Journal of Controlled Release 52 (1998) 221–225

Short Communication

Complications in investigations of the swelling of hydrogel matricesdue to the presence of trapped gas

*C.A. Fyfe , A.I. BlazekDepartment of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, BC V6T 1Z1, Canada

Received 30 July 1997; received in revised form 21 October 1997; accepted 11 November 1997

Abstract

In recent studies, NMR imaging has been used to investigate the swelling of hydroxypropylmethylcellulose (HPMC)tablets and to determine polymer concentration distributions in the swollen matrix. The total amount of polymer in thesystem was computed from these distributions and was up to 35% greater than the known weight of HPMC in the tablet.This deviation was traced to the presence of air bubbles in the swollen matrix which occupied a significant volumepreviously assumed to be occupied by polymer and water. When the air in the tablet was removed by vacuum, the swollengel contained no evidence of air bubbles and the calculated total polymer weights from the HPMC distributions were equalto the amount of HPMC in the tablet. 1998 Elsevier Science B.V.

Keywords: NMR Imaging; Controlled release; Hydrophilic matrix tablet; Swelling; Hydroxypropylmethylcellulose

1. Introduction In recent studies of HPMC tablet swelling usingNMR imaging methods [9], we have used imaging

Hydrogels such as hydroxypropylmethylcellulose and NMR data to calculate the time-dependent(HPMC) are important in a large number of different distributions of polymer in the swollen tablet withapplications including tablet formulations for con- the aim of developing a model to predict swellingtrolled drug release [1]. The swelling processes of behavior. However, when these distributions weresuch materials have been monitored in the past by used to calculate the total weight of HPMC in thethe measurement of dimensional changes, and more system, apparent overestimates of up to 35% wererecently by optical [2–4] and NMR imaging methods found compared to the known weight of the dry[5–8]. A detailed understanding of the swelling tablet before swelling. In the present work we havebehavior of these materials should provide insight traced this discrepancy to the presence of small gasinto the factors that regulate drug release from bubbles in the swollen hydrogel which are believedhydrophilic matrix tablets. to arise from air trapped in the tablet during com-

pression of the polymer powder.A simple experimental protocol to avoid the

*Corresponding author. problem and obtain quantitatively reliable hydrogel

0168-3659/98/$19.00 1998 Elsevier Science B.V. All rights reserved.PII S0168-3659( 97 )00217-4

222 C.A. Fyfe, A.I. Blazek / Journal of Controlled Release 52 (1998) 221 –225

concentration gradients in swelling studies of thesematerials is described.

2. Materials and methods

HPMC (Methocel K4M Premium) was obtainedfrom Dow Chemical and the HPMC tablets wereprepared by direct compression of the as-suppliedpowder using a rotary tablet press. The tablets,prepared with a peak compressional force of 6263MPa, weighed 16261 mg, were 12.7560.05 mm indiameter, and 1.3760.01 mm thick. They wereplaced at the bottoms of 12.8-mm i.d. NMR tubessuch that only one face was exposed to water therebyrestricting the tablet swelling to the axial direction.The bottom face and the edges of the tablets weresealed with a small amount of grease to preventleakage of water between the tablet and tube wall.Two variations of the swelling experiment wereperformed. In the first variation, 5 ml of distilled Fig. 1. Schematic representation of the experimental set-up forwater were added directly to the tablet. In the second vacuum-treating the HPMC tablets. Glass rods are used to hold

down the tablet in the NMR tube which is then placed in a longvariation, the tube containing the tablet was placedSchlenk tube and connected by a three-way valve to a vacuum lineinside a long Schlenk tube, as shown in Fig. 1, andand a source of water. The tablet is subjected to a vacuum of 2

kept under a vacuum of approximately 2 mm Hg for mm Hg for 1 min and then water is drawn into the Schlenk tube to1 min. (The glass rods, taped into place, were break the vacuum. Before the tablet is placed in the magnet forrequired to prevent the tablet from rising in the tube imaging, the glass rods are removed and the water drawn in by the

vacuum is replaced by fresh distilled water.during the vacuum treatment, an occurrence whichtended to cause water leakage between the tablet andthe tube wall.) After the treatment, water was ously [9]. The method involves the preparation ofallowed to enter the system to break the vacuum and mixtures of known HPMC composition, in units ofto prevent the reincorporation of air into the tablet. weight HPMC over total weight of the mixture, andThe water drawn into the tube by the vacuum was the calibration of the T relaxation parameter of the2

replaced with 5 ml of distilled water to make the two water component as a function of HPMC weightswellings as similar as possible from this point percent. The calibration from the mixtures was thenonwards. The vacuum-treated tablet system con- used to convert the T times measured from the2

tained little or no air bubbles during approximately imaging experiments into HPMC distributions. An40 h of swelling whereas the untreated system additional correction, necessary to eliminate anydeveloped numerous air bubbles. effect of diffusion on the measured T values from2

One- and two-dimensional NMR images were the imaging experiments, was performed using theacquired for both experimental variations at various known diffusion coefficients of water [9]. The two-times during the swelling of the tablet. The one- dimensional images of the water distribution in thedimensional images were acquired with a spin-echo swollen tablet were acquired with a standard slice-sequence that was used to monitor the water and selective spin-echo sequence with the plane of thepolymer concentration changes during the course of images orthogonal to the long axis of the tube andthe swelling. The details of the one-dimensional the direction of swelling. For the two-dimensionalimaging study and the method for calculating the imaging experiments, a solution containing 10 mMpolymer concentrations have been described previ- CuSO rather than distilled water was used to4

C.A. Fyfe, A.I. Blazek / Journal of Controlled Release 52 (1998) 221 –225 223

shorten the T relaxation time of the water protons1

and thus shorten the time to acquire the images. Each2563256 pixel image was acquired in 2.5 h with 32scans, a time-to-echo of 2.5 ms, and a repetitiondelay of 1 s. The slice thickness of the images was300 mm and the in-plane resolution was 70 mm.

3. Results and discussion

In the swelling of hydrogel matrix tablets, theabsorption of water causes a gradual increase in thevolume of the gel and a corresponding change in thepolymer concentration which decreases graduallyfrom the original dry-tablet face. Quantitative knowl-edge of these concentration gradient profiles asfunctions of the swelling time should provide suffi-cient information to model the swelling process Fig. 2. HPMC weight percent distributions calculated from NMRexactly and relate it to the release of a pharma- imaging data for the untreated tablet (s) and the vacuum-treatedceutical. As indicated above, we have recently used tablet (d) at a swelling time of 40 h. The dashed line at 0.137 cm

indicates the position of the initial water–tablet interface. Thethe technique of NMR imaging to determine thedrawing of the tablet in its initial state in the NMR tube isdistributions in polymer concentration during theincluded as an aid to interpret the distance scale.

swelling of an HPMC tablet [9]. Although NMRspectroscopy (and hence NMR imaging) does notdetect the HPMC polymer in these systems directlybecause of its lack of substantial molecular mobility, HPMC above, it was assumed that the hydrogelit is possible to determine the polymer concentration occupied the total volume inside the tube. We believeindirectly from its concentration-dependent effect on the local polymer concentrations measured from thethe T relaxation times of the water molecules in the variation of the water T values are completely valid2 2

polymer environment. In this manner, the local but only for the volume actually occupied by theHPMC concentrations of a swollen HPMC tablet hydrogel. The volume excluded by the air bubblesmay be determined at regular intervals during the converts directly into the substantial observed over-swelling process. An example of such a profile, estimate of the total polymer weight. This can bepresented in Fig. 2, shows the expected decrease in confirmed by swelling experiments in which there isthe polymer concentration from the original tablet no trapped gas present.location. The swelling time of 40 h was chosen so We have found that the effect of trapped gas maythat all calculated polymer concentrations were be efficiently eliminated by modifying a literaturebelow 40% w/w HPMC and could therefore be procedure that uses vacuum to remove air fromconsidered reliable [9]. within the tablet [10]. The two-dimensional cross-

However, the total polymer weight, calculated sectional images shown in Fig. 3 demonstrate thefrom these apparent concentrations by summing in substantial differences between the air content of ansmall slices over the known inner dimensions of the untreated versus a vacuum-treated tablet. In thesetube, was approximately 35% greater than the un- gray-scale images, the light areas indicate regionsambiguously known weight of the dry tablet before where the water concentration in the swollen tablet isswelling! This discrepancy was found to be due to high. The dark areas in the image of the untreatedthe presence of small air bubbles trapped within the tablet indicate regions where water is excluded byswollen hydrogel matrix, some of which were visible air. An accurate estimate of the volume occupied byto the eye. In the calculation of the total weight of air cannot be made from these images due to the fact

224 C.A. Fyfe, A.I. Blazek / Journal of Controlled Release 52 (1998) 221 –225

escaped from the weak gel or dissolved in the largeconcentration of water. At a swelling time of 40 h,the total polymer weight calculated for the untreatedtablet system was 218617 mg, about 35% higherthan the known weight of 16261 mg. The totalpolymer weight calculated for the vacuum-treatedsystem at the same swelling time was 180614 mg,only an 11% deviation, and, within 1.3 standarddeviations, equal to the known weight of the tabletbefore swelling. Although, we feel that the majoreffect of the trapped air was the result of occupiedvolume, there could also be a contribution fromsusceptibility effects which would lower the T value2

of water in the immediate vicinity of the air–gelinterface and also cause an increase in the apparentHPMC weight percent. HPMC concentration profilesfrom these experiments can now be used withconfidence in the interpretation of the hydrogelswelling and controlled drug release and studies ofthis type are currently in progress.

Although there may be some dissolved gas in theadded water, it was felt that the bulk of the gas wastrapped in the tablet during the compression of thepolymer powder. The degree of tablet porosity andthe amount of air trapped within those pores mayvary greatly between different experimental systemsas the porosity of a tablet depends on factors such aparticle size, compressional force and compressionalspeed [11]. The error which the incorporated gasintroduces can be substantial and will apply to all

Fig. 3. Two-dimensional images of the water distribution in an static swelling experiments of this type, including theuntreated tablet (A) and a vacuum-treated tablet (B) taken at

most common ones where only dimensional changesapproximately the same position in the swollen tablet at a swellingare monitored.time of 43 h.

that air bubbles appear larger than their true sizebecause of magnetic susceptibility effects. 4. Conclusions

When the calculation of the polymer weightpercent distribution was performed using NMR The presence of trapped air in swollen HPMCimaging data for the vacuum-treated tablet, a similar tablets causes significant deviations of the polymerbut not identical concentration profile resulted. The concentrations determined from NMR imaging ex-data in Fig. 2 show that the HPMC weight percents periments from the true values. Other methods usedcalculated for the vacuum-treated tablet were lower in the study of swelling hydrophilic matrix tabletsthan those calculated for the untreated tablet. The may be similarly susceptible to the presence oftwo distributions were similar in the lowest con- trapped air and investigators of these systems shouldcentration region where, for both systems, no air determine the effect of trapped air on their measure-bubbles were visible, most likely because the air had ments. A reevaluation of past data may be in order.

C.A. Fyfe, A.I. Blazek / Journal of Controlled Release 52 (1998) 221 –225 225

mobility and diffusion coefficients in the gel layer of HPMCAcknowledgementstablets using NMR imaging, Pharm. Res. 13(3) (1996) 376–380.

The authors would like to acknowledge financial [6] A.R. Rajabi-Siahboomi, R.W. Bowtell, P. Mansfield, A.support from the NSERC of Canada and Glaxo Henderson, M.C. Davies, C.D. Melia, Structure and behaviorWellcome Inc., Mississauga, Ontario, Canada. A.I.B. in hydrophilic matrix sustained release dosage forms: 2.

NMR-imaging studies of dimensional changes in the gelacknowledges the award of an NSERC PGS Fellow-layer and core of HPMC tablets undergoing hydration, J.ship and an I.W. Killam Predoctoral Fellowship.Control. Rel. 31 (1994) 121–128.

[7] R. Bowtell, J.C. Sharp, A. Peters, P. Mansfield, A.R. Rajabi-Siahboomi, M.C. Davies, C.D. Melia, NMR microscopy of

References hydrating hydrophilic matrix pharmaceutical tablets, Magn.Res. Imag. 12(2) (1994) 361–364.

[8] M. Ashraf, V.L. Iuorno, D. Coffin-Beach, C.A. Evans, L.L.[1] D.A. Alderman, A review of cellulose ethers in hydrophilicAugsburger, A novel nuclear magnetic resonance (NMR)matrices for oral controlled-release dosage forms, Int. J.imaging method for measuring the water front penetrationPharm. Tech. Prod. Mfr. 5(3) (1984) 1–9.rate in hydrophilic polymer matrix capsule plugs and its role[2] A. Thu Pham, P.I. Lee, Probing the mechanisms of drugin drug release, Pharm. Res. 11(5) (1994) 733–737.release from hydroxypropylmethylcellulose matrices, Pharm.

[9] C.A. Fyfe, A.I. Blazek, Investigation of hydrogel formationRes. 11(10) (1994) 1379–1384.from hydroxypropylmethylcellulose (HPMC) by NMR spec-[3] R.W. Korsmeyer, R. Gurny, E. Doelker, P. Buri, N.A.troscopy and NMR imaging techniques, MacromoleculesPeppas, Mechanisms of solute release from porous hydro-30(20) (1997) 6230–6237.philic polymers, Int. J. Pharm. 15 (1983) 25–35.

[10] S.J. Desai, P. Singh, A.P. Simonelli, W.I. Higuchi, Inves-[4] P. Gao, R.H. Meury, Swelling of hydroxypropyl methyl-tigation of factors influencing release of solid drug dispersedcellulose matrix tablets. 1. Characterization of swelling usingin inert matrices IV, J. Pharm. Sci. 55(11) (1966) 1235–a novel optical imaging method, J. Pharm. Sci. 85(7) (1996)1239.725–731.

[11] E. Shotton, J.J. Deer, D. Ganderton, The instrumentation of a[5] A.R. Rajabi-Siahboomi, R.W. Bowtell, P. Mansfield, M.C.rotary tablet machine, J. Pharm. Pharmacol. 15 (1963)Davies, C.D. Melia, Structure and behavior in hydrophilic106T–114T.matrix sustained release dosage forms: 4. Studies of water