Mechanism of facilitated saccharide transport through plasticized cellulose triacetate membranes

Mechanism of facilitated saccharide transport through plasticized cellulose triacetate membranes

Journal of Membrane Science 194 (2001) 165–175 Mechanism of facilitated saccharide transport through plasticized cellulose triacetate membranes Kimbe...

182KB Sizes 2 Downloads 57 Views

Journal of Membrane Science 194 (2001) 165–175

Mechanism of facilitated saccharide transport through plasticized cellulose triacetate membranes Kimberly M. White a , Bradley D. Smith a,∗ , Peter J. Duggan b , Sarah L. Sheahan b , Edward M. Tyndall b a

b

Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, IN 46556-5670, USA Centre for Green Chemistry and School of Chemistry, P.O. Box 23, Monash University, Clayton, Melbourne, Vic. 3800, Australia

Received 18 October 2000; received in revised form 4 May 2001; accepted 7 May 2001

Abstract Mechanistic insight is gained for saccharide transport through plasticized cellulose triacetate (CTA) membranes containing lipophilic ion-pair transport carriers. The molecular structures of the different membrane components are systematically varied and diagnostic transport characteristics such as saccharide–carrier diffusion constant and saccharide extraction constant are determined. The observed percolation thresholds support a jumping mechanism, however, the diffusion constants are found to decrease as the size of the saccharide, carrier cation, and carrier anion increase, indicating that the rate-limiting step in the transport process involves diffusion of a complex comprised of all three components. The data is reconciled in terms of mobile-site jumping mechanism where the saccharide is relayed along a sequence of ion-pair carriers that are locally mobile. In an attempt to improve saccharide selectivity, calix-[4]-arene dicarboxylates were evaluated as potential ditopic transport carriers. This produced no major change in saccharide extraction constants. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Facilitated transport; Sugar separation; Diffusion; Foods; Composite membranes

1. Introduction Facilitated transport through liquid membranes is an attractive separation method [1,2], but it suffers from several technological problems such as membrane instability and leaching of the membrane components into aqueous phases [3]. In an effort to overcome these problems, a number of research groups with an interest in metal cation separation have investigated plasticized polymeric membranes [4–14]. Our research focus is ∗

Corresponding author. Tel.: +1-219-631-8632; fax: +1-219-631-6652. E-mail address: [email protected] (B.D. Smith).

biomolecule separation [15], and we have examined the transport of small saccharides [16] and amino acids [17] through liquid and plasticized membranes. Facilitated sugar transport through lipophilic membranes is a particularly challenging problem, because very few carrier compounds are known to extract sugars from aqueous solution [15,18,19]. Thus, it was interesting to discover that plasticized cellulose triacetate (CTA) membranes containing large amounts of trioctylmethylylammonium chloride (TOMA-C, Fig. 1) are remarkably stable and selectively permeable to neutral mono- and disaccharides [16]. A fixed-site jumping mechanism was initially proposed, which envisioned the saccharide jumping along a sequence of relatively

0376-7388/01/$ – see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 3 7 6 - 7 3 8 8 ( 0 1 ) 0 0 4 8 7 - 2

166

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

transport characteristics such as saccharide–carrier diffusion constant and saccharide extraction constant. The observed percolation thresholds support a jumping mechanism, however, the diffusion constants are found to decrease as the size of the saccharide, carrier cation, and carrier anion increase, indicating that the rate-limiting step in the transport process involves diffusion of a complex comprising of all three components. The data is reconciled in terms of mobile-site jumping mechanism where the saccharide is relayed along a sequence of ion-pair carriers that are locally mobile.

2. Theory

Fig. 1. Structures and names of the membrane transport carriers.

immobile salt carriers. The saccharides are attracted to the carrier anions by hydrogen bonding interactions (Fig. 2) [20]. Evidence for a jumping mechanism comes from a plot of saccharide flux as a function of TOMA-C concentration [16]. The profile does not exhibit the linear relationship that is reflective of a transport system operating solely by carrier diffusion. Instead, the plot is non-linear with clear evidence for a percolation threshold at around 20% TOMA-C [21]. The aim of this paper is to provide additional mechanistic insight. The experimental approach is to vary the molecular structures of the different membrane components and to measure diagnostic

Saccharide transport through the plasticized membrane is strongly accelerated by the presence of lipophilic tetralkylammonium salts in the membrane. The added salts act as highly active transport carriers which means non-facilitated transport can be ignored in the kinetic analysis. Since the aqueous phases are rapidly stirred at the same constant speed in each experiment, diffusion through the unstirred aqueous boundary layers is assumed to be insignificant compared to diffusion through the membrane. The rates of saccharide–carrier complex formation and dissociation are assumed to be fast compared to saccharide movement through the membrane, so the concentration of the various species at the interfaces are related by Kex , the saccharide extraction constant. A plot of initial saccharide flux versus saccharide source concentration approaches a limiting maximum value at high saccharide source concentration, reflecting saturation of the carrier. This saturation curve can be fitted to the following equation: J =

Fig. 2. Saccharides form monotopic and ditopic complexes with carrier anions.

Dm CKex S , L 1 + Kex S

where J is initial saccharide flux, Dm the diffusion constant for the saccharide–carrier complex, C the carrier concentration, L the membrane thickness, S the saccharide concentration, and Kex the saccharide extraction constant [15]. For high values of S, the flux approaches a maximum value of J max = D m C/L which allows calculation of Dm . Eadie–Hofstee plots (J versus J/S) allow determination of Jmax (x intercept) and Kex (−slope).

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

3. Experimental 3.1. Preparation of plasticized CTA membranes CTA (100 mg), 2-nitrophenyl octyl ether (2-NPOE, 200 mg), and carrier (variable amounts) were dissolved in chloroform (5 ml) and the solution spread evenly in a 9.0 cm diameter glass petri-dish. The dish was loosely covered and the solution allowed to evaporate overnight at room temperature. The membrane was carefully peeled away from the glass to give a transparent film. The membrane thickness was measured using an optical microscope. Although, membrane stability is not an issue in this study which focuses on initial fluxes, a few comments are nonetheless pertinent. Since the plasticized membranes contain large amounts of quaternary ammonium salt carriers, it is not surprising that leaching of the less lipophilic carriers does occur. For example, a plasticized membrane composed of CTA (100 mg), 2-NPOE (200 mg), and TOMA-C (200 mg) was found to lose about 10% of its mass after 1 week of use, but this did not affect saccharide fluxes through this membrane which remained unchanged over the 1-week period. 3.2. Transport experiments The membrane was clamped between two cylindrical, water-jacketed half cells (volume 60 ml) that have been described before [22]. The side exposed to the air during membrane preparation always faced towards the source phase. Both source and receiving phases contained sodium phosphate buffer (60 ml, 0.1 M, pH 7.3) and the source phase also included saccharide. Unless indicated, the temperature was held constant at 25◦ C. Both aqueous phases were stirred at a constant 500 rpm using magnetic stir bars and external stir plates. Initial fluxes were determined by monitoring the change in saccharide concentration in the receiving phase. Control experiments showed that flux was inversely proportional to membrane thickness, so differences in membrane thickness were corrected by multiplying observed fluxes by (membrane thickness in presence of carrier/membrane thickness in absence of carrier). All transport experiments were run in duplicate and the measured fluxes were always within ±15%.

167

Saccharide concentrations were determined using the following colorimetric enzyme assays [22]. Glucose: a solution of hexokinase/glucose-6-phosphate dehydrogenase in 75 mM sodium phosphate buffer with 2 mM MgCl2 , pH 7.5 was added to an aliquot of receiving phase (0.5 ml) along with 1 mM ATP (5 ␮l) and 0.5 mM NADP (5 ␮l), prepared in 0.1 M sodium phosphate buffer, pH 7.5. The concentration of NADPH, which equals the concentration of glucose, was determined by UV absorption at 340 nm (ε = 6.23 × 103 M−1 cm−1 ) [23]. Sucrose: the procedure is the same as for glucose, except invertase and phosphoglucose isomerase were added to the solution. 3.3. Preparation of carrier salts Tridodecylmethylammonium chloride (TDMA-C) and Aliquat® 336, which is predominantly TOMA-C, were purchased from Aldrich. Mass spectral analysis of the Aliquat showed the presence of C10 chains. That is, the cation is a mixture of trioctylmethylammonium (31%), decyldioctylmethylammonium (46%), and didecyloctylmethylammonium (23%). This does not effect the interpretation of results. TOMA-DHP and TDMA-DHP were prepared in the following way. The appropriate phosphate was dissolved with NaOH (one equivalent) in H2 O (10 ml). Trioctylmethylammonium chloride (Aliquat® 336, one equivalent) was dissolved in CHCl3 (10 ml). The two solutions were vigorously mixed and after phase separation, the organic layer was washed with H2 O (2 × 20 ml), dried over magnesium sulfate, and the solvent removed under vacuum. The resulting samples of TOMA-DHP and TDMA-DHP were shown to be pure by 1 H NMR, mass spectroscopy, and combustion analysis. TOMA-TBC was prepared in three steps: 4-t-butylphenol (2.50 g, 0.017 mol) was refluxed overnight with one equivalent of t-butyl bromoacetate (2.69 ml) and 10 equivalents potassium carbonate (22.9 g) in acetone (50 ml). The solvent was removed under vacuum. The residual clear liquid was dissolved in ethyl acetate (10 ml) and washed with NaOH (2 × 10 ml) and H2 O (2 × 10 ml), then dried over magnesium sulfate. Removal of the solvent under vacuum gave a yellow liquid (4.18 g, 95%): 1 H NMR (CDCl3 ): δ 7.30 (2H, d, J = 9 Hz), 6.84 (2H, d, J = 9 Hz), 4.49 (2H, s), 1.50 (9H, s), 1.30 (9H, s). 13 C NMR (CDCl3 ): δ

168

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

168.20, 155.63, 144.10, 126.21, 114.01, 82.13, 65.77, 34.04, 31.45, 28.01. MS (FAB in NBA matrix); m/z 264 (M +H)+ . HRMS calcd for C16 H24 O3 : 264.1725; found: 264.1715. Trifluoroacetic acid (10 ml) was added to the above product (4.18 g) in chloroform (10 ml). One drop of anisole was added and the solution was stirred at room temperature for 6 h. Solvent removal under vacuum gave a grey solid. This solid was redissolved in hot chloroform, and cold petroleum spirit was added until a white precipitate formed. Suction filtration gave white crystals (1.54 g, 47%) mp 88–89◦ C (literature: 84–85◦ C [22]): 1 H NMR (CDCl3 ): δ 7.33 (2H, d, J = 9 Hz), 6.86 (2H, d, J = 9 Hz), 4.66 (2H, s), 1.30 (9H, s). 13 C NMR (CDCl3 ): δ 174.91, 155.07, 144.90, 126.47, 114.09, 64.91, 34.13, 31.43. MS (FAB matrix); m/z 208 (M + H)+ . HRMS calcd for C12 H16 O3 : 208.1103; found: 208.1099. This acid was converted into TOMA-TBC by the same method as that described for TOMA-DHP. TOMA-TOC was also prepared in three steps: 4-t-octylphenol (5.00 g, 0.024 mol) was refluxed overnight with one equivalent of t-butyl bromoacetate (3.91 ml) and 10 equivalents potassium carbonate (33.2 g) in acetone (100 ml). The solvent was removed under vacuum. The residual clear liquid was dissolved in ethyl acetate (10 ml) and washed with NaOH (2 × 10 ml) and H2 O (2 × 10 ml), then dried over magnesium sulfate. Removal of the solvent under vacuum gave a clear liquid (6.61 g, 86%): 1 H NMR (CDCl3 ): δ 7.26 (2H, d, J = 9 Hz), 6.81 (2H, d, J = 9 Hz), 4.49 (2H, s), 1.696 (2H, s), 1.47 (9H, s), 1.34 (6H, s), 0.70 (9H, s). 13 C NMR (75 MHz, CDCl3 ): δ 168.29, 155.56, 142.97, 127.05, 113.81, 82.12, 65.85, 56.99, 37.94, 32.29, 31.75, 31.63, 28.00. MS (FAB in NBA matrix); m/z 320 (M + H)+ . HRMS calcd for C20 H32 O3 : 320.2351; found: 320.2328. Trifluoroacetic acid (10 ml) was added to the above product (6.61 g) in chloroform (30 ml). One drop of anisole was added and the solution was stirred at room temperature for 6 h. Removal of the solvent under vacuum gave a grey solid. The solid was redissolved in hot chloroform and cold petroleum spirit was added until a white precipitate formed. Suction filtration gave white crystals (2.24 g, 41%, mp 107–108◦ C): 1 H NMR (CDCl3 ): δ 7.30 (2H, d, J = 8.7 Hz), 6.84 (2H, d, J = 8.7 Hz), 4.66 (2H, s), 1.70 (2H, s), 1.34 (6H, s), 0.70 (9H, s). 13 C NMR (CDCl3 ): δ 174.30, 155.06, 143.92, 127.33, 113.91, 65.02, 57.01, 38.08, 32.36, 31.79,

31.63. MS (FAB in NBA matrix); m/z 264 (M + H)+ . HRMS calcd for C16 H24 O3 : 264.1725; found: 264.1731. This acid was converted into TOMA-TBC by the same method as that described for TOMA-DHP. TOMA-HCC and TOMA-MCC were also prepared from the corresponding carboxylic acids by exchange with Aliquat® 336. The diphenol-diacid precursor to TOMA-HCC was prepared by the method of McKervey [24]: mp > 151◦ C (decomp.) (literature: mp > 220◦ C (decomp.) [25]): 1 H NMR (CDCl3 ): δ 7.05 (2H, s), 6.80 (2H, s), 4.62 (2H, s), 4.13 (2H, d, J = 13.5 Hz), 3.31 (2H, d, J = 13.5 Hz), 1.29 (9H, s), 0.95 (9H, s). 13 C NMR (CDCl3 ): δ 170.90, 149.62, 149.41, 147.87, 142.46, 132.68, 127.38, 126.05, 72.77, 34.08, 33.84, 32.19, 31.68, 31.57. MS (FAB in NBA matrix); m/z 764 (M + H)+ . HRMS calcd for C48 H60 O8 : 764.4288; found: 764.4280. The dimethoxy-diacid precursor to TOMA-MCC was prepared by basic hydrolysis of the corresponding diethyl ester [26], also following a method of McKervey [24]: mp 230–239◦ C (literature: 175–176◦ C [22]): 1 H NMR (CDCl3 ): δ 7.18 (2H, s), 6.61 (2H, s), 4.67 (2H, s), 4.20 (2H, d, J = 13.0 Hz), 3.83 (3H, s), 3.32 (2H, d, J = 13.0 Hz), 1.34 (9H, s) 0.84 (9H, s). 13 C NMR (CDCl3 ): δ 169.7, 152.9, 150.6, 147.6, 146.4, 134.5, 131.8, 126.2, 125.5, 72.1, 64.0, 34.5, 34.0, 31.8, 31.3, 31.2. MS (FAB in CH3 CN/H2 O+3% formic acid); m/z 793.5 (M +H)+ .

4. Results and discussion 4.1. Membrane preparation and transport measurements The plasticized membranes were prepared by evaporating chloroform solutions of polymer, plasticizer, and salt carrier [4]. In all cases, CTA was used as the polymer support and 2-NPOE as the plasticizer. The resulting films were clamped between two water-jacketed half cells containing 0.1 M sodium phosphate buffer (pH 7.3) with glucose or sucrose in the source phase. Initial saccharide fluxes into the receiving phase were determined by standard colorimetric enzyme assays [23] and corrected for differences in membrane thickness. Negligible transport

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

Fig. 3. Flux vs. membrane thickness for sucrose transport. Each stacked membrane was composed of 100 mg CTA, 200 mg 2-NPOE, 200 mg TOMA-C, and was 50 ␮m thick. Source phase: 2.0 M sucrose, 0.1 M sodium phosphate, pH 7.3. Receiving phase: 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

was observed using control membranes that did not contain carrier. 4.2. Membrane diffusion is the rate-determining step Sucrose flux was measured as a function of membrane thickness. This experiment was performed by stacking multiple membranes together. As shown in Fig. 3, flux decreased linearly with membrane thickness. This is unambiguous evidence that the slow step in the transport process is migration through the membrane and not decomplexation from the carrier [15,27]. In this study, membrane diffusion constants were determined by analyzing transport saturation curves

169

as described in Section 2. This approach was validated in one case by measuring the membrane diffusion constant by an independent method, namely lag time, which is the time for the saccharide–carrier complex to diffuse across the membrane from the source to the receiving phase [28]. The lag time is determined experimentally by initiating a transport experiment and measuring the time for sucrose to appear in the receiving phase. A lag time, tlag , of 340 ± 10 s was measured for sucrose transport through a membrane composed of 100 mg CTA, 200 mg 2-NPOE, and 200 mg TOMA-C (membrane thickness 50 ␮m, and each aqueous phase containing 0.1 M Na2 H2 PO4 buffer at pH 7.3). Using the relationship D m = L2 /6tlag , this corresponds to D m = (1.2 ± 0.3) × 10−12 m2 s which is within error of the value of (0.7 ± 0.2) × 10−12 m2 s obtained by analyzing the transport saturation curve (Table 1). 4.3. Percolation threshold A plot of sucrose flux and wt.% TOMA-C (Fig. 4) does not exhibit the linear relationship that is reflective of a transport system operating solely by carrier diffusion. Instead, the plot is non-linear with clear evidence for a percolation threshold at about 17% TOMA-C, a profile that is consistent with a saccharide jumping transport mechanism [21]. The percolation threshold of 17% TOMA-C for sucrose transport is slightly lower than the 20% TOMA-C previously found for fructose transport. This result indicates that the larger disaccharide does not require the carrier molecules to be as close together for transport to occur.

Table 1 Extraction constant (Kex ), maximum flux (Jmax ), and diffusion constant (Dm ) for saccharide transport through plasticized CTA membranesa Carrier

Saccharide

Kex (M−1 )b

Jmax (×10−8 mol m−2 s−1 )c

Dm (×10−14 m2 s−1 )d

TOMA-C TOMA-C TDMA-C TDMA-C TOMA-DBP TOMA-DBP TOMA-DHP TOMA-DHP

Glucose Sucrose Glucose Sucrose Glucose Sucrose Glucose Sucrose

0.2 0.5 0.8 0.3 0.5 0.3 0.3 0.3

18800 2600 1600 500 2500 1800 300 80

600 70 60 25 100 60 15 5

a

Values defined and determined as described in Section 2; each membrane composed of 100 mg CTA, 200 mg 2-NPOE, and 200 mg carrier, 50 ␮m thick; receiving phase: 0.1 M sodium phosphate, pH 7.3, T = 25◦ C. b ±30%. c ±10%. d ±30%.

170

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

Fig. 4. Sucrose flux vs. TOMA-C (wt.%). Membrane also composed of 100 mg CTA, 200 mg 2-NPOE. Source phase: 0.3 M sucrose, 0.1 M sodium phosphate, pH 7.3. Receiving phase: 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

Fig. 6. Glucose flux vs. TOMA-DBP (wt.%). Membrane also composed of 100 mg CTA, 200 mg 2-NPOE. Source phase: 0.3 M glucose, 0.1 M sodium phosphate, pH 7.3. Receiving phase: 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

Percolation thresholds were also observed for glucose transport through membranes containing the carriers TDMA-C (Fig. 5), TOMA-DBP (Fig. 6), and TOMA-DHP (Fig. 7), as well as the carboxylate carriers TOMA-TBC, TOMA-TOC, TOMA-HCC, and TOMA-MCC (data not shown). In most cases, saccharide flux does not immediately drop to zero when the carrier concentration is below the percolation threshold. As discussed previously [17], this is attributed to a residual low level of facilitated transport by a classical saccharide–carrier diffusion mechanism (see Fig. 12a).

4.4. Transport as a function of saccharide size

Fig. 5. Glucose flux vs. TDMA-C (wt.%). Membrane also composed of 100 mg CTA, 200 mg 2-NPOE. Source phase: 0.3 M glucose, 0.1 M sodium phosphate, pH 7.3. Receiving phase: 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

Fig. 7. Glucose flux vs. TOMA-DHP (wt.%). Membrane also composed of 100 mg CTA, 200 mg 2-NPOE. Source phase: 0.3 M glucose, 0.1 M sodium phosphate, pH 7.3. Receiving phase: 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

Saturation curves for the transport of glucose, a monosaccharide, and sucrose, a disaccharide, through plasticized CTA membranes containing phosphate and chloride carriers are shown in Figs. 8–11. The derived diffusion and extraction constants are listed in Table 1. Inspection of Table 1 reveals little difference in saccharide extraction constants, but in all cases the sucrose diffusion constants are lower than the corresponding glucose diffusion constants. A similar trend is observed with the carboxylate carriers

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

Fig. 8. Saturation curve for the transport of glucose and sucrose through a CTA membrane containing TOMA-C. Membrane composed of 100 mg CTA, 200 mg 2-NPOE, 200 mg TOMA-C and was 50 ␮m thick. Aqueous phases, 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

(Table 2). These results are consistent with a transport process that involves saccharide diffusion in the rate-determining step. 4.5. Transport as a function of carrier cation size The saccharide transport induced by TOMA-C was compared to that induced by the larger TDMA-C (Fig. 1). Both membrane systems exhibited percolation

Fig. 9. Saturation curve for the transport of glucose and sucrose through a CTA membrane containing TDMA-C. Membrane composed of 100 mg CTA, 200 mg 2-NPOE, 200 mg TDMA-C and was 50 ␮m thick. Aqueous phases, 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

171

Fig. 10. Saturation curve for the transport of glucose and sucrose through a CTA membrane containing TOMA-DBP. Membrane composed of 100 mg CTA, 200 mg 2-NPOE, 200 mg TOMA-DBP and was 50 ␮m thick. Aqueous phases, 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

thresholds (Figs. 4 and 5). Observation of saturation curves (Figs. 8 and 9) for glucose and sucrose transport indicate a carrier-mediated transport mechanism. The diffusion constants derived from these saturation curves (Table 1) show that saccharide diffusion through a TOMA-C membrane is faster than a TDMA-C membrane. In other words, saccharide diffusion is faster with the smaller carrier cation. This implies that the rate-determining transport step must

Fig. 11. Saturation curve for the transport of glucose and sucrose through a CTA membrane containing TOMA-DHP. Membrane composed of 100 mg CTA, 200 mg 2-NPOE, 200 mg TOMA-DHP and was 50 ␮m thick. Aqueous phases, 0.1 M sodium phosphate, pH 7.3, T = 25◦ C.

172

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

Table 2 Extraction constant (Kex ), maximum flux (Jmax ), and diffusion constant (Dm ) for saccharide transport through plasticized CTA membranes containing carboxylate anionsa Carrier

Saccharide

Kex (M−1 )b

Jmax (×10−8 mol m−2 s−1 )c

Dm (×10−14 m2 s−1 )d

TOMA-TBC TOMA-TBC TOMA-TOC TOMA-TOC TOMA-HCC TOMA-HCC TOMA-MCC

Glucose Sucrose Glucose Sucrose Glucose Sucrose Glucose

0.3 0.2 0.3 0.3 0.03 0.06 0.5

1080 680 340 240 1060 120 24

50 30 17 12 100 10 5

a

Values defined and determined as described in Section 2; each membrane composed of 100 mg CTA, 200 mg 2-NPOE, and 200 mg carrier, ranging from 32 to 72 ␮m in thickness; receiving phase: 0.1 M sodium phosphate, pH 7.3, T = 25◦ C. b ±30%. c ±10%. d ±30%.

involve diffusion of the carrier tetralkylammonium cations, which means that the cations are not fixed sites in the membrane. 4.6. Transport as a function of carrier anion size The transport ability of TOMA-DBP was compared to TOMA-DHP, which has a much larger phosphate diester anion (Fig. 1). Once again, saturation curves were obtained with each membrane system (Figs. 10 and 11), and the diffusion constants for membranes containing TOMA-DBP were lower than TOMA-DHP (Table 1). This shows that saccharide diffusion is faster when the carrier anion is smaller. A similar result was obtained for the carboxylate carriers TOMA-TBC and TOMA-TOC (Table 2). These results imply that the rate-determining transport step must involve diffusion of the carrier anions, which means that the anions are not fixed sites in the membrane. 4.7. Transport using ditopic carrier anions One of our research goals is to invent saccharideselective membranes which, in this case, means the development of saccharide transport carriers that select between glucose and sucrose. One approach is to use ditopic transport carriers that form size-selective complexes with the saccharides (Fig. 2). With this goal in mind, we evaluated saccharide transport promoted by the calix-[4]-arene dicarboxylates TOMA-HCC and TOMA-MCC (Fig. 1).

The glucose and sucrose extraction constants for TOMA-HCC (Table 2) are extremely low, most likely because the carboxylates in HCC form intramolecular hydrogen bonds with neighboring phenolic hydroxyls, which inhibit intermolecular interactions with the saccharides. A similar effect has been reported by Anslyn and coworkers [29]. This proposal is supported by the observation that Kex for glucose transport using the carrier, TOMA-MCC, which has its phenols capped with methyl groups, is an order of magnitude higher (Table 2). In terms of transport selectivity, the Kex for sucrose transport using TOMA-HCC is approximately double that of glucose (Table 2), which is a similar ratio to that observed with some of the monotopic carriers such as TOMA-C (Table 1). Thus, the calix-[4]-arene dicarboxylates produce no major change in saccharide selectivity. However, it is worth noting that Dm for glucose transport using TOMA-HCC is remarkably high (Table 2). One explanation is that the jumping pathway contributes more in this transport system, another possibility is that there is a change in the rate-determining step in the transport process. One of our future goals is to test these hypotheses.

5. Membrane structure and polarity A number of observations indicate that the difference in carrier transport abilities is not due to macroscopic differences in membrane structure. For

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

173

a start, each membrane has the same physical appearance of a clear transparent plastic film. FT-IR and mass balance measurements show that each membrane absorbs a small amount of water when exposed to aqueous phases (<5 wt.%). However, this does not result in a substantial change in membrane polarity as judged by measurements using Reichardt’s dye [27]. The λmax for this solvatochromatic dye is strongly dependent on solvent polarity. For example, λmax is 795, 622, and 515 nm in 1,4-dioxane, acetonitrile, and methanol, respectively [30]. To measure membrane polarity, 1 mg of Reichardt’s dye was incorporated into a series of standard plasticized membranes composed of 100 mg CTA, 200 mg of 2-NPOE, and 100 mg of carrier. Using a photospectrometer, the visible region of each dry membrane was scanned and λmax recorded. The membranes were then soaked in 0.1 M NaH2 PO4 buffer, pH 7.3, for 3 h (the length of a standard transport experiment) and the new λmax recorded. In all cases, the measured λmax values remained within a narrow range of 625–644 nm, which is close to the value obtained in neat 2-NPOE.

6. Conclusions Observation of a percolation threshold with each carrier indicates a jumping mechanism [21,31]. However, the saccharide diffusion constants decrease with increasing size of the saccharide, carrier cation, and carrier anion (Table 1), suggesting that the ion-pair carrier does not remain as a “fixed site” within the membrane. This data does not agree with the four transport mechanisms presented in Fig. 12. The first two (Fig. 12a and b) are discounted because they do not rationalize the observed carrier percolation thresholds, the next two (Fig. 12c and d) are discounted because they predict transport flux to be independent of the mobility of one or both of the carrier ions. The most consistent explanation is a mobile-site jumping mechanism (Fig. 13). The saccharides form hydrogen-bonded complexes with the carrier anions (Fig. 2), which in turn are weakly associated with their counter-cations. It is not clear whether the saccharide jumps from anion to anion or if the saccharide–anion complex moves from cation to cation, but the three-component saccharide–ion-pair complex is locally mobile. When the complex diffuses

Fig. 12. Unlikely saccharide transport mechanisms. The first two (a and b) are discounted because they do not predict carrier percolation thresholds, the next two (c and d) are discounted because they predict transport flux to be independent of the mobility of one or both of the carrier ions.

close enough to an unoccupied carrier ion-pair, the saccharide or saccharide–anion complex may “jump” across, and continue the transport process through the membrane.

Fig. 13. Proposed mobile-site jumping mechanism for saccharide transport mediated by ion-pair carriers. The saccharide “jumps” from ion-pair to ion-pair, and/or the saccharide–anion complex “jumps” from cation to cation, but the three-component saccharide–carrier complex is locally mobile.

174

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175

Acknowledgements This work was supported by the Herman Frasch Foundation of the American Chemical Society, the Australian Department of Industry, Science and Tourism, Monash University and the Australian Research Council.

[14]

References

[16]

[1] J. De Gyves, E.R. De San Miguel, Metal ion separations by supported liquid membranes, Ind. Eng. Chem. Res. 38 (1999) 2182–2202. [2] A.M. Sastre, A. Kumar, J.P. Shukla, R.K. Singh, Improved techniques in liquid membrane separations: an overview, Sep. Purif. Methods 27 (1998) 213–298. [3] A.M. Neplenbroek, D. Bargeman, C.A. Smolders, Supported liquid membranes: instability effects, J. Membr. Sci. 67 (1992) 121–133. [4] M. Sugiura, M. Kikkawa, Carrier-mediated transport of rare earth ions through cellulose triacetate membranes, J. Membr. Sci. 42 (1989) 47–55. [5] A.M. Neplenbroek, D. Bargeman, C.A. Smolders, Supported liquid membranes: stabilization by gelation, J. Membr. Sci. 67 (1992) 149–164. [6] T. Hayashita, M. Kumazawa, M. Yamamoto, Selective permeation of cadmium(II) chloride complex through cellulose triacetate plasticizer membrane containing triocylmethylammonium chloride carrier, Chem. Lett. (1996) 37–38. [7] A.J. Schow, R.T. Peterson, J.D. Lamb, Polymer inclusion membranes containing macrocyclic carriers for use in cation separations, J. Membr. Sci. 111 (1996) 291–295. [8] J.D. Lamb, A.Y. Nazarenko, Lead(II) ion sorption and transport using polymer inclusion membranes containing tri-octylphosphine oxide, J. Membr. Sci. 134 (1997) 255–259. [9] G. Arena, A. Contino, A. Magri, D. Sciotto, J.D. Lamb, Selective transport of cesium and strontium ions through polymer inclusion membranes containing calixarenes as carriers, Supramol. Chem. 10 (1998) 5–15. [10] M.F. Paugam, J.J. Buffle, Comparison of carrier-facilitated copper(II) ion transport in a supported liquid membrane and in a plasticized cellulose triacetate membrane, J. Membr. Sci. 147 (1998) 207–215. [11] K.L. Thunhorst, R.D. Noble, C.N. Bowman, Properties of the transport of alkali metal salts through polymeric membranes containing benzo-18-crown-6 crown ether functional groups, J. Membr. Sci. 156 (1999) 293–302. [12] M. Oleinikova, M. Munoz, Transport of rare earth metal ions through activated composite membranes containing DEHPA, Sol. Extract. Ion Exchange 18 (2000) 401–419. [13] T.G. Levitska, D.M. Macdonald, J.D. Lamb, B.A. Moyer, Predication of the carrier-mediated cation flux through polymer inclusion membranes via fundamental thermodynamic quantities: complexation study of bis(dodecyloxy)calix(4)

[15]

[17]

[18]

[19]

[20]

[21] [22]

[23]

[24]

[25]

[26]

[27]

arene-crown-6 with alkali metal cations, Phys. Chem. Chem. Phys. 2 (2000) 1481–1491. S.C. Lee, J.D. Lamb, M.H. Cho, C.H. Rhee, J.S. Kim, A lipophilic acyclic polyether dicarboxylic acid as Pb(II) carrier in polymer inclusion and bulk liquid membranes, Sep. Sci. Technol. 35 (2000) 767–778. S.J. Gardiner, B.D. Smith, Facilitated transport of small hydrophilic biomolecules through artificial membranes, in: G.W. Gokel (Ed.), Advances in Supramolecular Chemistry, Vol. 5, JAI Press, New York, 1999 (Chapter 4). J.A. Riggs, B.D. Smith, Facilitated transport of small carbohydrates through plasticized cellulose triacetate membranes. Evidence for fixed-site jumping transport mechanism, J. Am. Chem. Soc. 119 (1997) 2765–2766. T.A. Munro, B.D. Smith, Facilitated transport of amino acids by fixed-site jumping, Chem. Commun. (1997) 2167– 2168. M. Igawa, T. Sekimoto, H. Okochi, Facilitated transport of carbohydrates via complexation with borate ion fixed on an ion-exchange membrane, J. Membr. Sci. 98 (1995) 177–180. M. Barboiu, C. Guizzard, N. Hovnanian, J. Palmeri, C. Reibel, L. Cot, C. Luca, Facilitated transport of organics of biological interest. I. A new alternative for the separation of amino acids by fixed-site crown-ether polysiloxane membranes, J. Membr. Sci. 172 (2000) 91–103. J.M. Coteron, F. Hacket, H.J. Schneider, Interactions of hydroxy compounds and sugars with anions, J. Org. Chem. 61 (1996) 1429–1435. R.D. Noble, Facilitated transport mechanism in fixed-site carrier membranes, J. Membr. Sci. 60 (1991) 297–306. M. Di Luccio, B.D. Smith, T. Kida, C.P. Borges, T.L.M. Alves, Separation of fructose from a mixture of sugars using supported liquid membranes, J. Membr. Sci. 174 (2000) 217–224. P.R. Westmark, S.J. Gardiner, B.D. Smith, Selective monosaccharide transport through lipid bilayers using boronic acid carriers, J. Am. Chem. Soc. 118 (1996) 11093–11100. E.M. Collins, M.A. McKervey, E. Madigan, M.B. Moran, M. Owens, G. Ferguson, S.J. Harris, Chemically modified calix[4]-arene. Regioselective synthesis of 1,3-(distal) derivatives and related compounds. X-ray crystal structure of a diphenoldinitrile, J. Chem. Soc., Perkin Trans. 1 (1991) 3137–3142. J. Soedarsono, M. Bugard, Z. Asfari, J. Vicens, Liquid–liquid extraction of sodium and potassium using 25,27-dicarboxy26,28-dimethoxy-5,11,17,23-tetra-tert-butylcalix[4]arene, New J. Chem. 17 (1993) 601–605. P.J. Duggan, S.L. Sheahan, M.L. Szydzik, Crystal and solution structures of 25,27-diethoxycarbonylmethoxy-26,28dimethoxy-t-butylcalix[4]arene, Tetrahedron Lett. 41 (2000) 3165–3168. F. de Jong, H.C. Visser, Carrier-mediated transport through supported liquid membranes, in: J.L. Atwood, J.E.D. Davies, D.D. MacNicol, F Vögtle (Eds.), Comprehensive Supramolecular Chemistry, Vol. 10, Pergamon Press, New York, 1996 (Chapter 1).

K.M. White et al. / Journal of Membrane Science 194 (2001) 165–175 [28] L. Bromberg, G. Levin, O. Kedem, Transport of metals through gelled supported liquid membranes containing carrier, J. Membr. Sci. 71 (1992) 41–47. [29] C.-H. Huang, L.A. Cabell, E.V. Anslyn, Molecular recognition of cyclitols by neutral polyaza-hydrogen-bonding receptors: the strength and influence of intramolecular hydrogen bonds between vicinal alcohols, J. Am. Chem. Soc. 116 (1994) 2778–2792.

175

[30] C. Reichardt, E. Harbusch, R. Muller, Pyridinium-Nphenoxide betaine dyes as solvent polarity indicators. Some new findings, in: I. Bertini, A. Lunazzi, A. Dei (Eds.), Advances in Solution Chemistry, Vol. I, Plenum Press, New York, 1981. [31] E.L. Cussler, A. Rutherford, A. Brown, On the limits of facilitated diffusion, J. Membr. Sci. 43 (1989) 149–164.