Relation between polymer composition solvent and nonsolvent

mechanism of coacervate formation in a non-aqueous system has been investigated by Ruiz etal. (1989). Microcapsules of poly (DL-lactic acid-co-glycolic acid) were prepared by dissolving the polymer in methylene chloride and then adding various quantities of silicone oil to effect phase separation. The phase separation phenomenon was observed by taking photomicro graphs after increasing quantities of the incompatible polymer were added. At the first step when the amount of phase inducer is low (1-5%), a pseudoemulsion of the silicone liquid is formed. During the second step when more silicone oil is added, the beginning of the phase separation appears. The coacervate droplets appear to be unstable and merge together then break apart. In the third step the added quantity of silicone oil is sufficient to permit a stable dispersion of polymer coacervate droplets; this step is called the stability window. Finally, the fourth step occurs after further addition of silicone oil, which causes extensive aggregation of the coacervate droplets.

Four polymers with different compositions of lactic acid (LA) and glycolic acid (GA) were studied. The polymers with the highest percentage of lactic acid had the largest stability window, and required the largest amounts of silicone oil to reach that region. The polymer with the lowest content of lactic acid, namely 50% LA and 50% GA, is the least hydrophobic among the polymers studied and did not easily dissolve in the solvent and required only small amounts of silicone oil to effect phase separation, accounting for the small stability window. Silicone oil with a low viscosity did not yield a stability window with four polymers tested; however, silicone oils with higher viscosity, up to 12 500, increase the size of the stability window for all polymers. The stability window was increased by increasing the solubility of the poly(DL-lactic acid-co-glycolic acid) polymer in the solvent by the addition of methanol. As a result, more silicone oil was needed to reach the stability window, and thus to induce the appearance of the polymer droplets. The width of the stability window can be altered by changing the viscosity of the silicone oil and modifying the solvent for the polymer by adding a suitable percentage of a better solvent.

Shively and McNickle (1991) considered the effect of various solvent compositions on the coacervation process. Ternary phase diagrams were prepared using a biodegradable block copolymer prepared from tartaric acid and 1,10-decanediol, and ethanol and water, with or without NaCl. Microcapsules of kaolin or hydrocortisone-21-acetate were prepared by adding the core to an ethanol solution of the polymer and then titrating with the aqueous non-solvent. The microcapsules were then filtered and dried. At high polymer concentrations in the non-plait region, minimal or no solvent interaction occurred and the polymer was in the coiled configuration. In the plait region at low polymer concentration, the ionic strength of the non-solvent showed an effect on the coacervate, the adhesive forces were greater than the cohesive forces, and the polymer adopted a more linear configuration. Surface tension measurements, when the solvent composition was 30% water and 70% alcohol (non-plait) or 50% alcohol and 50% water (plait), showed that the area per polymer molecule decreased in post-coacervation compared with the precoacervation region. These results agree with the theory that the coacervation results in the reduction of the surface free energy of a system through a reduction of the molecular surface area. Analysis of the surface tension versus polymer composition graphs shows that coacervate phases resulting from 30% water, 70% ethanol compared with 50% water, 50% ethanol were very different suggesting differences in molecular configuration and interaction properties. Thus, the authors speculate that microcapsules made with different coacervation conditions would have different properties, such as diffusion or morphology. It was found that microcapsules produced with non-plait conditions had considerably slower rates of release of the drug and had rough and irregular surfaces compared with microcapsules prepared with plait conditions.

Solubility parameters. Robinson (1989) determined the solubility of ethylcellulose Type N10 in 122 solvents qualitatively and in 36 solvents quantitatively. The contribution of dispersive, polar and hydrogen-bonding intermolecular forces was determined and plotted on two-dimensional and triangular solubility graphs. The influence of dipole-dipole interactions on the solubility of ethylcellulose was shown by plotting the fractional polarity of the solvent against the solubility parameter. The diagram shows that ethylcellulose is soluble over a range of polarity from 0 to 0.75, but it is not soluble in solvents with either a low or high solubility parameter. In order to show the effect of the relative fractional contributions of the hydrogen bonding, polar and dispersion components, a triangular solubility diagram was prepared. The solvents were classified on their hydrogen bonding ability: weak, medium, and strong. The three areas of solubility overlap and they define a region which determines the intermolecular forces appropriate to dissolve ethylcellulose. Ethylcellulose occupies a central position within the defined solubility regions. The triangular diagram is useful for determining good solvents and non-solvents for microencapsulation purposes. Coacervation was observed after cooling a solution of ethylcellulose in a poor solvent which had a solubility parameter near, or just outside, the solubility region for the polymer. Gelation occurs after cooling with liquids usually considered non-solvents for the polymer and their solubility parameters are well outside the solubility region and have higher interaction parameters. Flocculation was observed after cooling solutions of the polymer in polar solvents where large values of the interaction parameters occur.

The selection of appropriate solvents and non-solvents may be ascertained through the use of solubility parameters. In a study by Moldenhauer and Nairn (1992), it was shown that microcapsules could be prepared by phase separation using a number of solvent-non-solvent pairs. The solubility parameter map was prepared using a number of solvents, both singly and in mixtures, to provide regions where the polymer ethylcellulose was

Fig. 3 Solubility parameter map for ethylcellulose showing the solubility border and the initial and final microencapsulation solubility parameters: □ using ethyl acetate and cyclohexane; ® using methyl ethyl ketone and cyclohexane; A using toluene and light liquid paraffin; ■ using ethyl acetate, cyclohexane and light liquid paraffin; A using methyl ethyl ketone, cyclohexane and light liquid paraffin; • using ethyl acetate and light liquid paraffin; and O using methyl ethyl ketone and light liquid paraffin. Reproduced with permission from Moldenhauer and Nairn (1992), I. Controlled Release 22, 205-218. Elsevier Science Publishers BV, The Netherlands.

Fig. 3 Solubility parameter map for ethylcellulose showing the solubility border and the initial and final microencapsulation solubility parameters: □ using ethyl acetate and cyclohexane; ® using methyl ethyl ketone and cyclohexane; A using toluene and light liquid paraffin; ■ using ethyl acetate, cyclohexane and light liquid paraffin; A using methyl ethyl ketone, cyclohexane and light liquid paraffin; • using ethyl acetate and light liquid paraffin; and O using methyl ethyl ketone and light liquid paraffin. Reproduced with permission from Moldenhauer and Nairn (1992), I. Controlled Release 22, 205-218. Elsevier Science Publishers BV, The Netherlands.

soluble (that is, gave clear solutions at definite concentrations) and regions where the polymer was insoluble (that is, where clear solutions were not obtained). This information was then used to prepare microcapsules of theophylline ion-exchange resin beads coated with ethylcellulose using a number of solvents. Partial evaporation of the solvent in a mixture leads to a change in solubility parameters effecting phase separation. These authors experimentally corroborated Robinson's (1989) studies that microencapsulation systems should be near the limit of ethylcellulose solubility where coacervation will occur and showed that microcapsules could be prepared by controlled evaporation of a solvent-non-solvent pair for ethylcellulose. The total amount evaporated was the same in all experiments. As evaporation of the solvent and non-solvent took place, the solubility parameter of the mixture generally changed, owing to the different vapour pressures, into a poor solvent for the polymer. The composition of the solvent pair, both before and after evaporation, was related to the phase diagrams and the solubility parameter map. Well-formed microcapsules were prepared from solvent-non-solvent pairs whose solubility parameters changed during evaporation from the soluble region to just at the other side or at the edge of the solubility region on the solubility parameter map. Even though different solvent mixtures were used, ethyl acetate or methyl ethyl ketone as solvents and non-solvents cyclohexane and light liquid paraffin, the solubility parameters were similar and evaporation produced microcapsules with similar characteristics (see Fig. 3).

Experiments conducted with solvents that had a poor solubility parameter for ethylcellulose yielded either no coat or a coat of poor quality. Evaporation of solvent pairs which had similar vapour pressures and which were in the solution region of the solubility parameter map did not change their solubility parameter during evaporation, and microencapsulation did not take place. Solubility parameter maps provide information about a number of solvent-non-solvent pairs whereas a phase diagram provides information about only one solvent-non-solvent pair.

Wall formation

The mechanism of wall film formation of ethylcellulose onto magnesium aluminium hydroxide hydrate was investigated by Kasai and Koishi (1977). Four different experiments were carried out in order to investigate the phenomenon of microencapsulation. With increasing amounts of ethylcellulose, in dichloromethane, added to the core and also addition of water, it was shown that the surface properties of the core changed from hydro-philic to hydrophobic, likely as a result of adsorption of the ethylcellulose onto the surface of the core. Photographs of ethylcellulose coacervate drops formed by the addition of increasing volumes of w-hexane to the ethylcellulose solution show an increase in size of the coacervate drops which corresponds to an increase in the weight of ethylcellulose in the coacervate. The authors indicate that the smaller-sized coacervate droplets are likely to be suitable for microencapsulation, and the larger-size coacervate droplets are not. The authors also relate the surface structure of the microcapsules to the weight of ethylcellulose coacervate as a result of the addition of nonsolvent.

The authors then suggest possible cross-sectional models for the deposition of ethylcellulose coacervate drops on the core and in the final state of the walls as shown in Fig. 4. It is noted in Fig. 4 that there is compression of the ethylcellulose on the core material and this is supported by the fact that despite an increase in percentage ethylcellulose concentration in the microcapsules from 27 to 46%, the wall thickness range is almost constant within the range of 15-17.5/¿m. In conclusion, the authors postulate a model for deposition of ethylcellulose coacervate based on the amount of coacervate on the core, as a result of increasing amounts of non-solvent added, scanning electron microscopy of the microcapsule surface, photographs of the coacervate droplets and the region of constant wall thickness.

STEP I STEP H

STEP I STEP H

Fig. 4 Possible cross-sectional models for the deposition of ethylcellulose coacervate drops on the core material and the final state of microcapsule walls. Key: step I: coacervate drops deposited at first; step II: final walls; ® ®and ®: different stages, C, E, and M, coacervate drops, ethylcellulose adsorbed initially on the core material, and the core material. Reproduced with permission from Kasai and Koishi (1977), Chem. Pharm. Bull. 25(2), 314-320. Pharmaceutical Society of Japan.

Fig. 4 Possible cross-sectional models for the deposition of ethylcellulose coacervate drops on the core material and the final state of microcapsule walls. Key: step I: coacervate drops deposited at first; step II: final walls; ® ®and ®: different stages, C, E, and M, coacervate drops, ethylcellulose adsorbed initially on the core material, and the core material. Reproduced with permission from Kasai and Koishi (1977), Chem. Pharm. Bull. 25(2), 314-320. Pharmaceutical Society of Japan.

Incompatible or non-wall-forming polymers

A number of papers have been written about the effect of polyisobutylene on the coacervation of ethylcellulose and the formation of microcapsules. In an early paper Donbrow and Benita (1977) describe ethylcellulose coacervation by dissolving the polymer in cyclohexane and slowly cooling with controlled agitation. Phase separation occurs over 24 h to yield a lower phase of coacervate droplets and a clear upper layer containing polyisobutylene. The authors noted the non-linear increase in volume of the coacervate with polyisobutylene concentration. At the same time, a decrease in the particle size of the coacervate droplets was related to the increase in phase coacervation volume effected by the change in polyisobutylene concentration. They attributed the rise in phase coacervation volume with polyisobutylene concentration to an increase in the volume of adsorbed solvated polyisobutylene. The adsorbed layer minimizes agglomeration of the droplets rather than the mixing of their solvated polyisobutylene layers. They suggest that polyisobutylene acts as a protective colloid in the coacervation process and prevents the formation of large aggregates of ethylcellulose. A free-flowing powder was obtained on drying when polyisobutylene was employed; however, in the absence of polyisobutylene an aggregate mass was produced. Benita and Donbrow (1980) extended their research on the role of polyisobutylene and its effect on coacervation. Microanalysis indicated that polyisobutylene was not coprecipitated with the washed ethyl-cellulose coacervate droplets and thus functions as a stabilizer by adsorption. The increase in phase coacervation volume with increasing polyisobutylene concentration was explained by a decrease in sedimentation rate as a result of combined effects of the smaller size of the droplet and the higher viscosity of the medium. The final phase coacervation volume is determined less by the ethylcellulose close-packed volume than by the repulsion forces between the stabilized droplets. During the cooling process in order to solidify the coacervate drops, the adsorbed layer of polyisobutylene increases the surface viscosity and it is expected that the rate of surface nucleation of the ethylcellulose decreases, thus explaining the formation of the smooth surface, characteristic of a structure of amorphous nature. The process would be promoted by an increase in adsorption of polyisobutylene and surface viscosity as the temperature falls. The authors concluded that polyisobutylene, a linear polymer, acts by forming a high-energy barrier as a result of adsorption of anchor groups on the surface of the droplet; the rest of the polyisobutylene molecule, bound by either looped segments or segments, is directed toward the outside of the coacervate droplet. This arrangement provides steric stabilization as a result of repulsion of solvated polymer chains.

The mechanism of aggregation prevention by polyisobutylene was studied using Eudragit RS or RL as the wall polymer, tetrahydrofuran as the solvent and cyclohexane as the non-solvent (Donbrow etal., 1990). Phase diagrams, phase volume ratios of the system and photomicrographs of various stages of microencapsulation were presented. The presence of polyisobutylene permits the formation of two liquid phases, which is an unstable emulsion and a significant volume fraction is occupied by the wall polymer phase. During addition of the non-solvent cyclohexane, the solvent tetrahydrofuran is removed from the wall polymer phase and its volume fraction decreases. The dispersed concentrated wall polymer phase remains fluid during deposition onto the core surface and then gelling occurs and the system is then composed of two liquids and one gel. If polyisobutylene is not present a viscous coacervate rapidly separates, which is adhesive, during the slower desolvation stage as the composition of the solvent changes, and in this case liquid and gel are formed. The authors suggest that secondary dispersion phase phenomena are more readily controlled in the two liquid dispersion compared with liquid gel dispersion and the polyisobutylene permits steric stability. The effect of polyisobutylene molecular weight was also investigated. Polyisobutylene (mol. wt 50000) did not permit the formation of microcapsules, but yielded matricized core particles. This was attributed to the formation of a low volume wall-polymer phase of high viscosity, even prior to non-solvent addition, which was not able to provide an appropriate coating similar to the condition of the gel formed in the absence of polyisobutylene. Polyisobutylene of higher molecular weight gave two incompatible fluid phases which progress through the appropriate changes on addition of cyclohexane to give microcapsules. Polyisobutylene solutions at a constant viscosity but using higher molecular weight polymers, that is, using smaller concentrations of polyisobutylene, gave smaller phase volumes at initial condition and smaller droplet diameters.

The influence of coacervation inducing agents, namely butyl rubber, polyethylene and polyisobutylene, was studied by Samejima etal. (1982). The core was ascorbic acid and the coat of ethylcellulose was deposited on the core drug as a result of temperature reduction from a solution of the polymer in cyclohexane. Scanning electron microscopy of the resulting microcapsule showed that the surface of ascorbic acid was poorly covered when butyl rubber was used as the coacervation inducing agent, a smooth coat with small holes was produced when polyethylene was used and a smooth surface with few holes was obtained with polyisobutylene. The wall thickness was in the range from 0.75 to 3.72 and was in the order butyl rubber < none < polyethylene < polyisobutylene. The authors suggested that the factors of smoothness and wall thickness influenced the dissolution rate in the order, butyl rubber > none > polyethylene > polyisobutylene. In conclusion, polyisobutylene is adsorbed on the coacervate wall which is on the surface of the crystal and functions as a stabilizer, preventing the agglomeration of single microcapsules into aggregates.

Viscosity and surface tension effects on microsphere size

The influence of viscosity and surface tension on the particle size of microspheres prepared by emulsification was investigated by Sanghvi and Nairn (1992). The microspheres were prepared by dissolving different amounts of polymer, cellulose acetate trimellitate, in solutions of acetone and ethanol and adding this solution to the external phase composed of mixtures of light and heavy mineral oil. The viscosities of the two phases were determined both before and after mixing and the interfacial tension between the two phases was also determined. The interfacial tension ranged up to 7 dynes cm-1 but did not affect the particle size appreciably. It was found that as the viscosity ratio of the internal phase to the external phase both before and after mixing increased, the particle size of the microspheres slowly increased from about 100 /¿m to about 200 /¿m until a minimum viscosity ratio of approximately 10 before mixing and approximately 1000

after mixing was achieved; subsequently there was a very rapid increase in the size to about 700 fim. The data were related to the theory of drop deformation as described by Becher (1965) and the ease with which particles coalesce (Gopal, 1968).

In a subsequent paper Sanghvi and Nairn (1993) were able to control the particle size of the cellulose acetate trimellitate microspheres by adjusting the ratio of the polymer to solvent concentration and by adjusting the internal phase volume fraction. The amount of polymer has a direct influence on the viscosity of the internal phase, and hence the viscosity ratio of the internal to external phase as described above. The phase volume ratio affects the particle size of the microsphere as it changes the probability of two droplets colliding and forming a larger droplet.

Was this article helpful?

0 0

Post a comment