A new route synthesis of immobilized-polysiloxane iminodiacetic acid ligand system, its characterization and applications

07) 4553–4558www.elsevier.com/locate/matlet

Materials Letters 61 (20

A new route synthesis of immobilized-polysiloxane iminodiacetic acidligand system, its characterization and applications

Nizam M. El-Ashgar a, Issa M. El-Nahhal b,⁎, Mohamed M. Chehimi c,Florence Babonneau d, Jacques Livage d

a Department of Chemistry, The Islamic University of Gaza, P O Box 108, Gaza, Palestinian Territoryb Department of Chemistry, Al-Azhar University, P O Box 1277, Gaza, Palestinian Territory

c ITODYS, Université Paris 7-Denis Diderot, associé au CNRS (UMR 7086), 1 rue Guy de la Brosse, 75005 Paris, Franced Laboratoire de Chimie de la Matière Condensée, Université Paris VI, 4 place Jussieu, 75252 Paris, France

Received 2 December 2006; accepted 23 February 2007Available online 1 March 2007

Abstract

A new route for the synthesis of immobilized-polysiloxane ligand system bearing iminodiacetic acid chelating ligand of the general formulaP-(CH2)3N-(CH2COOEt)2, (where P represents [Si–O]n polysiloxane network) was accomplished. The preparation of the immobilized iminodiethyl-acetate was achieved by the reaction of 3-aminopropyltrimethoxysilane with ethyl chloroacetate followed by hydrolytic polycondensation of thediethyliminodiacetate silane agent and tetraethylorthosilicate via the sol–gel process. The immobilized iminodiacetic acid P-(CH2)3N-(CH2COOH)2was obtained by hydrolysis of the ethyl acetate groups using diluted hydrochloric acid. The new functionalized ligand system exhibits high capacitiesfor uptake of the metal ions (Co2+, Ni2+, Cu2+ and Zn2+), and forms 1:1 metal to ligand ratio complexes in the case of Ni2+ and Cu2+.© 2007 Elsevier B.V. All rights reserved.

Keywords: Metal uptake; Iminodiacetic acid; Ethyl chloroacetate; Sol–gel process polysiloxanes; Polysiloxane-immobilized ligand systems

1. Introduction

The incorporation of chelating groups onto polysiloxanematrices is finding its way into an increasing number of ap-plications in areas such as extraction, recovery and separation ofmetal ions from aqueous solutions [1–5] and organic solvents[6,7]. Further applications of these systems are their use inchromatography [8–10] and catalysis [11 12]. Polysiloxanematrices are one of the most rapidly expanding areas of materialsresearch and development. One of the widely current usedmethods for the preparation of the polysiloxane ligand systems isthe sol–gel process [13]. This process involves hydrolysis andpolycondensation reaction of Si(OEt)4 and (MeO)3Si(CH2)3X(where X is an organic ligand group) leading to three dimen-sional cross linking silicalike matrix bearing a chelating ligandgroup. There are two strategies for the preparation of im-mobilized ligand systems: the first strategy is to react the silane

⁎ Corresponding author. Tel.: +972 82135612; fax: +972 82823180.E-mail addresses: [email protected] (N.M. El-Ashgar),

[email protected] (I.M. El-Nahhal).

0167-577X/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.matlet.2007.02.050

with the complexing group, and then to immobilized the com-plexing ligand with Si(OEt)4. The second strategy is to treat thepost-polysiloxane with the complexing group. The first strategyresults in higher amount of complexing ligands than the post-treatment polysiloxane and thus increases the metal uptakecapacities. Many techniques have recently been applied tocharacterize their chemical structures including high-resolutionsolid state nuclear magnetic resonance (NMR) X-ray photo-electron spectroscopy (XPS) and FT-IR [14–20]. This articledescribes an alternative synthesis route for the preparation of apolysiloxane with iminodiacetic acid and its capacity for uptakeof metal cations. The new preparationmethod enables the uptakeof a higher concentration of metal ions due to a higher amount ofimmobilized iminodiacetic acid ligands. The immobilizediminodiacetic acid was prepared by modification of aminosilaneagent with ethyl chloroacetate then the diethyliminodiacetatesilane agent was reacted with tetraethylorthosilicate (TEOS) toproduce the immobilized diethyliminodiacetate ligand system.The immobilized ligand exhibits higher potential for extractionof metal ions (Co2+, Ni2+, Cu2+ and Zn2+) than its analogousprecursor which was made by the reaction of iodopolysiloxane

4554 N.M. El-Ashgar et al. / Materials Letters 61 (2007) 4553–4558

and diethyliminodiaceteate [21]. The material was well char-acterized by high-resolution solid state nuclear magnetic res-onance (NMR), X-ray photoelectron spectroscopy (XPS), andFT-IR analysis.

2. Experimental

2.1. Reagents and materials

Tetraethylorthosilicate, 3-aminopropyltrimethoxysilane, metal(II) chloride and ethyl chloroacetatewere purchased from (Merck)and used as received. Diethyl ether and methanol (spectroscopicgrade) were used as received. Metal(II) solutions of theappropriate concentration were prepared by dissolving the metal(II) chloride (analar grade) in deionized water. pH (3.5–6) wascontrolled by using acetic acid/sodium acetate buffer solution.

2.2. General techniques

XPS spectra were obtained using a Thermo VG ESCALAB250 spectrometer (East Grinsted, UK). This instrument isequipped with a monochromated Al Kα X-ray source (650 μmspot-size) operating at 200 W, and a magnetic lens that permitsto increase the sensitivity. Static charge compensation wasachieved using a low energy electron source that ensures anegative but uniform charge compensation even for highlytopographic powdery insulating materials. The pass energy wasset at 150 and 40 eV for the survey and narrow scans, respectively.Ultimate high-resolution C1s spectra were recorded using a passenergy of 10 eV. The step size was 1.0 eV for the survey spectrumand 0.1 eV for the narrow scans. Take-off angle was 0° relative tothe surface normal. Spectrawere charge referenced to the aliphaticC–C/C–H C1s component set at 285.0 eV binding energy.

The apparent surface composition (in at.%) was computedusing the integrated peak areas and the corresponding Scofieldsensitivity factors corrected to the analyzer transmission function.

13C CP/MAS solid state NMR experiments were carried outat room temperature on a Bruker MSL-400 MHz spectrometerat a frequency of 100.6 MHz (13C) using a Bruker CP/MASprobe. Proton decoupling was always applied during acquisi-tion. Solid samples were spun at 5 kHz using 7 mm ZrO2 rotorsfilled in a glove-box under dried argon atmosphere.

Analysis for carbon, hydrogen, and nitrogen were carried out,using an Elemental Analyzer EA 1110-CHNS CE Instrument.

The infrared spectra for the materials were recorded on aPerkin-Elmer FT-IR, spectrometer using KBr disk in the range of4000 to 400 cm− 1.

All pH measurements were obtained using a HM-40 V pHMeter.

All ligand samples were shaken in aqueous metal ionsolutions using an ELEIA-Multi Shaker.

2.3. Preparations

2.3.1. Preparation of diethyliminodiacetatetrimethoxy silane agentEthyl chloroacetate (12.2 g, 0.1 mol) was added drop wise to

3-aminopropyltrimethoxysilane (5.4 g, 0.05 mol) in 15 cm3

ethanol with stirring. A stoichiometric amount of triethylaminewas added to trap all the liberated HCl. The mixture wasrefluxed with stirring for 48 h then cooled. The product wasisolated and identified using FT-IR spectra.

2.3.2. Preparation of polysiloxane-immobilized diethylimino-diacetate ligand system (P-DIDA)

Tetraethylorthosilicate (20.8 g, 0.1 mol) was added to 25 cm3

of absolute ethanol with stirring and the diethyliminodiacetateproduct (0.05 mol) was added drop wise followed by 4.95 g0.042MHCl as catalyst. Polymerization occurred within 3 days.The product was dried in a vacuum oven at 100 °C for 12 h, thencrushed, sieved, washed with successive portions 30 cm3 of0.025 M NaOH, water, methanol and diethyl ether and dried for12 h at 100 °C in a vacuum oven.

2.3.3. Preparation of polysiloxane-immobilized iminodiacetate(P-IDA)

The immobilized iminodiacetic acid P-IDA was obtainedwhen the polysiloxane-immobilized diethyliminodiacetate (5.0 g)was refluxed with 100 cm3 of 0.5 M HCl for 12 h. The productwas filtered, washed with 0.025 M NaOH, water, methanol anddiethyl ether and dried for 12 h at 100 °C in a vacuum oven.

2.4. Metal uptake experiments

A 100 mg sample of the functionalized polysiloxane-immobilized ligand system, P-IDA was shaken with 25 cm3

of 0.02 M aqueous solution of the appropriate metal ions (Co2+,Ni2+, Cu2+ and Zn2+) using 100-cm3 polyethylene bottles.Determination of the metal ion concentration was carried out byallowing the complex material to settle down, withdrawing anappropriate volume of the supernatant using a micropipette andthen diluting to the linear range of the calibration curve for eachmetal. The metal ion uptake was calculated as mmol of Mn+/gligand used. Each study was performed at least in triplicate. Theconcentration of metal ions in their aqueous solutions wasmeasured using a Perkin-Elmer AAnalyst-100. Metal uptakewas examined at different time intervals and pH values.

3. Results and discussion

3.1. Preparation of P-IDA polysiloxane-immobilized ligand system

The functionalized polysiloxane-immobilized ligand system P-IDA,was prepared using the sol-gel process as follows:

1– Preparation of diethyliminodiacetate silane agent by the reactionof 3-aminopropyltrimethoxysilane, with ethyl chloroacetate inthe ratio of 1:2 (Scheme 1).

Fig. 1. FT-IR spectra of (A) the immobilized P-DIDAand (B) the immobilized P-IDA.

4555N.M. El-Ashgar et al. / Materials Letters 61 (2007) 4553–4558

2– Hydrolytic polycondensation of the diethyliminodiacetate silaneagent with tetraethylorthosilicate (TEOS), in the ratio of 1:2(Scheme 2).

3– The functionalized ligand system P-IDA, was obtained by acidichydrolysis of the ethyl acetate groups of P-DIDA (Scheme 3).

3.2. Elemental analysis

The elemental analysis for the P-DIDA and P-IDA is given in Table 1.The elemental analysis results of P-DIDA show a lower C, H and Npercentages than the expected values (Table 1). This is probably due to theformation of low molecular weight small oligomers, which dissolvedduring the washing process. The decrease of carbon and hydrogenpercentages and the increase of nitrogen values upon treatment of the P-DIDAwith HCl are expected due to hydrolysis of ethyl acetate groups(–CO2Et) into the acid form (–CO2H).

The elemental analysis also showed that this route of synthesis of theimmobilized P-IDA shows a higher N content (N=3.2%) compared withthe previous route (N=1.4%) in which the P-IDA reacted with diethyliminodiacetate (Table 1) [21]. This is probably due to a higher amount ofimmobilized iminodiacetic acid groups. However the explanation whythis synthesis results in a higher amount of iminodiacetic acid groups, isthat the strategy of preparing the silane with the complexing group firstresults in a higher amount of complexing ligands than post-treatment ofthe polysiloxane.

3.3. FT-IR spectra

The FT-IR spectra of the P-DIDA ester derivative and the acid form(P-IDA) are given in Fig. 1. The spectra show three regions ofabsorption at 3500–3000 cm− 1 due to ν(OH) or ν(NH), 1750–1500 cm− 1 due to δ(OH), δ(NH), ν(C_O) and 1200–900 cm− 1 due toν(Si–O) [19,20].

The FT-IR spectrum for the immobilized-polysiloxane P-DIDA (esterform) shows strong absorption bands around 1748 cm−1 and 1643 cm−1

Table 1Elemental analysis data for the polysiloxane-immobilized diethyliminodiacetate(P-DIDA) and the modified form P-IDA

Polysiloxane Element %C %H %N C/N

P-DIDA Expected a 32.8 4.9 3.5 10.9Found 26.2 4.6 2.9 10.5

P-IDA Expected a 19.7 2.8 3.3 7.0Found 19.1 2.7 3.2 7.0[21] 16.7 3.5 1.4 13.9

a The expected values calculated in the case of P-DIDA assuming 100% reactionwhile in the case of P-IDA the values are based on the results found for P-DIDA.

due to ν(C_O) and δ(OH) vibrations respectively (Fig. 1A). After hy-drolysis the strong absorption band around 1748 cm−1 is shifted to a lowerfrequency and probably overlapped with δ(OH) vibration at 1643 cm−1

corresponding to the ν(C_O) of the acid form (–COOH) (Fig. 1B). Thepresence of a small shoulder at 1748 cm−1 after the hydrolysis mayindicate that not all the ester groups (–COOEt) were converted into theacid form (–COOH).

3.4. 13C CP/MAS NMR spectra

The CP/MAS 13C NMR spectra for the immobilized P-DIDA andP-IDA ligand systems are shown in Fig. 2. The spectrum of theimmobilized diethyliminodiacetate ligand system (P-DIDA) (Fig. 2A)shows three signals at 9.4, 21.5 and 43.1 ppm due to the threemethylene carbons C1, C2 and C3 (Formula 1). The broad signal at168 ppm is due to the carbonyl C5. The two shoulders at 63.1 and 14.5are due to C6 and C7 of the acetate group. The shoulder around 58 ppmis due to C4. These assignments were based on spectral data of similarsystems reported in literature [14–18,21].

Formula 1

The CP/MAS 13C NMR spectrum, of the immobilized P-IDA isshown in Fig. 2B. The spectrum of the immobilized iminodiacetic acidligand system (P-IDA) shows three carbon signals at 9.7, 21.3 and42.7 ppm due to three methylene carbons C1, C2 and C3 respectively(Formula 2). The signals at 170 and 60(sh) ppm are due to C5 and C4.The presence of two signals with intensities at 14.5 and 63.5 ppm issignificantly decreased due to the hydrolysis of (–OCH2CH3) groupswith HCl. This is also providing evidence that not all the ester groupsare hydrolyzed. Further evidence was confirmed by FT-IR discussed

Fig. 2. 13C NMR spectra of (A) the immobilized P-DIDA and (B) the immobilized P-IDA.

4556 N.M. El-Ashgar et al. / Materials Letters 61 (2007) 4553–4558

earlier. The extra signals or shoulders at 15–18 and 47–50 ppm areprobably due to the residual alkoxysilyl groups (Si–O–CH3 and Si–O–C2H5) from the tetraethoxy and the silane coupling agents, due toincomplete hydrolysis and condensation [14].

Formula 2

Fig. 3. Survey scans of P-IDA and P-IDA/copper complex.

3.5. XPS results

Fig. 3 shows the survey spectra of P-IDA as prepared and itscopper (II) complex. The main peaks Si2p, C1s, N1s and O1s are

centered at 102, 285, 400 and 532 eV, respectively. In the case ofP-IDA after retention of copper ions (Fig. 3b), an additional Cu2pdoublet is observed at the 930–960 eV.

Table 2XPS data and surface composition (in at.%) of P-DIDA, P-IDA and its copper (II) complex

System Element C Si O N Cl Na Cu

C1s Si2p O1s N1s Cl2p Na1s Cu2p3

P-DIDA Core-line 285 102 532 399.5 198.6 1073.8 934.3% Composition A 39.2 18.9 38.0 2.8 2.3 –

B 38.8 16.9 41.4 2.8P-IDA Core-line 285 102 532 399.5 0.59

% Composition A 41.9 18.2 36.1 2.64 0.62B 32.4 20.2 44.8 2.45

P-IDA-Cu Core-line 285 102 532 399.5% Composition A 37.0 19.9 38.8 2.82 0.94 – 1.55

B 33.6 18.4 43.6 2.9 1.34

A: This work.B: [21].

4557N.M. El-Ashgar et al. / Materials Letters 61 (2007) 4553–4558

The apparent surface compositions (at.%) is reported in Table 2.In contrast to elemental analysis, the XPS results could not accountfor the decrease of the carbon content as a result of the esterhydrolysis, probably because it affected more the bulk than thesurface of the sol–gel materials. Nevertheless, considering thechemical structure of the ligand (one nitrogen atom per ligand) andthe atomic percent of silicon and nitrogen, the ligand/Si ratio is about0.15, that is one ligand per six or seven silicon atoms. The high-resolution N1s region from P-IDA (Fig. 4) is fitted with twocomponents centered at 400.2 and 402.2 eV which are assigned tothe free, neutral nitrogen atoms and their corresponding protonatedform respectively. The fraction of the N+ form levels off to 40.6%and the neutral nitrogen form is about 59.4%. These results areclosed to the reported data described before [21].

Fig. 5 displays the fitted C1s region of the P-IDA system. Thecomplex spectrum is fitted with five components centered at 285, 285.7,286.5, 287.6 and 288.8 eV and assigned to the C–C/C–H, C–N, C–O,C_O and COOH, respectively. Interestingly the C–N/COOH ratio isabout 1.5, in line with the chemical structure of the ligand. The P-DIDAhas a similar structure (not shown). The high binding energy positiondue to the ester group is centered at 289.2 eV while that of the acidligand is at 288.8 eV. This slight negative shift could account for thechemical transformation of the ester to the acidic (COOH) or carbo-xylate (COO−) form.

Fig. 4. N1s fitted region of the P-IDA sol–gel material.

3.6. Metal uptake

The metal ion uptake capacity (Co2+, Ni2+, Cu2+ and Zn2+) as mmolM2+/g ligand, was determined by shaking the functionalized ligandsystem (P-IDA) with buffered metal ion solutions. The results in mmolM2+/g ligand are given in Table 3.

From the elemental analysis of N of the immobilized ligand (P-IDA) given in Table 1 (3.2% i.e. 2.28 mmol N/g ligand) andthe maximum metal ion uptake, it is possible to suggest that the1:1 metal to ligand complexes are expected in the case of Ni2+ andCu2+. Whereas in the case of Co2+ and Zn2+ the metal to ligandcomplex ratio is smaller than 1:1 indicating that less stable com-plexes are formed. It is clear that uptake of metal ions decreases inthe order:

Cu2þNNi2þNCo2þNZn2þ

3.6.1. Effect of shaking timeThe metal ion uptake capacity (Co2+, Ni2+, Cu2+ and Zn2+) was

determined by shaking the functionalized ligand system P-IDA withbuffered solutions of the metal ions. Measurements were carried out atdifferent time intervals. It is shown that the metal ion uptake isincreased as a function of shaking time and reached equilibrium nearlyafter 48 h where maximum uptake is obtained. Similar results wereobserved for the other metal ions.

Fig. 5. C1s region of the P-IDA ligand system.

Table 3Metal uptake capacities

Metal ion Co2+ Ni2+ Cu2+ Zn2+

Maximum uptake (mmol M2+/g ligand) 1.73 2.06 2.31 1.56

4558 N.M. El-Ashgar et al. / Materials Letters 61 (2007) 4553–4558

3.6.2. Effect of pHThe effect of the pH value on the uptake of Co2+, Ni2+, Cu2+ and Zn2+

ions by P-IDAwas conducted. The results show an increase of metal ionuptake with increasing pH value and reached its maximum at pH 5.5 inthe case of Ni2+ and Cu2+ ions. Low uptake capacity occurs at lower pHvalues. This is probably due to protonation of nitrogen atom and un-dissociation of the carboxylic acid groups.

4. Conclusion

The immobilized ligand system P-IDA was prepared byhydrolytic polycondensation reaction of Si(OEt)4 and (MeO)3Si(CH2)3N(CH2CO2H)2 followedby hydrolysis of ester groups. FT-IR, 13CNMRandXPS results of the P-IDA suggest that not all theester groups (–COOR) have hydrolyzed upon treatment of theimmobilized diethyl iminodiacetate ligand system (P-DIDA)withHCl. The XPS results showed that the amine nitrogen atoms existin both forms: the free amine form and the protonated aminecation or hydrogen bonded with surface silanol groups. Thisimmobilized ligand system exhibits higher potential for extractionof metal ions than the previously reported method in which theiodopolysiloxane is treated with diethyliminodiacetate. From themaximum metal uptake capacities it is possible to suggest that the1:1metal to ligand ratio complexes are obtained in the case ofNi2+

and Cu2+.

References

[1] I.S. Khatib, R.V. Parish, J. Organomet. Chem. 369 (1989) 9.[2] I.M. El-Nahhal, F.R. Zaggout, N.M. El-Ashgar, Anal. Lett. 33 (10) (2000)

2031.[3] N.M. El-Ashgar, I.M. El-Nahhal, M.M. Chehimi, F. Babonneau, J. Livage,

Monatsh. Chem. Mon. 137 (2006) 263.[4] I. Ahmed, R.V. Parish, J. Organomet. Chem. 452 (1993) 23.[5] R.V. Parish, D. Habibi, V. Mohammadi, Organomet. Chem. 369 (1989) 17.[6] G.S. Prado, L.N.H. Arakaki, C. Airoldi, J. Chem. Soc. Dalton Trans. 2206

(2001).[7] G.S. Prodo, C. Airoldi, Anal. Chim. Acta 432 (2001) 201.[8] N.M. El-Ashgar, Issa M. El-Nahhal, J. Sol–Gel Sci. Tech. 34 (2) (2005)

165.[9] A.A. El-Nasser, R.V. Parish, J. Chem. Soc., Dalton Trans. 3463 (1999).[10] L. Trojer, G. Stecher, I. Feuerstein, S. Lubbad, G.K. Bonn, J. Chromatogr.,

A 1079 (2005) 197.[11] J. Cermak, M. Kvicalova, V. Blechta, M. Capka, Z. Bastl, J. Organomet.

Chem. 509 (1996) 77.[12] P.M. Price, J.H. Clark, D.J. Macquarrie, J. Chem. Soc. Dalton Trans 101

(2000).[13] L.L. Hench, J.K. West, Chem. Rev. 90 (1990) 33.[14] J.J. Yang, I.M. El-Nahhal, I.S. Chung, G.E. Maciel, J. Non-Cryst. Solids

209 (1997) 19.[15] J.J. Yang, I.M. El-Nahhal, G.E. Maciel, J. Non-Cryst. Solids 204 (1996)

105.[16] J.J. Yang, I.M. El-Nahhal, I.S. Chuang, G.E. Maciel, J. Non-Cryst. Solids

212 (1997) 281.[17] M. El-Nahhal, M.M. Chehimi, C. Cordier, G. Dodin, J. Non-Cryst. Solids

275 (2000) 142.[18] N.M. El-Ashgar, I.M. El-Nahhal, M.M. Chehimi, M. Delmar, F. Babonneau,

J. Livage, J. Sol–Gel Sci. Tech. (2006) (online).[19] T. Lopez, P. Bosch, M. Moran, R. Gomes, J. Phys. Chem. 97 (1993) 167.[20] T. Lopez, L. Herrerra, J. Mendez, P. Bossh, R. Gomes, R. Gozalez, J. Non-

Cryst. Solids 147 (1992) 773.[21] I.M. El-Nahhal, F.R. Zaggout, M.A. Nassar, N.M. El-Ashgar, J. Maquet,

F. Babonneau, M.M. Chehimi, J. Sol–Gel Sci. Tech. 28 (2003) 255.