Incorporation of small quantities of surfactants as a way to improve the rheological and diffusional behavior of carbopol gels

Journal of Controlled Release 77 (2001) 59–75 www.elsevier.com / locate / jconrel

Incorporation of small quantities of surfactants as a way to improve the rheological and diffusional behavior of carbopol gels Rafael Barreiro-Iglesias, Carmen Alvarez-Lorenzo, Angel Concheiro* ´ Farmaceutica ´ , Facultad de Farmacia, Universidad de Santiago de Compostela, Departamento de Farmacia y Tecnologıa 15782 Santiago de Compostela, Spain Received 18 April 2001; accepted 20 August 2001

Abstract This paper analyzes the effects of Tween 80, Pluronic F-127, sodium dodecylsulfate (SDS), and benzalkonium chloride on the macro and microviscosity of Carbopol  934NF (0.25–0.50 g / dl) pharmaceutical gels. Carbopol / surfactant interactions, which were reflected in changes in the intrinsic viscosity of the polymer and in shifts of IR spectra bands of films, considerably modified the rheological properties of the gel (flow and oscillatory rheometry) and the diffusion coefficients of polystyrene particles (dynamic light scattering, DLS). At pH 4, any surfactant at a concentration of 0.01 g / dl promoted interpolymer connections producing an open three-dimensional network with maximum viscous and elastic moduli, which does not disturb the diffusive movement of polystyrene particles. An increase in non-ionic surfactant (0.05–0.50 g / dl) gradually decreased viscosity and elasticity since there were more surfactant molecules to surround each carbopol particle, forming intrapolymeric micelles and breaking the interpolymer connections. This macroscopic effect is, however, not reflected in a decrease but in an increase in microviscosity (estimated by DLS) owing to the formation of larger carbopol / surfactant aggregates and free micelles that contribute significantly to the obstruction of the diffusional path. Both ionic surfactants decreased macroviscosity owing to ionic aggregation (benzalkonium chloride) or increase in ionic strength (mainly SDS), while the repercussion on the diffusion of polystyrene particles was dramatically different, and was hindered (due to the carbopol / surfactant aggegates) or enhanced (due to the shrinking of carbopol microgels), respectively. At pH 7.4, the ionization of the carboxylic groups produced an expansion of the polymer chains accompanied by a huge increase in viscosity and elasticity and a decrease in diffusion coefficients in comparison with those obtained at pH 4. The effects of the surfactants were similar to those observed at pH 4 but less intense. Chloramphenicol release studies (Franz–Chien cells) revealed that 0.01 g / dl surfactant did not affect the diffusion while a change in pH dramatically altered the process. The results show that by choosing the appropriate proportion of the most suitable surfactant, it is possible to modulate the flow behavior, elastic properties, and diffusional microenvironment of carbopol gels, without losing the pH-dependent gelling ability, which could improve the suitability of carbopol gels for drug delivery through different routes.  2001 Elsevier Science B.V. All rights reserved. Keywords: Carbopol; Surfactants; Gelling in situ; Viscoelasticity; Microviscosity

*Corresponding author. Tel.: 134-981-594627; fax: 134-981-547148. E-mail address: [email protected] (A. Concheiro). 0168-3659 / 01 / $ – see front matter  2001 Elsevier Science B.V. All rights reserved. PII: S0168-3659( 01 )00458-8

60

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

1. Introduction Aqueous polymeric dispersions are very useful as platforms in drug delivery since they can resist the physiological stress caused by skin flexion, blinking, and mucociliar movement; adopting the shape of the application area and controlling the drug release [1–3]. Some of these systems combine gelling in situ behavior with high fluidity at the moment of administration: the low viscosity makes application easy, but later the temperature, pH, or ionic strength conditions of the application site cause its behavior to become viscous [2,4]. The incorporation of small amounts of surfactants into a polymeric dispersion can dramatically alter the polymer conformation and the viscosity of the dispersion. Aggregation processes can appear as a consequence of hydrophobic interactions between the non-polar surfactant tail and the polymer backbone, electrostatic interactions between the polar heads of the surfactant and the charged groups of the polymer, or both [5]. Usually, the importance of the aggregation effects on the rheological properties of the mixture depends strongly on the relative proportion of both components showing, for a given system, very different and even opposite effects [6]. Surfactants can also greatly modify the responsiveness of temperature-sensitive polymers. These effects were reported for cellulose ethers [7], poloxamer-coacrylic acid [8], or poly(N-isopropylacrylamide) [9]. The rheological parameters provide information useful for predicting the in vivo behavior of polymer dispersions [10,11]. However, most of the studies of the influence of surfactants on the rheological properties of liquid or semisolid polymer-based formulations focus on the stability of low polymer concentration suspensions or emulsions, to which surfactants are added as wetting agents or to prevent microbiological growth [12]. There is an important lack of information about the effects of surfactants on the rheological properties of pharmaceutical gels, into which surfactants are frequently incorporated to modulate the drug release rate or promote the drug flux through the cellular membranes [13]. In addition to the effects on the bulk properties of the gel, it is foreseeable that polymer–surfactant interactions dramatically alter the microenvironment in which the solute diffusion occurs. The formation of polymer– surfactant aggregates or free micelles could modify

the size of the water-filled regions or the mobility of the polymer chains. A reduction in the gel free volume or an increase in the path length due to obstructions restrict the diffusive movement of the drug and decrease the drug release rate [14]. In polymer dispersions or in surfactant solutions separately, it has been shown that the effects of the polymer concentration or the surfactant [15–18] on the macroscopic flow properties of the system do not necessarily correlate with the effects on diffusion, which means that it is necessary to take the microviscosity of the medium into account. Carbopol  is very useful as a major component of drug delivery gel systems for buccal [19], transdermic [20], ocular [4], rectal [21], and nasal [10] applications. The physical properties of the carbopol gels, the time they remain on the application area, and the drug release rate are extremely sensitive to the presence and concentration of additives [22]. The pH-induced gelling in situ capability and the control of drug release can be improved by addition of cellulose ethers [4] or polyvinylpyrrolidone [23], which causes structural changes that increase viscosity. In two recent papers [24,25], the influence of some nonionic surfactants on the viscosity of carbopol dispersions was evaluated for high concentrations of Tween 80 and Pluronic F-127, using continuous flow measurements. The shear stress applied can have the important drawback of breaking the aggregates or dissolving the entanglement of the polymer, leading to loss of information about the structure of the systems. Nevertheless, both studies revealed strong changes in the gel strength caused by the surfactants and their important practical repercussions. No references were found regarding the microviscosity of carbopol / surfactant systems. This paper reports on the analysis of the interactions and the effects exerted by the addition of small amounts of surfactants of different nature and low toxicities [26] (two non-ionic: Tween 80 and Pluronic F-127; and two ionic: sodium dodecylsulfate (SDS) and benzalkonium chloride) on the macro and microrheological properties and pH-responsiveness of Carbopol  934 gels. We evaluated the behavior of carbopol-surfactant systems when the polymer is non-ionized (acidic) or completely ionized (neutral conditions). To obtain information about viscous and elastic behavior under shear conditions closer to the physiological, flow vis-

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

cometry and oscillatory rheometry were used [11,27], while information about the microstructure of the gels was obtained from probe particle diffusion using dynamic light scattering. Diffusion assays using chloramphenicol, as a model drug, were also carried out. The final aim was to establish what possibilities the incorporation of small quantities of surfactants may offer in the modulation of the rheological and diffusional behavior of the gels with a view to applying them to the different commonly used routes of delivery.

61

before and after neutralization, was measured at 258C in a Cannon–Fenske capillary viscometer (Afora, Spain) (six determinations per concentration). Intrinsic viscosity was estimated by linear regression fitting of Fedors’ equation [29] to the results thus obtained: 1 / [2(h 1r / 2 2 1)] 5 1 /([h ]C) 2 1 /([h ]Cm)

(1)

where hr is relative viscosity, [h ] is intrinsic viscosity, C is polymer concentration, and Cm is a constant.

2.2.2. Preparation and characterization of carbopol /surfactant dispersions

2. Material and methods

2.1. Materials Carbopol  934NF (molecular weight 29 400– 39 400 Da, batch AB17796) was provided by BF Goodrich Europe, UK. Benzalkonium chloride, sodium dodecylsulfate (SDS), polyoxyethylene 20 sorbitan monoleate (Tween 80), polyoxyethylene– polyoxypropylene–polyoxyethylene triblock copolymer (Pluronic F-127), and chloramphenicol were from Sigma. Polystyrene latex nanospheres, diameter 162 nm, were provided by Duke Scientific Co. (Palo Alto, CA, USA). Purified water obtained by reverse osmosis (MilliQ  , Millipore, Spain) with resistivity ,1.82 mV cm was used. The other reactives were of analytical grade.

2.2. Methods 2.2.1. Characterization of Carbopol 934 NF 2.2.1.1. pKa and content on carboxylic groups. The procedure was that outlined in USP24-NF19 [28]. 2.2.1.2. Intrinsic viscosity. The viscosity of aqueous dispersions (0.010, 0.020, 0.050, 0.075, 0.10 g / dl)

2.2.2.1. Diluted dispersions. Intrinsic viscosity. To evaluate the effects of surfactant on the intrinsic viscosity of carbopol, the polymer was dispersed in solutions of each surfactant below and above its critical micellar concentration (cmc) (Table 1). The surfactant concentrations were as follows: 0.0075 and 0.050 g / dl for Tween 80; 0.50 and 2.0 g / dl for Pluronic F-127; 0.10 and 0.30 g / dl SDS; 0.002 and 0.30 g / dl for benzalkonium chloride. Intrinsic viscosity was estimated as described above in Section 2.2.1.2. 2.2.2.2. Concentrated dispersions Preparation Dispersions of carbopol (0.25 or 0.50 g / dl) without and with 0.01, 0.05, 0.1 and 0.5 g / dl surfactant were prepared by mixing solutions of carbopol and surfactant prepared in MilliQ  water. Additionally, a 0.50 g / dl carbopol dispersion in 0.10 g / dl NaCl was prepared. The pH of the dispersions was adjusted to 4 or 7.4 by adding 18% NaOH aq. After stirring for 1 h (800 rpm, RW20 DZM Janke & Kumkel, Germany), the systems were left to stand for 24 h before characterization. Characterization Infrared spectroscopy. IR spectra of films obtained by desiccation of 0.5 g / dl

Table 1 Molecular weight and critical micellar concentration (cmc) of the surfactants, and critical aggregation concentration (cac) in 0.25 g / dl Carbopol  934NF dispersions [51–53] Property

Tween 80

Pluronic F-127

SDS

Benzalkonium Cl.

Molecular weight cmc (g / dl) cac (g / dl)

1310 0.02 0.01

12 000 0.5 0.1

288.4 0.24 0.04

360 0.17 0.008

62

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

carbopol / 0.5 g / dl surfactant dispersions were recorded on a Bruker IFS 66V FT-IR spectrometer (Germany). To prepare the films, the dispersions were poured into a Teflon frame model and maintained at 408C for 24 h. IR spectra were obtained over the range 400–4000 cm 21 by the ATR (attenuated total reflection) technique. The results were compared with the IR spectra of 0.5 g / dl carbopol dispersions before and after neutralization with NaOH. Turbidimetry. The cloudiness of all dispersions was determined in triplicate by measuring transmittance (800 nm, Shimadzu UV-240, Japan) at room temperature against a blank of carbopol dispersion without surfactant. Rheology. Rheological behavior was evaluated in triplicate at 25 and 378C in a Rheolyst AR-1000N rheometer (TA Instruments, Newcastle, UK) equipped with an AR2500 data analyzer, fitted with a Peltier temperature control, and a 4 cm flat plate measuring geometry, or with a thermostated concentric-cylinder adapter and a medium coaxial cylinder-recessed end. (a) Ostwald’s equation was fitted to the flow curves:

h 5 m g~ n 21

(2)

where h represents viscosity, g~ the shear rate (up to 21 1400 s ), m the consistency index, and n the fluidity index. The thixotropy of each system was quantified by measuring the area of the flow cycle. When different rheological behaviors were observed in the same rheogram, the cycle was subdivided into different ranges and analyzed separately. (b) Oscillatory shear responses (G9 or elastic modulus, and G0 or viscous modulus) were determined at 0.1 Pa over the frequency range 0.05–50 rad / s. The linearity of viscoelastic properties was verified for all dispersions investigated. (c) The lifetime of the junction of the networks was obtained from the critical frequency of crossover, vc , in which G9 and G0 are equal, using the expression:

tc 5 1 / v c

(3)

2.2.2.2.3. Diffusion assays Diffusion coefficients of polystyrene latex nanospheres (162 nm) in all 0.25

g / dl carbopol / surfactant systems at pH 4 and 7.4 were estimated on the basis of 10 replicate assays, by dynamic light scattering (DLS) in a Zetasizer 3000HS apparatus (Malvern Instruments Ltd., UK) equipped with a He–Ne laser (633 nm) and a 7132 integrator–correlator. The measurement angle was 908, temperature 258C, and the data acquisition time 30 s. Microviscosity was estimated using the expression:

h /h0 5 D0 /D

(4)

where h and h0 are the viscosities (mPa s) of the polymer dispersion and of the medium without polymer, respectively, and D and D0 are the diffusion coefficients (cm 2 / min) for the nanospheres in the presence and absence of polymer, respectively. Assays for the characterization of chloramphenicol release from gels prepared with 0.25 g / dl carbopol and 0.01 g / dl surfactant were performed at 378C, in triplicate, in Franz–Chien vertical diffusion cells (Vidra Foc, Valencia, Spain) fitted with cellulose acetate membrane filters (0.45-mm pore size) between the donor and recipient compartments. A sample of 3.0 ml of the test formulation was placed in the donor compartment, while 6.0 ml of isoosmotic NaCl aqueous solution filled the recipient compartment and was stirred with a magnetic rod. Samples (0.50 ml) were taken from the recipient compartment at intervals over an 8-h period, for determination of chloramphenicol on the basis of absorption at 277 nm (Shimadzu UV-240, Kyoto, Japan); in each case, recipient medium volume was immediately made up with isoosmotic solution. Diffusion coefficients were estimated by applying the Higuchi equation [18,30]: Q /A 5 2 C0 (D t /p )1 / 2

(5)

where Q is the amount of chloramphenicol (g) released by time t (min), A is the diffusion area (0.785 cm 2 ), C0 is the initial concentration of chloramphenicol in the formulation (2.5310 23 g / ml), and D is the diffusion coefficient (cm 2 / min).

3. Results and discussion The strong dependence on pH of the degree of

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

ionization and the conformation of carbopol led us to evaluate the carbopol / surfactant interactions at pH 4, in which the polymer can be considered as nonionized and suitable for administration, and at pH 7.4, a physiological value at which the carboxylic groups are completely ionized.

3.1. Carbopol /surfactant interactions in diluted polymer dispersions The intrinsic viscosity of Carbopol  934 (62.6 g / dl carboxylic groups, with a pKa equal to 5.8) was estimated fitting Fedors’ equation [29] to the viscosity of diluted dispersions (Table 2). This equation, which was previously used with other polyelectrolytes such as butylmethacrylate and cationic polyethyleneglycols [29], provided a good fitting to the carbopol viscosity data as shown in the r 2 -values (Table 2). Martin or Schulz–Blaschke models, which are widely used for describing the viscosity of nonionic polymers in water or ionic polymers in salt solutions, were not adequate for carbopol since they assume that reduced viscosity increases with the concentration of polymer. This is not the case for carbopol dispersions in deionized water; generally,

63

the reduced viscosity of carbopol increases as the concentration of polymer decreases, owing to an expansion of the polymer chains when the free counterions become diluted [31]. The markedly higher intrinsic viscosity of the neutralized polymer is a consequence of the rise in the hydrodynamic volume of carbopol microgel particles when the degree of ionization of the carboxylic groups increases. The manufacturer estimates the particle size of carbopol microgels as around 0.8–12 mm at acidic pH and 2–30 mm at pH 7 [32]. Table 2 shows the influence of the presence of surfactants in the medium on the intrinsic viscosity of carbopol. Two concentrations of each surfactant were selected to be below (close to the critical aggregation concentration (cac)) and above the cmc of each pure surfactant (Table 1). In the case of Pluronic F-127, the lower concentration used is approximately its cmc. At concentrations below or close to the cmc and at acidic pH, both non-ionic surfactants (Tween 80 and Pluronic F-127) caused a strong decrease in the intrinsic viscosity of carbopol, which was particularly important in the case of Pluronic F-127. The dispersions prepared with this surfactant were also remarkably cloudy (Fig. 1) as

Table 2 Effect of different concentrations of surfactants on the intrinsic viscosity of carbopol in water at 258C pH

Surfactant

Concentration (g / dl)

[h ] (dl / g)

Coefficient of determination (r 2 )

4.0

– Pluronic F-127

– 0.5 2.0 0.0075 0.05 0.1 0.3 0.002 0.3

5.558 (0.752) 0.753 (0.152) 11.53 (0.184) 0.979 (0.015) 3.382 (0.228) 4.551 (0.492) 5.363 (0.217) 1.479 (0.281) 0.451 (0.501)

0.9984 0.9913 0.9096 0.9865 0.9021 0.9964 0.9994 0.9886 0.9063

– 0.5 2.0 0.0075 0.05 0.1 0.3 0.002 0.3

17.00 (0.407) 16.77 (0.300) 13.99 (0.457) 16.53 (0.411) 18.89 (0.658) 9.090 (0.078) 8.542 (0.694) 17.62 (0.125) 1.509 (0.230)

0.9995 0.9996 0.9991 0.9954 0.9992 0.9994 0.9999 0.9983 0.9058

Tween 80 SDS Benzalkonium chloride

7.4

– Pluronic F-127 Tween 80 SDS Benzalkonium chloride

Data obtained after fitting of Fedor’s equation [29]; mean of three replicates with S.D. in parentheses.

64

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

Fig. 1. Transmittance of dilute carbopol dispersions in the presence of surfactant at a concentration below cmc (pH 4).

the result of a strong associative process between carbopol and Pluronic F-127 through hydrogen bonds as shown by the IR spectra of films (Fig. 2). IR spectra of carbopol / surfactant films at pH 4 presented a broad carbonyl group peak and a shift from 1705 cm 21 to 1703–1696 cm 21 . These observations indicate a modification of the hydrogen bond interactions among the carboxylic groups [33]. The modification is less relevant in the case of Tween 80, indicating a weak interaction. Pluronic F-127 has a special effect in this band owing to its numerous ether groups that provoke the rupture of some hydrogen bonds among carboxylic groups of carbopol and formation of new polymer / surfactant hydrogen bonds [34]. Owing to the high molecular weight of Pluronic F-127, when its concentration increases above the cmc, the aggregation process might involve, especially for low concentrations of carbopol, the formation of hemi-micelles with the participation of various carbopol microgel particles [6]. This phenomenon increases the intrinsic viscosity of the system. Tween 80 molecules are smaller and their effect above cmc was less important. Since the polymer / surfactant interactions are mainly through hydrogen bonding, it is clear that when the carboxylic groups are ionized (pH 7.4), these interactions are reduced and the

Fig. 2. IR spectra of films made of 0.50 g / dl carbopol without (pH 4 and 7.4) and with (pH 4) 0.5 g / dl surfactant.

effect of the non-ionic surfactants on intrinsic viscosity is minimal. At acidic pH, the presence of SDS caused only a slight change in the intrinsic viscosity of carbopol (Table 2). The small decrease observed with 0.1 g / dl SDS (below the cmc) is a consequence of the increase in the ionic strength of the medium, which decreases the swelling degree of the microgels owing to a salting-out effect [31,35]. A similar effect was found in chitosan dispersions in the presence of a cationic surfactant [36]. The adsorption of SDS on carbopol may also play a role. Philippova et al. [37] have shown that hydrophobic interactions among acrylic polymers and anionic surfactants are more likely to occur at acidic pH, when there is no charge repulsion. This is shown in the IR spectra as a shift of the carbonyl band of carbopol (Fig. 2). To maintain electroneutrality conditions, the adsorbed surfactant is accompanied by its counterion, which increases the osmotic pressure inside the carbopol microgel particles and causes them to swell. Never-

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

theless, the hydrophobic interaction between carbopol and SDS is very weak and no significant modification in viscosity was observed for 0.3 g / dl SDS medium (above the cmc). At pH 7.4, the effect of the surfactant is mainly as a consequence of the change in the ionic strength of the medium that significantly decreases viscosity [31]. A comparison of SDS and NaCl effect on viscosity and elasticity is shown below in carbopol concentrated dispersions. Benzalkonium chloride of any concentration caused a decrease in the intrinsic viscosity at acidic pH, and only of those above its cmc at pH 7.4. Previous studies have found that the interaction between poly(acrylic acid) and cationic surfactants is stronger when the polymer is not ionized, since a shrunk coil conformation favors the cooperative binding of the surfactant [38,39]. IR spectra showed that benzalkonium groups can promote an ionic exchange reaction with the protons of carboxylic groups, resulting in the formation of surfactant / polymer pairs (Fig. 2). The appearance of the ionized carboxylic groups band (1534 cm 21 ) indicates the capability of the ammonium groups of benzalkonium chloride to interact with the carbopol carboxylic groups through a proton exchange [40]. The lower polymer charge and the establishment of hydrophobic interactions among the tails of bonded surfactant molecules can induce a decrease in the hydrodynamic volume of carbopol [35]. This effect was particularly important for the highest concentration of benzalkonium chloride (0.3 g / dl) at both pH values, and it even caused the precipitation of the polymer in the system with the highest carbopol concentration (0.1 g / dl), in which an almost complete charge neutralization occurred.

65

carbopol / surfactant medium in which drug diffusion occurs and may also be useful to predict the behavior of nanoparticular drug carriers incorporated in polymeric dispersions [41]. In the absence of surfactant, 0.25 g / dl carbopol dispersions had, at pH 4, shear-thickening flow above 500 s 21 (Fig. 3A), which is characteristic of a system constituted by individualized fuzzball units. In contrast, at pH 7.4, the flow of the dispersions was

3.2. Concentrated carbopol dispersions The repercussions of carbopol / surfactant interactions on the rheological behavior and microstructure of concentrated carbopol dispersions were evaluated by flow rheology in a wide range of shear stresses, by oscillatory rheometry, and by dynamic light scattering, which measures the resistance to the diffusion of small probe particles. This last technique, previously applied to other polymer and / or surfactant dispersions [15,16,18] can provide accurate information about the topological structure of the

Fig. 3. Influence of the change in pH on: (A) the flow properties and (B) elastic (G9, open symbols) and viscous (G0, full symbols) moduli of 0.25 and 0.50 g / dl carbopol dispersions at 258C.

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

66

pseudoplastic because of the ionization and swelling of the microgel particles, which favor interpolymeric interactions and increase the consistency of the network. In the 0.50 g / dl carbopol dispersions, the entanglement and swelling of carbopol particles are responsible for the pseudoplastic behavior of the gels [42]. The repercussions of a temperature increase from 25 to 378C on the behavior of 0.25 and 0.50 g / dl carbopol dispersions were negligible. The two systems are easy flowing fluids at pH 4 while they undergo a phase transition at pH 7.4 and form strong gels. Therefore, they are suitable for the preparation of a gelling in situ drug dosage form. These results agree with the findings obtained by Lin and Sung [25], who proposed 0.30–0.40 g / dl carbopol dispersions as the optimum concentrations for gelling in situ. We found that below 0.20 g / dl carbopol, the ability to form gels is scarce, and that above 0.50 g / dl, gels are formed even at pH 4. Oscillatory rheometry assays (Fig. 3B) showed that 0.25 g / dl carbopol gels without surfactant were mainly viscous at pH 4 (relaxation time lower than 0.01 s). The neutralization of the carboxylic groups provided a structured network with an important elastic component (relaxation time longer than 100 s). In the 0.50 g / dl carbopol dispersions, the elastic modulus was greater than the viscous modulus at both pH values, and almost independent of the angular frequency. This is characteristic of a rigid system in which the rearrangements of the structure are prevented by the cross-links and the side-chain entanglement [43]. The swelling of carbopol microgels and the formation of a structured network at pH 7.4 was reflected in an important decrease in the

diffusion coefficients of polystyrene particles (D5 1.42310 26 cm 2 / min at pH 4; D50.98310 26 cm 2 / min at pH 7.4). However, this relative increase in microviscosity (1.5 times) was much smaller than the increase observed in macroviscosity (e.g. 175 times for a shear rate of 100 s 21 ) when the pH changes from 4 to 7.4. This observation clearly indicates that it is the microstructure rather than the bulk properties of the gel that is responsible for the diffusional behavior of the dispersions. In the presence of surfactant in concentrations ranging from the cac to above the cmc (0.01, 0.05, 0.10, and 0.50 g / dl), carbopol dispersions (0.25 and 0.50 g / dl) were, in general, homogeneous. Significant cloudiness was observed only with the highest proportion of non-ionic surfactants (0.50 g / dl) at pH 4, and with the cationic surfactant (0.10 and 0.50 g / dl) at pH 4 and 7.4 (Table 3).

3.3. Carbopol /non-ionic surfactant dispersions At pH 4, the presence of low proportions of Tween 80 or Pluronic F-127 (0.01–0.10 g / dl) caused an important increment in the viscosity of 0.25 g / dl carbopol dispersions (Table 4). In these dispersions, the shear-thickening flow disappeared and the systems became pseudoplastic for all the range of shear rates assayed. As was expected from the surfactant / carbopol interaction reflected in the intrinsic viscosity data and IR spectra of films, these dispersions are more consistent owing to the establishment of surfactant bridges among several carbopol microgel particles. The stronger connectivity is reflected in a displacement of the rheograms to

Table 3 Transmittance of carbopol / surfactant dispersions at 258C and pH 4 % Carbopol

% Surfactant

Tween 80

Pluronic F-127

SDS

Benzalkonium chloride

0.25

0.01 0.05 0.10 0.50

106 (0.3) 93.5 (0.2) 76.8 (1.2) 9.3 (0.7)

91.7 77.9 74.5 16.6

(0.4) (0.4) (0.4) (0.8)

98.4 (0.3) 87.6 (0.4) 80.9 (0.8) 67.7 (0.3)

80.4 (0.6) 2.8 (0.1) Precipitation Precipitation

0.50

0.01 0.05 0.10 0.50

99.6 (0.2) 90.4 (0.7) 77.8 (3.0) 64.2 (2.8)

97.8 94.2 90.2 35.0

(0.3) (0.7) (0.4) (0.3)

99.5 (1.3) 94.3 (0.8) 92.4 (0.1) 77.2 (0.2)

91.6 (0.1) 12.0 (0.1) 0.6 (0.1) Precipitation

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

67

Table 4 Consistency index (m) and fluidity index (n) calculated by Ostwald fitting of 0.25 or 0.50 g / dl carbopol dispersions with different concentrations of surfactant, at pH 4 and 258C Carbopol

Surfactant

Parameter

Concentration of surfactant 0.01 g/dl

0.05 g/dl

0.10 g/dl

0.50 g/dl

Pluronic 0.25 g/dl a

(.646 s 21 ) 6.7310 25 1.89 0.9971 (.468 s 21 ) 9.6310 24 1.45 0.9977 (.351 s 21 ) 1.0310 23 1.43 0.9979

0.27 0.64 0.9921

0.11 0.76 0.9980

0.30 0.62 0.9928

m n r2

0.44 0.58 0.9992 (,360 s 21 ) (.360 s 21 ) 0.03 4.5310 24 0.88 1.55 0.9958 0.9995 (,360 s 21 ) (.360 s 21 ) 0.03 4.5310 24 0.88 1.55 0.9958 0.9995

0.33 0.61 0.9999 (,286 s 21 ) (.286 s 21 ) 0.01 4.7310 24 0.92 1.53 0.9970 0.9945

m n r2

0.34 0.61 0.9994 (,679 s 21 ) (.679 s 21 ) 0.19 2.1310 23 0.66 1.36 0.9996 0.9978 (,679 s 21 ) (.679 s 21 ) 0.19 2.1310 23 0.66 1.36 0.9996 0.9978

Precipitation

Precipitation

m n r2

8.15 0.38 0.9736

12.3 0.33 0.9870

10.1 0.35 0.9807

6.40 0.38 0.9833

m n r2

27.5 0.18 0.9631

12.4 0.34 0.9866

11.9 1.34 0.9834

8.92 0.37 0.9730

m n r2

11.4 0.34 0.9740

m n r2

7.61 0.36 0.9757

6.07 0.38 0.9780 (,835 s 21 ) (.835 s 21 ) 0.27 1.1310 23 0.65 1.49 0.9993 0.9857

3.76 0.40 0.9762 (,464 s 21 ) (.464 s 21 ) 0.35 9.4310 24 0.85 1.45 0.9996 0.9980

3.66 0.41 0.9816 (,266 s 21 ) (.266 s 21 ) 0.40 1.3310 23 0.16 1.58 0.9850 0.9994

Tween 80

SDS m n r2 Benzalk. chloride

(,646 s 21 ) 0.11 0.73 0.9995 (,468 s 21 ) 0.42 0.82 0.9994 (,351 s 21 ) 0.02 0.87 0.9987

m n r2

Pluronic 0.50 g/dl b

Tween 80

SDS

Benzalk. chloride

Values are means of three replicate samples. b Reference value: 0.50 g / dl carbopol: m513.6, n50.33, r 2 50.9656. a Reference value: 0.25 g / dl carbopol: m50.042, n50.82, r 2 50.9994 (shear rates below 499 s 21 ), and m52.9310 23 , n51.32, r 2 50.9960 (shear rates above 499 s 21 ).

higher values of viscous (G0) and elastic (G9) moduli (up to 1–2 orders of magnitude higher than carbopol dispersion alone, Fig. 4) with a dependence on the angular frequency similar to that observed with carbopol alone dispersions (Fig. 3B). In the case of 0.50 g / dl carbopol dispersions, since the polymer microgels are already in contact, the main effect of the surfactant was not on consistency or G0 but on G9. The maxima for G9 and G0 were reached with 0.01–0.05 g / dl surfactant and then the moduli decreased gradually as surfactant concentration in-

creased. The decrease in viscosity was maximum in 0.50 g / dl of surfactant (above the cmc, Table 1) systems (Table 3). In contrast, microviscosity, calculated from the diffusion coefficients of polystyrene particles, was not affected by the addition of 0.01 g / dl surfactant but increased considerably when surfactant concentration increased (Fig. 5). This different behavior may be explained as follows: when very small amounts of Tween 80 or Pluronic F-127 (below the cmc) are added to carbopol dispersions, surfactant molecules form interpolymeric mi-

68

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

Fig. 4. Elastic (G9) and viscous (G0) moduli of carbopol / non-ionic surfactant aqueous systems at 0.88 rad / s (continuous line, pH 4; dotted line, pH 7.4).

celles, and create a three-dimensional network of greater consistency than that created by carbopol dispersions alone. In doing so, the free volume among the carbopol particles is scarcely altered, and, in consequence, the diffusion of polystyrene particles is not reduced. When more surfactant molecules are added, the formation of intrapolymeric micelles becomes possible and the interpolymeric connections diminished, and the three-dimensional network disintegrates [6]. A schematic drawing of this process is shown in Fig. 6. Despite the decrease in macroviscosity, the diffusion of polystyrene particles became more difficult, which indicates that the formation of larger carbopol / surfactant aggregates or the appearance of free micelles make the medium more tortu-

ous, increasing the diffusion path length [16]. The hypothesis about the formation of individualized large carbopol / surfactant aggregates is also supported by the change from pseudoplastic to shear-thickening flow observed in the 0.25 g / dl carbopol / 0.50 g / dl Tween 80 systems, at 408–468 s 21 , and in the presence of 0.50 g / dl Pluronic F-127, around 562– 646 s 21 (Table 3). The dependence of the gel behavior on shear conditions may be explained as follows. First, an initial rearrangement of the polymeric chains towards the flow lines corresponds to the pseudoplastic region. Second, shear-thickening behavior is induced by: (a) the disintegration of the interpolymeric bridges that causes the three-dimensional network to disappear (carbopol / surfactant

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

69

size with a lower number of contacts among them, which are easy to break at high shear rates. At pH 7.4, the effects of the surfactants on G0 and G9 were similar to although less intense than those observed at pH 4. All dispersions presented a pseudoplastic flow similar to carbopol dispersions without surfactant. For all 0.25 g / dl carbopol / surfactant systems, when the pH changed from 4 to 7.4, there was an additional increase in the consistency index and G0 (50–70 times) and in G9 (four orders of magnitude), reaching values similar to those of carbopol gel alone. In the case of 0.50 g / dl carbopol, the pH change increases the consistency index, G9 and G0 6–10 times, both in the presence or absence of surfactant. Therefore, all gels maintain their gelling in situ capacity.

3.4. Carbopol /anionic surfactant dispersions

Fig. 5. Effects of the surfactants on the relative diffusion coefficients of polystyrene nanospheres (162 nm) in 0.25 g / dl carbopol dispersions (D0 51.72310 26 cm 2 / min).

aggregates exert a similar effect to particles in suspension; [44]); and (b) the surfactant desorption, more likely above the cmc (separately, the behavior of surfactant micelles and carbopol microgels is shear-thickening; [45]). Although in the cases of Tween 80 and Pluronic F-127, both mechanisms might occur, desorption may be more significant in the case of Tween 80, which interacts weakly with carbopol [24]. Pluronic interacts strongly with carbopol through hydrogen-bonding at pH 4 (Table 3 and Fig. 2; [34]). Above the cmc, the formation of hemi-micelles of Pluronic involving only one carbopol macromolecule create aggregates of a greater

The presence of SDS caused a decrease in the consistency of the systems at pH 4 and 7.4 (Table 4 and Fig. 7), except with the lowest concentration of surfactant (0.01 g / dl) that caused a slight increase in viscosity and elasticity (Fig. 8). All gels retained their pH-dependent gelling ability. The effect of this surfactant on carbopol dispersions may be attributed to two opposing mechanisms: hydrophobic adsorption and salt effect. Philippova et al. [37] have shown that SDS can be adsorbed on to hydrophobically-modified poly(acrylic acids). Nevertheless, this interaction might be scarce in the case of carbopol owing to its more hydrophilic character. On the other hand, the small molecular size and anionic character of SDS cause an increase in the ionic strength of the medium, which produces a shielding effect among the anionic charges of carbopol and its dehydration. As a consequence, the microgel particles shrink [35]. Therefore, the greater consistency of 0.01 g / dl surfactant systems might be attributed to a hydrophobic polymer / surfactant interaction, while the ionic strength effect appears with higher surfactant concentrations and is responsible for the considerable decrease in viscosity, G9 and G0. In 0.25 g / dl carbopol solutions this phenomenon is shown by the low value at which shear-thickening behavior appears (Fig. 7), which corroborates the hypothesis of the shrinking of the microgel particles due to the

70

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

Fig. 6. Schematic drawing of carbopol / surfactant interaction phenomena and of their repercussions on the volume of carbopol microgel particles in the aqueous medium.

shielding of the charges and the dehydration of the polymer. For further confirmation, a 0.50 g / dl carbopol dispersion was prepared in 0.10 g / dl NaCl (molarly equivalent to 0.50 g / dl SDS), at pH 7.4, and a decrease in G9 (from 396 to 65 Pa at 1 rad / s) and G0 (from 28.5 to 6.1 Pa at 1 rad / s) was observed. This decrease was even higher than in the case of 0.50 g / dl SDS (G95244 Pa, G0518 Pa, at 1 rad / s), which can be attributed to the lower saltingout efficiency of SDS compared to NaCl [46]. SDS hydrophobic adsorption can also play a role balancing the salt effect. Therefore, both hydrophobic adsorption and salt effect are responsible for the rheological behavior of carbopol / SDS dispersions. Under these conditions, the free volume among carbopol microgels increases and the microstructure of the systems is less tortuous and, as a consequence,

the diffusion of polystyrene particles occurs more easily as SDS concentration increases at both pHvalues studied (Fig. 5).

3.5. Carbopol /cationic surfactant dispersions The addition of 0.01 g / dl benzalkonium chloride (potentially able to neutralize less than 1 mol% of carbopol) had a negligible influence on the viscoelastic behavior of carbopol dispersions. In contrast, a 0.10 or a 0.50 g / dl surfactant (10–40 mol% of the amount needed to neutralize all carboxylic groups) produced an insoluble complex in 0.25 g / dl carbopol dispersions at pH 4, and the precipitation made the rheological evaluation of these systems impossible. In the case of 0.50 g / dl carbopol dispersions (neutralization degree 5–20 mol%), no precipitation

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

71

Fig. 7. Influence of ionic surfactants concentration (% w / v) on flow properties at pH 4 of 0.25 and 0.50 g / dl carbopol dispersions.

occurred but strong decreases in viscosity, G9 and G0 were observed (Table 4; Figs. 7 and 8). Philippova et al. [5] observed that poly(methacrylic acid) gels collapse at acidic pH upon absorption of only 4 mol% of cetylpyridinium chloride, and Ashbaugh et al. [47] reported that the swelling of poly(acrylic acid) gels decreases linearly with the fraction of dodecyl trimethylammonium bromide absorbed by the gel, and that for an amount of surfactant bonded equivalent to a 40 mol% of the carboxylic groups of the gel, this almost completely collapsed. In our case, the interaction between carbopol and a cationic surfactant involves, as was mentioned, hydrophobic and electrostatic forces. As a consequence of the electrostatic binding, the surfactant heads interchange with protons of carbopol carboxylic groups, which reduces the internal osmotic pressure of the

gel [38,40]. It also favors the hydrophobic binding of the tails among themselves and with the polymer backbone [37]. The aggregation phenomenon favored, at acidic pH, the appearance of shear-thickening flow even in the case of 0.50 g / dl carbopol dispersions. The aggregation process is mainly governed by the cationic / anionic groups ratio in the system more than by the fact that the surfactant concentration was below or above the cmc. As a consequence of this associative process, the potentially toxic effects of the surfactant could be reduced [40]. On the other hand, the aggregates created contribute to obstruct the medium, which makes the diffusion of polystyrene particles more difficult than in the absence of benzalkonium chloride (Fig. 5), again indicating a discrepancy between macro- and microviscosity.

72

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

prevented the precipitation process in 0.25 g / dl carbopol dispersions with 0.10 g / dl benzalkonium chloride.

3.6. Chloramphenicol release rate Since the size of the diffusant has a decisive influence on the diffusion process [48], experiments using a model drug were also carried out. We tested the diffusion rate of a small drug molecule, at the therapeutic dose, in the systems with the lowest proportion of surfactants that provided the maximum viscous and elastic moduli. Fig. 9 shows the diffusion coefficients estimated by the Higuchi model [30]. Eq. (5) is a simplified solution of Fick’s second law, which is valid to fit the diffusion profile of the first 30% of release from a planar surface, when the diffusant is completely dissolved in the donor gel [49]. The presence of 0.01 g / dl surfactant, non-ionic or ionic, did not cause a significant change in the chloramphenicol release rate from 0.25 g / dl carbopol dispersions, as was observed using polystyrene particles. This confirms that the diffusion process of chloramphenicol was not affected by the increase in macroviscosity caused by the presence of surfactant at any of the pH-values studied. The swelling of

Fig. 8. Elastic (G9) and viscous (G0) moduli of carbopol / ionic surfactant aqueous systems.

At pH 7.4, the effects were less intense but with the same tendency as at pH 4. The presence of sodium ions and the swelling of the microgels

Fig. 9. Effect of 0.01 g / dl surfactant on the relative diffusion coefficient of chloramphenicol in 0.25 g / dl carbopol dispersions at pH 4 (s) and pH 7.4 (d). D0 54.65310 211 cm 2 / min.

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75

carbopol microgel particles caused a decrease in the diffusion coefficients when pH rose from 4 to 7.4. The diffusion coefficient values obtained suggest that the systems can be useful in the control of drug release in a biological environment.

73

Acknowledgements This work was supported by the Xunta de Galicia (PGIDT 00PX120303PR). The authors also express their gratitude to Xunta de Galicia for an equipment grant (DOG 04 / 06 / 97) and to BFGoodrich Europe for providing free samples of Carbopol  .

4. Conclusions The addition of low proportions of surfactants commonly used in drug formulation has an important effect on the flow type, G9, G0, and microviscosity of carbopol gels at pH 4 and 7.4. Depending on whether a site-specific application is required or whether the drug has to be dispersed rapidly in a wide mucosal space, the formulation should be sufficiently consistent and elastic to remain at the application site or be easily spreadable, respectively [27,50]. The increases in viscosity, elasticity, and microviscosity observed with Tween 80, Pluronic F-127 or SDS below the cmc can be very useful to achieve a long mucosal and ocular contact time and may contribute to controlling the release at acidic pH (for example, for intravaginal delivery) or / and until gellation occurs (for ocular, rectal, or other physiological environments of neutral–basic conditions). Indeed, the pseudoplastic behavior obtained with the nonionic surfactants below cmc (instead of shear-thickening as happens with carbopol alone) constitutes another important advantage in the case of an ophthalmic formulation. The high viscosity at low shear rates of these systems might avoid drainage, while the decrease in viscosity at higher shear rates prevents irritation during blinking [11,23]. On the other hand, it is possible to obtain shear-thickening flow just by increasing the concentration of the non-ionic surfactants above the cmc or adding low proportions of SDS or benzalkonium chloride, which may be useful for prolonging the permanence of the dispersion on a surface continuously stressed by skin flexion. The diffusion results obtained with polystyrene nanoparticles and chloramphenicol showed that the effects of the surfactants on macroviscosity are different, and sometimes opposed, to those observed in microviscosity, emphasizing the need to evaluate this parameter to predict the release behavior of the formulation.

References [1] S.G. Deshpande, S. Shirolkar, Sustained-release ophthalmic formulations of pilocarpine, J. Pharm. Pharmacol. 41 (1989) 197–200. [2] S. Ding, Recent developments in ophthalmic drug delivery, Pharmaceutical Science & Technology Today 1 (1998) 328– 335. [3] M. Yonese, Sustained drug delivery by gels, in: Y. Osada, K. Kajiwara (Eds.), Gels Handbook, Vol. 3, Academic Press, San Diego, CA, 2001, pp. 230–240. [4] S. Kumar, B.O. Haglund, K.J. Himmelstein, In situ-forming gels for ophthalmic drug delivery, J. Ocul. Pharmacol. 10 (1994) 47–56. [5] O.E. Philippova, L.A. Chtcheglova, N.S. Karybiants, A.R. Khokhlov, Two mechanisms of gel / surfactant binding, Polym. Gel Networks 6 (1998) 409–421. [6] P. Hansson, B. Lindman, Surfactant–polymer interactions, Curr. Opin. Colloid. Interf. Sci. 1 (1996) 604–613. ¨ [7] H. Walderhaug, B. Nystrom, F.K. Hansen, B. Lindman, Surfactant–polymer interaction in thermo-induced gelling systems of ethyl (hydroxyethyl) cellulose (EHEC) studied by NMR self-diffusion, dynamic light scattering and rheology, Prog. Colloid. Polym. Sci. 98 (1995) 51–56. [8] L. Bromberg, M. Temchenko, R.H. Colby, Interactions among hydrophobically modified polyelectrolytes and surfactants of the same charge, Langmuir 16 (2000) 2609–2614. [9] E. Kokufuta, Y.Q. Zhang, T. Tanaka, A. Mamada, Effects of surfactants on the phase transition of poly(N-isopropylacrylamide) gel, Macromolecules 26 (1993) 1053– 1059. [10] J.S. Chu, R. Chandrasekharan, L.A. Gordon, D.W. Norman, H.G. Arthur, Viscometric study of polyacrylic acid systems as mucoadhesive sustained-release gels, Pharm. Res. 11 (1991) 1408–1412. [11] K. Edsman, J. Carlfors, K. Harju, Rheological evaluation and ocular contact time of some carbomer gels for ophthalmic use, Int. J. Pharm. 137 (1996) 233–241. ´ [12] R. Duro, C. Alvarez-Lorenzo, C. Souto, J.L. Gomez-Amoza, ´ R. Martınez-Pacheco, A. Concheiro, Effects of the interactions between cellulose ethers and polysorbate 80 on the stability of pyrantel palmoate suspensions, Chem. Pharm. Bull. 46 (1998) 1421–1427. [13] O.I. Corrigan, A.M. Healy, Surfactants in pharmaceutical products and systems, in: J. Swarbrick, J.C. Boilan (Eds.),

74

[14] [15]

[16]

[17] [18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75 Encyclopedia of Pharmaceutical Technology, Vol. 14, Marcel Dekker, New York, 1996, pp. 295–331. B. Amsden, Solute diffusion within hydrogels. Mechanisms and models, Macromolecules 31 (1998) 8382–8395. K. Al-Khamis, S.S. Davis, J. Hadgraft, Microviscosity and drug release from topical dispersion formulations, Pharm. Res. 3 (1986) 214–217. P.A. Hassan, C. Manohar, Diffusion of probe particles in surfactant gels: dynamic light scattering study, J. Phys. Chem. B 102 (1998) 7120–7125. G. Lu, H.W. Jun, Diffusion studies of methotrexate in carbopol and poloxamer gels, Int. J. Pharm. 160 (1998) 1–9. ´ ´ C. Alvarez-Lorenzo, J.L. Gomez-Amoza, R. Martınez-Pacheco, C. Souto, A. Concheiro, Microviscosity of hydroxypropylcellulose gels as a basis for prediction of drug diffusion rates, Int. J. Pharm. 180 (1999) 91–103. M. Dittgen, M. Durrani, K. Lehmann, Acrylic polymers. A review of pharmaceutical applications, Stp Pharma Sci. 7 (1997) 403–437. R. Ilango, S. Kavimani, K.S. Kumar, K.R. Deepa, B. Jaykar, Formulation and evaluation of transdermal preparations of nimesulide gel, East. Pharm. 41 (1998) 123–125. K. Morimoto, I. Hama, Y. Nakamoto, T. Takeeda, E. Hirano, K. Morisaka, Pharmaceutical studies on polyacrylic acid aqueous gel bases: absorption of insulin from polyacrylic acid aqueous gel bases following rectal administration in alloxan diabetic rats and rabbits, J. Pharm. Dyn. 3 (1980) 24–32. T. Ozeki, H. Yuasa, Y. Kanaya, Controlled release from solid dispersion composed of poly(ethylene oxide)-Carbopol interpolymer complex with various cross-linking degrees of Carbopol, J. Controlled Release 63 (2000) 287–295. M. Oechsner, S. Keipert, Polyacrylic acid / polyvinylpyrrolidone bipolymeric systems. I. Rheological and mucoadhesive properties of formulations potentially useful for the treatment of dry-eye-syndrome, Eur. J. Pharm. Biopharm. 47 (1999) 113–118. S. Simovic, S. Tamburic, J. Milic-Askrabic, D. Rajic, An investigation into interactions between polyacrylic acid polymers and a non-ionic surfactant: an emulsion preformulation study, Int. J. Pharm. 184 (1999) 207–217. H-R. Lin, K.C. Sung, Carbopol / Pluronic phase change solutions for ophthalmic drug delivery, J. Controlled Release 69 (2000) 379–388. P. Furrer, B. Plazonnet, J.M. Mayer, R. Gurny, Application of in vivo confocal microscopy to the objective evaluation of ocular irritation induced by surfactants, Int. J. Pharm. 207 (2000) 89–98. ¨ ¨ H. Hagerstrom, M. Paulsson, K. Edsman, Evaluation of mucoadhesion for two polyelectrolyte gels in simulated physiological conditions using a rheological method, Eur. J. Pharm. Sci. 9 (2000) 301–309. United States Pharmacopoeia 24, National Formulary 19, The United States Pharmacopeial Convention Inc., Rockville, MD, 2000, pp. 2426–2427. L. Ghimici, F. Popescu, Determination of intrinsic viscosity for some cationic polyelectrolytes by Fedors method, Eur. Polym. J. 34 (1998) 13–16.

[30] W.I. Higuchi, Analysis of data on the medicament release from ointments, J. Pharm. Sci. 51 (1962) 802–804. [31] J.O. Carnali, M.S. Naser, The use of dilute solution viscometry to characterize the network properties of carbopol microgels, Colloid Polym. Sci. 270 (1992) 183–193. [32] B.F. Goodrich, Bulletin 17, Brecksville, OH, 1994. [33] J. Dong, Y. Ozaki, K. Nakashima, Infrared, Raman, and near-infrared spectroscopic evidence for the coexistence of various hydrogen-bond forms in poly(acrylic acid), Macromolecules 30 (1997) 1111–1117. [34] M.L. Cole, T.L. Whateley, P.J. Harrison, Interaction of nonionic block co-polymeric (poloxamer) surfactants with poly(acrylic acid) and the application in the stabilization of multiple emulsions, J. Pharm. Pharmacol. 42 (1990) 136P. [35] C.H. Jeon, E.E. Makhaeva, A.R. Khokhlov, Swelling behavior of polyelectrolyte gels in the presence of salts, Macromol. Chem. Phys. 199 (1998) 2665–2670. ¨ [36] A-L. Kjøniksen, B. Nystrom, C. Iversen, T. Nakken, O. Palmgren, T. Tande, Viscosity of dilute aqueous solutions of hydrophobically modified chitosan and its unmodified analogue at different conditions of salt and surfactant concentration, Langmuir 13 (1997) 4948–4952. [37] O.E. Philippova, D. Hourdet, R. Audebert, A.R. Khokhlov, Interaction of hydrophobically modified poly(acrylic acid) hydrogels with ionic surfactants, Macromolecules 29 (1996) 2822–2830. [38] T. Shimizu, Changes of pH and counterion activity during the binding process of cationic surfactants to carboxylic polyions, Colloid. Surf. A 84 (1994) 239–248. [39] V.A. De Oliveira, M.J. Tiera, M.G. Neumann, Interaction of cationic surfactants with acrylic acid-ethyl methacrylate copolymers, Langmuir 12 (1996) 607–612. [40] C. Coquelet, N. Lakhchaf, B. Pages, M. Persin, L.S. Rao, J. Sarrazin, G. Tarrago, Association between benzalkonium chloride and a poly(acrylic acid) dispersion. Study by microfiltration and membrane dialysis, J. Membr. Sci. 120 (1996) 287–293. [41] A. Bochot, E. Fatal, A. Gulik, G. Couarraze, P. Couvreur, Liposomes dispersed within a thermosensitive gel: a new dosage form for ocular delivery of oligonucleotides, Pharm. Res. 15 (1998) 1364–1369. [42] B.W. Barry, M.C. Meyer, The rheological properties of carbopol gels. I. Continuous shear and creep properties of carbopol gels, Int. J. Pharm. 2 (1979) 1–25. [43] S. Tamburic, D.Q.M. Craig, The effects of ageing on the rheological, dielectric, and mucoadhesive properties of poly(acrylic acid) gel systems, Pharm. Res. 13 (1996) 279–283. [44] Y. Otsubo, Effect of adsorption affinity of polymers on the viscosity behavior of suspensions, J. Colloid Interf. Sci. 204 (1998) 214–216. [45] L-S. Choi, O-K. Kim, Fluorescence probe studies of poly(acrylic acid)–cationic surfactant interactions following high shear, Langmuir 10 (1994) 57–60. [46] N. Sarkar, Thermal gelation properties of methyl and hydroxypropyl methylcellulose, J. Appl. Polym. Sci. 24 (1979) 1073–1087. [47] H.S. Ashbaugh, L. Piculell, B. Lindman, Interactions of cationic / nonionic surfactant mixtures with an anionic hydro-

R. Barreiro-Iglesias et al. / Journal of Controlled Release 77 (2001) 59 – 75 gel: absorption equilibrium and thermodynamic modeling, Langmuir 16 (2000) 2529–2538. [48] J.H. Smidt, D.J.A. Crommelin, Viscosity measurement in aqueous polymer solutions by dynamic light scattering, Int. J. Pharm. 77 (1991) 261–264. [49] W.J. Addicks, N.D. Weiner, R.L. Curl, G.L. Flynn, Drug delivery from topical formulations: theoretical prediction and experimental assessment, in: D.W. Osborne, A.H. Amann (Eds.), Topical Drug Delivery Formulations, Vol. 42, Marcel Dekker, New York, 1992, pp. 221–243. [50] A.D. Woolfson, R.K. Malcolm, R. Gallagher, Drug delivery by the intravaginal route, Crit. Rev. Ther. Drug Carrier Syst. 17 (2000) 509–555.

75

[51] R. Barreiro-Iglesias, Influencia de algunos tensoactivos sobre ´ las propiedades reologicas de dispersiones de Carbopol 934NF, Minor Thesis, Department of Pharmacy and Pharmaceutical Technology, Universidad de Santiago de Compostela, Spain, 1999. [52] M.R. Porter, Handbook of Surfactants, 2nd Edition, Blackie Academic & Professional, Wales, 1994, pp. 118, 226, 250. [53] P. Alexandridis, J.E. Holzwarth, T.A. Hatton, Micellization of poly(ethylene oxide)-poly(propylene oxide)-poly(ethylene oxide) triblock copolymers in aqueous solutions: thermodynamics of copolymer association, Macromolecules 27 (1994) 2414–2425.