨g´1¨min´1 and 0.004 g¨mg´1¨min´1, respectively. It can be clearly noticed that the value of q e is relatively close to those acquired from the adsorption isotherm experiments (130.53 mg¨g´1) and from the Langmuir model (130.69 mg¨g´1), strongly supporting the validity of the Langmuir isotherm model.

Effect of Interfering Ions
The influence of complex matrices on the extraction of Zr(IV) was investigated in order to assess the applicability of the proposed methodology for analytical applications in real samples. For this purpose, 1 mg¨L´1 Zr(IV) in 25 mL sample solutions were prepared with either individual or mixed interfering ions. According to Table 2, it is motivating to note that Zr(IV) extraction was not influenced by the medium composition despite the presence of major cations like Na + , K + , NH 4 + , Ca 2+ and Mg 2+ in 3000-fold excess, Cd 2+ , Cu 2+ and Pb 2+ in 500-fold excess, Zn 2+ in 600-fold excess, Fe 3+ in 300-fold excess, and Al 3+ or Cr 3+ in 400-fold excess or anions (7000-fold excess of Cl´, F´, NO 3´, 6000-fold of CO 3 2´, SO 4 2´a nd 5000-fold of PO 4 3´) . This can be attributed to the low uptake capacity of SG-APTMS-N,N-EPANTf 2 phase towards these species. Based on the above results, one can easily conclude that the newly prepared SG-APTMS-N,N-EPANTf 2 phase has high selectivity for the adsorption of Zr(IV), and the proposed method can be applied to the detection of Zr(IV) in real samples. Table 2.

Real Sample Analysis
For checking the reliability of the method, real water samples, including drinking water, tap water, seawater and lake water, collected from Jeddah in Saudi Arabia were subjected to the determination of Zr(IV). The standard addition method was utilized for the analysis of water samples under the same batch conditions. The extraction of Zr(IV) in spiked water samples is reported in Table 3, which results reveal that the extraction of Zr(IV) is in the range of 94.47%-98.96%, confirming the feasibility and suitability of the proposed methodology.

Samples
Added (

Conclusions
In this study, a newly developed hydrophobic ionic liquid (N,N-EPANTf 2 ) was immobilized on the SG chemically bound amine (SG-APTMS) as a support material. The procedure offered simplicity, and effectiveness and this new reagent provided a selective extraction technique for the determination and extraction of Zr(IV) in aqueous samples with acceptable precision. Performance of the adsorption process of Zr(IV) onto SG-APTMS-N,N-EPANTf 2 phase followed a Langmuir isotherm model, confirming the growth of a single/monolayer on a homogeneous adsorbent phase. The kinetic isotherm results of Zr(IV) obeyed a pseudo-second order kinetic model. The determination or separation of Zr(IV) by SG-APTMS-N,N-EPANTf 2 was also not significantly affected by the phase or interferents. The anticipated technique can be employed for the determination of Zr(IV) ions in real water samples with reasonable and reliable results.