EUROPEAN POLYMER JOURNAL
European Polymer Journal 43 (2007) 3838–3848
www.elsevier.com/locate/europolj
Crown moieties as cation host units in model polyamide compounds: Application in liquid–liquid cation extraction and in membrane cation transport Marı´a J. Tapia a, Artur J.M. Valente b,*, Hugh D. Burrows b, Vero´nica Caldero´n c, Fe´lix Garcı´a a, Jose´ Miguel Garcı´a a a
Departamento de Quı´mica, Facultad de Ciencias, Universidad de Burgos, Plaza de Misael Ban˜uelos s/n, E-09001 Burgos, Spain b Departamento de Quı´mica, Universidade de Coimbra, 3004-535 Coimbra, Portugal c Departamento de Construcciones Arquitecto´nicas e Ingenierı´as de la Construccio´n y del Terreno, Escuela Polite´cnica Superior, Universidad de Burgos, Villadiego s/n, E-09001 Burgos, Spain Received 17 April 2007; received in revised form 12 June 2007; accepted 14 June 2007 Available online 28 June 2007
Abstract A new model polyamide compound that has a benzo-18-crown-6 moiety in the pendant structure is described. This model interacts with metal cations in the alkaline, earth alkaline, transition metal and heavy metal series. The interaction has been analyzed in terms of competitive cation extraction from aqueous solution by liquid model/dichloromethane phase. In each cation series, K(I), Ba(II), Cr(III), and Hg(II) have been selectively extracted by liquid model polyamide phases. The interaction of a dense composite model polyamide-cellulose acetate membrane with lead(II) has been studied through its adsorption isotherm, infrared spectra and scanning electron microscopy study of the membranes before and after Pb(II) adsorption. The transport of lead nitrate through the membrane together with that of sodium chloride (for comparison), have been evaluated. 2007 Elsevier Ltd. All rights reserved. Keywords: Polyamide; Crown compounds; Cellulose acetate; Metallic ions; Liquid–liquid extraction
1. Introduction The discovery of the ability of biological molecules, such as porphyrins or valinomycin, to selectively bind cations has opened new research fields * Corresponding author. Tel.: +351 239854459; fax: +351 239827703. E-mail address:
[email protected] (A.J.M. Valente).
in cation purification, extraction, environmentally toxic cation elimination or sensing. An early example of synthetic molecules which mimic Nature are the macrocyclic polyethers, which can selectively interact with cation through ion–dipole interactions of the positively charged cations involving the lone pairs of the ether oxygens. These macrocycles are cyclic polyethers made up of ethylene glycol units, (OCH2CH2)n, and were called crown ethers by
0014-3057/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.eurpolymj.2007.06.021
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
C.J. Pedersen in 1967 due to the appearance of crowns of the space filling models of the compound in the polyether–metal ion complex [1,2]. The specificity of the interaction of the crown ether with cations makes these compounds useful for preparation of novel materials as metal ion catalysts, molecular imprinting compounds, chirality inducing reagents, ion-exchange membranes, selective solvent extraction separation, selective membrane transport, ionoselective membranes for sensors, etc. [3–8]. We have previously studied the interactions of cations with methacrylic polymers having pendant crown ether moieties [9,10]. In this work, we extend the study to a new model compound of the polyamide bearing a pendant benzo-18-crown-6 ether moiety in each structural unit. The model compound of a polymer is a discrete, low molecular weight molecule that has a chemical structure, which resembles the structural unit of the polymer. The model compounds are used to study some characteristics of the polymers that cannot be studied in the macromolecule. The lack of solubility of the polyamide inhibits their study in liquid–liquid extraction. Thus, the model studied in this work is a host molecule for cations, which acts as probes of the polyamide and can be used in liquid–liquid extraction. Furthermore, the use of the model compound allows evaluation of the interaction and extraction of cations with a single 18-crown-6 host moiety, because in the polyamide, the crown moiety is a part of the pendant polymer backbone, so that the mobility of these subgroups is restricted and some kind of stacking is favored by the comb-like anchor of the pendant crown ether moieties to the main polymer chain [11–13]. In this manuscript, we describe the synthesis of a host molecule for cations, formally a model polyamide compound with a benzo-18-crown-6 moiety in the structure, together with its behavior in liquid– liquid extraction of cations in aqueous solution and its cation binding capabilities in a cellulose acetate matrix. Results of liquid phase of model compounds dissolved in organic solvents, in terms of selective cation uptake, have been obtained. Moreover, the interaction of a composite model polyamide–cellulose acetate dense membrane with lead(II) has been studied by means of its adsorption isotherm, infrared spectra and scanning electron microscopy studies. This membrane has also been tested for the transport of Pb(II) and Na(I). The data obtained should be useful both for applications in elimination, extraction or purifica-
3839
tions of cations, and also for a selectivity tests to prepare sensor membranes for applications in areas such as spectroscopic cation probes, ion-selective electrodes, or cation selective transport membranes. 2. Experimental 2.1. Materials All materials and solvents were commercially available and were used as received, unless otherwise indicated. N-Methyl-2-pyrrolidone (NMP) was vacuum-distilled twice, over phosphorous pent˚ molecular sieves. oxide, and then stored over 4 A Lithium chloride was dried at 400 C for 12 h prior to use. Triphenylphosphite (TPP) was vacuum distilled twice, over calcium hydride, and then stored ˚ molecular sieves. Pyridine was dried with over 4 A reflux over sodium hydroxide for 24 h, and distilled ˚ molecular sieves. The synthesis of 4over 4 A 0 0 (3 ,5 -dicarboxyphenylaminocarbonyl)benzo-18crown-6 was accomplished according to the procedures previously described [13]. 2.2. Model compound: 4-((3,5-bis(phenylcarbamoyl) phenyl)carbamoyl)benzo-18-crown-6 [MdC6] In a 50 ml three necked flask fitted with mechanical stirring, 22 mmol of aniline, 10 mmol of 4-(3 0 , 5 0 -dicarboxyphenylcarbamoyl)benzo-18-crown-6 and 1.4 g of lithium chloride were dissolved in a mixture of 6 ml of pyridine, 22 mmol of TPP and 20 ml of NMP. The solution was stirred and heated at 110 C under a dry nitrogen blanket for 4 h. The system was then cooled at room temperature and the product was precipitated in 300 ml of water. The product, MdC6 in Scheme 1, was purified by dissolving and re-precipitation (NMP and water, respectively), filtered off and dried overnight at 80 C in a vacuum oven. Yield: 80%; mp: amorphous solid (Tg = 75 C). 1 H NMR [400 MHz, deuterated dimethylsulfoxide (DMSO-d6), d, ppm]: 10,50 (s, 3H); 8.65 (s, 2H); 8.37 (s, 1H); 7.89 (d, 4H); 7.74 (m, 2H); 7.39 (t, 4H); 7.14 (d, 4H); 6.79 (m, 2H); 4.22 (s, 4H); 3.78 (s, 4H); 3.60 (s, 4H); 3.44 (m, 8H); 1.08 (t, 6H). 13 C NMR [400 MHz, deuterated dimethylsulfoxide (DMSO-d6), d, ppm]: 165.35; 165.19; 157.49; 151.68; 147.87; 139.91; 139.24; 135.97; 130.31; 129.44; 128.78; 126.73; 123.94; 122.81; 121.95; 121.86; 120.80; 118.86; 115.37; 113.88; 112.86; 70.25; 69.43; 69.12; 69.00; 68.86; 68.41; 65.73;
3840
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
Hg(II), Pb(II)] was carried out as follows. A solution of 50 mg of the model (MdC6) in 50 ml of dichloromethane was shaken thoroughly for a week at 25 C with a solution of the proper cation series in water. For every series, the sum of metal ions moles were equal to the moles of crown ether chemical groups, and all the metal ions of a series were introduced in equimolar quantities in such a way that direct information of model selectivity toward the metal ions in a series was obtained from the analysis of the cation concentration remaining in the water phase determined by ICP at the end of the liquid–liquid extraction process. 2.5. Membrane preparation
Scheme 1.
15.20. EI-LRMS (m/z, relative intensity): 326 (100); 325 (88); 233 (17); 232 (14); 215 (23); 170 (22); 169 (20); 77 (44); 65 (26). FT–IR (KBr, cm1): 3432, 2925, 1652, 1597, 1539, 1506, 1443, 1317, 1270, 1220, 1133, 756, 692. 2.3. Measurements 1
H and 13C NMR spectra were recorded with a Varian Inova 400 spectrometer operating at 399.92 and 100.57 MHz, respectively, with deuterated chloroform (CDCl3), or deuterated dimethylsulfoxide (DMSO-d6) as solvents. Low resolution electron impact mass spectra (EI-LRMS) were obtained at 70 eV on an Agilent 6890N mass spectrometer. Cation concentrations in the extraction studies and sorption isotherm were determined by inductively coupled plasma mass spectrometry (ICP, Agilent 7500 i). Consecutive dilutions of sample aliquots with ultra pure water/nitric acid (5% v/v) were made to reach concentration in the range of the calibration curve, from 0 to 40 ppb. 2.4. Liquid–liquid extraction The competitive liquid–liquid extraction of each cation series: alkaline [Li(I), Na(I), K(I), Rb(I), Cs(I)], earth alkaline [Mg(II), Ca(II), Sr(II), Ba(II)], transition metal [Cr(III), Mn(II), Co(III), Ni(II), Cu(II), Zn(II)] and heavy metal [Ag(I), Cd(II),
The cellulose acetate/model compounds films were prepared by dissolving cellulose acetate (CA) and the model compound (MdC6) in tetrahydrofuran and stirring for 24 h. A homogeneous colorless transparent film of 25 lm thickness was obtained by doctor-blade casting of a solution of MdC6 (2% w/v) and CA (15% w/v) on flat glass support, showing good miscibility and that no phase separation had occurred. After complete evaporation of solvent, membranes were removed from the glass support with the help of water. 2.6. Membrane characterization Membranes were characterised by their infrared spectra using a ATI Mattson Genesis Series FTIR spectrometer. The morphologies of the polymer films were analysed using a Jeol/Scanning Microscope, model 5310 under low vacuum, using a potential of 20 kV. 2.7. Transport experiments Permeability of salts was measured using a previously reported cell [14]. This consists of two 250 ml containers filled with salt solution (A) and water (B), respectively. These were connected by two 7 mm radius horizontal tubes, with the polymer membrane sealed, with silicone, between these two tubes. The change in the ionic solute concentration in cell B was determined during the permeability experiment by measuring the electrical resistance using a Wayne Kerr 4265 Automatic LCR Meter and a conductivity cell (WTW LR 325/01), 0.1 cm1 cell constant. This was calibrated prior to each experiment using at least five freshly prepared
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
standard solutions of the salt, with different concentrations. The same conditions were used for calibration and permeability experiments. Constant temperature (±0.1 C) was maintained by immersing the system in a thermostat bath (Velp Sientifica Multistirrer 6). Solutions in both cells were stirred at ca. 200 rpm to decrease the Nernst layer in the membrane-solution interface and to increase the reproducibility of the conductivity sensor. The permeability of ionic solutes through the polymeric membranes can be described in terms of Fickian diffusion oC=ot ¼ o=oxðDF oC=oxÞ
ð1Þ
with the boundary and initial conditions C(0,t) = C, C(l, t) = 0, and C(x, 0) = 0 (where, C(x, t) is the ionic concentration in the membrane at a distance x inside the membrane and at time t), resulting in the simple formulae for calculation of the permeability (P) and diffusion coefficient (DF) P ¼ Jl=c 2
DF ¼ l =ð6hÞ
ð2Þ ð3Þ
where l is the thickness of polymeric membrane, measured after each experiment at 25 C using a Mitutoyo micrometer (±0.001 mm), J is a steadystate flux through the membrane, h is the time-lag, and c is the bulk electrolyte concentration. 3. Results and discussion 3.1. Liquid–liquid extraction In liquid–liquid extraction, a distribution ratio of metal ions, M, between the organic phase (MO) containing MdC6 solved as extracting species and the aqueous phase (MA) initially containing equimolar quantities of metal ions of a series, is often quoted as a measure of how efficient the extraction process of a particular species is. The distribution ratio Kd is equal to the concentration of cations in the organic phase, MO, divided by its concentration in the aqueous phase, MA, (Eqs. (4) and (5)). In our systems, the MdC6 dissolved in the organic phase acts as an effective host molecule toward some cations, so the guest ions can be extracted from aqueous solutions into the organic phase remaining as host–guest couples. MA ¢ MO Kd ¼
½M O ½M A
ð4Þ ð5Þ
3841
The liquid–liquid simultaneous extraction of aqueous solutions of all the cations in each cation series by a dichloromethane solution of MdC6 has been studied in terms of extraction percentage, distribution coefficient and selectivity for each metal ion. The extraction percentage, %E, is the percentage of each ion extracted from the aqueous phase to the organic phase resulting from the presence of the host molecule. Since the study has been performed with equal volumes of organic and aqueous phases, the distribution coefficient as defined in Eq. (5) is the ratio of concentration of the cation on the two phases. Thus, Kd is a measure of the capacity of the host molecules to extract cations under such competitive conditions, see Eq. (6). %E Kd ¼ ð6Þ 100 %E The selectivity (a), Eq. (7), represents the ratio of two distribution coefficients. The subscripts indicate the liquid–liquid extraction (aL,L). The cation in each series with highest distribution coefficient (KdM2) is taken as reference for all the ions in that series. aL;L ¼
KdM1 KdM2
ð7Þ
Assuming an ideal interaction of one cation per host unit, the crown occupation level percentage (%O) has also been calculated for each cation series as shown in Eq. (8), where H-cation is the moles of {Host unit MdC6-cation} complexes and Ho the initial moles of crown model. The ratio gives an idea of the effectiveness of crown chemical structures in simultaneous cation ion liquid–liquid extraction. %O ¼
H cation 100 Ho
ð8Þ
Data of extraction percentage, distribution constant, selectivity and occupation level for each cation series are shown in Table 1. As the cavity of a 18-crown-6 is supposed to be ˚ and 3.2 A ˚ [15], if we consider it as a between 2.6 A ˚ , the effectiveness spherical hole of a radius of 1.45 A in cation extraction in alkaline and in alkaline earth cation series is related to the ionic radii of the cations. Thus, K(I) is extracted preferentially in series I, and Ba(II) in cation series II (Table 1). The ionic radii could also explain why the crown occupation level percentage (%O) is lower for alkaline and transition metal ions series than for alkaline earth and heavy metal series. Occupation is higher in the series
3842
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
Table 1 Liquid–liquid extraction data for four series of cations from aqueous phase to dichloromethane solutions of MdC6 model ˚) Cation series Cation Cation radii (A E (%) Kd aL,L
O (%)
I
Li(I) K(I) Rb(I) Cs(I)
0.76 1.38 1.52 1.67
8.4 35.1 4.6 3.2
0.09 0.54 0.05 0.03
0.17 1.00 0.09 0.06
12.81
II
Mg(II) Sr(II) Ba(II)
0.72 1.18 1.35
16.1 29.6 34.3
0.19 0.42 0.52
0.37 0.81 1.00
26.68
III
Cr(III) Mn(II) Co(III) Ni(II) Cu(II) Zn(II)
0.62 0.67 0.61 0.69 0.73 0.74
35.9 12.7 5.6 13.7 15.2 16.1
0.56 0.15 0.06 0.16 0.18 0.19
1.00 0.26 0.11 0.28 0.32 0.34
16.53
IV
Ag(I) Cd(II) Hg(II) Pb(II)
1.15 0.95 1.02 1.19
32.8 14.5 45.0 7.4
0.49 0.17 0.82 0.08
0.59 0.21 1.00 0.10
24.91
in which most of the metal ions radii are between ˚ and 1.4 A ˚ and lower if the majority of the metal 1A ˚ (transition ions of the series are smaller than 1 A ˚ metal) or larger than 1.4 A (alkaline). However, arguments based on the ionic radii cannot explain the relative percentage of extraction in transition metal and heavy metal series, indicating that in these cases the interaction between cations and host molecules depends probably on other factors, such as charge, degree of covalency of bonds, or, with transition metal ions, ligand field stabilization effects. The model polyamide compound, which acts as a host molecule for cations, is a promising molecule for selectively extracting cations. Thus, environmentally harmful cations, such as Cr(III) and Hg(II), can be recovered by this technique employing MdC6 as a host extractant. The molecule is also a promising structure to produce active transport membranes, or supported liquid membranes, for the purification/recovery of these cations, or for the preparation of sensing devices, such as ion-selective electrodes. Moreover, the data and results obtained herein could be extrapolated to the design of porous or dense re-usable polyamide membranes, such as homopolymer, copolymer or composite materials, for the separation, purification or recovery of environmentally harmful cations. Furthermore, polyamides bearing crown ether groups as pendant substructures have been recently described [11–13] by us and studies regarding cation transport
and extraction with these materials are currently being carried out. To gain an insight into the interaction of the crown moiety and metal ions in solid state, composite membranes of cellulose acetate containing the model compound (MdC6) were prepared, and lead(II) adsorption and transport properties were measured with several techniques. Lead(II) was selected for being one of the most widely spread toxic metal ions in developed countries, and one whose maximum levels are strictly regulated to protect health [16]. In addition, it has a stable divalent oxidation state, the nitrate salt is very soluble in aqueous solution and do not form aggregates, which facilitates the transport through membranes and the interpretation of sorption and transport results. 3.2. FTIR characterization of cellulose acetate membranes with 12 weight% of MdC6 (CA/MdC6) alone and with adsorption of lead Composite membranes of cellulose acetate (15% w/v) with 12 weight% MdC6 were prepared (CA/ MdC6, see Section (2). The main reasons and advantages for using cellulose acetate (CA) as a polymeric matrix for incorporation of MdC6 are: neutral properties, capacity for transparent film formation, ready processability from solution and low cost [16,17]. The infrared spectra of pure cellulose acetate membrane (Fig. 1a) show the characteristic bands
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
3843
Fig. 1. Infrared spectra of cellulose acetate (a) and cellulose acetate-MdC6 composite (b) films.
of cellulose acetate at 1739 (C@O stretching vibration), 1226 (CAOAC asymmetric stretching vibration), 1376 (CH3 deformation vibration of acetate group) and 1039 (CAO stretching vibration) cm1 [18]. The broad peak at ca. 3500 cm1 is assigned to OAH stretching vibration. However, in the infrared spectra of the MdC6-containing cellulose acetate film (Fig. 1b), the strongest bands occur at 1752, 1237, and 1059 cm1 (slightly shifted with respect to the pure CA film). In addition, all these bands show a greater intensity than the corresponding bands of cellulose acetate, suggesting an increased number of vibrational sites. The band shift of the most intense band and the increase in intensity observed at 1237 and 1059 cm1 can be justified by an increase of the fraction of vibrations stemming from the MdC6 CAOAC ether ring. However, the biggest alterations found on comparison of the IR spectra of the pure CA and the MdC6-CA membranes are: (a) a decrease of 25% in the intensity of CA band at 1374 cm1, and (b) a new distinct sharp band at 3469 cm1. By adding the model to the CA membranes, the relative intensity of the CH3 deformation vibration of acetate group (1374 cm1) diminishes with respect to those of C@O and CAO stretching bands because the model can contribute to the latter bands due to the amide and ether groups but not to the CH3 ones. The new band at 3469 cm1 may be explained by the existence of NAH stretching vibration bands of the amide groups of the model compound.
Experimental evidence of Pb(II) sorption by CA/ MdC6 composite films comes from FTIR spectroscopy (Fig. 2) of membranes which had been in contact with different Pb(II) solutions for a week at 25 C. The main differences in the spectra of CA/ MdC6 films upon Pb(II) sorption are found in the region 1380–1190 cm1, which can be assigned to CAOAC asymmetric stretching vibration. In the absence of lead, the CA/MdC6 FTIR spectrum show only one slightly broad peak in that region; however, in the presence of sorbed Pb(II), two specific bands at 1272 and 1235 cm1 can be found. Although, quantitative analysis is not possible, it is worthy of note that the intensity ratio at these two wavenumbers, I1272/I1235, shows a systematic decrease with an increase of initial Pb(II) concentration. That is, when [Pb(II)] is equal to 0.0398, 0.118 and 6.08 mM, the ratio I1272/I1235 decreases as: 1.40, 1.07, and 0.84. The changes in the relative intensity of these two bands can be attributed to differences in the CAO stretching bands of the ether and ester groups, and suggests that the lower wavenumber band (1235 cm1) corresponds to vibrations of the crown and CA ether groups. The interaction of Pb(II) with the oxygen atoms of these groups increases the dipolar moments of the CAO linkages due to the ion–dipole interactions, which will change the transition probability, and as a consequence, the intensity of the band will increase. However, the non-carbonyl oxygen of the ester group is not expected to interact with lead, as can be seen from the lack of specific membrane-lead interaction
3844
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
Fig. 2. FTIR spectra of CA/MdC6 films in equilibrium with Pb(II) aqueous solutions of different concentrations. (a) [Pb(II)] = 0, (b) [Pb(II)] = 0.0398 mM, (c) [Pb(II)] = 0.118 mM, and (d) [Pb(II)] = 6.08 mM.
with cellulose acetate, as is seen in the CA-lead isotherm (results below). As a consequence, no change in IR band intensity is expected with this group. The sorption isotherm of lead(II) with CA and CA/MdC6 composite films was determined. Five pieces of these films were left in contact over a week at 25 C with 5 ml of 10 lead solutions of different metal ion concentrations covering a range of metal to crown molar ratios varying from 0.1 to 2.0. To study the effect of the MdC6, a blank experiment was also carried out with five pieces of cellulose acetate membranes in contact with five lead solutions with a similar ratio of lead(II) concentration per gram of membrane to the solutions with the films containing MdC6. The Pb(II) concentrations in the supernatant were determined by ICP in properly diluted solutions and the results are shown in Fig. 3. The results of Pb(II) sorption by CA, in the tested Pb(II) concentration range, can be described by a Henry’s law type Eq. (9), where C is the sorbed Pb(II) concentration inside the polymeric matrix, in mol g1 of polymer, c is the initial lead(II) aqueous solution concentration (before sorption process takes place) in mol dm3, and K is the partition coefficient in dm3 g1. C ¼ Kc
ð9Þ
The K value computed from fitting experimental data to Eq. (9) is 0.30 (±0.01) dm3 g1 (correlation coefficient, 0.9938). These results suggest that both
poymer/permeant and permeant/permeant interactions are weak compared with polymer/polymer interactions, and consequently, the sorbed lead(II) is free inside polymeric matrix showing no specific membrane-lead interaction. When MdC6 is incorporated in CA the sorption isotherm changes and cannot be described by Eq. (9). Taking into account the discussion of the previous section, experimental data were treated in terms of dual mode sorption isotherm (Eq. (10)), where K 0 is an equilibrium constant involving the sorption and desorption processes in dm3 mol1, and Cp is the concentration of the sorbed molecules which can interact with the polymer (in mol g1). C ¼ Kc þ
CPK 0 c 1 þ K 0c
ð10Þ
Eq. (10) involves a combination of Henry’s law and Langmuir isotherm equations. This sorption isotherm suggest that some of the sorbed ions are present as a mobile population dissolved in the bulk of the polymer, whilst the remaining ions are bound at a fixed number of adsorption sites within the polymer. It should be noted that application of the Langmuir equation to model experimental data have failed. The fitting parameters of Eq. (10) obtained by a non-linear square minimum deviation method, using OriginPro 7.5 software, are: K = 0.0636 (±0.029) dm3 g1, Cp = 4.0 (±0.7) · 105 mol g1, and K 0 = 16,135 (±9802) dm3 mol1. Two
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
3845
Fig. 3. Lead(II) sorption isotherms of lead(II) in (h) CA and (s) CA/MdC6 (88% of cellulose acetate and 12% of MdC6 (w/w)) films. Solid lines represent the fitting of experimental data to Eqs. (9) and (10) (see Section (3).
different conclusions can be drawn from these values: (a) the partition coefficient have decreased with the presence of MdC6, in comparison with the CA system, showing that the free volume available for dissolution of sorbant species decreased, probably as a consequence of an alteration of polymeric matrix structure, and (b) there is a strong interaction between lead(II) and MdC6. Further evidence for complexation of lead by MdC6 come from SEM analysis, Fig. 4. CA/ MdC6 composite films (Fig. 4b) show a fairly compact and homogeneous surface morphology, showing porous diameter lower than 0.2 lm, in agreement with the low porosity of these films discussed elsewhere [13], and very similar to those found to cellulose acetate (Fig. 4a). However, in CA/MdC6-containing sorbed lead, there are indications of phase-separation, with a clear disruption of surface structure. However, no alterations on surface morphology of CA alone were observed when lead(II) was present, and the membranes showed a rather smooth and featureless surface. The phaseseparation in CA/MdC6-containing sorbed lead can be interpreted by the presence of some lead agglomerate at the surface of the membrane due to the specific bonding properties of the model. These alterations in surface morphology increase in number and area with an increase of initial lead(II) concentration (Figs. 4c and d). In Fig. 4c the large structure is around 10 lm extension whilst in Fig. 4d structures increase to ca. 30 lm extension,
suggesting increased aggregation when lead(II) is sorbed by polymer. It can be seen that sorption of lead(II) by the CA/MdC6 composite films, when in equilibrium with 6.08 mM lead(II) solution, promotes the formation of a finer morphology (Fig. 4e) characterised by an interpenetrating network of microcracks along the surface of the film. Another interesting feature is that some of the surface structures appear to be strongly reflecting, as seen by SEM. This can be justified by the existence of heavy metal Pb(II) encapsulated or surrounded by non-conducting species (crown). This supports the affinity of lead per the crown groups as was suggested from the extraction experiments and indicates that the crown groups show some tendency to be on the membrane surface. Experimental evidence that shows the role of the crown moieties in the interaction with metal ions comes from transport experiments. The effect of the model in the transport of lead(II) was studied and compared with that found to NaCl in CA and in CA/Md6 membranes. In the permeability experiments it is possible to calculate the diffusivities on the basis of the delay time (so-called time-lag, h), Eq. (3), which reflects the unsteady diffusion across a polymer membrane. Time-lag (h) can be associated with the dissolution of the permeant species to a constant level before a steady-state is achieved. In addition, an increase in the induction period will be found with an increase in the interaction between the permeant species and the polymeric matrix [18].
3846
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
Fig. 4. SEM micrographs of cellulose acetate (a) and CA/MdC6 films in equilibrium with Pb(II) aqueous solutions of different concentrations: (b) [Pb(II)] = 0, (c) [Pb(II)] = 0.0398 mM, (d) [Pb(II)] = 0.118 mM, and (e) [Pb(II)] = 6.08 mM. The black zones were burned by the filament.
Fig. 5 shows the flux of salt [NaCl or Pb(NO3)2] through cellulose acetate (CA) and cellulose acetateMdC6 (CA/MdC6) membranes.
From experimental data shown in Fig. 5 the timelag of NaCl in CA and CA/MdC6 films are 3.4 and 34.8 h, respectively, whilst the time-lag of lead
Fig. 5. Concentration of electrolyte that permeates polymeric membrane as a function of time, when an electrolyte concentration gradient of 0.1 M is applied. (h) NaCl, CA; (s) NaCl, CA/MdC6; and (D) Pb(NO3)2, CA/MdC6.
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
nitrate is one order of magnitude greater, 139.8 h. Taking into account the film thicknesses, these data can be ‘‘normalised’’ by using Eq. (3) and values can be calculated of apparent diffusion coefficients (DF), which are 1.7 · 1014, 1.6 · 1015, and 4.1 · 1016 m2 s1, for NaCl in CA and CA/MdC6 films and Pb(NO3)2 in CA/MdC6 films, respectively. From these results we can conclude that: (a) the presence of model has a great effect on the steady-state diffusion process, and (b) the interaction between lead nitrate and the model is much larger than sodium chloride. These interactions seem to decrease the ‘‘free volume’’ available for salt permeation once the permeability coefficients lead nitrate through the composite membrane (3.9 · 1015 m2 s1) is lower than that found to NaCl in both composite (8.5 · 1015 m2 s1) and cellulose acetate (5.6 · 1014 m2 s1). The decreases observed for both DF and P are significant and clearly suggest a transport mechanism involving interactions between electrolytes and the polymer, with particular relevance for MdC6. In order to estimate the amount of electrolyte which is retained by the polymeric matrix, the distribution coefficients, K, were calculated using the equation: K = P/DF. The K values for all systems are 3.4 (NaCl-CA), 5.2 (NaCl-CA/MdC6) and 9.7 [Pb(NO3)2-CA/MdC6]. These values show that interactions between electrolytes and MdC6 occur and some selectivity towards different salts can be quantified. 4. Conclusions Metal ions interact selectively with a novel model polyamide compound bearing an 18crown-6 moiety as pendant structure. The specific interactions with some cations in alkaline, earth alkaline, transition metal and heavy metal cation series lead to selective extraction in liquid–liquid extraction experiments. With the various cation series, K(I), Ba(II), Cr(III), and Hg(II) have been selectively extracted by liquid model polyamide phases. The interaction of a dense composite model polyamide-cellulose acetate membrane with lead(II) has been studied by means of infrared spectroscopy, adsorption isotherms and scanning electron microscopy. For comparison, the transport of sodium chloride through CA and CA/MdC6 membranes was studied as well as that of lead nitrate through CA/MdC6 membranes. All these techniques show the role of the polyamide model in the interaction
3847
with lead(II) and Na(I) ions. From FTIR is concluded that CA/MdC6 membrane–lead interactions increase the intensity of the band at 1235 cm1 attributable to the crown and CA ether groups interacting with lead(II). A mixture of Henry and Langmuir isotherms are necessary to explain the lead adsorption results in the case of CA/MdC6 membranes while only Henry isotherm explain the interaction in CA membranes, showing once again the role of the model in the specific interaction of these CA and CA/MdC6 membrane with lead ions. Similarly, SEM shows a phase separation when lead is adsorbed on CA/MdC6 membranes and not when it is adsorbed on CA ones. In addition, the transport experiments show lower diffusion and permeability of sodium for CA/MdC6 membranes than for CA ones. A stronger interaction of CA/MdC6 with lead compared with sodium ions has been confirmed. The materials and results presented suggest possible applications in decontamination, separation of mixtures of cations, or in the development of fixing sites carrier membranes for selective transport, or for the manufacturing of cation analytical sensors for sensing in UV–Vis absorption or luminescent devices, or in electrochemical application, such as ion-selective electrodes, etc. Acknowledgements Financial support from Accio´n Integrada Hispano-Portuguesa (HP2003-0077, Acc¸ca˜o E/405), Spanish Junta de Castilla y Leo´n, European Union (F.S.E.) (BU003A05) and from Ministerio de Ciencia y Tecnologı´a (FEDER-Plan Nacional de Investigacio´n Cientı´fica, Desarrollo e Innovacio´n Tecnolo´gica, MAT2005-01355) is gratefully acknowledged. HDB, AACCP and AJMV also thank POCTI, FCT and FEDER for financial support. References [1] Pedersen CJ. Cyclic polyethers and their complexes with metal salts. J Am Chem Soc 1967;89(26):7017–36. [2] Pedersen CJ. Cyclic polyethers and their complexes with metal salts. J Am Chem Soc 1967;89(10):2495–6. [3] Drake PL, Price GJ. Crown-ether containing copolymers as selective membranes for quartz crystal microbalance chemical sensors. Polym Int 2000;49(9):926–30. [4] Shinkai S, Takeuchi M, Ikeda A. In: Osada Y, De Rossi DE, editors. Polymer sensors and actuators. Berlin: SpringerVerlag; 2000.
3848
M.J. Tapia et al. / European Polymer Journal 43 (2007) 3838–3848
[5] Sata T, Sata T, Yang W. Studies on cation-exchange membranes having permselectivity between cations in electrodialysis. J Membr Sci 2002;206(1–2):31–60. [6] Watanabe H, Iijima T, Fukuda W, Tomoi M. Synthesis and phase-transfer catalytic activity of novel chiral crown ethers immobilized onto polystyrene supports. React Funct Polym 1998;37(1–3):101–9. [7] Gokel GW. In: Stoddart JF, editor. Crown ethers and cryptands. Monographs in supramolecular chemistry. Cambridge: The Royal Society of Chemistry; 1991. [8] Hiraoka M. Crown compounds their characteristics and applications, studies in organic chemistry, 12. Amsterdam: Elsevier; 1982. [9] Rubio F, Garcı´a F, Burrows HD, Pais AACC, Valente AJM, Tapia MJ, et al. Aqueous solution and solid state interactions of lanthanide ions with a methacrylic ester polymer bearing pendant 15-crown-5 moieties. J Polym Sci Part A: Polym Chem 2007;45(9):1788–99. [10] Tapia MJ, Burrows HD, Garcı´a JM, Garcı´a F, Pais AACC. Lanthanide ion interaction with a crown ether methacrylic polymer, Poly(1,4,7,10-tetraoxacyclododecan-2-ylmethyl methacrylate), as seen by spectroscopic, calorimetric, and theoretical studies. Macromolecules 2004;37(3):856–62. [11] Caldero´n V, Garcı´a F, de la Pen˜a JL, Maya EM, Garcı´a JM. Synthesis and characterization of new aromatic polyamides bearing crown ethers or their dipodal counterparts in the pendant structure. I. Benzo-12-crown-4 and ortho-bis(2ethoxyethoxy)benzene. Polym Sci Part A: Polym Chem 2006;44(7):2270–81. [12] Caldero´n V, Garcı´a F, de la Pen˜a JL, Maya EM, Lozano AE, de la Campa JG, et al. Synthesis and characterization
[13]
[14]
[15] [16]
[17]
[18]
of new aromatic polyamides bearing crown ethers or their dipodal counterparts in the pendant structure. II. Benzo-15crown-5 and ortho-bis[2-(2-ethoxyethoxy)ethoxy]benzene. Polym Sci Part A: Polym Chem 2006;44(13):4063–75. Caldero´n V, Schwarz G, Garcı´a F, Tapia MJ, Valente AJM, Burrows HD, et al. Synthesis and characterization of new aromatic polyamides bearing crown ethers and acyclic ethylene oxide units in the pendant structure. III. Benzo18-crown-6 systems and their open-chain counterparts. Polym Sci Part A: Polym Chem 2006;44(21):6252–69. Valente AJM, Polishchuk AYa, Lobo VMM, Burrows HD. Transport properties of concentrated aqueous sodium dodecyl sulfate solutions in polymer membranes derived from cellulose esters. Langmuir 2000;16(16):6475–9. Khopkar SM. Analytical chemistry of macrocyclic and supramolecular compounds. Berlin: Springer; 2002. Taylor A, Angerer J, Arnaud J, Claeys F, Kristiansen J, Mazarrasa O, et al. Differences in national legislation for the implementation of lead regulations included in the European directive for the protection of the health and safety of workers with occupational exposure to chemical agents (98/24/EC). Int Arch Occ Env Hea 2007;80(3): 254–64. Valente AJM, Burrows HD, Polishchuk AYa, Domingues CP, Borges OMF, Euse´bio MES, et al. Permeation of sodium dodecyl sulfate through polyaniline-modified cellulose acetate membranes. Polymer 2005;46(16):5918–28. Valente AJM, Polishchuk AYa, Burrows HD, Lobo VMM. Permeation of water as a tool for characterizing the effect of solvent, film thickness and water solubility in cellulose acetate membranes. Eur Polym J 2005;41(2):275–81.
