Swelling/deswelling of anionic copolymer gels

Biomaterids 16 (1995) 559-567

0 1995 Elsevier Science Limited Printed in Great Britain. All rights reserved

0142-9612/95/$10.00

Swelling/deswelling of anionic copolymer gels

Atul R. Khare* and Nikolaos A. Peppas School of Chemical Engineering, Purdue University, West Lafayette, IN 47907-1283, USA

Studies of dynamic and equilibrium swelling of ionic gels are important in understanding the diffusion

of physiologically important fluids in materials for site-specific controlled drug delivery applications.

The dynamic and equilibrium swelling properties of dry glassy poly(2-hydroxyethyl methacrylate-co-

methacrylic acid) and poly(2-hydroxyethyl methacrylate-co-acrylic acid) polymeric networks were

studied as a function of pH, ionic strength, nature of the counterion and buffer composition. The

mechanism of water diffusion in these gels became more anomalous as the pH of the swelling medium

increased and as the ionic strength decreased at a constant pH > PK,,~,,. The mechanism of water

diffusion was Fickian in all unbuffered swelling media at pH 4.0, which is lower than the PK~,~~,. The

pK,,,,, of these gels was between 5.5 and 6. At pH 4.0, the diffusion mechanism was independent of

ionic strength. This swelling behaviour is explained in terms of the concept of ion osmotic swelling

pressure and ion exchange kinetics.

Keywords: pH-sensitive hydrogels, Donnan equilibrium, ion exchange, polyelectrolytes, swelling

Received 29 June 1994; accepted 15 August 1994

Cross-linked polymeric networks are used for a variety of applications such as contact lenses, wound dressings, absorbents, monolithic drug delivery systems, membrane materials and chromatographic packing materials’. Polymeric networks containing ionic and/or hydrophobic moieties show a sudden or gradual change in their dynamic and equilibrium swelling properties as the external environmental conditions are changed. These environmental conditions include pHzs3, ionic strength4, solvent composition5, buffer composition, temperature6-8, pressureg, electromagnetic radiation” and/or photoelectric stimulus”. The dynamic swelling change can be used in the design of intelligent controlled drug release devices for site-specific drug delivery, or in design and analysis of artificial muscles and biosensors. Here, the change in the external environment will act as a stimulus whereas the response is the change in swelling properties of the network.

For example, in the case of anionic polymeric networks containing carboxylic or sulphonic acid groups, ionization takes place as the pH of the external swelling medium rises above the pK, of that ionizable moiety. The polymeric network becomes more hydrophilic as the degree of ionization increases. The opposite behaviour is observed in the case of cationic polymeric networks containing amine groups as shown in Figure 1. Most of the proteins and

Correspondence to Professor N.A. Peppas. *Present address: Materials & Membrane Technology Center, Baxter Healthcare Corporation, Round Lake, IL 60073, USA.

poly(amino acids) are ampholytic in nature, i.e. they contain both amine and carboxylic acid groups. Their dynamic and equilibrium swelling behaviour is more complicated because of the presence of both anionic and cationic groups.

The swelling behaviour of any polymer network depends upon the nature of the polymer, the polymer- solvent compatibility and the degree of cross-linking. The polymer elasticity combined with polymer- solvent mixing contributes to the overall swelling processl’. In the case of ionic networks, in addition to these factors, the ionization of fixed charges contributes to the swelling process. Therefore, the ionization equilibrium, the nature of the buffer composition and the nature of the counterion play a very important role in determining the swelling behaviour13. This informa- tion, although presented here for synthetic polymers, is equally important for naturally gelled macromolecular systems where charges and the ionization equilibrium control the solute permeability across such systems.

The kinetic swelling behaviour of ionic networks depends upon mass transfer limitations, Donnan equilibrium considerations, ion exchange and ionic interac- tions14. These factors are very important in the design and development of controlled drug release devices, which can come into contact with biological fluid having different ionic species present. The swelling behaviour of the network is dependent upon the nature of the surrounding biological fluid. This swelling in turn determines the diffusion coefficient of the solute, which is incorporated inside the network.

559 Biomaterials 1995, Vol. 16 No. 7

560 Swellingldeswelling of anionic copolymer gels: A.R. Khare and N.A. Peppas

F ._ = f

---

‘1 r--

Gels containing

SOaH, -COOH groups

PH

Figure 1 Equilibrium degree of swelling of ionic gels vs. PH.

The ionic contribution to the overall swelling process of ionic hydrogels can be described in terms of the ion osmotic swelling pressure15. For a weakly charged polymer network in a dilute electrolytic solution, it is given by:

Ton = RT C (Ci - Cr) (1)

Here, nion is the ion osmotic swelling pressure, Ci and C: are the concentrations of counterions inside and outside the gel, respectively, R is a general gas constant and T is the absolute temperature.

Swelling kinetics have also been reported for a number of cationic hydrogels as a function of pH, ionic strength and electrolyte composition’6B17. Similar to the anionic hydrogels’8-z0, cationic hydrogels can be used in the development of delivery systems or biomedical devices.

In this contribution, the dynamic and equilibrium swelling behaviour of anionic hydrogels are described as a function of pH, ionic strength and buffer composition of the swelling medium. Emphasis is given to the molecular reasons for the swelling behaviour in terms of ionization equilibrium considerations, the electroneutrality condition and ion osmotic swelling pressure.

EXPERIMENTAL DETAILS

Materials

2-Hydroxyethyl methacrylate (HEMA, Polysciences, Warrington, PA, USA) and methacrylic acid (MAA, Aldrich Chemical Co., Milwaukee, WI, USA) were vacuum distilled at 65” C/5 mmHg and 50” C/

50 mm Hg, respectively. In HEMA vacuum distillation, hydroquinone was added to prevent polymerization. Acrylic acid (AA, Aldrich, Chemical Co.), sodium chloride (Aldrich Chemical Co.), and sodium acetate (Aldrich Chemical Co.) were of analytical grade and were used as-received. Water used in all the studies was double-distilled and deionized with a pH x 7.00.

Synthesis of polymers

HEMA was added to appropriate quantities of MAA or AA to give 5g of mixture. Then 0.5 wt% ammonium persulphate and sodium bisulphite were added as redox initiators and 0.45mol.% ethylene

glycol dimethacrylate was used as a cross-linking agent. Solution polymerization was carried out in polyethylene vials using distilled deionized water (50wt%) as a solvent (to ensure absence of ions), in a constant temperature water bath at 37” C for 24 h, followed by gradual cooling to room temperature. The vials were cut away from the polymer gel. Swollen polymer gels were dried at room temperature under a hood and discs were cut from the dried glassy polymer cylinders with a lathe. These samples were further dried in a vacuum oven for a week at 40” C. The final samples had a diameter of lo-12 mm and thickness ranging from 0.8 to l.Omm. This range of dimensions (aspect ratio of more than 10) is consis- tent with the assumption of one-dimensional penetrant diffusion.

Dynamic and equilibrium swelling experiments

The thin glassy polymeric discs were swollen in a 3,3’- dimethyl glutaric acid (Sigma Chemicals, St. Louis, MO, USA) buffer at 37 f 0.2” C. The glutaric acid was chosen because it contains only one buffering component and gives a wide pH range of 3.2-7.6. Glutaric acid is a dibasic acid with two pK, values of 3.7 and 6.34 at 25°C. The other buffer used was an acetate buffer with acetic acid as a buffering agent having pK, = 4.76 at 25°C. All the swelling studies were carried out in 100 ml buffer. The concentration of the added buffering agent was 0.01 M. Ionic strength, Z, was calculated by standard techniques, and maintained at the appropriate concentration by the addition of NaCl. In the calculation, the degree of dissociation of the acetic and glutaric acids was taken into account at each pH value.

The swelling studies were carried out in duplicate by measuring the amount of water absorbed gravimetri- tally at specified time points. The average value of the water absorbed in grams of water absorbed per gram of dry polymer is reported. The equilibrium swelling was attained in 2 to 3 weeks.

RESULTS

Dynamic swelling

To investigate the influence of external conditions on the dynamic and equilibrium behaviour of anionic networks, thin disks of poly(HEMA-co-MAA) and poly(HEMA-co-MAA) were tested in various buffer solutions. Figure 2 presents the dynamic water uptake of poly(HEMA-co-MAA) (50:50 mol.%) glassy copolymers in an aqueous glutarate buffer, at a constant ionic strength of 0.1 M at 37”C, as a function of pH of the external swelling medium. It was observed that the mass uptake of water increased as the pH of the swelling medium increased in agreement with previous studies2p3. As the pH of the swelling medium increased through the pK, of monomeric MAA, the degree of ionization in the gel increased. That resulted in a more hydrophilic polymer network and contribu- ted to this higher water uptake.

The first 60% of the dynamic water uptake data was fitted’l using the following equation

Biomaterials 1995, Vol. 16 No. 7

Swellingldeswelling of anionic copolymer gels: A./?. Khare and N.A. Peppas 561

A 0 A 0

A o” A O0

0 0

;;Elo El80 80

0 1 2 3 4 5

Time (h)

Figure 2 Dynamic swelling behaviour of poly(2- hydroxyethyl methacrylate-co-methacrylic acid) (5050 mol.%) copolymers in various pH glutarate buffers of ionic strength, I = 0.1 M, at 37°C. Only the first 5 h of dynamic swelling behaviour are shown. Data are for buffers with pH 4.22 (0), 5.06 (O), 6.12 (0) and 6.81 (A).

Mt - = kt” MIX

Here, Mt is the mass of water absorbed at any time t; Mm is the amount of water absorbed at equilibrium; and k and n are constants. The values of the diffusional exponent n are presented with their 95% confidence limits as calculated from a non-linear statistical routine. For a planar geometry, the value of n = 0.5 signifies a Fickian water diffusion mechanism, while n = 1.0 indicates a Case II diffusional mechanism. For 0.5 < n < 1.0, the diffusional mechanism is non- Fickian, where both diffusion and polymer relaxation control the overall rate of water uptake.

The values of n are reported in Tables 1 and 2. It was found that n increased as the pH of the external

swelling medium increased. This indicates that the water transport mechanism becomes non-Fickian as gel ionization becomes prominent. As discussed previously’, the dynamic swelling behaviour of cross- linked polymers is dependent upon the relative magnitude of water diffusion and polymer relaxation times. Fickian transport is observed when water diffusion controls the process. Case II transport (n = 1.0) is the controlling mechanism when macromolecular relaxations predominate. In ionic polymeric networksz918, ionization may control the water diffusion process, thus affecting the relative magnitude of diffusion and relaxation times. Indeed, in preliminary dataIS we have shown that in anionic polymer networks, macromolecular relaxation becomes more prominent in alkaline solutions. Thus, non-Fickian (anomalous) transport is observed as the pH of the surrounding fluid increases above pK,. The data presented here amplify these conclusions. In general, an increase in the degree of ionization contributes to the electrostatic repulsion between adjacent ionized carboxylate groups leading to chain expansion, which in turn affects macromolecular chain relaxation. Contrary to that, swelling studies in buffers of pH 4.13, which is less than the pK, of monomeric acids, gave an exponent of approximately 0.5 indicating that the diffusion mechanism is Fickian. The value of the diffusional exponent n was found to be 0.5 in water for all the polymer composi- tions studied, and independent of the ionic content in the copolymer. Further analysis of this type of relaxation-controlled transport in ionic polymers is discussed by Hariharan and Peppa?.

Similar results were obtained for poly(HEMA-co-AA) copolymeric networks, as shown by the value of the diffusional exponent n in Table 3. The water transport mechanism deviated from a Fickian mechanism much more in AA-containing copolymeric networks than in MAA-containing copolymeric networks. This is because the m-methyl group of MAA gives a more

Table 1 Diffusional exponent n (calculated from fquation 2) for water transport in poly(2-hydroxyethyl methacrylate-co-methacrylic acid) gels swollen in 3,3’-dimethyl glutaric acid buffers of varying pH at constant ionic strength, I = 0.1 M and 37°C

pH of buffer Methacrylic acid content (mol.%) in copolymers

0 20 40 50 60 100

4.22 0.53 f 0.05 0.46 z!z 0.03 0.54 f 0.02 0.51 * 0.03 0.53 * 0.01 0.61 f- 0.01 5.06 0.48 * 0.02 0.55 * 0.03 0.56 f 0.01 0.64 k 0.04 0.44 f 0.06 0.60 zt 0.03 6.12 0.48 f 0.02 0.67 f 0.04 0.68 zt 0.01 0.69 f 0.01 0.72 * 0.02 0.73 It 0.01 6.81 0.47 It 0.04 0.73 * 0.01 0.65 zt 0.02 0.75 zt 0.02 0.77 * 0.01 0.76 zk 0.01 7.04 0.49 f 0.04 0.72 * 0.01 0.75 f 0.02 0.76 zt 0.02 0.81 zk 0.02 0.82 + 0.01

Table 2 Diffusional exponent n (calculated from Equation 2) for water transport in poly(Bhydroxyethyl methacrylate-co-methacrylic acid) gels swollen in acetate buffer of varying pH at constant ionic strength, I = 0.1 M and 37°C

pH of buffer Methacrylic acid content (mol.%) in copolymers

0 20 40 50 60 100

4.13 0.46 zt 0.05 0.48 Z!Y 0.03 0.52 f 0.02 0.43 zt 0.06 0.51 It 0.02 0.42 zk 0.04 4.64 0.43 It 0.04 0.50 f 0.03 0.48 z!z 0.02 0.49 f 0.01 0.55 * 0.04 0.57 f 0.02 5.19 0.50 * 0.03 0.57 f 0.02 0.58 f 0.02 0.56 f 0.02 0.63 f 0.01 0.71 f 0.01 5.67 0.48 zt 0.01 0.65 f 0.02 0.66 * 0.03 0.68 * 0.01 0.72 f 0.03 0.70 f 0.03 water 0.48 zt 0.03 0.50 i 0.01 0.48 z!z 0.02 0.47 f 0.03 0.48 IIZ 0.02 0.41 * 0.03

Biomaterials 1995. Vol. 16 No. 7

562 Swelling/deswelling of anionic copolymer gels: A.R. Khare and N.A. Peppas

Table 3 Diffusional exponent n (calculated from Equation 2) for water transport in poly(2-hydroxyethyl methacrylate-co- acrylic acid) gels swollen in acetate buffer of varying pH at constant ionic strength, I = 0.01 M and 37” C

pH of Acrylic acid content (mol.%) in copolymers buffer

20 40 60 80

4.14 0.52 i 0.02 0.52 f 0.03 0.62 f 0.01 0.51 * 0.02 4.71 0.64 -f 0.01 0.78 & 0.01 0.78 & 0.02 0.55 f 0.01 5.21 0.77 f 0.03 0.88 zt 0.01 0.86 It 0.02 0.69 I!= 0.02 5.66 0.78 ILL 0.04 0.86 It 0.01 0.87 + 0.03 -

4 5 6 7 8 pH of Buffer

Figure 3 Equilibrium water uptake vs. pH for poly(P hydroxyethyl methacrylate-co-methacrylic acid) copolymers in glutarate buffers of ionic strength I = 0.1 M, at 37°C. Methacrylic acid content in these copolymers was 20 (0), 40 (O), 50 (0), 60 (A) and 100 (0) mol.%.

hydrophobic character to the polymer network and this hinders gel ionization.

Figure 3 shows the equilibrium swelling behaviour as a function of pH at 37” C and I = 0.1 M for poly(HEMA- co-MAA) copolymers in a glutarate buffer. At a lower pH of 4.22, the equilibrium water uptake was found to be 1-2 g water per g dry polymer. However, as the pH increased the equilibrium water uptake increased continuously showing a large change. At pH 7.04, the equilibrium water uptake was approximately log water per g dry polymer, i.e. an order of magnitude higher than in acidic solutions as seen in Tables 4 and 5. The midpoint of the S-shaped curve of equilibrium water uptake against pH can be used to calculate an approximate pK, of the gel; that value lies between 5.5 and 6.5. The incorporation of HEMA and the formation of a network structure contributes to this increase in pK, of the gel.

Similar behaviour was observed for swelling in acetate buffers as well as for poly(HEMA-co-AA) copolymers in both acetate and glutarate buffer as shown in Table 6. The equilibrium water uptake values for AA-containing copolymers were higher than those for MAA-containing copolymers. The cc-methyl group of MAA provides a more hydrophobic character to the polymer network. As can be seen from Table 5, polyHEMA homopolymer did not show any pH- dependent equilibrium swelling behaviour. Its

Table 4 Equilibrium water uptake (grams of water per gram of

dry polymer) in poly(2-hydroxyethyl methacrylate-co- methacrylic acid) gels swollen in 3,3’-dimethyl glutaric acid buffers of varying pH at constant ionic strength, I = 0.1 M and 37; c

pH of Methacrylic acid content (mol.%) in copolymers buffer

0 20 40 50 60 100

4.22 0.75 0.83 1.07 0.94 1.04 2.65 5.06 0.73 1.34 1.95 2.21 2.46 4.20 6.12 0.71 5.70 7.66 6.86 7.99 9.57 6.81 0.72 7.78 9.84 9.19 10.56 13.07 7.04 0.72 7.55 10.27 9.31 10.88 13.85

Table 5 Equilibrium water uptake (grams of water per gram of

dry polymer) in poly(2-hydroxyethyl methacrylate-co- methacrylic acid) gels swollen in acetate buffers of varying pH at constant ionic strength, I = 0.1 M and 37” C

pH of Methacrylic acid content (mol.%) in copolymers buffer

0 20 40 50 60 100

4.13 0.68 0.91 1.10 1.90 2.89 3.16 4.64 0.69 2.40 4.05 4.68 4.76 8.13 5.19 0.69 5.54 8.08 8.75 8.43 9.66 5.67 0.68 8.31 11.12 10.36 12.67 12.73

Table 6 Equilibrium water uptake (grams of water per gram of dry polymer) in poly(2-hydroxyethyl methacrylate-co-acrylic acid) gels swollen in acetate buffers of varying pH at constant ionic strength, I = 0.1 M and 37’C

pH of buffer

Acrylic acid content (mol.%) in copolymers

20 40 60 80

4.14 1.15 3.82 5.24 5.04 4.71 3.09 5.90 7.34 6.77 5.21 4.06 7.77 8.82 8.25 5.66 6.08 9.29 9.82

dynamic swelling behaviour was found to be Fickian regardless of the pH of the external swelling medium.

Figure 4 presents the effect of cycling of pH on the swelling behaviour of these ionic networks. The pH was changed from 7 to 4 and the same cycle was repeated twice. The swelling time in pH 7 buffer was 24 h followed by 100 h in pH 4 buffer. The same cycle was repeated by keeping the gels for 24 h in each of pH 7 and 4 buffers, respectively. The water uptake values were the same after each 24 h period indicating reversible gel behaviour. The gels did not lose their elasticity during pH cycling. In general, the deswelling times were faster than the swelling times.

Effect of ionic strength

The dynamic and equilibrium swelling studies were also carried out as a function of ionic strength at pH 4.0 and 7.0 in a glutarate buffer. The ionic strength was changed from 0.0074 to 0.08M and the swelling studies were carried out at a constant temperature of 37’C. Figure 5 shows the dynamic uptake for poly(HEMA-co-MAA) (60:40mol.%) as a function of ionic strength at pH 4. The ionic strength did not affect


Swellingldeswelling of anionic copolymer gels: A.R. Khare and N.A. Peppas 563

a

--_--___

-I

0 t,,,,I,,,,I,,,,I,,,,l,,,,l

0 50 100 150 200 250 Time (h)

Figure 4 Cyclic swelling behaviour of poly(2-hydroxyethyl methacrylate-co-methacrylic acid) copolymers in glutarate buffers of ionic strength I = 0.1 M, at 37°C. The methacrylic acid content in these copolymers was 20 (A) and 40 (0, 0) mol.%.

BOA

0 1 2 3 4 5 Time (h)

Figure 5 Dynamic swelling behaviour of POlY(2- hydroxyethyl methacrylate-co-methacrylic acid) (60:40mol.%) copolymers in glutarate buffers of varying ionic strengths of pH =4.0, at 37°C. Only the first 5 h dynamic water uptake is shown. Data are for buffers with ionic strengths of / = 0.0074~ (0), / = 0.02~ (O), / = 0.05~

(0) and I = 0.08~ (a).

the dynamic swelling behaviour of these polymers. At pH 4.0 the gel was mostly in the un-ionized state. The value of diffusional exponent n was found to be approximately 0.5 in all cases indicating a Fickian diffusion mechanism.

Figure 6 shows the dynamic uptake for poly(HEMA- co-AA) (80:20mol.%) as a function of ionic strength at pH 7. As the ionic strength increased from 0.0074 to O.O8M, the dynamic water uptake rate decreased. Similarly, the value of the diffusional exponent n decreased from 1.01 to 0.87. At pH 7, the polymer network was mostly in the ionized form with a degree of ionization of 1. Therefore, as the ionic strength increased, the difference in concentration of mobile ions between the gel and the solution was reduced. This reduction in the difference in the concentration of these mobile ions reduced the ion osmotic swelling

0 1 2 3 4 5 Time (h)

Figure 6 Dynamic swelling behaviour of PotY (2- hydroxyethyl methacrylate-co-acrylic acid) (80:20 mol.%) copolymers in glutarate buffers of varying ionic strengths of pH = 7.0, at 37” C. Only the first 5 h dynamic water uptake is shown. Data are for buffers with ionic strengths of I = 0.0074~ (A), I = 0.02~ (0), I = 0.05~ (0) and I = 0.08~

(0).

pressure. Table 7 shows the values of the diffusional exponent n for both poly(HEMA-co-MAA) and poly(HEMA-co-AA) copolymers; the water diffusion mechanism was anomalous in both polymers.

The equilibrium swelling studies showed that at pH 4.0 the ionic strength did not affect the equilibrium water uptake values (Table 8), since the gel was mostly in the un-ionized state. However, at pH 7.00 the equilibrium water uptake values decreased as the ionic strength increased (Table 9). This is because of the reduction in the ion osmotic swelling pressure. The decrease in equilibrium water uptake values was not substantial. Our results corroborate other recent studies13,15 for cationic polymeric networks containing amine groups.

Effect of buffer composition

To study the effect of buffer composition on the dynamic and equilibrium swelling behaviour, two buffer systems were chosen. The acetate buffer was based on a monobasic acid and had a pH range of 3.6- 5.6, whereas the glutarate buffer was based on a dibasic acid and had a pH range 3.2-7.6. All the swelling studies were carried out at 37” C and Z = 0.1 M. The initial concentration of buffering acid was 0.01 M. Figure 7 shows the dynamic water uptake of poly(HEMA-co-MAA) gels at pH 4.86; the uptake was higher in acetate buffer than in glutarate buffer. Similar observations were made for poly(HEMA-co- AA) gels as shown in Figure 8.

Although these observations may seem contrary to the Donnan co-ion exclusion, Donnan exclusion applies only partially to loosely cross-linked polymeric networks14. For example, in such gels there is a large percentage of bulk water present inside the gel where Donnan exclusion does not apply completely. As the electroneutrality condition applies to the gel at equilibrium, a certain co-ion concentration is essential inside the gel. Since acetate ions carry a negative


564 Swellingldeswelling of anionic copolymer gels: A.R. Khare and N.A. Peppas

Table 7 Diffusional exponent n (calculated from Equation 2) for water transport in poly(Z-hydroxyethyl methacrylate-co-acrylic acid) and poly(2-hydroxyethyl methacrylate-co-methacrylic acid) gels swollen in 3,3’-dimethyl glutaric acid buffers of varying ionic strength at a constant pH 7.0 and 37” C

Ionic strength of buffer (M)

0.0074 0.02 0.05 0.08

Acrylic acid content (mol.%) in copolymers

20 60

1.01 It 0.01 1.06 f 0.03 0.91 It 0.02 0.92 f 0.01 0.88 * 0.01 0.88 + 0.02 0.87 f 0.01 0.87 i 0.02

Methacrylic acid content (mol.%) in copolymers

20 60

0.87 zt 0.02 0.81 I’L 0.01 0.81 l 0.02 0.78 i 0.01 0.74 f 0.01 0.73 z!z 0.03 0.78 zt 0.03 0.75 It 0.02

Table 8 Equilibrium water uptake (grams of water per gram of dry polymer) in poly(2-hydroxyethyl methacrylate-co-acrylic acid) and poly(2-hydroxyethyl methacrylate-co-methacrylic acid) gels swollen in 3,3’-dimethyl glutaric acid buffers of varying ionic strength at a constant pH 4.0 and 37” C

Ionic strength of buffer (M) Acrylic acid content (mol.%) in copolymers

20 60


20 60

0.0074 0.82 0.69 0.72 1.75 0.02 0.82 0.72 0.78 1.22 0.05 0.84 0.76 0.81 1.28 0.08 0.80 0.76 0.84 2.35

Table 9 Equilibrium water uptake (grams of water per gram of dry polymer) in poly(2-hydroxyethyl methacrylate-co-acrylic acid) and poly(2-hydroxyethyl methacrylate-co-methacrylic acid) gels swollen in 3,3’-dimethyl glutaric acid buffers of varying ionic strength at a constant pH 7.0 and 37” C

Ionic strength of buffer (M) Acrylic acid content (mol.%) in copolymers

20 60


20 60

0.0074 8.47 10.63 9.82 11.35 0.02 7.67 9.99 8.86 9.82 0.05 6.55 9.19 7.91 9.40 0.08 5.56 8.28 7.56

charge, while glutarate ions carry one or two charges, C acetate,gel is greater than Cglutarate,sel for the same counterion concentration. Therefore, there is a higher difference in acetate ion concentration between the gel and the external swelling medium compared with the glutarate ion concentration differential across the gel and the external solution. Thus, acetate ions create a higher ion osmotic swelling pressure.

Figure 9 presents the equilibrium water uptake as a function of pH for poly(HEMA-co-MAA) (60:40 mol.%) in acetate and glutarate buffers. The equilibrium water uptake values were higher in acetate buffer compared with glutarate buffer. The higher water uptake in gels swollen in acetate buffers was due to the partitioning of co-ions in the gel phase due to excess Na+ ions inside the gels. It is knownI that certain anionic exchange resins can form cationic gels. Therefore, co- ions such as the acetate or glutarate would become counterions due to this conversion. In the case of poly(HEMA-co-AA) polymer networks, a similar behaviour was observed (Figure IO). The equilibrium water uptake values for poly(HEMA-co-MAA) in these buffers were lower than those for poly(HEMA-co-AA).

Effect of counterions

As shown in Table 10, the nature of the counterion did not affect the Fickian diffusion mechanism of water

since the value of the diffusion exponent n was approximately 0.5. This was due to the limited ionization of fixed carboxylic acid groups present in the ionic network. However, the equilibrium water uptake values decreased as the valency of the counterion changed from 1 to 2 (Table 12). The decrease in the equilibrium water uptake values was not substantial. Obviously, calcium can bind to two carboxylate ions whereas only half the amount of sodium ions would be required to fulfill the electroneutrality condition for the same degree of ionization inside the polymeric network. That results in a decreased ion osmotic swelling pressure, and therefore the decrease in equilibrium water uptake values. Since the decrease in the equilibrium water uptake values is not very large, one concludes that the degree of ionization of the fixed ionizable groups inside the gel is very small. In addition, aqueous electrolytic solutions do not have any buffering capacity, and this contributes to a lower degree of ionization.

In conclusion, the higher swelling of MAA-containing copolymers as indicated in Table 1 I is due to the lower degree of cross-linking of these hydrogels. In an unbuffered solution of pH 7 this leads to limited ionization of the carboxyl groups. Therefore, the ionization contribution towards swelling is limited and the only major contribution is due to the mesh size of the network, which is, of course, controlled by the degree of cross-linking.


Swellingldeswelling of anionic copolymer gels: A.R. Khare and N.A. Peppas 565

0 0 1 2 3 4 5 6

Time (h)

Figure 7 Dynamic swelling behaviour of poly(2- hydroxyethyl methacrylate-co-acrylic acid) (60:40 mol.%) copolymers in two different buffers of pH =4.86, ionic strength I = 0.1 M, at 37” C. Only the first 5 h dynamic water uptake behaviour is shown. Data are for glutarate buffers (0) and acetate buffers (0).

*/-

U D 0 z 1 0 0

3

i

0 O0 & Zj 0 0 O0

O0 0.5

00 &IO

0 1 2 3 4 Time (h)

Figure 8 Dynamic swelling behaviour of PolY(2- hydroxyethyl methacrylate-co-acrylic acid) (60:40 mol.%) copolymers in two different buffers of pH =5.06, ionic strength I = 0.1 M and at 37°C. Only the first 5 h dynamic water uptake behaviour is shown. Data are for glutarate buffers (0) and acetate buffers (0).

DISCUSSION

The swelling behaviour of all samples studied could be described by a monotonic function. There was no sudden increase in the dynamic uptake after the dynamically swelling polymer became completely rubbery. The time required for polymer samples to reach the complete rubbery state was more than 3 h.

During the swelling process in a pH medium, the glassy polymer discs absorb water. Owing to the plasti- cization process, the polymer becomes rubbery and this process continues until the two advancing glassy- rubbery fronts meet at the centre. The velocity of these fronts depends upon the swelling as well as the rate of ionization at the glassy-rubbery interface. There are two ways a counterion can diffuse to this interface. It

E 4- cl .z 0 = s

2 - 0

W -CD

0 I,,,,I,,,,I,,,,-

4 5 6 7 8 pH of Buffer

Figure 9 Equilibrium water uptake of poly(2-hydroxyethyl methacrylate-co-methacrylic acid) (60:40 mol.%) vs. pH in two different buffers of ionic strength, I = 0.1 M, at 37°C. Data are for glutarate buffers (0) and acetate buffers (0).

9 I" I I” I I I I ’ I’ I

0 0.02 0.04 0.06 0.08 0.1

Ionic Strength (M)

Figure 10 Equilibrium water uptake of poly(2-hydroxyethyl methacrylate-co-acrylic acid) (80:20 mol.%) in glutarate buffers of varying ionic strengths of pH = 7.00, at 37” C. The data are shown for two different samples (0, 0).

can be carried by an un-ionized salt with a weak acid, or in the ionized form. At the interface, the released protons of the carboxylates are accepted by the buffering component to maintain a constant pH. These ion exchange steps explainz3 the pH-dependent swelling behaviour of these hydrogels.

Another important aspect of the phenomenon is the increase in the gel’s apparent equilibrium constant compared with that of the monomer acids. From the graph of equilibrium water uptake against pH, it can be estimated that the pK, of these gels is between 5.5 and 6.5 (Figure 3). This increase is due to the incorporation of HEMA which is a non-polar moiety, as well as due to the steric hindrance of the gel towards the proton and the counterions. That leads to the creation of a microenvironment around the fixed charges where the true pK, of the monomer acids is valid. However, because of these reasons, this change in the apparent equilibrium constant of the gel is a very useful tool for


566 Swelling/deswelling of anionic copolymer gels: A.R. Khare and N.A. Peppas

Table 10 Diffusional exponent n (calculated from Equation 2) for water transport in poly(2-hydroxyethyl methacrylate-co-acrylic acid) and poly(2-hydroxyethyl methacrylate-co-methacrylic acid) gels swollen in aqueous solutions of NaCl and CaCl* at constant ionic strength, I = 0.1 M and 37” C

Salt solution Acrylic acid content (mol.%) in copolymers Methacrylic acid content (mol.%) in copolymers

20 60 20 60

NaCl 0.54 f 0.06 0.56 f 0.06 0.53 It 0.02 0.54 i 0.03 CaC12 0.44 * 0.03 0.46 zt 0.02 0.51 It 0.02 0.59 f 0.04

Table 11 Equilibrium water uptake (grams of water per gram of dry polymer) in poly(2-hydroxyethyl methacrylate-co-acrylic acid) and poly(2-hydroxyethyl methacrylate-co-methacrylic acid) gels swollen in aqueous solutions of NaCl and CaCI, of constant ionic strength, I = 0.1 M and 37” C

Salt solution Acrylic acid content (mol.%) in copolymers Methacrylic acid content (mol.%) in copolymers

20 60 20 60

NaCl 1.19 1.05 i .4a 1.25 CaCI, 0.88 0.75 0.89 1.18

tailoring these gels for a particular site-specific drug mechanism remained Fickian in all the unbuffered delivery application. systems studied and at pH < pK,,s,l.

The mechanism of diffusion of water in these gels was found to be dependent upon whether or not the swelling medium is a buffered system. The diffusion mechanism was Fickian in all unbuffered swelling media. In unbuffered systems, due to very limited ionization, ion exchange between the gel and the swelling medium did not control the overall rate of mass transfer. However, in buffered systems, ion exchange controlled this overall rate inside the gel; this in turn determined the macromolecular relaxation contributing to the anomalous diffusion mechanism.

ACKNOWLEDGEMENT

This research was supported by grant GM43337 from the National Institute of Health.

REFERENCES

1

According to Donnan equilibrium at constant pH, as the ionic strength increases gel ionization increases in a weakly ionizable polyelectrolytic system. This leads to a higher equilibrium water uptake. After complete gel ionization, a further increase in the ionic strength results in a decrease in equilibrium water uptake because of the screening effect of the counterions. Once gel ionization is complete, the counterions are excluded from further entering into the gel. Thus, the concentration differential of counterions between the gel and swelling medium is reduced. In our case, we did not observe any initial increase in equilibrium water uptake. However, a decrease was observed at pH 7 due to complete gel ionization. A further increase in the ionic strength reduced the ion osmotic swelling pressure, and that resulted in a decrease in the equilibrium water uptake values.

2

3

4

5

6

Gehrke SH, Lee PI. Hydrogels for drug delivery systems. In: Tyle P, ed. Specialized Drug Delivery Systems, New

York: Marcel Dekker, 1990: 333-392. Brannon-Peppas L, Peppas NA. Solute and penetrant diffusion swellable polymers. IX. The mechanism of drug release from pH-sensitive swelling controlled release systems. I Contr Release 1989; 8: 267-274. Kuo JH, Amidon GL, Lee PI. pH-Dependent swelling and solute diffusion characteristics of polylhydroxyethyl methacrylate-co-methacrylic acid) hydrogels. Pharm Res 1988; 5: 592-597.

Siegel RA, Firestone BA. pH-Dependent equilibrium swelling properties of hydrophobic polyelectrolyte copolymer gels. Macromolecules 1988; 21: 3254-3259. Tanaka T. Hydrogels. In: Mark HF, Kroschwitz JI, eds. Encyclopedia of Polymer Science and Technology,

New York: Wiley, 1985: 514-531.

Bae YH, Okano T, Kim SW. Insulin permeation through thermo-sensitive hydrogel beads. / Conk Release 1989;

9: 271-276. 7

CONCLUSIONS

Dong LC, Hoffman AS. Thermally reversible hydrogels. In: Russo PS, ed. Reversible Polymer Gels and Related

Systems, ACS Symposium Series No. 350. Washington D.C.: ACS, 1987: 236-244.

Ionization equilibrium, polymer morphology and buffer composition play a very important role in controlling the swelling properties of ionizable gels such as poly(HEMA-co-MAA). Their swelling behaviour was dependent on pH, ionic strength, nature of the counterion and buffer composition. The pK, of these polymers was in the range of 5.5-6.0. The mechanism of water diffusion in these gels was anomalous in all the buffered swelling media. However, the diffusion

8

9

10

11

Freitas RFS, Cussler EL. Temperature sensitive gels as extraction solvent. Chem Eng Sci 1987; 42: 97-103. Lee KK, Cussler EL, Marchetti M, McHugh MA. Pressure dependent phase transitions of hydrogels. Chem Eng Sci 1990; 45: 766-767. Grodzinsky AJ, Weiss AM. Electric field control of membrane transport and separations. Sep Purif Methods 1985; 14: l-40. Irie M, Photoresponsive polymers. Adv Polym Sci 1990;

94: 28-67.


Swelling/deswelling of anionic copolymer gels: A.R. Khare and N.A. Peppas 567

12

13

14

15

16

17

Peppas NA, Barr-Howell BD. Characterization of the cross-linked hydrogels. In: Peppas NA, ed. Hydrogels in Medicine and Pharmacy, Vol. 1. Boca Raton, FL: CRC Press, 1987: 27-56.

18

Firestone BA, Siegel RA. Kinetics and mechanisms of water sorption in hydrophobic, ionizable copolymer gels. JAppl Polym Sci 1991; 43: 901-914. Helfferich F. Ion Exchange, New York: McGraw-Hill, 1962.

19

20

Ricka J, Tanaka T. Swelling of ionic gels: quantitative performances of the Donnen theory. Macromolecules 1984; 17: 2916-2921. Beltran S, Baker JP, Hooper HH, Blanch HW, Prausnitz JM. Swelling equilibria for weakly ionizable, temperature-sensitive hydrogels. Macromolecules 1991; 24: 549-551. Huglin MB, Rego JM. Influence of a salt on some properties of hydrophilic methacrylate hydrogels. Macromole- cules 1991; 24: 2556-2563.

21

22

23

Brannon-Peppas L, Peppas NA. Dynamic and equilibrium swelling behavior of pH-sensitive hydrogels containing 2-hydroxyethyl methacrylate. Biomaterials 1990; 11:635-644. Nakano Y, Seida Y, Uchida M, Yamamoto S. Behavior of ions within hydrogel and its swelling properties. ] Chem Eng Jpn 1990; 23: 574-579. Sawai T, Yamazaki S, Ikariyama Y, Aizawa M. pH- Responsive swelling of the ultrafine microsphere. Macromolecules 1991; 24: 2117-2118. Ritger PL, Peppas NA. A simple equation for description of solute release. I. Fickian and non-Fickian release from non-swellable devices in the form of slabs, spheres, cylinders or discs. JContrRelease 1987; 5: 23-26. Hariharan D, Peppas NA. Swelling of ionic and neutral polymer networks in ionic solution. J Membr Sci 1993; 4: 275-289. Gehrke SH, Cussler EL. Mass transfer in pH-sensitive hydrogels. Chem Eng Sci 1989; 44: 559-556.


Swelling/deswelling of anionic copolymer gels

Documents